Newnes Power Engineering Series: Industrial Power Engineering ...

Inaustrial PowerEngineering andApplicationsHandbookIK.C. Agra WalQiBNewnes

دانای آشکار و نهان استببه نام او که صفحه آشنايي باچشم داشتي ساعتها زحمت ككشيده شده است،‏ لطفا با حفظاين فايل و قراردادن آن در ت سايت بدون هيچ با سلام،‏ براي تهيهسايت،‏ ما را د در اين مسير ياري نماييد.‏طاق دانشتشكر صميمانه از شما،‏ مديريت آموزشگاه () بادر عرصه آموزشآموزشگاه ممجازي طاق دانش انقلابيكسب رتبهي برتر در كنكور هستيد؟به دنبالصرف زمان كمتر هستيد؟ آياآيا به دنبال يادگيري بهتر و يافتنيادگيري با مشكل مواجه هستيد و به دنبالآيا درنياز به يادگيري زبانهاي خارجي در فرصت كوتاه داريد؟ آيا آيا نياز به راهنمايي در انجام پايان نامه داريد؟با سابقه هستيد؟يك استاد خصوصي خوب وهستيد؟مشكل مواجه آيا سابقه آموزشي استاد خصوصي براي شما ممهم است؟ آيا در انجام پروژهي خود با هستيد و فرصت حضور در كلاس هاي درس را نداريد؟آيا شاغلمشكل مواجه شدهايد؟ آيا در شبيهسازي مقالات با نياز به يادگيري نرمافزار خاصي داريد؟آيا نياز به مشاوره تحصيلي توسط نخبگان دانشگاهي داريد؟ آيايابيدطاق دانشرا در آموزشگاه مجازي پاسخ تمام نيازهاي خود آموزش نرمافزار ميباشد.‏ گروه آموزشي برقسافت ومجازي طاق دانش نتيجه سالها تجربه در زمينه تدريس خصوصي وآموزشگاهطرحي نودانشجوي قديم با پنج سال تجربه ر در زمينههاي مختلف آموزشي از جمله تدريس خصوصي و آموزش نرمافزار،‏ درپارساز يك سال تلاش مداوم درپس1391/ 10/30آموزشگاه تحت وب با امكانات حرفهاي براي مدرسان،‏موفق به تاسيس يكپژوهشگران،‏ دانشجويان و دانشآموزان شد.‏ در طراحي اين آموزشگاه سعي شده تا تمامي نيازهاي آموزشي گروههاي مختلفبرآورده شود.‏ عضويت در اين آموزشگاه ي براي همه ‏(مدرسان،‏ دانشجويان و ...) رايگان مميباشد.‏شبيهسازي مقالات ودانشجويان كه در زمينه تدريس،‏ آموزش نرمافزار،‏ راهنمايي در اانجام پروژه،‏ بازسازي نتايج همهصصاس -... تخصميتوانند به عنوان مدرس در سايت ثبت نام كنند.‏دارند،‏به داشتن كنترل پنل اختصاصي براي مشاهدهي درخواستهاي تدريسمزاياي اين آموزشگاه براي مدرسان ميتوانجملهازتنظيم پروفايل شخصي،‏ آگاهي از درخواستها از طريق ايميل و اآموزش نرمافزار و ... به صورت تفكيك شده؛ خصوصي،‏اس ؛ امكان ددانلود كتابهاي تخصصي و ججزوههاي آموزشي؛ و ... اشاره كرد.‏اماكنون برايميدهد.‏ همكل كشور پوششمقاطع تحصيلي و تمامي رشتههاي دانشگاهي را در آموزشگاه طاق دانش تماميثبت نام و برخورداري از امكانات فراوان اين آموزشگاه به آدرس اينتررنتي www.TagheDanesh.comمراجعه فرماييد.‏

NEWNES POWER ENGINEERING SERIESIndustrial PowerEngineering andApplications Handbook

NEWNES POWER ENGINEERING SERIESSeries editorsProfessor TJE Miller, University of Glasgow, UKAssociate Professor Duane Hanselman, University of Maine, USAProfessor Thomas M Jahns, University of Wisconsin-Madison, USAProfessor Jim McDonald, University of Strathclyde, UKNewnes Power Engineering Series is a new series of advanced reference texts covering the core areas ofmodern electrical power engineering, encompassing transmission and distribution, machines and drives, powerelectronics, and related areas of electricity generation, distribution and utilization. The series is designed for awide audience of enginccrs, academics, and postgraduate students, and its focus is international, which isreflected in the editorial team. The titles in the series offer concise but rigorous coverage of essential topicswithin power engineering, with a special focus on areas undergoing rapid development.The series complements the long-established range of Newnes titles in power engineering, which includes theElectrical Engineer’s Reference Book, first published by Newnes in 1945, and the classic J&P TransformerBook, as well as a wide selection of recent titles for professionals, students and engineers at all levels.Further information on the Newnes Power Engineering Series in available from bhmarketing@repp.co.uk andwww.newnespress.comPlease send book proposals to Matthew Deans, Newnes Publishermatthew.deans @repp.co.ukOther titles in the Newnes Power Engineering SeriesAcha, Agelidis, Anaya & Miller Electronic Control in Electrical System 0 7506 5 126 1Miller (ed.) Electronic Control of Switched Relunctance Machines 0 7506 5073 7

NEWNES POWER ENGINEERING SERIESIndustrial PowerEngineering andApplications HandbookK.C. AgrawalNewnesBOSTON OXFORD AUCKLAND JOHANNESBURG MELBOURNE NEW DELHI

I TO my parents 1Copyright 0 200 1 by Butterworth-HeinemannAll rights reserved.member of the Reed ElsevierNo part of this publication may be reproduced, stored in a retrieval system, or transmitted inany form or by any means, electronic, mechanical, photocopying, recording, or otherwise,without the prior written permission of the publisher.Recognizing the importance of preserving what has been written, Butterworth-Heinemann printsits books on acid-free paper whenever possible.A.l.ll...,.,.l.GL88N, Butterworth-Heinemann supports the efforts of American Forests and the Global ReLeaf programin its campaign for the betterment of trees, forests, and our environment.Library of Congress Cataloging in Publication DataAgrawal, K.C.Industrial power engineering and applications handbooWK.C. Agrawal.p. cm.Includes bibliographical references and index.ISBN 0 7506 7351 61 Factories - Power supply - Handbooks, manuals, etc. I Title.TK 4035 F3. A37 2001658.2'6-dc21 2001016211British Library Cataloging in Publication DataAgrawal, K.C.Industrial power engineering and applications handbook1 Electrical power transmission - Handbooks, manuals, etcI Title621.3'1The publisher offers special discounts on bulk orders of this bookFor information, please contact:Manager of Special SalesButterworth-Heinemann225 Wildwood AvenueWoburn, MA 01801-2041Tel: 781-904-2500Fax: 781-904-2620For information on all Butterworth-Heinemann publications available, contact our World Wide Webhome page at: http://www.bh.com10987654321Typeset in Replika Press Pvt Ltd., Delhi 110 040, IndiaPrinted and bound in Great BritainPOR EVERY TITLE THAT WE PUBLISH, BUWERWORTH~HEINEMANNWILL PAY FOR HTCV TO PLANT AND CARE FOR A TREE.

ContentsPrejkiceAckiiowledgernentsTechnical supportIntroductionPart I Selection, Testing, Controlsand Protection of Electric Motorsixxi ...Xlllxv1 Theory, Performance and ConstructionalFeatures of Induction Motorslntroduction . Brief theory of the operation of apolyphase motor. Motor output and torque . Motorratings and frame sizes . Preferred ratings at differentvoltages . Influence of service conditions on motorperformance . No-load performance . Effect of steel oflaminations on core losses . Circle diagram . Types ofinduction motors . Mounting of motors Enclosures *Weatherproof motors (WP) . Degree of protection .Cooling systems in large motors . Single phasemotors Theory of operation2 Motor Torque, Load Torque and Selectionof MotorsMotor speed-torque curve . NEMA rotor designs .Special designs of rotors . Effect of starting current ontorque . Load torque or opposing torque 1 Selection ofmotors . Time of start-up and its effect on motorperformance . Thermal withstand time3 Duties of Induction MotorsDuty cycles Continuous duty (CMR) (SI) Periodicduties . Factor of inertia (FI) Heating and coolingcharacteristic curves ' Drawing the thermal curves .Rating of short motors . Equivalent output of shorttime duties . Shock loading and use of a flywheel4 Starting of Squirrel Cage Induction MotorsDirect on-line starting (DOL) . Reduced voltage starting5 Starting and Control of Slip-ring InductionMotorsImportant features of a slip-ring motor . Starting of slipringmotors . Hypothetical procedure to calculate therotors resistance. Speed control of slip-ring motors 'Moving electrode electrolyte starters and controllers6 Static Controls and Braking of MotorsSpeed control in squirrel cage motors . Speed controlthrough solid-state technology . Vu control (speedcontrol at constant torque) . Phasor (vector) control .Use of phasor control for flux braking . Control andfeedback devices . Application of solid-statetechnology ' Conduction and commutation . Circuitconfigurations of semiconductor devices . Smoothingripples in the d.c. link ' Providing a constant d.c.voltage source 4 Providing a constant current source .Generation of harmonics and switching surges in astatic device switching circuit . Protection ofsemiconductor devices and motors . Energyconservation through solid-state technology .Application of static drives . Speed variation throughvariable-speed fluid couplings . Static drive versus fluidcoupling . D.C. drives Braking . Induction generators' Inching or jogging . Number of starts and stops7 Special-Purpose MotorsTextile motors . Crane motors . Determining the sizeof motor . Sugar centrifuge motors Motors for deep-well pumps . Motors for agricultural application .Surface-cooled motors 8 Torque motors or actuatormotors . Vibration and noise level . Service factors .Motors for hazardous locations . Specification ofmotors for Zone 0 locations . Specification of motorsfor Zone 1 locations . Motors for Zone 2 locations .Motors for mines, collieries and quarries . Intrinsicallysafe circuits, type Ex. 'i' . Testing and certifyingauthorities . Additional requirements for ciriticalinstallations Motors for thermal power stationauxiliaries . Selection of a special-purpose motor8 Transmission of Load and Suitability ofBearingsDirect or rigid couplings . Flexible couplings .Delayed-action couplings . Construction and principleof operation . Belt drives . Checking the suitability ofbearings . Suitability of rotors for pulley drives9 Winding Insulation and its MaintenanceInsulating materials and their properties . Ageing ofinsulation . Practices of insulation systems . Procedurefor vacuum pressure impregnation . Maintenance ofinsulation . Monitoring the quality of insulation of HTformed coils during manufacturing

VI Contents10 Installation and Maintenance of ElectricMotorsInstallation of bearings and pulleys . Important checksat the time of commissioning Maintenance of electricmotors and their checks . Maintenance of bearings .General problems in electric motors and their remedy .Winding temperature measurement at site . Analysis ofinsulation failures of an HT motor at a thermal powerstation11 Philosophy of Quality Systems and Testingof Electrical MachinesPhilosophy of quality systems . Testing of electricalmachines . Procedure for testing . Load test . No-loadtest . Tolerances in test results Certification of motorsused in hazardous locations12 Protection of Electric MotorsPurpose . Unfavourable operating conditions . Faultconditions . Protection . Single-device motorprotection relays . Summary of total motor protection .Motor protection by thermistors . Monitoring of amotor’s actual operating conditions . Switchgears formotor protection . Selection of main components .Fuse-free systemPart II Switchgear Assemblies andCaptive Power Generation13 Switchgear and Controlgear AssembliesApplication . Types of assemblies . Conventionaldesigns of switchgear assemblies (also referred to asswitchboards) . Design parameters and serviceconditions for a switchgear assembly . Deciding theratings of current-carrying equipment, devices andcomponents . Designing a bus system . Designing aswitchgear assembly . HT switchgear assemblies .General guidelines during installation and maintenanceof a switchgear or a controlgear assembly . Powercircuits and control scheme diagrams . Painting .Procedure of switchgear and control gear assembliesand treatment of effluent14 Testing of a Metal-enclosed SwitchgearAssembliesPhilosophy of quality systems . Recommended tests .Procedure for type tests . Procedure for routinetests . Procedure for field tests An introduction toearthquake engineering and testing methods15 Instrument and Control Transformers:Application and SelectionIntroduction . Types of Transformers . Commonfeatures of a voltage and a current transformer .General specifications and design considerations forvoltage transformers Precautions to be observedwhile installing a voltage transformer . Currenttransformers . Short-time rating and effect ofmomentary peak or dynamic currents . Summary ofSpecifications of a CT Precautions to be observedwhen connecting a CT . Test Requirements16 Captive (Emergency) Power GenerationIntroduction . DG set . Operating parameters . Theoryof operation . Guidelines on the selection of a DG set. Types of loads . Starting of a DG set . Protection ofa DG set . Parallel operation . Procedure of paralleloperation . Recommended protection for asynchronizing scheme . Load sharing by two or moregenerators . Total automation through PLCsPart 111 Voltage Surges,Overvoltages and GroundingPractices17 Voltage Surges - Causes, Effects andRemediesIntroduction . Temporary overvoltages . Voltage surgeor a transient . Transient stability of overhead lines .Causes of voltage surges . Definitions . Causes ofsteep rising surges . Effect of steep-fronted TRVS onthe terminal equipment (motor as the basis). Determining the severity of a transient . Protection ofrotating machines from switching surges . Theory ofsurge protection (insulation coordination)18 Surge Arresters: Application andSelectionSurge arresters . Electrical characteristics of a ZnOsurge arrester . Basic insulation level (BIL) .Protective margins . Protective level of a surge arrester. Selection of gapless surge arrester . Classification ofarresters . Surge protection of motors . Pressure relieffacility . Assessing the condition of an arrester19 Circuit InterruptersCircuit interrupters . Theory of circuit interruptionwith different switching mediums (theory ofdeionization) . Theory of arc plasma . Circuit breakingunder unfavourable operating conditions . Circuitinterruption in different mediums . Current chopping .Virtual current chopping . Containing the severity ofswitching surges . Comparison of interrupting devices20 Temporary Overvoltages and SystemGroundingTheory of overvoltages . Analysis of ungrounded andgrounded systems . The necessity of grounding anelectrical system . Analysis of a grounded system .Arc suppression coil or ground fault neutralizer .Ground fault factor (GFF) . Magnitude of temporaryovervoltages . Insulation coordination . Application ofdifferent types of grounding methods (for HT, HV andEHV systems) . Important parameters for selecting aground fault protection scheme21 Grounding Theory and Ground FaultProtection SchemesProtection of a domestic or an industrial single phasesystem . Ground fault on an LT system . Ground faultprotection in hazardous areas. . Ground leakage in anHT system . Core-balanced current transformers(CBCTs). Ground fault (GF) protection schemes

Contents vii22 Grounding PracticesGrounding electrodes Resistivity of soil (p) .Measuring the ground resistance Metal for thegrounding conductor Jointing of groundingconductors . Maintenance of grounding stationsGrounding practices in a power generating stationTolerahle potential difference at a location . Voltagegradients . Determining the leakage current through abody . Measuring the average resistivity of soil .Improving the performance of soil . Determining theground fault current . Designing a grounding gridPart IV Power Capacitors23 Power Capacitors: Behaviour, SwitchingPhenomena and Improvement of PowerFactorIntroduction . Application of power capacitors Effectof low PF . Other benefits of an improved powerfactor . Behaviour of a power capacitor in operation .Generation of triple harmonics in an inductive circuit.Generation of harmonics by a power electronic circuit.Resonance . Effective magnitude of harmonic voltages;~nd currents When harmonics will appear in a\ystem Filter circuits : suppressing harmonics in apower network . Excessive charging currents(switching inrush or making currents) Limiting theinrii\h currents . Capacitor panel design parameters .Capacitor rating for an induction motor. Location ofcapacitors Automatic PF correction of a system .Switching sequences . PF correction relays24 System Voltage RegulationCapacitors for improvement of system regulationSeries capacitors Rating of series capacitorsAdvantages of series compensation . Analysis of asystem for series compensation . Reactive powermanagement Influence of line length (ferranti effect)To optimize the power transfer through reactivecontrol . Transient stability level Switching of largereactive banks25 Making Capacitor Units and Ratings ofSwitching DevicesMaking a capacitor element . A critical review ofinternally protected capacitor units . Self-healingcapacitors . Making a capacitor unit from elements .Making capacitor banks from capacitor units Ratingand selection of components for capacitor duty Fastdischarge devices26 Protection, Maintenance and Testing ofCapacitor UnitsProtection and safety requirements . Installation andmaintenance of capacitor units Test requirements27 Power ReactorsIntroduction Selection of power reactors . Magneticcharecteristics . Design criterion and I . q5characteristics of different types of reactor .ApplicationPart V Bus Systems28 Carrying Power through Metal-enclosedBus SystemsIntroduction . Types of metal-enclosed bus systems .Design parameters and service conditions for a metalenclosed bus system Short-circuit effects . Serviceconditions Other design considerations ‘Skin effect .Proximity effect . Sample calculation for designing a2500 A non-isolated phase aluminium busbar system29 Recommended Practices for MountingBuses and Making Bus JointsPrecautions in mounting insulators and conductors .Making a joint Bending of busbars30 Properties and Ratings of Current-CarryingConductorsProperties and current ratings for aluminium andcopper conductors . Current-carrying capacity of’copper and aluminium conductors31 An Isolated Phase Bus SystemAn isolated phase bus (IPB) system Constructionalfeatures . Special features of an IPB system .Enclosure heating . Natural cooling of enclosuresContinuous rating . Forced cooling . Influence of aspace field on the metallic structures . Fault level .Voltage drop . Forming of sections for IPB systemsDetermining the section and size of conductor andenclosure Sample calculations32 Testing a Metal-enclosed Bus SystemPhilosophy of quality systems . Recommended testsProcedure for type tests Routine tests . Field testsIiidex 959

PrefaceThe author has had a long association with the machinesdescribed in this book. The book is the result of thisexperience and the overwhelming help and supportextended to him by his colleagues, friends and businessassociates over the last twelve years.The purpose of this book is to share the experience ofthe author with those in the field. It is an attempt to makethese subjects simple and interesting. The book shouldprovide an easy approach to answer the problems anengineer or engineering student may face when handlingthese machines.The author is sure that the readers will find ampleopportunity to learn from his experience and apply thisinformation to their field of activities. The book aims toprovide a bridge between the concept and the application.With this book by his or her side, an engineer should beable to apply better, design better and select betterequipment for system needs and ambient conditions. Itshould prove to be a handy reference to all those in thefield of design and application, protection and testing,production, project engineering, project implementationor maintenance, in addition to the sales and purchase ofthese products.Engineers have done an incredible job by inventingnew technologies and bringing them, over the years totheir present level. Research and development work by adedicated few scientists and engineers has been an untiringprocess which has provided us with yet more advancedforms of technology. The credit for this book goes tothese engineers and scientists throughout the world.The author is not an inventor, nor has he done anythingnew in these fields. He has only attempted to bring togethersuch advances in a particular field in one book for theirbetter application. The author’s contribution can beregarded as an appropriate selection and application ofthe available technology and products for their optimumutilization.All relevant aspects of a machine, including its design,have been discussed but greater emphasis is laid onselection and application. Since this is a reference bookthe basic theory is assumed to be known to a student ora practising engineer handling such machines and/ortechnologies.In the academic world the derivation of a formula fromfundamentals is regarded as most important. In practice,this formula matters more rather than its origin. But forthose who wish to know more of the reasoning and thebackground, care is taken that such subjects are alsocovered. The author hopes that readers will be satisfiedto have most of their queries answered.The book has been written so that it should refreshand awaken the engineer within a reader. The author iscertain that this is what readers will feel as they progressthrough this book. A cursory reading will bring themabreast of the subject and enable them to tackle problemswith ease and simplicity. The author’s efforts will bedefeated if this book falls short of this aim.The endeavour has been to provide as much informationas possible on the application of available technologyand products. It should help application engineers toselect and design a more suitable machine or power systemfor their needs. As mentioned above, this text will notcover the full engineering derivations, yet all fundamentalshave been provided that are considered relevant to engineerany machine or system covered in this book. To augmentthe information, ‘further reading’ has also been providedto support the text and to answer queries that may ariseon a particular subject. For detailed engineering, themanufacturers are still the best guide. Detailing andengineering must be left to them. In this book, the authorhas tried to make the subject comprehensive yet conciseand easy to understand so that, one can easily refer to itat any time. The references drawn are brief, but pertinent,and adequate to satisfy a query.This book may prove to be a boon to young engineersentering the field. With it they can compare the theory oftheir studies with application in the field.Whereas all aspects that were thought necessary havebeen considered, it is possible that some have been omitted.The author would be grateful to receive suggestions fromreaders for any additions, deletions or omissions to makethis book even more useful and up to date.K.C. Agrawal

AcknowledgementsThe author acknowledges with gratitude the help extendedto hiin by his colleagues and friends and for the cataloguesand leaflets from various manufacturers. Also helpfulhave been a score of references and textbooks on thesub,jrct as well as product standards such as IEC, IEEE,ANSI and NEMA. ISO, BS and IS and other nationaland international publications.Thc author's most sincere thanks are due to the following,who have given their whole-hearted support and help incompiling this book and making it up to date. They haveprovided details of the latest practices followed by leadingconsultants. engineers and end users, as well as the nationaland international practices of the various manufacturers.Special thanks for formatting the book go to:Anjuna AgrawalThe author's daughter forher unflinching supportthroughoutMaiiish ChandComputer support andassistanceAsad MirzaEditing and proof rendingA.K. ShringhiI 11 us trati o n s and drawingservicesBesides the above, the following have bccn a source ofencouragement and continued support:Bhanu BhushanPower Grid CorporationHarbhajan Singh Crompton ~reaGesDr Abraham Verghese English Electric (now Alstom) Indu Shekhar Jha Power Grid CorporationAni1 KamraCrompton GreavesN.K. BhatiaEnglish Electric (nowAshwani Agrawal ABBAlstom)Dr Ashok Kurnar University of Roorkee Dr P.C. Tripathi BHELB. RamanBHELY. Pal Ex EILD.K. DuwalNTPCM.R. At&ECSas well as a number of friends and well-wishers in addi-P.S. GokhlcSyntron Controlstion to those mentioned in the Technical Support sectionS.P. Shnrma L&T below.Dr Subir SenPower Grid CorporationI would like to apologize to my family and friends forV.P. SharmaEILneglecting them during the time it has taken me to writethis book; in particular. my wife Madhu and my daughterThc International Electrotcchnical Commission (IEC)" Anjuna.Thz author thanks thc IEC lor permission to use the following material. All extracts are copy!-lght 0 IhC Geneva. S\vit/erland. All right\Iesened. All IEC Publication\ are available from www.icc.ch. IEC takes no rmponsihility for and will not assume linhility fol- darnazesresulting from the readers misinterpretation of the referenced inaterial due to ils placement and context in thi\ puhlication. The material ISireproduced or written with their permis\ion.

Technical SupportThe following companies and individuals have providedinvaluable technical support.Chapter 1R.K. Gupta, NGEF Ltd, New Delhi; J.R. Mahajan, VoltasLtd, Mumbai; Ani1 Kamra, Crompton Greaves Ltd, NcwDelhi; B. Raman, BHEL, Bhopal; S.K. Sharma, GEMotors India Ltd, Faridabad; Anirudh Singh, BHEL, NewDelhiChapter 3R.K. Gupta, Bhartiya Cutler Hammer Ltd, FaridabadChapter 5Y.H. Kajiji, AOYP Engineering Company, Mumbai; RamChandran. Bhartiya Cutler Hammer Ltd, FaridabadChapter 6Sanjay Sinha, Hemant Agrawal, Vijay Arora, MunishSharma and Ajay Kumar, Allen Bradley India Ltd,Sahibabad; Shasidhar Dhareshwar, ABB, New Delhi;Ashwani Kaul, Kirloskar Electric Co. Ltd, New Delhi;Siddhartha Ghosh, Siemens Ltd, New Delhi; S.D. NandaKishore, Cegelec (India), Noida; Arun Gupta, EISA Lifts,New Delhi; Rajiv Manchanda, Usha (India) Ltd,Faridabad; Atul Grover, Allen Bradley Ltd, New Delhi;Ashok Jain, Fluidomat Ltd, Dewas (MP); AK Advani,Greaves Ltd, New Delhi; Nitin Sen Jain, Kirloskar ElectricCo. Ltd. New Delhi; M/s Subhash Projects India Ltd;Sarvesh Kumar, Vestas RRB India Ltd, New Delhi;Bal Gopal, Dynaspede Integrated Systems (P) Ltd,Tamil NaduChapter 7Y.I.P. Sehgal, International Pumps and Projects, Noida(UP); S.U. Submersible Pumps, Noida (UP); V.P. Sharma,EIL, New DelhiChapter 8Pembril Fluidrives Ltd, Aurangabad:Ashok Jain, Fluidomat Ltd, Dewas;Eddy Current Controls (India) Ltd, Kerala;A.K. Advani, Greaves Ltd, New Delhi;Sunil K. Kaul and K. Sukhija. Fenner (India) Ltd.Hyderabamew Delhi;S. Kumar and P.K. Das, Dunlop (India) Ltd. Kolkata/New DelhiChapter 9B. Raman, BHEL, Bhopal, IndiaChapter 10Soman Dhar, Castrol India Ltd, New Delhi; ChandraKumar, National Engineering Company, Jaipur; B.B. Rao,SKF Bearings India Ltd, Mumbai; B. Raman, BHEL,Bhopal; Ani1 Kamra, Crompton Greaves Ltd, New DelhiChapter 11Y.D. Dosaj, Crompton Greaves Ltd, Ahmed Nagar;Y. Pal, Formerly with EIL, (GG-1/13-A, Vikaspuri. NewDelhi)Chapter 12P.B. Dabholkar and lndrajit Jadhav, ABB, Vadodara; SanjivBahl, ABB, Delhi; Pankaj Sachdeva and Balamourougan,Alstom (India), Chennai; V. Mohan, Rajiv Tandon, PremKumar and Chand Chadha, Larsen and Toubro, NewDelhi; Kapil Grover, Allen Bradley. Sahibabad; S.P.Sharma, L & T, Mumbai; B. Raman, BHEL. BhopalChapter 13M.R. Atter, ECS, Noida; S.P. Sharma, L & T, Mumbai;D.K. Duggal, NTPC, Noida; N.K. Agrawal and P.K. Garg,BHEL, Haridwar; Abid Hussain, Consultant, New Delhi,Dr V.K. Anand, Anand CIS India, New Delhi; Stat-fieldsystems (Coating) Pvt. Ltd: Mitsuba Electricals (P) Ltd,

Technical support xivIndia; Sajal Srivastava, Berger Paints Ltd, New Delhi;Sajjathe Sulthan, SBA Enviro Systems, New Delhi; R.R.Bagri, Clear Water Ltd, New DelhiChapter 14Dr. Ashok Kumar Mathur, Department of EarthquakeEngineering, University of Roorkee; Dr H.K. Chand,formerly Professor of Geography, Bhagalpur UniversityChapter 15S. Raghavan, Indcoil Transformers (P) Ltd, Mumbai;P.U. Patwardhan, Prayog Electricals (P) Ltd, Mumbai;Dr Abraham Verghese, English Electric Co. now Alston(India), Sharjah UAEChapter 16Kamal Singhania, Crompton Greaves Ltd, New Delhi;Ashawni Kaul, Kirloskar Electric, New Delhi; AtulKhanna, AVK-SEGC & Controls (India) Ltd, New Delhi;V.K. Chaudhary, Cable Corporation of India Ltd, NewDelhi; 0 P Kwatra, Havell’s India Ltd, New DelhiChapter 17Ashwani Agrawal, ARB, New Delhi; T.K. Modak, JyotiLtd. New DelhiChapter 18Elpro International Ltd, Chinchwad Gaon, Pune; StationSurge Arresters and Thyrite Magne - Valve StationLightning Arresters, M/s IEG (India) Ltd; TransiNor As,Norway, through Vijay Khanna, New Delhi; N. V. AnanthaKrishnan, W.S. Industries (India) Ltd, New Delhi; A. S.Kushwaha and Arvind M. Khurana, Power GridCorporation of India Ltd, New Delhi; V.P. Singh, ElproInternational Ltd, New Delhi; Ashwani Agrawal, ABBIndia Ltd. New DelhiChapter 19V. Raghavan, BHEL, Bhopal; N. Biswas, Alstom, NewDelhi; P.P. Sreekanth, Alstom, New Delhi; Manjeet Singhand Man 0. Misra, ABB, New Delhi; Sandeep MathewSiemens India, New DelhiChapter 20Y.K. Sehgal, Power Grid Corporation of India, New DelhiChapter 21P. U. Patwardhan, Prayog Electricals (P) Ltd, Mumbai;Dr Abraham Verghese, Alstom (India), Sharjah, UAE,Dr Girish Chandra, KGMC, LucknowChapter 22V. P. Sharma, EIL, New DelhiChapter 23Neutronics Manufacturing Co.; Syntron Controls; M. R.Attar, ECS; P. S. Gokhle, Syntron Controls, Mumbai;N. P. Singh, Sigma Control, New Delhi; Vijay Jain,Neutronics Manufacturing Co. Ltd, Mumbai; S.P. Sharma,L & T, Mumbai; Y. P. Likhyani, Crompton Greaves Ltd,New Delhi; M. K. Pandey, Kapsales Electricals (KhatauJunker), New Delhi; A.S. Kushwaha, Power GridCorporation of India, New Delhi; P. K. Mittal, IndianRailways Telecommunication, New DelhiChapter 24S. P. Singh, CEA, New Delhi; Dr Subir Sen and InduShekhar Jha, Power Grid Corporation of India Ltd, NewDelhi; Brajesh Malviya, ABB India Ltd, New Delhi;M. Hanif, BHEL, New DelhiChapter 25M. K. Pandey, Kapsales Electricals (Khatau Junker), NewDelhiChapter 26M. K. Pandey, Kapsales Electricals (Khatau Junker), NewDelhiChapter 27P. T. Pandyan, Western High Voltage Equipment,Jaisinghpur, MaharashtraChapter 29Russy N. Master. L & T, Mumbai; J. K. Juneja,M/s J. K. Plastics, Noida; P.H. Kakade, M/s VinayakCorporation, MumbaiChapter 30Regie Paul, Indian Aluminium Co. Ltd (Indalco), Cochin,KerdlaChapter 31B. S. R. Patnaik, Best & Crompton Engineering Ltd,Chennai; D. K. Chaturvedi, NTPC Ltd, Noida; V. K.Mehta, BHEL, Jhansi; R. N. Khanna, Controls andSwitchgears Ltd, New Delhi

IntroductionThis book is split into five parts. A summary of each partfollows.Part I Selection, testing, controlsand protection of electric motorsThis part deals with three- and single-phase ax. machines,and their protective switchgears. However, reference ismade to and comparisons drawn of a d.c. motor with anax. motor, to assist a user to make a proper choice ofmachine.However simple a motor is, it requires careful handlingto ensure optimum performance and long years of troublefreeoperation. A small drive, failing while in operation,may bring the entire process to a halt. One can visualizethe loss of production that can result. Power plants,chemicals, fertilizers, petrochemicals, paper and cementmills all require careful selection of equipment to avoidbreakdown or malfunctioning during operation. Motorsand their controlgears are core components that requirespecial attention:This part deals with the specifications, performance,characteristics and behaviour of motors under differentoperating conditions, their application and selection.It also covers aspects such as shock loading, motorsfor hazardous locations and open transient conditionsin HT motors during a switching sequence.In Chapter 12 a detailed analysis is made of allunfavourable operating conditions and their effect onthe performance of the motor and its protection foroptimum utilization. The precautions also cover surgeprotection for HT motors. The details provided coverthe smallest-influence that a particular parameter canhave on a machine.This part-also deals with static controls and drives,soft starting and process control through solid-statetechnology (phasor and field-oriented controls) usingIGBTs as well as energy conservation.There is special coverage of fluid couplings for softstarting and speed control. A comparison between staticdrives and variable-speed fluid couplings is made.Windmills (induction generators) as an unconventionalenergy source, vertical hollow shaft motors and submersiblepump sets, selection of belts for transmissionof load. the phenomenon of shaft currents are discussed.The text especially covers testing requirements and anintroduction to quality assurance systems and applicationof IS0 9001.Special coverage of impulse testing of resin-rich formedcoils and their in-house testing requirements is given.Part II Switchgear assemblies andcaptive (emergency) power generation(Including instrument transformers and cableselection)The subjects covered aim at providing methods to formspecifications and then design a switchgear assembly forall power distribution needs. lt also provides coverage ofdraw-out assemblies. Establishing the fault level of asystem is described including the electrodynamic andelectromagnetic forces that arise.Testing procedures are informative and elaborate.Seismic effects and earthquake engineering is coveredin this part to study the behaviour of an object underseismic conditions and its suitability for criticalinstallations. The formation of the earth and movementsof tectonic plates that cause earthquakes and volcaniceruptions are described.Instrument transformers (CTs, class PS, CT,, VTs andCVTs etc.) form important components of a switchgearassembly for measurement and protection. They arecovered for their specifications, selection andapplication.Design of class PS CTs (special coverage) is provided.Captive (emergency) power generation covers theapplication of a diesel generating set, its starting,protection, synchronizing and load sharing. This formsan important part of power distribution at anyinstallation to provide a standby source of supply.The entire painting rocedure and effluent treatment iscovered for those in the field of manufacturing suchassemblies.In an attempt to provide as much information on therelated subjects as possible and to make the bookmore complete for a project or a design engineer wehave provided data and tables on cables and describedin detail the procedure for the selection of the typeand size of control and LT and HT power cables.

www.taghedanesh.irxvi IntroductionPart 111 Voltage surges, overvoltagesand grounding practices(including causes, effects and remedies andtheory of overvoltages, ground fault protectionschemes and grounding practices)This part is complementary to Part I1 and providestechnical support to switchgear assemblies and machinesfed by them for surge and overvoltage protection. It is avery useful part for all those handling HV and EHVpower systems and their surge and overvoltage protection.The part deals with the BIL of a system, protectivemargins and insulation coordination.It also deals with electric motors as they are typicalfor their surge behaviour and protection.It also covers the steepness of TRVs, their significanceand methods of taming them. Reflections of travellingwaves and surge transferences are also described.This part specifically considers the application andselection of surge capacitors and surge arresters. Sincethe internal causes of surge generation are a consequenceof switching and type of interrupter, the part providesdetails of the various types of interrupters in use, theirswitching behaviour, current chopping and quenchingof arc plasma. It also makes a detailed comparison ofthe various types of interrupters available to facilitatetheir selection and adaptation to a more appropriatesurge protection scheme.Temporary overvoltages are different from surges asare their causes. Therefore temporary overvoltages alsoform an important parameter in a system design andits grounding method. This topic is therefore complementaryto surge protection and has been dealt in detailto make a practising engineer or engineering studentmore aware of the behaviour of an HT system,particularly on a ground fault.Exposure of a human body to touch and step voltagesand methods to deal with these are also covered.Grounding and ground fault protection schemes aredescribed in detail with illustrations to help an engineerto select the most appropriate grounding method andground fault protection scheme for a machine or system.The use of CBCTs is covered.Part IV Power capacitors: powerfactor improvement and systemvoltage regulation: application ofshunt and series capacitorsReactive control is an important tool for voltage regulationand for optimizing available power utilization. It canalso be used for attaining better stability of the system.It has therefore become a very important technique toimprove an old distribution network that is beingoverutilized and is ailing with recurring problems suchas flickering of voltage, frequent system outages and anormally low voltage at the consumer end. The authorhas attempted to apply reactive control to improve powerdistribution networks which are overloaded and whichpresent these problems.In this part the author provides all relevant aspects ofa reactive control and carries out an exhaustive analysisof a system for the most appropriate control. Harmoniceffects and inductive interferences as well as use of filterand blocking circuits are covered. Capacitor switchingcurrents and surges and methods of dealing with theseare also described.This part considers reactive power control with theuse of shunt and series capacitors. The controls may bemanual or automatic through electromagnetic or staticdevices. Protection of capacitors and capacitor banks aswell as design, manufacturing and test requirements,installation and maintenance are also covered, the mainthrust being on the application of power capacitors.Application of series capacitors and analysis of anuncompensated transmission line and the capabilityof power transfer and system regulation with and withoutseries compensation are also presented.To clarify the subject the basics and the behaviour ofpower capacitors in operation are also discussed.This part also briefly describes different types of powerreactors required to control inrush currents or suppressthe system’s harmonic disorders, besides absorbingthe excessive charging currents on an EHV system.Part V Bus systems in includingmetal-enclosed non-isolated andisolated phase bus systemsPower transfer is a very important area of a power system.In this part it is dealt with in detail for both LT and HTsystems and for all current ratings. For large to verylarge ratings, skin and proximity effects are also discussedto arrive at a design to transfer large amounts of power,without great loss, voltage drop or voltage unbalance.Technical data and current ratings for various sizes andsections of aluminium are provided with more emphasison aluminium as it is most commonly used. The textprovides material to design, engineer, manufacture andtest a bus system of any current and voltage rating.This part specifically deals withDesign parametersShort-circuit effectsElectrodynamic and electromagnetic forcesEffects of proximity and reducing this by phaseinterleaving or phase transpositionDesigning a reactor for the middle phase to balance alarge unbalanced current-carrying systemRecommended practices to mount buses and make busconnectionsA detailed discussion of the isolated phase bus systemconcentrating on types of isolated enclosures and theirdesign aspectsSample calculation to design an IPBTesting of bus systems.

www.taghedanesh.irPART ISelection, Testing,Controls andProtection ofElectric Motors

www.taghedanesh.ir

I,"www.taghedanesh.irTheory,Performance andConstructionalFeatures ofInduction Motors1.4 Motor ratings and frame sizes 1/8I 5 Preferred ratings at different voltages 1/91.7 No-load performance 1/171.8 Effect of loading on motor performance 1/17induction motors 1/20hoice of voltage 1/20otors 1/241. I8 Theory of operation 1/27List of formulae uwd 1/34

www.taghedanesh.ir

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/51.1 IntroductionThe age of electricity began with the work of HansChristianOersted (1777-1851), whodemonstratedin 1819that a current-caving conductor could produce a magneticfield. This was the first time that a relationship betweenelectricity and magnetism had been established. Oersted’swork started a chain of experiments across Europe thatculminated in the discovery of electromagnetic inductionby Michael Faraday (1791-1867) in 1831. Faradaydenionstrakd that it was possible to produce an electriccurrent by means of a magnetic field and this subsequentlyled to the development of electric motors, generatorsand transformers.In 1888 Nikola Tesla (1 856-1943) at Columbus, Ohio,USA, invented the first induction motor which has becomethe basic prime mover to run the wheels of industrytoday. Below, for simplicity, we first discuss a polyphaseand then a single-phase motor.Figure 1.2disposition‘\, R4kPhasor representation of current and flux phase1.2 Brief theory of the operation ofa polyphase motorAs noted above, electromagnetic induction takes placewhen a sinusoidal voltage is applied to one of two windingsplaced so that the flux produced by one can link theother. A polyphase winding when arranged in a circularform produces a rotating field. This is the basic principleof an electric motor, appropriately termed an inductionmotor. Here applies the theory of the ‘left-hand rule’ todefine the relative positions of the current, field and force.The rule states that when the thumb, the forefinger andthe middle finger of the left hand are arranged so thatthey all fall at right angles to each other then the forefingerrepresents the flux 4 or the magnetic intensity H, themiddle finger the current and the thumb the force or themotion (Figure 1.1). The field thus induced would rotateat a synchronous speed and the magnitude of flux builtup by the stator current would be equal to 4m in 2-4windings and 3/2$m in 3-4 windings. For brevity, we arenot discussing the basics here. Figures 1.2-1.4 illustratea current-flux phasor representation, the flux waveformand the magnetic field, respectively, in a 3-4 winding.The winding that is static is termed a stator and thatwhich rotates is a rotor. If lrr is the rotor current and $ thcinstantaneous flux, then the force in terms of torque, T,produced by these parameters can be expressed bytB$1 = & sln of@ = dm sin (ut - 120)@3 = @ , sin (ut - 240)- utFigure 1.3 Magnetic flux waveformTAt any instant3$3 + $2 + $3 = 2 $mA constant field rotating at synchronous speed NsFigure 1.1 Fleming’s left hand ruleFigure 1.4Production of magnetic field in a 3 4 winding

www.taghedanesh.ir1/6 industrial Power Engineering and Applications Handbookwhere4 = &, sin utand Q,,, = maximum field strengthIn a 3-4 winding, therefore, for the same amount of current,the torque developed is 50% more than in a 2-4 winding.The rotor power P developed by torque Tat a speed Ncan be expressed byp=- T.N974whereP = rotor power in kW7' = torque in mkgN = speed in r.p.m.Since the kW developed by a 3-4 winding is 50% morethan by a 2-4 winding for the same value of stator currentI,, the economics of this principle is employed in aninduction motor for general and industrial use. As standardpractice, therefore, in a multi-phase system, only 3-4induction motors are manufactured and employed, exceptfor household appliances and applications, where mostlysingle-phase motors are used.The magnetic field rotates at a synchronous speed, soit should also rotate the rotor. But this is not so in aninduction motor. During start-up, the rate of cutting offlux is the maximum and so is the induced e.m.f. in therotor circuit. It diminishes with motor speed due to thereduced relative speed between the rotor and the statorflux. At a synchronous speed, there is no linkage of fluxand thus no induced e.m.f. in the rotor circuit, consequentlythe torque developed is zero.Since,(1.3)s= 1-Speed +Shp(a) Typical for an LT motorthe impedance considered represents only the rotor side.For simplicity, the stator impedance has been ignored,being too small with little error.In equation ( 1.3)T = torque developedS = slipR2 = rotor resistance per phase\\X2 = standstill rotor reactance per phase, and,,e2 = standstill rotor induced e.m.f. per phaseThe last two parameters are maximum during start-up,diminish with speed and become zero at the synchronousspeed (when S = 0). Therefore T = 0 when ,\ez = 0.CorollaryThe speed-torque characteristics of a motor will largelydepend upon its rotor parameters such as R2 and \\X2.The higher the rotor resistance R2, the higher will be thetorque. From equation (1.3) we can draw a speed-torquecurve of a motor as shown in Figures 1 .5(a) and (b).During start-up or at high slips, the value of ,,X2 willbe too high compared to R2 and equation (1.3) willmodify tos- 1Speed --c+ Slip(b) Typical for an HT motor4 'N,s = 0Figure 1.5 Speed-torque and speed-current curves at the ratedstator voltageS. ,,e: . R2T\, Dc s? , ,\X;where T,, is the torque during start-up or(1.3a)and at lower slips or at near the rated speed, whenS . ,,X?

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/7or(1.3b)7', is referred to as the rated torque.R? and ,,X, are thus the vital parameters that are usedat the design stage to accomplish the desired characteristicsand performance of an induction motor. The normalstarting torque T,, for a medium-sized LT squirrel cagemotor, say up to 400 kW, can be attained up to 200-250% and even more of the full-load torque T,, and thepull-out torque Tp up to 200-350% of T, (see also Chapter2). In slip ring motors, the starting torque (TJ can bevaried up to its pull-out torque (Tpo). (See Chapter 5.)For HT motors (2.4 kV and above) these figures arequite low compared to LT motors, due to the designeconomics for such machines. One consideration is therotor slots that are normally not made with a doublecage hut with tapered or deep bars, to reduce rotor sizeand hence, the overall size of the machine and thus thecost. The T,, is now of the order of 70-130% of T, andTpo 170-250% for motors up to 3000 kW. For yet largermachines these figures may be lower, say T,, of the orderof 33-70% and Tpo of the order of 1.50-225% of T,.Large motors are normally switched at no-load throughstatic drives (Section 6.16) or hydraulic couplings, (Section8.3). A low starting torque therefore should not matter inthe majority of cases.These figures are for general reference only. For actualvalues the reader should refer to the motor manufacturer.Motors can, however, be designed to suit a particularapplication. Large LT and all HT motors are generallycustom bui It.If e? is the induced e.m.f. in the rotor circuit at anyspeed thenwhere synchronous speed120.fN, = - r.p.m Pf= frequency of the supply system in Hz andp = number of poles in the stator winding.1.2.1 Stator current(I .6a)An induction motor draws a very high current duringstart-up as a result of magnetic saturation (Section1.6.2A(iv)). The rapid voltage change from one peak toanother (2Vm) (Figure A) within one-half of a cyclesaturates excessively the iron core of the stator and therotor. The saturation induces a very low inductance, L,and hence a low switching impedance (R being lowalready). The inductance of the circuit L varies with thelevel of saturation. Since e = - L (dildt), a high e and lowL cause a very high starting current. This is seen to be ofthe order of six to seven times the rated current andexists in the system until the rotor picks up to almost itsrated speed. We will notice sub-sequently that theperformance of a motor is the reflection of its rotorcharacteristics. As the rotor picks up speed, it reducesthe secondary induced e.m.f. S . sse2, which in turn raisesthe inductance of the rotor circuit and diminishes thestart-up inrush current. The duration of start-up currentthus depends upon the time the rotor will take to pick-upspeed to almost its rated speed.The negative expression of voltage is according to Lenz'slaw. which states that 'The direction of the induced e.m.f.is such that it tends to oppose the change in the inducingflux'. In equation (1.4)Z, = number of turns in the rotor circuit per phaseandd$ldt = rate of cutting of the rotor flux.An illustration of this expression will givee2 = 4.44 Kw . $", . z, ' f, (1.5)whereKw = winding factor andJ; = rotor frequency = S.f.At synchronous speed, dqdt = 0 and therefore e2 = 0.This is why an induction motor ceases to run atsynchronous speed. The rotor, however, adjusts its speed,N,, such that the induced e.m.f. in the rotor conductors isjust enough to produce a torque that can counter-balancethe mechanical load and the rotor losses, includingfrictional losses. The difference in the two speeds is knownas slip, S, in r.p.m. and is expressed in terms of percentageof synchronous speed, i.e.Figure ACorolluly: The case ($ u transformerThe situation in the case of a transformer is somewhatdifferent. Its primary and secondary circuits form acomposite unit and behave as one winding only. Sincethere is no air gap between the primary and the secondarywindings, the combined winding impedance is less thanthat of a motor on switching (considering secondaryopen-circuited or connected on load). Consequently theswitching current is a little higher, of the order of 8-12times the rated current. If the secondary is short circuited,the short-circuit current will be much more than this, asindicated in Table 13.7. As will stabilize the voltageinitial spikes, so it will diminish the level of saturation,raising the value of L. It will provide a high dampeningeffect (RlL) initially and slowly thereafter. The currentwill also decay rapidly initially and then slowly. Thesame will be true for any inductive circuit other than amotor.

www.taghedanesh.ir1/8 Industrial Power Engineering and Applications Handbook1.2.2 Rotor currentWith the same parameters, the rotor current of a motor,Zrr, can be expressed bys ' sse2I, =JR; + S2 . ,,X;(1.7)at high slips, i.e. during the starting region when S . ,,X2>> RZ:(1.7a)This is a vital relationship, which reveals that duringstart-up and until such speed, the reactance of the motorwindings ,,X2 >> R2, the rotor current will also remainalmost the same as the starting current and will fall onlyat near the rated speed. (Refer to the current curves inFigures 1.5(a) and (b)). The initial inrush current in asquirrel cage induction motor is very high. In a slip-ringmotor, however, it can be controlled to a desired level.(Refer to Section 5.2.1.)Note For all practical purposes the stator performance data areonly a replica of rotor data for torque and current. The performanceof a motor is the performance of its rotor circuit and its design.1.3 Motor output and torqueMotor rating, i.e. power available at the motor shaft, P,,can be expressed in kW bywhereN, = rotor speed in r.p.m.The power transferred by the stator to the rotor, P,, alsoknown as air gap power at synchronous speed, N,, can beexpressed in kW by:(1.8a)whereN, = synchronous speed in r.p.m.The difference in the two is the electric power loss in therotor circuit and is known as slip loss, i.e.Slip loss = P, - P, = S . P, (1.9)and from equation (I .8)(1.10)1.4 Motor ratings and frame sizesThe standard kW ratings are internationally adopted andare based on the recommendations made by IEC 60072-1 and 2.* The ratings in kW up to 110 kW and theircorresponding h.p. are shown in Table 1.1. Preferredratings beyond 110 kW are given in Table l.l(a). Forrecommended frame sizes, adopted by national andinternational manufacturers to harmonize and ensureinterchangeability between all makes of motors, refer toIEC 60072-1 and 2. For easy reference we have providedthese frame sizes in Table 1.2. These standards suggestthe vital dimensions of a motor such as the shaft height,extension, its diameter and the mounting dimensions,etc.Table 1.1 Preferred kW ratings and their nearesthorsepowerkW HP kW HP kW HP0.060.090.120.180.250.310.550.751.11.5According to IEC 60072-10.08 2.2 3.0 310.12 3.7 5.0 450.16 5.5 7.5 550.25 7.5 10.0 750.33 9.3 12.5 900.50 11 15 1100.75 15 20 -I .o 18.5 25 -1.5 22 30 -2.0 30 40 -506015100125150"aBeyond 1104000 kW, the preferred ratings can follow the 'RenardSeries', R-20 of preferred numbers as in IS0 3. Refer to Tablel.l(a).Table l.l(a) Preferred ratings beyond 110 kW according tothe Renard Series, 'R-20' of IS0 3kW kW kW1121251401601802002242502803153554004505005606307108009001000112012501400160018002000224025002800315035504000Example 1.1For a 150 h.p., 1480 r.p.m. (N,) motor, the torqueNote Of these, ratings up to 1000 kW with some modificationshave now been standardized by IEC 60072-1. For details refer toIEC 60072- 1.T, = '10x974(15~h.p. = 1lOkW)1480= 72.5 mkg *IEC: International Electro-technical Commission, Switzerland.

www.taghedanesh.irTheory, performance and constructional features of induction motors 119Tabte 1.2Frame size5663718090s90LIOOS1 OOL112s112M112L132sI32M132L160s160M160L180s180M180L200s200M2001,2259225MRecommended frame sizesShaft height for B3 Frame size Shaft height for B3(Figure 1.17) (Figure 1.17)motorsmotors(mm)Imm)566371809090I00100112I12I12132132132I60160160I80180I80200200200225225225L250s250M250L280s280M280L315s315M315L355s355M355L400s400M400L4505 00560630I108009001000-According to IEC-60072-1 and 60072-222525025025028028028031531531535535535s4004004004505005606307108009001000-Note Designations S, M and L denote the variation of length inthe motor housing for the same shaft height.1.5 Preferred ratings at differentvoltagesThe preferred ratings at different rated voltages accordingto publication MG-I are given in Table 1.3.1.6 Influence of service conditionson motor performanceThe performance of an induction motor is influenced bythe service conditions, when these differ from the designTable 1.3 Preferred horsepower rating for different voltagesSupply system (kV)0.415*3.3G.611Preferred rating (h.p.)Up to 600200-60001000- 12 0003500-25 000*This voltage is now revised to 0.4 kV as in IEC 60038.‘MG- 1 : Reference to a publication on Motors and Generators byNEMA (National Electrical Manufacturers Association, USA) whichis adopted universally.assumptions. The parameters that may affect the performanceof a motor are1 Voltage unbalance2 System harmonics3 Voltage variation4 Frequency variation5 Ambient temperature and6 Altitude.They may influence the performance of the motor by its1 Output and2 Torque. IEC 60034-1 stipulates for a minimum inbuiltcapacity of a machine (all ratings and voltages)to sustain an excessive torque of 60% for a minimumof 15 seconds, without stalling or an abrupt change inits speed. This stipulation is to meet the need for anexcessive torque of a transitory nature due to the aboveparameters or for a momentary excessive load torqueitself during operation. This stipulation, however, wouldnot apply to motors designed and manufactured tospecific requirements.3 Efficiency4 Power factor5 Speed6 Slip and7 CurrentNote According to the stipulations of IEC 60034-1, any factornoted above which may influence the performance and cause anexcessive current than rated, or because of an excessive load itself,motors rated up to 3 15 kW and rated voltage up to 1 .O kV shouldbe capable of withstanding a current equal to 150% of the ratedcurrent for a minimum of two minutes. No such tolerance is, however,recommended for motors beyond 315 kW and all HT motors.This stipulation would not apply to motors that arc dcsigncd andmanufactured to specific requirements.The extent to which these performance data may beinfluenced, by the non-standard opcrating conditions isdiscussed briefly below.1.6.1 Voltage unbalance and system harmonicsAs standard practice, all motors are designed for a balancedand virtually sinusoidal supply system, but it may notbe feasible to obtain the designed supply conditionsin practice. Hence, a motor is designed with a certainin-built capacity to sustain small amounts of voltageunbalances and some degree of harmonic quantities, suchthat the voltage waveform may still be regarded assinusoidal.Voltage unbalanceThe performance of a motor is greatly influenced by avoltage unbalance in the supply system. It reduces itsoutput and torque and results in a higher slip and rotorloss. This subject is covered in more detail in SectionI2.2(v). For likely deratings, refer to Figure 12.1. A systemwith an unbalance of up to 1 % or so calls for no derating,whereas one having an unbalance of more than 5% is notrecommended for an industrial application, because of a

www.taghedanesh.ir1/10 Industrial Power Engineering and Applications Handbookvery high derating and the highly unstable per-formanceof the motor.The system may be regarded as balanced when thenegative sequence component does not exceed 1% of thepositive sequence component over a long period, or 1.5%for short durations of a few minutes and the zero sequencecomponent does not exceed 1% of the positive sequencecomponent. Refer to Section 12.2(v) for more details onpositive, negative and zero sequence corn-ponents.System harmonicsA supply system would normally contain certain harmonicquantities, as discussed in Section 23.5.2. The influenceof such quantities on an induction motor is also discussedin Chapter 23. To maintain a near-sinusoidal voltagewaveform, it is essential that the harmonic voltage factor(HVF) of the supply voltage be contained within 0.02for all 1-4 and 3-4 motors, other than design N* motorsand within 0.03 for design N motors, whereHVF= V'T c-(1.11)Hereuh = per unit summated value of all the harmonic voltagesin terms of the rated voltage V,n = harmonic order not divisible by 3 (presuming thatthe star-connected motors (normally HT motors)have only isolated neutrals) in 3-41 motors, i.e. 5, 7,11 and 13, etc. Beyond 13, the content of harmonicquantity may be too insignificant to be considered.For example, for a system having uh5 = 5%uh7 = 3%Uhll = 2% anduh13 = 1%ThenTable 1.4 Combined permissible voltage and frequencyvariations1 1: : I ;i rzVoltage +5% to -5% to +IO% to -10% tovariation -3%Frequencyvariation -28 -5 %According to IEC 60034- 1Nore IS 325 specifies voltage variation f 6% and frequencyvariation k 3% or any combination of these.Chapter 7 for special design considerations for certaintypes of load requirements.A Effect of voltage variationVoltage variation may influence the motor's performanceas shown below.(i) TorqueFrom equation (1.3), T = ,,e;, and since the standstillrotor voltage is a function of supply voltage V,:. T = VI2During start-up, at lower voltages, the starting torque is(V, + 5%, f - 5%)(Vr + 3%, f- 2%)Voltage V,HVF = (7 +-0.026131121.6.2 Voltage and frequency variationsThese have a great influence on the performance of amotor and the driven equipment, as analysed later. Themotors are, however, designed for a combined variationin voltage and frequency according to zone A of Figure1.6. The maximum variation during service must fallwithin this zone. It will permit a variation in these parametersas indicated in Table 1.4.Where, however, a higher variation in voltage andfrequency is envisaged, motors suitable to fall withinzone B in Figure 1.6 can also be manufactured. See alsoFrequency fr 2Rated ~parameter!(V, - 3%, f + 2%) -t1.03-(Vr-7%, f+3%-*IEC 60034- 12 has recommended four rotor designs, i.e. N, HandNY, HY, to define starting performance for DOL and Y/A startingsrespectively. They are along similar lines, ones to those in NEMAMG- 1. They define minimum torques, though the manufacturercan produce better ones.Figure 1.6 Voltage and frequency limits for motors

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/11reduced squarely and one should ensure that this issufficient to accelerate the load within a reasonable time,1 2 3 Iwithout injurious heat or causing a stalled condition.This aspect is discussed at length in Chapter 2. The torquehowever. improves in the same proportion at highervoltages.Example 1.2During a run, if the supply voltage to a motor terminal dropsto 85% of its rated value, then the full load torque of themotor will decrease to 72.25%. Since the load and its torquerequirement will remain the same, the motor will start to dropspeed until the torque available on its speed-torque curvehas a value as high as 100/0.7225 or 138.4% of T,, to sustainthis situation. The motor will now operate at a higher slip,increasing the rotor slip losses also in the same proportion.See equation (1.9) and Figure 1.7.CorollaryTo ensure that the motor does not stall or lock-up duringpick-up under such a condition, it should have anadequately high pull-out torque (TJ.Since the motor now operates at a higher slip, the sliplosses as well as the stator losses will increase. A circlediagram (Figure 1.16) illustrates this.Judicious electrical design will ensure a pull-out torqueslip as close to the full-load slip as possible and minimizethe additional slip losses in such a condition. See Figure1.8. A motor with a pull-out torque as close to full load slipas possible would also be able to meet a momentarilyenhanced load torque during a contingency without anyinjurious heat or a stalling condition.An increased voltage should improve the performanceof the machine in the same way by reducing slip and theassociated slip losses and also stator losses as a result oflower stator currents, but this would hold good only upto a certain increase in voltage, say up to 5%. Beyondthis, not only will the no-load losses, as discussed above,assume a much higher proportion than the correspondingreduction in the stator current and the associated losses,the winding insulation will also be subject to higherSpeed + s4 s3 s2Figure 1.8 Significance of lower Tpo slipdielectric stresses and may deteriorate, influencing itsoperating life, while at overvoltages of about 10% andhigher the insulation may even fail. Moreover. the statorcurrent may start to rise much more than the correspondingincrease in the output to account for higher no-load lossesand a poorer power factor. Figure 1.9 illustrates theapproximate effect of voltage variation on the motoroutput. Higher voltages beyond 5% may thus be moreharmful, even if the insulation level is suitable for suchvoltages. The circle diagram of Figure 1.16 explains thisby shifting the semicircle to the right, because of higherIn( and and a larger circle diameter, due to higher I,,120TPOt2t I1Os 100,40-0-F 90.+.880112 5107 5s= 1 Speed- sz SI+ Slip s2' S IFigure 1.7 Higher full-load slip at lower voltages70 095 105% Apphed voltage +* Motors are designed to develop rated output at 95% voltageFigure 1.9Effect of voltage variation on the motor output

www.taghedanesh.ir1/12 Industrial Power Engineering and Applications Handbookthereby increasing I, a proportion of which will dependupon the magnitude of voltage.Important note on starting torque, T,,It is important to note that at voltages lower than ratedthe degree of saturation of the magnetic field is affectedas in equations (1.1) and (1.5). There will also be a furtherdrop in the supply voltage at the time of start-up. The neteffect of such factors is to further influence the startingperformance of the motor. See also IEC 60034-12. Forrisk-free duties and large motors, HT motors particularly,which have a comparatively lower starting torque, it wouldbe appropriate to allow margins for such factors duringstart-up. Table 1 .5, based on the data provided by a leadingmanufacturer, shows such likely factors at various voltages.The square of these must be applied to derive the morerealistic operating starting torques at lower voltages thanrated, rather than applying only the square of the appliedvoltage. Such a precaution, however, may not be necessaryfor normal-duty and smaller motors.Example 1.3Consider a 750 kW, 4-pole motor, having a rated startingtorque Td as 125% of the rated and starting current /* as600% of the rated. At 80% of the rated voltage the startingtorque T. will reduce toTst at 80% V, = 125 x 0.745'= 69% of rated (and not 125 x 0.82 or80% of rated)(ii) Starting currentIn the same context, the starting current will also decreaselinearly according to this factor and not according to thesupply voltage. In the above example, the starting currentIst at 80% of the rated voltage will therefore reduce toISt = 600 x 0.745= 447% (and not 600 x 0.8 or 480%) of the ratedcurrent.(iii) Loud currentIn addition to the increased stator current necessitatedTable 1.5 Multiplying factor for starting torque at lowervoltagesSystem voltage as% of ratedIO09590858075706050Source Siemens Catalogue M20-1980Approx. multiplying factor(apply square of this)For 2- and 4-pole For 6- and &polemotorsmotors1 .o 1 .O0.93 0.9250.87 0.860.805 0.790.745 0.730.68 0.6650.625 0.6050.51 0.490.40 0.38by the lower voltage, the stator must draw a higher currentfrom the supply to compensate for the higher losses atlower voltages. Sinceif the supply voltage falls to 85%, the stator current forthe same load should increase byI: = IJ0.85,i.e. by roughly 18% and if the voltageincreases by 15%, the stator current should decrease toIJl.15 or 87% of Ir, i.e. areduction of 13%. These wouldbe the values when the core losses are ignored. But thecore losses will also vary with the voltage as discussedbelow. Moreover, at a higher voltage, the p.f. will alsobecome poorer and the motor will draw a higher currentto account for this. All such factors, therefore, must beconsidered when making a selection of the rating of amotor, particularly for critical applications.(iv) Core and load lossesA current-canying conductor produces two types of losses,i.e.1 Resistive losses = 12R and2 Core losses or magnetizing losses, which compriseEddy current losses andHysteresis lossesMagnetizing losses, however, as the name implies, are aphenomenon in electromagnetic circuits only. They areabsent in a non-magnetic circuit. A motor is made ofsteel laminates and the housing is also of steel, hencethese losses. Some manufacturers, however, usealuminium die-cast stator frames in smaller sizes, wheresuch losses will be less (the bulk of the losses being inthe laminations),Since the resistive loss would vary in a square proportionof the current, the motor will overheat on lower voltages(drawing higher currents). At higher voltages, while thestator current may decrease, the core losses will be higher.To understand magnetizing losses, we must first identifythe difference between the two types of losses. Bothrepresent losses as a function of the electric field, generatedby the current-carrying conductors. A current-carryingconductor generates an electric field in the space aroundit such that I = 4. The higher the current through theconductor, the stronger will be the field in the space.Some of the current will penetrate through the conductoritself, because of the skin effect and the rest will occupythe space. In an electric motor it will penetrate throughthe stator and the rotor core laminations and the housing -of the motor and cause losses in the following way:1 Resistive losses within the current-carrying conductors,i.e. within the electrical circuit itself, caused by theleakage flux (Figure 2.6), as a result of the deepconductor skin effect. This effect increases conductorresistance and hence the losses. For more details referto Section 28.7.2 Losses as caused by its penetration through themagnetic structures (core) and components existingin the vicinity. These losses can be expressed by:

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/13(u) Eddy current loss, L,:Because of the skin effectt,z .f? .B?L, = (1.12)Pi.e. = B', other parameters remaining same.where,f = supply frequencyt, = thickness of steel laminations. To reduce suchlosses, particularly in large motors, it is imperativeto use thinner laminations. To economize on thesize and bulk of windings, because of the skineffect, motor manufacturers adopt the methodof transposition when manufacturing statorwindings. See also Section 28.8.4.B = flux densityp = resistivity of the steel laminations.(h) Hysteresis loss, LhTo understand this, consider the rotor as magnetizedup to a certain level u, as illustrated in Figure 1.10.When the current (i.e. flux 9 or flux density B sinceI = 4 B 0~ H etc.) is reduced to zero, the magneticcircuit would still have a quantum of residualmagnetism, @ or B. For instance, at no current, i.e.when H = 0, there would still be a residual magneticfield in the circuit as indicated by ob and this wouldaccount for the hysteresis loss.See the curve abcda in the shape of a loop, whichis drawn at different values of H after the magneticcore was magnetized up to its saturation level oa.This is known as a hysteresis loop and the magneticarea represented by daebd as the energy stored. Thisenergy is not released in full but by only a part of it(bae) back to the magnetizing circuit when themagnetic field H (or current) is reduced to zero. Theloss of this energy (dab) is termed hysteresis lossand appears as heat in the magnetizing circuit, i.e.the stator and the rotor of an induction motor. Thisloss may be attributed to molecular magnetic friction(magnetostriction) and is represented byLh = f . (B,)'.s toi.e. = (B,)' '(1.13)Eddy current and hysteresis losses are complexquantities and can be estimated in a laboratory, on noload, in the form of power input, less I,,,?.R and frictionand windage losses etc. as shown in Section 11.5. BasedtNon-linear mag-Linear mag- netization (tendingnetization to saturation)----- &I---H 0 +HH (-I,,,) Amp/metref-Intensity of magnetlc held-Figure 1.10 Hysteresis loop and magnetizing curve illustrating hysteresis loss

~www.taghedanesh.ir1/14 Industrial Power Engineering and Applications Handbookon this, a designer can take corrective action to minimizethese losses.For all practical purposes, the core losses may beregarded as proportional to the square of the flux density.With a reduction in voltage, the flux 41 and so the fluxdensity B will decrease in the same proportion as thevoltage and so will the core losses.As a rough estimate, except for a rise or fall in thecurrent drawn by the stator, at lower and higher voltagesrespectively, the variation in the Z2R loss will be more orless counterbalanced by the fall or rise in the core loss.The core loss is proportional to B2, provided that noloadiron loss and full-load resistive loss are roughlyequal, which is so for most designs. A comparativelyhigher resistive loss to core loss will result in a higherf2R loss at lower voltages compared to a correspondingreduction in the no-load loss and vice versa. While onno-load, the no-load current will be more or less accordingto the square of the voltage.(v) Effect on power factor, efficiency and speedAt lower voltages the rotor will adjust its speed at ahigher slip. The extent of slip will depend upon the speedtorquecharacteristics of the motor and the supply voltage.The efficiency is now low but the power factor better,due to the lower inductive core loss. These figures arereversed when the voltage is high. Table 1.6 shows thevariation in the various parameters with the voltage.(vi) Pegormance of the driven equipmentThe speed of the motor is affected slightly as is thespeed of the driven load. Since the output of the load isa function of speed, it is also affected although onlymarginally, unless the variation in the voltage is substantial,which may also cause a substantial reduction in speed.The poorer the speed-torque characteristic of the motor,the higher will be the speed variation, as illustrated inFigure 1.8.B Effect of frequency variationOn a 34 system the frequency variation is normallywithin limits, which according to IEC 60034-1 is k 2%(Figure 1.6). The reason for keeping the frequencyvariation low is that it is not influenced by any conditionoutside the generating point, and at the generating point,it is maintained constant through automatic speed regulationof the prime mover (frequency is directly proportionalto the speed, equation (1.6a)). The effect of frequencyvariation, however small, is discussed below:(i) Speed Since the speed is proportional to frequency,it is affected the most and in turn influences theperformance of the driven equipment in the sameproportion as its relation to the speed.(ii) Slip The motor adjusts its speed according to thefrequency and therefore there is no change in theslip.(iii) Motor output and torque From equation (1.5)Table 1.6 Approximate effects of voltage and frequency variations on motor performanceParametersVoltage (as percentege of the rated voltage)120 110 YOFrequency (as percentage of therated frequency)I05 9.5Torque"Starting and maxrunningIncrease 44%Increase 2 1 %Decrease 19%Decrease 10%Increase 1 15%SpeedSynchronousFull loadPercentage slipEflciencyFull loadThree-quarter loadHalf loadPower factorFull loadThree-quarter loadHalf loadNo changeIncrease 1.5%Decrease 30%Small increaseDecreaqe 0 5-2%Decrease 7-20%Decrease 5-156Decrease 10-30%Decrease 1540%No changeIncrease 1%Decrease 17%Increase 0.5 to 1%Little changeDecrease I-2%Decrease 3%Decrease 4%Decrease 5-6%No changeIncrease 1.5%Decrease 23%Decrease 2%Little changeIncrease I-2%Increase 1%Increase 2-3%Increase 4-5%Increase 5%Increase 5%Little changeSlight increaseSlight increaseSlight increaseSlight increaseSlight increaseSlight increaseDecrease 5%Decrease 5%Little changeSlight decreaseSlight increaseSlight increaseSlight increaseSlight increaseSlight increaseCurrentbStartingFull loadTemp. riseMax. overloadCapacityMagnetic noiseIncrease 25%Decrease 11 %Decrease 5-6°CIncrease 44%More significantincreaseIncrease 10-12%Decrease 7%Decrease 34°CIncrease 21 %Slight increaseDecrease 10-12%Increase 11%Increase 67°CDecrease 19%Slight decreaseDecrease 5-6%Slight decreaseSlight decreaseSlight decreaseSlight decreaseIncrease 5 - 6%Slight increaseSlight decreaseSlight decreaseSlight decrease'See also the important note in Section 1.6.2A(i).hSee also Section 1.6.2A(ii).

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/15Vf - $, .J i.e. for the same applied voltage, V,, anincrease or decrease in the system frequency willdecrease or increase the flux in the same proportion.Consider the variation in the rotor current I,, withfrequency from equation (1.7):S' ,,P?I,, = ____________ -I R: + S' ,,Xisince ,,X2 = 27r. f . LwhereL = inductance of the rotor circuitor -I-.fTorqueFrom equation (I. 1)1nr --.fThe torque of the motor would approximately varyinversely as the square of the frequency.P0w.erThis will also be affected, being a multiple of T.N,i.e. by roughly l/f(iv) Effect on power factorThe no-load losses will slightly decrease or increase withan increase or decrease in the supply frequency sinceL, H' . .f' and (from equation (I. 12))Lh Bt:.f (from equation (I. 13))where the cumulative effect of flux densit is in thesquare proportion and that of frequency fY(approx.).With an increase in frequency, therefore, the core lossesdecrease slightly and vice versa. The efficiency improvesslightly and so does the power factor at higher frequenciesand vice versa. Table 1.6 shows approximate variationsin these parameters with frequency.C Effect of ambient temperatureThe motors are universally designed for an ambienttemperature of 40°C according to IEC 60034- I , unlessspecified otherwise. For temperatures below 40"C, themotor will run cooler by the amount the ambient temperatureis lower. To utilize a higher output because oflower ambient temperature may. however, not be worthwhile.For a higher ambient temperature, the end temperatureof the winding will exceed the permissible limit by theamount the ambient tempcraturc is higher. For example,for a class E motor, an ambient temperature of 50°C willcause the end temperature to reach 125°C as against115°C permissible by the resistance method. For detailsrcfcr to Sections 9.1 and 1 1.3.2.To restrict the end temperature to less than the permissiblelimits, it is essential that the motor output bereduced, or for a required output, a higher-capacity motorbe chosen. Table 1.7 gives the approximate deratingfactors for different ambient temperatures. Figure I. 1 Iis based on these figures from which the derating factoreven for intermediate temperatures can be quickly deterini ned .To consider a higher ambient temperature, we mustassess the temperature rise so affected. For instance, anambient temperature of 50°C will require the temperaturerise to be restricted by 10°C or 65/75, Le. 86.67% inclass E insulation and 70/80, Le. 87.5% in class Binsulation. The derating for insulation B will be less thanE for the same ambient temperature. But for simplicity,a derating graph has been drawn in Figure 1.12. basedon the temperature rise restrictions for class E insulation.For all practical purposes, this curve will also hold goodfor class B insulation. These values are only for guidanceand may vary from one manufacturer to another dependingon the design and the reserve capacity available in aparticular frame.Example 1.4A 100 h.p. motor is required to operate at 50°C and limit thetemperature rise to 60°C. The derating and h.p. to be chosenfor class B insulation can be determined as follows:Table 1.7 Derating for temperature rise restrictionC'iass E! Pernii,s.sible Percentage rise; temp. rise "CCluss BPerm issihlePermissible Percrnragc. rise orrtpur %i temp. rise "CNoreUnless specified otherwise, all temperature measurements of motor arc referred to by the resistance method

www.taghedanesh.ir111 6 Industrial Power Engineering and Applications Handbook10080t 6o.-400'D2Fij 208-040 50 60 70 80 90 100Ambient temperature ("C)Figure 1.11Derating curve for higher ambient temperaturesfor insulation class E or B10080- 1609s?F\408 20I I I I I I I I I I I I I I I I I I I0 t I I I ! I I I I ! ' I I ' ! I -I I I ! I t I I20 40 60 80% Temperature rise ("C)Figure 1.12 Derating curve for temperature rise restriction(drawn for class E insulation)Total permissible temperature for class B,40 + 80 = 120°CEnd temperature required 50 + 60 = 110°CNote If the manufacturer can ascertain that a 100 h.p. framehas a reserve capacity such that at full load the temperaturerise will not go beyond 60°C, the derating as calculated belowwill not be necessary.If the machine is derated for a 60°C temperature rise above4OoC, then the total temperature the motor would attain willbe 100OC. In this case, even if the ambient temperature risesto 50°C (the temperature rise remaining the same) the totaltemperature the motor will attain will be;= 50" + 60" = 110°C as desiredTherefore, derating the machine only for limiting the temperaturerise to 60°C will be adequate.Derating from Table 1.7 for restricting the temperature riseto 60°C is 84%.100 ':. h.p. required = - I e 119 h.p.0.84 . 'The next nearest size to this is a 125 h.p. motor.Effect of cooling water inlet temperatureFor large motors, which use water as the secondary coolantin a closed circuit, the temperature of the cooling air, i.e.of the primary coolant, varies with the temperature ofthe cooling water inlet temperature and its rate of flow.For the performance of the motor output, this primarycoolant, temperature has the same significance as theambient temperature for an air-cooled motor. The motoroutput is unaffected by the ambient temperature. Forsuch motors the output graph is shown in Figure 1.13 atdifferent coolant temperatures and altitudes. The ratingat 25°C inlet water temperature for water-cooled machinesis the same as for air-cooled machines at an ambienttemperature of 40°C.D Effect of altitudeMotors are designed for an altitude up to 1000 m frommean sea level, according to IEC 60034-1. At higheraltitudes, cooling reduces due to lower atmosphericpressure and should be compensated as for higher ambienttemperature. It is estimated, that for every 100 m above1000 m, the cooling is affected to the extent of 1% andthe required output must be enhanced accordingly. Table1.8 shows temperature rise restrictions for differentaltitudes above 1000 m and the corresponding deratingsbased on Table 1.7. With these values, a graph is alsodrawn in Figure 1.14, from which the derating forintermediate values can also be ascertained.Note A slightly higher derating for higher-speed motors, 1500r.p.m. and above, is required in view of the cooling which in higherspeedmotors will be affected more than in lower-speed motors.IIO]-"V I10 20 30 40Cooling water inlet ternperatue ("C)(1) Altitude up to -1000 m(2) Altitude above 1000-1500 m(3) Altitude above 1500-2000 m(4) Altitude above 2000-2500 m(5) Altitude above 2500-3000 mFigure 1.13 Effect of the cooling water inlet temperature onoutput, at different altitudes1

0.75www.taghedanesh.irTheory, performance and constructional features of induction motors 111 7Table 1.8Derating for higher altitudesAltitudetm)Permissible temp. riseabove 40°C( "C) (%)Class EPermissible outputfrom the curves(%)up to 1000 1 75a 100 100Reduction in temperature ~ I .o -rise for every 100 m above1000 m2000 I 67.5 90 943000 , 60.0 80 874000 I 52.5 70 79Class BPermissible temp. riseabove 40°C( "C) (%)80" IO0 1000.8 1 .o -72 90 9464 80 8756 70 79Permissible outputfrom the curvesm)aIEC 60034-1 has now substituted this by reduction in the ambient temperature.100t 8oc,P0'bs!e 60c8-401000 2000 3000 4000Altitude in metresFigure 1.14 Derating cutve for higher altitudesExample 1.5In the previous example, if altitude be taken as 2000 m, thena further derating by 94% will be essential.:. h.p. required = loo = 126.8 h.p.0.84 x 0.94A 125 h.p. motor may still suffice, which the manufactureralone can confirm, knowing the maximum capacity of a 125h.p. frame.1.7 No-load performanceA study of no-load performance suggests that no-loadcurrent, power factor and losses may vary in the followingproportions, depending upon the type, size and design ofthe motor:1 No-load currentThis may be 25-50% of the full-load current andsometimes even up to 60%. The higher the rating, thelower will be this current and it will depend upon themagnitude of magnetic induction (type and depth ofslots) and air gap, etc.2 No-load lossesWindage losses: A result of friction between themoving parts of the motor and the movement ofair, caused by the cooling fan, rotation of the rotor,etc.Core losses (Section 1.6.2A(iv))- Eddy current losses, caused by leakage flux- Hysteresis losses, caused by cyclic magnetizationof steel.These may be considered up to 3-8% of the ratedoutput.3 No-load power factorAn induction motor is a highly inductive circuit andthe no-load p.f. is therefore quite low. It may be ofthe order of 0.06-0.10 and sometimes up to 0.15.The above values give only a preliminary idea of theno-load data of a motor. The exact values for a particularmotor may be obtained from the manufacturer.1.8 Effect of loading on motorperformanceThe declared efficiency and power factor of a motor areaffected by its loading. Irrespective of the load, no-loadlosses as well as the reactive component of the motorremain constant. The useful stator current, Le. the phasecurrent minus the no-load current of a normal inductionmotor, has a power factor as high as 0.9-0.95. But becauseof the magnetizing current, the p.f. of the motor does notgenerally exceed 0.8-0.85 at full load. Thus, at loadslower than rated, the magnetizing current remaining thesame, the power factor of the motor decreases sharply.The efficiency, however, remains practically constant forup to nearly 70% of load in view of the fact that maximumefficiency occurs.at a load when copper losses (Z'R) areequal to the no-load losses. Table 1.9 shows an approximatevariation in the power factor and efficiency withthe load. From the various tests conducted on differenttypes and sizes of motors, it has been established that the

www.taghedanesh.ir111 8 Industrial Power Engineering and Applications HandbookTable 1.9 Approximate values of efficiency and power factor at three-quarter andhalf loads corresponding to values at full load% EfficiencyFull load Three-quarter loud Half load94 93.5 9293 92.5 9192 92 9191 91 9090 91 9089 90 8988 89 8887 88 8786 87 8685 86 8584 85 8483 84 8382 82 8181 81 1980 80 I179 79 1678 77 I477 76 1376 75 1215 74 7172 70 6470 67 6065 62 6160 57 4655 51 4550 47 35Power factorFull load Three-quurter loud Half load0.91 0.89 0.830.90 0.88 0.8 10.89 0.86 0.780.88 0.85 0.760.87 0.84 0.750.86 0.82 0.120.85 0.8 I 0.700.84 0.80 0.690.83 0.79 0.670.82 0.77 0.660.8 I 0.76 0.650.80 0.75 0.640.79 0.74 0.620.78 0.12 0.610.71 0.70 0.590.76 0.69 0.570.75 0.68 0.560.74 0.61 0.540.73 0.65 0.520.72 0.63 0.500.70 0.63 0.500.68 0.58 0.480.65 0.56 0.460.63 0.55 0.450.60 0.5 I 0.420.55 0.45 0.35extent of variation in efficiency and power factor withload is universal for all makes of motors.1.9 Effect of steel of laminations oncore lossesThe steel of laminations plays a very significant role indetermining the heating and the power factor of a motor.See Section 1.6.2A(iv). A better design with a judiciouschoice of flux density, steel of laminations and its thicknessare essential design parameters for a motor to limit thecore losses to a low level.For a lower range of motors, say up to a frame size of355, the silicon steel normally used for stator and rotorcore laminations is universally 0.5-0.65 mm thick andpossesses a high content of silicon for achieving betterelectromagnetic properties. The average content of siIiconin such sheets is of the order of 1.3-0.8% and a core lossof roughly 2.3-3.6 Wkg, determined at a flux density ofI Wdm2 and a frequency of 50 Hz. For medium-sizedmotors, in frames 400-710, silicon steel with a still bettercontent of silicon, of the order of 1.3-1.8% having lowerlosses of the order of 2.3-1.8 W/kg is preferred, with athickness of lamination of 0.5-0.35 mm.For yet larger motors of frame sizes 710 and above,core losses play a more significant role, and require veryeffective cooling to dissipate the heat generated. Coolingof larger machines, complicated as it is in view of theirsize and bulk, necessitating core losses to be restrictedas low as possible.To meet this requirement, the use of steel with a stillbetter silicon content and lower losses is imperative. Acold-rolled non-grain oriented (CRNGO) type of sheetsteel is generally used for such applications, in thethickness range of 0.35-0.5 mm, with a higher siliconcontent of the order of 2.0-1.8% and losses as low as1 .O-1 .S W/kg.When the correct grade of steel is not available thecore losses may assume a higher proportion and requirea reduction in output or a larger frame than necessary.The data provided above are only for general guidance,and may vary slightly from one manufacturer to anotherand according to the availability of the silicon-grade steelat the time of manufacture.1.10 Circle diagramThis is a very useful nomogram to determine theperformance of a motor with the help of only no-loadand short-circuit test results. In slip-ring motors, it alsohelps to determine the external resistance required in therotor circuit to control the speed of the motor and achievethe desired operating performance. Slip-ring motors arediscussed in Chapter 5. The concept behind this nomogramis that the locus of the rotor and the stator currents is acircle. Consider the equivalent circuit of an inductionmotor as shown in Figure 1.15, whereR, = stator resistanceX, = stator reactance

www.taghedanesh.irTheory, performance and constructional features of induction motors 111 9. _ ~I,, = /I;, + I ,/A = Loss or active component supplying the hysteresis and eddycurrent losses to the stator core.I,,, = Magnetizing or reactive component producing the field (flux).la = Active or torque producing component.Figure 1 .I 5 Simple equivalent circuit diagram of a motorRi = rotor resistance referred to statorS.s,X2 = X; = rotor reactance referred to statorI,( = no-load currentRe' = RS . I-s is the external rotor resistanceSreferred to the stator.All these values are considered on per phase basis.1.10.1 Drawing the circle diagram (Figure 1.16)Take V, on the vertical axis and draw I,( at an angle$ne obtained from the no-load test.From a short-circuit test draw the start-up current I,,at an angle &.Join AB and determine the centre C and draw the circle.The diameter of the circle can also be determined by:BB' will determine the locked rotor torque and powerloss while the rotor is locked.Divide BB' at M in the ratio of R; : RI and join AMwhere Ki = R2( $)'IR2 = rotor resistance per phaseZ, = number of turns in the stator circuit per phaseZ, = number of turns in the rotor circuit per phase5 At load current I,, i.e. OP, the rotor current I,, is AP.PP, will determine the power input per phase and theperformance of the motor as follows:(i)(ii)(iii)(iv)(v)(vi)Power input = 3 . V, . PPI wattsCore and friction loss = 3 . V, . DI D2 wattsStator copper loss = 3 Vi . MI Po wattsRotor copper loss = 3 . V, . /MI wattsMotor output = 3 . Vt . Ph wattsAB is known as the output line, since the outputis measured above this.Running torque,T = 3 . V, . PM, synchronous wattsT.N,and since 3 ' Vt . PMl = ~0.9743 V( '0.974' PMI:. T = mkg.N,This is the maximum torque that can bedeveloped by the rotor during a run, but theuseful torque will be in accordance to output,1.e.3. V, . Pb'0.974 mkgN,Since the maximum torque is measured by lineAM, it is known as the torque line.(vii) Starting torque or short-circuit torqueBMT,, = 3 . V( ' ~ ' 0.974Nr(viii) Full-load slip, S = - bM,PMI1.10.2 Inference from the circle diagramThe maximum value of the output and torque of themotor can be obtained by dropping perpendiculars CCIand CC3 on the output and torque lines respectively fromthe centre C.CIC2 and C3C4 indicate the magnitude ofthe maximum output and torque, respectively, that themotor can develop. This torque is the pull-out torqueTpo In slip-ring motors it can be obtained at any speedon the normal speed-torque curve by inserting a suitableresistance into the rotor circuit to vary the slip.' L0 I&9 4 Figure 1.16 Circle diagram

www.taghedanesh.ir1/20 Industrial Power Engineering and Applications HandbookUnder normal conditions, the starting current is toohigh, whereas the corresponding starting torque is not sohigh. If we can shift the Zst point to the left, or in otherwords increase the slip, the torque will increase and IStwill decrease. This is used, in a slip-ring motor to controlboth Tst and Ist with suitable external resistances. Forinstance, if the starting torque is required to be equal toTpo, Le. C3C4, then the start-up current should be controlledto OC3. The starting resistance, R,, therefore should besuch as to achieve this current during start-up:ss e2Le. R,, = -AC3and external resistance Re = RSt - R2CorollaryThe higher the full load slip, the higher will be the rotorlosses and rotor heat. This is clear from the circle diagramand also from equation (1.9). An attempt to limit thestart-up current by increasing the slip and the rotorresistance in a squirrel cage motor may thus jeopardizethe motor’s performance. The selection of starting currentand rotor resistance is thus a compromise to achieveoptimum performance.1.1 1 Types of induction motorsThese are of two types:I Squirrel cage rotor2 Slip-ring or wound rotorIn a squirrel cage motor the rotor winding is made ofsolid metallic rods short-circuited at both ends of therotor. Short-circuiting of rotor bars leads to fixed rotorparameters. In slip-ring motors the rotor is also woundlike the stator and the six winding terminals are connectedto a slip-ring assembly. This gives an opportunity to varythe rotor circuit impedance by adding external resistanceand thus vary the rotor circuit parameters to achieve therequired performance.Although it may seem easy to alter the speed-torqueand speed-current characteristics of such a motor throughits rotor circuit, the use of such motors is recommendedonly for specific applications where the use of a squirrelcage motor may not be suitable. The reason is its slipringsand the brushes which are a source of constantmaintenance due to arcing between the rings and thebrushes, besides a much higher initial cost and equallyexpensive control gears.Specific applications for such motors are rolling mills,rice mills, paper mills and cranes etc. for one or more ofthe following reasons:1 To contain the start-up inrush current, as a result oflow start-up impedance and to control the same asneeded through external resistance in the rotor circuit.2 To provide a smoother start.3 To meet the requirement of a high start-up torque andyet contain the start-up inrush current.4 To meet the load demand for more frequent starts andreversals, as for cranes and other hoisting applications.S To achieve the required speed variation throughvariation in rotor circuit impedance.Nore The latest trend, however, is to select only squirrel cagemotors as far as practicable and yet fulfil most of the above loadrequirements. Fluid couplings and static (IGBT or thyristor) drivescan meet the above requirements by starting at no load or light loadand controlling the speed as desired, besides undertaking energyconservation. See also Chapters 6 and 8.1.11.1 Choice of voltageBecause of heavy start-up inrush currents, the use of LTmotors should be preferred up to a medium sized ratings,say, up to 160 kW, in squirrel cage motors and up to750 kW in slip-ring motors. For still higher ratings, HTmotors should be used.1.12 Mounting of motorsThe common types of mountings are illustrated in IEC60034-7, and the most frequently used are shown in Figure1.17. The legends to represent these mountings are givenbelow:1 Floor mounting2 Wall mounting3 Ceiling mounting4 Wall mounting5 Wall mounting6 Flange mounting7 Flange mounting8 Flange mounting9 Foot-cum-flange mounting10 Foot-cum-flanges on both sidesB3B6fB7B8v5V6B5v1v3B3fB5B17Nore For brevity, the prefix IM, representing ‘InternationalMounting’, has been omitted from thcsc legends. In common practicethe different types of mountings are represented as noted above.While designation B represents a horizontal mounting, designationV represents a vertical mounting. For more details and numericaldesignation of motors refer to IEC 60034-7.Care should be taken in selecting the mounting for applicationswhere the motor weight would apparently fall on its flange or thefoot, as in mountings 55, B6, B8, V3, V5 and V6. In such mountings,the foot or the flange is subject to a shearing force due to a cantilevereffect and is vuinerable to breakage. For small sizes, where thcweight of the motor may not matter so much, these types of mountingscan be employed, otherwise reinforcement may be necessary eitherto the foot or to the flange, if possible. For larger motors it isadvisable to select an alternative mounting.1.1 3 EnclosuresA motor can be constructed in different enclosures tosuit a particular location as follows.1.13.1 Protected motors (degree of protectionIP 12, 21 or 23)1Screen protected (SP): Not a common enclosure.For locations clear of dust and water.

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/21Foot mounting (83)Foot mounting (88) Flange mounting (835) Foot mounting (86)Foot mounting (V5)Flange mounting (VI)Foot mounting (87)Foot mounting (V6). - . -. - .Foot-cum-flange mounting (817)Flange mounting (85)Flange mounting (V3)Figure 1.17 Types of mounting arrangements2 Screen protected, drip proof (SPDP): As above, butwith additional protection against dripping of waterFigures 1. I8(a)-(d).3 Splash proof: As in (2) but, unless the degree ofprotection according to IEC 60034-5 is specified,this remains a vague term and is not in frequentuse.1.13.2 Totally enclosed fan-cooled motors(TEFC) (degree of protection1P 44, 54 or 55)These motors are suitable for locations prone to dust,coal dust and metal particles etc. and occasional waterspray and rain (Figures 1.19(a) and (b)).Nofr SPDP motors also have a cooling fan but their surface isplain. whereas TEFC motors have fins on their housings that add totheir cooling surface. Refer to Figures 1.18(at(c) and 1.19(a) and (b).1 .I 4 Weatherproof motors (WP)(degree of protection IP 55)These are an improvised version of a totally enclosedmotor to protect its live parts from ingress of water, rain,snow or airborne particles etc. This is achieved byproviding them with additional protection such aslabyrinths at joints such as bearing end covers, centrifugaldiscs, end shields and terminal box and giving the outersurface a special coat of paint to protcct against unfavourableweather conditions.1.14.1 NEMA enclosuresThese have only a historical significance, in the light ofbetter defined enclosures now available. They are brieflydefined here:Type I A weather-protected Type I machine is anopen machine with its ventilating passages so constructedas to minimize the entrance of rain, snowand air-borne particles to the electrical parts and havingits ventilated openings so constructed as to preventthe passage of a cylindrical rod 3/4-inch in diameter.Type I1 A weather-protected Type I1 machine willhave, in addition to the enclosure defined for a weatherprotectedType 1 machine, ventilating passages at bothintake and discharge ends so that high-velocity airand air-borne particles blown into the machine bystorms or high winds can be discharged withoutentering the internal ventilating passages leadingdirectly to the electrical parts of the machine. Thenormal path of the ventilating air which enters theelectrical parts of the machine is so arranged by bafflingor separate housings that it provides at least threeabrupt changes in its direction, none of them beingless than 90". In addition. an area of low velocity notexceeding 600 feet per minute is also provided in theintake air path to minimize the possibility of moistureor dirt being carried into the electrical parts of themachine.

www.taghedanesh.ir1/22 Industrial Power Engineering and Applications HandbookFigure 1.18(a) Screen protected drip proof (SPDP) squirrelcage motor (Cooling system ICOAI)Figure 1.18(b) Screen protected drip proof slip ring motor(Cooling system ICOAI)Figure 1.18(c)Large SPDP squirrel cage motor (enclosure IP 12) (Cooling system ICOAI)1 Access for checking air gap2 Air-deflecting baffle3 Coil bracing ring4 Fan5 Rotor end ring6 Rotor bars7 Stator core8 Fully-formed coils of the twolayer stator winding10 Core duct separator11 Preformed coil in section12 End winding connections13 Bearing endshield14 Terminal box with bolted oncable sealing end15 Shaft16 Grease ejector handle17 Grease collector18 Anti-friction bearing withgrease regulator19 Grease impeller19181716 15 14Figure 1.18(d) Cross-sectional view of a large screen protected motor showing the cooling circuit (Cooling system ICOAI)(Courtesy: NGEF Ltd)

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/231.1 5 Degree of protectionSquirrel cage rotorThe nomenclatures used above to define an enclosurewere earlier interpreted in different ways by differentmanufacturers. To achieve harmonization, IEC 60034- 1has eliminated the use of these codes. Instead, designationIP, followed by two characteristic numerals according toIEC 60034-5, is now introduced to define an enclosure.The first characteristic numeral defines the protection ofpersonnel from contact with live or moving parts insidethe enclosure and of machines against the ingress ofsolid foreign bodies. The second numeral defines thetype of protection against ingress of water. Tables 1.10and 1.11 show these requirements.Table 1.10 Types of protection against contact with live ormoving partsFirstcharacteristicnumber as inIEC 60034-5Type of protectionFigure 1.19(a) TEFC squirrel cage motor (Cooling system ICOAI)(Courtesy: NGEF Ltd)Slip ring rotorFigure 1.19(b) TEFC slip ring motor (Cooling system KOA1(Courtesy: NGEF Ltd)No special protection of persons against accidentalor inadvertent contact with live or moving partsinside the enclosure.No protection of equipment against ingress ofsolid foreign bodies.Protection against accidental or inadvertentcontact with live and moving parts inside theenclosure by a larger surface of the human body,for example a hand, but not against deliberateaccess to such parts.Protection against ingress of large solid foreignbodies (diameter greater than 50 mm).Protection against contact with live or movingparts inside the enclosure by fingers.Protection against ingress of small solid foreignbodies (diameter greater than 12 mm).Protection against contact with live or movingparts inside the enclosure by tools, wires or objectshaving a thickness greater than 2.5 mm.Protection against ingress of small solid foreignbodies (diameter greater than 2.5 mm).Protection against contact with live or movingparts inside the enclosure by tools, wires, or suchobjects of thicknesses greater than 1 mm.Protection against ingress of small solid foreignbodies (diameter greater than 1 mm) excludingthe ventilation openings (intake and discharge)and the drain hole of the enclosed machine whichmay have degree 2 protection.Complete protection against contact with live ormoving parts inside the enclosure.Protection against harmful deposit of dust. Theingress of duct is not totally prevented, but dustwill not be able to enter in an amount sufficientto harm the machine.Totally dust-tight. No ingress of dust.

www.taghedanesh.ir1/24 Industrial Power Engineering and Applications HandbookTable 1.11SecondcharacteristicnumberTypes of Protection against ingress of waterType of protectionNo special protectionDripping water (vertically falling droplets) willhave no harmful effect.Droplets of water falling at any angle up to 15"from the vertical will have no harmful effect.Water falling as a spray at an angle equal to orsmaller than 60" from the vertical will have noharmful effect.Water splashed under stated conditions againstthe machine from any direction will have noharmful effect.Water injected under stated conditions through anozzle against the machine from any directionwill have no harmful effect.Water from heavy seas will not enter the machinein a harmful quantity.Ingress of water in the machine immersed inwater under stated conditions of pressure andtime will not be possible in a harmful quantity.Ingress of water into the machine immersed inwater under specified pressure and for anindefinite time will not be possible in a harmfulquantity.1.16 Cooling systems in largemotorsThe cooling system in large motors becomes vital, asone fan cannot cover the entire length of the motor bodyor cool the inside bulk of the motor windings. Now amore judicious design is required for adequate coolingto eliminate any hot spots in the rotor, stator or theoverhangs of the stator windings and bearings etc. Thereare many cooling systems adopted by various rnanufacturers,depending upon the size of the machine andthe heat generated in various parts during full-loadcontinuous running. The cooling system may be selfventilated,closed circuit, not requiring any external sourceto augment the cooling system, or a forced cooling system,employing an external source, to basically work as heatexchangers to dissipate the heat. Thus, there may be avariety of cooling systems to cool a large machine.IEC 60034-6 has specified a number of probable coolingsystems, as adopted by various manufacturers. The morecommonly used practices are shown in Table 1.12.According to this specification any cooling system maybe expressed by the letters IC (international cooling)followed by1 A number to indicate the arrangement of the coolingcircuit as in column 1 of Table 1.12.2 Each cooling circuit is then identified for the primarycooling medium by a letter A, H or W etc. whichspecifies the coolant as noted below:34For gasesFor liquidsAir - AFreon - FHydrogen - HNitrogen - NCarbon dioxide - CWater - WOil - UThe letter is then followed by a number, describingthe method to circulate the coolant as in column 3 ofTable 1.12.Another letter and a number are added after the aboveto describe the secondary cooling system.ExampleCoding arrangementas in column 1, Table 1.12Primary cooling systemMethod of circulating thecoolant as in column 3 ofTable 1.12Secondary cooling system 1JI C 3 A l W 6Depending upon its size, a machine may adopt morethan one cooling system, with separate systems for thestator and the rotor and sometimes even for bearings. Todefine the cooling system of such a machine, each systemmust be separately described. For more details refer toIEC 60034-6.The following are some of the more prevalent systemsfor totally enclosed large machines:Tube Ventilated Self Cooled (TV)Closed Air Circuit Water Cooled (CACW)Closed Air Circuit Air Cooled (CACA)The above cooling systems will generally comprisethe following:1 Tube ventilation In this system cooling tubes whichwork as heat exchangers are welded between the corepacket and the outer frame and are open only to theatmosphere. See to Figures 1.20 (a)-(c). One fan insidethe stator, mounted on the rotor shaft, transfers theinternal hot air through the tube walls which form theinternal closed cooling circuit. A second fan mountedoutside at the NDE blows out the internal hot air ofthe tubes to the atmosphere and replaces it with freshcool air from the other side. This forms a separateexternal cooling circuit.2 ClosedAir Circuit Water Cooled (CACW) The motor'sinterior hot air forms one part of the closed air circuitthat is circulated by the motor's internal fans. A separateheat exchanger is mounted on top of the motor as thecooling water circuit. This forms the second coolingcircuit.I

~~ ~www.taghedanesh.irTable 1.12 Normal systems of cooling for totally enclosed large machinesTheory, performance and constructional features of induction motors 1/25First characteristicnumber to indicatethe cooling system1 20IJ56DescriptionFree circulation of the coolant from themachine to the surrounding mediumInlet pipe-circulation: The coolant flows tothe machine through inlet pipes from a sourceother than the surrounding medium and thenfreely discharges to the surrounding medium(as in the use of separately driven blowers)Outlet pipe circulation: The coolant is drawnfrom the surrounding medium but is dischargedremotely through the pipesInlet and outlet pipe circulation: The coolantflows from a source other than the surroundingmedium through the inlet pipes and isdischarged remotely through the outlet pipesFrame surface cooled (using the surroundingmedium): The primary coolant is circulatedin a closed circuit and dissipates heat to thesecondary coolant, which is the surroundingmedium in contact with the outside surface ofthe machine. The surface may be smooth orribbed, to improve on heat transfer efficiency(as, in a TEFC or tube ventilated motor (Figures1.19 and 1.20)Integral heat exchanger (using surroundingmedium): As at No. 4 above, except that themedium surrounding the machine is a heatexchanger, which is built-in as an integral partof the machine like a totally enclosed. tubeventilatedmotor (Figure 1.20)Machine-mounted heat exchanger (using thesurrounding medium): As at No. 5 above,except that the heat exchanger is neitherexternally mounted nor forms an integral partof the machine. Rather it is mounted as anindependent unit, directly on the machine(Figures 1.21 and 1.22)Integral heat exchanger (not using thesurrounding medium): As at No. 5 above,except that the cooling medium is differentfrom the surrounding medium. It can be liquidor gasMachine-mounted heat exchanger (not usingthe surrounding medium): As at No. ‘6’ aboveexcept that the cooling medium is differentfrom the surrounding medium. It can be liquidor gas (Figures 1.21 and 1.22)Separately mounted heat exchanger: Theprimary coolant is circulated in a closed circuitand dissipates heat to the secondary coolant.It can be a heat exchanger as an independentunit separately mountedSecond characteristicnumber for means ofsupplying power tocirculate the coolant3 40123456789DescriptionFree convection: No external power sourceis essential. Heat dissipation is achievedthrough natural convection like a surfacecooled motorSelf-circulation: Movement of the coolantis normally through a fan mounted on therotor shaft, like a normal fan cooled motor(Figures l.lS(a)-(d) and 1.19(a) and (b)Circulation by integral independentcomponent: Like a fan, driven by anelectric motor, and the power is drawn froma separate source. rather than the mainmachine itselfCirculation by independent componentmounted on the machine: As at No. 5above, but the movement of the coolant iathrough an intermediate component andmounted on the machine and not an integralpart of the machineCirculation by an entirely separatesystem: As at No. 6 above, but thecirculation of the coolant is by an entirelyindependent system, not forming a part ofthe main machine in any way and mountedseparately like a water-distribution systemor a gas-circulation systemCirculation by relative displacement: Asat No. 0 above, except that instead ofsurface cooling the cooling is achievedthrough the relative movement of thecoolant over the machineThis numeral is used for circulation byany means other than stated above

www.taghedanesh.ir1/26 Industrial Power Engineering and Applications HandbookFigure 1.20(a) Totally enclosed tube ventilated (TETV) squirrel cage motor (Cooling system IC5A111) (Courtesy: BHEL)Figure 1.20(b) A typical cooling circuit type IC5AIAI1 Lifting lug 12 Handle for emptying 1 2 3 4 5 6 7 8 9Air baffleCoil bracing ringCooling tubesShort-circuiting ringStator core packetTwo-layer fully formed13141516grease collecting boxWelded frameTerminal box withcable sealing boxRotor core packetSection bars of thecoils of stator winding squirrel cage8 Air guide shell9 Bearing endshield1718Rotor end plateGrease collecting box1011Fan for outer aircircuitFan hood withprotective grid forcooling air intake192021Grease thrower oflabyrinth sealOpening for checkingair gapFan for inner aircircuit11011121312 17 16 15 14Figure 1.20(c) Cross-sectional view of a large tube-ventilated squirrel cage motor showing the cooling circuit (Cooling systemIC5AIA1) (Courtesy: NGEF Ltd)

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/27The heat exchanger consists of a large number ofcooling tubes connected to the stator through headers/ducts. The tubes may have coils of copper wire woundaround them to enhance their cooling capacity. Filteredwater (soft water), to avoid scaling of tubes, is circulatedthrough these tubes. The hot air circulating throughthe motor stator and rotor ducts passes through theseheat exchangers and becomes cooled. See Figure 1.21.3 Closed Air Circuit Air Cooled (CACA) This coolingsystem is the same as for CACW except that, insteadof water, air flows through the top-mounted heatexchangers. See Figures 1.21 and 1.22.1 .I7 Single-phase motorsApplication - Domestic appliances- Small machine tools- Industrial and domestic fans, pumps,polishers, grinders, compressors andblowers etc.1 .I 8 Theory of operationA single-phase winding cannot develop a rotating field,unlike a multiphase winding. But once it is rotated, itwill continue rotating even when the rotating force isremoved so long as the winding is connected to a supplysource. To provide a rotating magnetic field, an auxiliarywinding or start winding is therefore necessary acrossthe main winding. It is placed at 90" from the mainwinding and connected in parallel to it, as shown inFigures 1.23 and 1.24. The impedances of the twowindings are kept so that they are able to provide a phaseshift between their own magnetic fields. This phase shiftprovides a rotating magnetic field as already discussed.The auxiliary windings may be one of the following types:I Split phase windingWhen another inductive winding is placed across themain winding (Figure 1.23(a) and (b)) so that RIX,,of the auxiliary winding is high, a phase shift willoccur between the two windings. This shift will below and much less than 90°, as explained in the phasordiagram (Figure 1.23(c)). But it can be made adequateby increasing the R, so that a rotating field may developsufficiently to rotate the rotor. The higher the ratioRIX,,, the higher will be the starting torque, as RIX,,will move closer to the applied voltage V, and help toincrease the phase shift. In such motors the startingtorque, T,,, is low and running speed-torquecharacteristics poor as illustrated in Figure 1.23(d).Figure 1.23(e) shows a general view.2 Capacitor start windingIf the inductive auxiliary winding is replaced by acapacitive winding by introducing a capacitor unit inseries with it (Figure 1.24(a) and (b)) the phase shiftwill approach 90" (Figure 1.24(c)) and develop a highstarting torque. When this capacitor is removed on arun, the running torque characteristics become thesame as for a split-phase motor. Figure 1.24(d)illustrates a rough speed-torque characteristics of sucha motor.In both the above methods a speed-operatedcentrifugal switch is provided with auxiliary windingto disconnect the winding when the motor has reachedabout 75-85% of its rated speed. Figure 1.24(e) showsa general view.3 Capacitor start and capacitor run windingsWhen the running torque requirement is high butthe starting torque requirement not as high then aFigure 1.21 Closed air circuit, air cooled (CACA) squirrel cage motors (likely cooling systems IC6AlA1 or IC6AlA6)(Courtesy: BHEL)

www.taghedanesh.ir1/28 Industrial Power Engineering and Applications HandbookframeFigure 1.22 Cooling cycle for a CACA (IC6AlA6) or CACW(IC9A6W7) motorcapacitor of a low value, so that the capacitor currentmay remain less than the magnetizing componentsof the two windings, may be provided and thedisconnecting switch removed. Figures 1.25(Ul)and (b,j are drawn with the switch removed. Thestarting torque in this case may not be very highbut the running torque would be higher as required.The value of capacitor C1 would depend upon thevalue of L1 and the running torque requirement.We can improve the starting performance of theabove method by providing C in two parts, one forstart Cz, of a much higher value, depending uponthe requirement of TFt, through a disconnect switch(Figures 1 .25(u2) and (b2)), and the other C,, for arun of a much lower value (so that IC, < Zmj.Notes1 The size of capacitors C, C, or C2 will depend upon thehorsepower of the motor and the torque requirement of theload. For starting duty capacitors generally in the range of 30-100 pF and for a run of 2-20 pF will be adequate.2 Whenever frequent switchings are likely, high transient voltagesmay develop and harm the motor windings and the capacitors.Fast discharge facilities must be provided across the capacitorterminals to damp such transients quickly. See Section 25.7, formore details on discharge devices.4 Shaded pole motorsApplications requiring extremely small motors, in bothsize and horsepower, may be designed for shadedpole construction. Electronic drives, cassette players,recorders and similar applications need an extremelysmall size of motor, as small as 1 W (1/746 h.p.j. Suchmotors can he designed in shaded pole.The stator is of a salient pole type that protrudesoutwards within the stator housing similar to a d.c.machine but is made of steel laminations. A smallside end portion of each pole is split and fitted with aheavy copper ring as shown in Figure 1.26(a). Thisring is called a shading coil, as it shades the normalflux distribution through that portion of the pole andsubstitutes for a split phase and provides the requiredsecond winding. The stator poles are wound as usualand the end terminals are brought out to receive thea.c. supply. Figure 1.26(b) illustrates a simple twopolemachine. When the voltage is applied across thestator windings, a magnetic flux is developed in theentire pole, which cuts the copper ring arranged atthe tip of the pole. The main flux, thus cutting thecopper coil (ring), induces a current in the ring.The current in the copper ring opposes the mainflux in that area of the pole and behaves like an artificialsecond winding, and develops a rotating field. Althoughthe torque so developed is extremely low, it is enoughto rotate such small drives, requiring an extremelylow starting torque, of the order of 40-50% of thefull load torque.

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/290- Vr(1-4) -0MainIm windingCentrifugal switchopens hereDisconnectswitchStart hndingXL > XL,RNote Phase shift is obtained by increasing -XL1(a) Schematic diagramt3PCOT,Tsi1 ; "1 I,",.StartYwinding(b) General arrangementISpeed Nr ---+ 7585% N,(d) Speed-torque characteristics of a split phase motorshiftb .. ..lr(c) Phasor diagramLow starting and running torquesFigure 1.23Split-phase winding(e) Split phase 1-4 motor [Courtesy: AUE (GE Motors)]Since there is only one winding and the poles arealready shaded at one particular end, the direction ofthe rotating flux is fixed and so is the direction ofrotation of the rotor. The direction of rotation cannotbe altered as in the earlier cases. Since there is onlyone winding and no need of a speed-operatedcentrifugal switch, these motors require almost nooperational maintenance.5 Universal motorsThese are series motors and are relatively compactand lightweight compared to an a.c. motor. The useof such motors is therefore common for hand toolsand home appliances and also for such applicationsthat require a high speed (above 3200 r.p.m) which isnot possible in an a.c. machine. Likely applicationsare polishers, grinders and mixers. This motor runsequally well on both a.c. and d.c. sources of supply.The motor is designed conventionally, with alaminated stator, a static magnetic field and a rotatingarmature, as shown in Figure 1.27(a) and (b). Thearmature and the field windings are connected in seriesthrough two brushes, fitted on the armature extendedcommutator assembly, to obtain the same direction offield and armature currents. Thus, when the directionof the line current reverses, the field and armaturecurrents also reverse. When operated on a.c., the torqueproduced is in pulses, one pulse in each half cycle asillustrated in Figure 1.27(c). The normal characteristicsfor such motors are also illustrated in Figure 1.27(d).The no-load speed may be designed very high, to theorder of 2000-20 000 r.p.m. but the speed on loadmay be around 50-80% of the no-load speed due towindage and friction losses, which constitute a higherpercentage for such small to very small motors (l/loto 1 h.p.). The required output speed for the type ofapplication can be obtained through the use of gears.

www.taghedanesh.irMainI, Im winding- - -Start hinding(a) Schematic diagram(c) Phasor diagramHigh start but low running torquesCentrifugal switcheET-Disconnectswitch! -I,r -ImiLYStart winding(b) General arrangement01@ Capacitor start and run windings@ Run winding@ Capacitor start and capacitor run windingsNr(d) Speed-torque characteristics of capacitor start andcapacitor run motors(e) Capacitor start or capacitor start-capacitor run 1-0 motorFigure 1.24Capacitor start winding

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/31r- VW-S) -7 y- Vr(1-d -7- -LMainL1wCII -(ai) Schematic diagramStart windingCz = 5 to 6 times Cl(a2) Schematic diagramvr(l-$) windingStart winding -General arrangement(bl) Low start but high running torquesStart windingCl = Run capacitorC, = Start capacitorGeneral arrangement(b2) High start and high running torquesFigure 1.25 Capacitor start and capacitor run windingsShading coil (copper ring)Laminated stator coreSquirrel cage rotor 7Shading coil (copper ring)Figure 1.26(a) General arrangement ofa shaded pole motorFigure 1.26(b) Shaded pole 1-g motor [(Courtesy: AUE (GE Motors)]

www.taghedanesh.ir1/32 Industrial Power Engineering and Applications HandbookRotatingStatic armatureseries field0 m .* Vr(a.c. or d.c.)(a) Schematic diagramII1/ No-load ,feedPositive half cycle'I I i'-GIII0TrTorque (T) -T,, I, and N, are the rated values(d) Characteristics of a universal motorNegative half cycle* ax. 1-@or d.c.(b) Theory of operation-TimeNegative half cycle(c) Pulsating torqueon 1-0 a.c. supplyFigure 1.27 Theory of operation of a universal motor

www.taghedanesh.irTheory, performance and constructional features of induction motors 1/33Relevant StandardsIEC Title IS BS IS060034-1/199660034-5/199 16OO34-6/199 160034-711 99260034- 12/198060038/199460072- 1 /I 99 160072-2/199060072-3/199460529/198960617-1 to 13--NEMAlMG 111 993NEMAlMG 211989NEMNMG 10/1994NEMA/MG 11/1992NEMA/MG l3/1990ANSI C 84.111995Rotating electrical machinesRating and performanceRotating electrical machines.Classification of degrees of protectionprovided by enclosures for rotatingmachineryRotating electrical machines. Methodsof cooling (IC Code)Rotating electrical machines. Classificationof types of constructions and mountingarrangementsRotating electrical machines. Startingperformance of single-speed three-phase cageinduction motors for voltages up to andincluding 660 VIEC Standard voltagesDimensions and output series for rotatingelectrical machines. Frame number 56 to 400and Flange number 55 to 1080Dimensions and output series for rotatingelectrical machines. Frame number 355 to 1000and flange number 1 180 to 2360. Dimensionsand output series for rotating electricalmachinesSmall built-in motors. Flange number BF IOto BF SOSpecification for degrees of protectionprovided by enclosures (IP Code)Graphical symbols for diagramsPreferred numbersDimensions of motors for general use4122/1992325/19964691/19856362/19952253119748789/1996123 VI991123 1/1991996/ 199 I469111985120321076119858223/1976Related US Standards ANSJINEMA and IEEEBS EN 60034-1/1995BS 4999- 105/ I988BS EN 600346/ 1 994BS EN 60034.7/1993BS EN 60034-12/1996BS 5000-1O/1989 -BS 4999- 14 1 /I 987BS 5000-10/1989 -BS 4999-103/1987BS 5000-11/1989BS EN 60529/1992BSEN 606 17-2 to 13BS 2045/1982 3,17,497BS 2048-VI989Motors and generators ratings, construction, testing and performanceSafety Standards (enclosures) for construction and guide for selection, installation and use of rotating machinesEnergy management guide for selection and use of three-phase motorsEnergy management guide for selection and use of single phase motorsFrame assignments for ax. integral horsepower induction motorsElectric power systems and equipment - voltage ratings (60 Hz)Notes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.AbbreviationsANSI American National Standards Institute - USABS British Standards - UKIEC International Electro Technical Commission - SwitLerlandIEEE The Institute of Electrical and Electronics Engineers - USAIS Indian Standards - IndiaIS0 International Standards Organisation - SwitzerlandNEMA National Electrical Manufacturers’ Association - USA

www.taghedanesh.ir1/34 Industrial Power Engineering and Applications HandbookList of formulae used for S high, I, = -5s e2Theory of operationT = (Pm . [rr$ = $, sin (utT = torque developed&, = maximum field strengthI, = rotor currentT.N(1.1)E< x2Motor output and torqueP, = rotor powerP, = synchronous powerp=- 974P, - P, = s . P,P = rotor power in kWT = torque in mkgN = speed in r.p.m.(1.3)P, - P, = slip loss(1.7a)(1.8)(1.8a)(1.9)(1.10)S= slipR2 = rotor resistance per phase,,X2 = standstill rotor reactance per phase and,,e2 = standstill rotor induced e.m.f. per phase.System harmonics(1.3a)n(1.1 1)Tst = starting torqueHVF = harmonic voltage factorV, = per unit harmonic voltagesn = harmonic order not divisible by 3(1.3b)(a) Eddy current lossT, = rated torqueLe = t:.f2.B2Ip(1.12)e2=-zI.-dt)dt(1.4) tl = thickness of steel laminationsB = flux densityp = resistivity of the steel laminationszr = number of turns in the rotor circuit per phase andd@dt = rate of cutting of rotor fluxe2 = 4.44 Kw . @ , . zr . ,f, (1.5)Kw = winding factorf, = rotor frequency = S . f(b) Hysteresis lossLh = f . (Bm)'3 to 2 (1.13)N, = synchronous speedN, = rated speed Further reading120. fN, = - r.p.m.Pf= frequency of the supply system in Hzp = number of poles in the stator windingRotor current(1.6a)1 Golding, E.W., Electrical measurements and measuringInstruments, The English Language Book Society and Sir IsaacPitman & Sons Ltd, London.2 Humphries, J.T., Motor and Controls for l-$ motors. Merrill,Columbus, OH.3 Puchstein, A.F., Lloyd, T.C. and Conard, A.G., AlternatingCurrent Machines, Asia Publishing House, Bombay (1959).4 Say, M.G., The Perjkrmancr und Design ofAlternating CurrentMachines, Pitman & Sons Ltd, London (1958).5 Smeaton, R.W., MotorApplication and Maintenance Hand Book,(1.7)McGraw Hill, New York (1969).

www.taghedanesh.irMotor Torque,Load Torque andSelection ofMotorsontents.2 NEMA rotor designs 22.7.3 Heating during an on-load start-up 2/442.8.1Heating phenomenon in a motor duringa stalled condition 2/45of the motor 2t of formulae used 2/48

www.taghedanesh.ir

www.taghedanesh.irMotor torque, load torque and selection of motors 2/372.1 Motor speed-torque curveRefer to Figure 2.1 whereT,, = starting torque or breakaway torque.T, = minimum, pull-in or pull-up torque.Tpo = pull-out, breakdown or maximum torque, obtainableover the entire speed range. In a good design thisshould occur as close to the rated slip as possibleto ensure that the motor runs safely, even duringmomentary overloads, load fluctuations exceedingthe load torque, or abrupt voltage fluctuations,without harmful slip losses (equation (1.9)). Insome specially designed rotors, however, to achievea high starting torque sometimes the pull-out torqueTpo may not be available on the speed-torque curve.It is possible that in such cases the T,, may be thehighest torque developed by the motor in the entirespeed range (Figure 2.2).T, = rated or the full-load torque and should occur asnear to the synchronous speed as possible to reduceslip losses.S = rated slip at which occur the rated torque andcurrent.2.2 NEMA rotor designsFigure 2.2curve- Speed NrT,, too high to have Tpo on the speed-torqueAs a further step towards standardization and to achievemore harmony in motor sizes and designs, for betterinterchangeability in the motors produced by differentmanufacturers, in the same country or by other countries,Speed --CT,, : Starting torqueT, : Pull-in or DUII UD toraueTi : Pull-out 0; breakdown torque (maximum torque)Tr : Rated torqueFigure 2.1 Defining a motor torqueTrNEMA,* in its publication MG-I for Induction Motors,has prescribed four rotor designs, A, B, C, and D, coveringalmost all sizes of LT motors, to possess a prescribedminimum T,,, Tpo and pull-up torques. These torques aregenerally as drawn in Figure 2.3 to meet all normalindustrial, agricultural or domestic needs. (Refer to thesaid publication or IEC 60034-12 for values of thesetorques. IEC 60034-12 has also provided similarstipulations.)However, motor manufacturers may adopt more flexibledesigns with more reserve capacity and better speedtorquecharacteristics to suit the requirements of aparticular sector. These are particularly for installationswhere the distribution system may have wider voltagefluctuations or the load itself may have varying loaddemands. It is possible that the same motor may have todrive more than one type of loads at different times. Anagricultural pump motor may be one such applicationwhere it may also have to drive a thrasher or a winnowerat different times. A motor with higher flexibility wouldbe more desirable for such applications.Manufacturers, depending upon market needs, mayadopt all or a few such designs or even have their owndesigns, still conforming to such stipulations. Specialapplications may, however, call for a custom-built motoras noted later. As a standard practice all HT motors arecustom-built for each application and no rotor dcsignsare prescribed for these.*NEMA - National Electrical Manufacturers Association, USA.

www.taghedanesh.ir2/38 Industrial Power Engineering and Applications Handbookmade of these two cages, having a low effective resistance,being in parallel. In such designs, therefore, the speedtorquecurve can be achieved to take any desired shapeby suitably choosing the resistances of the two cages,the width of the slot opening and the depth of the innercage. The equivalent circuit diagram of a motor with asingle and a double cage rotor is illustrated in Figure 2.4(a) and (b) respectively. To draw the speed-torque curvefor such a motor theoretically, consider the two cagesdeveloping two different torques separately. The effectivetorque will be the summation of these two, as shown inFigure 2.5.Figure 2.3 Speed-torque characteristics of motors as perNEMA standardNotesThe inncr and outer cages are separated by a narrow slit tofacilitate linking of the main flux with the inner bars which arequite deep.HT motors are also manufactured with double cage rotors. Theyare designed especially to match a particular load requirementwhen the load characteristics are known, or as in NEMA classC, or as the manufacturer’s own practice, when the startingtorque requirement exceeds 150% of the full-load torque (FLT).The likely applications for a high starting torque may be induceddraughtfans, blowers, coal crushers, mill motor5 and coalconveyor motors.Generally. depending upon the type of load, differentmanufacturers may adopt to different design practices, such ashigh T,, and low thermal withstand time or moderate T,, andhigh thermal withstand time.2.3 Special designs of rotors2.3.1 Double squirrel cage motorsIf the torque requirement of a load is high, an ordinarysquirrel cage motor, even on a DOL* switching, may notbe suitable to meet the stringent starting requirements.If, however, the resistance of the rotor circuit is increasedthe starting torque can be improved as discussed in Section1.2 (equation (1.3)). But high rotor resistance will meanhigh running slip, causing greater rotor losses and heatin the rotor circuit. The solution to this problem is foundin a double squirrel cage motor. In such motors the rotorhas two cages, one closer to the periphery of the rotorand the other deeper and nearer to the core.The one closer to the periphery has a high resistanceand the one nearer to the core a low one. To accomplisha high rotor resistance, high-resistivity materials such asbrass is generally used. The inner cage has a high leakagereactance due to its depth, while the outer one has a highresistance and a low reactance like an ordinary squirrelcage rotor.During start-up the inner cage has a very high impedanceand thus, the larger portion of the current passes throughthe outer cage only. Because of high resistance and high12R loss in the rotor circuit, it develops a high startingtorque and accomplishes an analogue to a slip-ring motor.When the rotor reaches the rated speed, the reactances ofboth the cages are almost negligible because of low slipand the rotor current is carried into two parallel paths*DOL - Direct On-lineFigure 2.4(a)cage motorFigure 2.4(b)cage motor- RisEquivalent circuit diagram of a single squirrel1st cage2nd cageEquivalent circuit diagram of a double squirrel

www.taghedanesh.irMotor torque, load torque and selection of motors 2/39Leakagefl, I“Leakage(a) Deep bar (b) Taper bar (c) Double cageSpeed -NrFigure 2.6effectDifferent types of rotor slots, making use of skinFigure 2.5cage motorSpeed-torque characteristics of a double squirrelPerformanceIn such motors the pull-out torque is normally less thanthe starting torque. This is because the pull-out torquesby the two cages occur at different speeds. Such motorswould possess a low power factor and efficiency comparedto an ordinary squirrel cage motor, because of the highleakage reactance of inner cage and comparatively higher12R losses. Such motors would have a slightly higher slipthan an ordinary squirrel cage motor due to higher rotorresistance.LimitationsDuring start-up since only the outer cage is in the circuitwith a very high current, the motor is heated up quicklyby every start and may not be suitable for frequent startsand reversals.There are several other designs available to achieve aconsiderably high staring torque and yet overcome theabove limitation. It is possible by employing a deep cage,tapered cage or special types of rotor materials such asbrass and selenium to increase the starting resistance ofthe rotor circuit, and hence the starting torque. Thesemethods are discussed briefly below.2.3.2 Other designs of rotor cageUse of skin effectThe basic concept used in the design and selection ofother types of rotors to provide better startingcharacteristics is the high rotor resistance during startup.Other than the double cage rotors, this can also heachieved by making deep or taper rotor bars as shown inFigure 2.6 (See also Figure 2.7). At different frequencies,the rotor has different effective resistances, due to a changeFigure 2.7 Other designs of a few double cage slotsin inductive reactance (S.ssX2), which in an inductionmotor varies with rotor frequency (i.e. speed). This effectof change of resistance is termed the ‘skin effect’. Formore details, see Section 28.7. To make use of this effect,the slot, irrespective of its configuration, may be madedeep to create higher eddy currents and correspondinglyhigher eddy current losses, to add to the effective resistanceof the rotor during start-up and to diminish this withspeed. (See also Section 2.4.) In this way the depth, indeep bars, and depth and taper, in tapered bars, can bevaried to achieve the desired performance. For the sametorque characteristics either of these types of cages canhe employed which, for one characteri\tic, will requirethe same area of cross-section but the depth will varydepending upon the type. The deep bars will be deeperthan a taper bar. Moreover, the taper slot will have abetter grip for rotor conductors during a run than a deepparallel bar and also better cooling properties.Angle of skew in squirrel cage rotorsThe movement of rotor teeth around the stator producesa clogging effect, resulting into vibrations and noise. Toreduce this effect, the common practice is not to providethe rotor slots parallel to the shaft axis but at an angle.This practice is known as ‘rotor bar skew’. A properskewing can also improve the starting torque and reducethe starting current, in addition to the effects of space

www.taghedanesh.ir2/40 Industrial Power Engineering and Applications Handbookharmonics and slot losses. The angle of twist (skew) is amatter of experience, by results obtained over the years.The most common skew angles, for various combinationsof stator and rotor slots in practice, are given in Table 2.1.Table 2.1 Typical angles of skew for cage rotorsNumber Number Number Skew angleof poles of stator of rotor (degrees)slots slots2 [!4 {6 3614 2616 2028 1618 2028 13 to 1433 11 to 142.4 Effect of starting current ontorqueIgnoring the friction and core losses, the torque developedin synchronous watts,or -3.-I: R,S--Speed --+Slip N,Figure 2.8 Starting (locked rotor) currents corresponding todifferent starting torquesdue to the skin effect, will also diminish the full-loadpower factor. (See the circle diagram, Figure 1.16,corroborating this statement.) The T,, and I,, are, therefore,a matter of compromise to achieve a good Tp, a betterpower factor and a lower slip. Figure 2.9 shows fordifferent starting torques the corresponding pull-outtorques and their occurrence of slip, maintaining the sameSince the stator current is a function of the rotor current,the motor torque is proportional to the square of thestator current. Generalizing,TS,,= [k)2 (for the same rotor resistance R2)(2.3)or = (&)2 . 5 (for different rotor resistances)Ts*2 R;Analysing equation (2.2), the higher the starting torque,the higher will be the starting current for the same motorparameters (Figure 2.8). An attempt to keep the startingcurrent low and yet achieve a higher starting torque maybe feasible, but only up to a certain extent, by suitablyredesigning the rotor with a higher resistance (equation(2.1)). However, the results of such an attempt mayadversely affect the other performance of the motor. Forexample, the Tpo will be reduced due to a higher rotorresistance and may occur at a higher slip, even if thefull-load slip is the same. The increased slot leakage,Figure 2.9Effect of starting torque on T,, and slip

~2.7www.taghedanesh.irMotor torque, load torque and selection of motors 2/412.4.1 NEMA recommendations onstartingcumnts 2.5 Load torque Or Opposing torqueWith a view to achieve yet more standardization in motordesign, NEMA Standard MG- 1 has also recommendedthe maximum locked rotor current of single-speed threephasemotors for the various rotor designs A, B, C, andD, for various recommended torque values. These havebeen derived for a 415 V a.c. system and are shown inTable 2.2.Table 2.2 Recommended maximum locked rotor currents forvarious rotor designsHPI1 .52357.5IO1520253 040so6075100125150200Approx. maximumlocked rotor current18253139567198141I782222653s444 152966 I88411051319Rotor designB.D.B.D.B.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.B.C.D.For smaller loads, say up to 20/30 kW, it may not beessential to pre-check the load curve with that of themotor. But one should ensure that working conditions orthe load demand are not so stringent that they may causea lock-up of rotor during pick-up due to a very low appliedvoltage or accelerating torque, or a prolonged startingtime as a consequence or due to a very large inertia ofrotating masses etc. For critical applications and for largermotorb it is essential to check the speed-torquerequirement of the load with that of the motor. Loadscan generally be classified into four groups. Table 2.3indicates the more common of these and their normaltorque requirements, during start-and variation with speed.The corresponding curves are also drawn in Figures 2.10-2.13. To ascertain the output requirement of a motor, fordifferent applications a few useful formulae are given inAppendix 1 at the end of Part I of this book.2.6 Selection of motorsThe recommended practice would require that at eachpoint on the motor speed-torque curve there should be aminimum 15-20% surplus torque available. over and abovethe load torque, for a safe start (Figure 2.14). The torquethus available is known as the accelerating torque.1764 B.C. ~ Time of start-up and its effectNote For motors beyond 200 h.p., NEMA has not covered these on motor performancedata. It is, however, recommended that larger motors may be designedto have even lower locked rotor current; than the above to reducethe starting transient effects on the distribution system as well ason the motor windings.This depends upon the applied voltage, i.e. type ofswitching, starting torque of the motor, counter-torqueTable 2.3 Types of loads and their characteristicsSeriai Load Characteristics of Sturting torque Opposing torque with Figurt. no.110. load speedPresses, punches, latchesand drilling machinesFans. blowers, centrifugalpumps and compressorsRolling mills, ball mills,hammer mills, calendardrives and sugarcentrifugesConveyors and hoists- Light duty 20-309The power isproportional to thethird power of thespeed (P = N3)The power isproportional to thesquare of the speed(P = Nz)The power isproportional to thespeed (P - N)Medium duty 1040%Heavy duty 3040%.May be more and have toaccelerate large massesof heavy moment ofinertia, requiring aprolonged time of start-upHeavy duty lOO-llO%Torque remains constant 2.10and at a very low value.since the load is appliedwhen the motor has run tospeedTorque rises with square of 2.1 1the speed (T Q N')Near full-load torque 2.12Torque remains constant 2.13throughout lhe speed rangeand at almost the full-loadtorque

www.taghedanesh.ir2/42 Industrial Power Engineering and Applications Handbook20 -% Speed-Figure 2.10Light duty10080t2Pe 40ae100t 8o3P 60eae402cC1. Torque = (speed)'2. Power w (~peed)~% Speed-200/1, Torque constant2. Power speed-% SpeedFigure 2.1 1 Medium dutyFigure 2.13 Heavy dutyof the load and the inertia of the rotating masses etc. Itis expressed byGD; . N ,t, = (2.5)375 . T,wheret, = time of start-up in secondsGD; = total weight moment of inertia of all the rotatingmasses, referred to the motor speed in kg.m2= CD; + GD;( GDL is motor and GD: is load weight momentof inertia referred to the motor speed)whereGD~ = 4.g . M . K~g = 9.81 m/s2M = mass and,q . M = W (weight in kg)K = radius of gyrationTo = average accelerating torque in mkg (Figure 2.14),i.e. average (Tst- TL) in mkgTL = opposing torque (load torque)GD; at motor speedIf the load is driven through belts or gears at a speeddifferent from that of the motor, the effective value of

~ ~~ _.www.taghedanesh.irMotor torque, load torque and selection of motors 2/43Heat generated;H = Is: . R ' t, watt. s. (W.S.)also H=W.6.8Minimumf 15-2070 Of T,Figure 2.14Speed 4 N,Accelerating torque (Ta)GD2 of the load, as referred to the motor speed, will bedifferent. Equating the work done at the two speeds:GO; . N,' = GO,' . Nfor 8=- "CW. 6whereW = weight of heated portion in kg6 = specific heat of the material of windings, in watt . s./kgm ("C)8 = temperature rise in "C (Table 1 1. I )A possible way to restrict the temperature rise is theuse of a material having a high specific heat. An increasein the weight would be futile, as it would require morematerial and prove to be a costly proposition. A motor'sconstructional features should be such as to provide goodheat dissipation through its body.Sharing of heatThe rotor and stator heats, during start-up and run, areinterrelated and vary in the same proportion as theirrespective resistances. (See circle diagram Figure 1.16in Section 1.10.)IfandH, = the heat of rotor in W.S.H, = the heat of stator in W. s.or GD; = GD; ($)?whereGD; = weight moment of inertia of load at a speed N,*.Example 2.1A 100 kW, 750 r.p.m. motor drives a coal mill, having GD: as600 kg.m2 through belts, at a speed of 500 r.p.m. Then itseffective GD: at motor speed will be500GDf = 600 x ( 75(1)= 600 x 0.445= 267 Kg m2Note For simplicity, the synchronous speed of the motor isconsidered, which will make only a marginal difference incalculations.2.7.1 Motor heating during start-upIrrespective of the type of switching adopted or the loaddriven by the motor, each time it is switched it generatesheat. in both the rotor and the stator components. Themagnitude of the start-up heat will depend upon the inertiaof the rotating masses, the type of switching, the torquedeveloped by the motor and the opposing (load) torqueetc., as can be inferred from equation (2.5). The higherthe time of start-up, the higher will be the heat generated.The corresponding temperature rise of the stator or therotor windings can be measured as below:(2.9)While the total heat generated in the stator and therotor may be comparable, there may be a significantdifference in the temperature rise of the respective partsas a result of the bulk of their active parts and area ofheat dissipation. For the same material, the rotor willhave a much higher temperature rise compared to thestator, in view of its weight, which may be several timesless than the stator. During start-up, therefore, the rotorwill become heated quickly and much more than thestator. Repeated start-ups may even be disastrous. Duringa run, however, when the temperature has stabilized, anoverload will render the stator more vulnerable to damagethan the rotor. The rotors, as standard practice, are designedfor much higher temperature rises (200-300°C) and maybe suitable to withstand such marginal overloads.CorollaryDuring start-up thc rotor, duc to its lighter weight comparedto the stator, and during a run, the stator, due to overloadare more vulnerable to damage through excessive heat.Example 2.2A rotor fails during start-up, possibly due to a lower supplyvoltage than desired or a smaller accelerating torque thanrequired or reasons leading to similar conditions. In suchcases the rotor fails first, due to higher rotor currents and aprolonged acceleration time or a locked rotor. At this instant,unless the motor control gear trips, the stator may also faildue to excessive heat. Instances can be cited where even

www.taghedanesh.ir2/44 Industrial Power Engineering and Applications Handbookthe short-circuit end rings of a squirrel cage rotor melted, andthe molten metal, through its centrifugal force, had hit thestator overhangs, damaging them through its insulation,causing an inter-turn fault.2.7.2 Heating during a no-load start-upDuring a no-load start-up, i.c. when the motor shaft isfree, half the energy drawn from the supply appears asheat in the rotor and the stator windings. In slip-ringmotors the bulk of the rotor heat is shared by the externalrcsistance, a feature which makes it a better choice forfrequent starts and stops, and for driving loads that possesslarge inertia. It has been seen that most of the stringentload requirements can also be met with high torque squirrelcage motors, manufactured with a judicious design ofstator and rotor resistances, an efficient means of heatdissipation and a proper choice of active material. Theheat generated during a no-load start-up can be expressedby;300%200%fP)eco100%uSpeed +GD; .N:H,l = w.s(2.10)730This expression, except for the mechanical design, istotally independent of the type of start and the electricaldesign of the motor. Electrically also, this is demonstratedin the subsequent example. The expression, however,does not hold good for an ON-LOAD start. On load, theaccelerating torque diminishes substantially with thetype of load and the method of start, as can be seenfrom Figure 2.14, and so diminishes the denominator ofequation (2.5), raising the time of start.Example 2.3A squirrel cage motor is started through an auto-transformerstarter with a tapping of 40%. Compare the starting heat witha DOL starting when the motor shaft is free.With DOL T, = 100%With an auto-transformer T, = (0.4)' or 16%GD; N:. Starting time with DOL, t, = - X-375 T,GDiand with auto-transformer, t,, = - x -375 O.16TaLe. 6.25 times of DOLSince the heat during start-up -(/,J2 . f:. Heat during start on a DOL =(Ist)' . t,and on an auto-transformer = (0.4/,# . t ,~ori.e.N~0.16(/,,)' x 6.25tS= (kl)' ' tsThus at no load, irrespective of the motor torque and the typeof switching, the starting heat would remain the same.2.7.3 Heating during an on-load start-upAgainst an opposing torque, the accelerating torque ofthe motor, which hitherto had varied in proportion to thetype of switching, will now diminish disproportionatelyFigure 2.15 Variation in T, with Y/A switchingwith a switching other than DOL (Figure 2.15). Thestarting time rises disproportionately and so does thestarting heat. Care should therefore be taken when selectinga motor for a particular type of switching and magnitudeof the opposing torque. This is to avert possible damageto the motor due to prolonged starting time, as aconsequence of an inadequate accelerating torque.Maintaining a minimum accelerating torque at eachpoint, during the pick-up may also not be adequatesometimes, when the starting time exceeds the lockedrotor or thermal withstand time of the motor, as discussedbelow.2.8 Thermal withstand timeThis is also known as safe stall time or the locked rotorwithstand capacity of the motor. This is the time duringwhich the motor can safely withstand electromagneticeffects and consequent heating in a locked condition.These are drawn for the cold and hot conditions of themotor in Figure 2.16. Evidently, the motor must come tospeed within this time, irrespective of type of load ormethod of switching. In a reduced voltage start-up orslip-ring motors the starting current would be low andthese curves would signify that for any reason if therotor becomes locked during start or run, or takes aprolonged time to come up to speed, the protective devicemust operate within the safe stall time. Generally, thesecurves are drawn for the stator to monitor the actualrunning condition and not the condition during start-up.The rotor can withstand much higher temperatures duringa run. With the help of these curves, knowing the startingtime and the starting current of the motor, one can ascertainthe number of starts and stops the motor would be capableof undertaking. These curves also help in the selectionof the protective relays and their setting as discussed inChapter 12.

Iwww.taghedanesh.irI IJ- Locked rotor currentMotor torque, load torque and selection of motors 2/45C = heat capacity of the motor= heat required to raise the temperature of thewindings by 1 "C in Joules6= w.whereW = weight of the stator windings in kg= volume of stator windings x specific gravity ofthe metal of the windings= L,, . Z, . A,, . dLn,t = length of a mean turn of the winding in metresZ, = number of stator turns per phaseA,, = area of the whole windings in cm'd = specific gravity of the winding material ingm/cm36 = specific heat of winding metal in watt . s/kg m "C- c-ISafe stall time 'tS; (seconds)-D-- .A - Maximum withstand time under hot condition (on DOL)B - Maximum withstand time under cold condition (on DOL)C - Maximum withstand time under hot condition during YD - Maximum withstand time under cold condition during Y----cIiNore I In equation (2.11) it is presumed that the heating of thewindings is adiabatic Le. whatever heat is generated during a stalledcondition IS totally consumed in raising the temperature of thestator windings by 8. An adiabatic process means that there is noheat transfer from the system to the surroundings. This is alsoknown as the heat sink process. The presumption is logical. hecausethe duration of heating is too short to be able to dissipate a part ofit 10 other parts of the machine or the surroundings.Figure 2.16 Thermal withstand curves2.8.1 Heating phenomenon in a motor during astalled condition(a) For the statorStalling is a condition in which the rotor becomes lockeddue to excessive load torque or opposing torque. Stallingis thus a replica of a locked rolor condition arid canoccur at any speed below the Tpo region, as illustrated inFigure 2.17. The figure also shows that the stator currentduring stalling will generally correspond to I,, only, dueto the characteristic of the motor speed-current curve.Whenever the rotor becomes locked in a region that almostcorresponds to the I,, region of the motor (Figure 2.17)it will mean a stalling condition.In such a condition, if the heat generated in the windingsraises the temperature of the windings by 8 above thetemperature, the motor was operating just before stalling.Then by a differential form of the heat equation:and p = pJ0( 1 + - h)where, = = temperature coefficient of resistivity- 1 "C234.5p4" = resistivity of copper at 40°CIt is known as the middle temperature during the entire temperaturevariation in the locked rotor condition.whereHstI =heat generated during stalled condition per secondin watts= power loss= 1\21, RI= current at the point of stalling in AmpsRf = resistance of the stator windings per phase in RtSti = time for which the stalling condition exists insecondsI Load torque ~I Speed --FStalling(Locked rotorcondition)Figure 2.17 Stalled or locked rotor condition

www.taghedanesh.ir2/46 Industrial Power Engineering and Applications HandbookTherefore the permissible rise in temperature in a stalledcondition will be as follows:andAm= J,, = current density during start in A/cmZ andwherek = material constant for the metal;(i) for aluminium = 0.016(ii) for copper = 0.0065(iii) for brass = 0.02761o,oo65I:. ts1t = -' 0. -(for copper windings)J;:Note 2 Since no system can be heat adiabatic in practice there isa certain amount of heat dissipation from the impregnated windingsto the stator core and housing. This heat dissipation is consideredas 159 of the total heat generated as in IEC 60079-7.:. actual H, = 85% of what has been calculated aboveand t,,, =OJi x 0.0065 x 0.85eor f\tl = seconds(s)(2.12)Jf x 0.00552ApplicationFor safe stall conditions t,,e should be less than the thermalwithstand time of the motor under locked or short-circuitcondition.(i) 8 is called the permissible rise in temperature inthe stalled condition.(ii) For class B insulation, the maximum limitingtemperature is 185°C and for class F 210°C (shorttimepermissible temperature). The permissible risein temperature in class B is 80°C above an ambientof 40°C.:. 8 = 185 - (40 + 80)= 65°C for hot conditionsand 8= 185 -40(b) For the rotor= 145°C for cold conditionsTo ascertain whether the stator or the rotor would failfirst during a stalled condition, the thermal withstandtime of the rotor should also be determined separatelyfor the rotor bars and the end rings. The lowest valuesfor the stator or the rotor will be the safe stall time forthe entire motor. The limiting temperatures in rotorcomponents may be considered as follows:Limiting temperature for bars 450°CLimiting temperature for rings 100°COperating temperature for bars 150°COperating temperature for rings 70°C8 for bars in cold conditions = 450 - 40 = 410°C8 for bars in hot conditions = 450 - 150 = 300°C8 for rings in cold conditions = 100 - 40 = 60°C8 for rings in hot conditions = 100 - 70 = 30°C2.8.2 Plotting thermal withstand characteristicsof the motorCalculate the thermal withstand times tSte's under coldand hot conditions and also at different Ist, say 200%,300% and 400% etc. of I, as shown below. Afterdetermining the corresponding safe thermal withstandtimes, according to the above formula, draw the graph(Figure 2.16), Ist vs tSt:Stalled current late Thermal withstand time, f,,, in secondsas % of I,2003 004005 00600and for J,, (for A windings) =Note J,, is a design parameter and more details may be obtainedfrom the motor manufacturer.Iste. (Area of windingdtum) .Z,Example 2.4A 250 kW motor has a cold thermal withstand time of 30seconds and a hot thermal withstand time of 25 seconds. Ifthe starting time is 7 seconds, determine the consecutivecold or hot starts that the motor will be able to sustain safely.Number of consecutive cold starts = -30= 4.37i.e. 4 startsandi.e. 3 startsnumber of hot starts = -25= 3.67The period after which this can be repeated will depend uponthe heating curve and the thermal time constant of the motor,i.e. the time the motor will take to reach thermal equilibriumafter repeated starts (See Chapter 3).Example 2.5A centrifugal compressor driven through V-belts at a speedof 4500 r.p.m. having the torque curve as shown in Figure2.18 and a moment of inertia M@ of 2.50 kgm' employs asquirrel cage motor with the following parameters:kW = 350N, = 1485 r.p.m.speed-torque characteristic as in Figure 2.18GD; = 30 kgm'Safe stall timehot - 30 S.Cold - 40 S.

www.taghedanesh.irMotor torque, load torque and selection of motors 2/471751501 125al; 1000.96T,b?10075503525Speed -?N, Figure 2.18 Determining the accelerating(1485 rpm) torqueCalculate the starting time and consecutive cold and hotstarts for which the motor will be suitable with a DOL starting.SolutionTo determine the exact accelerating torque, measure theordinates of torque as shown at different speeds and calculatethe average torque as follows:(a) Average load torque- 35+40+50+60+70+80+85+95+100i -9=- 6159= 68.33%(b) Average motor torqueT=150+120+110+110+130+170+180+120+1009=- 119092 132.2%(c) :. Average accelerating torqueT, = 132.2 - 68.33= 63.87%(d) Motor rated torque- 350x974I-1485* 230 mkg:. T, = 230 x 0.6387= 146.9 mkg(e) Total GD: at motor speed = 30 + (4 x 9.81 x 2.5) ___( ::::)zwhere [GD: = 4 . g. MK2] (at the compressor speed)1.e. GD: = 30 + 901= 931 Kgm'931 x 1485(f) Stating time t, =375 x 146.9= 25.1 secondsTake roughly 10% more to account for any tolerance andvariations,:. ts = 25.1 x 1.1= 27.6 secondsThis motor is therefore suitable for only one cold or one hotstart at a time until the temperature rise stabilizes again.If this motor is started with an auto-transformer with atapping of 80%, the motor average torque will be= 132.2 x 0.64 (curve 2, Figure 2.18)or T= 84.6%and acceleration torque T, = 84.6 - 68.33i.e.931 x 1485and t, =375 x 37.42= 16.27%230 x 0.1627 = 37.42 mkg.= 98.52 secondswhich is much more than the safe stall-withstand time.

www.taghedanesh.ir2/48 Industrial Power Engineering and Applications HandbookInferenceOn an ON-LOAD start, the starting time increasesdisproportionately, depending upon the type of switching. Thisload therefore cannot be accelerated within a safe stall timethrough an auto-transformer, even with a tapping as high as80% although the motor possesses some accelerating torqueat each Point during Pickup (cuwe of Figure 2.18).Relevant StandardsIEC Title IS BS60034- I2/ 1980 Rotating electrical machines 8789/1996 BS EN 60034-starting performance of single-speed three-phase cage12199660079-7/1990induction motors for voltages up to and including 690 VElectrical apparatus for explosive gas atmosphere. 638111991 BS 5501-611977Increased safety protection eNEMA/MG-1/1993NEMA/MG-2/1989Related US Standards ANSVNEMA and IEEEMotor and generators ratings, construction, testing and performanceSafety Standards (enclosures) for construction and guide forselection, installation and use of rotating machinesNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevant23organizations for the latest version of a standard.Some of the BS or IS standards mentioned against IEC may not be identical.The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedEffect of starting current on torqueI: xR,T, -S&- = (&)'(for the same rotor resistance R,)Tct2(2.3)T,, 1 = (F)' (for different rotor resistances)or -Tstzst2Time of start-upx 2GD? . N,ts = 375 . T,t, = time of start in secondsGD; = GD; + GD;GDi = motor andGD; = load weight moment ofinertia referred to the motor speed(2.4)(2.5)GD: = weight moment of inertia of load at a speed NLMotor heating during start-upH = I$ .R.t, (W.S.) (2.7)also, H = W . 6 . 8W = weight of heated portion in kgm6 = specific heat of the material of windings, inwattdkgm. "C.8 =temperature rise in "CSharing of heatH, = rotor heat in W.S.H, = stator heat in W.S.Heating during a no-load startHeating in a motor during a stalled condition:for the stator(2.8)(2.10)HSte . tstc = 8 . C (2.1 1)Hste = heat generated during stalled condition per secondin wattstste = time for which the stalling condition prevails insecondsC = heat capacity of the motoror tstc =secondsJA X 0.00552J,, = current density during start in A/cm2Further reading(2.12)Machinery Hand Book, Industrial Press Inc., 200 Madison Avenue,New York, USA.

www.taghedanesh.irDuties of InductionMotors3.3 Periodic duties 3/5 11Y (S3)y withy with3.3.5 Coiitinuous duty wtth intermittent periodic loadingt and brake (S,)3/53odic Ypeed changes3.3.9 Duty with discrete constant loads (SI") 3/55Rating of short-time motors 3/S9f formulae used 3/67ored by the flywheel 316

www.taghedanesh.ir

www.taghedanesh.irDuties of induction motors 3/513.1 Duty cyclesUnless otherwise specified, the rating of the motor willbe regarded as its continuous maximum rating (CMR),defined by duty S, as noted below. But a machine is notalways required to operate at a constant load. Sometimesit must operate at varying loads, with a sequeiice ofidentical operations, involving starts, stops braking, speedcontrol and reversals, with intermittent idle running andde-energized periods etc. (e.g. a hoist, a crane, a lift orother applications.) Using a CMR motor for suchapplications, with a rating corresponding to the maximumshort-time loading will mean an idle capacity during noloadrunning or de-energized periods and a constant drainon energy. in addition to a higher cost of installation. Toeconomize on the size of machine for such applications,IEC 60034-1 has defined a few duty cycles, as notedbriefly below. These may be considered while selectingan economical size of machine and yet meet the variableload demands safely. Such motors may be runningoverloaded during actual loading but for shorter durationsnot sufficient to exceed the permissible temperature riselimits. They dissipate excessive heat during idle runningor de-energized periods to reach a thermal equilibriumat the end of the load cycle. These duties are described inthe following sections.3.2 Continuous duty (CMR) (SI)The operation of a motor at a rated load may be for anunlimited period to reach thermal equilibrium (Figure3.1) and possible applications are pumps, blowers, fansand compressors.3.3 Periodic duties3.3.1 Short-time duty (S,)In this case the operation of the motor is at a constant1No loaalosscs.load during a given time just adequate to attain themaximum permissible temperature rise, followed by arest and de-energized periods of long durations to reestablishequality of motor temperature with the coolingmedium (Figure 3.2). The motor should restart for thenext cycle only when it has attained its ambient condition.The recommended values for short-time duty are 10,30,60 and 90 minutes. The type designation for a particularrating, say for 30 minutes, will be specified as S1_ - 30minutes. Likely applications are operation of lock gates,sirens, windlasses (hoisting) and capstans.3.3.2 Intermittent periodic duty (S,)This is a sequence of identical duty cycles. each consistingof a period of operation at constant load and a rest andde-energized periods. The period of energization mayattain the maximum permissible temperature rise (On,).The period of rest and de-energization is sufficient toattain thermal equilibrium during each duty cycle (Figure3.3). In this duty the starting current I,, does notsignificantly affect the temperature rise. Unless otherwisespecified, the duration of each duty cycle should be IOminutes. The recommended values for the cyclic durationfactor CDF are IS%, 2596, 40% and 60%. The typedesignation for a particular rating, say for 40%. will bespecified as S3 - 40%.Cyclic duration factor CDF =N~N+RwhereN = operation under rated conditionsR = at rest and de-energized and0, = temperature rise reached during one duty cycle(= 0)Likely applications are valve actuators and wire drawingmachines.3.3.3 Intermittent periodic duty with start (S,)This is a sequence of identical duty cycles. each consistingLoadNoloadf24E,Amb tempI L V A m (H= 0) temp bTime -*CMR: Continuous maximum ratingFigure 3.1 Continuous duty, S,a0+-Long rest+T,me '0Amb temp (0 = 0)Note Short-time loading IS higher than the CMR and it is true for allduties S2 - S,,Figure 3.2 Short-time duty S,

Iwww.taghedanesh.ir3/52 Industrial Power Engineering and Applications HandbookLoadOne cycleP N R -INo load 4trl 8One cycleFD, N ,t-084I-ITime0Amb. temp. (e = 0)Figure 3.3 Intermittent periodic duty, S,of a period of start, a period of operation at constant loadand a rest and de-energized periods. The starting,operating, rest and de-energized periods are just adequateto attain thermal equilibrium during one duty cycle (Figure3.4). In this duty the motor is stopped, either by naturaldeceleration, after it has been disconnected from the supplysource, or by mechanical brakes, which do not causeadditional heating to the winding:D+NCDF= D+N+RwhereD = period of startingN = operation under rated conditionsR = at rest and de-energized,6, = temperature rise reached during one duty cycle(= 0)For this duty cycle, the abbreviation is followed by theindication of the cyclic duration factor, the number ofduty cycles per hour (CAI) and the factor of inertia (FZ).(See Section 3.4 for FI.) Thus, for a 40% CDF with YOoperating cycles per hour and factor of inertia of 2.5, thecycle will be represented byS4 - 40% - YO c/h and FI - 2.5Likely applications are hoists, cranes and lifts.-1I 'Time10Amb. temp. (e= 0)Figure 3.4 Intermittent periodic duty with starting, S,whereD = period of startingN = operation under rated conditionsF = electric brakingR = at rest and de-energized, and6, = temperature rise attained during one duty cycle(= 0)For this duty cycle also, the abbreviation is to be followedby the indication of the CDF, the number of duty cyclesper hour (c/h) and the FI, e.g.S5 - 40% - YO c/h and FI - 2.5Likely applications are hoists, cranes and rolling mills.One cyclekMLoadStarting 5losses4No loadlosses I ~ 1 I 1 I 1t3.3.4 Intermittent periodic duty with start andbrake (S,)This is a sequence of identical duty cycles, each consistingof a period of start, a period of operation at constantload, a period of braking and a rest and de-energizedperiods. The starting, operating, braking, rest and deenergizedperiods are just adequate to attain thermalequilibrium during one duty cycle (Figure 3.5). In thisduty braking is rapid and is carried out electrically:CDF =D+N+FD+N+F+R-TimeFigure 3.5 Intermittent periodic duty with starting andbraking, S,temp. (e = 0)0

www.taghedanesh.irDuties of induction motors 3/533.3.5 Continuous duty with intermittent periodicloading (S,)This is a sequence of identical duty cycles, each consistingof a period of operation at constant load and a period ofoperation at no-load. The repeat load and no-load periodsare just adequate to attain thermal equilibrium duringone duty cycle. There is no rest and de-energizing period,(Figure 3.6). Unless otherwise specified, the duration ofthe duty cycle will be IO minutes.One cycler V-iP-&-+-Timeet - (Corresponding to no-load heating)Figure 3.6 Continuous duty with intermittent periodic loading, S,0The recommended values of CDF are I5%, 25%. 40%and 60%:CDFN= -N+VwhereN = operation under rated conditionsV = operation on no-load and0, = temperature rise attained during one duty cycle= corresponding to the no-load heatingThe designation in this case will be expressed asS6 - 40% CDFLikely applications are conveyor belts and machine tools.3.3.6 Continuous duty with start and brake (S,)This is a sequence of identical duty cycles, each consistingof a period of start, a period of operation at constant loadand a period of electric braking. The start, operating andbraking periods are just adequate to attain thermalequilibrium during one duty cycle. There is no rest andde-energizing periods (Figure 3.7):CDFD+N+F=D+N+F='whereD = period of startingN = operation under rated conditionsF = electric braking andOne cycleLoad -Startinglosses -No loadlossest-u-1 0I l l-TimeFigure 3.7 Continuous duty with starts and brakes, S,fs c$3?ccCorrespondingto no-load/ heating-6Amb. temp.- (0= 0)

www.taghedanesh.ir3/54 Industrial Power Engineering and Applications Handbook6, = temperature rise attained during one duty cycle= corresponding to the no-load heating.For this duty cycle also, the abbreviation is followed bythe indication of number of cycles per hour and the FI.For example, for 300 c/h and FI 2.5S, - 300 c h FI - 2.5Likely applications are machine tools.3.3.7 Continuous duty with periodic speedchanges (Ss)This is a sequence of identical duty cycles, each consistingof a period of operation at constant load, correspondingto a determined speed of rotation, followed immediatelyby a period of operation at another load, correspondingto another speed of rotation, say, by change of number ofpoles. The operating periods are just adequate to attainthermal equilibrium during one duty cycle. There is norest and de-energizing period (Figure 3.8):D+NICDF =D+NI +F, +N2 +F2 +N3at speed Nr, for load PIat speed Nr2 for load P2at speed Nr3 for load P3whereF,, F2 = changeover of speed by acceleration.D = electrical braking, Nr3 to Nr,N,, N2, N3 = operation under rated conditions and0, = temperature reached during one duty cycle= corresponding to the heating under ratedconditions (PI as in Figure 3.8)One cycle4I 1Speed change-over 'I(braking oracceleration) lossesNo loadIItPw2I- Corresponding- to load P,

~~ ~~~~~ -~~ Onewww.taghedanesh.irDuties of induction motors 3/55We have considered three different speeds (lower Nr,to higher Nr3) for this duty cycle, having three CDFs forone cycle, each corresponding to a different speed.For this duty type, the abbreviation is followed by theindication of the number of duty cycles per hour, the FIand the load at the various speeds.As an illustration the CDF must be indicated for eachspeed as follows:c /I1 FI kW Speed CDF(rp m ) %20 25 10 1440 6020 25 6 960 4020 25 1 730 40Likely applications are where the motor is required torun at different speeds.3.3.8 Non-periodic duty (S,)This is a type of duty in which load and speed both varynon-periodically, unlike the periodic duty cycles notedabove. The motor now supplies variable load demands atvarying speeds and varying overloads, but within thepermissible temperature rise limits. It is a duty similar toduty cycle S8, axcept that sometimes the overloads mayexceed the full load but are within the thermal withstandlimit of the motor (Figure 3.9):. ~~D = period of startingN1, N2, Ni = operations within rated load (PI)N4 = operation during overload (P2)F = changeover of speed by electrical brakingR = at rest and de-energized,6, = temperature rise reached during one dutycycle (= 0)3.3.9 Duty with discrete constant loads (Sl0)This is a type of duty consisting of a number of varyingloads, not more than four in each cycle. Each load isperformed for sufficient duration to allow the machineto attain its thermal equilibrium (Figure 3.10). It is,however, permitted that each load cycle may not beidentical, provided that each discrete loading during oneparticular load cycle is performed for a sufficient durationto attain thermal equilibrium. The temperature attainedduring each discrete loading is within permissible limitsor within such limits that if it exceeds the permissiblelimit, the thermal life expectancy of the machine is notaffected. For example, performing one discrete loadingP2, as in Figure 3.10, the temperature reached (e2) mayexceed the permissible limit (en,) for a short duration(t2), but the final temperature at the end of the cycle isstill such that the next duty cycle can be performed. Theshort duration excess temperature, €$, reached whileperforming the load duty P2, will not however, bedetrimental to the thermal life expectancy of the machine.Referring to Figure 3.10,I i I I !L L J 1 1duty cycle ~Tme-Figure 3.9 Duty with non-periodic load and speed variations, S,

www.taghedanesh.ir3/56 Industrial Power Engineering and Applications HandbookIExample 3.1If GD; = 0.16 kg m'and GD: = 0.8 kg m' at motor speed3.5 Heating and coolingcharacteristic curves"tTemperature riseattained duringone duty cycleFigure 3.10Time -(B= 0)Duty with discrete constant loads, Sl0t,, t,, tg and t4 = duration of operation during discreteconstant loads PI, P,, P3 and P4respectivelyP = equivalent rated load as for continuousduty - SiF = electrical losses6, = maximum permissible temperatureattained for load P01, e,, e,, 6, = temperature reached during differentdiscrete loads6, = temperature rise reached during one dutycycle.3.4 Factor of inertia (FI)This is the ratio of the total moment of inertia referred tothe motor shaft to the moment of inertia of the motor. Ifthe motor moment of inertia is GD; and the load momentof inertia at motor speed, GD; , thenFI =GD; + GD;GD;(GO2 values are weight moments of inertia)(3.1)The heating and cooling behaviour of an induction motor,up to around twice the rated current, may be consideredas exponential, as a part of the heat generated is offset bythe heat sink (heat dissipation) through the windings.But beyond 21, it should be considered adiabatic (linear),as the heat generated is now quick and the windinginsulation may not be able to dissipate this heat equallyquickly, when it occurs for a short duration. Since amotor would normally operate at around I , except duringabnormal operating conditions, the exponential heatingand cooling characteristics are more relevant during anormal run. They determine the performance of a motor,particularly when it is required to perform intermittentduties, and help determine safe loading, starts and brakingsetc. (See curves (a) and (b) of Figure 3.11). They alsoassist in providing a thermal replica protection to largemotors. With the help of these curves a motor protectionrelay (Section 12.5) can be set to closely monitor thethermal conditions prevailing within the machine, andprovide an alarm or trip when the operating temperatureexceeds the safe boundaries. These curves are known asthermal withstand curves and are provided with the motorsas a standard praztice by motor manufacturers. But whenthese curves are not available at a site and a thermal,IDMT or a motor protection relay (Chapter 12) is requiredto be set during commissioning, then the proceduredescribed in Section 3.6 can be adopted to establish them.To determine them it is, however, essential to know theheating and cooling time constants of the motor, whichare provided by the motor manufacturer.3.5.1 Time constantsThese are the times in which the temperature rises orfalls by 0.632 times its maximum value e,,, and areprovided by the machine manufacturer. They are alsoshown in Figure 3.1 1.Significance of thermal time constantsThe short time rating of a CMR motor varies with itsthermal time constant and may differ from onemanufacturer to another depending upon the coolingdesign adapted and its effectiveness. The shorter thethermal constant, the lower will be the short time ratinga CMR motor can perform. It is not, however, practicalto achieve the thermal time constant infinitely high, whichis a compromise with the economics of the motor's designsuch as size, wall thickness of the housing, number anddepth of cooling fins and efficiency of the cooling fan.

6,www.taghedanesh.ir1-T,meDuties of induction motors 3/57~-- Heating Cooling --@ - Heating curve@ - Cooling curve@ - Curve with short time rating, but output not exceeding C.M.R@ - Curve with short time rating, but output more than C.M.R.Figure 3.1 1 Heating and cooling curves-t&cN mx&(D mx13.5.2 Heating curvesExponential heating on a cold startThe temperature rise corresponding to the rated currentof the machine can be expressed exponentially by6, = 8, (1 - e-"Z ) (3.2)where6, = temperature rise of the machine on a cold start abovethe ambient temperature, after t hours ("C)If 0, is the end temperature of the machine in "C aftertime t and Ox the ambient temperature in "C then6, = 6, ~6, = steady-state temperature rise or the maximumpermissible temperature rise of the machine undercontinuous operation at full load in "C, e.g. for aclass B motor, operating continuously in asurrounding medium with an ambient temperatureof 45°C6, = 120 - 45 = 75°C (Table 9.1)For intermittent temperature rises between 0, and e,, asNoteapplicable to curves (c) and (d) of Figure 3.1 1, e,, may be substitutedby the actual temperature on the heating or cooling curves.t = time of heating or tripping time of the relay (hours)t = heating or thermal time constant (hours). The largerthe machine, the higher this will be and it will varyfrom one design to another. It may fall to a low of0.7-0.8 hour.The temperature rise is a function of the operating currentand varies in a square proportion of the current. Theabove equation can therefore be more appropriately writtenin terms of the operating current asK.Z,? = I:(1 -e-'T) (3.3)whereI, = rated current of the motor in A andK = a factor that would depend upon the type of relayand is provided by the relay manufacturer. Likelyvalues may be in the range of 1 to 1.2I, = actual current the motor may be drawingTest check(i) For rated current at t = 0Bc = 6, - 8, = Ir2 (1 - e-" )or(ii) At= 0, (e-" = 1)0, = e, which is truet = m0, = 8, (1 - C)= e,, (e? = 0)which is also true.The relative temperature rise in a period t after theoperating current has changed from Io to I,Oc(re1ative) = (I: - I: )(I -If I, is higher than Io, it will be positive and will suggesta temperature rise. If I, is lower than I,, then it will benegative and will suggest a temperature reduction.

www.taghedanesh.ir3/58 Industrial Power Engineering and Applications HandbookExponential heating on a hot startThis can be expressed by61, = e, + (e, - &)(i - e-"')and in terms of operating currente,, = 102 + (I: -I: )(I - e t/i)(3.4)where61, = temperature rise of the machine on a hot start abovethe ambient temperature, after t hours in "C= 6, - 6,When this quantity is required to monitor the health of amachine, say, for its protection, it can be substituted byk. I;, where I, is the equivalent maximum current atwhich the motor can operate continuously. It may alsobe considered as the current setting of the relay up towhich the relay must remain inoperative.0, = initial temperature rise of the machine above theambient in O C= e, - e,where 0, is the initial temperature of the hot machine in"C before a restartIo = initial current at which the machine is operating8, = end temperature rise of the machine above theambient andI, = actual current the motor may be drawingHence equation (3.4) can be rewritten asK . I =I; ~ +(I? -I;)(] -e-"')For the purpose of protection, t can now be consideredas the time the machine can be allowed to operate at ahigher current, 11, before a trip:. t = tripping time.Simplifying the above,1; - 1;and t= zlog, If - kl,?(3.5)With the help of this equation the thermal curve of amachine can be drawn on a log-log graph for a known Z,t versus Il/Ir for different conditions of motor heatingprior to a trip (Figure 3.12). The relay can be set for themost appropriate thermal curve, after assessing the motor'sactual operating conditions and hence achieving a truethermal replica protection.Equations (3.2) to (3.5) are applicable only when theheating or cooling process is exponential, which is trueup to almost twice the rated current as noted above.Beyond this the heating can be considered as adiabatic-0 1 2 3 4 5 6 7IOverloading conditions (L)1,Figure 3.12 Thermal withstand curves(linear). At higher operating currents the ratio tlzdiminishes, obviously so, since the withstand time of themotor reduces sharply as the operating current rises. Atcurrents higher than 24, the above formulae can bemodified as below.Adiabatic heating on a cold start1e, = e, - e, = 1: .-zAdiabatic heating on a hot starte,, = e, - e, = e, +- (e, - e,)t/~or= I; + (If - I,2)t/z3.5.3 Cooling curves(3.6)(3.7)The residual temperature fall in terms of time, after themotor current is reduced to zero, can be expressedexponentially bye= e, . e-uf(3.8)where .--. -z' = cooling time constant in hours. It is higher than theheating time constant Z. When the machine stops,its cooling system also ceases to function, exceptfor natural cooling by radiation and convection. Themachine therefore takes a longer time to cool thanit does to heat.

www.taghedanesh.ir3.6 Drawing the thermal curvesThese can be drawn for temperature versus time or currentversus time as desired, depending upon the calibrationof the device, such as a motor protection relay. Belowwe provide a brief procedure to draw these curves.3.6.1 From cold conditions(a) For I, I200% I,(b) For I , > 200%(3.9)(3.10)where e,, and z are design parameter, and are providedby the machine manufacturer. The curves can now beplotted in the following ways.For thermal settingsOcj6,,l versus tlr different ratios of (lllI,.) as shown inTablc 3.l(a) and Figure 3.13(a), for l,/Zr I 200%, andTable 3. I (b) and Figure 3. I3(b) for 1,/11- > 200%. SinceO,,, and r arc known, f can be calculated for different0:s corresponding to different motor currents. The relaycan then be set to provide a thermal replica protection.For overload settingsI,/lr versus tlzdif'ferent ratios of 0,/6,,, as shown in Table3.2(a), and Figure 3.14(a) for i,/lr 5 200% and Table3.2(b) and Figure 3.14(b) for I,/I, > 200%. Since I, andr are known, t can be calculated for different overloadconditions. corresponding to different temperature rises.The relay can then be set for optimum utilization of themachine.Example 3.2For an overload of 25%, a class B motor, operating at anambient temperature of 45°C. the relay corresponding to 10%over temperature rise can be set to trip as follows:i.e. - HC = 1.1 = (1.25)'(1 -e-' r,0,or..= 1 - 0.704= 0.296-3.378t = T log, 3.378= 1.5 x 1.217 or 1.82 hoursDuties of induction motors 3/59Considering r = 1.5 hoursThe relay may therefore be set to trip in less than 1.82 hours.3.6.2 From hot conditionsSimilar curves can also be plotted for hot conditionsusing equation (3.5) and assuming 0,) = 0, for ease ofplotting and to be on the safe side. The relay may then beset accordingly. For brevity these curves have not beenplotted here.One may appreciate that by employing such a motorprotection relay it is possible to achieve a near thermalimage protection for all ratings, types and yakes of motorsthrough the same relay by setting its I- - t and 0 - tcharacteristics as close to the motor's characteristics aspossible. The O/C condition is normally detected throughthe motor's actual heating, rather than current. for optimumutilization. Moreover, the starting heats or the heat of theprevious running if it existed, when the motor wasreswitched after a rest or a shutdown. are also accountedfor by measuring the thermal content.Example 3.3The motor is operating hot, say at an end temperature of130°C. If the motor is of insulation class B and ambienttemperature at 50°C then,Hc = 130 - 50 = 80°Cand 0, = 120 - 50 = 70°C.. ---=- 114%0, 70Referring to the curves of Figure 3.13(a) the relaycorresponding to an overload of 150°/0 will trip in about0 0tit = 0.53, for curves of 100% or 0.70 for 2 curve ofem 6"120%.If 7 is taken as 1.5 hours, the motor may be set to trip inabout 0.60 x 1.5 hours, i.e. about 54 minutes, consideringthe average value of tls as 0.60, for a 0,/6, as 114% byinterpolation. It will be noted that a setting corresponding toa thermal curve of 100% will underutilize while correspondingto 120% will overutilize the motor, while the optimum trueutilization will correspond to Qc/O,,, as 114% only. It is thereforeadvisable to draw closer curves or use extrapolation whenevernecessary to obtain a closer setting and plot a more truereplica of the motor thermal characteristics.3.7 Rating of short-time motorsIf a short-time duty is performed on a ContinuousMaximum Rating (CMR) motor with some no-load oridle running, the temperature rise 0 may not reach itsmaximum value, e,,,, as shown in curve (c) of Figure3. I I. A CMR motor therefore can be operated at higheroutputs on short-time duties as shown in curve (d). Theextent to which a CMR motor can be over-rated to performa particular short-time or intermittent duty is consideredin the following example. While evaluating the ratingfor such duties, the heat during star-up and during brakingand their frequency of occurrence should be considered.

~~www.taghedanesh.ir3/60 Industrial Power Engineering and Applications HandbookTable 3.1 In terms of thermal settings (z versus L)I(a) For l- 5 200%I,tlr1511-- e1 -- 1= 1 - 1 = 0.632e' 2.7182 0.8657.38934I 1--=1-1-e5 148.4- 0.9933.158 0.355 0.6323.216 r0.486 0.8653.237 1 0.534 1 0.953.2455 I0.552I 0.9823.248 I 0.558 1 0.9930.9871.351.481.531.55Refer to the curves inFigure 3.13(a)1.422 I 2.528I1.946 I 3.46 1L2.1372.209 3.9282.234These high values of e,/@,, are not permissible.Plot curves for tlz < 1 also for settings athigher currents.I(b) For 2 > 200%I,0.2 1.2521 High values of 6,/8,,, indicate that the motor can sustain such high currents for shortdurations only, i.e. at low tl7.2 These are danger areas and the machine must be prevented from operating in these areasas far as possible. If absolutely essential, the maximum permissible temperature may beexceeded for only a short period to protect the insulation from a rapid deterioration ordamage (Section 9.2).

~ 150'7,www.taghedanesh.irDuties of induction motors 3/610.20.180.160.14t O.I2-8 0.1(I)E0.080 1 ' 2 3 4 6 7YOnly these portionsof curves are relevant1.(a) For s 200%LL2.0@m- L 1 . 5Om- BC,, .2@m0.060.04~Bc=l .o 0.02om%=0.75ern00 1 2 3 4 5 6 7I(b) For > 200%4' IOnly these portionsof curves are relevantNote For ease of illustration two graphs are drawn ($5 200% and > 200% , For actual use, the relevant portions of the graphs (as marked)1,alone must be drawn on one common graph. More points can be plotted in the required region for a closer setting of the relay.Figure 3.13Thermal curves to set the relay for over-current protection corresponding to different operating temperaturesExample 3.4(a) If a CMR 25 h.p. motor, with a thermal heating constantof 1.5 hours reaches a maximum temperature of 115°C incontinuous operation with an ambient temperature of 40°C,then the half-hour rating P of this motor can be determinedas below.Compare the temperature rises which are proportional tothe losses at the two outputs and the losses are proportional tothe square of the load. Ignoring the mechanical losses thenOm for load 'P' = P'and @(, ,ph,,(a)for 25 h.p. = (25)' when run just for half an hourSince a 25 h.p. motor now operates only for half an hour(b)where Om = 115 - 40= 75°CFrom (a) and (b)om = 0. (3or om = e,[ior 1 = (1- 0.716).or25P=-210.284= 47 h.p.- e-% ) . (51'

www.taghedanesh.ir3/62 industrial Power Engineering and Applications HandbookTable 3.2In terms of current settingsI(a) For -L 5 200'37, using the same equationI,t /T11--et I?1 - - = 0.0099O.O1 I I.bl10.05 I 1 -- = 0.04761.0510.10 I 1 -- = 0.0951.1051 - - = 0.1800.20 I 1.;20.50 I11 -- = 0.3941.651.0 1 1 -2,718 1 -- 0.6323.178 4.495XqG-1.026 I 1.4510.39810.5621.667 I 2.04 I 2.36 1 2.58 1 2.89 1 3.331 , L1.126 1.38 1.59 1.74 1.950.889 I 1.089 1 1.26 1 1.38 I 1.54 1 1.78 I*Notes (1) For a closer overload protection, more curves should be drawn for f/r > 1(2) *These points are not relevant for ll/lr < 200%.I(b) 1 > 200'37, using equation (b)I,11 lB,].e. - =I, 4K.t0.100.201 .0 1.41 2.24 2.74 1.16 3.46 3.87 4.470.7 1 I .0 1 .58 1.94 2.24 2.45 2.74 3.16**These conditions may not occur even on a fault in the motor.Nore For obtaining a true replica of the motor thermal characteriatics, I* - I and 0 - t mure curves may be plotted for rlz < 0.02

www.taghedanesh.irDuties of induction motors 3/63200150t 100mE .m"50-Tune (UT)II(a) For C 200%(b) For 1 > 200%1, I,INote For actual use combine curves - S 200% and I > 200% on one graphI, I,00 002004006008 01 012 014016018 02Tune (Vr)-Figure 3.14 Thermal curves to set the relay for over-temperature protection corresponding to different overload conditions(b) Similarly, if the rating is 1 hour, then- 25__-.G0.487= 35.8 h.p3.8 Equivalent output of short-timedutiesFor varying loads (Figure 3.15) or for short-time duties(Figure 3.16) it may not be necessary to select a motorcorresponding to the maximum load during one cycle.Consider a motor that is always energized under thefluctuating loads of Figure 3.15. Then the equivalentrequirement can be determined as below, ensuring thatthe output achieved and the motor chosen will be sufficientto develop a torque, during all conditions of voltages,adequate to drive even the highest load and meet itstorque requirement. Consider heating to be proportionalto the square of the loading, ignoring the mechanicallosses. ThenP,Z . t, + Pz' ' 12 +Peq (rms.) =. tj(3.1 1)tl f I2 + t3Instead, if the load values represent the torque requirement,then7 1Fmb,Y----Time* 6- Corresponding to load P30- Thermal curve of a motor with a rating of Peqt5-1Loadlossestemp.(e= 0)0@- Heating curves for varying loads, the average heating notexceeding the permissible temperature rise 0,Figure 3.15 Equivalent output of short-time duties (varyingloads)

www.taghedanesh.ir3/64 Industrial Power Engineering and Applications Handbook54Loadlosses-TimeI0t, and tr, periods of rest.@- Thermal curve of a motor with a rating of Peq.@- Thermal curves for intermittent loads, the average heating not exceeding the permissible temperature rise 6,Figure 3.16Equivalent output of short-time dutiesTeq (r.m.s.) =and motor output. tl +-T: . t, -+ T: .t3tl + t2 + t3(3.12)If the cycle has a short-time rating, with a period ofenergization and one of rest, the motor will obviouslycool during the de-energizing period, and depending uponthe peak load and the rest periods, a comparatively loweroutput motor can perform duties at higher loads. Theequivalent output for the load cycle of Figure 3.16 isP? . t, + P; . t2 +pi . t3Pq =tSince total time tis more in this instance, the equivalentpower required will be less.Example 3.5Determine the motor rating, for a 10-minute cycle operatingas shown in Figure 3.17. There is no rest period, but in onecycle, the motor runs idle twice at no load for 2 minutes each.The cycle starts with a load requirement of 10.5 kW for 4minutes followed by an idle running, a load of 7.5 kW for 2minutes, again with an idle running and then the cycle repeats.SolutionAssuming the no-load losses to be roughly 5% of the motorrating of, say, 10 kW, thenpeq =(10.5)' x 4 + (0.5)' x 2 + (7.5)' x 2 + (0.5)' x 210_ 441 + 0.50 + 112.50 + 0.5010=.J5545= 7.45 kWThe nearest standard rating to this is 7.5 kW, and a motor ofthis rating will suit the duty cycle. To ensure that it can alsomeet the torque requirement of 10.5 kW, it should have aminimum pull-out torque of 10.5ff.5 or 140% with the slip atthis point as low as possible so that when operating at 140%

www.taghedanesh.irDuties of induction motors 3/65(One cycle)on its speed-torque curve, the motor will not drop its speedsubstantially and cause high slip losses.Note For such duties, the starting heat is kept as low aspossible by suitable rotor design to eliminate the effect offrequent starts and stops. The margin for starting heat andbraking heat should be taken into account if these areconsiderable. The manufacturer is a better guide forsuggestions here.3.9 Shock loading and use of aflywheelThe application of a sudden load on the motor for shortduration, in the process of performing a certain loadduty, is termed 'shock loading'. This must be taken intoaccount when selecting the size of a motor. Electricmp. temp hammers, piston pumps, rolling mills, cane crushers and(e = O) cane levellers, sheet punching, notching, bending and0 cutting operations on a power press, a brake press or aTime +shearing machine are a few examples of shock loading.*e,- Corresponding to no-load lossesThey all exert a sudden load. although for a very short0- Actual heating curveduration, during each load cycle, and may damage the@- Thermal curve for 7.5 kW motormotor as well as the machine. Such machines. therefore,Figure 3.17the motor rat,ng (CMR) for a short- experience a sharp rise and fall in load. Figure 3.18 depictstime dutywch a load cycle, having excessive load P2 for d shortp3TtiNo loadlosses= Correspondingto no loadlossesP, - Light loadingP2 - Severe to very severe loading-TimeFigure 3.18 A typical shock loading dutyr

www.taghedanesh.ir3/66 Industrial Power Engineering and Applications Handbookduration t2 a very light load P1 for a duration t, and at noload for rest of one cycle.For such load requirements, one may either choose acomparatively larger motor to sustain the load and torquerequirements during shock loading or a smaller motor,depending upon the average equivalent loading Peq asdiscussed earlier. When choosing a smaller motor it wouldbe advisable to absorb and smooth the shocks first tocontain the additional shock burden on the motor, aswell as on the main machine. This is made possible byadding more moments of inertia to the drive by introducinga flywheel in the system, as shown in Figure 3.19. Theflywheel will now share a substantial jerk of the peakload, because it possesses a high inertia, on the one hand,and is already in motion, on the other, before the loadjerk is applied. The motor now has to share only a moderatejerk and a smaller motor can safely perform the requiredshock duty. During peak load, the stored kinetic energyof the flywheel is utilized to perform the load requirement.This energy is regained when the motor picks up afterperforming the task. Motors for such applications canbe built with larger air gaps which may mean a low powerfactor and a higher slip, but a higher capacity to sustainshocks.3.9.1 Size of flywheelThis is a mechanical subject, but is discussed briefly formore clarity. The size of the flywheel, as well as the sizeof the motor, will depend upon the speed variation thatwill be permissible for the type of duty being performed.It should be such that by the time the machine mustFly wheel/7F-perform the next operation it has gained enoughmomentum and regained its consumed energy capableof performing the next operation without undue stresson the motor. This permissible speed variation may be aslow as 1-2% in steam engines and as high as 15-20%for punches and shears, etc.3.9.2 Energy stored by the flywheelwhereF = energy stored by the flywheel in JoulesW = weight of the flywheel in kgV1 = velocity of the flywheel in m/sg = 9.81 m/s2(3.13)After performing the duty, if the velocity of the flywheeldrops to V2 then the energy shared by the flywheel whileabsorbing the shock load-W(V,' - v,' )2.gJoulesFrom the peak load P2 and from the available h.p. of themotor Peq, we can determine the energy to be shared bythe flywheel, i.e.(3.14)(T2 and Teq are in Joules)From this one will be able to ascertain the weight of theflywheel in kg. The velocity V of the flywheel is a designparameter of the basic machine and is derived from there.Based on the speed of the flywheel and weight W, thediameter and width and other parameters, as required todesign a flywheel, Figure 3.20 can be easily determinedwith the help of any mechanical engineering handbook.AFigure 3.19 A brake press illustrating the use of a flywheel(Courtesy: Prem Engineering Works)Figure 3.20 Flywheel

~_____~e~~~ ~~www.taghedanesh.irDuties of induction motors 3/67Relevant StandardsIt c Rile, IS R\- - - - -60033- I / 1996 Rotating electrical machine\ 3122/ 1992 BSEN 60014 11995R'iting and performance 3251199660072- I /I 99 IDimenuons md output wries tor rotating electricalmachine\ Frame number 56 to 400 and tlangenumber 55 to 10801 23 I / I99 I BS 5000-10/19XYB\ 4999 1411198760072 21 IO90 Dinien\ion\ and output \erie\ tor rotating electrical I21 I /I 991 BT 5000 lO/19X960072-?/l993machine\ Frdme number 355 to 1000 and flangenumber I I XO to 2360Diinensions and output \erie\ tor rotating electrical machine\Sindll built-in moton Flange number BF 10 to BF50996/1991BS 3999-101/1987BS 5000- I I / 19x9- ~-NE I\.lA/MG - L/ 1993NbhlA/MG-2/198YNEMAfilG 10119Y4Related US Standards ANSI/NEMA and IEEEMotor\ and generator\ ratings, con\truction testing. and performance~~ -Satety Standard\ (enclosure\) for conctruction and guide tor \election. in\tCillation and use ot rotating machine\Energy man,igement guide tor \eleLtion and u\e 01 three-phdse nio1or\-_____ ~-No re \I In the table\ of relevant Standards in this book while the latest editions of the standards are probided. it is possible that rebised editionshave become available. With the adlance\ of technology and/or its application, the updating of standards is a continuous pi-ocesn bydiflerent standards organizations. It is therefore advisable that for more authentic references. readers should consult the relevantorganiration\ for the latest version of a standard.7 Sonic of the BS or IS standards mentioned against IEC may not be identical.3 The yur noted apain\t each standard may also refer to the year of its last amendment and not nece\\arily the year of publicationList of formulae usedFactor of inertiaGD;, = M.1 of motorGDf = M.1 of load at motor speedM.1 = moment of interiaK = a factor that depends upon the type of relay (generally1 to 1.2)I, = actual currentExponential heating on a hot start(3.1) 0,, = I; +(I; -Ii)(I -c.-"T) (3.4)S, = temperature rise on a hot start above 0:, after t hourin "C.= 0, - 0,Io = initial current at which the machine was operatingI, = actual cui-rent of the machineHeating curvesI,? - Ift = rlog, ~If - kl,'Exponential heating on a cold startAdiabatic heating on a cold starte, = 0,J 1 - P)(3.2)8, = Oc - e:, = If . IO, = temperature rise on a cold start above 0, after t shours in "C.= 8, - 0, Adiabatic heating on a hot start0, = end temperature of the machine in "C after time t0,, = 0, - 0, = 1; i (If- O,, = ambient temperature in OCIf )tlre, = steady-state temperature rise at full load in "Ct = tripping time of the relay in hoursCooling curvesr = heating or thermal time constant in hoursK. If = If ( I ~I") (3.3) 0 = 0 ,inExponential temperature fall when I, = 0I, = rated current of the motor in A f = cooling time constant in hours(3.5)(3.6)(3.7)(3.8)

www.taghedanesh.ir3/68 Industrial Power Engineering and Applications HandbookTo draw the thermal curvesFrom cold conditions(a) When I, 5 200% I,Shock loading(3.9)Energy stored by the flywheel(b) When I, > 200%(3.13)Equivalent output of short-time dutiesP,,(r.m.s) =P? ' tl + P;- ' t2 + P? ' tjtl + t2 + t,(3.10)(3.1 1)F = energy stored by the flywheel in JoulesW = weight of the flywheel in kgVI = velocity of the flywheel in m/sg = 9.81 m/s2Energy to be shared by the flywheel(3.14)Teq (r.m.s) =t, + t 2 + t 3(3.12)T2 = corresponding to peak load P2Teq = corresponding to the h.p. of the motor

www.taghedanesh.irStarting of SquirrelCage InductionMotors

www.taghedanesh.ir

www.taghedanesh.irStarting of squirrel cage induction motors 4/71An induction motor can be regarded as a transformerwith a small air gap in the magnetic circuit. When atrest. having no induced back e.m.f., it can be regarded asa transformer with a short-circuited secondary. A squirrelcage induction motor therefore draws a very high startingcurrent, as noted in Section 1.2.1. However, it reducessubstantially in a slip-ring motor due to the high impedanceof the rotor circuit.The starting of an induction motor does not relate tosimple switching alone. It also involves its switchgearsto control its starting inrush current, starting torque, orboth, and its overload and short-circuit protection.The following are common methods to start a squirrelcage motor, depending upon the limitation, if any, on themagnitude of switching current, I,,.4.1 Direct on-line starting (DOL)This is an ideal type of starting, and is simple andeconomical. In this type the full voltage is applied acrossthe stator windings. The torque developed by the motoris maximum. The acceleration is fast and the heat ofstarting is low (Figure 4. I (a)). For heavy rotating masses,with large moments of inertia, this is an ideal switchingmethod. The only limitation is the initial heavy inrushcurrent, which may cause severe voltage disturbances tonearby feeders (due to a large I,, . Z drop). With this inmind, even local electricity authorities sometimes restrictthe use of DOL starting beyond a certain rating, say, IOh.p., for small installations. For large installations, thiscondition may be of little significance as, in most cases,the power from the electricity authorities may be availableon HT 3.3-33 kV. When it is so, a transformer is providedat the receiving end for the distribution of power on theLT side thus eliminating the cause of a line disturbancedue to a voltage dip. On the HT side, the effect of avoltage dip, caused by small motors, is of littlesignificance. However, the method of DOL switchingfor larger motors is recommended only where the supplysource has enough capacity to feed the starting kVA ofthe motor, with a voltage dip of not more than 5 9 on theLT side.For a large LT motor, say, 300 h.p. and above, there isno economical alternative other than DOL starting. Onecan, however, employ delayed action coupling (Section8.3) with DOL starting to start the motor lightly andquickly. (See also Example 7.1, in Chapter 7 for moreclarity.). Soft starting through a solid-state device (Section6.16.1) or a liquid electrolyte (Section 4.2.3) are costlypropositions. Autotransformer starting i s also expensivedue to the cost of its incoming and outgoing controlgears and the cost of the autotransformer itself. This isalso the case with Y/A starting. Moreover, AIT and Y/Astartings are reduced voltage startings and influence thestarting torque of the motor. It is possible that the reducedstarting torque may not be adequate to drive the loadsuccessfully and within the thermal withstand time ofthe motor. Unlike in smaller ratings, where it is easy andeconomical to obtain a high ir;, characteristic, in largemotors, achieving a high T,,, say, 200% and above, maybe difficult and uneconomical. See Section 2.2.With the availability of large contactors up to 1000 Aand breakers up to 6400 A, DOL starting can be used forLT motors of any size, say, up to 1000 h.p. The use ofsuch large LT motors is, however, very rare, and isgenerally not recommended. Since large electricalinstallations are normally fed from an HT network,whether it is an industry, residential housing, an officeor a commercial complex, DOL switching. even when______c tf30I, r-- -i_11 3 On h2 Off h3Speed + N,(a) Torque and current characteristicstrip(b) Power and control circuit diagramFigure 4.1Direct on-line starting

www.taghedanesh.ir4/72 Industrial Power Engineering and Applications Handbooklarge LT motors are used, may not pose a problem orcause any significant disturbance in the HT distributionnetwork. It is, however, advisable to employ only HTmotors, in large ratings.HT motors have little alternative to be switched otherthan a DOL or a soft starting. Y/A switching in an HTmotor is neither advisable nor possible for its windingsare normally wound in a star formation to reduce thewinding's design voltage and economize on the cost ofinsulation. Autotransformer switching is possible, butnot used, to avoid a condition of open transient during achangeover from one step to another, as discussed inSection 4.2.2(a) and for economic reasons as noted above.Moreover, on a DOL in HT, the starting inrush current isnot very high as a result of low full-load current. Forexample, a 300 h.p. LT motor having a rated full-loadcurrent (FLC) of 415 A on a DOL will have a startinginrush of approximately 2500 A, whereas a 3.3 kV motorwill have an FLC of only 45 A and starting inrush on aDOL of only 275 A or so. An HT motor of 3.3, 6.6 or1 1 kV will thus create no disturbance to the HT distributionnetwork on DOL starting.For power and control circuit diagrams refer to Figure4.1 (b).4.2 Reduced voltage startingWhen the power distribution network is available on LT,and the motors are connected on such a system, it becomesdcsirable to reduce the starting inrush currents, for motorsbeyond a certain rating, say, 10 h.p., to avoid a large dipin the system voltage and an adverse influence on otherloads connected on the same system. Sometimes it mayalso be a statutory requirement of the local electricityauthorities for a consumer to limit the starting inrush ofmotor currents beyond a certain h.p. to protect the systemfrom disturbances. This requirement can be fulfilled byadopting a reduced voltage starting.Sometimes the load itself may call for a soft start anda smoother acceleration and a reduced voltage switchingmay become essential.A few common methods to achieve a reduced voltagestart are described below.4.2.1 Staddelta (Y/A) startingThe use of this starting aims at limiting the starting inrushcurrent, which is now only one third that of DOL starting.This is explained in Figure 4.2. This type of starting is,however, suitable for only light loads, in view of a lowertorque, now developed by the motor, which is also onethird that of a DOL. If load conditions are severe, it islikely that at certain points on the speed-torque curve ofthe motor the torque available may fall short of the loadtorque and the motor may stall. See the curves in Figure4.3(a), showing the variation in current and torque in thestar and delta positions and the severe reduction in theaccelerating torque.When employing such a switching method precautionsshould be taken or provision made in the starter to ensurethat the windings are switched ON to delta only when(a) Line currentin case of Y connected windingsI--_ 1-31or I, = I,,(b) Line current IE2 in case of A connected windingsFigure 4.2the motor has run to almost its full speed. Otherwise itmay again give almost a full kick as on a DOL anddefeat the purpose of employing a staddelta switching.Figure 4.3(a) explains this.Limitations1 Due to the greatly reduced starting torque of the motor,the accelerating torque also reduces even more sharplyand severely (Figure 4.3(a)). Even if this reducedaccelerating torque is adequate to accelerate the load,it may take far too long to attain the rated speed. Itmay even exceed the thermal withstand capacity ofthe motor and be detrimental to the life of the motor.One considers this aspect when selecting this type ofswitching.To achieve the required performance, it is essentialthat at every point on the motor speed-torque curvethe minimum available accelcrating torque is 15-20%of its rated torque. In addition, the starting time mustalso be less than the thermal withstand time of themotor. For more details see Section 2.8.2 This type of switching is limited to only LT systemfor HT motors are normally wound in star. However,in special cases, HT motors can also be designed fordelta at a higher cost to the insulation system whichmay also call for a larger frame size. Also a provisioncan be made in the switching device to avoid thecondition of an open transient during the changeoverfrom Y to A as discussed in Section 4.2.2(b).

www.taghedanesh.irI Minimum \ \Load torque \\II ISpeed -60%Note Current jumps to almost DOL current even at around 60%speed, if switched to A position.mIGMU8IR Y B NT T T IStarting of squirrel cage induction motors 4/73O/COn Offtrip(a) Torque-current characteristics on a VA startingFigure 4.3Star-delta startingMotor windings in A(b) Power and control circuit diagram3 This type of switching requires six cable leads to themotor as against three for other types of switching toaccomplish the changeover of motor windings from Yto A.Ratings of contactorsSince all the contactors now fall in phase, they may berated for the phase current only, i.e. IrIa 58% of&The star contactor is in the circuit only for a short periodand is generally chosen a size lower than the line and Aconlactors.For power and control circuit diagrams refer to Figure4.3( b) .4.2.2 Autotransformer (AIT) startingFor smoother acceleration and to achieve a still lowerstarting current than above, this type of switching, althoughmore expensive, may be employed. In this case also thestarting current and the torque are reduced in a squareproportion of the tapping of the autotransformer. Thenormal tappings of an autotransformer are 40%, 60%and 80%. At 40% tapping the starting current and thestarting torque will be only 16% that of DOL values. At40% tapping, therefore, the switching becomes highlyvulnerable as a result of greatly reduced torque andnecessitates a proper selection of motor.To determine the tapping of the autotransformerConsider an autotransformer with a tapping at 40%. Thenby equating the powers of the primary and the secondarysides of the autotransformer (Figure 4.4)or I,, = (0.4)' . V, IZwhile the starting current on DOL:I,,,7= 2:3 . V,IZ... IAT = (0.4)' . IDOLi.e. proportional to the square of the tapping, whereIAT = starting current on an autotransformer switchingIDOL = starting current on a DOL switchingZ = impedance of the motor windings referred to thestator side per phaseGeneralizing the above equation, autotransformer tappingfor a particular starting current, IAT is

www.taghedanesh.ir4/74 Industrial Power Engineering and Applications HandbookR Y B Ni-tau0.4K1 1 1 1hl1oicTripNotes(1) C, and C, switches or contactors are interlocked among themselves, so that only one is ON at time.(2) C, is essential to isolate the transformer after performing the starting duty, to eliminate avoidable transformerlosses. Moreover the transformer is only short-time rated.Figure 4.4Auto-transformer starting (or 40% tapping)Tapping = 45 . 100%(4.1)From this equation, the desired tapping of the autotransformer,to limit the starting current to a desired valuecan be determined.Example 4.1A squirrel cage motor has its lSi on DOL as six times its ratedcurrent. Find the required tapping on an autotransformer tolimit the starting current to 1.5 times.or 50%Rating of autotransformerSince an autotransformer is in the circuit for only a shortperiod (during the start only), it can be short-time rated.The rating of the autotransformer can be calculated fromkVA( ,= continuous rating of the autotransformer.Since the transformer will be in the circuit for only 15to 20 seconds, the approximate short-time rating of thetransformer can be considered to be 10-15% of itscontinuous rating. The manufacturer of the autotransformer would be a better judge to suggest the mostappropriate rating of the transformer, based on the tappingand starting period of the motor.Rating of contactorsStar contactor C, and AIT contactor C2 must be rated forthe square of the percentage tapping. For a tapping of80%, for instance, the rating of the contactors Cl and C2will be (0.Q2 or 64% of the full-load current of the motor.The main contactor C,, however, will be rated for thefull-load current.For power and control circuit diagrams refer to Figure4.4.kVA (CMR, = 43 . kV ' I,,wherekV = applied voltage, and(4.2)Example 4.2For a 3.3 kV 450 kW motor, with a full-load current of 100 Aand starting current on DOL as six times the rated current,the kVA rating of the transformer for 50% tapping will be

www.taghedanesh.irStarting of squirrel cage induction motors 4/75/AT = 0.!j2 x 100 x 6 = 150 A from equation (4.1)-:. kVA(CMR) = 43 x 3.3 x 150= 860 kVAAn autotransformer of nearly 100 kVA continuously-ratedshould be sufficient for this application.Note The above example is only for a general reference.The CDF of the transformer, for the short-time rating, shouldbe increased with the starting time and the number of startsper hour. Refer to the transformer manufacturer for a moreappropriate selection.4.2.2(a) Open transient condition during areduced-voltage switching sequenceWhenever a changeover of a switching device (contactorsgenerally) from one condition to another takes place, asdiscussed above, in changing over from one tapping toanother, as in an autotransformer switching, or from starto delta as in a Y/A switching, there appears a small timegap of, say, 20-80 ms before the next contactor closes,after the first has dropped. During this period, while themachine will drop its speed only very marginally, andwhich may not influence the load, the power from acrossthe motor terminals will cease for this duration (exceptits own induced e.m.f.). This time gap, when switchingHT motors particularly, may cause switching transientsand prove disastrous for the motor windings, as discussedin Section 17.7.2. This is termed an open transientcondition. The situation is aggravated further because ofthe motor's own induced e.m.f., which may fall phaseapart with the applied voltage and magnify the voltagetransients, in addition to causing current transients. Thecurrent transients may far exceed even 14-20 times therated current of the motor, as illustrated in Figure 4.5,depending upon the transient recovery voltage (TRV)(Section 17.6.2). We will describe the effect of an opentransient condition on an LT and HT machine separately.1 LT motors In LT motors, such a situation may notbe a matter of concern as no switching surges wouldgenerally occur. The voltage would be too low tocause a re-strike between the interrupting contacts ofthe contactors (Section 17.7.6) and cause surges. Themotor's own induced e.m.f. may, however, fall phaseapart with the applied voltage and the voltage acrossthe motor windings may double. In all likelihood thewindings of a motor would be suitable to withstandeffect of the same (Table 1 1.4). In locations, however,that are humid or chemically contaminated, or wherethe motor is likely to be switched on after long gaps,it is possible that the windings may have attained alow dielectric strength to withstand a voltage up totwice the rated one. Open transient conditions mustbe avoided in all such cases.2 HT Motors Y/A switching in HT motors is rare,while an A/T switching may become necessary whenthe capacity of the feeding transformer is not adequateto withstand the start-up inrush of DOL switching, orwhen the drive calls for a frequent switching, such asin a large compressor or pump house, and the feedingtransformer is not adequate for such a duty, or whena number of large drives are to be switched in quicksuccession and the feeding transformer may not beadequate to sustain such heavy inrush currents.c TEaISt(A)lst(Y)0-0SpeedFigure 4.5\\\.Noteob = Peak up to twice (= 14/J the actual /(A) currentod (it may even exceed 14-201,)ob, = No current transient during a closed transientswitchingc'c = 1 to 2 ms. (opening + closing time of Y and 3contactors respectively)Explanation:A. (i) ob is the current transient during changeoverfrom Y to A in an open transient condition.(ii) Voltage transient across the windings: 2.45 to4.1 V,B. Sequence of a closed transient changeover,(i) When the bridging resistor is introduced inparallel to Y windings at point, a, the totalimpedance of the windings gets reduced andcurrent has small overshoot to ob comparedto its normal current, oa.(ii)There is no voltage transient now and thevoltage across the windings remains at V,.(iii) The Y contactor drops at bl, the impedance ofthe windings becomes high, current drops to9.(iv) The bridging resistor drops at b3 and the motor$3 \current traces back its normal A current curve\~1 d.Open and closed transient conditions in YlA switching1J8

www.taghedanesh.ir4/76 Industrial Power Engineering and Applications HandbookHowever, unlike an LT distribution system, whichmay impose a limitation while switching large motorson DOL, the HT feeding lines in all probability maynot pose any such limitation, as it may be feedingmany more loads and may already be of a sufficientcapacity. In HT systems an open transient conditionmay lead to severe voltage transients, which may provedisastrous for the motor windings. All motors that areswitched A/T are therefore recommended to have asurge suppressor on each interrupting pole as notedin Section 18.8 which will take care of these surges.Precautions such as adopting a closed transientswitching method noted below will be essential wheresurge suppressors are not provided.4.2.2(b) Closed transient switching1 In a Y/A starter When desired, the above situationcan be averted by inserting a bridging resistor in themotor windings through an additional contactor, whichcan be called a transition contactor, C. A typical powerand control scheme is shown in Figure 4.6 for a YfAswitching. This contactor is energized through a timer,TI, just before the desired time of changeover (beforethe Y contactor opens) and energizes the auxiliarycontactor, d, which de-energizes the Y contactor Sthrough its NC contact and energizes the A timer, T2.Timer T2 energizes the A contactor D, thus bridgingthe time of the second contactor and eliminating thecondition of an open transient. In fact, the use oftimer T2 becomes redundant with the introduction ofthe auxiliary contactor, d, which introduces the requireddelay (by its closing time) to close the A contactor D.It is, however, provided to allow only for an additionaldelay. The time of this timer, when provided, may beset low to account only for the transient time. As soonas the changeover is complete, the resistor contactor,C, drops through the NC contact of D.The scheme is termed a closed transient switching.A comparison of the two methods in terms of voltagetransients and current overshoots is given in Table 4.1.2 In an A/T starter The same logic can be applied asdiscussed above. The star point of the AfT is openedand connected through the main contactor C, to providea near replica to a Y/A switching. Figure 4.7 illustratesthe revised scheme.Pressing the start P.B. will energize the auxiliarycontactor d and timer T. The star contactor C, isswitched on and energizes the AIT contactor C2 at thedesired tapping. The motor starts at the required1OCR Reset-1'"IPower circuitM = Main contactorD = Delta contactorS = Star contactorC = Resistor contactorControl circuitITripT1 = Resistor ('ON delay timer)R = Delta ('OFF' delay timer)d = Auxiliary contactorFigure 4.6Circuit diagram for a closed transient Y/A switching

www.taghedanesh.irStarting of squirrel cage induction motors 4/77Table 4.1Comparison between an open and a closed transient switching in terms of voltage and currentSerial no. Condition Open transient Closed transient.__-_____12_.~Voltage transient across the Up to 3 to 5 p.u. i.e. 3 to 5 (+)motor windingsor 2.45 to 4.1Vr Section 17.7.2Current overshoot during The current curve becomes a b c dchangeover from Y to A, (Figure 4.5) and current overshootspoint, a, on Y current curve from oa to ob momentarily, which(Figure 4.5) may exceed 14-201,__ _. .-There is no voltage transient. The voltageacross the motor windings remain at ‘V;The current curve becomes ah,h,h,c,d(Figure 4.5). There is no current overshootbeyond the normal current in A. .-Note Since the surge impedance of a circuit is normally very high, as noted in Section 17.8, it is the voltage transient that is the cause ofconcern in the above case than the current transient.I I ’fi-3L~@Power circuitControl circuitc1 = Star contactorC2 = Transformer contactorC3 = Main contactord = Auxiliary contactorT = Star connection-ON delay timercTrip1h3fFigure 4.7Circuit diagram for a closed transition ATswitchingreduced voltage. After the preset time of timer T thestar contactor C1 falls out. The motor is still energizedthrough the transformer winding without any interruptionduring the changeover. The main contactorC, now energizes and the motor runs at full voltage.The AIT contactor C2 also falls out thus achieving aclosed transient switching sequence.4.2.3 Soft startingSoft starting minimizes the starting mechanical andthermal stressedshocks on the machine and the motor. Itresults in reduced maintenance cost, fewer breakdownsand hence longer operating life for both. Reduced startingcurrent is an added advantage.

www.taghedanesh.ir4/78 Industrial Power Engineering and Applications HandbookThrough semiconductor devices (static drives)In a solid-state static switching device the voltage can bevaried smoothly to any required value from high to lowor low to high without creating a condition of opentransient. For HT motors particularly and large LT motorsgenerally, it provides a more recommended alternativeover an autotransformer or a YIA starting. For details seeSection 6.16.1.Through static electrodes liquid electrolyte orchemical resistance startingThis is a primary resistance starting and has a well-provenFrench technology for soft starting of all types of inductionmotors. The device works on the principle of a decreasein resistance of an electrolyte (chemical) having a negativetemperature coefficient. The passing of the starting currentthrough the electrolyte raises its temperature inside astatic electrode chamber. The rise in temperature of theelectrolyte decreases its own resistance progressively.This device thus provides a natural variable resistanceduring the start-up period and hence the desired variableresistance control. The resistance of the electrolyte variessmoothly, and helps to start up the motor smoothly.The electrolyte normally consists of sodium-based saltsmixed with distilled, de-mineralized (DM) or soft drinkingwater. These salts are neutral and non-corrosive and remainstable throughout the life of the electrodes, which canbe many years. Evaporation and outside contaminationare minimized by providing anti-evaporationIsealant oil.The electrolyte is filled in separate tanks for each phase,each with two electrodes (Figure 4.11 below). Theseelectrolytic resistances are used in series with the motor’sstator windings. During the start-up period, the currentpasses through them and causes a voltage drop, whichallows a reduced voltage to the motor’s stator windings.The current through the electrolyte causes its temperatureto rise and resistance to drop, and thus reduces the voltagedrop. Gradually the voltage applied to the stator windingsbuilds up until it almost reaches the rated voltage. At thisstage the residual electrolyte can be totally cut off fromthe circuit with the help of a shorting contactor. A timercan also be introduced in to the electrolyte circuit toautomatically cut off the electrolyte circuit after a presetstarting time.% Speed ---+Variation in stator resistance during starting@ Initial resistance(@ Resistance at short-circuitingFigure 4.8Variation in electrolyte resistance with speedFigure 4.9 illustrates transient-free switching through suchelectrolyte starters. This is a definite advantage of electrolyticswitching over conventional YlA or autotransformerswitching.Important features of electrolyte switchings1 They have in-built safety features to prevent excessivefrequent starting, by means of thermostats and lowlevelelectrolyte monitors, .1IOStarting characteristicsWith this type of switching we can also obtain similarspeed-torque or speed-current characteristics as withreduced-voltage starting, in a star-delta or an autotransformerstarting. Since the variation in the resistance ofthe electrolyte with the starting heat, is very smooth, asshown in Figure 4.8, the speed-torque and speed-currentcharacteristics are also very smooth. The characteristicsare now without any torque, current or voltage spikes,unlike in YIA or autotransformer startings. The YIA orAIT startings exert voltage and current transients on thedrive during the changeover sequence from star to deltaor from one tapping to the other as noted in Table 4.1. Italso eliminates the changeover open transient condition.Figure 4.9-00% SpeedResidualcut-offSmooth acceleration through liquid electrolyte starters

www.taghedanesh.irStarting of squirrel cage induction motors 4/792 It is important to maintain the level of the electrolyteto retain the desired characteristics on a repeat start.However, this may be necessary only once a year asa result of very little evaporation. In the event of alower level the electrolyte can be filled up with drinkingwater, as in a car battery.3 This type of switching provides very smooth acceleration.This is an advantage of electrolyte switchingsover other conventional types of switchings. It exertsno kicks and calls for no special coupling arrangementto transmit the power smoothly to the drive if therequirement of the drive is to be precise and to havea smoother acceleration.4 Since the starting characteristics will depend uponthe initial resistance of the electrolyte, the concentrationof electrolyte and the active area of the electrode mustbe determined beforehand for a particular type of drive,and the requirements of starting torque and current.Small adjustments at site are, however, possible byvarying the depth of electrodes, adjusting the activearea of the electrode, repositioning the flanges andchanging the concentration of the electrolyte etc.5 Electrode assembly - a general arrangement of anelectrode assembly is shown in Figure 4.10. The valueof the resistance is preset at the works, according toload requirements, starting current, torque limitation,starting time etc. Small adjustments are possible atsite as noted above.6 The electrolytes are non-corrosive and the electrodesdo not corrode with time. This feature is of specialsignificance when compared with an ordinary liquidresistance starter used commonly for slip-ring motors.7 Electrolytes do not deteriorate and therefore do notrequire replacement. The evaporated liquid can bereplenished with drinking water when the level of theelectrolyte falls as a result of evaporation. In Europesuch starters have been used for over 15-20 years.8 Electrolyte switching is a costlier proposition comparedto direct on-line or staddelta switching due to additionalshorting contactor and timer, and the cost of electrolyte,its tank and thermostatic control etc. The cost may,however, be comparable with autotransformerswitching.Application, ratings and sizesElectrolyte switchings are simple in construction andpossess high thermal capacity. They are ideally suitedfor difficult starting duties and remotely located plants,where expert services, such as are required for staticdrives, are not easily available. These starters are notbulky and rating is no bar. The common range is from 1h.p. to 1000 h.p. for LT as well as HT squirrel cagemotors (Figure 4.1 1).Unit first-Unit second-ConnectionLF+Note One single unit is normally designed for 10 to 40 H.P.(a) Typical arrangement of two electrolyte units being used inparallel one above the otherElectrodeTank coverRegulating insulationElectrode for electrolytelevel indicating lampsNylon flangesNylon baseBase clipFigure 4.10 A typical liquid electrolyte electrode assembly(Courtesy: AOYP Engineering)(b) Overview of the starterFigure 4.1 1 Electrolyte starter

www.taghedanesh.ir4/80 Industrial Power Engineering and Applications HandbookRelevant StandardsIEC Title IS BS60076- 1/1993 Power transformers - General Specification for 2026- 111991 BS EN 60076-1/1997tapping and connectionsNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedTapping of autotransformer = 42 . 100% (4.1)ZAT = starting current on an autotransformer switchingZDoL = starting current on a DOL switchingRating of autotransformerI

www.taghedanesh.irStarting andControl of Slip-ringInduction MotorsSelection of rotor resistaiice 5/8Determining external recictanceNumber of steps 5/90Duty cycle and duty rating of recistance units 5/90Temperature rise limit.; 5/91Speed control of slip-ring motor? 5/935.4.1 Resictance for speed control 5/94Moving electrode electrolyte ctarters and controllers 5/945 5.1 As a rotor tesiPtance for slip-ring motors 5/945.5.2 Automatic speed control of slip-ring motors 5/95

www.taghedanesh.ir

www.taghedanesh.irStarting and control of slip-ring induction motors 5/83The use of wound motors is on the wane, for reasons ofinherent advantages of a squirrel cage motor over a woundmotor and the availability of static drives (Chapter 6)which can make a squirrel cage motor perform the sameduties as a wound motor and even better. Nevertheless,the use of wound motors exists and will continue worldwidefor years. Where slip-ring motors are employedwhose only purpose is starting or a limited speed controlthe resistance or electrolyte method of starting has beenmost commonly adopted, for reasons of cost. Static drivesgenerate harmonics and distort the supply voltage, andcall for larger sizes of cables. It is also cumbersome andcost-inhibiting to provide filter circuits, particularly whenthe installation is small.But where accurate speed control is the processrequirement, static controllers, termed 'slip recoverysystems' (Section 6.16.3) are recommended. which inaddition to exercising extremely accurate speed control,also conserve slip losses. Static drives are discussed inChapter 6. Below we will describe a procedure to determinethe value of resistance, its steps and switching andcontrol schemes for these steps for a rotor resistance starter.An electrolyte starter is almost a standard product likea motor and the manufacturer, depending upon the numberof starts and the speed control requirement, can adjustthe quantity of electrolyte, depth of electrodes etc.5.1 Important features of a slip-ringmotorThese motors are switched through their rotor circuit byinserting suitable resistances and then removing themgradually. In view of their varying characteristics throughtheir rotor circuit, they can provide the following features:1 The external resistance adds up to the total impedanceof the motor windings and limits the starting current.It also improves the starting power factor.2 Since the performance of an induction motor can bevaried by altering the rotor parameters, a slip-ringmotor, through its rotor circuit, can be made to suitany specific torque and speed requirement.3 The speed of a slip-ring motor can be vaned throughan external resistance. Therefore the torque can bemaintained at any value up to the pull-out torque inthe entire speed range by suitably varying the externalresistance. (See circle diagram in Figure 1.16 andSection 1.10.2). At lower speeds, however, theefficiency of the motor will be poor, as the output isproportional to the speed. The efficiency would beroughly in the ratio of the two speeds, i.e.In fact, it would be even worse as a result of theequally reduced cooling effect of the fan at lowerspeeds. Since kW = Nr . T, kW would vary withspeed, the torque remaining almost the same throughoutthe speed range. The motor would draw the samepower from the supply as before, which, proportionalto speed variation, would appear as slip loss in therotor circuit. For instance, at 25% slip, the poweroutput will be 75% minus the cooling effect and this25% will appear as a slip loss in the rotor circuit.4 Restriction in starting current and a requirement forhigh starting torque to accelerate heavy rotating massessometimes limit the use of a squirrel cage motor. Forsuch applications a slip-ring motor provides a betteralternative.5 As discussed in Section 2.7.1. during start-up, therotor is more vulnerable to damage due to excessiveheat in the rotor compared to the stator. But in slipringmotors a major portion of this heat is shared bythe external resistance, in proportion to its resistivevalue. Therefore. a slip-ring motor can be switchedON and OFF more frequently, compared to a squirrelcage motor. It can also withstand a prolonged startingtime, while accelerating heavy loads. Now the externalresistance will have to be suitable for such duty/loadrequirements.Slip lossFrom equation (1.9), slip loss = S . P,. If the full-loadslip is Sand the speed varied to slip S, the additional sliploss due to the increased slip= P,(S, - S)= kW (S1 - S) (ignoring rotor losses).Example 5.1If the speed of a 125 kW, 1480 r.p.m. motor is varied atconstant torque to 750 r.p.rn., then the additional slip lossE 125 x (0.50 - 0.01 33)= 125 x 0.4867or 61 kWwhere S = "O0 - 1480 = 0.0133 and SI = 0.51500DisadvantagesA slip-ring motor is expensive, as are its controls,compared to a squirrel cage motor. It also requiresmeticulous and periodic maintenance of the brushes, brushgear, slip-rings, external rotor resistances etc. A squirrelcage motor is thus prefemed to a slip-ring motor. A slipringmotor also requires a larger space for the motor andits controls.5.2 Starting of slip-ring motorsThese can be started by adopting either a 'current limiting'method or a 'definite time control' method. In a currentlimiting method the closing of contactors at each step isgoverned by the current limiting relays which permit theaccelerating contactor of each step to close when themotor current has fallen from its first peak value to thesecond pre-set lower value. The relays determine theclosing time by sensing the motor data between eachstep and close only when the current has fallen to a predeterminedvalue of the current relays. The closingsequence is automatic and adjusts against varying loads.

~www.taghedanesh.ir5/84 Industrial Power Engineering and Applications HandbookThe starting time may be shorter or longer dependingupon the load. The disadvantage is that if for any reason,say, as a result of excessive load due to friction ormomentary obstructions if the motor takes a longer timeto pick up, the second contactor will not close until themotor has picked up to a certain speed, and may thuscreate a false stalling condition and allow persistence ofa higher starting current. In a 'definite time' start theother contactor will close after a pre-set definite time,cutting a piece of external resistance and hence increasingthe torque, giving the motor a chance to pick up. Thus,only this method is normally employed for modem slipringmotor starts and controls. Below we will describeonly with this type of control.5.2.1 Selection of rotor resistanceThe choice of the external resistance to be introduced inthe rotor circuit during start-up will depend upon thetorque requirement or limitation in the stator current,without jeopardizing the minimum torque requirement.Since T,, and Ist are interrelated (Section 2.4) limitationin one will determine the magnitude of the other. Makinguse of the circle diagram or the torque and current curvesavailable from the motor manufacturer, the value of thestator current corresponding to a particular torque andvice versa can be determined. The slip at which thistorque will occur on its operating region can also be foundfrom these curves (Figure 5.1). For this stator current,the corresponding rotor current can be ascertained andthe rotor circuit resistance calculated to obtain this current.If stator full load current = I,Corresponding rotor current RA = I,Figure 5.1- c L gSpeedSmaxOperating region of torque and current curvesRotor standstill voltage = ,,e2Rotor voltage at a particular speed = RV = S . sse2Then, for a stator current of Zrl, corresponding to a startingtorque of T,,,, the required rotor current will beIF1I,1 = - ' I,I,To obtain this rotor current, the required rotor circuitresistance can be calculated as below for the variousconfigurations of the rotor windings and the resistanceunits. It may be noted that, when the resistanceconfiguration is not the same as that of the rotor, it mustSI. no.Rotor configurationExternal resistanceEquivalent configurationsof the rotorTotal rotor circuit resistance3 R p y34sse2 R2:+qsse2 ZRzR2'he 5.2(c)Re'Ae 5,2(d)Re(5.1~)

www.taghedanesh.irI-IStar connected rotor 1'BrushesmStarting and control of slip-ring induction motors 5/85JInI-1Star connected rotorStar connected resistanceFigure 5.2(a) External rotor resistance&Delta connected rotorSliprings17, ----I,, -Delta connected rotor( I/-IStar connected resistanceFigure 5.2(b) External rotor resistanceDelta connected resistanceFigure 5.2(d) External rotor resistancefirst be converted to the equivalent configuration of therotor (see Example 23.4 in Chapter 23 for conversion) tofacilitate calculations.NotesI Normally the external resistances are connected in star. But forlarger motors, calling for high resistance grids, they may also beconnected in delta, to reduce their current rating and hence thecost. Insulation is not a limiting factor usually, in case of resistancegrids.Z The rotor voltage, \,e2 refers to the standstill secondary inducede.m.f.. between the slip-rings. Whereas the rotor current, I,,refers to the full load rotor current, when the slip-rings are shortcircuited.Refer to Figure 5.3(a) for a typical cast iron and Figure5.3(b) for a stainless steel resistance grid used for suchpurposes. Cast iron grids are, however, not preferred dueto their resistance change with temperature, which mayalter the predefined performance of the motor forwhich the resistance was designed, particularly whenthe resistance is used to effect speed variation. Castiron grids are also brittle and may break during trans-portation, installation or maintenance. They may alsonot be able to absorb shocks and vibrations during normalservice.The stainless steel punched grids, generally alloys ofaluminium and chromium, are normally preferred. Figure5.3(c) shows an A1-Cr alloy steel punched grid resistancein a multi-tier arrangement. They are unbreakable, onthe one hand, and possess a high specific resistance (about120 $2-cm), on the other. A high specific resistancehelps in saving the material required to make the grid ofa particular resistance value.More importantly, such alloys also possess a very lowtemperature coefficient of electrical resistance (of theorder of 220 pQIQ/"C, typical), which causes only amarginal change in its resistance value with variation intemperature. They can therefore ensure a near-consistentpredefined performance of the motor for which theresistance grid is designed, even after frequent starts andstops. They are also capable of absorbing shocks andvibrations during stringent service conditions and aretherefore suitable for heavy-duty drives, such as steelmill applications.

www.taghedanesh.ir5/86 Industrial Power Engineering and Applications HandbookMade of1. tough iron casting free from brittleness or2. stainless steel or3. aluminium chromium corrosion and vibration resistantPossesses high specific resistance = 120 FR-cm and lowtemperature coefficient of 0.00022 R/R/"C. Causes negligiblechange from start to run and therefore provides a fairly accuratespeed controlFigure 5.3(a) Cast iron grid resistance (Courtesy: BCH)Figure 5.3(b) Stainless steel punched grid resistance (Courtesy: BCH)Figure 5.3(c)AI-Cr alloy steel punched grid resistor in multitier arrangement (Courtesy: BCH)

www.taghedanesh.irStarting and control of slip-ring induction motors 5/87Example 5.2A 125 kW, 41 5 V, 3@ wound rotor has the following parameters:l, = 230 Asse2 = 500 VI,, = 180 ATpo = 250%R2 = 0.09 ohm (star connected)The torque and current curves are as shown in Figure 5.4.Determine the external resistance required to achieve a startingtorque of 200%.SolutionFrom these curves, for a torque of 200% the stator currentshould be 250%, Le. 230 x 2.5 A.:. Corresponding rotor current,I,,,-180x 230 x 2.5 A230= 180 ~ 2.5 Aand rotor circuit resistance, with a star-connected resistanceunit,500R2, = & x 180 x 2.5= 0.641:. External resistance required, Re = 0.641 - 0.09= 0.551 R per phaseMethod of cutting off the external resistanceThe simplest method is performing it manually. Oncethe total external resistance is known, the resistance unitcan be built with a hand-operated mechanism to manuallycut-off the external resistance. However, this method issuitable only for applications where the magnitude oftorque during the pick-up period (torque at different'r18speeds) is of little consequence. In such starters, there isno control over the time of start or resistance in thecircuit (torque at different speeds) during the start-upperiod. Liquid rotor starters are one type and are suitableonly for light duties. For heavy loads, requiring specifictorque values during pick-up, specific resistances mustbe introduced into the rotor circuit at specific speeds.The specific resistances so required can be determinedas discussed below and manufactured in the form of agrid. These resistance grids are then controlled throughcontactors and timers. The duration between each stepof the resistance grid is pre-determined, the torque demandis already ascertained and the resistance required at eachstep is pre-calculated. The starter can then be madepushbutton-operated fully automatic. With the help ofpre-set timers at each step, the entire resistance is cut offgradually and automatically, maintaining the predeterminedtorque profile.5.2.2 Determining external resistance andtime of startConsider Figure 5.5 with six steps (rotor resistance unitwith five steps) and assume the maximum and minimumtorques as T,,, and Tmin between each step, to suit aparticular load demand (Figure 5.6(a)). Let thecorresponding rotor currents be I,,, and Imin. Then by asimple hypothesis using equation (I .7).lR2Rotor -YShortedduring run2ecoT,-1R2 15-step externalrotor resistanceSpeed +Figure 5.4 Speed-torque and speed-current curves of a125 kW motorFigure 5.5Star connectedresistanceFive-step rotor resistance

etc.www.taghedanesh.ir5/88 Industrial Power Engineering and Applications HandbookTorque curve/- obtainedWith rotor'sown R2-T\S152 s3 s4Speed -NsLFigure 5.6(a)Torque curve with a five-step external resistanceFrom (a) and (b)where R2,, RZ2, ... RZs etc. are total rotor resistances perphase after introducing the external resistances, orAt the start, when SI = 1Similarlyor - R21 = R22 R,,~ = -s2 s3 Smax(5.2)... R22 = p . R2,RZ3 = p R,, = P'andR2 = P . RZ5= P ' P4 . R2,= ps . R2,R. R,, etc.or R 2- p 5 . 2 or p5 = s,,,:. p =SmaxGeneralizing this for a number of resistance steps a,P= G (5.3)

www.taghedanesh.irStarting and control of slip-ring induction motors 5/89C35 dlTId2 /d3d4dd5d2d4d3CIujdd2dl‘Contactors connected in A to reduce their sizeFigure 5.6(b) Power and schematic diagram for a five-step stator-rotor starterOnce p is known, the resistance value for each step Example 5.3can be determined. In the above case, the slip S, Consider a rotor resistance unit with five sections (total numbercorresponds to the slip at the operating point ofthe torque of steps is Six). Then in the previous example (Figure 5.4),considering Tst as 200% and the corresponding slipor current curve, where the desired torque T,,, or currentSm, asroughly 14% thenI,,, occurs. This can be obtained from the torque andcurrent curves available from the motor manufacturer 0.0921(Figure 5.4). --0.14

www.taghedanesh.ir5/90 Industrial Power Engineering and Applications HandbookStep no.Total rotor circuit resistance RRExternal resistance ReRResistance between stepsR1 R2j = 0.6432R22 = p . R21 = 0.676 x 0.643 = 0.4353R23 = p. R22 = 0.676 x 0.435 = 0.29441324 = /?. R23 = 0.676 x 0.294 = 0.1995R25 = p . R24 = 0.676 X 0.199 = 0.1356 R2 = 0.09 = 0.09Total rotor resistance R2, = 0.643 R per phase. See Figure 5.50.643 - 0.09 = 0.5530.435 - 0.09 = 0.3450.294 - 0.09 = 0.2040.199 - 0.09 = 0.1090.135 - 0.09 = 0.0450.09 - 0.09 = 00.643 - 0.435 = 0.2080.435 - 0.294 = 0.1410.294 - 0.199 = 0.0950.199 - 0.135 = 0.0640.1 35 - 0.09 = 0.0450.09 - 0 = 0.090= 0.643 Rand i.e. p =Illlax= 0.676The power and schematic diagram for this configuration ofresistance unit is given in Figure 5.6(b).5.2.3 Number of stepsGenerally, the number of steps is decided by the limitsrequired in the maximum and minimum torque values. Itis therefore not an arbitrary figure as assumed in theabove example. We will, however, conclude in the nextarticle, that the lower the limits of T, and Tmin the higherwill be the required number of steps and vice versa.The recommended number of steps, i.e. the number ofaccelerating contactors for general-purpose (light-duty)application, may be chosen as indicated in Table 5.1.However, for specific load demands and accurateminimum and maximum torque requirements, the numberof steps can be calculated on the basis of actual requirementas discussed in Section 5.2.2. Sometimes, for aneconomical design of the resistance unit, the number ofsteps can be reduced without jeopardizing the basicrequirement of torque except for slightly higher or lowerlimits in its values as shown in Figure 5.6(a). Also, as weapproach the rated speed, the closer become the stepsand the greater the tendency to become infinite. In sucha situation it becomes essential to liberalize the torquelimits to achieve a quicker objective. Moreover, by nowthe motor will have almost run to its full speed and willpose no problem of a current or torque kick. Sometimes,however, when the motor possesses a substantially highpull-out torque than is required by the connected load, aquick removal of external resistance may cause an abruptjump in the torque and may exert a jerk on the load,which may not be desirable. To overcome such a situationand to economize on the resistance design, a smallresistance is sometimes left permanently connected inthe rotor circuit, even when the motor is running at fullload. Such a practice will apparently add to the rotorresistance and increase the full-load operating slip. Thehigher slip losses will, however, be shared by the rotor’sown resistance and the external resistance, which is nowpermanently connected in the rotor circuit. The externalresistance has to be continuously rated. One should,however, make sure that this slightly increased slip doesnot influence the performance of the driven machine orcause overheating of the rotor windings.5.2.4 Duty cycle and duty rating ofresistance unitsThe resistance units, when used only for starting purpose,are in the circuit for only a short time and are thus shorttimerated. However, when they are employed for speedcontrol, braking or plugging operations, in addition tothe starting duty, they may bc rated for continuous duty.The resistance units are thus classified according to theirduty demand, Le. number of operations per hour (c/h).See Chapter 3 on the duty cycle.The following information is essential to decide theduty class of the resistance units:1 The duty cycle of the motori.e. which one out of S-1 to S-10 (Chapter 3)Say, for a duty cycle ofS4 - 40% - 90 c/h and FI as 2.5i.e. each cycle of ~ 6o 6o- 40 seconds902 The starting current and the peak current duringeach cycleBy referring to NEMA publication ICs 2-213: with theabove details, the resistance class can be found. Theresistance units are designated by class number, currentand resistance. NEMA tables ICs 2-213 to ICs 2-213-5can be referred to for this purpose.Table 5.1Recommended number of starting steps of a resistance unit to switch a slip-ring motorMotor output in kW for different starting torquesStandard no. ofstarting stepsTorque 50% T, 100% T, > 100% T,kW I20 kW I 10 kWI7 320 < kW I 200 10 < kW S 100 7 < kW I 704kW > 200 kW > 100 kW > 70 6

www.taghedanesh.irStarting and control of slip-ring induction motors 5/915.2.5 Temperature rise limitsAs in NEMNICS-2-2 13. the exposed conductor resistanceunit:, can attain a temperature rise up to 375°C. whereaswhen they are enclosed in a housing, the temperaturerise must be restricted to 350°C. Accordingly, thetemperature rise of the air in the vicinity of the housing,when measui-ed at a gap of 1 inch from the ericloaure,should not exceed 175°C.Corollar;i,Very high temperature-rise permissible limits of resistanceunit:, render them unsuitable for installations which aretire-prone. such as pulp and paper industries, chemicalindustries, refineries, textile mills, etc. For specificapplications and surroundings, however, resistance designcan be altered (derated) to restrict the temperature rise towithin desirable limits.Moreover, high-temperature variation may result inlarge variations in the resistance of the grid and mayvary the performance of a variable-speed drive if care i snot exercised in selecting a proper alloy of the resistance.An alloy such as aluminium-chrome steel should bepreferred when it is required to perform a speed controlor a speed variation. as discussed above.5.3 Hypothetical procedure tocalculate the rotor resistanceThe tollowing is il nioi-e appropriate method to dctcrrriinethe number of steps and resistance of each step of theresistance grid, making use of only rotor data and desiredlimits of T,,:,, and T,,,, as ascertained from the availableload curve. The concept used in arriving at the numberof steps is based on the fact that the rotor current variesin direct proportion of torque (equation ( I .I)). For moreclarity we will discuss this method using a practicalexarriple. The procedure is generally the same as thatadopted in Example 5.3.Example 5.4Consider a conveyor system, requiring an average torque of100% during pick-up. The motor data are as follows:kW= 450N, = 980 r.p.m.V, = 6.6 kVI, = 50 Asse2 = 750 VI,, = 435 ATpa = 250%R2 = 0.02 R (star connected)SolutionFor conveyors, the torque should not be very high, toeconomize on belt size and cost. Otherwise a higher safetyfactor must be considered for the belts, resulting in a wider orthicker belt at an extra cost. Therefore considering a torquedemand of T, as 180% and T,,, as 120%, for a conveyortorque of 1 OO%, providing an average starting torque of 150%and an accelerating torque of 50%, (Figure 5.7(a)):For T, (1 80%) R2, = 750 12,3 x 435 x 1.8= 0.553 Rand for T,,, (120%) R2, = 750 '2 (1 etc>3 x 435 x 1.2Determine slip S, to give the same value of R,, as above.Then for this slip, determine one step on the torque curve.Calculate for this slip S,, R,,, etc. as shown in Table 5.2.Since the torque at synchronous speed is zero and the endportion of a torque curve is almost a straight line, the sameresults are obtained if a straight line is drawn from startingpoint 'a' to zero torque 'A'. Where it meets the T,,, linedetermines one step. Here again, the torque goes up to T,by another resistance value and this can also be joined to A.Continuation of such a procedure gives a point where it meetsthe motor curve. To reduce the number of steps, slight variationsin T, and T,,, can be made as shown. Note that completeresistance cannot be removed at step f otherwise the torquewill jump to about 212%, which may not be desirable. Thus,the total number of steps amount to seven (resistancesegments six), which is reasonable.We must also check whether the starting time of the motorwith this profile of starting torque would be safe for the motorto pick up to the rated speed. Considering the same data asfor Example 7.1, GD: = 1866 kgm'and T,450 x 974 o,5=980= 223.62 mkg:. Accelerating time1866 x 980t, =375 x 223.62= 21.8 secondsThis seems to be high and must be checked with thethermal withstand time of the motor (Section 3.5).The upper limit of the starting torque chosen at 180%appears to be too low and can be raised to, say, 210%. Thiswill also help in reducing the number of steps. Number ofsteps at 4-6 are preferable to economize on the cost ofswitchgears, and yet provide a reasonably smooth (free fromovershoots) starting torque.Considering T, = 210%= 65%andfs = 16.77 secwhich is quite reasonable. A modified accelerating torquediagram is drawn in Figure 5.7(b), providing a smoothacceleration. Now we can cut off the last resistance at e,giving a jump in T, of even less than 210%.To obtain the starting characteristics according to Figure5.7(b), it is essential to calculate the total rotor resistance,R,, and resistances between each step along the lines ofTable 5.2.5.3.1 Calculation of time betweeneach stepTo make the whole starting sequence automatic in acontactor type automatic starter unit it is essential toknow the time the motor will take to accelerate from oneslip to another between each step. It is required to select

www.taghedanesh.ir5/92 Industrial Power Engineering and Applications Handbook250212180I150P)p.& 120c8100-100 67 44 30 18 11 6 2TAS1 52 s3-s4 S5 S6 S7 N,% SlipSpeedFigure 5.7(a) Determining the number of steps and accelerating torque between each step250210 23r5180;$208100100-S1 s2 53 54 s5 Nf+-- % SlipSpeedFigure 5.7(b) Determining the number of steps and accelerating torque between each step

~~ 1.8RZ5www.taghedanesh.irStarting and control of slip-ring induction motors 5/93Table 5.2Stepno.I~-7I34567Determining the resistance between each stepTotal resistance (R) for T,,,SymbolValueRAt slip750R21 li'5 x 435 x 1.8 SI = 100%= 0.553Rzz R, . S2 = 0.371 Sl = 61%R23 R1 S3 = 0.244 s3 = 44%R24 R,. S, = 0.166 S, = 30%RZh R, . S6 = 0.061 S6 = 11%R2 R, = 0.020 S7 = 6%Evaluation of slip for T,,,resistances, = =s, - 44%Resistancebetweeneach stepRR22Rz2 - Rz3 = 0.1271.8R23S, = 1.2 . S3 308 Rz3 - R,, = 0.0781.8s -'.2.s, - 18% R,, ~RZ4= 0.070R25s 6 ----.s,1.21.8= 11% Rz~ - R26 = 0.035R26S, ='.2.S,1.8-6% RZ6 - RZ = 0 041R? (at rated speed) = 2% R, = 0.020NoteTotal resistance, R = 0.553Rand set the timer relay to automatically remove, one byone, each resistance step and provide a smooth acceleration.The procedure for calculating the acccleration timebetween each step is dealt with in Chapter 2 and aschematic diagram is given in Figure 5.6(b).5.4 Speed control of slip-ringmotorsThe speed of a slip-ring motor can be varied by up to25% of the rated speed. A further reduction may greatlydiminish the cooling effect and reduce the output in amuch larger proportion and will not be worth while.Moreover, it will now operate in a region which is unstableand may thus stall (see any speed-torque curve). Asdiscussed in Section 5.1, during speed reduction, withthe torque remaining almost constant, the motor willdraw the same power from the supply lines while theoutput will be proportional to the speed minus the coolingeffect. On the subject of cooling, Table 5.3 shows theapproximate values of h.p. and torque a motor will beable to develop at various speed reductions. Figure 5.8gives the curves for output and torque. From these curvesit is evident that the losses increase in a much largerproportion than the speed variation. The speed in suchmotors can be vaned in the following four ways:1 Torque and output varying as in the curves, in whichcase no derating is necessary.2 Keeping torque constant throughout the speed range.At reduced speed the torque is low and therefore themotor rating should be derated accordingly, e.g. at50% speed, the torque is 73%. To obtain 100% torque,the motor should be rated for 100/0.73, i.e. 137%.Table 5.3 Variation in torque and output with speed inslip-ring motors56781 %Seriai~of ratedno. speed1 260504030outputRatingrequired forconstantoutput (5%)% 3100.0 10086.8 11573.6 13661.0 16448.0 20836.5 27426.0 38416.5 605Torque or current ratingRatingrequired forconstanttorque (Si8 4100 10096 10492 10987 11580 12573 13165 15455 182Notes1 These value are only indicative and may vary from onemanufacturer to another, depending upon their design and coolingefficiency and also the normal speed of the motor. The higherthe motor rated speed, the lower will be the cooling at lowerspeeds, compared to a slow-speed motor, and will require yethigher deratings.2 The rotor current also varies in the same proportion as the torque.Table 5.3 shows the derating values and Figure 5.9the derating curve. The variation in the rotor currentis also in the same proportion as the torque.3 Torque varying with the square of the speed as forfans and centrifugal pumps etc. as discussed in Chapter2.4 Keeping the output constant throughout the speedrange. As a result of still higher derating, the torqueof the selected higher size motor (column 3 of Table5.3) will become very high and is generally notpreferred. See also Figure 6.5 1.

www.taghedanesh.ir5/94 Industrial Power Engineering and Applications HandbookYTo achieve a better torque, the slip-ring rotors are normallywound in star, in which case the rotor current is 8 timemore than in delta for the same output. Also since thetorque is proportional to the rotor current equation (1. 1),the torque developed will be greater in this case.Example 5.5For the 125 kW motor of Example 5.2, if a speed reduction isrequired by 50% at constant torque (see also Figure 6.51)and the rotor current is now 73% of its rated value (Table5.3), then the total rotor circuit resistance- 500~0.5- x 180 x 0.73= 1.098 !2and external resistanceRe = 1.098 - 0.09or 1.008 Q04 %Speed-Shaded portion indicates power loss in the rotor circuitFigure 5.8 Variation in torque and output with speed5.5 Moving electrode electrolytestarters and controllers5.5.1 As a rotor resistance for slip-ring motorsThese are similar to stator resistance starters, as discussedin Section 4.2.3 and can be used in the rotor circuit tocontrol the rotor side resistance. Figure 5.10 shows thesmooth variation of resistance by electrolytic vaporizationcompared to a conventional metallic resistance variation.The self-variable resistance of electrolyte is equivalentto almost three or four steps of a metallic resistance andmakes such starters economical. Normally one step issufficient for motors up to 160 h.p. (For speed-torque0 10 20 30 40 50 60 70 80 90 100% Speed- 4Figure 5.9 Multiplying factor for motor rating for speed variationat constant torque5.4.1 Resistance for speed controlIn this case the regulating resistance grids are normallycontinuous duty, unlike those for start-up, which are shorttimeduty. Equations (5.1 a-d) can be used, for determiningthe total rotor circuit resistance for a particular speedvariation, i.e.ss e2R21 =-'2/7. I,%Speed reduction%Current at the reduced speed(5.4)0 25 50 75' 1 )O% Speed- NrVariation in resistance during starting1. 3-Step metallic resistance2. Liquid electrolyteFigure 5.10 Variation of electrolyte resistance with speed

~~ ~~ ~~www.taghedanesh.irStarting and control of slip-ring induction motors 5/95Electrode at theback (not visible)0 25 50 75 + io0% Speed- N,1. DOL torque2. Liquid electrolyte (rotor circuit)3. 5-Step metallic resistance4. Load torqueFigure 5.11 Smooth variation of torque with smaller numberof steps in liquid electrolyte starterscharacteristics see Figure 5.1 I). Starting requirementssuch as electrolyte quality and electrode depth, the activearea of the electrode and the positioning of the flangesetc., are determined by the requirement of the drive.The remaining details are the same as for statorresistance starters. Here also, by adding individualelectrolyte stacks, starters of any rating up to 25 000 h.p.can be produced from a smaller unit of 10 h.p. or so(Figure 5.12). These starters are almost 20-25% moreeconomical than a conventional contactor and timeroperatedmetallic rheostatic starters discussed earlier.Figure 5.12 Liquid rotor starter (Courtesy AOYP Engineering)5.5.2 Automatic speed control of slip-ringmotorsBy making the electrodes move through a geared motor,it is possible to achieve even automatic speed control ofslip-ring motors through such starters.Relevant StandardsIEC Title IS BS-. ~60947- 1/1998 Specification for low voltage switchgear and control 13947-1/1998 BS EN 60947-1/199260947-4-1 - 1990gear, General rulesElectromechanical contactors and motor starters, 13947-4-1/1993 BS EN 60947-4-1/1992including rheostatic rotor startersNEMA/ICS-2-213/1993Related US Standards ANSINEMA and IEEERewtors and rheostatsNotesI In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

www.taghedanesh.ir5/96 Industrial Power Engineering and Applications HandbookList of formulae usedSelection of rotor resistancessR21 = R2 + Re = - e243 ' 1,l(Rotor Y, external resistance Y)R21 = R2 + 3Re =43 . sse2___Id(Rotor A, external resistance Y)(Rotor Y, external resistance A)(Rotor A, external resistance A)R2 = Rotor resistance WphaseRe = External resistance RlphaseRZ1 = Total rotor resistance R/phaseR = required rotor circuit resistancesse2 = rotor standstill voltageIml = required rotor current(5.la)(5.lb)(5.1~)(5.ld)To determine external resistance andtime of startR2l = Si .R2-sm,(5.2)R21, R22, are total rotor resistances per phase afterintroducing the external resistancesSI = slip at the beginning of a stepS, = slip at the end of a stepP=-R22-R23-....R2lR22a= no. of resistance stepsResistance for speed controlssR21 = -. e2$j . I,(5.3)% Speed reduction% Current at the reduced speed (5.4)

www.taghedanesh.ir6/97Static Controls andBraking of MotorsContents 6.16.2 Soft %tanning 61140L I Y6.16 3 Slip-recoverj system (to control nouiid rotor6.1 Speed control in squirrel cage motors 6/99motors) 611406 1 1 One winding 6/996.16.4 Application of solid-state technolocy in the6 1.2 Two windings 6/996.26.36.46.56.66.76.86.9Speed control through solid-stat6 2.1 Theory of application 6/6 2.2 Effectc of vaiiablc-supply parameters on theperformance of an induction motor 61101V/f control (speed control at constant torque) 61101Phacor (vector) control 6/1036.4 1 Single phasor (tector) control 6/1046 4 2 Field-oriented control (FOC) 6/1066 4.3 Direct torque control (DTC) 6/108Uqe of phasor control for flux braking 6/11 1Control and feedback devices 611 116 6 1 Speed sensor5 611 1 1Application of solid-state technology 611 116 7 1 Power diodes 611 126.7.2 The power transimr famil6 7.3 The thyristor family 61114Conduction and commutation 6/11Circuit configurations of semicondn6 9 1 Coni erter or rectifier unit 616.9.2 Inverter unit 611 196.9.3 Voltage source iiivertei (VSI) using IGBT5 611256 9.4 Current source inverter (CSI) 611 266 9.5 Cyclo converters (frequency converters) 6/1276.9.6 The regenerative schemes 611276.10 Smoothing ripples in the d.c link 6/126.1 I6.12 Providing a constant current source 6/1296.13 Generation of harmonics and xwitching surges in astatic device switching circuit 6/1296.13 1 Supprehmg the harmonic\ 6/1306.146.156.16Providing a constant d.c. voltage sourcProtection of semiconductor devicez and motors 6/136.14 I Overvoltages and voltage surges caused bydisturbance\ in an LT 5ystem 61130Energy conservation through did-state te6.15 1 Illustration of energy conservation6 15.2 computation of energy saving 61135Application of static drives 611386.16 I Soft staiting (VJcontrol) 6/13operation of a process plant 611426.5 Other application5 61144peed variation through variable-5peed fluid couplin,nq 6il-I.i6.18 Static drive verbu\ fluid coupling 611356.19 D.C. drives 611476 20 Braking 611476.20 1 Types of braking 6/15 IInduction generators 6i1.56Inching or logging 611616.23 Number of starts and stops 61161List of forinulae used 61163Further reading 61163State agencies in India for micro citing of windmill\ 61163

www.taghedanesh.ir

www.taghedanesh.ir6.1 Speed control in squirrelcage motorsSpeed control in slip-ring motors has been discussed inthe previous chapter. Squirrel cage motors have limitationsin their speed control in view of their fixed rotor parameters.Speed variation, in fixed steps, however, is possiblein such motors if the stator is wound for multipoles andsuch motors are known as pole changing motors. Up tofour different speeds can be achieved in such motorseconomically, in combinations of 1/4,4/6,4/X, 6/8,6/1’.?,1/4/6,4/6/X, 1/4/6/12 and 4/6/8/13 poles etc. or any othersimilar combination. For limitation in the motor size andflux distribution, winding sets of more than two are notrecommended. The two windings can be arranged fortwo. three or (maximum) four different speeds.6.1.1 One windingThe hgle winding can be connected in delta/doublestar (NYV) to give two combinations of poles in the ratioof 2: I . i.e. 4/2. 8/4 or 1Y6 poles etc. as shown in Table6.1.6.1.2 Two windingsWhen more than two or non-multiple speeds are required(2.g. 46 or 6/8 etc.) then two windings are necessary.Each can further be connected in NYYas noted above togive one additional speed for each winding and can thusLe arranged Tor three or four different speeds as shownin Table 6. I .6.2 Speed control throughsolid-state technologyIn the following text we have discussed how, with theapplication of varying supply parameters (V andf), onecan alter the characteristics of a fixed parameter inductionmotor in any desired way. We then deal with the applicationof solid-state technology to obtain the variations in thefixed supply parameters to achieve the required controlsin an a.c. machine.The static drives also provide a few more advantagessuch asI They transform an unbalanced supply system automaticallyto a balanced supply system through theswitching logistics of the ICBTs* or the SCRs*. Thefeature is termed dynamic phase balancing.2 Since the starting inrush current is kept moderate forall types of drives. it can economize not only on ratingsof the switchgears and cables but also on the size ofthe generator when a captive power is required tofeed the load.*IGBTs - Insulated gate bipolar transistorsSCRs - Silicon-controlled rectifiers (thyristors)Both are discussed in the subsequent text.Static controls and braking of motors 6/99The method for speed control as discussed earlier areonly conventional and can only provide a fixed speedvariation say, from 3000 r.p.m. to 1500 r.p.m. to 750r.p.m. or vice versa. They cannot provide a smootherspeed variation between any two speeds. The applicationof variable voltage is also not practical nor advisable, forit means a poor performance by the machine at lowervoltages. whereas a higher voltage (more than 5% of therated) is not permissible. Moreover, through this methodthe speed can be varied only within a very limited spandue to very unstable conditions below the T,,, region,and a more than proportionate reduction in the h.p.developed. The torque also reduces in square proportionof the voltage. For such applications, therefore. whichrequired smoother speed variation and over a wide range,one had no choice but to select d.c. drives. These driveswere costly and needed higher maintenance becausecommutators, sliprings and brushes etc. caused continuousarcing and required constant checks and maintenance aswell as more downtime, which a process industry couldleast afford. Systems were used that required very elaboratearrangements, using two or more a.c. machines, renderingthe whole system very cumbersome, vulnerable and yetmore expensive. Since the speed was normally changedthrough the variation in frequency (Nm.0, these systemswere basically frequency changers (converters) and wereknown asCascade connections (concatenation)Use of two motors and the prime mover was a slipringmotor.Frequency convertersSchrage type motor (commutator brush shiftingarrangement)Leblanc or Scherbius Advancers.These systems were evolved to provide a variable -frequency supply source to feed directly the statorterminals of the a.c. motor or its rotor through the sliprings.The motors had to be invariably a combination oftwo or more slip-ring motors to receive the rotor frequencyvoltage from the other machine or feed back the rotorfrequency voltage to another machine. The easiest methodwas to have a variable-frequency supply source, whichwas not possible, unless the supply source itself wascaptive and specified for this drive alone or a combinationof these drives on the same bus.There was thus a practical limitation in employing ana.c. motor for all such applications that required frequentspeed variation. Since these drives are no longer in practice.we have not considered it relevant to provide more detailsof these systems.The above methods provide speed variation in steps,as in squirrel cage motors or in two machines or more,as in frequency converters, and cannot be used for aprocess line, which requires frequent precise speedcontrols. Until a few years ago there was no other optionwith all such applications and they had to be fitted withd.c. motors only. D.C. motors possess the remarkableability of precise speed control through their separatearmature and field controls. In d.c. motors the speed

www.taghedanesh.ir6/100 Industrial Power Engineering and Applications HandbookTable 6.1Connection diagrams for multispeed motorsNo. of wdgs Poles Connection Connection diagramOne 014 or 412 etc. Deltddouble star aCb’bTwo016 or 614 etc.Staristar or deltaldeltaMotor winding(i)A.b 8-pole0)b’8-pole(ii)6-pole(ii)4-pole(iii)Two01614 or 61412 etc. One deltddouble starand one star of deltaa‘Winding no. 1814-pole0)8-pole(ii)4-pole(iii)f Winding A no. 2 e6-poleTwo121016/4 etc. Two deltddouble starWinding no. 112/6-pole(i)A dC be’12-pole(ii)6-pole(iii)Winding no. 28/4-pole(iv)4-pole(vi)control below the base speed can he achieved throughthe armature control at constant torque and above thebase speed through the field control at constant h.p. (Figure6.7(b)). But a d.c. machine also has a few limitations:1Frequent maintenance due to continuously rubbingbrushes mounted on a commutator.2 A continuous arcing as a result of the above, givingrise to a source of fire hazard, particularly at installationsthat are contaminated with explosive gases,vapour or volatile liquids or are handling materialsthat are hazardous.An induction motor, particularly a squirrel cage, is

~ Single-phasorwww.taghedanesh.ircheap, robust and is devoid of any such operatinglimitations and, has an obvious advantage over d.c.machines. It alone can provide an immediate answer tosuch limitations. With the advent of static technology asdiscussed later, it has now become possible to make useof cage motors with the same ease and accuracy of speedcontrol and that are even better than d.c. machines. Staticdrives response extremely fast as they can be microprocessorbased. They can compute process data andprovide system corrections almost instantly (called ‘realtimeprocessing’) as fast as within 1-2 ms and even less.In Table 6.5 we show a broad comparison between a d.c.machine and a static drive using cage motors. It gives anidea of applying static technology to all processrequirements with more ease and even better accuracy.With the advent of this technology, the demand for d.c.machines is now in decline as noted in Section 6.19.Static controls and braking of motors 6/1016.3 Vlfcontrol (speed control atconstant torque)This is also known as variable frequency control. Considerthe following equations from Chapter 1 :and el = 4.44 KCL’ . $, . Z, ‘,f, (13):. for the same supply voltage VI6.2.1 Theory of applicationThe application of solid-state technology for the speedcontrol of a.c. motors is based on the fact that thecharacteristics and performance of an induction motorcan now be varied, which until a few years ago wereconsidered fixed and uncontrollable. This concept is nowa matter of the past. With the advent of solid-statetechnology, which was introduced around 1970 forindustrial application, the motor’s parameters and thereforeits performance can now be varied by varying the supplyparameters of the system, for example the voltage andfrequency in cage motors, and rotor resistance or rotorcurrent in slip-ring motors, as discussed in Chapter 1.This technology can also provide a varying resistance inthe rotor circuit of a slip-ring motor by varying the rotorcurrent as discussed in Section 6.16.3 without the loss ofpower in the external resistance. It is thus also suitableto provide speed control in a slip-ring motor. Speed controlof slip-ring motors with the use of solid-state technologyis popularly known as dip recovery systems, as the slippower can also be fed back to the source of supply througha solid-state feedback converter bridge, discussed later.6.2.2 Effects of variable supply parameters onthe performance of an induction motorHere we analyse the effect of variation in the incomingsupply parameters (voltage and frequency) on the characteristicsand performance of an induction motor (such asits flux density, speed, torque, h.p., etc). We also assessthe effect of variation of one parameter on the other, andthen choose the most appropriate solid-state scheme toachieve a required performance. We generally discussthe following schemes:1 Vlfcontrol (speed control at constant torque)2 Phasor (vector) control(vector) control- Field-oriented control (FOC), commonly known asdouble-phasor or phasor (vector) control and- Direct torque control (DTC)i.e. for the same design parameters ($, remaining thesame) and ratio ezlf,, the torque of the motor, T, willremain constant. Since both e2 andji are functions of thesupply system, a variation in V, and f can alter theperformance and the speed-torque characteristics of amotor as required, at constant torque. By varying thefrequency smoothly from a higher value to a lower oneor vice versa (within zero to rated). an almost straightline torque can be achieved (Fig. 6.3). This type of acontrol is termed variable voltage, variable frequency(v.v.v.f or Vu) control. At speeds lower than rated, thenatural cooling may be affected, more so at very lowspeeds, and may require an appropriate derating of themachine or provision of an external or forced cooling.The practice of a few manufacturers, up to mediumsizedmotors, is to provide a cooling fan with separatepower connections so that the cooling is not affected atlower speeds.Note The speed of the motor can be varied by varying the frequencyalone but this does not provide satisfactory performance. A variationin frequency causes an inverse variation in the flux. $I, for thesame system voltage. The strength of magnetic field, $I, develops,the torque and moves the rotor, but at lower speeds, f would bereduced, which would raise $I, and lead the magnetic circuit tosaturation. For higher speeds, f would be raised, but that wouldreduce @ , which would adversely diminish the torque. Hencefrequency variation alone is not recommended practice for speedcontrol. The recommended practice is to keep Vlf as constant, tomaintain the motor’s vital operating parameters, i.e. it\ torque and&,, within acceptable limits.The above is valid for speed variation from zero to therated speed. For speed variations beyond the rated speed,the theory of VI’ will not work. Because to maintain thesame ratio of Vlfwould mean a rise in the applied voltagewhich is not permissible beyond the rated voltage, andwhich has already been attained by reaching the ratedspeed. The speed beyond the rated is therefore obtainedby raising the supply frequency alone (in other words,by weakening the field, $,) and maintaining the voltageas constant at its rated value. We can thus achieve a

www.taghedanesh.ir6/102 Industrial Power Engineering and Applications Handbookspeed variation beyond the rated speed by sacrificing itstorque and maintaining the product T . f as constant.Since P = T . N, and N, = h, therefore speed variationcan now be achieved at constant power output (see Figure6.4). We have combined Figures 6.3 and 6.4 to produceFigure 6.5 for more clarity, illustrating the speed variationat constant torque below the base speed and at constanth.p. above the base speed.Vlfis the most commonly used method to control thespeed of a squirrel cage motor. The fixed frequency a.c.supply, say, at 415 V, 50 Hz from the mains, is firstrectified to a constant or variable d.c. voltage, dependingupon the static devices being used in the inverter circuitand their configuration. This voltage is then inverted toobtain the required variable-voltage and variable-frequencya.c. supply. The basic scheme of a Vlfcontrol is illustratedin Figure 6.6. The approximate output voltage, currentand torque waveforms are shown in Figure 6.7(a). Thetorque curve is now almost a straight line, with onlymoderate pulsations, except for some limitations, asdiscussed later, enabling the drive to start smoothly. When,however, a higher starting torque is required to start themotor quickly, it is also possible to boost the startingtorque to a desirable level (up to the designed T,, of themotor) by raising the voltage to V,, through the pulsewidth modulation (PWM), discussed in Section 6.9.2.The starting current can also be reduced to only 100-150% of the rated current or as desired, to the extentpossible, by suitably varying the Vlf. The control canthus also provide a soft-start switching. It may, however,be noted that there is no control over the starting current,which is a function of the applied voltage, and theminimum voltage during start-up will depend upon themotor, the load characteristics and the thermal withstandtime of the motor (Section 2.8). With the use of staticcontrol drives, it is however, possible to minimize thestarting inrush current to a reasonable level but with aloss of starting torque. To do so, it is advisable to matchthe load and the motor starting characteristics anddetermine the minimum starting torque required to pickup the load. For this starting torque is then adjusted thestarting voltage. The magnitude of the starting currentwill then depend upon this voltage.The static drives too are defined by their overcurrentcapacity and its duration. E C 60146-1-1 has defined itas 150-300%, depending upon the type of application,for a maximum duration of one minute. For instance, forcentrifugal drives, which are variable torque drives(Figures 2.10 and 2.1 I ) such as pumps, fans andcompressors etc. an overcurrent capacity of the order of1 10% for one minute would be adequate. For reciprocatingdrives, which are constant torque drives (Figures 2.12and 2.13), such as ball mills and conveyors, a higherovercurrent capacity say, 15&300% for one minute, wouldbe essential. For higher starting currents or longer durationsof start, the drive may have to be derated, which themanufacturer of the drive alone will be able to suggest.Drive ratings are basically selected on the basis of motorrating and the starting currents and are rated for a startingcurrent of 150% for one minute. If more than 150% isrequired, then the rating of the drive is worked out on thebasis of the starting current requirement and its duration.Vlfis a concept to vary the speed, maintaining a constanttorque. Through a static drive, however, it is possible tovary one parameter more than the other, to obtain anyspeed-torque characteristic to meet a particular duty cycle.By varying the rectifier and inverter parameters, a wholerange of Vlfcontrol is possible. High-inertia loads, callingfor a slip-ring motor for a safe and smooth start, can nowmake use of a standard squirrel cage motor. Open- orclosed-loop control systems can be employed to closelymonitor and control the output voltage and frequency sothat the ratio of Vlfis always maintained constant.Limitations of V/f controlFigure 6.1 illustrates theoretical speed-torque curves. Infact at very low speeds, say, at around 5% of N, (at 5%f ) or less, the motor may not be able to develop itstheoretical torque due to a very low stator voltage, on theone hand, and relatively higher proportion of losses anda lower efficiency of the machine, on the other. Figure6.2 illustrates a realistic torque characteristic at differentsupply frequencies. They display a sharply drooping andrather unstable performance at very low speeds. For moreaccurate speed controls at such low speeds, one mayhave to use phasor-controlled drives, discussed later.Phasor-controlled drives are available at reasonable costand can provide extremely accurate speed controls evenat very low speeds, 5% of N, and less, or cyclo-converters,which are relatively very costly, for large drives. Typicalapplications calling for such a high torque at such lowspeeds could be a steel plant process line or materialhandling (e.g. holding a load stationary at a particularheight by the crane, while shifting material from onelocation to another).At Ns = 5%and less, thef= 1v= 1motor is not f decreasing f risingable to maintainSpeed or frequencyits Tstand TpolevelsFigure 6.1 Approximate theoretical representation of speedversus torque for V/f control of a motor at different values of f

www.taghedanesh.irStatic controls and braking of motors 611 031 10 20 30 40 50 60 70 80 90 HzBelow the base speed -Abovethe base speedFrequency (Speed) 4* Drooping torque at lower speeds. At higherspeeds too, the torque profile is variableFigure 6.2 Actual speed-torque characteristics by a conventionalfrequency control (V/f control)0 0 25f1 O.5Of1 0 75f1 1 OOf,Ns4 Ns3 Ns2 Ns 1Speed0 Tst1 = 0 Tst2 =1 11 10 Ts13= 0.5f 0 Ts,4= 0.25f(V/f control &, constant)Figure 6.3 Speed variation at constant torques= 1Figure 6.4Speed -(Variation in frequency)Speed variation at constant HPApplicationIn a V/fcontrol generally, only the frequency is vaned toobtain the required speed control. Based on this frequency,the switching logistics of the inverter control circuit controlthe inverter’s output voltage using the PWM techniqueto maintain the same ratio of V/f. A Vlfcontrol is, however,not suitable at lower speeds. Their application is limitedto fan, pump and compressor-type loads only, where speedrcgulation need not be accurate, and their low-speedperformance or transient response is not critical and theyare also not required to operate at very low speeds. Theyare primarily used for soft starts and to conserve energyduring a load variation (see Example 6.1 for more clarity).Rating, however, is no bar.6.4 Phasor (vector) controlA simple V/fcontrol, as discussed above, will have thefollowing limitations:Control at very low speeds is not possible.0 Speed control may not be very accurate.0 Response time may not be commensurate with thesystem’s fast-changing needs.

www.taghedanesh.irT- !f/f HP- Tf0 Constant torque region I Constant HP regionSpeed (f) -Figure 6.5 Speed control in an a.c. motorWhile all these parameters are extremely essential fora process line, with the R&D in the field these limitationshave been overcome with the use of phasor controls. Toimplement these controls different manufacturers haveadopted different control and feedback systems to monitorand control the torque and field components. They havealso given these controls different trade names. The basictechnological concept remains the same but processimplementation may vary from one manufacturer toanother. Below we attempt to identify the more cQmmonphasor controls introduced by a few leading manufacturers.1 -6.4.1 Single phasor (vector) controlLet us consider the simple equivalent motor circuit diagramas shown earlier in Figure 1.15. The no-load componentof the current, In,, that feeds the no-load losses of themachine contains a magnetizing component, I, . I,produces the required magnetic field, $, in the stator andthe rotor circuits, and develops the rotor torque so thatT .x $m . In (1.1)The magnetizing current, I,,,, is a part of the motor statorcurrent Il (Figure 1.15). The rotor current is also a reflectionof the active component of this stator current, as can beseen in the same figure, so that7, = 7", + 7,_ _ _= I; + I, +I, (6.1)All of these are phasor quantities. I, is the active componentresponsible for developing the rotor torque and I, themagnetic field. Varying I, would mean a correspondingvariation in the torque developed.Variation of speed below the base (rated) speedThe machine now operates in a constant torque region(see Figure 6.7(b)). The frequency is reduced as is thevoltage to maintain the same ratio of Vlf. At lower voltages,II and therefore I, will diminish, while $, and I, willrise, so that 4,. I, is a constant.Equation (6.1) can be rewritten, for better analysiswith little error as- -I, = I, +I,3-4a.c. supply -Fixed ax.@ IsolatorDiode bridge rectifier (converter)Inverter unit IGBT or thyristor,depending upon the size of machine.CTTacho-generator for open loop orencoder for a closed loop controlSpeed potentiometerSpeed comparatorSpeed amplifier and controllerCurrent comparatorCurrent amplifier and controllerGate control in case of thyristorinverters only.3-0 Motor(2'Figure 6.6 Typical block diagram of a V/f control scheme with open- or closed-loop control scheme

www.taghedanesh.ir- fConsant torque regionStatic controls and braking of motors 6/105Constant HP regionVoltageI 1 @ =Torque curveArmature voltage fixedfield current reduced2 = Output curveI, =Armature currentI0CurrentTorque-NBase meedSpeed (= f)Base speed - It is normally the rated speed at which the ratedparameters are referred (T,, HP and V,)Figure 6.7(a) Approximatecharacteristics of vital parametersafter pulse width modulationFigure 6.7(b)machine)Variation of torque with speed in a d.c. machine (same for an a.c.where,I,-T = V, andTherefore a normal Vlfsan also be transformed into aphasor control, I, and I, being torque- and flux-producingcomponents respectively. These components arerepresented only theoretically. In fact they are not separateand hence the difficulty in controlling each of them moreprecisely.Variation of speed beyond the base (rated) speedSee Figure 6.7(b). The machine now operates in a constanth.p. region. The frequency is raised but the voltage iskept constant at its rated value (as it should not be raisedbeyond rated). The flux will diminish while I, and alsoI,, will remain almost the same. The torque thereforereduces so that the h.p. developed remains almost aconstant (h.p. 0~ T.W. This is also known as the fieldweakeningregion.In the above schemes the two quantities (I, and I,) arenot separated. Initially it was not easy to separate themand the whole phasor I, was varied to achieve a speedvariation. Yet close speed control was possible but themotor’s basic parameters were essential to achieve moreaccurate speed control. Since it may not be practical toobtain all the parameters of each motor promptly, thedrive software is designed so that the name plate particularsof a motor (V, I,, Nr, and kW) alone are enough to determinethe machine’s required parameters through the motor’smathematical model. The motor is calibration* run onno-load at different voltages and speeds and the drive isable to establish near-accurate vital parameters of themachine in terms of the equivalent circuit diagram shownin Figure 1.15 earlier. With these parameters known, it isnow possible to achieve the required speed controls notedabove. Since these parameters are fixed for a motor, themotor has to be selected according to the load duty andmay require a pre-matching with the load.ApplicationSuch a control is good for machines that are required tooperate at low speeds with a high accuracy. Now thephasor I,, in terms off,,,, is varied according to the speedrequired. Figure 6.2 now changes to Figure 6.8, which isa marked improvement on the earlier characteristics. Thetorque variation with speed is now almost constant, exceptat very low speeds. The reason for poor torque at lowspeeds is the method of speed variation which is stillbased on V/f Now a motor’s mathematical model is used______*1 Calibration run or autotuning is a feature of the motor’smathematical model that can establish the motor parameterbwith its test run.*2 If the actual motor data are available frnm the motor manufacturer,a calibration run will not be necessary. The motor’s mathematicalmodel can now be fed with the explicit data to achieve moreprecise speed control.*3 All the motor models are implemented by mlcrocomputersoftware.

www.taghedanesh.ir6/106 Industrial Power Engineering and Applications HandbookNearly constant torque speed Icontrol with ver!012 5 10 20 30 40 50 60Hz-Below the base speed t+----) Above the base speedFrequency (Speed)Figure 6.8 Speed-torque characteristics by flux (I,,,) control(single phasor control)to vary the speed of the machine by sensing the component,I,, of the machine. Any variation in the actual I,,, thanthe desired pre-set value in the inverter switching logisticsis made up by the PWM technique. The field-orientedblock diagram illustrated in Figure 6.12 below can besuitably simplified for I, control. Tachogenerator or pulseencoder feedback devices can be employed to achievehigher accuracy in speed control.With years of research and development in the field ofstatic drives, it is now possible to identify and separatethese two parameters (I, and I,,,) and vary themindividually, as in a d.c. machine, to achieve extremelyaccurate speed control, even slightly better than in d.c.machines. In d.c. machines the armature current and thefield strength are also varied independently. A.C. machinescan now be used to provide very precise speed control,as accurate as +_ 0.001% of the set speed, with closedloopfeedback controls. This technique of speed controlis termed field-oriented control (FOC) and is discussedbelow.6.4.2 Field-oriented control (FOC)This is commonly known as double phasor or phasor(vector) control. If we analyse equation (1.1) in Chapter1 we will observe that @is a function of stator magnetizingcurrent, I,, and I, is a transformation of the stator activecurrent, I,.Hence, equation (1.1) can be rewritten asT oc I, '7,Both of them are phasor quantities, and are shown inFigure 6.9. In absolute terms, they can be represented byT = k . I, . I, sin 8(6.2)where 8 represents the electrical position of the rotorfield in space with respect to the stator. In other words,it is the- phasor displacement or slip angle betweenI,,, andI, and will continue to vary with variation inload (torque). For instance, referring to Figure 6.9, thesmaller the load I,,, the smaller will be sin 8, and thelarger the load Ia2, the larger will be sin &. Thus toachieve a required level of speed control the stator current,I,, field current, I,,,, and phasor angle, 8, can be suitablyvaried. Since it is the phasor of the rotor flux (rotatingfield), i.e. the magnitude and its angular position withrespect to the active current of the stator, which is beingvaried, to achieve the required speed control, this phasorcontrol is called field oriented control (FOC). The theoryof field orientation was first introduced by F. Blaschkein 1972 (see Blascke (1972) and EPE Journal (1991)).Having been able to identify the rotor field phasor it isnow possible to vary this and obtain a speed control in asquirrel cage machine similar to that in a d.c. machine.For field-oriented controls, a mathematical model ofthe machine is developed in terms of rotating field torepresent its operating parameters such as N,, I,, I, and8 and all parameters that can influence the performanceof the machine. The actual operating quantities are thencomputed in terms of rotating field and corrected to therequired level through open- or closed-loop controlschemes to achieve very precise speed control. To makethe model similar to that for a d.c. machine, equation(6.2) is further resolved into two components, one directaxis and the other quadrature axis, as discussed later.Now it is possible to monitor and vary these componentsindividually, as with a d.c. machine. With this phasorcontrol we can now achieve a high dynamic performanceand accuracy of speed control in an a.c. machine, similarto a separately excited d.c. machine. A d.c. machineprovides extremely accurate speed control due to theindependent controls of its field and armature currents.Different manufacturers have adopted different methodswith minor changes to achieve almost the same objective.For example, field-oriented control was first introducedby Allen Bradley in the USA in 1981 and a similartechnique was introduced at the same time by ABB ofFinland. ABB claim their technique to he still faster inresponding, as it eliminates the modulation section ofFigure 6.9 Phasor representation of field current (I,,,) andstator active current (I,)

www.taghedanesh.irStatic controls and braking of motors 611 07the drive (which we will discuss later, when discussingdrives), and the torque could be controlled directly. Theycall it direct torque control (DTC).The phasor I, and I,, are separated and then controlledseparately as discussed later. For more precise speedcontrol a pulse encoder feedback device can also beemployed. The characteristics now improve to Figure6. IO. The torque can now be maintained constant at anyspeed, even at zero speed.With different approaches to monitor and control thebasic parameters of the motor, Le. I,, I, and sin 8, manymore alternatives are possible to achieve the requiredspeed variation in an a.c. machine. Control of theseparameters by the use of an encoder can provide anaccuracy in speed control as good as a d.c. machine andeven better.Figure 6.11Rotor field reference frameTo implement the FOCThe field phasor is a continuously rotating phasor in thespace, whose angular position keeps changing with theposition of the rotor with respect to the stationary stator.Let the rotor field displacement under the stationarycondition with respect to the stator be denoted by anglepas shown in Figure 6.11. This displacement will continueto change and will rotate the rotor (field frame). All thephasor quantities of the stator are now expressed in termsofthe field frame. Figure 6.11 shows these two equivalentstator side phasors transformed to the rotor frame.I; = corresponding stator phasor for active current,referred to the rotor sideI,:, =corresponding stator phasor for the magnetizingcurrent, referred to the rotor side8 = phase displacement between the stator active andmagnetizing current componentsp = angular displacement of the rotor field with respectto the stationary stator at a particular instant. It wcontinue to vary with the movement of the rotorThe phasor diagram would suggest that:T5eco01 5 10 20 30 40 50-60 70 80 90HzBelow the base speed _I_ Above the base speedFrequency (Speed)Figure 6.10Speed-torque characteristics by field-orientedcontrol (FOC) (flux and torque control) (Source: Allen Bradley)1: sin 0 can be regarded as the quadrature component.It is the torque component on which will depend thetorque developed by the rotor. It is this componentthat will be varied for speed variations below the basespeed, maintaining the field current constant accordingto the rated condition. It is similar to the armaturecurrent control in a d.c. machine.I& can be regarded as the direct axis field componentresponsible for the field flux. This can be weakened(reduced) for speed variations above the base speed,which is the constant-output, constant-voltage region.Now the torque component will diminish. It is similarto the separately excited field control in a d.c. machine.Both these phasors can be regulated separately like ad.c. machine to achieve any speed variation with highprecision and accuracy and provide a high dynamicperformance. Leading manufacturers have developedmathematical models with microprocessors to determinethe modulus and the space angle p of the rotorflux space phasor through I, and speed. The spaceangle of the rotor flux space phasor is then obtainedas a sum of slip angle Band the field angle p. The slipangle Bcan be calculated from the reference values I,and I, (an indirect method, as no sensors are used).With these parameters known, it is now possible toidentify the position of the rotor flux phasor and thenorientate the stator current phasor to determine therelative displacement between the two in the space.to achieve a phasor diagram as shown in Figure 6.11.The phasor diagram provides crucial parameters andmust be established accurately to obtain accurateresults from the control of 1: and 16, . The relationshipbetween these two phasors, It: andl,;, is thenmonitored closely and controlled by adjuhling thesupply parameters to the stator of the machine. Thesupply parameters can be controlled through a VSI(voltage source inverter) or CSI (current sourceinverter), which will be discussed latcr, dependingupon the practice of the manufacturer.The microprocessor plays the role of an electroniccontroller that transforms electrical quantities such as V.I and N etc. into space flux phasors, to be compared withthe pre-set data. It then creates back V. I and N etc.,

www.taghedanesh.ir6/108 Industrial Power Engineering and Applications Handbookwhich are the controlling variables, in terms of correctionrequired and feeds these to the machine through a VSI orCSI to achieve the required controls.Mathematical modelling of the machine is a complexsubject and is not discussed here. For this, research anddevelopment works carried out by engineers and thetextbooks available on the subject may be consulted. Afew references are provided in the Further reading at theend of this chapter. In the above analysis we haveconsidered the rotor flux as the reference frame. In factany of the following may be fixed as the reference frameand accordingly the motor’s mathematical model can bedeveloped:Rotor flux-oriented control - when the rotor is consideredas the reference frame.Stator flux-oriented control - when the stator is consideredas the reference frame andMagnetic field-oriented control - when the field isconsidered as the reference frame.The rotor flux-oriented control is more popular amongdifferent manufacturers to achieve high prccision of speedcontrol in an induction machine. With this technology(any of the three methods noted above), it is now possibleto obtain a high performance of the machine, i.e. torqueup to 100% of T, at speeds down to zcro.Since the motor’s fixed parameters can now be variedto suit a particular load requirement, there is no need topre-match a motor with the load. Now any motor can beset to achieve the required characteristics to match withthe load and its process needs. Full-rated torque (T,) atzero speed (during start) should be able to pick up mostof the loads smoothly and softly. Where, however, a higherTst than T, is necessary, a voltage boost can also be providedduring a start to meet this requirement. (See also Section6.16.1 on soft starting.) The application of phasor (vector)control in the speed control of an a.c. motor is shown ina block diagram in Figure 6.12.ApplicationFOC drives are capable of providing precise speed controland are used for applications calling for high performanceand precision (e.g. machine tools, high-speed elevators,mine winders, rolling mills, etc.). These drives are capableof regulating a number of variables at the same instantsuch as speed. position, acceleration and torque.6.4.3 Direct torque control (DTC)This is an alternative to FOC and can provide a very fastresponse. The choice of a static drive. whether through asimple Vlfcontrol, field-oriented phasor control or directtorque control with open or closed-loop control andfeedback schemes, would depend upon the size of themachine, the range of speed control (whether required tooperate at very low speeds, 5% N, and below), the accuracyof speed control and the speed of correction (responsetime). The manufacturers of such drives will be the bestguide for the most appropriate and economical drive fora particular application or process line.This technology was introduced by Ws ABB of Finlandto achieve an extremely fast and highly accurate speedcontrol in an a.c. machine. This is also based on phasorcontrol, but the field orientation is now obtained withoutusing a modulation (PWM) control circuit. The manufacturermakes use of only motor theory, through a highlyaccurate mathematical motor model, to calculate the motortorque directly. There is no need to measure and feedback the rotor speed and its angular position through anencoder to the motor model. The controllable variablesnow are only 4 and T. It is therefore called a direct torquecontrol (DTC) technique. There is no need to control theprimary control variables V or I and N now as in a fluxphasor control. This scheme eliminates the signalprocessing time in the absence of a PWM and also anencoder circuit, both of which introduce an element ofdelay. As these drives control @and Tdirectly, they respondextremely quickly and it is possible to achieve a responsetime as low as 1-2 ms. They are thus a corollary to thesensor-less flux control drives. For reference, a roughcomparison between the various types of drives is givenin Table 6.2.Basic scheme of a DTC driveA simple block diagram as shown in Figure 6.13 illustratesthe operation of a DTC drive. It contains two basic sections,one a torque control loop and the other a speed controlloop. The main functions of these two control circuitsare as follows:1 Torque control loopSection 1This measuresthe current in any two phases of the motorthe d.c. bus voltage, which is a measure of the motorvoltagethe switching position of the inverter unit.Section 2This is a highly advanced motor model, which is firstmade to read and store the machine’s vital parameterssuch as R,, L,, saturation coefficients and its momentof inertia during an autocalibration run. The motor isrun under a locked rotor condition and the mathematicalmodel is capable of computing its basic characteristicsin terms of these parameters or any data that may be ofuse to actuate the control logistics. The block diagramis drawn in an open-loop condition, which would sufficefor most process lines. For still higher accuracy in speedcontrol, an encoder may be introduced into this circuit,as shown by the dotted line in Figure 6.13. The quantitiesmeasured in Section 1 are also fed into this section,which are able to compute the actual operating valuesof T and @ about every 25 p s, i.e. around 40 000 timesevery second. These are the output control signals ofthis section.Section 3The above actual operating data are fed into a torque

www.taghedanesh.irStatic controls and braking of motors 6/109AC supply0 Speed comparator: to determine the speed error (wet- w) where,q,, = required speed reference andy = actual speed.@ Speed control amplifier to feed the reference torque datala sin to the torque error detector. This is the quadraturecomponent.@ Current comparator: It is the torque error detector, to detect thetorque component error Ala sin 8 (/,sin 8,,ef, - I, sine)@ Torque error amplifier to feed the torque error to carry out thedesired correction through block 6.@ Field error A/,,, amplifier.@ Phasor rotator to transform the field frame coordinates to the statorframe coordinates.@ Field weakening unit to command the field strengthregulation above the base speed.Figure 6.12for speed@ Field current comparator (A/,,,) to determine (/,c,fi- I,,,).@ Motor flux mathematical model to determine prevailing I, sin 8andI, and also the space field displacement angle with respect to thestator 'p'.@ Current comparator to compute the error between the actual linequantities and the desired quantities from block 6 and givecommand to the control unit 11.@ Switching block controls the switching of the inverter unit to regulateits output to the preset reference line quantities.@ Inverter unit.@ Pulse encoder-To feed back actual speed of the motor and theangular position of the rotor with respect to the stator at a particularinstant.Block diagram for a flux-oriented phasor controlcomparator and a flux comparator, which are alreadysupplied with predefined reference data to which themachine is required to adjust. Any errors in these twodata (reference and actual) are compared by thesecomparators at an extremely fast speed as noted aboveand provides Tand @error signals to the switching logisticsof the inverter unit.Section 4This determines the switching pattern of the inverter unit,based on the T and @ error signals, obtained from thetorque and flux comparators. Since these signals areobtained at very high speed, the inverter IGBTs are alsoswitched with an equally high speed to provide a quickresponse and an accurate T and N.2 Speed control loopSection 5This is a torque reference controller which controls thespeed control output signal through the required torquereference signal and the d.c. link voltage. Its output torquereference is fed to the torque comparator (section 3).Section 6This is the speed controller block that consists of both aPID (proportional integral derivative, a type of programming)controller and an acceleration compensator. Therequired speed reference signal is compared with theactual speed signal obtained from the motor model (section2). The error signal is then fed to both the PID controller

www.taghedanesh.ir6/11 0 Industrial Power Engineering and Applications HandbookSpeed control loop -II Torque control loop AC supplyI.-cPID - proportional integralderivative (a type of programming)Figure 6.13 Block diagram of a direct torque control inverter circuitTable 6.2 Comparison of conventional V/f control with different modes of phasor control drivesSerial no PerformanceparametersVariable-frequencydrivesV/f controlWithoutencode?Withencode#Phasor (vector) control drivesFlux (I,,,) control Double-phasor (vector) or(single-phasor field-oriented controlcontrol) ( F Wa, c Without Withencode? encoderbDirect torque control(DTC)Withoutencode?Withencode#Response time inadjusting thespeed (regulation)200-400 ms 100-200 ms125-250 ms10-20 ms 5-10 ms1-2 msdAccuracy of speedcontrol (regulation)Speed rangeAbility to maintain100% rated torqueat zero speed* 1.0% * 0.1%40: 1 40: 1Not possible Not possiblet (Figure 6.2) +f 0.5%120:lNot possible(Figure 6.8)f 0.5% +0.001%120:l 1000: 1Yes Yes+(Figure 6.10) +*0.1% +0.001%100: 1 1001Yes Yest (Figure 6.10)+Notes1 All values are approximate and are for reference only. For exact values consult the manufacturer.2 All these drives are based on pulse width modulation (PWM) and hence would produce overvoltages at the inverter output and requireovervoltage protection for cable lengths of 100 m (typical) and above, depending upon the steepness of the wave (Section 6.14.1).3 The performance of the drive would also depend upon the accuracy of the motor’s mathematical model used for the phasor control.4 The choice of the type of drive would depend upon the degree of speed regulation required by the process.a Open-loop controlsClosed-loop controlsThese drives are normally open loop (sensor-less) without encoder. For higher regulation, it is better to adopt a two-phasor control, suchas a field-oriented control (FOC) or a direct torque control (DTC) drive.Response time without an encoder is sufficiently low. Where response time alone is the prime consideration, the encoder is not necessary.

www.taghedanesh.irStatic controls and braking of motors 6/111and the acceleration compensator. The sum of these twois the output signal that is fed to the torque referencecontroller (section 5).Section 7This is the flux reference controller which provides theabsolute value of stator flux to the flux comparator (section3). The value of this absolute flux can be varied to fulfilmany functional requirements from the inverter unit such;ISField strengthening - to obtain speed variation belowthe base speedFlux braking - to carry out braking dutiesFicld weakening - to obtain speed variation above thebase speed.6.5 Use of phasor control forflux brakingIt is possible to perform braking duties by the motor byraising the level of magnetization (field strengthening).By raising flux, the speed reduces (Figure 6.7(b)) andthe stator current rises. This is an apparent advantage inthis kind of a speed control. as the heat generated by themotor during braking appears as thermal energy in thestator rather than in the rotor. Also it is easier to dissipateheat from the stator than from the rotor due to its (stator)bulk and its outer surface. which is open to the atmosphere.6.6 Control and feedback devicesThe control and feedback circuits are also solid-statedevices and offer high reliability and accuracy. The outputfrom these devices can be interfaced with a microprocessorto carry out the required corrections in the system’sparameters through the inverter controls. A microprocessoris a semiconductor device and consists of logic circuitsin the form of ICs (integrated circuits), capable ofperforming computing functions and decision making.These capabilities are used in carrying out processcorrections by providing the necessary timing and controlcorrective signals to the switching logistics of the inverterunit.A microprocessor is capable of solving complexmathematical problems very rapidly and can analyse asystem more closely and accurately. It can be fed with avariety of measuring and control algorithmic softwarethat may be necessary for system monitoring and control.They are also capable of performing supervisory anddiagnostic functions and carry out historical recordingio store information related to process and drive conditions,io facilitate reviews of data for continuous processmonitoring. fault analysis, diagnostics, trend analysis,etc.The control logistics such as PWM or frequency controlsare diFitd1 circuits and are microprocessor based. Theycan compare the actual inverter output parameters withpre-set reference parameters and help to implement therequired precise adjustments in the system’s parametersinstantly by providing corrective command signals tothe switching circuits of the inverter unit. This in turnadjusts the system variable parameters within the requiredlimits by adjusting V andfas in a simple Vlfcontrol orI,,, andf as in a flux phasor control, or I,, I,,,, p and f asin a field-oriented control or the torque phasor control asin a direct torque control technique etc. As a result oftheir accuracy and speed, they are capable of achievingprompt corrections, high reliability. better flexibility andhence a high dynamic performance of the drive.There are many types of sensors used to feed-back theprocess operating conditions to the switching logisticsof an inverter unit. They can be in terms of temperature,pressure. volume, flow, time or any activity on whichdepends the accuracy and quality of the process. Directsensing devices used commonly for the control of a driveand used frequently in the following text are speed sensors.as noted below.6.6.1 Speed sensorsThese are closed-loop sensing devices and are mountedon the machine or a process line. They are able to sensethe operating parameters and provide an analogue ordigital feedback input to the inverter switching logistics.For example:Tacho generator (TG) -This is an analogue voltagefeedback device and can provide a speed input to theinverter control circuit. It provides only a low level ofspeed regulation, typically i 0.4% of the set speed.Pulse encoder - This is a digital voltage feedbackdevice and converts an angular movement into electricalhigh-speed pulses. It provides speed and also theangular position of the rotor with respect to the statorfield when required for field-oriented control to theinverter control circuit to achieve the required speedcontrol. It is a high-accuracy device, and providesaccuracy up to k 0.001% of the set speed. (See thesimplc feedback control scheme shown in Figure 6.12.)6.7 Application of solid-statetech no1 ogyThis field is very large and a detailed study of the subjectis beyond the scope of this handbook. We will limit ourdiscussions to the area of this subject that relates to thecontrol of a.c. motors and attempt to identify the diffcrcntsolid-state devices that have been developed and theirapplication in the control of a.c. motors. Only the morecommon circuits and configurations are discussed. Thebrief discussion of the subject provided here, however,should help the reader to understand this subject in generalterms and to use this knowledge in the field of a.c. motorcontrols to achieve from a soft start to a very precisespeed control and, more importantly, to conserve theenergy of the machine which would be wasted otherwise.For more details of static controllers see the Further reading(Sr. nos. 2, 4. 5, 8 and 12) at the end of the chapter. To

www.taghedanesh.ir6/112 Industrial Power Engineering and Applications Handbookbring more clarity to the subject, passing references to ad.c. machine are also provided. To make the discussionof static drives more complete different configurationsof converter units are also discussed for the control ofd.c. motors.In the past five decades solid-state devices such asdiodes, transistors and thyristors have attained a remarkablestatus and application in the field of electronic powerengineering. Diodes and thyristors were introduced inthe late 1950s, while the basic transistor (BJT -bipolarjunction triode) was introduced in 1948. In India theyappeared much later (thyristors were introduced in the1970s and power transistors in the 1980s). This technologyis now extensively applied to convert a fixed a.c. powersupply system to a variable ax. supply system, which inturn is utilized to perform a required variable duty of afixed power system or a machine. They are all semiconductorswitching devices and constitute two basicfamilies, one of transistors and the other of thyristors.The more prevalent so far is discussed here to give readersan idea of the use of this technology in today’s domesticand industrial applications, power generation, distributionand their controls etc.More emphasis is provided on the control of inductionmotors. Research and development in this field is acontinuous process and is being carried out by agenciesand leading manufacturers. This aims to advance andoptimize the utility of such devices by improving theircurrent-handling capability and making them suitablefor higher system voltages, switching speeds etc. Theremay be more advanced versions available by the timethis book is published and readers should contact theleading manufacturers for details of the latest technology.Diodes are purely static power switching devices andare used extensively with thyristor and transistor powerschemes. Transistors are relatively cheaper and easy tohandle compared to thyristors. The latter are moreexpensive and more complex as noted below. This textdeals with the application of such devices for the controlof induction motors, which can now be employed toperform variable duties through its stepless speed controlby close monitoring of load requirements during aparticular process or while performing a specific dutycycle. The controls are assisted by microprocessor-based,open- or closed-loop control techniques, which can senseand monitor many variables such as speed, flow ofmaterial, temperature, pressure or parameters important€or a process or a duty cycle. With these techniques, it ispossible to achieve any level of automation. Open-loopsystems are used where high accuracy of controls andfeedback is not so important and closed-loop where ahigh degree of accuracy of control is essential. Withsolid-state technology it is now possible to utilize aconventional machine to perform a variable duty.Transistors so far have been developed to handle currentsup to 2000 A and voltages up to 1200 V and are utilizedfor low-capacity power requirements. Thyristors havebeen developed up to 3000 A and voltages up to 10 kVand are employed for large power requirements such asHV d.c. transmission and static VAr controls. With thevariety of such devices and their number of combinations,it is possible to achieve any required output variation inV and f of the fixed a.c. input supply or I, and I, (phasorcontrol) in the machine’s parameters and use these toperform a required duty cycle with very precise speedcontrol.6.7.1 Power diodesThese are unidirectional* and uncontrollable? staticelectronic devices and used as static switches and shownin Figure 6.14. A diode turns ON at the instant it becomesforward biased and OFF when it becomes reverse biased.By connecting them in series parallel combinations, theycan be made suitable for any desired voltage and currentratings. Whether it is a transistor scheme or a thyristorscheme, they are used extensively where a forwardconduction alone is necessary and the scheme calls foronly a simple switching, without any control over theswitching operation. They areusedextensively inarectifiercircuit to convert a fixed a.c. supply to a fixed d.c. supply.ilA (anode)K(cathode)Figure 6.14 Circuit symbol for a power diode as a switch6.7.2 The power transistor familyThe solid-state technology in the field of transistors inparticular has undergone a sea-change, beginning in the1950s from the basic bipolar junction transistor (BJT) tothe more advanced insulated gate bipolar transistor (IGBT)by the 1990s. The following are some of the moreprominent of the power transistor family that are commonlyused in power circuits.Bipolar junction transistors (BJTs)These are the basic transistors (triodes) and are illustratedin Figure 6.15. They are unidirectional and controllableand are capable of handling large currents and highvoltages and also possess high switching speeds (fasterthan thyristors). However, they require a high base currentdue to the high voltage drop across the device, whichcauses a high loss and dissipation of heat. This adversefeature of their characteristics renders them unsuitableas power switching devices for efficient power conversion.Therefore they are generally used as electronic controldevices rather than power devices in electronic controlcircuits and are not produced at higher ratings.Two-junction transistors or power DarlingtonsThese also have three terminals as illustrated in Figure*A unidirectional switch is one that can conduct in only one directionand blocks in the reverse direction.tA controllable switch is one that can be turned ON and OFF byswitching a control circuit ON and OFF.

www.taghedanesh.irY"A npnjunctionPCCircuit symbolsNectrical representation6 baseE emitterC collector(Note A diode has only one p n junction)P"A PnPjunctionFigure 6.15 Circuit symbols and electrical representation of abasic triode or power transistor (BJT)6.16. They are fabricated of two power transistors andare used as a single transistor and are suitable only forcontrol circuits. They are used to reduce the control currentrequirement and hence cause lesser heat dissipation,particularly during switching operations.MOSFETsThese are metal oxide semiconductor field effectivetransistors and are shown in Figure 6.17. They are capableof switching quickly (but are more sluggish than BJTs)and handle higher switching frequencies. But they candeal with only lower currents and withstand lower voltagesand thus possess a low power-handling capability. Theyare bi-directional and can operate as controlled switchesin the forward direction and uncontrolled switches in thereverse direction. MOSFE'Ts are composed of a diodeand a BJT or IGBT in anti-parallel. They are voltagecontrolleddevices and require a negligible base currentat their control terminals to maintain the ON state andhence are low-loss devices. MOSFETs are used extensivelyPCStatic controls and braking of motors 6/113as fully controllable power switches, where they arerequired to handle only low powers. The latest trend isto use them only as control devices.The power BJTs and power MOSFETs have providedtwo very useful static switching devices in the field oftransistor technology. But while the former can handlelarger powers, they dissipate high heat. the latter poses alimitation in handling large powers. As a result of theseshortcomings, they are used mostly as control circuitswitching devices. Such limitations were overcome byyet another development in the 1990s, in the form of anIGHT.Znszilated gate hipolar transistors (ZGBTs)These are unidirectional transistors and have an insulatedgate (G) instead of the base (R) as in a bipolar transistor(RJT) and are represented in Figure 6.18. They are ahybrid combination of a przp bipolar transistor which isconnected to a power MOSFET like a two-junctiontransistor (power Darlington, Figure 6.16). A positivevoltage between the gate and the emitter switches ONthe MOSFET and provides a low resistance effect betweenthe base and the collector of a pnp bipolar transistor andswitches this ON as well. The combination of twotransistors offers an insulated gate that requires a lowbase current which makes it a low-loss device. When thevoltage between the gate and the emitter is reduced tozero, the MOSFET switches OFF and cuts off the basecurrent to the bipolar transistor to switch that OFF aswell. With slight modification in construction andupgrading the bipolar and MOSFET transistors. it ispossible to produce a low-loss IGBT. suitable for fastswitching, handling large currents and withstanding highervoltages. Present ratings have been achieved up to 1600V and 2000 A, but ratings up to 600A are preferred. dueto their easy handling and making power connections. Inhigher ratings, because of their small size, they maypose a problem in making adequate power connections.proper handling, thermal stability, adequate protectionetc.The development of this hybrid combination in a bipolartransistor has greatly enhanced the application of powertransistors in the field of power conversion and variablespeeddrives. It possesses the qualities of both the powerbipolar transistor (RJT) and the power MOSFET. Like apower MOSFET, it is a voltage-controlled switching device(kind of a base)ODrain P"JL'a SourceFigure 6.16 Circuit symbol for a twojunctiontransistor or power DarlingtonFigure 6.17 Circuit symbol for apower MOSFETFigure 6.18IGBTCircuit symbol for an

www.taghedanesh.ir6/114 Industrial Power Engineering and Applications Handbookthat permits fast switching and gate voltage control andas a bipolar transistor it allows a large power handlingcapability. The switching speed of the IGBT is also higherthan that of a bipolar transistor. It thus provides an efficientpower conversion system. With gradual and consistentdevelopment of their design, it has been possible to achievehigher ratings of these devices. Presently single IGBTunits have been used to handle power up to 650 kW orso. Their series-parallel combination, that initially createdlimitations as a result of their complex switching, is alsobeing overcome and it is hoped that much larger ratingsfrom such devices will be possible in the near future.They are used extensively in an inverter circuit to converta fixed d.c. supply to a variable ax. supply. Since theyare more expensive compared to power diodes, they arenot used generally in a rectifier circuit where power diodesare mostly used. However, when the power is to be fedback to the source of supply, then they are used in therectifier circuit to adjust V and fof the feedback supplyto that of the source.With this development, thyristor technology is nowbeing applied to handling large to very large powerrequirements, where there is no option but to use thyristorsalone, for example, for very large motors, reactive powercontrols etc., as discussed in Section 24.10. A typicalinverter circuit using IGBTs is illustrated in Figure 6.19.In addition to being suitable for high switching frequencies,these transistors virtually retain the sinusoidal waveformof the motor currents. The motor current now containslesser harmonics, and causes less heating of the motorwindings. It also causes less pulsation in torque and lowmotor noise. The motor also runs smoother, even at lowerspeeds. Now V and f can both be varied through thissingle device and a fixed voltage power diode bridgerectifier is sufficient to obtain a fixed d.c. voltage, ratherthan to use a phase-controlled thyristor rectifier to obtaina variable d.c. voltage.However, they may generate switching surges. Althoughmoderate, they have caused failure of motor insulationin some cases. Depending upon the type of installation,a surge protection, in the form of dvldt protection throughchokes, may become mandatory with such drives, particularlywhen the cable length from the drive to the motoris too long or when the motor is rather old and may notpossess a sound dielectric strength. More details of thisaspect are discussed in Section 6.14.MOS controlled thyristors (MCTs)The latest in the field of static devices are MOS-controlledthyristors (MCTs), which are a hybrid of MOSFETs andthyristors. There is yet another device developed in thisfield, i.e. insulated gate-controlled thyristors (IGCTs).Implementation of these devices in the field of staticdrives is in the offing.6.7.3 The thyristor familyThe thyristor is a semiconductor device made of germaniumor silicon wafers and comprises three or morejunctions, which can be switched from the OFF state tothe ON state or vice versa. Basically it is apnpn junction,as shown in Figure 6.20(a) and can be considered ascomposed of two transistors with npn andpnp junctions,as illustrated in Figure 6.20(b). It does not turn ON whenit is forward biased, unlike a diode, unless there is a gatefiring pulse. Thyristors are forced commutated (a techniquetransistor Power q*iesistor dynamic forfor brakingbraking* Uncontrolled line side diode bridge rectifier. When a variable d.c. is required, it can be replaced by thyristors.*' Mechanical braking or non-regenerative braking:For small brake power, resistance unit is small and can be located within the main enclosure. But for higher power thatmay call for large resistance units and have to dissipate excessive heat, it is mounted as a separate unit. Theresistance units are short-time rated depending upon the duty they have to perform.Figure 6.19A typical inverter circuit using IGBTs, also showing a feedback unit

+www.taghedanesh.irAnodeg GateCathode(a) ThyristorPAG 9N KO AStatic controls and braking of motors 6/115PA(b) Equivalent 2-transistor representationFigure 6.20 A basic thyristor [silicon controlled rectifier (SCR)]for switching OFF) and hence call for a complex circuitry,more so in a 3-4 system, where six of them have tooperatc simultaneously. The control device has to be veryaccurate to trigger all the thyristor devices at the sameinstant and a slight error in the firing angle may cause ashort-circuit, whereas a transistor can be switched OFFsimply by removing the base signal. Thyristors aretherefore also known as phase-controlled rectifiers. Thephase angle delay is known as the firing angle of theswitching element. In this book we have denoted thisangle by a. For diodes a = 0, while for thyristors it canbe adjusted as illustrated in Figure 6.23. We will not gointo more detail on the construction of this device andwill limit our discussion only to its application in a powersystem. The device constitutes a large family, but onlythe more prevalent of them are discussed here, i.e.Silicon-controlled rectifiers (SCR). These are basicallythyristors and unless specified, a thyristor will meanan SCRTriacsGate turn-off switches (GTO)MOS-controlled thyristors (MCTs) and insulated gatecontrolledthyristors (IGCTs) (discussed in Section6.7.2).Silicon-controlled rectifiers (SCRs)The most popular of the thyristor family is the SCR,which was first developed in 1957 by General Electric,USA. The SCR is similar in construction to that of ajunction diode, except that it has three junctions insteadof one, and a gate to control the flow of power. The SCRis commonly represented as shown in Figure 6.20(a) andcan be regarded as a semiconductor switch, similar to atoggle switch. An SCR is unidirectional and conducts inone direction only and can also be termed a reverseblocking triode thyristor. When anodc A is positive withrespect to K, it is in the conducting mode and is termedforward biased. The V-I characteristics are similar tothose of a semiconductor diode, as shown in Figure 6.21.When K is positive with respect to A, it is in the nonconductingmode and conducts a very low leakage current.In this mode it is termed reverse biased. In this condition,when the reverse voltage is raised a state is reached whenthe low-leakage reverse current increases rapidly as aresult of the dielectric breakdown. This stage is calledthe reverse avalanche region (Figure 6.2 1 ) and may destroythe device.When a load and a power source is connected acrossthe anode and the cathode of the SCR, there will be noconduction and no current will flow, even when the anodeis made positive with respect to the cathode unless thegate is also made forward biased with the application ofa positive potential at the gate. After the conductioncommences, the gate potential can be removed and the* /-- Reverse voltage (V)i regionfFigure 6.21gate voltageP5f1\Reverse blocking Ea,-regionNormal operation(“ON” state)Forwardblocking regionL ,’J > +4 F0 Forward voltage (V)V-/ characteristics of a thyristor (SCR) without a

www.taghedanesh.ir6/116 Industrial Power Engineering and Applications Handbookdevice will continue to conduct. It is the gate signal thatplays the most vital role in achieving the desired variationin the voltage. The main power connections to the deviceare made to its terminals A and K and a turn-on signal isapplied between the gate and K. An SCR can easilyprovide a variable voltage source by varying its firingangle. In view of its simplicity, it is the most commonlyused thyristor in a phase-controlled rectifier unit (converter).Gate control is now simple, as it is connected onthe a.c. or the line side, which provides it with a naturalcommutation. The thyristor gets switched OFF at everycurrent zero. This may therefore also be termed a linecommutated rectifier.The use of SCRs in an inverter circuit is intricate becauseof the absence of a natural commutation. Now only aforced commutation is possible, as it is connected to ad.c. source which provides no current zeros and hencefacilitates no natural commutation. A forced commutationcalls for a separate switching circuit, which is cumbersome,besides adding to the cost. As a result of this feature,they are also called forced commutated thyristors.TriacsUnlike an SCR, which is unidirectional, a triac is abidirectional thyristor switch and conducts in bothdirections. It can be considered as composed of two SCRs,connected back to back with a single gate, as shown inFigure 6.22(a). Since the thyristor now conducts in bothdirections there is no positive (anode) or negative (cathode)terminals.The triac may, however, have some limitations inhandling frequencies higher than normal. In such cases,they can be simulated by using two SCRs in inverseparallel combinations as illustrated in Figure 6.22(b).Now it is known as a reverse conducting thyristor. AnSCR has no frequency limitations at least up to ten timesthe normal. The required voltage and current ratings areobtained by series-parallel connections of more than onethyristor unit,iA I gFigure 6.22(a) Schematic Figure 6.22(b) Use of tworepresentation of a triac SCRs in inverse parallelcombination to simulate atriac (reverse-conductingthyristors)Gate turn-off switches (GTOs)The gate can only turn the thyristor ON but it cannot turnit OFF (commutate). Switching OFF can be accomplishedonly by reducing the conducting current to less than thethyristor's holding current. A device that allows the gateto switch OFF is called the gate turn-off switch (GTO)or gate control switch (GCS). The GTO turns it OFF byfiring (applying) a negative potential between the gateand the cathode. It is the most commonly used device ina thyristor inverter circuit.6.8 Conduction and commutationA thyristor can be turned ON by the gate at any angle a,with respect to the applied voltage waveform as shownin Figure 6.23(a) and (b) for half-wave and full-wavecontrolled rectifiers respectively. By varying the firingangle, which is possible through the firing circuit, thed.c. output voltage through a converter circuit can bevaried, as illustrated in the figure. The voltage is full(maximum) when the firing angle is zero. Now theconduction angle is 180". As the firing angle increases,the conduction angle decreases and so does the outputvoltage. The output voltage becomes zero when the firingangle becomes 180" and the conduction angle becomeszero. Thus the conduction, i.e. the power through athyristor, can be varied linearly by varying the gate voltageand its firing angle. Such control is termed phase control,and a rectifier or converter unit, employed to convert a.c.to a variable d.c., is called a controlled rectifier or acontrolled converter.In thyristor technology the switching OFF of a thyristoris conventionally termed commutation. In a.c. circuits,when the current through a thyristor passes through itsnatural zero, a reverse voltage appears automatically andturns OFF the thyristor. This is a natural commutation.No external circuit is now required to turn OFF thethyristor. They are therefore commonly called linecommuting thyristors, like those used for a.c.-d.c.converters. But this is not so in d.c. circuits, as the currentwave now does not pass through a natural zero. Theforward current can now be forced to zero only throughsome external circuit to turn the thyristor OFF. This istermed forced commutation, such as when used for d.c.-a.c. inverters. Unlike a thyristor, a transistor can beswitched OFF simply by removing the base signal and itrequires no separate circuit to switch it OFF. In Table 6.3we show a brief comparison between transistor devicesand the basic thyristor (SCR). The triacs and GTOs andother thyristor devices fall in the same family, but withdifferent features of conduction and commutation to suitdifferent switching schemes. Other importantcharacteristics are also shown in the table.6.9 Circuit configurations ofsemiconductor devicesSemiconductor devices, as noted above, are used widely

www.taghedanesh.irStatic controls and braking of motors 611 173a= 0"0OL = 30"0a= 60'0u= 90'0a= 180"v= 0Magnitude of rectifiedvoltage can be variedi'60"A,!+-L1/J/90'(Y -the firing angle@ 3-0 half wave controlledrectifier. (Current flowunidirectional)'@ 0a=30 , n=60 , a=90", a= 18030" I 60" 1 v= 030 60" 90"Magnitude of rectifiedvoltage can be variedn the firing angle@ 3-6full wave controlledrectifier (Current flow inboth directions)Figure 6.23Phase control of voltage through different gate firing anglesto convert a fixed a.c. supply to a fixed or variable d.c.supply, and then from a fixed d.c. supply to a variableax. supply. for example, for variable ax. drives. A variabled.c. supply is used for the variable d.c. drives. Theconventional nomenclature to identify the various typesof these circuits and their applications isConberter or rectifier unitInverter unit6.9.1 Converter or rectifier unitA converter unit is used for the control of d.c. machinesand also to provide a d.c. source to an inverter unitcontrolling an ax. machine. In d.c. drives the d.c. voltageafter the converter unit should be variable, whereas foran a.c. drive It is kept fixed. The voltage is varied by theinverter unit. A converter unit is the basic power conversionscheme to convert an a.c. supply to a d.c. supply.Conventionally they are also known as rectifier unitsand can be arranged in four different modes to suit differentapplications of a motor as follows:I Uncontrolled rectifier- units These can beHalf wave similar to Figure 6.24(a) configurations (b)and (c), using one diode per phase instead ofa thyristor.orFull wave similar to Figure 6.24(a) configuration (a),using two diodes in anti-parallel per phase instead ofthyristors or in the form of a centre-point configuration.2 Controlled rectifier umrr These can also beHalf wave (Figure 6.24(a)). configurations (b) and (c)using one thyristor per phase orFull wave (Figure 6.24(a)), configuration (a), usingtwo thyristors in anti-parallel per phase or in the formof a centre-point configuration.Note A half-wave rectifier is a single-bridge rectifier and is witablefor only single-quadrant operations I or 111. A full-wa\~ rectifiei- isa double-bridge rectifier and suitable for multi-quadrant operationsparticularly quadrant? I1 and IV. See Table 6.3.

~~~~~~~~ ..L-l~ ~__~ .-www.taghedanesh.irTable 6.3 A brief comparison between a transistor and basic thyristor technologyParametersTransistor technology 1 Thyristor technology1 As a static switchUnidirectiorJBidirectional-ControllableJUncontrollableIGBTJuncontrollable in reverseTRIAC- Reverse conductingthyristorJ-r‘JJControllable inboth directionsControllable inboth directions2 Switch OFFcharacteristics3 Conrrols4 For conduction irboth direction55 Switchingfrequency6 Rating7 To vary V andf-8 Heating effect______9 Cost factor~~~~This calls for no switching OFF circuitry, since a transistor can be switched OFFsimply by removing the base signal.These require only the base signal to switch ON. Thus provide a simpler technology.Generally two circuits are usedPower MOSFETs and IGBTs can handle much higher switching frequencies,compared to a thyristor. In an a.c. motor control, fast switching is mandatory andtherefore transistors are preferred.(a) Can handle only moderate currents and voltages. A BJT is used mostly inelectronic control circuits. As they are small, hundreds can be placed on a smallPCB (printed circuit board). Some manufacturers, however, also use them for thecontrol of small motors, say, up to 1615 HP.(b) MOSFETs and IGBTs alone are used for power applications. Rating ofsingle-piece IGBT is possible up to 650 kW after considering all possible deratings.Typical V and I ratings for a single unit achieved so far areV= 1600Va I=2000AaOnce fired, a thyristor cannot be controlled. It requires a forced commutation to switch itoff and the gate control is quite cumbersome. To switch OFF, the conductinp > rnrrent is --reduced to less than its holding current. The commutation circuitry i\ therefore highlycomplex and also influences the reliability~~~ of a thyristor application.There are six thyristor firing circuits required for a 3$ system, as there are two thyristorsconnected back to back each phase. The whole scheme is therefore complex and lessreliable.These can be connected in anti-parallelVery low switching frequency, but a GTO is suitable for frequent switchings.~Can handle much larger powers Typical V and I ratings for each unit dchieved so far dreV= 10kVaI = 3000 A”a Subject to applicable deratingsThese ratingsvcan be enhanced bv connecting them in series-parallel combinations. In series to enhance the voltage rating and in parallel to enhance the current rating. But theLcontrols may not be so accurate as with a single device.Both V andfcan be varied with the help of pulse width modulation (PWM) in theinverter circuit. The converter unit normally is an uncontrolled power diode rectifier.Low heat dissipation due to low voltage drop across the device (up to I V only)Much more economical compared to a thyristor drive in this range_____~By varying the gate firing angle, V can also be varied. With SCRs frequency variation isnot possible.Note Since an inverter is not line commutated, the SCRs have a switching limitation andhence a limitation in frequency variation. When thyristors are to perform frequent qwitchings,GTOs are used in the inverter circuit.This device has a high voltage drop across it (up to 3 V) and therefore generates excessiveheat and poses a cooling problem, particularly for large systems, which may even call foran external cooling arrangement.In smaller ratings they are economical. In an a.c. to d.c. converter, for instance, for thecontrol of a d.c. motor, where a variable d.c. voltage is desirable, SCRs are used extensivelyand the voltage variation is obtained by varying the firing angle. Since the SCRs are nowline commutated, they pose no switching OFF problem.

www.taghedanesh.irStatic controls and braking of motors 6/119The uncontrolled rectifier units are simple rectifiersand use power diodes universally. The rectificationobtained is uncontrollable and provides only a fixed d.c.voltage output from a fixed a.e. supply. Thc diodes haveno control over their switching instants. These rectifierunits are used extensively to provide a fixed d.c. sourceto an inverter unit when being employed to control ana.c. machine. Since there is no switching of diodesinvolved, there arc no voltage surges across the diodes.There is thus no need to provide a snubber circuit acrossthe diodes to protect them asainst such surges, as discussedlater.When, however. a variable d.c. voltage output isrequired, controlled rectifiers are used. Now the diodesare replaced by one or two phase-controlled thyristors(SCRs) in each phase, one thyristor per phase for a halfwave and two per phase in anti-parallel for a full waverectification. The required voltage variation is achievedby adjusting the delay time of the gate-firing pulse (a) toeach thyristor unit. Thyristor circuits are possible up toany rating by arranging them in series-parallelcombinations. With the advances in this technology, it ispossible to achieve a total matching of switching sequencesof all thc thyristors to ensure an accurate and fully cohesiveoperation (switching of all thyristors at the same instant).Unlike diode, a thyristor does not turn ON automaticallyat the instant it becomes forward biased but requires agate pulsc at its gate terminal to switch it OK. This isobtained through a control circuit which is a part of therectifier unit. The gate-firing pulse is provided at theappropriate instant to each thyristor in each switchingcycle. positivc half to negative half, to obtain the requiredvoltage. The switching is automatic as the thyristors areline commutated and at each half cycle the voltagewaveform passes through its natural zero. The instantsmay bc delayed from 0" to 180" electrical to obtain therequired infinite control in the output supply, as illustratedin Figure 6.23(b). The d.c. output is controlled by adjustingthe delay time of the gate-firing pulse. A few voltagewaveforms are illustrated in Figure 6.2Xb) at differentfiring angles. Phase control of positive and negative halfwaves of each phase can be infinitely varied to meet anypower demand.A half-wave rectifier is able to provide only a unidirectionald.c. power source which may also containmany a.c. ripples (Figure 6.24(a)). A full-wave rectifieris employed to reduce such ripples, on the one hand, andprovide a d.c. power in forward as well as reversedirections, on the other. A fixed forward and reverse d.c.power is required for an inverter unit when it is employedto control an a.c. machine. Now an uncontrolled rectifierunit is adequate as V and ,f control is obtained throughthe inverter unit.A controlled rectifier unit is necessary when it has tocontrol a dc. machine. which would call for a variabled.c. voltage. When the d.c. machine has to opcrate inonly one direction (quadrants I or 111) a half-wavecontrolled rectifier will be adequate and when the machinehas to operate in either direction. a full-wave controlledrectifier will be essential.For operations in quadrants I1 and IV it is essential tohave an unrestricted flow of reverse power and hence anadditional feedback inverter unit would alw he es\ential.as shown in Figures 6.31-6.33. depending upon theconfiguration of the converter unit.Now It is possible that at some locations there is no ax. sourceavailable. such as for battery-operuted lifts and motor vehicles.Such applications may also call for a variable d.c. source. When itis so. it can be achieved with the use of a chopper circuit whichuses the conventional semiconductor devices. The devices areswitched at high repetitive frequencies to obtain the required variationin the output voltage as with the use of a phase-controlled thyristorrectifier. A typical chopper circuit is shown in Figure 6.15. usinpdiodes and a controlled unidirectional semiconductor switch. whichcan be a thyristor or an ICIRT.Relevance of different quadrants of aconverter unitDepending upon the mode of operation of a motor, thetype of converter unit can be decided. For simplicity, theoperation (conduction) of a motor can be reprewnted byfour quadrants as illustrated in T;lble 6.4.Quadrant I Both V and Tare positive. The machinecan be run only in one direction (say. forward). Brakingoperations are possible. It is a converter modeand either a half-wave or a full-wave rectifier can beused.Quadrant I1 Now T is in the reverse direction andthe machine can be run in the reverse direction. Brakingand regeneration are possible. For regeneration anadditional bridge will be essential as discussed later.For current to flow in either direction. a full-waverectifier will also he essential.Quadrant 111 Now both the voltage and the torqueare in the reverse direction otherwise it is similar toQuadrant I. The machine can now be run in the reversedirection.Quadrant IV Now the voltage alone is in the reversedirection and the machine can be run in the forwarddirection. Braking and regeneration are possible. Themachine in the forward direction can generate, whenit is brought down from a higher speed to a lowerspeed or when feeding a load going down hill etc. Forregenera-tion. an additional bridge will be essential asnoted above, and for the current to flow in eitherdirection, a full-wave rectifier.The changeover from motoring to regeneration, 1.e.from Quadrant I to Quadrant 1V or from Quadrant I11 toQuadrant I1 or vice versa. is achieved by first bringingthe torque to zero by ceasing the firing of one bridge andcommencing that of the other.6.9.2 Inverter unitThe purpose of an inverter unit is to invert a fixed d.c.power to a variable ax. power which can be achieved intwo ways:1 Using IGBT devices These are the latest in the fieldof static power control. They are easy to handle and

www.taghedanesh.ir6/120 Industrial Power Engineering and Applications HandbookThyristorconfigurationComparison of a few common thyristor rectifier CIWinding arrangement and thyristor @rectifier configurationderter circuitsNo. of phases (pulse number,n and voltage phasor)(a) Bi-phase(Single phasecentre point)(Full waverectifier)i(b) 3-phase,(Half waverectifier)1:l rn=33@ neutral pointn=6Notes (1) Considering the firing angle K = 0, the values are same for power diodes also.(2) (i) This ratio is true for power diodes. Or thyristors with a= 0,(ii) V, is considered as the phase voltage of each transformer limb. Even Vph/2 is considered as V, for simplicity.(iii) To derive this equation refer to Vithayathil (1995).(3) kdc(h) - d.c. output with ripplesI

www.taghedanesh.irStatic controls and braking of motors 6/121Voltage and currentwave formsfora=Od.c. output -:yL~y a.c voltagevoltageripplesvdc(h)d.c. outputvoltage after)'I, I ' ,I$ smoothingI ' ,I \\ iwaveform I aF uIa b-1- 2d2 cCurrent output waveform. Eachphase conducting for 2d2 = 180"j,c. output voltageI 1 vbI IIIh2x13Current output waveform eachphase conducting for 21d3 = 120Input voltageV, I waveform.Output voltage withfiring angle a42.nV& = -x v,, sin E cos a@3e cos a3 5 cos aCunent on 1 a.c'urrentpeak=bbc- Idc6?atio of d.c.to a.caltagewa=00.91.171.35(4)(5)(6)IC - d.c. output with minimal ripplesThe six secondaly phases are obtained by shorting the centre points of each of the three-phase windings of a 36 transformer secondary.For use of L and C refer to Figure 6.34.Figure 6.24(a) A few configurations of controlled rectifier units (for uncontrolled rectifier units the thyristors (SCRs) are replacedwith diodes)

www.taghedanesh.ir6/122 Industrial Power Engineering and Applications HandbookRear viewFront viewFigure 6.24(b)Thyristor cubicles for high-power static converter units (Courtesy: Siemens)I/CACSupplyLSI and S, are staticunidirectional switches,which may be an IGBTor a thyristor.Figure 6.25A typical bi-phase chopper schemecontrol, besides being inexpensive in the presentlydeveloped ratings compared to GTOs. A fixed d.c. voltageis obtained through a power diode converter. Theconversion from a fixed a.c. supply to a variable a.c.supply is thus cheap and easy to handle. A completesystem composed of a converter and an inverter unit, asused to control an a.c. motor, is more commonly calledan inverter. In Figure 6.26(a) we show a basic IGBT orthyristor (GTO) inverter unit. Single IGBTs developedso far can handle machines up to 650 kW. There aresome deratings of an IGBT on account of a lower r.m.s.value of the inverted power, compared to the power inputto the inverter, the configuration of the inverter (whichdefines the inverted waveform), the ambient temperature,

dri\~ ireversewhen~ halfconverter~ forward~ an~ onlywhen~ whenconverterwww.taghedanesh.irStatic controls and braking of motors 61123Table 6.4 Operation of a motor in different modes and the corresponding conducting quadrants of a controlled converter unitQuadrant I1Operation\ ~('onli~uratiotiMock (it' opelation ~ing and braking (revme motoring)regenerationN7)T- onl) a full wabe- an additional bridge for regenerationmotoring ~when regenerating - inverterQuadrant IllOperations - dri\ tiig and braking (reverse motoring)Configurationwave or full waveMode of nprratim - cun\t'itetQuadrant IOperations - driving and braking (forward motoring)Configuration - half wave or full waveMode of operation - converterQuadrant IVOperationsConfigurationMode of operation ~3 T)+T- driving and braking (forward motot-ing)regenerationa full waveadditional bridge for regenerationmotoring ~regeneration - inverterKe\,erse-running quadrantsForward-running quadrant\-Vsafet) margins etc. Their overcurrent capacity is definedby the overload current and its duration. IEC 60146-1 -1has defined it as 150-300%. depending upon theapplication. for a duration of one minute. The manufacturercan derate a device for a required load cycle (overloadingand its duration) according to the application. IJnlessspecified. the present normal practice is to produce suchdevices for an overcurrent of 150% for one minute. Ahimher starting current than this or a start longer than onePminute may call for a higher derating.ICBTs can be used for still higher ratings (> 650 kW)by connecting them in series-parallel combination as notedearlier. but for higher ratings, thyristors (GTOs) arenormally preferred for better reliability. More than oneIGBT in series-parallel combination may sometimes acterratically and perform inconsistently. They are, however,being used for higher ratings also in the light of thetechnological advances in this field.2 C/.iirig rhyvistor- dc,ixic.e.v (GTOs) Thyristor (GTO)inverter circuits are used for higher ratings of machinesthan above and to control larger powers such as for thosefor reactivc power control.To obtain variable V and J'In IGBT\ thr-oiigli p rl vr bvidtli rnodulatiorz (PWM) Thefrequency in the inverter circuit is varied by frequentlyswitching the IGBTs ON and OFF in each half cycle.While the voltage is controlled with the help of pulsewidth modulation, which is a technique for varying theduty cycle or the zeros of the inverter output voltagepulses. The duty cycle or CDF (cyclic duration factor) ofthe pulses is the ratio of the period of actual conductionin one half cycle to the total period of one half cycle.For Figure 6.27(a)CDF =t, f t, 4- t? + tJ -k t, i t,T(6.3)where t,, t2, ... t6 are the pulse widths in one half cycle.If V is the amplitude of the output voltage pulses. thenthe r.m.s. value of the output a.c. voltageor V, = V . ,/CDF (6.4)By varying the CDF, Le. the pulse widths of the ax.output voltage waveform, the output, V,,, , can be varied.The CDF can be controlled by controlling the periodof conduction, in other words, the pulse widths (periodictime period, Tremaining the same). Thus the a.c. outputvoltage in an IGBT inverter can be controlled with thchelp of modulation. The modulation in the inverter circuitis achieved by superposing a carrier voltage waveform

www.taghedanesh.ir6/124 Industrial Power Engineering and Applications HandbookDedicated transformerIAC 47"supply 4% 3Rectifier or converter unitConverts a.c to d.c. and can be a ower diode fixed voltage for a.c. drives (phase controlled thyristor converter for d.c. drives)D.C.link circuit connectingat06Inverter unitInverts d.c. to variable a.c. and can be an IGBT or a thyristor (GTO) circuit.Variable Vwith an VSI (Figure 6.28(a)) or variable I with a CSI (Figure 6.29), and variable fax. power outputInductor L(i) It is necessary to smooth a.c. ripples, whether it is a power diode or a thyristor rectifier. It also suppresses harmonics when, @is acontrolled rectifier, producing a.c. rip les and harmonics(ii) Can be a large size inductor, when 6 is a current source inverter (CSI) (Figure 6.29)(iii) Provides a short-circuit protection for a fault in d.c. link, by adding to its impedanceCharging capacitor, to hold the charge, by smoothing the output ripples and providing a near constant voltage source to the inverter circuit,when it is a voltage source inverter. When the converter is a thyristor converter, a resistance R is also provided with C to make it suitable toperform its duty under frequent thyristor switchings, by quickly discharging it through R. Now it becomes a snubber circuit, to also protectthe inverter devices from duldt.(a) Current limiting reactor on I/C side, to control dildtduring switching of thyristor units when @is a phase controlled rectifier. Notnecessary when the unit is supplied through a dedicated transformer.(b) Also required to limit dildt to the solid-state circuits when the source of supply is large and is protected by current limiting device(Section 6.14, Figure 6.35).Inverter unit (conventional name). Converts fixed ax. to variable ax.Figure 6.26(a)Basic IGBT or thyristor (GTO) inverter unitof much higher frequency on the natural voltage waveform.Figure 6.27(b) is a simple block diagram for a PWMscheme, the natural voltage being the voltage obtainedby the switching of the IGBTs. The carrier wave can beof any shape, the frequency of which is altered, to obtainthe required degree of modulation, and hence the voltage,while the amplitude is kept fixed. The amplitude is amatter of scheme design. (For more detail refer to thetextbooks in the Further reading.) Generally, a triangularwave is used as shown in Figure 6.27(b) to obtain amore uniform sinusoidal voltage waveform. By Fourieranalysis we can establish the amplitude of voltage andquality of waveform (distortions), and by controlling thepulse widths through the frequency of the carrier wave,we can decide the best modulation to obtain the requiredamplitude and a near-sinusoidal output voltage waveform.(For details of Fourier analysis, refer to a textbook.)This is the most commonly used technique in the invertercircuit to obtain the required Vlf pattern. It is alsoeconomical and can be used to control multi-motor drivesthrough a single unit. Since the variation is based onvoltage, the inverter may be called a voltage source inverter(VSI). To obtain an accurate Vlfcontrol, it is essentialthat the voltage is maintained uniform (without ripples)as much as possible. This can be achieved by providinga capacitor across the d.c. link as shown in Figure 6.26(a).The purpose of the capacitor is to hold the charge andsmooth the output a.c. ripples of each diode and henceprovide a near-uniform d.c. voltage. The charge retainedby a capacitor can be expressed byFigure 6.26(b) A small rating IGBT inverter unit(Courtesy: Kirloskar Electric)Q=Cdudt

www.taghedanesh.irPulse widthsStatic controls and braking of motors 61125AC outputvoltage pulsesAC inputinstantaneousvoltagewaveformC D ztl+fP+t3+t4+t5ft6~rIand V,,, = V,iC D FFigure 6.27(a) Varying the output a c voltage with PWM technique@)* FiringModulatorfIcircuitReference voltageInverter natural voltage waveform before modulation improvedto a near sinusoidal waveform, with the use of Land C@ Carrier voltageTriangular voltage waveform of fixed amplitude@ Variable frequency and modulated voltage output (Vif ) as desiredFigure 6.27(b)Block diagram for a PWM schemewhere Q = charge stored by the capacitor unitC = capacitance of the capacitordu- = rate of voltage change or a.c. ripples in thedtd.c. linkThe higher the value of C, the lower will be the voltageovershoots in the rectified voltage and the inverter circuitwould be fed by an almost constant voltage source. Thecapacitor in the circuit also provides an indirect protectionfrom the voltage surges.The above method is used to vary the frequency andthe voltage ot the inverter output (motor side) accordingto the process needs, irrespective of the electronics schemeadopted to obtain the required speed control.,Vrur The variation of frequency is generally up to its fundamentalvalue. ].e. 0-50 or 0-60 Hr. in vicw of the fact that the motor isgenerally required to operate below the base speed. At higherfrecpencie\ the motor \n ill overspeed. for v hich its own suitab~lityai wcll as the suitability of the mechanical sy\tein must he prechecked.When. however, such a situation be desirable. the frequencymay be varied to the desired lebel by twitching. kccping the outputvoltage to the rated value. Since the torque of the motor will nowreduce a IIN, this must be checked with the load requirement.In GTOsThe frequency in the inverter circuit is varied by switchingthe GTO pairs ON and OFF repeatedly through theirgate control in each one half cycle. The rate of frequencyvariation will depend upon the frequency of switchingof the GTO pairs. The voltage variation is obtained byvarying the gate firing angle, a.By using converter-inverter combinations in differentconfigurations and by applying a proper gate control. avariety of fixed and variable output parameters of a fixedparameter a.c. input power can be obtained. When themotor is operating at very low speeds, say, below 5% ofN,, the motor voltage demand is also low. If the invertercircuit is load commutated (motor side), its phase currentwill have to commutate with a very low voltage at theload side. It is difficult to guarantee reliable commutationat such low voltages. Pulse width commutation is thereforealso employed in thyristor drives when the motor has tooperate at very low speeds. Where the motors are verylarge, cyclo converters can also be employed. Below wediscuss a few inverter configurations. Generally PWM(for IGBTs) and gate control schemes (for thyristors)may be applied to these inverter circuits to obtain therequired variable a.c. supply parameters at the outputline to suit a particular requirement.6.9.3 Voltage source inverter (VSI) using IGBTs(to vary V andf)This is the most commonly used inverter for the controlof a.c. motors and is shown in Figure 6.28(a). The fixedd.c. voltage from the uncontrolled rectifier converter actsas a voltage source to the inverter. The voltage in theinverter unit is varied to the required level by using apulse width modulation, as noted earlier. Through theswitching circuit of the inverter the frequency of the

www.taghedanesh.ir6/126 Industrial Power Engineering and Applications HandbookHigh voltage LOW current ~~~~l~ constantand current ripplesvoltage sourceripples/ /A Dower diodeInverterfixed voltageunitrectifierFigure 6.28(a) A voltage source inverter (VSI)- Use of Cis to hold the charge and providea near constant voltage source to theinverter- Use of LIS to improve the quality of d.c.power to the inverter, in turn to improvethe quality of power to the machine.output supply is varied by repeated switching of the IGBTs.The frequency of switching of the IGBTs determines thefrequency of the output a.c. voltage. The inverter unit(converter-inverter unit combined) can be considered ascomprising:1 Rectifier unit (converter) This is a fixed voltageuncontrolled diode bridge rectifier.2 Smoothing circuit To obtain a near-constant voltagesource for the inverter circuit a smoothing capacitanceacross the d.c. link is used to smooth the a.c. ripplespresent in the d.c. link after conversion. The capacitorretains the charge and provides a near-constant d.c.voltage output.3 Inverter unit This inverts the fixed d.c. voltage to avariable Vandfa.c. voltage. Such a system can controlmulti-motor drives, operating on the same bus andrequiring similar speed controls. The inverterparameters can be closely controlled with the help offeedback controls and sensing devices as illustratedin Figure 6.28(b).0 Isolator@ Fuse@ Diode bridge rectifier (converter)@ Inverter unit IGBT or thyristor depending upon the size of machine@ CT@ Current comparator@ Current amplifier and controller@ Gate control in case of thyristor inverters only.Figure 6.28(b) Single line block diagram showing cascadeconnections of motors on a variable voltage common bus,using a VSI6.9.4 Current source inverter (CSI)(to vary 1, andf)This is similar to a voltage source inverter, except thatnow it is the rectified current that is varied rather thanthe voltage. On the input side of the inverter it acts as acurrent source. Since this current is already pre-set forthe required a.c. output current I,, the motor current isalways within its permissible loading limits. The currentcontrol, therefore, provides self-control to the motor. Asbefore, as shown in Figure 6.29, a fixed d.c. voltage isprovided through an uncontrolled diode bridge rectifier.This voltage is converted to a constant current sourcewith the help of a large series inductor in the d.c. link.The purpose of this inductor is to provide a near-constantcurrent source by reducing the dildt ripples. The inductorplays the role of a current source and acts like a currentchopper. Since, the voltage across the inductor can beexpressed bydi .V 2- L - (ignoring R of the circuit) (6.6)dtthe higher the value of L, the lower will be the currentovershoots, i.e. the rate of the current change dildt, throughthe inductor. A high value of the inductor would make itpossible to provide a near-constant current source forthe inverter circuit. With more modifications, the above

www.taghedanesh.irStatic controls and braking of motors 6/127A power diodefixed voltagerectifierInverterunitLarge L IS essential to smoothen ripples andprovide a near constant current source to theinverter.Figure 6.29 A current source inverter (CSI)can be made to operate according to a pre-defined currentwaveform for very strict speed control of a motor. Nowthey may be called current regulated inverters. Withfeedback controls, precise control of a motor can beachieved. A current source inverter provides a simplerand better control and may be preferred for large drives.particularly where regenerative controls are involved.Now the frequency of the a.c. output current is alsovaried through the switching of the IGBTs in the inverterunit, as noted earlier, and the current is varied by varyingthe output a.c. voltage, using the same PWM techniqueas for a VSI. Through this scheme only single-motorcontrol is possible, as different motors will have differentcurrents, as they may be of different ratings. However, itis more suitable for larger drives, as it is easy to handlecurrents rather than voltages.6.9.5 Cyclo converters (frequency converters)In addition to the above inverter systems there is onemore system, called a cyclo converter system. Thesedrives are employed for very large motors, when IGBTsin such ratings are a limitation. It converts the fixed a.c.supply frequency to a variable frequency, generally lowerthan rated, directly and without rectifying it to a d.c.source. They are basically frequency converters. Thissystem is more complex and expensive and has onlylimited application, such as for very large motors thatare to operate at very low speeds (e.g. in cement factoriesand steel mills). For details refer to the literature availableon the subject. A few books are mentioned in the Furtherreading.6.9.6 The regenerative schemesA motor can fall in a generator mode when the machineis energized and is run beyond its synchronous speed,such as when driving a load. travelling down hill orwhen its speed is reduced to perform a specific duty. Thesame conditions will appear when a running machine isreversed, whether it is an as. or a d.c. machine.In any of the above generating modes if the surplusenergy is not fed back to the supply source it may haveto be dissipated in some other form. Otherwise it mayraise the d.c. link voltage beyond its acceptable level.and lead to an unwanted trip of the machine or overheating.It may also endanger the static devices used in the invertercircuit or the components in the d.c. link. One simpleway to do this is to consume these in a resistor as shownin Figure 6.30. This is known as dynamic braking, andthe regenerative energy is wasted. The resistor isintroduced in the circuit through a bus voltage sensor.As soon as the bus voltage rises beyond a pre-set limit,the resistor is switched into the circuit. in smaller motorsPower diodeconverterDC linkInverter unitFigure 6.30 An IGBT inverter unit with dynamic braking

www.taghedanesh.ir6/128 Industrial Power Engineering and Applications Handbookit is common practice to dissipate the heat of regenerationin this way but in larger machines it can be a substantialdrain on the useful energy, particularly when the machineis called upon to perform frequent variations of speed,reversals or brakings. It is, therefore, advisable to conservethis energy by feeding it to the other drives or bytransferring it back to the source of supply, which can bedone in the following ways.When controlling an ax. machine, the converter isusually a full-wave, power diode fixed-type rectifier andthe Vlf is controlled through the IGBT inverter. Forregenerative mode, the d.c. bus is connected in antiparallelwith a full-wave voltage-controlled inverter asshown in Figure 6.3 1.During a regenerative braking, the d.c. voltage startsto rise, which the inverter regulates to the required level(V andf) and feeds back the regenerative energy to thesource of supply. This is known as synchronous inversion.When the inverter is of a lower voltage rating than thesupply source then a transformer between this and thesupply source as shown will also be necessary to regulatethe feedback voltage to the required level. The delta sideof the transformer may be connected to the supply sideto eliminate the third harmonic quantities to the source.Overvoltage, overload and short-circuit protections maybe provided on the d.c. bus, on which may occur a faultor whose voltage may rise beyond the pre-set limits. Asimilar inverter circuit is used in a d.c. machine alsowhen the regenerative energy is to be fed back to thesource of supply (Figure 6.32).It is also possible to use an IGBT converter instead ofa power diode or a thyristor converter. A single IGBTconverter unit is capable of performing both the jobs,converting the fixed a.c. supply to a fixed or variable d.c.voltage to control an ax. machine and during a regenerativemode, feeding the regenerative energy back to the sourceof supply. A separate transformer will not be necessarynow, as the same IGBT circuit will act as a regenerativeinverter. The switching of an IGBT is now an easy feature.The harmonics are also too low, and the p.f. can bemaintained up to unity. Now the IGBT converter can becalled a sinusoidal converter, as it will provide a nearsinusoidalwaveform of voltage and currents during afeedback. The d.c. bus can also be made a common busto feed a number of drives through their individual IGBTinverters to cut on cost. This is possible when a numberof such drives are operating in the vicinity or on thesame process line. Figure 6.33 illustrates a simple schemewith a common d.c. bus.6.10 Smoothing ripples in thed.c. linkA power diode rectifier unit feeding a fixed d.c. powerto an inverter unit to control an a.c. motor, or a thyristorrectifier unit, directly controlling a d.c. motor, both containax. ripples in their d.c. outputs, as illustrated in Figure6.24(a). It is essential to smooth these ripples to improvethe quality of d.c. power. To achieve this, a series inductorL is provided in the d.c. link as shown in Figures 6.24(a)and 6.28(a). In the process it also reduces the harmonicson the input side. To cut on cost, it is possible to limit theDC linkI-... -I ,Thyristor inverter> in antiparallel,to feed energyback to sourceDedicated transformerto match the supplysource voltageFigure 6.31 Regenerative energy feedback arrangement for an inverter unit

www.taghedanesh.irStatic controls and braking of motors 61129-6'* Fully controlled thyristorinverter in anti-parallel tofeedback energyFigure 6.32 Regenerative energy feedback arrangement fora converter unituse of such inductors to larger drives only, say, IO h.p.and above. In smaller drives the ripples may not significantlyinfluence the performance of the machine.The inductance in the d.c. link may cause a reversevoltage spike across the power diodes or thyristors as aresult of the decay of the reverse current (release of itsstored energy). A power device may be protected againstsuch voltage spikes through an R-C snubber circuit, asshown in Figure 6.37. (This circuit is discussed later.)-co'CircuitinterrupterIGBTinverter unitsu~vA group of AC drives beingfed from a common DC bus6.11 Providing a constantd.c. voltage sourceAfter smoothing the d.c. voltage may contain moderateripples not desirable when a constant voltage d.c. sourceis needed. To achieve this. a charging capacitor C is alsoprovided across the d.c. link for all sizes of drives asqhown in Figures 6.24(a) and 6.281a).6.12 Providing a constant currentsourceInstead of a charging capacitor C, a large size seriesinductor L is introduced in the d.c. link (Figure 6.29).Since V = L dildt, the larger the value of L, the lower willbe the current overshoots (di/dt) and a near-constant d.c.link current source is obtained for the inverter unit.(+ve) (-ve) IGBT converter andv feedback inverter unitsCommonDC busFigure 6.33 An lGBT converter-cum-inverter unit to feed backregenerative energy6.13 Generation of harmonics andswitching surges in a staticdevice switching circuitA switched static device (particularly a thyristor) producesvoltage and current transients similar to inductive or

www.taghedanesh.ir6/130 Industrial Power Engineering and Applications Handbookcapacitive switching (Section 17.7). They also produce A high value of L will limit the rate of rise of faultharmonics. However, a power diode converter unit having current for the same voltage and save the circuitno switching sequence is devoid of such a phenomenon. components.A thyristor (SCR) switched phase-controlled converter 2 A low dildt will also help to smooth the d.c. linkunit produces large quantities of harmonics on the supply current waveform.(a.c.) side and also in the d.c. link and also voltage and 3 It has a disadvantage in that it may have a sluggishcurrent surges on the incoming supply side. An IGBT, response to the control circuit demands due to itson the other hand, as used in an inverter circuit, causes high time constant (z = L/R)only moderate harmonics during switching, but doesproduce voltage surges on the load side. All these factors Current harmonics on the incomingare not desirable and must he suppressed or tamed at the ax. supply sidepoint of occurrence to save the connected equipmentand the devices. Below we discuss these phenomena, The presence of harmonic quantities in the electronicparticularly for thyristor circuits and their possible circuit distorts the sinusoidal incoming supply system toremedies.a non-sinusoidal one the magnitude of which will dependupon the configuration of the converter circuit and the6.13.1 Suppressing the harmonics (in phase- variation in the connected load. The line side convertercontrolled rectifier units)unit draws a somewhat squarish waveform current fromthe mains, as analysed in Figure 23.7. It may adverselyPhase-controlled rectifier circuits generate excessive odd influence the power equipment operating on the incomingharmonics such as 5th, 7th, llth, 13th etc., depending supply side of the system, which may be a motor, aupon the pulse number, n, of the circuit configuration transformer or a generator, due to higher no-load lossesadopted, as discussed in Section 23.6(h). These harmonics as a consequence of high harmonic frequencies (equationsadd to the inductive loading of the circuit since XL 0~ fh (1.12) and (1.13)). It may also cause overloading of theand diminish the p.f. of the system, although they hardly capacitor banks connected on the incoming side andinfluence an induction motor, (Section 23.5.2(B)). The subject all this equipment to higher voltage stresses. The3rd harmonic3 are totally absent because mostly six pulse higher inductive loading also diminishes the p.f. of thethyristor converters are employed, which eliminate all system. To contain the influence of these features, thethe 3rd harmonics from the voltage and the current output use of filter circuits to suppress the harmonics and powerwaveforms. Thyristors in other configurations such as capacitors, to improve the system p.f. on the incoming12, 18 and 24 pulses are also possible, which can eliminate side are mandatory to maintain a healthy supply system,most of the harmonics from the output waveforms. The particularly when it is feeding a few phase-controlledhigher the pulse number, the closer it approaches the converter units, handling large machines and generatingmean and effective (r.m.s.) values of the rectified voltage high harmonics. Figure 6.34 shows the use of an inductorand the voltage approaches a near peak value (Section in the incoming circuit to suppress the harmonics and23.6(b)). (See also Figure 23.10.) However, higher pulse limit current overshoots. Power capacitors are not shown,thyristor converters become very expensive and are which can be provided for the whole system at a centralizedemployed only for very large power applications. location. The design of filter circuits and the size ofpower capacitors, to adopt a more appropriate correctiveCurrent harmonics in a d.c. linkmethod, will require a meticulous network analysis todetermine the actual numbers and magnitudes of suchTo limit the current harmonics generated in the d.c. link, harmonics present in the system. The subject is dealtseries smoothing reactors are inserted on the d.c. side as with in more detail in Section 23.5.2. In Figure 6.24(a)shown in Figure 6.24(a). They are large iron core un- we have shown a few more common types of thyristorsaturable reactors (L).(For details on reactors see Chapter configurations, their voltage and current wave-forms and27.) They provide high impedance paths to the different the application of reactors to suppress the harmonics andharmonic quantities and suppress the more prominent smooth ax. ripples.of these at the source, and provide a near smooth d.c.output voltage waveform. For large power applications,requiring a near-constant d.c. output, more accurate L-Ccircuits (even more than one) may be provided in the 6.14 Protection of semiconductord.c. link to suppress the more prominent of the harmonic devices and motorsquantities.A large inductor in the d.c. link may also play the 6.14.1 Overvoltages and voltage surges causedfollowing roles:by disturbances in an LT systemI In the event of a fault in the d.c. link it will add to the Semiconductor devices are irreparable after a failure andcircuit impedance and limit the rate of rise of fault hence require extra precautions for their protection.current, since under a transient conditionAlthough a voltage transient generally is a phenomenonof HT systems, as discussed in Chapter 17, moderatediv= L-long-duration switching surges (voltage spikes), otherdt than lightning and the transference of surges, are noted

~ Transferencewww.taghedanesh.irStatic controls and braking of motors 6/131Switch FuseHigh voltage LOW current Nearly constantand current ripples voltage L, - To limit harmonics and dddfripples / L - To smoothen harmonics and/ripplesC - To hold the charge and act asDCa constant voltage source forthe d c machineR - The Combination of G-Ractsas snubber circuit to protectelectronic (static) componentsPhase controlledvariable voltagefrom high dL/dtrectifierFigure 6.34 Application of inductor and capacitor with a controlled bridge rectifier (for control of d c machines)on an LT system also (see IEEE-C62.41). We summarizethe likely causes for such surges on an LT system andtheir remedial measures below.External causes- Lightningof surges from the HV to the LV side of'a transformer (Section 18.5.2).Proteclion from such surges is achieved by using adistribution class surge arrester at the receiving end ofthe supply. Generally no protection is therefore necessaryfor the semiconductor devices. For more details seeChapter 18.Internal causesLT systems that are prone to frequent faults and outagesor constitute a number of inductive or capacitive switchedloads and welding transformers may generate temporaryovervoltages (TOVs) and voltage surges. Such systemsmust be studied carefully, and when felt necessary, ametal oxide varistor (MOV) or a large inductor be installedat the incoming side of the semiconductor circuits, asnoted below:outgoing feeder, as illustrated in Figure 6.35. theprotective device will trip and trap an inductive chargewithin the transformer and the connecting cables, withan energy of 1/2 L . (1;. - is: ). L is the inductance ofthe transformer and the interconnecting cables up tothe static circuits and is, the cut-off current of theprospective fault current I,,, at the instant of fault, asshown in Figure 6.35. The clearing of fault by a currentlimiting device is a transient condition and is synonymouswith a switching condition and may generateswitching surges. This energy is discharged into thecircuits located downstream of the feeding source. Thetrapped energy is a source of danger to all healthycircuits that are located near the source. and the worstaffected are the feeders, that may be switched at suchinstants.All the above surges and even the transference of alightning surge from an overhead line through inter-Switching of motors: this is generally not harmful asthcy can generate only overvoltages, not exceeding2 - v,.Sw,itching of capacitors: these also generate onlyovervoltages, generally not exceeding 2 * V,. However,it may be ufconcern when there is a parallel switchingof large capacitor banks, causing a high switchingfrequency (Section 23.5) and correspondingly a shorterwave and ;I shorter rise time that may take the shape ofa surge.current)O/G feedersSome of them may-3 hCurrent limiting+ deviceBoth the\e switchings on an LT system do not cause are-\tri keSuitching of welding transformers: these may causedangerous voltage surges.0 Fault condition: particularly when the LT distributionIS fed through a large transformer and the outgoingfeeders are protected by a current limiting device, HRCfuses or breakers. In the event of a fault, on a largestate circuitsvTrapped energy getsdischarged into thehealthy feederskircuitsFigure 6.35 Trapped energy distribution of a large feedingsource during a fault clearing by a current-limiting device

www.taghedanesh.ir6/132 Industrial Power Engineering and Applications HandbookLLI/CACsupplyLarge feedingsourceL Inductance of the supply sourceL, - Inductor to smooth ripples1L2- Inductor to absorb the trapped energy up to 2 L (If - &)[partly absorbed by the feeder's own impedance and other feedersconnected on the same line]Figure 6.36 Use of inductor on the supply side of a static drive to absorb the trapped energyconnecting cables are unidirectional and reflect in ncarlyfull at a junction and cause a doubling effect, hence theyare more dangerous. The semiconductor devices can besaved from such harmful effects by absorbing the trappedenergy. The effect of a trapped charge is somewhat areplica of a discharge of a surge arrester. In a surgearrester, the energy above the protective level of thearrester is discharged through the ground (Section 18.5).In this case, it is discharged into the healthy circuitsahead. Normal practice to tackle such a situation is toprovide an inductor, sufficient in size, to absorb thisenergy at the receiving end of the static circuit as illustratedin Figure 6.36. This protection is applicable to all typcsof electronic circuits. It is equally applicable even in apower diode converter unit, involving no switchingoperations.Transients occurring within the converter unitThis is applicable to thyristor (SCR) circuits to protectall the semiconductor dcvices used in the switching circuit,such as diodes (also power diodes) or TGBTs, in additionto SCRs. The same protection can be applied to all thesemiconductor circuits likely to experience high duldt.The role of SCRs is to vary the supply parameters,which require frequent changes in V, i.e. dvldt and in I,i.e. dildt in an energized condition. Because of momentaryphase-to-phase short-circuit, duldt occurs during switchingOFF and dildt during switching ON sequences. Both aretransient conditions and may damage the semiconductordevices used in the circuit. To protect the devices, thetransient conditions can be dealt with as follows:Voltage transients (duldt) When a thyristor switchesfrom a closed to an open condition, i.e. from aconducting to a non-conducting mode, a transientrecovery voltage (TRV) appears. This is a transientcondition and the rate of change of voltage can beexpressed byQ=C'.-dudtwhere, Q = charge stored within the devices beforeoccurrence of the switchingC' = leakage capacitance of the thyristorbetween its junctionsduldt = rate of rise of recovery voltage (r.r.r.v.)During a switching OFF sequence this charge must bedissipated quickly otherwise it may cause dangerousovervoltages, which may damage the static devicesused in the circuit or turn them ON when this is notwanted. A thyristor is switched ON only when there isa current pulse applied to its gate. It is possible thatthe gate may turn ON without this pulse as a result ofexcessive forward duldt due to leakage capacitancebetween the thyristor junctions. The leakage capacitancemay cause a charging current through the gate. Whenit exceeds its threshold value, it can turn the gate ON.duldt is therefore a very important limiting parameterto avoid an erratic turn ON of the thyristors, whichmay corrupt the output parameters and lead tomalfunctioning of the whole system or cause a shortcircuitand damage the static devices used in the circuit.It is therefore important to suppress such transientswithin safe limits. It is possible to contain them,provided that the stored energy can be dissipated quicklyinto another source. This is achieved by providing asnubber circuit across each static device as noted below,similar to the use of a quenching medium in an HTinterrupter (Section 19.2).Snubber circuit More conventional protection fromhigh duldt is to provide an R-C circuit across eachdevice, as shown in Figure 6.37. The circuit providesa low impedance path to all the harmonic quantitiesand draws large charging currents and absorbs theenergy released, Q, and in turn damps duldt withinsafe limits across each device. Now Q = C (dvldt)AIt may r R1th;:&ror an IGBT,.uKFigure 6.37 Use of a snubber circuit across a power switchingdevice

www.taghedanesh.irStatic controls and braking of motors 6/133where C is the capacitance used across the device.During a turn OFF operation the stored energy, Q, ofthe circuit will discharge into this capacitor and chargethe same to its optimum level (charging time constantr= R C) and slow down the rate of rise of TRV (r.r.r.v.).i.c. duldt across the static circuit and limit the voltagespikes. similar to motor protection discussed in Section17. IO. 1. The higher the value of C, the lower will bethe voltage (commutation) overshoots. During a switchON the capacitor discharges its total energy into the Rand prepares for the next switching operation. Thepower dissipation into R is proportional to the switchingfrequency. R also limits the peak value of the dischargecurrent through the static device and damps theoscillations. Here the use of C is to hold the chargeand then release the same into R and not to smooth theripples.Ciirrerlr rrnrisirrits A similar situation will arise whena switching ON operation of the rectifier unit occurswhen it is a thyristor rectifier. Under load conditions.the stored magnetic energy in the incoming supplysystem. which can be the feeding transformer and theline reactances similar to a fault condition discussedearlier. may cause a current transient which can beexpressed bywhereV =applied voltageL = inductance of the total circuit up to the d.c.link anddi/d/ = rate of change of current, as the switchingON is a transient condition and causes overloadand short-circuits. This is rnaxirnurn at thecommencement of switching ON and becomeszero on its completion. It is analogous tocontact making in an interrupter (Section19.1. I). The same situation will arise evenduring a fault condition. Excessive rate ofchange of current may cause an overload andeven a short-circuit.The rate of current change must therefore be controlledto a safe limit by providing a dampening circuit on thesupply side. This can be a series inductance as shownin Figures 6.2h(a), 6.34 and 6.36. This inductance maynot be necessary when the unit is being supplied througha dedicated transformer. The inductor will absorb themagnetic charge and damp the rate of rise of current.diNOW V = L -dtL, is the additional series rcactmce. The higher thevalue of L,, the lower will be the rate of rise of current.The inductor on the input side also suppresses theharmonics in the incoming supply, as high Li willprovidc a high impedance path to higher harmonics.For suppression of harmonics, where the supply systemis already substantially distorted, additional L-Cfilter circuits may be provided on the incoming sidefor more prominent harmonics. (For details of filtercircuits see Section 23.9.) The main purpose ofinductance here is protection. rather than suppressionof harmonics.Due to a high time constant of the dampening circuitt = L,/R (R being the resistance of the circuit) it willalso delay occurrence of the fault by which time thecircuit’s protective scheme may initiate operation.It would also add to the line impedance to containthe severity of the fault conditions.From the above we notice that the current surgescan be caused either by the tripping of a current limitingdevice. when the distribution is through a largetransformer on which is connected the static circuits.or by switching of the SCRs within the converter circuititself. The protective scheme for both remains the sameand is located at the incoming of the semiconductorcircuits. There can be two situations. When the staticcircuits are being fed through a dedicated transformerin all probability no additional inductor will beary. Not even when there is a large transformerg a large distribution network on which isconnected the semiconductor circuits. It is. however.better to carry out the trapped energy calculations tocompare these with the inductance already availablein the switching circuits of the semiconductor devices.When there is no dedicated transformer and thesecircuits are connected on the system bus directly alarge inductor will be essential at the incoming of thestatic circuits, sufficient to absorb the trapped chargewithin the transformer and the interconnecting cablesup to the converter unit. The size of the inductor canbe calculated depending on the size (kVA) of thedistribution transformer, its fault level and thecharacteristics of its current limiting protective device.An inductor sufficient to absorb i,: . L of the transformerand the cables may be provided at the incomingof the static circuits.Voltage surges in the inverter circuitGenerally, voltage surges on an LT system are of littlerelevance as analysed in Section 17.7.6. Instances can.however, be cited of motor insulation failures, even onan LT system, when the machine was being controlledthrough a static drive, which might be an IGBT switchedor a thyristor (GTO) switched inverter. the reason beinga steep rising switching wave generated through theinverter circuit. The output of the inverter unit being inthe shape of a non-sinusoidal voltage waveform also addsto the switching transients. To visualize the effects offast switching in a static circuit, it is relevant to corroboratethese with the switching of a conventional HT interruptingdevice, discussed in Section 17.7. The static devices alsocause switching surges and their severity is also definedby their amplitude and the rise time (Figure 17.2). Thesedevices are seen to produce voltage surges with anamplitude up to two to three times the voltage of the d.c.link and a rise time as low as 0.05-0.4 ps (typically) inIGBTs and 2 to 4 ps (typically) in GTOs (see Lawrenceet al. (1996) and the Further reading at the end of the

www.taghedanesh.ir6/134 Industrial Power Engineering and Applications Handbookchapter). Their severity increases when they cause areflection (doubling phenomenon) at the motor terminals.The amplitude of the reflected wave will depend uponthe length of the interconnecting cables, between theinverter and the machine and the switching frequency ofthe inverter unit. The amplitude of the wave after reflectionmay exceed the BIL (basic insulation level, Table 1 1.4for LT and Table 11.6 for HT motors) of the machine,particularly when the machine is old and its insulationhas deteriorated (Section 9.2), or when it is required toperform frequent switching operations. In IGBTsparticularly, the rise time is too short and the surgesbehave like an FOW (front of wave, Section 17.3.3),which are all the more dangerous for the end turns of theconnected machine. The length of cable from the inverterto the machine plays a vital role. The longer the cable,the higher will be the amplitude of the voltage surge atthe motor terminals. This aspect is discussed in greaterdetail in Section 18.6.2. The leading manufacturers ofstatic drives specify, as a standard practice, the amplitudeand the rise time of the switching surges of their devicesat a particular voltage and switching frequency of theirinverter unit (normally 2-4 kHz) and the maximum safecable lengths. While most installations may not needseparate surge protection, it is advisable to take precautionsagainst any contingency during actual operation to protectthe machine against these surges under the most onerousoperating conditions. The remedies are the same as thosediscussed in Section 17.10 on the protection of electricmotors. For example, surge capacitors for most of themotors will prove sufficient and economical to protectthe machine by taming the steepness of fast rising surges atthe motor terminals. (See curves 1 and 2 of Figure 17.21 .)The surge capacitor may be provided with a dischargeresistance as standard practice, as also noted in the snubbercircuits. The resistance would help the capacitor todischarge quickly and prepare for the next operation.In addition, it would also help to dampen the amplitudeof the arriving surge. The use of inductor on the loadside to provide an impedance to the arriving surges witha view to suppressing them is not good practice, for itmay diminish the p.f. of the circuit and also cause avoltage drop across it, which may affect the machine’sperformance.Note A manufacturer of static drives would normally give anoption to the user to operate their inverter circuit fIGBTs normally)at high PWM carrier frequencies (typical 2-8 kHz) to smooth theoutput (load side) voltage. But at high frequencies, the propagationof surges becomes faster and may cause quick reflections, whichwould require either a shorter cable or the use of a surge suppressor.High-frequency operations also raise the noise level in the groundpath and can cause sensitive devices like PLCs, sensors and analoguecircuits to behave erratically, as they are all connected through theground circuit. It is therefore desirable to operate the IGBTs orGTOs at lower frequencies, preferably 24 kHz, as this will causelow ripples as well as a low noise level. A moderate carrier frequencywill also help in taming the aniving surges at motor terminals withonly moderate steepness.Conclusion1 Induction motors, both, squirrel cage and slip-ring,can be easily controlled to achieve the required characteristicsby applying solid-state technology.With the availability of phasor control technology, bywhich one can separate out the active and magnetizingcomponents of the motor’s stator current and varythem individually, it is now possible to achieve higherdynamic performance and accuracy of speed controlin an a.c. machine similar to and even better than aseparately excited d.c. machine.With this technology it is now possible to achieveextremely accurate speed control of the order off 0.01% to f 0.001%. To achieve such high accuracyin speed control, closed-loop feedback control systemsand microprocessor-based control logistics can beintroduced into the inverter control scheme to sense,monitor and control the variable parameters of themotor to very precise limits.A very wide range of speed controls is available throughthis technology as it is possible to vary frequency onboth sides (k) of the rated frequency.Controls are available in the range: IGBTs 1600 V,2000 A and thyristors 10 kV, 3000 A (ratings are onlyindicative) and can cover the entire voltage rangesand ratings of a motor.6.15 Energy conservation throughsolid-state technologyWhile the motor is operating under loadedIn the various types of static drives discussed so far, thesupply voltage would adjust automatically at a level justsufficient to drive the motor to meet its load requirements.Hence it is not necessary for the motor to be appliedwith full voltage at all times: the voltage adjusts with theload. This is an in-built ability of a static drive that wouldsave energy and losses. One would appreciate that mostof the industrial applications consider a number ofderatings and safety margins while selecting the size ofthe motor to cope with a number of unforeseen unfavourableoperating conditions occurring at the same time.This is discussed in Chapter 7 (see also Example 7.1).The size of the motor is therefore chosen a little largerthan actually required. As a consequence even when themotor may be performing its optimum duty, it may hardlybe loaded by 60-80% of its actual rating causing energywaste by extra iron and copper losses and operating at areduced p.f. Static drives are therefore tangible means toconserve on such an energy waste.While performing a speed controlA very important feature of solid-state technology is energyconservation in the process of speed control. The sliplosses that appear in the rotor circuit are now totallyeliminated. With the application of this technology, wecan change the characteristics of the motor so that thevoltage and frequency are set at values just sufficient tomeet the speed and power requirements of the load. Thepower drawn from the mains is completely utilized indoing useful work rather than appearing as stator losses,rotor slip losses or external resistance losses of the rotorcircuit.

www.taghedanesh.irStatic controls and braking of motors 6/1356.15.1 Illustration of energy conservationIn an industry there may be many drives that may not berequired to operate at their optimum capacity at all times.The process requirement may require a varying utilizationof the capacity of the drive at different times. In aninduction motor, which is a constant speed prime-mover,such a variation is conventionally achieved by throttlingthe flow valves or by employing dampers.There may be many types of the drives in an industry,particularly when it is a process industry. The mostcommon drives are fan$, pumps, and compressors etc.,employed for the various utilities. storage and processactivities of the plant. The plant may be chemical or apetrochemical. water treatment or sewage disposal, paperand pulp unit or even a crane or a hoist application.The method of speed or flow control by throttling,dampening (vane control) or braking, indirectly reducesthe capacity of the motor at the cost of high power lossin the stator and slip loss in the rotor circuit, as discussedabove. These losses can now be eliminated with theeffective use of static control variable-speed drives orfluid couplings. We will show, through the followingillustrations. the energy saving by using such controls.Throttle, dampening or vane controlFor ease of illustration we will consider the characteristicsand behaviour of a centrifugal pump which is similar inbehaviour to radialhxial flow fans and centrifugakrewcompressors. Figure 6.38 shows the mechanical connectionof a flow valve to control the output of the pumpor the discharge of the fluid through the throttle of thevalve. Figure 6.39 illustrates the characteristics of thepump:Dijcharge versus suction head, i.e. Q versus Hd andDischarge versus pump power requirement, Le. Q versush.p.The rated discharge is Ql at a static head of Hdl and amotor h.p. PI. In the process of controlling the dischargefrom Q, to Q2 and Q3, the valve is throttled, whichinci-eases the head loss of the system (or system resistance)from Hdl to Hd, and Hd3 respectively. The operatingpoint on the Q-Hd curve now shifts from point A, to A2and A, as a result of back pressure. The pump powerrequirement now changes from PI to P, and P, on theQ-h.p. curve. We can see that. due to added resistance inthe system. while the discharge reduces, the correspondingpower requirement does not reduce in the same proportion.Flow control through static controlThe same operating control. when achieved through theuse of a solid-state control system will change themechanical system to that of Figure 6.40. We have useda simple. full-wave, phase-controlled, variable-voltagesolid-state device, employing a triac (two SCRs in antiparallel).The voltage to the motor is monitored througha flow sensor, which converts the flow of dischargethrough a venturi meter to electrical signals. These signalsR Y BrVenturimeter - to measure the velocity of fluidProbe to sense the velocity of fluid0Flow meter or sensor - To convert the velocity of fluidto the rate of flowMotorized sluice valve - To throttle the flow of fluidFigure 6.38 Conventional throttle control- DischargeIcontrol the voltage and adjust the speed of the motor tomaintain a predefined discharge flow. The use ofa throttlevalve is eliminated, which in turn eliminates the extrahead loss or system resistance. Figure 6.41 illustratesthe corresponding characteristics of the pump with thistype of flowispeed control.To reduce the discharge from Q, to Q2 and Q3 in thiscase, the speed of the pump, and so also of the motor,reduces from N,, to Nr2 and Nr3. The Q-Hd characteristicschange according to curves Nrz and Nr3, at a correspondingpump power requirement of Py and P;l respectively.according to power curve P'. These power requirementsare significantly below the values of P, and P, of Figure6.39 when discharge control was achieved by the throttle.The system resistance curve remains unaffected, whereasthe pump power demand curve traverses a low profile asin curve P' due to lower speeds N,? and Nr3. The powerrequirement diminishes directly with speed in such pumps.The energy saving with this method is considerable,compared to use of the throttle. which is also evidentfrom curve P' of Figure 6.41.6.15.2 Computation of energy savingConsider Figure 6.42 with typical Q-HCI curves at differentspeeds and different system resistances, introduced bythe throttle. Point A refers to the rated discharge QI atrated speed N,, and head Hdl when the throttle valve ihfully open. Lets us consider the condition when thedischarge is to be reduced to say, 0.67 QI.

www.taghedanesh.ir6/136 Industrial Power Engineering and Applications HandbookHd3Hd2t HdlPP9v)ccurves-0Discharge 'Q'Q3 Q2 QI(a) Discharge versus suction head.0Q3 Q2Discharge 'Q'-Q,(b) Discharge versus pump power.Figure 6.39 Power requirement and rate of discharge on a throttle controlThrottle controlThe system resistance increases and discharge reducesat the same rated speed Nrl. This condition refers topoint B, to which the earlier point A, has now shifted.The system now operates at a higher head Hd2, whereasthe actual head has not increased. This condition hasoccurred due to higher system resistance offered by thethrottle. The pump and the prime-mover efficiency willnow reduce to 73% from its original 85%.Static controlThe same discharge of 0.67 Q, is now obtained withoutthrottling the valve, but by controlling the speed of themotor to 0.65 Nrl. Point A now shifts to point C on theQ-Hd curve for 0.65 Nr,. For this reduced discharge of0.67 Q, the suction head Hd also reduces to 0.45 Hdl andso also does the system resistance. The efficiency of thesystem is now set at 82%.There is thus an obvious energy saving by employingstatic control over the conventional and more energyconsumingthrottle control.Extent of energy savingExample 6.1This can be determined by

www.taghedanesh.irStatic controls and braking of motors 611 37R Y E(iii) Power required when discharge is controlled throughstatic control, i.e. by speed variation;..Head Hd3 = 0.45 Hd,discharge = 0.67 Qq = 0.820.45 Hdl x 0.67 Q . dPower36 x 0.82Ratio of power through speed variationRated power requiredII00.45 x 0.67 o,850.82= 0.31i.e. the power reduction through speed control is 69% that ofthe rated power requirement.Thus the energy saving in this particular case by employingthe method of speed control over that of throttle control willbe= 69%-11% = 58%0 Venturimeter -To measure the velocity of fluid@ Probe to sense the velocity of fluid@ Flow meter or sensor - To convert the velocity of fluidto the rate of flow@ Variable voltage (fixed frequency) static controlFigure 6.40 Energy conservation through static controlwhereP = shaft input in kWHd = head in barQ = discharge in m3/hourd = specific gravity of the liquid in gm/cm3q = efficiency of the pump(i) Rated power required by the pump(ii) Power required when discharge is controlled through thethrottle,head Hd, = 1.138 Hdldischarge = 0.670q = 0.731.138 H,, x 0.67 Q. d:. Power =36 x 0.73Ratio of Dower throuah throttleRated power required1.138 x 0.67 o,850.73= 0.89i.e. the power reduction with this method is around 11% ofthe rated power requirement.Example 6.2To determine saving of costs through speed control, considerthe following parameters, for the above case:P = 100 kWDischarge = 67% of the rated flowDuration of operation at reduced capacity = say, for 25%of total working hours.Energy tariff = say, Rs 1.2 per kWh:. Total energy consumed per year, considering 300 operatingdays/year= 100 x 24 x 300= 720 000 kWhEnergy consumed while operating at the reduced capacityfor 25% duration = 0.25 x 720 000= 180 000 kWh:. Energy saving = 0.58 x 180 000= 104 400 kWhAnd total saving in terms of cost = 1.2 x 104 400= Rs 125 280 per yearExample 6.3Energy saving in a slip recovery systemConsider an I.D. fan of 750 kW, running at 75% of N, for atleast 20% of the day. Considering the load characteristic incubic ratio of speed,:. P at 75% speed= (0.75)3 x 750 kW= 316 kWand slip power,f, = (1 -0.75) x 316 kW= 0.25 x 316= 79 kW

www.taghedanesh.ir6/138 Industrial Power Engineering and Applications HandbookPump head04 4Capacity 'Q'-(a) Discharge versus suction head-04 4Capacity 'Q'(b) Discharge versus pump powerQ,Figure 6.41Power saving with the use of static control compared to throttle controlConsidering the efficiency of slip recovery system as 95%and 300 operating days in a year, the power feedback to themain supply source through the slip recovery system in 20%of the time,= [0.95 x 791 . [0.2 x 300 x 241= 108 072 kWhAnd in terms of costs at a tariff rate of Rs 1.2 per unit,= RS 1.2 x 108 072= Rs 129 686.4 per year6.16 Application of static drives6.16.1 Soft starting (Vlf control)This facilitates a stepless reduced voltage start and smoothacceleration through control of the stator voltage. Thesmooth voltage control limits the starting torque (T,Jand also the starting inrush current (IJ a5 illustrated inFigure 6.43. This is achieved by arranging a full-wavephase-controlled circuit by connecting two SCRs in antiparallelor a triac in each phase as illustrated in Figure

www.taghedanesh.irStatic controls and braking of motors 611391 25Hd2 1 138t Hd‘O0--.$ 0752P050Hd3 0 450 250-Hd - HeadH, - Rated headN - Actual speedN, - Rated speed025 050 067075Speed -c-1 II 1 QiQ - DischargeQn - Rated dischargeq - Pump efficiency-RatedpointFigure 6.42 Discharge versus head curves of a pump atdifferent speeds and resistances introduced by throttle6.44. Varying the firing angle and the applied voltagewill also affect the dynamic phase balancing in eachphase, and controls the I,, and T,, as programmed.Increasing the firing angle decreases the angle ofconduction and causes the voltage to decrease. as seen inFigure 6.23. The starting voltage, i.e. the angle of firing,is pre-set according to the minimum voltage required, toensure a permissible minimum T,, and maximum I,, whichcan be predetermined with the help of the motorcharacteristics and the load requirements. The voltage israised to full, gradually and smoothly, but within a presettime, as determined by the motor and the loadcharacteristics. The size of the static starter will dependupon the starting current chosen and the correspondingstarting time. Generally, the normal practice of the variousmanufacturers, as noted earlier, is to define the size oftheir soft starters, based on a starting current of I SO% of1,and a starting time of up to one minute. A higher startingcurrent or a higher duration of start may call for a largerstarter. The starter therefore provides no control over thestarting current, which is a function of the applied voltage.It is also possible to perform a cyclic duty havingsome no load or a light load and some fully loaded periods,as discussed in Section 3.3. The firing angles of theSCRs or the triacs can also be programmed accordinglyto reduce the applied voltage to the motor to a minimumpossible level, during no-load or light-load periods, andhence conserving on otherwise wasted energy by savingon the no-load losses. Different mathematical algorithmsare used to achieve the desired periodic T-Ncharacteristicsof the machine.During start-up, the firing angle is kept high to keepthe V and I,, low. It is then reduced gradually to raise the(b)controlFigure 6.43Speed control by varying the applied voltage (use of higher-slip motors)

www.taghedanesh.ir6/140 Industrial Power Engineering and Applications HandbookR Y B0 Rotor000Variable V, constant fcontrolFigure 6.44 Stator static voltage control (soft starting)voltage to the required level. There is thus a controlledacceleration. The maximum I,, can also be regulated to arequired level in most cases, except where the T,, requirementis high and also needs a higher starting voltage,when it may not be possible to limit the I,, to the requiredlevel. The scheme is a simple reduced voltage start anddoes not facilitate any speed control. Since the startingcurrent can now be regulated to a considerable extent, thescheme is termed a soft start. Now T,, =V2 and I,, - V.This may be a good alternative to Y/A or autotransformerstarting by eliminating current overshoots during thechangeover from Y to A in a YlA or from one step toanother in an autotransformer start. It also prevents anopen transient condition (Section 4.2.2(a)). Also in asoft start the switching SCRs or the triacs are bypassedthrough a contactor after performing a starting duty similarto a YlA, AIT or a wound rotor start. In conventionalstartings also the star contactor, the autotransformer orthe external resistances, respectively, are cut off fromthe circuit, after performing their starting duty. Such aprovision protects the starter unit from avoidable voltagestrain, saves heat losses and cools it to prepare it for thenext switching operation. The starter is also free to performswitching duties on other machines of a similar sizeconnected on the same line, if so desired, to save costs.The starter can be used as a switching kit for a group ofsimilar-sized motors.The voltage can be varied from 0% to 100% and hencethe ZSt can be limited to any required level. Such a linearvoltage variation may be suitable in most cases, particularlywhen the motor has to pick up a light load or isat no-load. Motors driving heavy inertia loads or loadsrequiring a high starting torque or both may not be ableto pick up smoothly with a linear voltage rise from 0volts. This may cause a locked rotor or stalling until thevoltage rises to a level sufficient to develop an acceleratingtorque, capable of picking up the load. It is also possiblethat it may need a prolonged starting time notcommensurate with the thermal withstand time of themotor or a larger starter. In such cases, it is essential tohave a minimum base or pedestal voltage as illustratedin Figure 6.45(a). The voltage is adjusted to the lowestpossible level, so that the I,, is kept as low as possible. InFigure 6.45(b) we illustrate a motor with a normal startingcurrent of 650% I, at the rated voltage. To limit this to amaximum of 350% of I,, we have provided a base voltageof about 350/650 or 54% of the rated voltage. Theminimum T,, is, however, matched with the loadrequirement to attain the rated speed within its thermalwithstand time. For more details see Section 3.5 andExample 7.1. The voltage is then raised so that duringthe pick-up period the I,, is maintained constant at 350%,until the motor reaches its rated speed.Since it is not practical to custom build each drive, thenormal practice by manufacturers of soft starters is toprovide a variety of motor parameters to which nearlyall motors will fit and the user may select the motor thatbest suits the load requirements. The other advantages ofa solid-state soft start are that an unbalanced power supplyis transformed to a balanced source of supply automaticallyby suitably adjusting the firing angle of each SCR ortriac through their switching logistics. Also a low startingcurrent can economize on the size of switchgears, cablesand generator where a captive power is feeding the load.6.16.2 Soft stoppingThe normal method to stop a motor is to do it instantly.This is accepted practice in most cases and the machinestops as shown in equation (6.8) discussed later. A highermoment of inertia (MI) of the driven masses will mean areasonably longer duration to come to a standstill, whilea low MI will mean a faster stop. But in loads requiringhigh braking torques, such as conveyor systems, escalatorsand hoists, the stopping time may be too short. In somecases, such as a pump duty, the stoppage may be nearabrupt.Such a situation is not desirable and may causeshocks to the motor. In a pump, an abrupt stoppage maycause severe shocks and hammering effects on pipelines,due to backflow of the fluid, Shocks may even burstpipelines or reduce their life when such stoppages arefrequent. They may also damage non-return valves andother components fitted on the lines and thus weaken thewhole hydraulic system.A gradually reducing voltage rather than an instantswitch OFF is therefore desirable for all such applications.This would also gradually reduce the flow of the fluid,leading to a strain-free and smooth stopping of themachine. In such cases, a soft stop feature, similar to asoft start, can be introduced into the same starter, whichwould gradually reduce the stator voltage and facilitatea smooth and shock-free stop.6.16.3 Slip recovery system (to control woundrotor motors)As discussed earlier, the motor speed-torquecharacteristics depend largely upon the rotor current

www.taghedanesh.irStatic controls and braking of motors 61141t1100t100 58=54546 NrSpeed (in terms of pre-defined bme) +I , I ,,Approximate torque curve during a soft startNormal torque curveLoad torqueBase or pedestal voltageSoft starter can be removed from theother motors, if desired.(a) Torque characteristicsandII , I,1 ' A 'to start6 NrSpeed (in terms of pre-defined time) 4(1) Approximate current curve during a soft start(2) Voltage can be adjusted to maintain the starting current constantat 350% (or anv desired value). I(3) Normal current curve(4) Base or pedestal voltage(5) Soft starter can be removed from the circuit and used to startother motors, if desired.(b) Current characteristicsFigure 6.45 Current and corresponding torque characteristics of a motor during a soft start(equation ( 1. I)) and rotor resistance (equation (I 3)).The higher the rotor current or resistance, the higher willhe the starting torque as illustrated in Figure 6.46, andthe higher will be the slip and slip losses as well asrcduced output. The maximum torque is obtained whenR3 and ,,X2 are equal. Speed control can be achieved byvarying the rotor resistance or by varying the rotor current/rl. The slip recovery system provides an ideal control,Ts, = max.I when, R2 = ,,X, It 4 -- SpeedSlipszFigure 6.46 Effect of rotor resistance on torquesiemploying a basic converter unit, supplemented by aninverter unit in the rotor circuit of the motor, as illustratedin Figure 6.47. The inverter unit controls the power flowfrom the rotor to the mains, thus acting as a variableresistance. The stator operates at a fixed frequency.The inverter may be a current source inverter, ratherthan a voltage source inverter (Section 6.9.4) since itwill be the rotor current tu that is required to be varied(equation (1.7)) to control the speed of a wound rotormotor, and this can be independently varied through thecontrol of the rotor current. The speed and torque of themotor can be smoothly and steplessly controlled by thismethod, without any power loss. Figures 6.47 and 6.48illustrate a typical slip recovery system and its controlscheme, respectively.The major difference in this configuration from that ofa V/fcontrol is the variable voltage and frequency fromthe rotor circuit that is first converted to a d.c. voltageand then inverted to a fixed frequency supply voltage inorder to feed the slip power back to the supply source.The converter-inverter combination acts like a variablecurrent source and in turn like a variable resistance. Thepower saving by this method is twofold. First, the powerloss in the external resistance is totally eliminated andsecond, the rotor power is fed back to the main supplysource. This system has a very high initial cost and istherefore preferred for large wound motors above 250kW. Nevertheless, it is advisable to employ a slip recoverysystem even for lower rated motors which are requiredto perform frequent speed variations. Also the regularpower saving would offset the heavy initial cost in the

www.taghedanesh.ir6/142 Industrial Power Engineering and Applications HandbookMain supply- -TT-T IVariable AC voltageand frequencyconverter(IGBT or thyristor)DC voltage tofixed frequencyand voltagesupply inverterFigure 6.47 Slip-ring motor control showing slip recovery systemfirst few years only and then provide a recurring energyand cost saving.6.16.4 Application of solid-state technology inthe operation of a process plantWith the use of static drives for speed control of inductionmotors, through open or closed-loop feedback controlsystems, it is now possible to monitor and control aprocess line automatically, which would not be feasibleif carried out manually. We will consider a simple processline of a continuous galvanizing plant to demonstrate theapplication of this technology in automatic and accuratecontrol of a process industry.The total engineering of such a system will first requirea thorough study of the process, dividing this processinto various activities and then monitoring and controllingeach activity through these controls to achieve the requiredprocess operation.Figure 6.49 illustrates a continuous hot-dip galvanizingline to perform zinc coating of MS sheets so that theproduction line has no discontinuity even when the supplyof sheet is exhausted, or during a changeover from onefeeding route to another or at the finishing line during achangeover from a completed roll to an empty one. Allthis is possible with the use of this technology as describedbelow.The process line indicates only the vital areas. Theremay be many more auxiliary drives and controls, interconnectedwithin the same process line, to adjust theprocess and its quality more closely. They have not beenshown in the figure for the sake of simplicity. We do notdiscuss the duration of one cycle, its speed, the temperatureof the furnace or the hot dip zinc vessel etc. and otherimportant parameters. These all are a matter of detailedengineering and process requirements. We describe theprocess only broadly, to give an idea of applying thistechnology to a process line for very precise control. Weemploy encoders to give a pulse output of speed of aparticular drive, and PLCs* to implement the processlogics through the various drives.*PLC - Programmable logic controller (the registered trade markof Allen Bradley Co. Inc., USA).3lsoiators 0@ 3-g slip-ring motor@ Tacho generator@ Starting resistor@ Variable ax. voltage and frequency IGBT or thyristorconverter@ Inductor@ D.C. voltage to fixed voltage and frequency diodebridge inverter@ CTs@ Dedicated transformer@ Speed potentiometer@ Speed comparator@ Speed amplifier and controller@ Current comparator@ Current amplifier & controllerFigure 6.48 Typical block diagram of a large slip recovery system using IGBTs or thyristors

www.taghedanesh.irStatic controls and braking of motors 61143The use of PLCs is essential in the control of motorsto closely monitor the operating parameters of the processline on which the motors are connected. The inverterunit controlling the motors then conducts the requiredcorrection in each motor speed through its switchinglogistics, which may be activated by the motor side V,f,I,,,, I;, or p etc.. depending upon the inverter logisticsbeing used. All such parameters are predefined for aparticular process and are preset.A brief description of the process and the use ofstatic drivesWe have divided the total process line into three sections:Uncoiler sectionI Pay-off reel no. 1 feeds the raw MS sheet to theprocess section via feed pinch rolls nos I and 2which straighten the sheet before it enters the welder.2 To maintain continuity and achieve an uninterruptedprocess line a second parallel feed route is providedthrough a second pay-off reel no. 2.3 These rolls are driven by motors MI and MZ whosespeed is controlled through the tension of thetravelling sheet. The tension of the sheet is adjustedby monitoring the diminishing diameter of the payoffroll and the thickness of the sheet.3 The pay-off roll is unwound by the tension of thesheet, caused by the speed of the recoiler at thefinishing line and the bridles positioned at differentlocations. The pay-off roll motors therefore operatein a regenerative mode and can feed-back the energythus saved to the source of supply, if desired. Thiscan be done by using a full-wave synchronousinverter, as shown in Figures 6.31 or 6.33.However, this is a more expensive arrangement.A more economical method is to use a full-wave,diode bridge converter and an IGBT inverter unitcombination as shown in Figure 6.50 in place of' anadditional thyristor or IGBT feedback circuit. Thed.c. link bus is now made a common bus for all thedrives operating on the process. During a regenerativemode, such as during uncoiling the pay-off reels,the voltage of the d.c. bus will rise and will be utilizedto feed the other drives. This process will draw lesspower from the source. The regenerative energy isnow utilized in feeding the process system itselfrather than feeding back to the source of supply.There is now only one converter of a higher rating,reducing the cost of all converters for individualdrives and conserving regenerative energy again ata much lower cost.There is no need to introduce a resistance for thepurpose of dynamic braking for the individual drives,but a large resistance will be necessary on the d.c.bus to absorb the heat energy during shutdown(braking) of the process. The useful energy duringshutdown cannot be fed back to the source due tothe configuration of the converter-inverter combination.This arrangement can feed the regenerativeenergy to its own process only.The scheme will also facilitate conservation ofregenerative energy if there are more of such driveswithout any additional cost.5 Shear no. I is used to shear the edges of the sheet ofpay-off reel no. I at the beginning as well as at theend of the coil before it enters the welder to smooththis for a correct welding with the outgoing edge.At the beginning of the coil this edge is welded withthe tail end of the previous coil and at the end it iswelded with the edge of the fresh coil at the beginningfrom pay-off reel no. 2. Pay-off reel no. 2, driven bymotor M1, is arranged parallel to pay-off reel no. Ito provide a second feed route for an uninterruptedand continuous process flow.6 The deflector roll guides the sheet to another pinchroll no. 4 to carry out precise alignment of the edgesbefore they enter the welder.7 Pinch roll no. 4 aligns the edge of the sheet througha feedback control.8 For further alignment of the edge width just beforeentering the welder the sheet is guided again throughan entry-side guide.9 With the help of bridle no. 1 driven by motors Miand M4, the uncoiler section speed is controlled bymonitoring the tension of the travelling sheet andhence maintaining constant speed of the sheet in theuncoiler section. The tensile difference of T, and T2determines the speed of the uncoiler. Speed andtension of the sheet must remain constant for absolutesynchronization between the uncoiler process andthe recoiler sections.10 To allow for welding time, a buffer of a certainsheet length, in the form of entry accumulator, ismaintained, generally in a vertical formation, to savespace. This feeds the line ahead until the weldingoperation is completed and the second route isinstalled to feed the process.11 The time gap in carrying out the welding 1s . compensatedby raising the speed of the second routenow introduced until the predefined buffer of anexcess length of sheet is produced with the help ofaccumulator drive motor M5.Process section12 The tension and speed in the process section ismaintained again with the help of bridle no. 2. drivenby motors M, and M,.13 The sheet is now fed through a pair of guide rollersto a furnace section through a degreasing tank, whereit is preheated for drying and raising the temperatureof the sheet up to a required level (40046SoC typical)before it enters the hot galvanizing pot for the desiredthickness of zinc coating. The movement of the sheetthrough the furnace is helped by motors M8 and My.14 The hot-treated sheet is cooled to the required levelby fans, driven by motors M,o and MI, before itenters the molten zinc pot.15 The required thickness of zinc coating is achievedby dipping it in the molten zinc pot for a presettime. The thickness of the coating is monitored bytwo rollers through which the coated sheet is passed.The time of welding. degreasing. heating and hot

www.taghedanesh.ir6/144 Industrial Power Engineering and Applications HandbookSetposition Tensiometer4Tensioi( I ,I l lI l lI l lI l lI l lI l lI l lI l lI l l1roll-I roll-2 rollpitsV ’-Uncoiler section@ E - EncoderProcess section(Fixed speed and tension)dipping are synchronized so that the weldingoperation takes the same time as the degreasing,heating and hot dipping.16 Bridle no. 3, driven by motors MI2 and MI3, is theprocess section master controller and controls thespeed and travel of the sheet in the process section.Coiler or the finishing sectionThis section is almost the same as the uncoiler section.17 The hot galvanized sheet is cooled by blowers andfed to an exit accumulator driven by motor MI4,similar to the entry accumulator, via a deflector rollto adjust its position.18 Before the finished sheet is finally cut into lengthsas required or rolled into recoilers its exit speed andtension is monitored and controlled again by bridleno. 4, driven by motors MI5 and MI619 The sheet may also be inspected for the quality ofwelding and checked for pin-holes by a weld-holedetector before it is finally cut into lengths or woundonto rolls.20 It then passes through a pair of guide rollers, pinchrollers no. 5, a shear, an edge guide unit (to alignthe width of the sheet) and a final alignment pinchroll no. 6 as shown.21 The recoiler is driven by motor MI-, that adjusts itsspeed and tension as calculated for the whole processline. It is this drive and the bridles that maintain therequired tension throughout the process and makethe pay-off reel drives operate as regenerative units.NoteWe have considered all drives as a.c. although d.c. drives arealso in use. Earlier only d.c. drives were used.IGBT inverters (the latest in the field of semiconductor devices)have been considered.All activities can be monitored through a control desk.All controls and precise adjustments are carried out by a PLC.6.16.5 Other applicationsIn view of the low maintenance cost and high reliabilityof solid-state devices, combined with precision and

www.taghedanesh.irStatic controls and braking of motors 6/145accuracy through a feedback network, this technologyfinds many applications in such process industries as thefollowing:TextilesMaterial handling0 Steel rollingOn- and off-shore drilling platforms0 Machine toolsProcess industries such as sugar, printing machinery,cement mills, chemicals, paperThermal power plant auxiliaries such as flow controlof primary air fan, ID fan and forced-draught fans,boiler feed pumps circulating water pumps andcondensate pumps, coal handling plant (e.g. ball mill,wagon tippler, and stacker reclaimer)Mining. A solid-state semiconductor device has nophysical contacts to make or break the current. Thereis thus no arc formation during switching of thesedevices. Therefore they have a very wide applicationin mines and other hazardous areas using flameproofa.c. motors where d.c. machines cannot be used.6.17 Speed variation throughvariable-speed fluid couplingsThe speed of the motor can also be varied by a variablespeedfluid drive as discussed in Section 8.4.6.18 Static drive versus fluidcouplingVariable-speed drives are essential for many industrialapplications requiring variable operating parameters duringthe course of operation. Such variations can be in theflow of fluid and pressure of air or gas etc. The con-

www.taghedanesh.ir6/146 Industrial Power Engineering and Applications Handbookl/C-lsw* Interlocked so that only twocan be made on at a time.CT >ASSwire A.C. busFuse swlCT p@ASSOEmergency PBOStop PBswlFuse,,FASS0 Emergency PB0 Stop PBBus-couplerswor MCCB7’ventional method of throttle control through a vane or adamper causes a considerable waste of energy. To obtaina variable speed and yet save on energy, one can useeither a static drive as discussed earlier or a variablespeed fluid coupling (Chapter 8). The choice betweenthe two will be a matter of system requirements and anoverall assessment of the ease of application, economyand accuracy of speed control in addition to the amountof energy it will be able to conserve. The applicationengineer would be a better judge to make a moreappropriate choice between the two, based on systemrequirements. We give a brier comparison between thesedrives in Table 6.5. This comparison should help in makinga more judicious choice of drive.Fluid couplings larger than, say, 400 kW can beemployed with ease in achieving the required speedcontrol. They also provide a mechanical coupling betweenthe drive and the load. A variable-speed fluid couplingalso results in energy saving along the lines as discussedin Example 6.1. In some countries this device is classifiedas ‘energy saving’ and attracts subsidies from the state.They are simple to operate and are highly economical intheir initial cost compared to static drives. The only caseagainst their use, despite the exorbitant costs of staticdrives, is their recurring power losses, as mentioned inTable 6.5 (items 19 and 20). It is a constant and recurringdrain on useful energy. A static drive, irrespective of itscost, has a pay-off period of three to four years, dependingupon the size of drive, frequency and duration of speedcontrol and its accuracy. Thereafter it achieves regularand high energy saving. This advantage is not possiblein a fluid coupling due to higher power losses.Conservation of energy is therefore the main criterion inthe selection of static drives.Until a few years ago, when static technology was stillin its infancy, variable-speed fluid drives had very wideapplication. With the advent of static technology, thetrend is shifting in favour of static controls, particularlyfor drives that have to undergo frequent and wide speedvariations during their normal course of operation, orwhich require very accurate speed control. For applicationsnot needing very precise speed control or widc variationsin speed (e.g. high-capacity pumps or ID fans) variablespeedfluid couplings are still the best choice.

www.taghedanesh.irStatic controls and braking of motors 6/147CT t-w ASSInterlocked so that only twocan be made on at a time.“iFusesw$0 FuseCTASS8 Emergency PB8 Stop PBsw$0 FuseC T p ASS @0 Emergency PB8 Stop PBIFigure 6.50 Power distribution arrangement for the galvanizing process of Figure 6.496.19 D.C. drivesThe use of d.c. motors for precise speed controls is longpractised and it had been a unanimous choice until a fewyears ago. It still is, in a few applications, purely on costconsiderations. But its use is now gradually waning outin the face of a more advanced technology in static controlsto control an a.c. motor (for very wide and accuratespeed variations, a variable-speed fluid coupling maynot be suitable). But older installations still use d.c. motorsand may continue to do so for a few more years, until thenext modernization of the installation, although manyleading manufacturers have discontinued the manufactureof such machines due to a sharp decline in demand.There are a few that are still in the field and may continueuntil there is a demand for replacements and extensionsof existing load lines. A few manufacturers who havediscontinued the production of this machine haveestablished links with those still in the field to cater forreplacements. Therefore this book has not dealt with thismachine in more detail.To make a better comparison between an a.c. and ad.c. drive we illustrate in Figure 6.51 for a d.c. motor,the likely variation in its torque, with variation in theapplied voltage, below the base speed and with a constantvoltage but variable field strength, above the base speed.6.20 BrakingBraking results in heating, irrespective of the methodused. When the braking is external, the heat will appearin the external circuit and the motor windings will remainunaffected. But when it is internal, the entire brakingheat will be generated within the motor windings. Dueconsideration of this must be made when selecting themotor rating, particularly when the loads are heavy andthe braking frequent. An analogue to the starting timegives the braking time tb asGD: -Ntb = 7 seconds (s)375 Tbwhere N = I?, - Nrl, i.e. speed reduction in r.p.m.Tb = braking torque in mkg

~~~~~www.taghedanesh.ir6/148 Industrial Power Engineering and Applications HandbookTable 6.5 Comparative study of performance of an a.c. drive, variable speed fluid coupling and a d.c. driveFeatures1 Manufacturing range2 Starting accelerationA.C. drive Variable speed.fluid coupling D.C. driveAny range-Soft and stepless start40 to IO 000 kW by GE, USAIn India up to 3000 kW byPembril Engineering andFluidomatSoft and stepless start0.1 to IO 000 kWSoft and stepless start3 Starting current (I,J~.Can be controlled to any desired levelsubject to the minimum T,,requiredCannot be controlled. Althoughduration of thc starting inrushis very short because of no-loadstartCan be controlled to anydesired level. Normally up to1.5 times the rated current4 Starting torque (TJ As desired As desired As desired up to 5-6 times theT,, through field forcing (i.e. byraising the field voltage duringthe start up period)5 Variation in torque withspeedThrough Vlfcontrol, the torquecan be keut constantup to a desired speed, N. Sinceh.p. = T.N., h.p. varies withspeedBeyond N, the h.p can be keptconstant by keeping the voltagefixed and raising f (h.p. = T.N).The speed-torque characteristic issimilar to a d.c. machine as shownin Figure 6.51T - N'Characteristics are similar to aninduction machine as shown inFigure 6.51:(i) Up to the base speed N, torquecan be kept constant througharmature voltage control,V = Id, field current (I, = @),remaining constant(I-..-- @"a - constant, I ~ ,Nbeing the' armature current)(ii) Beyond the base speed N, h.p.can be kept constant (h.p. =T.N.) by reducing the fieldcurrent (I, = $) and keepingthe armature voltage asconstant (I, = constant).Torque will now diminishexponentially, since @diminishes and N rises, Nbeing in denominator, has thesame effect as @ (Fig. 6.51)6 Magnetising losses ofthe motorVary with f Remain at 1004% Remain constant for speedvariations within the requiredspeed N, as the field current iskept fixed and only thearmature voltage is varied. Forspeed variations beyond N,however, when the armaturevoltage is kept constant and thefield current is varied, themagnetizing losses also vary7 Copper (PR) losses Low at reduced speeds, due to lowmagnetizing current (I,) andcorrespondingly lower I,No reduction because of samemagnetizing losses andtherefore relatively higher I,At lower speeds more than theax. drives as I, remains thesame and a relatively higher I,8 Power factorAlthough I, is low, overall p.f. may Low as I, remains the same Lower than the a.c. drivebe slightly lower on the line side,because of same field currentbecause of harmonic contents andinclusion of L. L is introduced to(i) Limit the current harmonics and(ii) Limit the rate of rise of current,i.e. ripples (dildr)9 Combined efkiency ofthe motor and the driveAt rated speed 90% and more.Reduces slightly at lower speedsbecause of poor efficiency of the(i)At rated speed = 87-90%(coupling efficiency ashigh as 97-98%)At rated speed up to 80-90%.At lower speeds reduces morethan the a.c. drives because of

www.taghedanesh.irStatic controls and braking of motors a149Features A.C. drive Variable-Speed fluid coupling D.C. drivemachine at lower speeds. Losses in (ii) At two thirds of the inputthe a.c. controls do not normally speed 50%exceed 0.5-1.5%(iii) at 20% of the input speed= 66% See Figure 6.52fixed field lossesIO Voltage dip during start- Nil High because of same Zst but NilUPfor a very short duration as themotor picks up lightly11 Fault level LOW High because of high Zst LOW12 Any cost reduction in Yes; because of lower capacity of Similar cost reduction possible Yes, as in a.c. drives.electricals (motor, motor, cables and switchgears and a but all requirements to becables, switchgears low fault level suitable for slightly higher faultetc.).level because of high Z,.~13 Range of speed control Very wide and stepless up to zero Moderate to accurate, Very wide and stepless as forsped depending upon the accuracy of a.c. drivescontrols. Stepless up to 20% ofN, at constant h.p. and up to33% of N, at constant torque ispossible. Pumps, ID fans etc.,that call for speed variationduring a process need may notnecessarily be too accurate. Orvariation in flow of fluid, gasor temperature etc. not callingfor very accurate controls, thatsuch drives find their extensiveuse. It may be made moreaccurate, but at higher cost ofcontrols14 Accuracy of speed Up to & 0.01 % in open-loop and Moderate to precise controls as Very accurate speed controls upcontrol f 0.001% in closed-loop control for ax. drives possible with the to k 0.01%systemsuse of microprocessor-basedcontrol systems15 Monitoring of operating Very accurate controls through Moderate to microprocessor- Same as for a.c. drivesparameters microprocessor-based closed-loop based, fully programmablefeedback control systems logic controls and feedbackcontrol systems are available,to provide smooth speedcontrols as good as a.c. drives16 Acceleration and braking Possible Possible Possible17 Reversal Possible Not possible. Although coupling Possibleis bidirectional, it can be run inany one direction only18 Inching Possible Possible Possible19 Power loss No loss except in the form of motor Relatively higher losses Losses are high because ofinefficiency at lower speeds because of field system, but comparatively(i) High starting current much less than fluid and eddy(ii) Coupling slip up to 15-16%at two thirds the input speedand about 20% at 20% of theinput speed (slip reduces atlower speeds as illustrated inFigure 6.52current couplings. At lowerspeeds the losses rise in theform of motor inefficiency20 Energy saving Optimum saving up to 100% Good saving. But low compared Slightly low, 90-94%. because(no loss)to a.c. drives because of high sliplosses. As it saves energy, it isalso entitled to state suhsidiesof field losses (up to 5-776)

www.taghedanesh.ir6/150 Industrial Power Engineering and Applications HandbookFeatures A.C. drive Variable Speed fluid coupling D.C. drive21 Harmonics They generate high harmonics nk ? There arises no such1 (Section 23.6(b)) and must be phenomenonsuppressed by using filter circuitsHigher than even a.c. drives,because of commutator arcingwhich it produces at highfrequencies22 Unfriendly environmental Not suitable. The static controls mustconditions such as be located away from such areas industladen areas, fire well-protected roomshazardous, corrosive andcontaminated locations23 Heating (i) Moderate in transistor controlsbut(ii) Excessive in thyristorcontrols. The higher the ratingthe higher the temperaturerise. They require extensivecooling arrangements whichmay be external or forcedNo problems as it is a sealedunit. But for controls whichmay be located separatelyVery high at lower speeds andrequires extensive coolingNot suitable at all, as the motoritself cannot be relocated insafer areasModerate24 Maintenance and May be high. For efficient Very low maintenance. Where adowntimemaintenance high-skilled operatorsand their proper training are essential,besides stocking enough spares asrecommended by the manufacturer.Yet the expert services of themechanical ruggedness isrequired, a fluid couplingprovides a more reliablesolutionmanufacturer may sometimes benecessary. At times, the spares maynot be readily available or it maynot be possible to repair themimmediately. In such cases either itis a total shutdown or the controlshave to be bypassed and the machine__run on DOL_. -~25 Cost consideration A costly arrangement It is generally cheaper than astatic drive in all rangesDowntime due to brushes isabout 1 hour only in 6 months’to a year’s time. However,meticulous monitoring ofbrushes and commutatorcondition is important and soalso the environmentalconditions. Controls will alsoneed similar attention as forax. drives. In a breakdown, themotor cannot be run in any way(unlike ax. motors) and wouldlead to a total shutdownQuite economical compared toan a.c. drive. The followingmay give a rough idea:Up to 10 kW - price differencenot significantAbove 10-40 kW- a.c controlsmay be costlier by 3MO%Above 40-130 kW - a.c.controls may be costlier by40-50%In the range of 500 kW andabove - a.c. controls may becostlier bv 2.5 to 3 times andeven more in still higher ranges

www.taghedanesh.irStatic controls and braking of motors 6/151Armature vottage control(Field strengtheniq region)Constant toque region- Va e, Ta 0 ./, = ConstantField contml(Field weakening region)Constant HP regionArmature voltage fixedfield current reducedla = Armature current@ = Torque cutve@ = output cum-0' NBasa speedSPeed(=9Base speed - It is normally the rated speed at whichthe rated parameters are referred(T,, HP and V,)Figure 6.51 Variation of toque with speed in a d.c. machine (the same as for an a.c. machine)251-l-rlI I IA. External: mechanical or friction brakingThis type of braking is suitable for small motors and canbe achieved through1 Solenoid-operated brakes,2 Electro-magnetically operated brakes, or3 Magnetic particle brakes.Figure 6.52 An approximate illustration of 7 and loss variation,with change in speed in a variable-speed fluid couplingAn analogue to starting heat (equation (2.10)) gives thebraking heat Hb asGD;H --. (Np - N i ) watt-seconds (Ws) (6.9)- 730The wider the speed range of braking, the greater will bethe heat generated.6.20.1 Types of brakingThere are several methods of braking, external or internal,and they are briefly discussed below. Any of them can beemployed, depending upon the torque requirement, i.e.size of motor, its speed, the type of load, etc.In the first two types a brake shoe, operated by an externalauxiliary supply, is mounted on the extended shaft at theNDE (non-driving end) of the motor. These brakes arenormally operated after the motor is switched OW. Theheat of braking appears in the external circuit and themotor windings &e not affected. For motors with thisbraking, only the starting heat need be considered,depending upon the frequency of starts and not the heatof braking.Nofe Friction braking may be employed for all sizes of drives,either as the only braking means as noted below, or as a supplementarysafety means to keep the drive locked stationary when required.1 AC solenoid brakes These are employed for smallmotors, say, up to 15-20 h.p. They are suitable forapplications such as conveyors, hoists, cranes, machinetools, lock gates and dumb waiters (Figure 6.53). Thebrakes are spring loaded and mounted on twomechanically opposing brake shoes. They grip a brakedrum or disk, coupled rigidly at the NDE of the motorshaft. The brakes are applied mechanically and releasedelectrically. The braking action takes place by deenergizingthe spring. The brakes are normally appliedin the OFF position for reasons of safety in the eventof a power failure. They are released only when thesolenoid is energized.

www.taghedanesh.ir6/152 Industrial Power Engineering and Applications Handbooklowers the load to the ground loading station or the desiredplatform, as the situation may require in the event of a powerfailure.The ratings of the brakes noted above are only indicative.The braking torque of the shoe brakes may diminish with thenumber of operations. The heat of braking wears out the brakelinings. The extent of fading will depend upon the braking torqueto decelerate the heavy loads and frequency of its operations.They may also need replacement of the brake linings, similar toan automotive vehicle.Figure 6.53 AC solenoid brake (Courtesy: BCM)Magnetic particle brakes One type of these brakes isillustrated in Figures 6.55(a) and (b). They are alsoknown as powder brakes and have a main body (stator)that houses a drive cylinder, forming the main rotatingpart of the brake. Through its extended shaft is coupledthe main drive that requires the braking facilities.Within and concentric to the drive cylinder is a rotorrigidly fixed with the housing. There is a space betweenthe drive cylinder and the rotor, which is filled withsmall granules of steel in the form of powder, withexcellent magnetic properties. This powder, whenmagnetized, condenses into a solid mass between the2 Electromagnetic shoe-brakes These are similar tothe above, but are used for still higher motor ratings,say, 5-800 h.p. (Figure 6.54). In this case instead ofa solenoid coil, an electromagnetic coil is employed.This releases the brakes and develops a torque at leastequal to the motor torque, to brake or hold the fullload. In this case also, the brakes are applied on themotor shaft when the holding coil (electromagnet) isde-energized and is released only when the electromagnetis energized to make it safe against failure.Possible applications include cranes, hoists, elevators,conveyors, machine tools, rolling mills and ball mills,etc. and also holding of loads in conveyors, hoistsand elevators, etc.Note1 In both the above types of braking systems, a hand-operateddevice is also provided, to release the mechanical brakes inapplications such as lifts, elevators, cranes, and winders. ThisFigure 6.55(a)A magnetic particle brake (Courtesy: Dynaspede)*Figure 6.54!-Electro-magnetic shoe brake (Courtesy: BCM)Figure 6.55(b) Cross-section of a typical magnetic particlebrake

www.taghedanesh.irStatic controls and braking of motors W153drive cylinder and the rotor and provides the requiredbraking effect. This is possible with the help of amagnetic field which is provided through a stationarymagnetic coil placed in the main housing outside thepexiphery of the drive cylinder as shown. The fieldstrength of this coil can be varied with the help of avariable current source to obtain a variable brakingtorque and thus achieve more precise braking control,even remotely. Depending upon the type of applicationand accuracy of the speed control desired, extremelyprecise and accurate electronic controls are available.These can infinitely vary the torque and hence thespeed of the motor. Such braking devices are availablein the range 0.1 kW-60 kW.Strength of brakesThe brakes should be suitable to counter at least thetorque developed by the motor. They must thereforedevelop at least this amount of torque. To find the leastbraking torque, the brake drums must be able to develop,in either of the above types of mechanical brakes, thetorque shown in equation (1.10) may be used i.e.0 25 50 75 100%Speed---,Figure 6.56 Typical braking torque curves for a wound rotorfor different external resistances but Same excitation currentThe brakes must develop at least this amount of torqueor slightly more, i.e.where Tb is the torque of braking(6.10)B. Internal type1 Electrodynamic or d.c. electrical braking When a d.c.voltage is applied to the motor windings, a steady fluxis produced since f = 0. The theoretical synchronousspeed of the motor, N,, now reduces to zero. Whenthis steady flux is cut by the rotor conductors, as therotor is rotating, it induces a steady (d.c.) e.m.f. in therotor circuit, which produces the required braking effect.In slip-ring motors, the braking torque can be controlledby inserting suitable resistance in the rotor circuit andvarying the excitation voltage (Figure 6.57), keepingthe excitation current the same. Braking in slip-ringmotors by this method is more accurate and simple.Some typical braking curves are shown in Figure 6.56for a slip-ring motor.In squirrel cage motors, in the absence of externalresistance, the stator windings can be arranged indifferent configurations such as series, parallel, star ordelta, as shown in Figure 6.57, to achieve the varyingeffects of excitation voltage. This type of braking isuseful for both squirrel cage and slip ring motors, butis rarely used.For applying the brakes, the stator is disconnectedfrom the supply and a d.c. excitation voltage is appliedto the windings as shown in Figure 6.57. The windingscan be arranged in any configuration, as illustrated, toobtain the required braking torque. If the ampere turnsduring braking are maintained as during normal running,Fig.abdkRequireddc voltagee1.225 i&. 2R1.41idc .-3R22.45uShortede = Excitationvoltage.& = Excitation or braking currentR = Stator resistance per phase. For slip-ring motors,external resistance can be added and R varied. RI& ' - 2Figure 6.57 Stator or rotor connections for d.c. electric braking

www.taghedanesh.ir6/154 Industrial Power Engtneering and Applications Handbookthe braking torque curve will almost take the shape ofthe motor's normal speed-torque curve.If an independent d.c. source is not available a singlephasetransformer and a rectifier bridge as shown inFigure 6.58 can also be used to obtain the requiredd.c. voltage. Although the requirement of d.c. excitationvoltage is not high, the rating of the rectifier transformexand the bridge should be commensurate with the brakingforce required. This braking force would depend uponthe size of the motor and the time of braking. If thebraking current, idcr is known, which is a measure ofthe braking torque necessary to fulfil a particular loadduty requirement, the excitation voltage e can bedetermined for different winding configurations, asindicated in Figure 6.57. The ik can be determinedfrom the following equation, considering the sameampere turns as for a standard motor:wherez& = braking currentIst(phl = phase value of the starting current(4.1 1)= I, (for a delta-connected stator or rotor)4-3kl = factor to determine the equivalent ampere turnsv-for a particular configuration, as indicated inFigure 6.57, To avoid overheating and excessiveelectromagnetic forces, i& is normally notallowed to exceed ZaW)Te = average load torque between the running speedand the final speed {Figure 6.59)Tb = average braking torque between the runningspeed and the final speed (Figure 6.59). Thiswill depend upon the braking duty the motoris required to perform such as the final speed,Nrl (which we have considered as zero in Figure6.541, and the duration within which the motormust brake to this speed from N,. This can bedetermined from equation (4.8)T, = braking torque of the external brakes, if providedotherwise it may be. considered to be zeroT,, = locked rotor (starting) torque of the motorka = a factor to account for the average braking torque.This may be considered to be 1.3-1.7 (consultthe manufacturer for a more accurate value)In addition to electrical br ng, a mechanical brake,as discussed in Section 6.20.1(A) may also be essentialif the motor is required to be stopped completelybecause, at any value of excitation current, the motorwill never reach a standstill condition, The heat ofbraking up to the standstill condition {Nrl= 0) is roughlyequal to one start and is expressed by equation (6.9).2 Plugging By changing any two of the phases themotor will develop a torque in the reverse directionand provide the necessary braking. The voltage acrossI l l$ 1 8-SpsedTotel braking torque at point A = Te + TbIFigure 6.58Obtaining d.c. voltage through a bridge rectifierFigure 6.59Braking torque during d.c. electric braking

www.taghedanesh.irthe windings at the instant of plugging becomes twicethe rated voltage, and slip as 2S, for the changedmagnetic field. With these changed parameters, thecurrent and torque curves can be approximatelydetermined from equations (1.7a) and (1.3a) respectively,for high slip conditions.Current and voltage will both give a transitory kickat the instant of plugging, depending upon the effectivevoltage across the windings, under the influence ofthe motor’s self-induced e.m.f. and the applied voltage.The transitory state will last only a few cycles andthen the curves will generally take the shape as in theequations noted above and illustrated in Figure 6.60.Generally, except for the initial kick, there will be nosignificant variation in the current and torque valuescompared to their starting values at S = 1. These valuesCM be varied in slipring motors by altering the rotor’scircuit resistance. During plug-ging, if the supply isnot switched OFF at the instant of reaching the standstillposition, the motor will start rotating in the reversedirection, tracing the same speed-torque and speedcurrentcurves as in the forward direction. But a reversedirection may damage the driven load. Precautionsare essential to prevent such a situation by providingan electrical interlock-ing andor a reverse ratchetarrangement in the load coupling.The windings may, however, be subject up to twicethe rated voltage and must be suitable to withstandthis voltage repeatedly when necessary. The heat generatedduring braking will be roughly three times theheat generated during start-up as determined below:Rotor losses per phase W = I: . RZI: . RzRotor torque per phase T = -S:. Rotor loss per unit torque - W= ST3Static controls and braking of motors W155Average loss between slip SI and S2 =(i) During a normal running,when S1 = 1 and S2 = 0starting heatTa starting loss = - 2(ii) During plugging,when, S1 = 2 and S2 = 1Heat generated during plugging(Sl +S2) .*2Therefore the heat of the motor during plugging isthree times that of during a normal start. Stator heatand thus the total motor heat is a function of the rotorheat (see also Section 2.7.1). Such a method is thereforenot suitable for larger motors or for frequent brakings.Note This is an approximate derivation for a simple illustrationof the mtio of heats. The time of start and haking is not consideredin the above derivation, whereas both would be different and sowill be the heat generated. The time of start would be muchhigher than the time of braking, as the latter is much higherthan the former. Figure 6.60 illustrates this. But in view of thehigh current during plugging the ratio of heat as noted above isa near approximation.Regenerative braking If the motor be run beyondsynchronous speed by some external means it willwork as a generator and feed back useful energy tothe supply system. It will draw only the necessaryexcitation current, Im, for the generator action fromthe source of supply. In such a condition, the motorN, 0C Plugging (braking) -- Running ATotal braking torque at pdnt A = Tt + TbFigure 6.80 Approximate motor torque and current characteristic curves during plugging

I &www.taghedanesh.ir6/156 industrial Power Engineering and Applications Handbookwill exert a countertorque, the magnitude of whichwill depend upon the motor speed above synchronous.Such braking conditions may occur automatically indownhill conveyors, lifts and hoists etc. whiledescending with the load, i.e. operating as an inductionmotor while ascending and as an induction generatorwhile descending. The generator and the braking actionceases at synchronous speed. For speed control belowsynchronous speed, therefore, it will be essential toemploy a multi-speed motor which, at a higher speed,can be switched to the lower speed winding to makethe motor work as a generator between the high andthe low speeds. Such a braking method, however, hasonly limited commercial applications, as in a sugarcentrifuge motor (Section 7.4).With the application of solid-state technology,however, as discussed above, the potential energy ofthe loads in hoists, lifts and conveyors during descentscan be saved and fed back to the source.6.21 Induction generatorsDuring generator action, the slip, and currents of thestator and the rotor are negative. The motor draws reactivepower from the source for its excitation (magnetization)since it is not a self-excited machine. However, it feedsback to the supply system an active power almost equalin magnitude to the motor rating or slightly less or more,depending upon its supersynchronous speed. As aninduction generator, it can feed back to the supply sourceroughly equivalent to its h.p. at the same negative slip,say, 3-5%, as the positive slip, at which it operates undernormal conditions and delivers its rated h.p. As a resultof the absence of the reactive power, which is now fedby the source and the mechanical losses that are fed bythe wind, the power output of an induction generator isusually more than its power consumption when workingas a motor (Figure 6.61).The power factor, however, is poor because of highernegative slip. The power output is expressed in the sameway as the motor input, i.e.GI = J ~.I,.V.COS$.K (6.12)whereGI = generator output at the same negative slip asfor the motor. See also Figure 6.61 and thecircle diagram of Figure 1.16, redrawn in Figure6.62 for an induction generatorK = factor to account for the lower p.f. at highernegative slips when working as an inductiongenerator (say, 0.97).IG = generator rated current in A.cos Q = generator rated p.f. which is quite high comparedto a motor, as the reactive power is now suppliedby the external source.Example 6.4For a 100 kW, 380 V induction motor operating at anapproximate r~ of 92.2%, having I, as 185 A and cos @ as0.89, the output as an induction generator will beJr0B JQF520 100% 200%Motoring-Generating(Super synchronousspeed region)Speed/M - Motoring current drawn from the sourceIC - Generating or braking current fed to the sourceFigure 6.61 Current, torque and speed characteristics of aninduction generatorG = 43 x 380 x 185 x 0.89 x 0.97 w(considering r~ as = 100%)= 105 kWand the kVAR consumed by the induction generator from thesource of supply,x 380 x 185 x 4--0.97 x 1000= 57.24 kVArCorollaryThe power output of an induction motor, when operating asan induction generator, is usually more than its output whenworking as a motor. There are no mechanical or windagelosses which are fed by the mechanical power that makesthe motor run as a generator, such as the potential energyaccumulated during downhill conveying or wind power etc.The power output is approximately equal to the effectivepower intake except for the lower power factor and resistive(copper) losses, that are ignored above.The circle diagram (Figure 1.16) reverses in this case andthe magnitude of braking torque and corresponding statorand rotor currents can be ascertained at any particular speedfrom this diagram redrawn in Figure 6.62. A study of thisdiagram will reveal that for a motor’s normal running speedN, to reach the synchronous speed Ns, the motor will behaveas a generator without output (region DIPl) and will drawpower from the main supply to meet its no-load core lossesand friction losses etc. (DIPl). This will deliver active powerback to the main supply as soon as it exceeds its synchronousspeed. The maximum power that can be delivered is measuredfrom the no-load line of the motor to the output line of thegenerator (DlP2) minus the no-load losses (DIP,), i.e. thedownward hemisphere from the centre line or the generatoroutput line (PlP2).

www.taghedanesh.irStatic controls and braking of motors 611 570Motor output h e(generator input byFigure 6.62Motoringregion-Motor's no-load line B,DlP.Generator output lineCircle diagram for an induction generatorFigure 6.63 shows the generating region and torqueand current variation for a dual-speed motor when athigher speed the supply is changed over to the low-speedwinding. From the torque curve it is evident that thetorque reverses twice in quick succession through tworeverse peaks (curve area nbcd) and the motor must besuitable to withstand this. All the parts, subject to thisphenomenon should be carefully designed and selectedand rotor bars and end rings should be tightly securedand braced. Precautions against heavy centrifugal forcesto which the rotor conductors will be subject at supersynchronousspeed should also be taken into account. Ata higher speed beyond the synchronous (a higher rated-Dnegative slip) the higher will be the stator current andthe generator will run overloaded. To safeguard againstoverloading, relays or similar protections must be providedin the supply lines to disconnect the motor beyond aspecific speed (generally beyond the rated negative slip).after a certain time delay, if such a situation arises.In downhill conveyors, running mostly on storedpotential energy by gravity, the motor may overspeedbeyond excessive limits unless prevented by a brake or atachogenerator relay.Critical speedAt certain speeds, rotating masses become dynamicallyunstable and cause deflection and vibration in the rotorwhich may damage the motor. The speed at which suchinstability occurs is known as critical speed and occursat different multiples of the rated speed. The mmust therefore rotate within 20% below or above thecritical speeds to avoid such a situation. These vibrationssettle down again at higher speeds above critical andrecur at the next higher critical speed.To achieve a vibration level of 0 micron in a rotatingmass is practically impossible. It will possess some runout,however precisely and accurately it has been balanced.While rotating, therefore, the shaft on which the othermasses rotate, will deflect to the side which is heavierand the opacity around the centre of the shaft will becomeuneven. The masses will rotate in small circles abouttheir own geometrical axis rather than the axis of theshaft.In a motor the first critical speed is several times higherthan the operating speed and may not reach in operation.In downhill conveyors, without any mechanical speedcontrol, such speeds may be reached when the motoroverspeeds while descending. When so, the motor shouldbe disconnected from the supply lines, otherwise therewill be excessive overloading. There may be several criticalspeeds in a rotating mass which tend to become infinitewith the number of loads on the same shaft. But the firstcritical speed alone is of significance, as other criticalspeeds much higher than even the first critical speed areof no relevance. No rotating mass may possibly reachthis during operation in an induction motor.0 1000Speed (r.p.m.) +I, - Motoring current drawn from the source/G - Generating or braking, current fed to the sourceI 1 1Fti,l 500Figure 6.63 Motoring and generating torque curves for a1500/1000 r p.m. dual-speed motor (torque and current curvesfor generating region are drawn for only low-speed winding)Induction generator as a wind-powered generatorThe main application of this machine is found in nonconventionalenergy generation, such as for wind power,gas turbines and mini- or microhydro power generationetc. They are used extensively to convert wind powerinto electrical power, The wind power is Pirat convertedinto rotating kinetic energy by aerodynamically designedblades. This energy is then converted into electrical energythrough these machines. They are thus known as windelectricity generators (WEC). Presently such machincsare in use from 50 kW to 6 MW worldwide. In India theyare in use for up to I MW (mostly in the range 400-600 kW). Potential locations for such machines are coastalareas, whcrc thcre are high and continuous winds. InIndia examples of these locations are in the states ofMaharashtra, Kerala, Tamil Nadu, Karnataka and GLi.jarat.

www.taghedanesh.ir6/158 Industrial Power Engineering and Applications HandbookThe electricity generated depends primarily on the speedof the wind at the site of installation. A conventionalformula to determine the wind energy, based on the designof the rotor (rotating blades) and the site conditions isgiven byP = 0.5 . C, . A * p . V3* (6.13)whereP = power generated by the turbine (windmill) in wattsCp = coefficient of performance which depends uponthe aerodynamic efficiency of the rotor and varieswith the number of blades and their profile. Thisfactor is provided by the mill supplier and generallyvaries between 0.35 and 0.45A = swept area of the rotor in m2--KO2- 4whereD = diameter of the rotor (blades) in mp = air density = 1.225 kg/cm3V = velocity of wind at the site of installation, at theheight of the hub in mls* The ideal condition would be when the rotor outputis a cubic function of wind speed. But in practicethis may not be so. It is found to be linear or anear quadratic (square) function of the wind speed,as shown in Figure 6.64Typical specification for a KEC (India) 400 kW machineis provided below for a general reference:Cut-in speed = 4.5 mlsApproximate output at the cut-in speed from themanufacturer’s data (Figure 6.64) = 8 kWRated wind speed = 11.5 m/s, - Ratedspeed-0 415 6 7 8 9 1011 12 14 16 18 204.5Wind speed m/s (hub)Figure 6.64 Typical power curve for an induction generatorof 400 kW at 11.5 m/s wind speedMean wind speed = 25 kmph- 25 x 1000 m,s- 60 x 60= 6.94 rn!~Note Generally, the ratio of rated to mean wind speed may bequite high due to long lean periods, when the machine may stayidle. reducing the value of the mean speed.Shutdown speed = 20 mlsRotor diameter including hub = 39.35 mRotational speed of the rotor at the rated wind speed= 38 r.p.m.Example 6.5For the above machine the wind power considering the C,as0.35 will be:P = 0.5 x 0.35 x 1.225 x x x - 39.35’ 11.534= 396.3 kWBelow we give a brief idea of the mechanical system ofsuch a mill and its various controls, as a passing reference.For more details on the subject, see the Further readingat the end of the chapter.Mechanical systemSee Figure 6.65, illustrating the general arrangement ofa windmill.I Tower This may be tubular or lattice type to mountthe mill’s mechanism. The structural design is basedon the cutout wind speed.2 Nacelle This is the main housing of the mill, madeof metal or FRP and contains a rotating hub on whichis mounted the blades. Inside the housing is a gearbox,one end of which is coupled to the rotating bladesand the other to the generator. For optimum efficiencyof the mill, it is essential that the blades fallperpendicular to the direction of the wind. Toaccomplish this, the nacelle is made to rotate at thetop of the tower. It aligns to the required directionthrough a yawing mechanism which adjusts the rotorassembly so that when the blades are in motion theyfall perpendicular to the wind direction and, when atrest, at low and high cut-out speeds, they fall in linewith the wind direction for minimum stress.3 Blades These are made of wood and epoxycompound composite material or fibre glass. Theirmechanical strength is also commensurate with thecut-out wind speed. These blades are connected to arotating shaft coupled to the generator through a gearassembly. They may also be fitted with an additionalfeature of a pitch control mechanism through aservomotor or a hydraulic system. Such a systemcan rotate the blades around their own axis to adjusttheir angular speed with the speed of the wind. Thisfeature assists the gear system to provide a nearconstantspeed to the rotating shaft. It also helps in

www.taghedanesh.irStatic controls and braking of motors 6/159@ Main hub@ Main shaft@ Main shaft brake@ Gear mechanism0 Gear box@ Brake disc0 Coupling@) Generator@ Frame@ cup anemometer@) Lattice tower@ Foundation@ Aerofoil blades@ Control panel@ Transformerpower gridFigure 6.65General arrangement of a windmill'slipping' the excess power of wind to some extent,to optimize the operation and running hours of themachine during excessive wind speeds exceeding thecut-out speed. This fcature also protects the machinefrom excessive wind pressure (for example, duringstorms ).4 Knis This is a mechanism that helps the nacelle tomove in the right direction with the help of a yawingmotor or a hydraulic system. In this case the hydraulicsystem provides a smoother movement.5 Hydruulic power supply This actuates brakes andalso feeds the yawing and pitch control (if they arealso hydraulically operated).6 Brnkes These block the rotor at the cut-in and cutoutspeeds and also during maintenance:Cut-in wind speed is the minimum wind speed atwhich the generator commences the generation ofpower. At this speed the brakes release and theprime mover (blades) starts rotating.Cut-out wind speed is the maximum wind speedbeyond which the prime mover may overspeedabove its permissible limits. As the structure andthe blades are designed for a particular maximumspeed, a wind speed higher than this may exceedtheir mechanical endurance and become unsafe.At this speed the brakes apply and the machine isdisconnected from the grid. The cut-in and cut-outspeeds define the wind speed limits within whichthe turbine will work safely through the generator.Rated wind speed is the speed at which the primemover rotates at the rated negative slip and generatesthe rated power.Note To avoid unnecessary wear and tear of the machine.the blades are braked when the machine is not in operation.In fact the main protcction to the rriiichine is through thebrakes only. During an overloading or system fault conditionthe blades are braked and the machine ceases to generate.The control panel monitors closely all the operating

www.taghedanesh.irW160 Industrial Power Engineering and Applications Handbookparameters and sends out warning signals or stops the machinewhen operating conditions exceed the permissible limits.7 Gear system Since the wind speed is never thesame, and the rotor is never permitted to rotate at ahigher speed than its designed parameters, a gearsystem is provided between the blades and thegcnerator rotor to protect it from excessive wear andtear. The gear system helps the generator rotor to runat the generator’s rated speed. The normal gear ratioto boost the speed of the prime mover to the generatorspeed is from roughly 30-40 r.p.m. to a little overthe synchronous speed of the induction generator.The gear system is a two- or three-stage fixed ratiogearbox giving fixed output speeds. Fine adjustmentis achieved through pitch control of the blades.8 SensorsAnemometer -Temperature -sensorsWind vane -Vibration sensors -Pressure sensors -to monitor wind speedto monitor the operatingtemperature of the gearbox,generator windings and otherimportant parts of the machineryto monitor wind directionto monitor correct alignment,loose or worn-out parts etc.for the hydraulic system thatactuates the yawing mechanism,brakes and pitch ofthe blades9 Electrical systemMain power panel to receive power from the generatorand feed this back to the power gridStep-up transformer if the voltage of the generatoris different from that of the gridRelated switchgearsControl and relay panel (which may be microprocessorbased), to record, display, monitor andcontrol the generated power, voltage and frequencyand also detect fault conditions.Frequency inverters to regulate the voltage andfrequency etc.10 Induction generator This is a standard squirrel cagemotor with additional treatment to weather the siteconditions. The normal specifications are generallythe same as for a standard motor. The permissiblevoltage and frequency variations are, however, wideras noted below:Voltage f13%Frequency -3% + 3%or as may be required to suit a particular grid system.A frequency limit of +3% signifies that the machineshould not overspeed beyond 101.5% of its synchronousspeed. This is to avoid overloading of the machineas well as retaining synchronism with the grid. Forbetter use the machines are often wound for a dualspeed, such as 416 pole, so that when the wind speedfalls below the minimum cut-in speed, the lower speedwindings take over and the generator may still operateand feed back to the power grid at a slightly loweroutput (in the ratio of motor outputs at the two speeds).To achieve this, two switching circuits may beprovided, one for each speed as shown in Figure13.59 (Table 6.1).The changeover is obtained by measuring theaverage power generated during a particular timeperiod, say, one minute or so, rather than the speedof the wind. When this average power falls below apreset level the machine changes over to the lowerspeed windings and vice versa. Due to the unpredictablenature of the wind speeds, this may requirefrequent changeovers and may affect the reliabilityof a double-speed system.As soon as the wind speed reaches the minimumrequired speed (cut-in speed), i.e. a lower speed mode,in dual-speed machines the brakes release and theprime mover picks up speed. At about 95% of therated speed of the machine, the stator of the inductiongenerator is supplied with the necessary reactive powerfrom the grid through an inverter, to generate therequired magnetic field. At this speed, which is almostthe rated speed of the motor, the machine draws justenough reactive power from the grid to meet itsmagnetizing requirements and a small amount ofmechanical power from the wind power to meet itsfriction and windage losses (area DIP, of Figure 6.62).The magnetizing current is now much less than theno-load losses of the machine when it operates as amotor. The inverter is normally thyristor controlledto provide the machine with a soft start and to controlthe starting current to almost the rated current incase of excessive machine speed. It also avoids anexcessive voltage dip at the generator terminals andtime to hook up to the grid. As soon as the machineexceeds its synchronous speed, it starts generatingand feeding the active power back to the grid, similarto the output curve of the machine in Figure 6.64.(The curve is provided by the machine supplier.)The generator, when hooked up to the grid, will followthe grid voltage and frequency. The generator isregulated not to overspeed bcyond 101% of itssynchronous speed. Small speed variations are camedout through pitch control to optimize the power outputand running hours of the machine. The powergenerated is proportional to thc negative slip at whichthe machine operates (area PIPz of Figure 6.62).The total reactive power that it draws from the gridVAr =V, . I , wattswhereV, = rated voltage of the machine in volts, andI, = magnetizing current of the machine in amps.(Figure I. IS)Use of capacitorsPower capacitors are installed to compensate for thereactive power, particularly that which the machine drawsfrom the grid. The capacitors must be suitable for aminimum voltage of V, + 13% (generally V, + 15%).

www.taghedanesh.irStatic controls and braking of motors 6/161Active powerThe power the machine feeds back to the grid is expressedby equation (6.12) discussed earlier.Note A normal motor is designed for a slightly lower voltage thanthe system voltage to account for cable drops from the source ofsupply up to the motor. But as an induction generator, it is designedfor a slightly higher voltage than the grid voltage or the primaryvoltage of the transformer (when a transformer is installed to raisethe voltage of the wind generator to the grid voltage) to account forthe voltage drop from the generator to the grid or transformer.Micro sitingTo identify the correct location and size of a mill it isessential to ensure that there is adequate wind at highspeeds at the site. The process of identifying the mostappropriate locations is termed micro siting. Localgovernment and private agencies conduct surveys andcompile wind data for potential areas for these mills. Alist of these agencies in India is provided at the end ofthe chapter. These mills are generally installed away fromareas of habitation for maximum wind and safety forpeople and animals or in remote areas not connectedwith power lines. The power so generated may be captiveto supply nearby areas or fedback to a power grid toaugment its capacity. In remote areas where no powergrid exists the power so generated may also be stored inbatteries through a voltage source inverter and utilizedwhen required. Such a system, however, is more convenientfor small mills, say, up to 5 kW, otherwise the cost factormay act as a deterrent to such an arrangement.Normally such mills are installed in groups known aswind farms to provide a sizeable power source, except inremote areas, where power demand may be restricted toa very limited area and small mills may suffice. Whenmills are installed in groups, precautions are necessaryto ensure that there is enough distance between any twomills so that there is no hindrance to routine maintenance,on the one hand, and obstruction of wind to other mills,on the other. For more details, refer to the literatureavailable on the subject in the Further reading at the endof the chapter.6.23 Number of starts and stopsDue to excessive starting and braking heat losses it isnot advisable to switch an induction motor ON and OFFfrequently. The number of starts and stops a motor iscapable of performing will depend upon its workingconditions such as type of switching, braking and loaddemand etc. and can be determined from(6.14)whereZL= number of starts and equivalent stops per houron load. For example, in plugging one start andone stop will mean four starts and if reversal isalso involved then five starts.ZNL = permissible number of starts per hour for a motorwith a free shaft, using mechanical braking, thusplacing no strain on the motor. This factor willdepend upon the electrical and mechanical designof a motor and will vary from one manufacturerto another. The cooling capacity, its effectiveness,i.e. heating and cooling characteristics and startingtorque of a motor, are the parameters that woulddetermine this factor. The smaller the motor, thegreater number of starts it will be capable ofperforming. For a lower-speed motor, the averagestarting torque will be normally less and the inertiamore. Therefore the permissible number ofswitching operations will be comparatively lessfor a low-speed motor than for a high-speed motorof similar rating. As a rough guide, small motors,say, up to 20 h.p., may have a factor as high as1000-2000.1 .a0.8I-6.22 Inching or joggingThis means repeated short-duration application of powerto the motor to cause small movements of the shaft fromrest to perform certain load requirements. The motormay normally not reach its full speed, nor at timescomplete even one full revolution, and can be rotated ineither direction. Likely applications may relate to liftingor hoisting which may call for delicate handling and arather slow, smooth and more accurate final movementfor exact positioning, lifting or unloading etc.This is a severe duty for the switching contactors asthey have to endure repeated arcing of the interruptingcontacts every time they make or break. (Select onlyAC-4 duty contactors: see Section 12.10.)T o’620.40.20 0.2 0.4Loading Tt I c-5 0.8Figure 6.66 Average load factor KL when started againstloadD

www.taghedanesh.irW162 Industrial Power Engineering and Applications HandbookKb = factor of braking, which depends upon the typeof braking used, such as(a) Mechanical braking Kb = 1(b) D.C. braking Kb = 0.6-0.5(c) Plugging Kb = 0.3-0.4(d) Regenerative braking Kb = 0.4-0.5KL= mean load factor, i.e. the ratio of the averageload torque to the motor torque which dependsupon the loading on the motor during start-up.For most applications (e.g. cranes, lifts or machinetools) this factor is based on a loading of 0.5 or0.75. This factor is also determined by themanufacturer and may have a shape as shown inFigure 6.66:FI =GD& + GD:GD&Relevan. StandardsIEC Tftle IS BS60034- 111996 Rotating electrical machinesRating and performance60034-17/1998 Cage Induction motors when fed from convertersApplication guide4722i1992 BSEN 60034-325/1996 11199560146-1-1/1991 Specifications of basic requirements for power 14256/1995 BSEN 60146-1-converters /199360439- 1/1992 Low voltage switchgear and controlgear assemblies. 8623 (Part-I) 1993 BS EN : 60439Type tested and partially type tested assemblies10994Code of practice for selection, installation and 10118 (Part-HI) 1982 -maintenance of switchgear and controlgear: Part 3InstallationRelated US Standards. ANSVNElMA and IEEEANSMEEE-C62.4 1- 1991ANSYIEEEC-37.21/1998ANSMEEE-5 19/1993NEMAIMG-7/1993NEMMCS-1.111988NEMMCS-7/1993NEMMCS7. 111995NEMAIICS-9/1993NEWS- 1/1992Recommended practice for surge voltages in low voltage a.c. power circuitsStandard for control switchboardsGuide for harmonic control and reactive compensation of static power convertersMotion/position control motors and cablesSafety guidelines for the application, installation and maintenance of solid state controlIndustrial control and systems. Adjustable speed drivesSafety standards for construction and guide for selection, installation and operation of adjustable speeddrive systemsIndustrial control and systems. Power circuit accessories (requirement for brakes)Low voltage surge protection devicesNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

www.taghedanesh.irStatic controls and braking of motors 6/163List of formulae usedSpeed control through phasor control- -j -71I - m +Im +I,II = line currentI; = loss componentI,,, = magnetizing componentI, = active componentField-oriented control(6.1)T = k . I, . I, sin 8 (6.2)8 = phasor displacement between I, and I, (electricalposition of rotor field in space with respect to stator)To obtain variable V and f in IGBTsthrough PWMt, + t2 + t3 + t4 + t, + t6CDF =T(6.3)tl, t2 . . . t6 are the pulse widths in one half cycleT = one half of a cycleor v~,~,~ = v2/CDF (6.4)V = amplitude of output, voltage pulsesV,,,, = r.m.s. value of the output, a.c. voltageTo smooth output a.c. ripplesQ=CdvdtQ = charge stored by the capacitor unitC = capacitance of the capacitordv/dt = rate of voltage change or a.c. ripples in the d.c.linkCurrent source inverter (CSI) to vary ZI and fVdi .= L - (ignoring R of the circuit)dtV = voltage across the inductorL = large series inductordi- = a.c. ripplesdtComputation of energy saving in a pumpP = shaft input in kWHd = head in barQ = discharge in m3/hourd = specific gravity of the liquid in g/cm317 = efficiency of the pumpBraking timeGD; .Ntb = ~ seconds(s)375 . TbN = N, - N,, (i.e. speed reduction in r.p.m.)Tb = braking torque in mkgBraking heatGD;Hb=-.730(N; - N:l) W.sMinimum braking torqueElectrodynamic or d.c. electric braking(6.9)(6.10)(6.11)idc = braking current= phase value of the starting current= IJ43kl = factor to determine the equivalent ampere turnsfor a particular configurationTI = average load torque between running speed andthe final speedTb = average braking torque between running speedand the final speedTe, = braking torque of the external brakes, if provided;otherwise this may be considered as zeroT,, = starting torque of the motork2 = a factor to account for the average braking torqueInduction generator power outputGI = generator output at the same negative slip as forthe motorK = factor to account for the lower p.f. at highernegative slips, when working as an inductiongeneratorIG = generator rated current in Acos @J = generator rated p.f.Wind energyP = 0.5 ' Cp. A ' p ' V3 (6.13)P = power generated by the windmill in wattsCp = coefficient of performanceA = swept area of the rotor in m2D = diameter of the rotor (blades) in mp = air density = 1.225 kg/m3V= velocity of wind at the site of installation, at theheight of hub in mtsNumber of starts and stopsZL = ZNL. x' Kb KL (6.14)Z, = number of starts and equivalent stops per hour onloadZNL = permissible number of starts per hour with freeshaftKb = factor of brakingKL = mean load factor

www.taghedanesh.ir6/164 Industrial Power Engineering and Applications HandbookFurther reading123456789IOABC ofDrives, Siemens Catalogue DA 662 (1993).Berde, M.S., Thyristor Engineering. Khanna Publishers,India (1900).Blascke, E, ‘The principle of the field oriented transvertor asapplied to the new closed loop control system for rotatingfield machines’, Siemens Review 34, 217-220 (1972).Humphries, J.T., Electronic AC Motors und Controls, MerrillPublishing, Columbia, OH. (19-0).Indo-British Workshop on Power Electronics, Energy SavingMachine Control and Simulation Vol. I and 11, 14-1 8 December1992. Organized by IIT, Delhi, India.Leonard; W., ‘30 years space vectors, 20 years field orientationand 10 years digital signal processing with controlled a.c. drives’,EPE Journal, 1 No. 1, July (1991) and 1, No. 2, Oct. (1991).Microprocessor Application Programme. Department ofElectrical Engineering, Indian Institute of Technology, Delhi,India.Nath, R., ‘Saving energy with variable speed drives’, SiemeizsCircuit, XXII, Oct. (1987).Publication 1204-1.0, March 1996, Allen Bradley, USA.Saunders, L.A., Skibinski, G.L., Evon. S.T. and Kempkes, D.L.‘Riding the reflected wave’, IEEE PCIC, Sept. (1996).11 Suresh, R., Wind Technologies, Tata Energy Research Institute,New Delhi, India (19.0).12 Vithayathil, J., Power Electi-onics-principles nndllpplications.McGraw-Hill, NewYork (1995).13 Wind Energy Te~h~log,~,Kirloskar Electric Co., India.State agencies in India for micrositing of windmillsI The Agency for Non-conventional Energy and Rural Technology(ANERT), PB No. 442, Thaycaud PO, Trivandrum 695 014,Kerala.2 The Tamil Nadu Energy Development Agency (TEDA), JhaverPlaza, IV Floor, 1 -A, Nungambakkam High Road, Chennai 600034, Tamil Nadu.3 The Karnataka State Council for Science and Technology,Indian Institute of Science, Bangalore 560 012, Karnataka.4 The Non-conventional Energy Development Corporation ofAndhra Pradesh (NEDCAP), 5-8-207/2, Pisgah Complex,Nampally, Hyderabad 500 001, Andhra Pradesh.5 The Gujarat Energy Development Agency (GEDA), Suraj Plaza11, 2nd Floor, Sayajiganj, Vadodara 390 005, Gujarat.

www.taghedanesh.irSpecial-Pu rposeMotors7.17.27.37.47.57.1.1 Lo 77.1.2 Ca7.1.3 Ring frame or spinning frame motors 7/167Crane motors 7/168Determining the size of motor 7/168vertical wet pit pump 7/1707.67.77.87.97.1 1.1 Clascification of hazardous locations 7/1797.11.2 Classification of gases, chemical vapour andvolatile liquids 7/1797.12 Specification of motors for Zone 0 locations 7/1807.13 Specification of motor? for Zone 1 locations 7/180f motors, type Ex. ‘d’ 7/180type Ex. ‘e‘ 7/I80e Ex. ‘i’ 7/1827.18 A cal installation\ 7/1837.18.1 Powerhouse treatment of insulation 7/1837.18.2 Terminal boxes 7/1837.18.3 Phase Eegregation 7/1837.18.4 Applied voltage up to 200% 7/1847.18.5 Re-acceleration of motors 7/1847.19 Motors for thermal power station auxiliaries 7/186

www.taghedanesh.ir

www.taghedanesh.irSpecial-purpose motors 7/167Loads and installations that cannot use a standard motordue to their constructional needs, operational demands,special functions, unfavourable location of installation,hazardous items of process, etc. require a special motor,either in mechanical construction or in performancecharacteristics or both. During performance, such loadsmay require a prolonged starting time, a high startingtorque, smoother acceleration, frequent cold or hot starts,stops and reversals etc. For all such applications,meticulous selection of the motor is essential, which shouldmeet all the load requirements without excessive costand yet achieve a higher efficiency and conserve energyin addition to fulfilling environmental needs. Specialfeatures of a few such applications are discussed below.7.1 Textile motors7.1.1 Loom motors (IS 2972 Part I)Figure 7.l(a)Surface cooled loom motor, without finsElectrical featuresLooms for weaving require high torque and motors forsuch applications in a 6-pole design must possess aminimum starting torque, T,,, of 230% and a pull-outtorque, Tpo, of 270% of the rated torque TI. For an 8-poledesign these values must be Tst - 200% and Tpo - 230%of T,. The recommended poles for such motors are 6 and8. For light fabrics such as cotton, silk, rayon and nylonetc., the kW requirement of looms may vary from 0.37to 1.5, while for heavy fabrics (canvas, woollens, juteetc.) from 2.2 to 3.7 kW. The looms may be driven directly,requiring a high torque as above, or through a clutch,which may engage after the motor has run to speed,when a normal torque motor may also be suitable. Unless,the motor is coupled through a clutch it should be suitablefor frequent starts and stops.Constructional featuresA textile mill is normally humidified up to a predeterminedlevel with a view to smooth the process and diminishbreakage of threads. Fluff and cotton dust is wet andadheres to the motor’s surface. It may accumulate on thefan and inside the cooling ribs (fins) and obstruct naturalcooling. These motors are therefore unventilated andsurface cooled (without cooling ribs) or have radial coolingribs (Figures 7.l(a) and (b)). For easy mounting on theloom frame and also to make them adjustable, they aremade either flat based or cradle mounted (Figures 7.2(a)and (b)).7.1.2 Card motors (IS 2972 Part 11)These are similar to loom motors but must possess a stillhigher torque, Le. a starting torque of the order of 350%and 275% of TI and a pull-out torque 375% and 300% ofTI for a 6-pole and an 8-pole motor respectively. Thecard drum is a heavy rotating mass and has a high momentof inertia. The motor, therefore, undergoes a prolongedstarting time and must be capable of withstanding 2.5times the rated current for a minimum of two minutes.Figure 7.l(b) Loom motor with circular ribs (flat basemounted)But unlike a loom motor, which requires too many startsand stops, the operation of carding is continuous. Suchmotors are also required to have circular fins and a flatbase or lug mounting as for loom motors.7.1.3 Ring frame or spinning frame motors(IS 2972 Part 111)These are required to make threads, i.e. the final drawing,twisting and winding of cotton. Such motors must possessvery smooth acceleration to eliminate breakage of threads.They are recommended to have a starting torque of 150-200% of TI and a pull-out torque of 200-275% of TI witha mean acceleration torque of 150-175%. A normalacceleration time of 5-10 seconds is recommended. Fasteracceleration may cause more breakages, while a sloweracceleration may result in snarls and knots in the yarn asa result of insufficient tension.Since card and ring frame motors are normally mountedinside the machine frame, there is an obvious obstructionin the cooling. With this in mind and to meet the torquerequirements, the common practice is to choose the next

www.taghedanesh.ir7/168 Industrial Power Engineering and Applications Handbook225-275% of T, and a low starting current, up to amaximum of five times the rated at the rated voltage aswell as frequent starts, stops and reversals. They arenormally short-time rated. To make these motors suitablefor frequent starts and stops, the rotor is designed so thatacceleration is quick and the heat generated during astart is low. This now limits the temperature rise of therotor even after frequent starts, without sacrificing theframe size. It is possible to achieve this by keeping theGD2 of the rotor low. In such motors also, fan coolingmay be obstructed when a brake is mounted on theextended shaft at the non-driving end (NDE). For suchinstallations also, sometimes a surface-cooled motor maybe preferred. Alternatively, to increase the cooling surface,the housing may be designed with circular ribs, as shownin Figure 7.l(b) and 7.2(b).Figure 7.2(a) Surface-cooled loom motor without fins (cradlemounted)Lift motorsGenerally same as the crane motors, but comparativelysilent in running and have a very low vibration level.For general requirements of other types of lifting andhoisting applications see Table 7.1.7.3 Determining the size of motorFor lifting/hoistingThe mechanical output of the motor for cranes and hoistsin lifting the hook load is the useful work done by it. Thelosses produced in the crane or hoist mechanism aretaken into account by the mechanical efficiency of thehoisting mechanism.The output Ph of such motors is expressed byFigure 7.2(b) Loom motor with circular ribs (cradle mounted)higher frames for such motors, compared to a standardmotor. The latest practice is to employ a variable-speeddrive. As for ‘Cop Bottom Build’ and ‘Nose formation’,the frame must operate at a slower speed to minimize theend breakage, while for the remaining yam it may operateat the higher speed.7.2 Crane motorsCrane and hoist motorsSuch duties require a high starting torque, of the order ofPh = - F.V kW102(7.1)‘ qwhereF = useful load in kgfV = lifting speed in m/sq = efficiency of the mechanismThis output corresponds to a continuous duty of drive. Itmust be suitably corrected for the duty cycle the motorhas to perform (see equation (3.1 l)), Le.Pheq =For traversetl + t* + t3 + ...1.027 . T, . N,Pt = . kW1000 . q(7.2)whereT,,, = maximum torque, consisting of torque, resultingfrom weight load, friction and acceleration inkgf.mq = efficiency of the whole mechanism for traversing.Correct this output also depending upon the duty cycleas noted in equation (3.12), i.e.

AsAswww.taghedanesh.irSpecial-purpose motors 711 69Table 7.1Electrical requirements for lift and crane motorsType of dllh Storfing torquP (TJ Pull-uut torque (T,,J Over .\perding If’freqiientuccrlercition,braking andreverscils arere y uIredI Portal and semi-portal Standard as in t 2.25 of T, Up to 2.5 Nr or Yeswharf cranes manufacturer’s design 2000 r.p.m.whichever is les\2 Oberhead travelling > 2.25 of T, Up to 2.75 of T, A, above Yescranes3 I.iits (a) For squirrel cage ~ above -motors 2 2.25 5 2.75of T, (at any speed)(b) For 4ip-ring motors, ~ above -t 2.25 of T, (withsuitable resistances)4 Power-driven winches E 2.0 of T, 2 2.25 of T, for squirrel cage andtor lifting and hauling2 2.75 of T, for slip-ring motorsYesT2.t, iT;.t2iT;Z.t,+...Teq (r.m.s.) =i t, +I? i tz i ...Duty cycleDuty types Sj, S, and Ss as discussed in Chapter 3 arenormally applicable to crane and hoist motors. For dutytypes S4 and Ss, the duty cycle per unit time is greaterthan S3. The most important factor is the number ofswitching operations per hour. A temperature rise in themotor occurs during acceleration, braking and reversing.Many crane manufacturers specify that the motor shouldbe suitable for half an hour or one hour duration accordingto the British practice still followed in some countries.In fact, it is not possible to correlate precisely theseratings with any of the duty factors. Hence the motorsare designed for any of the duty factors of IS%, 2S%,40% and 60%. In fact the duty factors for different typesof cranes have been standardized, depending upon theiroperation. after several years of experience. For example,the cranes operated in steel industries have different typesof duty factors as follows:Hoisting 60%Traversing 40%Tral elling 60%Slewing 40%For steel mill auxiliary drives or for material handlingequipment, the duty factor normally chosen for slip-ringmotors is either 40% or 60%.StandardizationThe fixing dimensions of the motors are standardized atnational and international levels. However, the outputframerelationship is not yet established for duty typerated motors. The Indian Electrical and ElectronicsManufacturers Association (IEEMA) has issued a standardon crane duty motors in which an attempt is made to listthe outputs against the IEC frames for S3-6, S,- I SO andS5-300 startdhow duty types.Static drivesWith the availability of Vlf drives and other advancedtechnologies through static controls, as discussed inSections 6.2 to 6.4, the use of standard squirrel cagemotors for such applications is a preferred choice.7.4 Sugar centrifuge motorsIn sugar mills a rapid separation of sugar crystals frommolasses is achieved through the use of massecuite* andcentrifugal force. The motor drives a basket full of molasseswhich undergoes repeated cycles of operation, i.e.Charging of massecuite: at a low speed, to preventspillage, normally by a 24-28 pole motor.Intermediate spinning: at half the maximum speed ofspinning, i.e. at 8 or 12 pole.Spinning: at a very high speed compared to the above,generally at 4 or 6 pole.Regenerative braking: When the process is completeand the residue molasses are purged, an oversynchronousbraking is applied by changing the motor fromspinning (4 or 6 poles) to ploughing (48 or 56 poles).This brake energy is then fed back to the mains (seeSection 6.20.1(B)).Ploughing of sugar crystals: at very low speeds of 50r.p.m. or so. This is achieved by a 48- or a 56-polemotor. A further reduction in speed is obtained byconventional electrodynamic or d.c. electric braking.*Massecuite is used to form and then remove sugar crystals frommolasses by a centrifugal technique.

www.taghedanesh.ir7/170 Industrial Power Engineering and Applications HandbookThe centrifugal basket is a dead weight to be acceleratedto the maximum spinning speed. The motor operates forshort durations at different speeds and varying loads. Itis required to accelerate heavy inertia loads at each speed,and is normally designed for multi-speeds such as, 4/8124/48 poles, 6/12/28/56 poles or 6/12/24/48 poles etc.,depending upon the type of centrifuge. The rotor is givenspecial consideration at the design stage to take accountof the excessive heating due to rapid speed changes,braking and acceleration of heavy masses of massecuiteand basket etc. (e.g. by better bracing, high-resistancerotor bar material, better heat dissipation etc.). Duringone complete cycle of massecuite there is a widefluctuation in load and the motor speed and the motoroperates at different h.p. A normal overcurrent release(OCR) therefore cannot protect the motor. Use of RTDsand thermistors as discussed in Section 12.7, can providetotal protection against such variable load drives.The method discussed above is a conventional one toachieve required speed variations. With the applicationof newer technologies, speed variations may be achievedmore accurately and promptly with a single-speed motor,by the use of the following:1 Variable drive fluid couplings (see Section 8.4.1(2))These may not prove to be as effective from the pointof view of energy conservation, as the motor willalways be running at its rated speed and engagementof the coupling alone will vary the output speed.2 Static drives using solid-state technology (see Section6.2) This is the best method for achieving the requiredspeed variations, not from the point of view of quickerand smoother speed variations, but of total energyconservation even at low outputs.NoteApplication of solid state technologyFor all the applications discussed above, which may require specialstarting and/or pull-out torques or speed variations, the use of staticdrives is more appropriate today. With the use of this tcchnology,a standard motor can be made to perform any required duty, exceptthe constructional features and the applicable deratings as discussedin Chapter 1. See also Example 7.1.The use of special motors was more relevant until the 1980s,when solid-state technology was still in its infancy and was not sowidely applied. With the advent of static drives, as discussed inSections 6.2-6.4, the use of standard motors is gradually becomingmore common for all these applications. The drive itself can alterthe supply parameters to the required level to make a standardmotor operate and perform within desired parameters, besidesconserving energy. The purpose of describing a few of theseapplications is only to indicate their non-standard features, wherea standard motor with normal controls may not be able to performthe required duties.7.5 Motors for deep-well pumpsThese pumps are used to lift deep groundwater or anyother liquid from hard or rocky soil. Moreover, the liquidlevel may be so deep that it may prevent the use of acentrifugal pump. Theoretically, the maximum depth fromwhich water can be lifted against atmospheric pressureis 9.8 m (32 feet). To lift water or any other liquid fromgreater depths than this, one has no option but to lowerthe pump into the well, in other words, to lower theentire pump house below ground level, which is noteconomical, practical nor advisable. Moreover, as thegroundwater table may be receding rapidly one cannotbe certain of an ideal depth for such pump houses. Adepth considered ideal today may not be so with in a fewyears as the water level may recede further. Betteralternatives are found in a deep well turbine and asubmersible pump, described briefly below.7.5.1 Deep-well turbine or a vertical wetpit pumpUse of vertical hollow shaft motorsWith the use of such pump sets, the pump alone is loweredinto the pit and the prime mover is mounted above groundlevel. These pumps can lift water or any other liquidfrom a depth of more than 10 m. They are used extensivelyfor irrigation, domestic use, sewage disposal, etc., andare easy to install and maintain. They have an extra-longdrive shaft and an extra number of bearing assemblies tohold the long drive shaft in position and to eliminate therisk of excessive shaft vibration and hunting around its ownaxis. They are built to maintain permanent shaft alignment,have high thrust capacity and are compact in size.The shaft of the pump goes through the motor shaft tothe top of the motor and is bolted there. The pump shaftcan be adjusted at the top to set the impeller by tighteningor loosening the nut holding it. This also eliminates theuse of a flexible coupling between the motor and thepump. These motors are always constructed in a verticalflange design and are provided with heavy thrust bearingsto take the additional load of the pump impeller, its shaftand the fluid in the shaft. These motors are produced insquirrel cage design for simplicity and also because theyhave to drive only a light-duty load and operate at afixed speed. They are also provided with an anti-reverseratchet arrangement to prevent the rotor from rotating inthe reverse direction caused by backflow of liquid in theevent of an abrupt shutdown and during an accidentalphase reversal. A reverse rotation may cause the pumpshaft to unscrew. Since the motor is vertical flangemounted and the pump shaft passes through the motor’shollow shaft, it is called a vertical hollow shaft motor(see Figures 7.3 and 7.4).7.5.2 Submersible pumps using submersible motorsA more economical alternative is found in a submersiblepump where the pump, directly coupled with the primemover, is slid into the tubewell through narrow pipes.Narrow pipes are easy to sink into rocky terrain or verydeep water levels. They are less expensive and are easyto install due to the elimination of the need for a pumphouse. Once the unit is slid into the well it requires littlemaintenance. (See Figures 7.5-7.7.) Such pumps have astandard centrifugal multistage arrangement, and themotors are required to work under water or any otherliquid. These motors have an exclusive application forsubmersible pumps.

www.taghedanesh.irITop shaft adjusting nutHollow shaft motorNon-return valve arrangementSpecial-purpose motors 7/171This is provided to prevent reverse rotation of the pumpin the event of a power failure or a deliberate shutdowndue to backflow of liquid from the rising mains (pipelines).This is located immediately after the last pump stagecasing/discharge outlet to prevent the shaft from rotatingin the reverse direction. The provision of a non-return valvealso ensures that the pump always starts in a shut-offcondition, when the power requirement is at a minimum.Enclosed line shaft bearingColumn pipe couplingGuide spiderLine shaftLine shaft couplingColumn pipeBowlsImpellersimpeller shaftSuction case sand collarSpecial features of a submersible motorThese motors are comparatively long and slender requiringa smaller bore diameter to slide easily into the bore hole/bore well with the pump.Stator winding and insulationThe conductors are waterproof PVC (polyvinyl chloride)-insulated winding wires. These are sprayed with polyamideand conform to IEC 60851 and IEC 60182. For statorwindings, both open and closed type (tunnel type)laminations with a PVC lining are used. Closed typelaminates provide a smooth bore and reduce frictionallosses. The windings are in the shape of ready-madecoils or pull-through wires. For LT sub-mersible motors,the winding cable must be suitable for 1000 V, whereasthe windings are wound for 415 V +6% or any otherdesigned voltage.Deepwell turbine pump fitted with a hollow shaft motorFigure 7.3(a)The application and use of deep-well turbine andsubmersible pumps, is extensive and a choice will dependupon the depth of liquid and the rate of discharge. Inrocky areas, where the digging of larger well cavity is adifficult task, submersible pumps provide an easy alternative.Similarly, for higher heads and where only a smallquantity of liquid is to be pumped, these pumps arepreferred. We discuss below the characteristics of thesemotors and the application of these pumps.ConstructionThe pump is placed above the motor and the water inletis provided between the pump and the motor (Figures7.5-7.7). The discharge cover or case contains the topjournal bearing and small thrust pad to cope with theupward axial thrust during start-up. The pump shaft issupported in the journal bearings. The weight of thepump shaft and the hydraulic axial thrust is borne by themotor shaft and thrust bearing through a rigid meshcoupling.RotorThe rotor is squirrel cage with short-circuited copperrings at the ends. Here also, to vary the starting characteristicsof the motor, the skin effect is used by providingdeep bars, flat bars, tapered bars or other types of slots(discussed in Section 2.3).TorqueThe minimum value of pull-out torque (Tpo) at the ratedvoltage should be 150% of the rated torque according toIS 9283.CharacteristicsThe normal characteristics of these motors are generallythe same as those of a standard squirrel cage motor.EfJiciencyThese motors have a lower efficiency as a result of runningin liquid, causing more liquid drag and also axial thrustbearing loss, which is also a part of the motor. However,this lower efficiency of the motor is compensated byfewer mechanical and hydraulic losses in a submersiblemotor-pump installation, compared to a vertical turbinepump installation.PerformanceThe effect of frequency, voltage variation and ambient

www.taghedanesh.ir

www.taghedanesh.irSpecial-purpose motors 71173Terminal blockUpper ball bearingLower ball bearingInner mechanical sealFigure 7.5(b)Sewage submersible pump and its cross-sectional view (Courtesy: SU Pumps)temperature etc. on the performance of such motors is asfor standard motors, discussed in Chapter 1. The generalroutine tests and methods of conducting them are alsosimilar to those for standard motors discussed inChapter 11.ProtectionThe protection for such motors is also as for standardsquirrel cage motors. See Chapter 12.Rewinding of statorThe maintenance of such motors at site is easy, since thestator can be wound with readymade PVC insulatedwinding coils, and does not need a varnish impregnationand subsequent baking etc., unlike a standard motor. It isthus easy to rewind them at site.BearingsThe bearings are water lubricated. The typical materialsof construction are carbon, copper alloys, bakelite andceramics. The mechanical seals, like a double oil sealprotected with a cap called a Sand Guard, are robust andperfect in sealing the motor to prevent the entry of pumpedliquid into the bearing housing and vice versa to preventany possible contamination.CoolingThe motor stator windings and bearings are cooled throughthe pumped liquid passing externally around the motorbody through heat conduction. The interior of the motorbody (housing) is normally filled with a fluid to facilitaterapid dissipation of heat. The general practices are:1 Water-filled motors: These are initially filled withclean water. Distilled water was originally used as aninitial filler, but with the passage of time, clean waterwas found to be a better substitute and now onlyclean water is used as an initial filler.2 Oil-filled motors: Instead of water oil is now usedwith oil and mechanical seals at the bearings to preventleakage of oil through them.3 Air-filled: The latest practice is to keep the housingempty. This arrangement is found to be economicaland the motor operates at higher efficiency.Motor shaftThe motor with a short shaft length has a shaft of stainless

www.taghedanesh.ir7/174 Industrial Power Engineering and Applications HandbookIswWater levelindicatorIilFl-3Water levelguard?Supporting clampsNon return valveFigure 7.6 Typical arrangement of a submersible pumpsteel generally while those with longer shafts have shaftsmade of carbon steel with stainless steel protective sleevesin the bearing portion.Sizes of submersible motorsSubmersible motors up to 3.50 h.p. and suitable for voltagesup to 3.3 kV are being produced in India and as large asSO00 kW, 11 kV by KSB in Germany (Figure 7.8).Applications0 Water extraction from bore holes and river beds forwater supply and irrigation0 Flood water controlsDe-watering of coal, tin, copper or gold mines, coalwashing and sludge pumpingPumping industrial effluentsSewage duties, industrial slurries and ash handlingDe-watering of dirty water when building roads, dams,harbours, tunnels, etc.Seawater services on onshore platforms, for drinking,washing and firefighting services on off-shore platformsSeawater lift pumps for cooling gas compressors onoil platformsSeabed trenching from remote-operated vehicles (Figure7.9)Oil extraction

www.taghedanesh.irIREhrh-1 ,-CableNon-return valveSpecial-purpose motors 7/175-p7Cable llrDischarge casin,Bearing sleeveBearing bushMining type-cable box'mpellerYPump bowlCasing wear ringBearing sleeveSuction strainerCouplingarrangement,if provided(details not shown)Suction casingSand guardMotor bearingbushRadial shaft seal,- Cable glandMotor bearinghousing (upper)Figure 7.8 Sectional arrangement of a submersible squirrel cagemotor of 860 kW, 30, 50 Hz, 3.3 kV, 1470 r.p.m. (Weir pumps)3Thrust bearina11 Ill I IBlplat6Thrust bearinghousin 9 -Breath( irdiaphragmWindingMotor bearing bushThrust bearing ringWater fillingcum-drainplugRefineriesChemical and process plant duties7.6 Motors for agriculturalapplicationFigure 7.7 Nomenclature of a submersible pump setaccording to IS 9283Deep-sea mining for manganese nodules etc. (Figure7.9). Many riches lie on the seabed. To explore thishidden wealth, submersible pumps have proved to bea boon. Figure 7.9 shows the arrangement of a generalseabed exploration. Here three submersible pumps havebeen used in series, encased in a shrouding to extractand lift manganese nodules from a depth of over5000 m to the sea surfaceIn developing countries particularly, rural powerdistribution networks normally suffer from wide voltagefluctuations. This is generally because of LT loads whichcause a high voltage dip. Consequently, the motors aredesigned for voltages as low as 415 V -20% to +5% andgenerally 415 V-12 3 %, i.e. 360V. For such high voltagefluctua-tions, the general practice is to select a motorlarger than normal. These motors may be designed for alow starting torque, in view of their applications for pumps,winnowers, threshers etc., all requiring low startingtorques. The rotor can also be designed for a low startingcurrent, and thus protect the network from a high startingvoltage dip.

www.taghedanesh.ir7/176 Industrial Power Engineering and Applications HandbookR-Riser pipe3 submersible’ pump sets in tandem-0 500 1000 1500 2000 2500 3000Speed (rpm)Figure 7.10 Approximate output of a surface-cooled motormotor (see also Figure 7.10). For an exact derating contactthe manufacturer.Table 7.2 Approximate output of surface cooled motorsSerial no. Number of Poles % output5250etres2468IO12286070758082Figure 7.9 Deep-sea mining through submersible pumps(typical)Nore IS 7538 recommends a voltage variation of -15% to +5%.But in view of the excessive voltage fluctuations in rural areas, thepractice is to design a motor even for -20% to +5% of 415 V. Avoltage of 415 V is considered for the sake of illustration only, forseries I voltage systems (Table 13.1). It will depend upon the voltageof the LT system being used.7.7 Surface-cooled motorsThese motors are without the cooling fan and may bewith or without the cooling fins shown in Figures 7.l(a)and 7.2(a). Where cooling by the external fan is likely tobe obstructed, for reasons discussed in Sections 7.1 and7.2, the application of such motors is recommended. Anormal motor without a fan will become over-heatedand will require a larger frame to provide extra copperand a larger surface area for heat dissipation. The higherthe speed of the motor, the higher will be the coolingeffect of the fan, and the higher will be the derating ofthe motor frame to compensate for the reduced cooling.As a rough guide, the outputs shown in Table 7.2 may beadopted for a surface-cooled motor over a fan-cooled7.8 Torque motors or actuatormotorsThese motors are designated by their ‘stall-torque’ ratherthan kW ratings. They are fitted with a brake with anadjustable torque and are normally surface cooled. Theirapplication is for auxiliary loads, which sometimes requirea very small movement of the motor, even less than awhole revolution such as to actuate a motorized flowvalve to control the flow of liquid or gas. They thus havea high slip characteristic (high rotor resistance) and arecapable of providing the maximum T,, with the minimumpower input. The design of such motors is thus differentfrom that of a normal motor and may possess some ofthe following features:They may be short-time rated.Their moment of inertia is kept low to restrict theirtimes of acceleration and deceleration (equation (2.5))to facilitate frequent starts and stops.The rating is designated by torque rather than kW.Maximum torque occurs at the locked rotor condition,i.e. when S = 1 (Figure 7.11).Zst is kept much lower than a standard motor, as it may

www.taghedanesh.irSpecial-purpose motors 7/1770Figure 7.11SoeedCharacteristic of a torque motorbe required to operate frequently and a high I,, wouldbe a deterrent.The speed-torque characteristics are almost linear, withtorque falling with speed (Figure 7.1 I).There is no Tpo region, thus no unstable region. Themotor can operate at any speed up to the rated onewithout any stalling region.The locked rotor thermal withstand time is much higherthan for a standard motor.Likely applicationsWire-winding machine.;Roller tablesFrequent starts and reversesMaterial handlingValve actuatorVane controlBagging machinesFa\t tapping machinesClamping and positioning devices7.9 Vibration and noise levelExcessive vibrations according to international codes cancause mechanical failure in the insulation by looseningwedges, overhangs, blocks and other supports that holdthe stator and the rotor windings or rotor bars in theirslots. Vibrations also tend to harden and embrittle copperwindings and may eventually break them when theybecome loose (see also Sections 11.4.6 and 11.4.7).There are some essential features that a good motorshould possess, and among these are vibrations and noiselevel. The noise may be magnetic or aerodynamic,assuming that there are no frictional or other noisesemanating from the motor. Magnetic noise is due toresonant excitation of the stator core by slot harmonics,caused by the electro-magnetic circuit of the machine(Section 23.6(a)) and loose clamping of the stator’s steelstampings and the motor’s rotor core. Or loose rotor barsand magnetic unbalance, when the magnetic andgeometrical axes of the motor are not concentrias tooth ripples and magneto-striction. Aerodynamic noiseis caused by the flow of cooling air at a higher peripheralspeed over the cooling fins and the rotor bars ax well asunbalanced rotating masses, aerodynamic loads and somesecondary effects such as noise from the bearings.instability of the shaft in the bearings. passive resistanceand aerodynamic expansion.In an electric motor, although such parameters areinherent, they can be tolerated only to a certain extent.IEC 60034-14 gives these levels as indicated in Table1 I .3. For applications such as household appliances.escalators in residential buildings, offices and hospitals,and for machine tools, the vibration levels must be evenlower to eliminate the transmission of these as far aspossible through the driven structure to the building orto the cutting tool in the case of a machine tool. Unlessthe vibrations are reduced to a reasonably low level. theymay cause a noise nuisance to the occupants of a buildingor affect the accuracy of the cutting tool and the machine.Research on the effect of noise on a human body hasrevealed that noise level (sound pressure level) must belimited to 85 dB for a human body to work safely withoutfatigue, for eight hours a day and without causing a healthhazard on a continuous basis. In the OSHA (OccupationalSafety and Health Administration, USA) regulations themaximum industrial noise level is 90 dB. Table 7.9 showsthe likely noise levels and their sources of origin.Vibrations, when transmitted from the source to theconnected appliance or structure or a machine tool, becomemagnified, depending upon the contact surface area andcreate noise. In fact, vibrations are the primary source ofnoise. In machine tools, the vibrations are transmittedfrom the motor body to the cutting tool and affect itsaccuracy. For precision work and for sophisticated machinetools, a vibration level as low as 2 microns is sometimespreferred.Vibrations and noise levels are mainly associated withthe mechanical construction and electrical design of themotor. Sound mechanical construction and a balancedrotor, a tightly fastened core and smooth bearings caneliminate vibrations and mechanical noise to a large extent.Also a better electrical design can eliminate electricalvibrations in the stampings. due to magnetic forces andhigher harmonics. A better electrical and mechanicaldesign will thus mean:1 Rigid fastening of the stator core to the housing2 Smooth and frictionless bearings with proper greasing3 Precision dynamically balanced rotating parts. Theleading Indian manufacturers recommend and maintaina vibration level, peak to peak as shown in Table 7.3.See also Table 11.3, according to IEC 60034.144 A uniform air gap between the stator and the rotor5 Judicious selection of stator and rotor slots, with angleof skew and glue density6 Magnetic loading, i.e. tlux density7 Windage noise (noise at the suction and the exhaustof the cooling airj. This makes a large contribution tothe total noise emanating from a motor. To suppressthis, some manufacturers provide the followingadditional features or noise-suppression devices in

~ ~ ~www.taghedanesh.ir7/178 Industrial Power Engineering and Applications HandbookTable 7.3 Maximum vibration levels, as practised by IndianmanufacturersSpeed Nsrp.m. at 50 HZ LT(microns)Vibration level, peak to peak(double amplitude)HT motors(microns)1. 3000 1s 152. 1500 40 253. Up to 1000 40 25Note In higher speed ranges and for HT motors, these levels ofshaft vibration are generally of the same order or slightly betterthan prescribed in IEC 60034-14, corresponding to Table 11.3.the fan and fan cover. The basic purpose of all theseis to reduce to a minimum the windage (air friction)noise at the suction and exhaust points:By providing a unidirectional axial flow fanBy providing a sound-absorbing fan cover at thenon-driving end, as shown in Figure 7.12By transforming the intake axial air flow to a radialair flow, as illustrated in Figure 7.13, thussignificantly reducing frictional and hence suctionnoiseBy changing the geometry of the fan cover, i.e. byproviding muffling cones (noise-hod) at the drivingend (Figure 7.14) and providing felt wool on theNon-driving endcAir inccFigure 7.12 Sound-absorbing fan cover at the non-drivingendFigure 7.14 Motor with muffling cone and lining of round feltwool to absorb the exhaust noise of airdriving end fan cover to absorb the friction andintensity of exhaust air.7.1 0 Service factorsWhen a motor is expected to operate in unfavourableconditions such as:Intermittent overloadingHigher ambient temperaturesA restricted temperature rise as for a spinning mill, arefinery or a hazardous areaFrequent starts, stops and reversesor any such conditions during operationand when it is not possible to accurately define theirlikely occurrences or magnitudes, it becomes desirablefor the motor to have some in-built reserve capacity. Toaccount for this, a factor, known as the ‘service factor’,is considered when selecting the size of the motor. A‘service factor’ in the range of 10-15% is consideredadequate by practising engineers. With this service factor,no more derating would normally be necessary. See alsoExample 7.1 at the end of the chapter.7.1 1 Motors for hazardous locationsFigure 7.13Radial flow of air to reduce suction noiseAreas prone to or contaminated with explosive gases,vapours or volatile liquids are at risk from fire orexplosions. A hazardous area is a location where there isa risk of fire or explosion due to the formation of anexplosive mixture of air and gas or inflammable vapour.Normal motors may emit sparks or some of their partsaccessible to such environments may reach a temperaturehigh enough to ignite inflammable surroundings duringnormal running. Special motors have thus been developedfor such locations and may be one of the followingtypes:1 Flameproof (FLP) or explosionproof type (Ex. ‘d’)2 Increased safety type (Ex. ‘e’)3 Pressurized type (Ex. ‘p’)4 Non-sparking type (Ex. ‘n’)

www.taghedanesh.irSpecial-purpose motors 7/179IEC 60079-14 provides general guidelines for the selectionof electrical equipment for hazardous areas.7.11.1 Classification of hazardous locationsIt is important to identify areas in accordance with theexpected degree of fire hazard to facilitate an appropriateand economical selection of electric motors. These areas,according to IEC 60079-10, are classified into threecategories as follows.Zone 0This is a location which is continuously contaminatedwith explosive gases, chemical vapours or volatile liquidsand thus is highly susceptible to fire hazards. Installationof electrical machines in such areas should be avoidedas far as possible, to reduce cost, facilitate maintenanceand take other precautions.IEC 60079-14 recommends the use of only intrinsicallysafe apparatus in such locations. Intrinsically safeapparatus and their electrical circuits are basically lowenergydevices and release only small amounts of energy,insufficient to ignite the surrounding atmosphere. Thedevices may be to record, sense or monitor the operatingparameters of equipment operating in such locations orthe condition of the surroundings. They may be mountedon the surface rather than the interior of an enclosure.Circuits connecting such devices and instruments arealso made intrinsically safe for use in such areas. Weshow in Section 7.12 that an electric motor is not suitablefor such locations (see also Section 7.16).Zone 1This is a location which is not permanently contaminatedbut is likely to be prone to fire hazards during processing,storage or handling of explosive gases, chemical vapouror volatile liquids, although under careful and controlledconditions. For such locations in addition to a flame- orexplosion-proof enclosure, type Ex. ‘d’, an increasedsafety motor, type Ex. ‘e’ or a pressurized motor, typeEx. ‘p’ may also be considered to be safe.Zone 2This is a location safer than Zone 1 with a likelihood ofconcentration of explosive gases, chemical vapour orvolatile liquids during processing, storage or handling.This would become a fire hazard only under abnormalconditions, such as a leakage or a burst of joints or pipelinesetc. Such a condition may exist only for a short period.A standard motor with additional features, as discussedbelow, may also be safe for such locations. A non-sparkingtype, Ex. ‘n’, or an increased safety motor, type Ex. ‘e’,may also be chosen for such locations.Nore Some applications creating hazardous conditions arepetrochemical or fertilizer plants, refineries, coal mines etc., whereinflammable gases and volatile liquids are handled, processed andstored.Mines, collieries and quarriesIn view of the magnitude and intensity of explosion andfire damp hazards at such locations, a motor with superiorelectrical and construction design is recommended. IEC60079-0 categorizes such areas as Group I and recommendsonly flame-proof motors, type Ex. ‘d’, dependingupon the nature and the ignition temperature of the gases,chemical vapour or volatile liquids at such locations.The maximum permissible temperature at the externalsurface of the machine must also be limited so that itdoes not ignite the inflammable substances at theinstallation.7.11.2 Classification of gases, chemical vapourand volatile liquidsBased on the ignition temperature of these inflammablesubstances, Table 7.4 shows the groups of substancesrequiring specially constructed motors.Table 7.4 Grouping of gases, chemical vapour and volatile liquids, their ignition temperature and limits of permissible temperatureat the external surface of the motor based on IEC 60079-20Ignition group TI T2 T3 T4 T5 T61 Ignition temp. (“C) Above 4502 Grouping AcetoneEthaneEthyl-acetateAmmoniaBenzeneAcetic acidCarbon monoxideMethane (fire damp)NaphthalenePropaneCoal gas (town gas)PetroleumTolueneWater gasHydrogen300/450AcetyleneEthyl amineEthyleneIso-amyl acetateButanen-Butyl alcoholn-Propyl alcoholButanolMethanol200/300 135/200 100/135 85/100Acetaldehyde Ethyl-ether CarbonEthyl glycoldisulphideCrude oilEthyl Nitriten-HexaneTurpentineMineral oilsCyclo hexene

www.taghedanesh.ir711 80 Industrial Power Engineering and Applications Handbook7.12 Specification of motors forZone 0 locationsFor Zone 0 locations only intrinsically safe, low-energyapparatus is recommended. Induction motors, being largeenergy sources, release high energy, particularly duringswitching or a fault condition, as discussed in Section6.14.1, are not suitable for such locations. See Section7.16 for more details.7.13 Specification of motors forZone 1 locationsFor Zone 1 locations, the following types of enclosuresare recommended7.13.1 Flame- or explosion-proof motors,type Ex. ‘d’IEC 60079-1 defines the basic requirements for suchmotors which, besides limiting the maximum temperatureof any part of the motor, accessible to the contaminatedarea, as shown in Table 7.4, also maintain definite lengthsof paths, air gaps, widths, and diametrical clearancesbetween various rotating and stationary parts to avoidany rubbing and arcing. The following design considerationsmay also be noted.Design considerationsThese motors should be able to withstand an internalexplosion of inflammable gases, chemical vapour orvolatile liquids without suffering damage or allowingthe internal inflammation to escape to external inflammablesubstances through joints or other structuralopenings in the enclosure. (The explosion may have beencaused by the gases. vapour or volatile liquids that mighthave entered or originated inside the enclosure.)Apart from withstanding the internal explosion, theconstruction must be such, that the flame escaping fromthe interior is cooled down to such an extent that it isrendered incapable of igniting the surrounding hazardousatmosphere. This is achieved by providing joints withextra long surfaces (flame paths) and special clearances(gaps). The flame path is the breadth or the distanceacross the face of the flange, and the gap is the distancebetween the two faces of the flange, as shown in Figure7.15. The requirements of minimum lengths of flamepaths and maximum gaps for various gas groups arespecified in IEC 60079-1 1. Some of the importantconstructional features of such enclosures are as follows:All components such as stator housing, end shields,terminal box and covers etc. are pressure tested beforeuse.No light metal such as aluminium is used as an externalsurface to avoid frictional arcing.The maximum surface temperature must remain belowthe temperature class specified in Table 7.4 for a specificapplication.7.13.2 Increased safety motors, type Ex. ‘e’These motors will also suit areas defined for Zone 1.Use of HT Ex. ‘e’ motors, however, should be avoided inthis zone. Such enclosures do not produce arcs internallyand also restrict the temperature rise of any part accessibleto such an environment to a limiting value, during startupor run, in accordance with the applicable class ofinsulation shown in Table 7.5. The limiting temperaturemust be less than or equal to the ignition temperature ofthe prevalent atmosphere shown in Table 7.4, otherwisethe limiting temperature will become the same as theignition temperature, according to Table 7.4. The rotortemperature is also restricted to 300°C during start-up,unless Table 7.4 shows a lower limiting temperature.Design considerationsIEC 60079-7 outlines the basic requirements for increasedsafety, type Ex. ‘e’ motors as follows:The enclosure must have a high degree of protectionShrouding?aring covererior of enclosureof pathLerior of enclosureL = Length of flame pathFigure 7.15Flame paths and gaps in a flame-proof or explosion-proof motor

Special-purpose motors 7/181Table 7.5Limiting temperature and limiting temperature rise for type Ex. 'e' motorsMeasuringmethodClass of insulationA E B F HLimiting temperature R 90' 105" 110" 130a 155aunder rated service conditions ("C) T 80 95 100 115 135Limiting temperature rise under ("C) R 50 65 70 90 115rated service condition (referred to T 40 55 60 75 95an ambient temperature of 4OOC)Limiting temperature at the end of ("C) R 160 175 185 210 235time tkLimiting temperature rise at the ("C) R 120 135 145 170 195end of time t," (referred to anambient temperature of 40°C)As in IEC 60079-7R - by resistance method.T - by thermometer method (only permissible if the resistance method is not practicable).a These values are restricted by 10°C, compared to the working temperature prescribed for standard motors, as in Table 9.1.These values are composed of the temperature (or the temperature rise) of the windings in rated service and the increase of temperatureduring time tE.to prevent entry of dust, water or moisture. Theminimum protection specified is IP:54 (IP:55 ispreferred) according to IEC 60034-5.All terminals must be the anti-loosening and antirotatingtypes.The minimum clearance and creepage distances mustbe maintained for conductors as specified.The temperature rise of windings must be 10°C lowerthan that specified for normal machines.The mechanical clearance between rotating parts, e.g.,fan and fan cover or the radial air gap between thestator and the rotor, should not be less than specifiedto prevent sparking.The temperature of the windings and other parts mustnot exceed the limiting temperature specified in Table7.5, even if the motor, after a prolonged operating period,remains energized in a stalled condition, for a specifiedtime of tE seconds while tE will not be less than 5 seconds.All gases are classified according to their ignitiontemperatures as in Table 7.4. Table 7.6 recommends theTable 7.6 Temperature class and limiting surface temperaturewith regard to gas ignitionIgnition temperature ("C)TemperatureLimiting surfaceclass Above UP to andincludingtemperature, ( T )T6 85 100 85T5 100 135 100T4 135 200 135T3 200 300 200T2 300 450 300T1 450 - 450As in IEC 60079-14limiting surface temperatures of motors for these ignitioncategories in terms of temperature classes TI to T6. Themotor's surface temperature must not exceed the limitingtemperature specified in Table 7.6 under the same conditionand time tE. The protective switch will thus be set to operatewithin the heating-up time tE for the relevant temperatureclass. Embedded temperature detectors or thermistors arerecommended to give the required signal to the trippingcircuit during an emergency. Figure 7.16 illustrates thegeneral requirements for an increased safety motor.Heating-up time tEThis is the time taken by the stator or the rotor, whicheveris less, to reach the limiting temperature rise, as specifiedin Table 7.5, when the starting current Zst is passed throughthe stator windings after the motor has reached thermalequilibrium, under rated conditions. For increased safetymotors, this time should not be less than 5 seconds(preferably 10 seconds or more).7.13.3 Pressurized enclosures, type Ex. 'p'These may be standard TEFC motors suitable for operatingunder an internal pressure of 0.05 kP, or a pressure slightlyabove atmospheric. The minimum specified is 5 mmwater-gauge above atmosphere. During operation thispressure is maintained with inert gas or air through anexternal closed circuit, preventing inflammable gases fromreaching the motor's inner components. Before reswitchingsuch a motor after a shutdown, inflammablegases which may have entered the enclosure must firstbe expelled. For pressurizing, two holes are normallydrilled on the motor's end shields, one for entry and theother for exit of the air or gas. For large motors, a totallyenclosed dual-circuit type enclosure is normally used sothat the motor's interior is pressurized with air or nitrogen

7/182 Industrial Power Engineering and Applications HandbookPerfectly sealed joint forprotection against dust,moisture corrosionand waterSpecial windingfor 10°C lower7.14 Motors for Zone 2 locationsLocations falling within this category can employ motorsthat are more economical than the other types discussedabove. IEC 60079-14 has defined the basic requirementsfor such enclosures which are obviously less stringentthan the others. In addition to maintaining specifiedcreepages and clearances between the rotating and thestationary parts, the following are the two mainrequirements specified for such enclosures:and rotorAnti-loosening, anti-rotating typeterminal arrangement with specifiedclearance and creepage distances12They should produce no arcing during a start-up orrun.Surface temperature should not exceed the ignitiontemperature noted in Table 7.6 for a particulartemperature class under any conditions of operation.There is no limit to the temperature rise to thepermissible limits for a particular class of insulationof windings or other parts of the machine, except thelimiting surface temperature as in Table 7.6. For suchan application, a normal 1P:SS enclosure may also beemployed.Figure 7.16 General requirements for an increased safetymotorto about 75 mm water-gauge above that of the atmospheresurrounding the enclosure. The enclosure is sealed andthe leakage rate controlled to about 250 litredmin. Someof the basic safety features and devices are noted below.Such enclosures are recommended only when there area number of these machines installed at the same locationso that a common pressurizing and piping system maybe employed to reduce costs on piping network andpressurizing equipment and its accessories, in additionto its regular maintenance.Safety featuresThe equipment is usually fitted with interlocks to ensurethat if the internal pressure or flow rate of air or inertgas falls below a certain minimum level, the supply tothe motor is cut off.Pressure-measuring device: This is provided for theoperation of alarm or trip devices in the event thepressure within the casing falls below the permittedminimum.Safe-starting device: This is provided to ensure thatno apparatus within the enclosure is energized untilthe initial atmosphere within the casing has beencompletely displaced.7.14.1 Non-sparking motors, type Ex. ‘n’A subsequent study of construction features of motorsfor Zone 2 locations, resulted in the development of nonsparking,type ‘n’ motors. The basic design considerationfor such motors is similar to that of type ‘e’ motors butnow there is no restriction in the limiting temperature,by 10”C, as in type ‘e’ motors. Frame sizes for thesemotors are generally the same as for general-purposemotors. Thus they tend to be smaller and less expensivethan type ‘e’ motors for the same output.7.15 Motors for mines, collieries andquarriesSpecial provisions are laid down in IEC 60079-0 andIEC 60079-1 for motors required for such locations inview of fluctuating degrees of humidity and temperature.Such locations are defined with a surface temperaturelimit of 150°C where coal dust can form a layer, or450°C where it is not expected to form a layer. Otherwise,other details are generally the samc as for flameproofmotors type Ex ‘d’, according to IEC 60079-1. Forvariations in length of paths, gaps, widths, creepage andclearance distances, the reader should consult theseStandards.7.1 6 Intrinsically safe circuits,type Ex ‘i’IEC 60079-14 specifies two categories of intrinsicallysafe circuits and apparatus, i.e.0 Intrinsically safe category ‘ia’ as in IEC 60079-1 1 forinstallations in Zone 0, and

Special-purpose motors 7/183lntrinsically safe category ‘ib’ for installations in Zones1 and 7,. The apparatus for this category must be ofgroups UA. nB. and nC (IEC 60079-0).The main stipulation for such systems is that they willnot emit sparks in normal or fault conditions. This thereforerestricts the use of only low-energy auxiliary and controlcircuits connecting instruments and devices at theselocations. Such instruments and devices may record, senseor monitor certain operating parameters of machines orapparatus installed at these locations or the surroundingatmosphere itself, through temperature detectors (RTDs).instruments to measure humidity and pressure, andvibration detectors.To comply with the requirement of intrinsic safety, allelectrical circuits installed at such locations should besafe and must produce no spark or heat under normal orfault conditions, sufficient to cause ignition of thewrrounding medium. The parameters of the circuit suchas b” I, R. L. and C, which can release heat energies byI’R. I’L and CV‘ have a recognisable bearing on theignitability of an explosive atmosphere. For an inductivecircuit, for instance, with a magnetic or solenoid coil, Vand L would be the most potential parameters. During afault condition the circuit may release high stored energyby heating the wires or any other part of the device towhich it is connected. This energy may be sufficient tocause ipnition of the surrounding atmosphere.For the safety of these circuits, therefore. the mainconsideration is defined by the level of this energy, whichshould always be less than that required to ignite explosivegases. chemical vapour or volatile liquids in thewrrounding atmosphere. Accordingly, minimum safeignition currents (MICs) have been established by laboratorytests for different control voltages and R, L, and Cof the circuit. and are the basis of testing the circuits ofthese devices and instruments at such locations todetermine their compliance for safe installation. Sincesuch circuits would be required for other zones and groupsof gases also with different limiting temperatures, IEC60079- I 1 would also apply to all such zones and groupsof gases. The test requirements. however, would varyfor different locations, as stipulated.To contain the temperature of the electrical circuitswithin safe limits for a particular temperature class of. the maximum current rating for aa conductor is also stipulated in IEC60079- I I. The constructional requirements also stipulatethe rnininium clearances and creepage distances in airbetween the conducting parts of all the intrinsically safeelectrical circuits.7.1 7 Testing and certifyingauthor it iesInstallations categorized as hazardous locations, requiringthese types of motors, would be regarded as critical applications.where stringent safety measures would bemandatory. Normally, the government of a country wouldauthorize some agencies to independently grant approvalfor the use of this equipment at such locations. This ismandatory for the installation and operation of anyelectrical equipment in hazardous areas. These authoritiesmay also specify construction requirements for equipmentand issue regulations for its installation and operation.In India, the Central Mining Research Institute, Dhanbadcarries out this testing and provides the necessarycertification for motors used in explosive atmospheres.But for approval of the equipment, whether it is worthyof use in a particular hazardous area, there are accreditedagencies. Some of these are Directorate General MinesSafety, Dhanbad, Chief Controller of Explosives. Nagpurand Directorate General of Factory Advice Service andLabour Institute. Bombay.7.1 8 Additional requirements forcritical installationsA process industry, a refinery, a petrochemical, a fertilizerplant or a power station are installations that can beclassified as critical. They cannot afford a breakdownduring normal operation and would prefer to incorporatemore safety features into the drive motors, even if theseare expensive. These features can be one or more of thefollowing.7.18.1 Powerhouse treatment of insulationSpecial insulation and coating of stator windings andoverhangs are sometimes essential to ensure protectionagainst tropical weather, fungus growth, moisture, oilabrasives, and acid and alkali fumes. Powerhousetreatment is one insulating process that can meet all theserequirements (see Section 9.3).7.18.2 Terminal boxesIt is recommended that the motor should have separateterminal boxes for the main supply and for the accessoriessuch as space heaters, embedded temperature detectors,bearing temperature indicators and moisture detectorterminals, etc. In LT motors, however, if it is not possibleto provide a separate terminal box for these accessories,the main terminal box may be adequately spaced tosegregate these terminals from the main terminals withinthe box.In HT motors, however, these teiminal boxes are alwaysseparate because two or more voltages (main andauxiliary). For main terminals there are normally twoterminal boxes - one on one side of the stator to housethe main three-phase stator terminals and the second onthe other side to form the star point. These boxes aregenerally interchangeable to facilitate cable routing.7.1 8.3 Phase segregationIt may be necessary to separate the live phases withinthe terminal box to eliminate any occurrence of a tlashoveror a short-circuit and a subsequent explosion. Such explosionsmay be very severe, depending upon the short-circuitlevel of the system and the surrounding atmosphere and

7/184 Industrial Power Engineering and Applications Handbookmay cause a damage to life and property. Such a provisionis normally desirable for HT motors, where such incidentsare more likely due to a higher short circuit level. For aterminal box with a less stringent design and adequateair and creepage distances between phases and phaseand ground, capable of withstanding system faults, phasesegregation may not be necessary. An explosion diaphragmwill, however, be essential at a suitable location on theterminal box to allow the high-pressure gases to escapein the event of a fault inside the terminal box.A normal phase segregation arrangement is shown inFigure 7.17. Each individual phase lead of incomingcable separates within a compound-filled separatingchamber and terminates in the main terminal box with aseparate enclosure for each phase, thus eliminating thepossibility of flashover inside the terminal box. For moredetails see Section 28.2.2. A few designs of different typesof terminal boxes are illustrated in Figure 7.18(a)-(c).7.18.4 Applied voltage up to 200%When a motor is energized there is an induced e.m.f. inthe stator windings which takes time to decay to zeroafter it is switched OFF (z= LIR, an analogue to capacitivedischarge, Section 25.7). In ordinary switching operations,the sequential delay may be sufficiently long for theinduced e.m.f. to persist and affect the motor’s performance.But, a system that has an emergency source ofsupply, through an automatic changeover scheme, mayhave a changeover time of just a few seconds on failureof the main supply. This period sometimes can be tooshort for thc motor-induced e.m.f. to decay to a safelevel. As a consequence, the impressed voltage of theother system, at the instant of closing, may fall phaseapart with the motor’s own induced e.m.f., and the motorwindings may be subject to a momentary overvoltage.The effect of such changeovers is felt up to 200% of therated voltage of the motor windings, and may occur whenthe applied bus voltage and the residual motor voltageTerminal boxSteel enclosure,(for each phase)Cable lugBushingSeparating chamberor cable end box(compound filled)Figure 7.17 Phase-segregated terminal boxfalls phase apart by 180’. This voltage may cause aflashover inside the terminal box if the motor terminalsare not adequately spaced, and damage the inter-turninsulation and the overhangs of the motor windings etc.To withstand such overvoltages that may result in higherdielectric and electrodynamic stresses the overhangs canbe treated by an additional coat of varnish, and a tightbinding. A similar treatment is desirable for the slot coils,as well as liberal spacings and creepage distances betweenthe phases and the ground inside the terminal box. Inaddition the motor and the driven equipment shafts mayalso be braced to withstand transient torques. This is thecase when the interrupter remains closed during thechangeover period. If it is not closed the changeovermay also cause switching surges in HT motors (Section17.7.2(ii)).7.18.5 Re-acceleration of motorsIt may be essential for certain critical process drives tohave an autostarting feature, to re-accelerate them aftera momentary main power failure. This is required to savethe process and the downtime, and prevent re-switchingof these drives on a rapid restoration of power. Thisscheme can also be useful for critical processes where arestart of a drive may take a long time due to its torquecharacteristics or process requirements and resulting ina long downtime. Such a scheme can achieve fasterstabilization of the process by retaining the drive in motion,and picking it up quickly on a rapid restoration of power,as in polypropylene plants and gas crackers. A papermill, for instance, would require the whole length ofpaper to be removed from its drying cylinders if the millis to be re-started after a shutdown. This is a waste ofpaper, in addition to a longer downtime.Re-acceleration can be achieved by introducing an OFFdelaytimer T, (Figure 7.19) into the control circuits ofall these drives with a time setting of, say, 040 seconds,so that the contactors of the critical drives restart automaticallyafter restoration of power within the set period.The time setting, however, has to be such that the drivesare still in motion at speeds so that the process can berestored. Figure 7.19 illustrates a typical scheme. Thiscan also be achieved by placing a capacitor across theoperating coil of the main contactor which can provide atime delay up to 1 4 seconds by holding the contactor.Contactors with built-in capacitors are available, and arecalled OFF-delay release contactors. Such a short-durationhold-on feature may be adequate when the system isrequired to hold without a trip against momentary heavyvoltage dips, arising from system disturbances or simultaneousswitching of large motors or during an auto-manualbus transfer.Re-acceleration may be restricted to only critical drivesto avoid a high switching inrush current on resumptionof power. This can be achieved by grouping all thesedrives in two groups, one with a delay of TI and theother through an ON-delay timer T2. Thus extending aprogressive time delay so that each subsequent group ofdrives accelerates only after the previous group is switched,to avoid simultaneous switching of all motors. Figure7.20 illustrates a typical scheme.

Special-purpose motors 7/185(1) Bushing insulator (cast resin)(2) Bottom view of terminal box(3) Terminal stud(4) Pressure relief flap(5) Top view of the box(6) Cable entry(7) Grounding terminal(8) Weather guard for upper joint and fitting aid(9) Sealing r\ng(1) Bushing insulator(2) Pressure relief plate(3) Terminal stud(4) Cover(5) Cable box for compound filling(6) Cable entry(7) Grounding terminalTerminals for cable connectionsPressure relief diaphragmPost insulator (cast resin)(c)NoteThe terminal boxes upto a fault level of 350 MVA for 5 cycles at 6.6 kVare generally non-segregated and beyond 350 MVA segregated type.Figure 7.18 Typical designs of a few HT terminal boxes

7/186 Industrial Power Engineering and Applications Handbookinst. - instantaneous7C431 1(Off-delay) On Off+ tripinst.Udinst.Figure 7.19 A typical scheme illustrating re-acceleration of a critical motor7.19 Motors for thermal powerstation auxiliariesThese applications have considerably more stringentperformance requirements than any other application.Circulating water pumps, boiler feed pumps, forceddraught(FD) and induced-draught (ID) fans, pulverizers(ball mills) and condensate pumps are components in athermal power station that may require extra safety in astandard motor to make it able to fulfil these requirementsand withstand abnormal service conditions and systemdisturbances. Abnormal operating conditions may be oneor more of the following:High ambient temperatureHigh humidityDust-laden (coal dust and fly ash) environmentLong continuous operationHigh fault levelsOvervoltages, caused by a fast bus transferRe-acceleration in some drivesVoltage surges, caused by system disturbances orswitching operations (for HT motors).In a large power station, connected to a transmissionnetwork through a power grid, system disturbances are acommon feature. For details see Section 17.5.With a view to standardizing the basic requirementsof an electric motor for essential services, electricityauthorities may sometimes stipulate norms and outlinetheir basic minimum requirements for the benefit of usersand the manufacturers. In India, for instance, the CentralElectricity Authority (CEA) publishes manuals based onthe actual feedback obtained from various thermal sites,electricity companies, leading consultants and the NTPC(National Thermal Power Corporation), in addition tomotor manufacturers. Below are extracts from thesemanuals :I Motors must be generally squirrel cage type and notslip-ring.2 The preferred ratings for motors above 1104000 kW,must follow the ‘Renard Series’, R-20, according toIS0 3, as shown in Table l.l(a).3 The preferred synchronous speeds is as shown in Table7.7.4 The enclosure and type of cooling is as shown inTable 7.8.5 Space heaters: motors above 30 kW must be providedwith anti-condensation space heaters suitable for240 V 50 Hz (or 60 Hz), a.c. supply. The space heatersmust switch ON automatically when the motor is outof service. Motors of 30 kW and below must be suitablefor connection to a 240 V, 50 Hz (or 60 Hz) a.c.supply. Here also the windings must be connectedautomatically to the 240 V supply when the motor isout of service.6 Terminals and terminal boxHT motors must be provided with phase-segregatedterminal boxes as illustrated in Figure 7.17.

Special-purpose motors 7/187F4- YNL1 1‘61st motorI(Off-delay) On Off O/C+ tripinst.1 1);id,2nd motor I Ih213(On-delay) (Off-delay) On Off O/C+ tripinst.inst. - instantaneousFigure 7.20 A typical scheme illustrating re-acceleration of two motors (or two groups of motors) with a time gapTable 7.7 Preferred speeds for thermal power station auxiliariesAuxiliuryInduced-draught fan (ID fan)Forced-draught fan (FD fan)Primary air fan (PA fan)Ball mill (coal crusher)Vapour fanBoiler feed pump (BFP)Condensate pumpCirculating water pump (CW pump)Synchronousspeed (rp.rn.)750, 10001000, 1500150010001000, 15001500, 30001000, 1500375,429,500Table 7.8 Recommended enclosures and type of cooling forpower station auxiliariesEnclosure1 TEFC2 TETV3 CACA4 CACWType of coolingus inIEC 60034-6IC 0141IC 0151IC 0161ICW 37A 81 01ICW 37A 9 1

7/188 Industrial Power Engineering and Applications HandbookIn HT motors the terminal box for space heatersand embedded temperature detectors must beseparate from the main terminal box.The terminal box must be capable of withstandingthe system fault current for at least 0.25 second.The terminal box must be suitable for being turnedthrough 180”, for bottom or top cable entry.The terminal box must have the same degree ofprotection as the motor.For large motors, 1000 kW and above, if they areprovided with star windings the three neutral endleads must be connected to a separate terminal boxto enable mounting of CTs for differential protection.The neutral terminal box need not be phasesegregated.7 PerformanceMotors must be designed for an ambient temperatureof 50°C.HT motors must be generally wound with class Finsulation and the temperature rise should not exceedthe prescribed limits of class B insulation. LT motorsup to 250 kW, however, can be wound with class Binsulation.The performance of the motors must conform toIEC 60034- 12.8 Voltage and frequency variationsDuring start-up, the motors must be suitable foraccelerating to 80% of the rated voltage.Motors must run at full h.p. under the followingconditions:Variation in frequency: f5%Variation in voltage: flO%Combined variation in frequency and voltage: f10%Motors must not stall due to a momentary dip involtage up to 70% of the rated voltage, i.e.7’ to be more than T,./(0.625)’, i.e. 256% for a1200 r.p.m. motor and TJ(O.605)’, i.e. 273% for a1000 r.p.m. motor and less, as shown in Table 1.5.Motors must run satisfactorily for 5 minutes at asupply voltage of 75% of the rated value.Nore This operating condition is, however, not specific forit does not stipulate the frequency of occurrence of such acontingency. It may be assumed that this condition will notoccur more than once before thermal stability is reached.See also Section 3.7. In normal practice a motor meeting theother operating conditions noted above in all likelihood willsatisfy this requirement also without needing yet anotherderating. See also Example 7.1,9 Starting currentOn DOL the I,, must not exceed 600% of I, but forboiler feed pumps it must be limited to 450% of I,,subject to the tolerance stipulated in IEC 60034-1.To determine the starting current, when the test isconducted at a reduced voltage (1/& Vr ),allowance must be made for the saturation effectwhile estimating the value of I,, at the rated voltage.See also Section 1.6.2(A) also (the important noteon starting torque).NoteAt full voltage, the stator core will offer a lowerimpedance, because of the saturation effect, than at a reducedvoltage. The Zst at full voltage therefore would be slightlyhigher than that calculated from the starting current, measuredat a reduced voltage. Hence the allowance for this. See alsoFigure 27.2(b).10 Frequency of start: The HT motors must be suitablefor two starts in quick succession when the motor ishot or three equally spread starts in an hour (not to berepeated in the second successive hour).11 Stresses during a fast bus transfer: A fast bus transferbetween the unit and station supplies is an essentialfeature in a power station to maintain an uninterruptedpower supply to the unit and station auxiliary services(Figure 13.21). The changeover may be as fast asthree or four cycles which is adequate to break beforemake in the modern interrupting devices and hence issafe to adopt. The motors should, however, be suitableof withstanding excessive voltages on account of thisup to 150% of the rated voltage for at least one secondduring such bus transfers.Nure The time gap between a fast bus transfer is so small thatit will not permit the various drives to slow down sufficiently toaffect their performance or the system if they are made to reaccelerate(reenergize) on restoration of power.In a fast bus transfer scheme, the phase angle ismonitored through special relays (Section 16.11) whichinitiate the changeover, so that the self-induced e.m.f.of the motor does not slip too far in the phase angle,A@, between the two voltages and the motor is notexposed to excessive overvoltages. At higher A6 therelays will block the transfer and protect the systemfrom excessive voltages. If the fast bus transfer failsto change over automatically, slow changeover takesplace and the motor-induced e.m.f. is allowed to decayby up to 40-20% before the bus transfer takes place.For insulating and bracing the windings, such bustransfers may occur up to 500 times in the total lifespanof the motor.12 Starting time and locked rotor withstand timeFor motors with a starting time of up to 20 seconds,with the minimum permissible applied voltage(80%), the locked rotor withstand time under hotconditions at rated voltage must be at least 2.5seconds more than the starting time.For motors with a starting time of more than 20seconds but within 45 seconds, at the minimumpermissible applied voltage (SO%), the locked rotorwithstand time, under hot conditions at rated voltage,must be at least 5 seconds more than the starting time.13 Bearings: The bearings will preferably have an arrangementfor self-lubrication.14 Bearings must be insulated wherever necessary, toprevent them from shaft currents (Section 10.4.5).15 Over-speed: Motors will be designed to withstand120% of ff, for at least 2 minutes without anymechanical damage.16 Noise level: Motors must conform to the requirementsof IEC 60034-9. A safe sound pressure level for ahuman body to perform better, during an 8-hour

Special-purpose motors 7/189Table 7.9levelSources emitting sound and their likely loudnessNoise level (dB)Sound source0 Threshold of hearing20 Rustling sound, whisper, homes(quiet places)40 Motor cars60 Normal conversation80 Street traftic100 Light engineering workshop120 Thunderllightningworking day, without undue fatigue, is considered at90 dB as noted in Section 7.9. For recommendedvalues of sound level for airborne noise, refer to theIEC publication. See also Table 7.9 for sources inday-to-day life that emit noise and their likely noiselevels.17 Temperature detectorsHT motors must have a minimum six numbers ofresistance temperature detectors (RTDs)(See section12.8).All bearings of HT motors must be provided withembedded temperature detectors (ETDs).18 Polarization index testMotors rated 7500 kW and less must be consideredsuitable for dielectric tests or operation only whenthe polarization index or the value of the insulationresistance (at 40°C) is at least the minimumrecommended values.Motors rated above 7500 kW must have both thepolarization index and the insulation resistance abovethe minimum recommended values.The recommended minimum value of thepolarization index for motors having class B or Finsulation must be 2 when determined according toIEC 60034- 18- 1.Note The above recommendations are those of an electricityauthority of one country and may vary for other countries.7.20 Selection of a special-purposemotorExample 7.1Select a suitable motor for a ball mill with the following loaddetails:kW = 450r.p.m. = 400 through V-beltsStartina - torque . = 40% rising to 100% at full load(Figure 7.51).Gd of rotating masses = 10 000 kgm’Supply system:Voltage = 3.3 kV f 10%Frequency = 50 Hz ? 3%Combined voltage and frequency variation not to exceedf 10%. Voltage may fall to 80% during running for about 20s= 1Figure 7.21- SpeedSlipat 80%voltageDetermining the accelerating torqueminutes and can be assumed to occur once an hour. Alsolikely overloading by 20% for the same period in the sameduration. Only one contingency occurring at a time. The motorshould also be capable of running, without stalling if the voltagedrops to 70% momentarily, say, for 25 cycles.Ambient temperature = 50°CPermissible starting = DOL in a squirrel cage or rheostat ina slip-ring motorDesired starts = six equally spread starts per hour andthree consecutive hot starts not tobe repeated for one hourService factor = 1.1, alternatively 1.15SolutionDeratings(a) For +lo% voltage and frequency variation: 90%(Section 1.6.2(C)).(b) For 50°C ambient temperature: 92% (Section 1.6.2).(c) At 80% voltage, the full load torque of the motor will dropto 0.73*, Le. 0.53 times the rated torque, as shown inTable 1.5. Also the motor will tend to stall unless adequatetorque is available on the motor torque curve, i.e. Tpomust not be less than 1/0.53 or 189%. The motor will thusstart to drop its speed until it reaches a point where amotor torque of 189% of 450 kW is available. The motorwill now operate at a higher slip, causing higher slip losses.Assuming the mill torque at the reduced speed to be thesame as at 100% speed, then the kW requirement of themillkW- N. TThis will also decrease in the same proportion as theincrease in slip. For a rough estimate, we may also ignorethe higher slip losses for an equal reduction in the requiredkW. However, due to the lower voltage the motor currentwill increase proportionately and will be IJ0.8 or 1.251,.The motor will thus run overloaded by 25% for 20 minuteswhich is likely, and not more than once an hour. Thefrequency of occurrence must be known. A higher deratingmay be necessary if such a condition is frequent.(d) Overloading of 20% for 20 minutes per hour need not be

7/190 Industrial Power Engineering and Applications Handbookconsidered in view of condition (c) which is more severe.Therefore, average loading on motor in one-hour cycle/(1.25Ir)’ x20+/: x40=I 60= I, \lm= 1.091,i.e. the rating of the motor should be higher by 9% toaccount for this stipulation.Note Normally, the heating-up time constant T for an HTmotor varies between 0.9 to 1.5 hours, i.e. heating-uptime under normal running condition from 2.5 to 4.0 hoursapproximately. In one hour, therefore, the motor will notreach its thermal equilibrium and the equivalent rating asdetermined above is in order. One should not consider acycle longer than heating time, which may give incongruousresults::. kW to be chosen, considering all deratings- 450 x 1.090.92 x 0.90= 600 kWA 1000 r.p.m. 600 kW motor should be an economicalchoice.(e) To meet the condition at 70% voltage, the motor shouldhave a T of more than 189% and should be minimum1/(0.605)go (Table 1.5) or 273% of 450 kW or 204.75% of600 kW, so that the motor does not stall.(f) To consider the service factor (SF) the motor should havea continuous reserve capacity as may be desired, forexample for a SF of 1.15, a motor of at least 1 .I5 x 450,i.e. a 517.5 kW rating must be required. We have alreadyselected a motor of 600 kW, therefore this factor need notbe considered again. Moreover, too large a motor thanthe load requires would make it operate underloaded anddiminish its operatinq efficiency and P.f., which is a drainon the usable energy and is not desirable. See also Sections23.3 and 7.10.Checking the suitability of a squirrel cage motorAssume that the motor is designed for an average speeduptorque of 135% and Tpo of 220% (Figure 7.21). If theaverage load torque is assumed as 68%, the averageaccelerating torque, T,, available will be 67% on DOLstarting, i.e.T, =450974 x 0.67 rnkg (considering N, = 980 r.p.rri.)980- 300 mkgSay, the GD; = 200 kgm22and GD; = 10000 (%) (at motor speed)-- 1666 kgm2:. GD; = 1866 kgm2and starting time,t, =1866 x 980375 x 3002- 16.25 secondsConsidering the permissible tolerances in the declaredtorque values, the safer starting time should be taken asroughly 1&15% more than the calculated value. Therefore,consider the safe starting time as 16.25 x 1.15, i.e. 19.0seconds. For a single start under hot conditions the motorshould have a minimum thermal withstand time of 19.0seconds, and for three consecutive starts, a minimum57.0 seconds, which, for this size of motor, may not bepracticable in view of design economics. For this application,therefore, a squirrel cage motor is not recommended,unless a fluid coupling is employed to transmitthe power.CorollaryThe use of a squirrel cage motor is, however, inevitableas discussed earlier particularly for a process plant or apower house application, where downtime for maintenanceof a slip-ring motor is unwelcome, or a chemical plant orcontaminated locations, where the application of a slipringmotor is prohibitive. To meet such load requirementswith a squirrel cage motor, the use of fluid couplings tostart the motor lightly and reduce starting time, is quitecommon and economical, as discussed in Chapter 8, andmust be adopted in the above case.Let us use a fluid coupling and start the motor lightly,similar to Figure 8.3. The revised accelerating torqueand approximate clutching sequence of the coupling isillustrated in Figure 7.22.1 Consider the clutching of the coupling with the loadat roughly 0.8Nr, by which time the motor would beoperating near its Tpo region:. N,, = 0.8N,= 800 r.p.m.Considering an average coupling opposing torque asroughly O.4Tr and average motor torque as 1.35Tr,between 0 to 0.8 Nl., the exact torque can be calculatedby measuring the torque ordinates at various speedsand then calculating the average. See Example 2.4for more details.... Tal = 1.35Tr - 0.4Tr= 0.95Tr- 450 x 974 o,95980= 424.88 mkgGD2 during a light start, considering the GD: of thecoupling (impeller) as equal to the motor (the exactvalue should be obtained from the manufacturer)... GD; = 200 + 200= 400 kgm2and t,, up to 0.8Nr =400 x 800375 x 424.88= 2.01 seconds2 The coupling average accelerating torque after it has

Special-purpose motors 711 91Coupling speedFigure 7.22 Accelerating torque with the use of fluid couplingengaged with load, up to Nr, Tr2 z 1 6ST,(b) calculating average ordinate\ at points I to 6= i[XO + 135 + 185 + 215 + 215 +;(215 + 1001= 165%Lo'id aberage torque E 0.68Tr.. TCi2 = 1.65T, - 0.68T,and= 0.97Tr == 433.33 mkg450 x 974 x 0.97980GD; = GD

~7/192 Industrial Power Engineering and Applications Handbook‘2=[m] 600 x8= 32 secondsTo determine the number of equally spread starts andstops, heating and cooling curves of the motor should beavailable and the manufacturer may be consulted.The above exercise is only for a general guidance. Insuch cases, the manufacturer should be consulted whomay offer a better and more economical design.Relevant StandardsIEC Title IS BS IS060034-1/199660034-3/198860034-5/199160034-6/199160034-7-199260034-9/199760034- 12/198060034-14/199660034-18- 1/199260072- 1/199 I60072-2/199160072-3/ 199460079-0/199860079-l/199060079-2/198360079-7/ 199060079-10/199560079- 1 I/ 199 160079- 12/197860079- 14/1996Rotating electrical machinesRating and performanceRotating electrical machinesSpecific requirements for turbine typesynchronous machinesRotating electrical machinesClassification of degrees of protectionprovided by enclosures for rotating machineryRotating electrical machincsMethods of cooling (IC Code)Rotating electrical machinesClassification of types of constructions andmounting arrangements (IM Code)Rotating electrical machinesNoise limitsRotating electrical machinesStarting performance of single-speed three-phase cageinduction motors for voltages up to and including 660 VFor agricultural applicationRotating electrical machines - mechanical vibration ofcertain machines with shaft heights 56 mm and above.Measurement, evaluation and limits of vibrationRotating electrical machines. Functional evaluation ofinsulation systems. General guidelinesDimensions and output series for rotating electricalmachines. Frame number 56 to 400 and Flange number55 to 1080Dimensions and output series for rotating electricalmachines. Frame number 355 to 1000 and Flangenumber 1 180 to 2360Dimensions and output series for rotating electricalmachines. Small built-in machines. Flange numberBF 10 to BF 50Electrical apparatus for explosive gas atmospheresGeneral requirementsConstruction and verification test of flameproofenclosures of electrical apparatusElectrical apparatus. Type ‘p’Electrical apparatus for explosive gas atmospheresIncreased safety motors, type ‘e’. GeneralrequirementsElectrical apparatus for explosive gas atmospheresClassification of hazardous areasSpecifications for intrinsically safe electrical system ’i’General recommendations on the selection of theappropriate group of electrical apparatus based on safegaps and minimum igniting currentsElectrical apparatus for explosive gas atmospheresElectrical installations in hazardous areas (other thanmines)472211 9928 15 1/1990469 1/19856362/19952253/197412065/19878789/ 19967538/1996-78 I6/199l123 1/199 1I23 1/1991996/199 12 148/19932 148/19937389/199 1638 1/199 15572/19943682/19915571/1991BS EN 60034-1/1995 -BS 5000-99/1986BS EN 60034-3/1996 -BS 4999-105/1988 -BS EN 60034-6/1994BS EN 60034-7/1993BS EN 60034-9/1998 -BS EN : 60034-12/1996 -- -BS 4999-l42/1987 2373BS EN 60034-18-1/1994BS 5000-10/1989BS 4999-141/1987BS 5000-10/1989BS 4999-103/1987BS 5000-1 1/1989BS 5501-1/1977,BS 5000-17/1986BS 4683-2/1993BS 5501-5/1997BS EN 50016/1996BS 5501-6/1977BS 5000-15/1985BS 5501-7/1977BS 5501-9/1982BS 5345-1/1989BS EN : 60079-14/1997

Special-purpose motors 7/193IEC Title IS BS IS060079- 15/1987 Electrical apparatus with type of protection ‘N’ 9628/19888289/199160079-1711996 Inspection and maintenance of electrical installations inhazardous areas (other than mines)60079- 19/1993 Repair and overhaul for apparatus used in explosiveatmospheres (other than mines or explosives)6OO79-20/1996 Data for flammable gases and vapours for electricalapparatus60317(0-47)Requirements and dimensions for round copper wire6085111996Winding wires for submersible motorsTextile motors: Part 1 Loom motorsTextile motors: Part 2 Card motorsTextile motors: Part 3 Spinning frame motorsMotors for submersible pump setsSubmersible pump setsSpecification for general purpose electrical powerdriven winches for lifting and haulingCode of practice for safe use of cranesSpecification for power driven overhead travellingcranes, semi-goliath and goliath cranes for general useSpecification for power-driven mobile cranesSpecification for electrically driven jib cranes mountedon a high pedestal or portal carriageSpecification for high voltage busbars and busbarconnections8783/19952972-1 19902972-11 19902972-Ili 19909283/19958034/19899507119953177/19904137/1995457311990459411995Related US standards ANSI/NEMA and IEEEBS 6941/1988BS 5000-16/1997BS EN 60079-17/1997BS EN 60317 (0-47)BS EN 60851/1998808411992 BS 159/1992-BS 574411979BS 466/1984 4301-18306BS 175711986BS 2452/1954NEMA/MG 111993NEMA/MG 2/1989NEMAiMG 3/1990NEMA/MG 1011994NEMA/C50.41/1982Motors and generators ratings, construction, testing and performanceSafety Standards (enclosures) for construction and guide for selection, installation and use of rotating machinesSound level prediction for insulated rotating electrical machinesEnergy management guide for selection and use of three phase motorsPolyphase induction motors for power generating stationsNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedDetermining the size of motorFor lifh’ng and hoisting%=- F ’V kWF = useful load in kgf(7.1)V = lifting speed in mls77 = efficiency of the mechanismFor traverse1.027 . T,,; N,p,=kW1000. q(7.2)T, = maximum torque77 = efficiency of the whole mechanism for traversing

7/194 Industrial Power Engineering and Applications HandbookFurther readingI Anderson, H.H., Submersible Pumps and their Appliccrtion,The Trade and Technical Press Ltd, Morden, Surrey.2 Battiwala, N., ‘Centrifugal motors for the sugar industry,’Siemens Circuit, 111, No. 4, July (1968).3 Brown Boveri & CIE, Explosion Protection Mnnuul,Aktiengesellschaft Mannheim Verlag, W. Giradet, Essen,Germany.4 Central Board of Irrigation and Central Electricity Authority,Standardisution of Motors for Thermal Power StationAuxiliaries in India, Power Publication No. 140, March (1983).5 The Clzemicd Times, ‘Design and application of electric motorsin explosive atmosphere’, IV, No. 15, April (1977).6 Ghosh, S.N., ‘Electric motors for hazardous environments’,Siernem Circuit, VI, No. 4, July (1971).7 Ghosh. S.N.. ’Electric motors for special applications’,Siemens Circuit, VII, No. 4, July (1972).8 Ghohh, S.N., ‘Induction motors for auxiliary drive for thermalpower stations,’ Siemens Circuit, XVII, No. 2, April (1982).9 Jain, S.K., ‘Sugar centrifuge motors’, Siernrns Circuit, VI,No. 2, Jan. (1971).IO Singhvi, N., A paper on ‘Motors for hazardous applications’.Crompton Greaves Ltd, India.

-"Transmission ofLoad andSuitability ofBearings-

Transmission of load and suitability of bearings 8/197The power from the motor shaft to the driven shaft canbe transmitted in many ways, depending upon the powerand load application. In this chapter we deal with themost common types of driving systems, their influenceon the starting characteristics of the motor and their effecton bearings.8.1 Direct or rigid couplingsThese are suitable for smaller ratings only because oftheir weight, which also includes the weight of the loadwhich falls directly onto the motor and its shaft. Thedriven equipment, such as a mono block pump is mounteddirectly on the motor shaft (Figure (8.1). Such couplingsoffer the advantage of design simplicity but do not allowfor any misalignment. They are used for low-speedapplications or where only marginal misalignment isanticipated.8.2 Flexible couplingsThe load is now transmitted through ordinary couplings,one half mounted on the motor shaft and the other halfon the driven shaft. They are bolted together with a rubberpad between the two. The motor and the driven machineare mounted on a common bed (Figure 8.2). They arenow able to provide margin for misalignment of the twoshaft ends and thus extend more flexibility.8.3 Delayed action couplingsIn the above two types of couplings the transmission oftorque is linear, but now the drivin torque rises as thesquare of the input speed (T = N 5). These couplingsfacilitate engagement of the motor shaft with the drivenshaft after a pause, by when the motor nearly picks up itsrated speed. They are of two types:Figure 8.1 Direct coupling (mono block pump)(Courtesy: NGEF Ltd.)IFigure 8.2 Flexible coupling (Courtesy: Crompton Greaves Ltd)1 Fluid or hydraulic couplings These are available intwo designs:0 Constant filled or traction type andVariable filling or scoop control type2 Magnetic couplingsDelayed action clutching helps the load in the followingwavs:fi enables the motor to have a soft start. The motorpicks up lightly and quickly and reduces the pressureon the supply source.The motor itself undergoes less stress.This is a type of a mechanical clutch that enablesthe motor start almost at no-load, irrespective ofthe type of load.These couplings are adjustable and therefore canalso facilitate speed control.Since they work like a mechanical clutch, and engageonly slowly and gradually, these couplings enablethe motor to have a light start, on the one hand, anda smoother engagement of the load shaft, on theother. It avoids a jerk on the load and an excessivetorque demand on the motor. In normal couplings,a high inertia load results in a longer start anddemands a high starting torque, and requires eithera higher rating or a specially designed high startingtorque motor. With the application of these delayedaction couplings, it is possible to use a lower ratingand a normal torque motor. Proper application ofthese couplings can thus result in low initial cost ofelectrical accessories (motor, cables, control gearetc.) and a substantial saving on energy by- Selecting a normal low h.p. motor, even forstringent load requirement and adopting a DOLswitching.- Minimizing the starting losses and strain on thefeeding lines, as the motor now starts up lightlyand accelerates quickly.- Eliminating damper, vane or throttle controls,which are a constant source of energy drain fordrives having variable load demands. This ispossible by achieving automatic speed regulationof the drive through variable-speed couplings.For automatic control of the couplings, sensingdevices may be placed in the flow circuit of pumps,fans or any drives that have variable load demands

8/198 Industrial Power Engineering and Applications Handbookduring their run. (See Figure 6.38.) The couplingscan thus control the flow of air or fluid byautomatically regulating the speed of the drive andusing only the amount of power that is needed bythe load, which results in substantial energy saving.See Section 6.15.2, which analyses the amount ofenergy saved.Environment-friendly couplingsThese couplings are totally enclosed and are suitable forany environment prone to fire hazard, corrosion, dust orany other pollutants. The controls can be providedremotely in a safer room and the pushbutton stations,which can be easily made suitable for such environments,located with the drive.8.4 Construction and principle ofoperation8.4.1 Fluid or hydraulic couplingsThese couplings are made in two parts, the impeller asthe inner part and the runner (rotor) as the outer. Bothare enclosed in a casing. The impeller and runner arebowl shaped and have large number of radial vanes asshown in Figure 8.4. They are separated by an air gapand have no mechanical interconnection. The impeller ismounted rigidly on the motor shaft while the runnerrotates over the impeller, with a uniform air gap. Theimpeller is filled with fluid (mineral oil) from a fillingplug. A fusible plug is provided on the impeller, whichblows off and drains out oil from the coupling in theevent of sustained overloading, a locked rotor or gtalling.As the motor picks up speed, it builds up a centrifugalforce in the impeller, which causes the fluid to spill outand fill the air gap between the two couplings. As the airgap is filled only gradually, the runner engages with theimpeller gradually. The impeller converts the mechanicalpower into hydraulic power and the runner converts thisback to mechanical power. The transmission of powerfrom the runner to the load may be through flexiblecouplings or belts. The torque transmitted during thestart is ro ortional to the square of the motor speed? P(T N-). There may be small speed variations atthe secondary side, which will be due to slip betweenthe impeller and the runner (generally of the order of 2- 5%). Thus the clutching at full speed is rigid. Thefollowing illustrates the clutching sequence.Consider a heavy inertia load such as a conveyor, havingcharacteristics according to Figure 2.13. Choose a standardmotor having the characteristics shown in Figure 8.3.The fluid coupling will have centrifugal characteristicsduring start (T 0~ N2) as shown. The load-transmittingcharacteristic of the coupling is pre-set (by varyingthe quantity of oil in it) so that the load will clutch atpoint A in the graph figure by which time the motor willreach about 80-90% of its rated speed with a torquesomewhere in its Tpo region. By now the motor's initialinrush current will also droop and it will be operatingalmost at its normal running condition. The motor willthus pick up lightly and smoothly with an acceleratingtorque, Tal, with no strain on it or the supply system.After point A the coupling will gradually clutch with theload with an accelerating torque, T,?, as illustrated inFigure 8.3. Ta3 being high, it will accelerate the loadquickly once again, without any strain on the motor orthe supply system.These couplings have been produced up to 10000 kW,N,, = Coupling speed(Indicates coupling slip)T, = Load starting torque'The load 1;clutches ,here I II 1IIIIII0 80% of N, N'1 Speed1- Motor speed(primary speed)(se~~~adda~~~~ed)Figure 8.3 Smoother and quicker acceleration of heavy-duty loads with the use of a delayed-action coupling

Transmission of load and suitability of bearings 8/199Radial vaneslmoeller Rotor CasingFigure 8.4Constant filled fluid couplings (Courtesy: Fluidomat Ltd)by GE, USA. In India they are available up to 2500-3000 kW.1 Constant filled or traction type couplings(constant-speed couplings)These are pre-filled fluid couplings. The quantity of oilfilled in the coupling cannot be varied during running,hence the name. The quantity of fluid can, however, bechanged depending upon the starting or the loadrequirement. They have a safety device in the form of afusible plug. The plug blows off and drains out oil fromthe coupling to provide protection against an excessivetemperature rise, which may occur due to sustainedoverload or stalled conditions, as well as protecting themotor. Figures 8.4 and 8.5 illustrate these couplings andFigure 8.6 shows a general coupling arrangement. Theoperating slip is in the range of 2-5%. Since they arepre-filled, they are constant-speed couplings.Applications include electric motor drives for conveyorsand other material handling equipment such as stackerreclaimers, crushers, haulages, ball mills, cranes, hammermills, rotary dryers, centrifuges, reciprocating pumps,winches, fans and wire drawing machines.In view of their easy adaptability, they are manufacturedin standard sizes suitable for common types of loadapplications. They are therefore available on shortdeliveries, in smaller ranges, say, from 1 kW to 200 kW.They can be selected for the required service conditions,as a motor is selected to suit a particular load requirementfrom the available sizes of couplings.2 Variable-filling or scoop control-type couplings(variable-speed couplings) (Figure 8.7)Through such couplings the output speed of the runnercan be changed by varying the volume of oil in the workingcircuit through the scoop operating lever, as shown inFigure 8.7(b). When the oil volume is full, slip is atminimum and the output speed is maximum. As the oilcircuit is emptied, the slip gradually increases. A constantspeedmotor can thus be used to provide a stepless variablespeeddrive. Speed variation is possible up to 15-20% ofN, for centrifugal loads such as fans and pumps.Figures 8.7(a) and (b) illustrate variable-speed couplingswhich can provide a stepless speed variation over a widerange. The impeller and runner of the couplings are housedin a stationary housing with a built-in oil sump. Oil is

8/200 Industrial Power Engineering and Applications HandbookFigure 8.5 Constant filled fluid couplings (Courtesy: Pembril Fluid Drives)Figure 8.6 General arrangement of a constant filled couplingcontinuously introduced into the working circuit througha separately driven oil pump. A scoop tube is provided toregulate the volume of oil in the working circuit andhence control the speed and the transmitting torque. Theposition of the sliding scoop tube can be governed by aservo-actuator or manually. The coupling can be attachedwith the motor on one side and load on the other throughflexible couplings as shown in Figure 8.7(a).Regulation of the quantity of oil in the working circuitbetween the impeller and the runner is effected by anadjustable scoop tube, which slides in the stationaryhousing (Figure 8.8). When the fluid coupling is at rest,the oil level is below the opening in the casing. With thescoop tube retracted radially inwards, on starting up theset the oil forms a rotating ring in the reservoir casingdue to centrifugal force. The casing is of sufficient capacityto contain the full quantity of oil, clear of the tip of thescoop tube, so that the working circuit remains empty,when the drive is disconnected (Figure 8.8(a)).By sliding the scoop tube radially outwards its tipenters the rotating ring of oil and a quantity of oil ispicked up by the scoop tube and transferred to the workingcircuit. With the working circuit partially filled (Figure8.8(b)) the output shaft is driven at a reduced speed orthe torque capacity of the drive is restricted. With thescoop fully extended (Figure 8.8(c)) all the oil is transferredto the working circuit, bringing the coupling output shaftup to full speed. While the coupling is running, oil escapesfrom the working circuit into the reservoir casing throughsmall leak-off holes, and is returned to the circuit by thescoop. The radial position of the scoop tube determinesthe depth of the oil ring in the reservoir casing and therebythe volume of oil in the working circuit.Fluid couplings have a proven record of reliability andease of operation. They thus have found wide applicationwherever a light load start and capacity control of drivesis required as in fans and pumps.The range is from 90 kW to 10 000 kW and applicationsare as follows:Steel works conveyors, crushers, wagon tipplers, skidgears, furnace charges, bogie drives in bogie hearthfurnaces, furnace winch drives, cranes, pumps, compressorsand fans.Power house auxiliaries ID fans, FD fans, primaryand secondary air fans, boiler feed pumps and slurry

Transmission of load and suitability of bearings 8/201pumps, coal mill drives, conveyors, pulverizers andwagon tipplers.Docks and harbours Material handling, mining,railway traction, process and chemical plants.Most of these applications are heavy inertia loads andrequire a light load start. The use of such variable drivescan save energy.Advantages of variable-speed fluid couplings1 It is possible to switch the motor at no load by selectinga variable-speed fluid coupling. The oil need not beprefilled which can be varied at site during operationas required.The couplings can be designed to develop torqueseven higher than the Tpo of the motor. They can thusbe made compatible for transmitting loads up to theoptimum capacity of the motor and sustaining abnormalload conditions without stalling or damage.Basically a variable drive fluid coupling is a tailormadeclutch to suit a specific load duty for moreaccurate application.They are suitable for stepless speed variation and canbe controlled through speed, torque, temperature orflow of a process. They can also be programmed forany sequence of operation.Conventional throttling or damper control of flow iswaste of energy. The use of such couplings can varyspeed through oil control, relieve strain on the systemand save on the energy consumption.1. Input shaft2. Upper housing3. Impeller4. Rotor5. Primary casing6. Secondary casing7. Scoop tube8. Output shaft9. Bottom housing10. Flexible coupling11. Flexible coupling12. Working oil circuitFigure 8.7(a) Variable-speed scoop control fluid coupling (Courtesy: Fluidomat)

8/202 Industrial Power Engineering and Applications Handbook8.4.2 Magnetic couplings and eddy currentcouplingsThe principle of operation of these couplings is almostthe same as of fluid couplings except the air gap betweenthe impeller and the runner of the coupling, which isnow filled with iron granules or iron powder instead offluid. The iron powder condenses into a solid mass whenmagnetized through an external exciter. The exciter ismounted on the same coupling, and clutches the runnerwith the impeller. The power of transmission as well asthe speed of the runner can be controlled by varying theexcitation.Eddy current couplings have become more common.In this case the impeller of the coupling is a ferromagneticdrum, coupled to the induction motor and housed in anouter shell (Figure 8.9(a)). The shell is fitted with amagnetic yoke and an excitation coil. Within this drum,a multipole inductor, made from special alloy steel, rotatesfreely, maintaining a close and constant air gap. Thisforms the driven part or the runner of the coupling. Whenthe excitation coil is energized, it develops a magneticfield in the coupling’s runner and impeller. The relativemotion between them causes the magnetic field toconcentrate on the surface of the drum. This results in aflow of eddy currents in the metallic drum which causesa torque transmission from the prime mover to the runner.The transfer of load from impeller to runner is performedwithout any mechanical connection. The torquetransmitted is a function of the d.c. excitation. By varyingthe d.c. excitation, the output speed of the drive can bevaried smoothly even up to 7-10% of N,. Similarly, theoutput speed can be maintained constant, even on a changeof load, by sensing the output speed and monitoring theexcitation level through a closed-loop speed feedbackcontrol system. The speed can be sensed through atachogenerator mounted integrally on the load shaft. Veryaccurate feedback control systems are also possiblethrough microprocessor-based analogue and digitalcontrols, as discussed in Sections 6.6 and 13.2.3 to achievetotal automation of speed, torque and power or any otherprocess parameter such as, flow of liquid, gas or temperatureetc.Figure 8.9(b) illustrates a general scheme to achievespeed control through such couplings. These couplings,up to 90 kW, are easily available.ApplicationThis is as for fluid couplings but at lower ratings.Generally, they are used in the cement, rubber andchemical, paper, chemical fibre, electric wire makingand mining industries, as well as in material handling,conveyors and thermal power plants.8.5 Belt drivesBelt drives are employed to transmit load from the drivingshaft of the motor to the driven shaft when they areseparately located. There are two types of belting availablefor industrial use.

Transmission of load and suitability of bearings 8/203Inputshaft(a) Oil in reservoir, circuit empty, drive is disengagedFigure 8.9(a) Eddy current drive (Courtesy: Dynodrive)Inputshaft. outputshaft(b) Circuit partly filled, drive is partly engagedoutputshaftA@ DC-excitation voltage regulator.@ Speed setting rheostat.@ Speed comparator.@) Excitation coil.@ Tacho-generator.NR - Required speed possible up to 7-1 0% of N,.NA - actual speed.AN - Speed error.Figure 8.9(b) A general scheme illustrating speed control byfeedback control system using an eddy current couplingFlat beltsV-belts(c) Oil in circuit, drive is fully clutchedFigure 8.8 Illustration of clutching of a scoop control variablefluid couplingThe flat belt is a friction drive transmitting load throughthe friction between the belt and the pulley while the V-belt is a positive drive which is flat on one side and hasa projection like a geartooth on the other. These

8/204 Industrial Power Engineering and Applications Handbooktoothshaped belts seat into pulley-matching grooves.For transmission of smaller loads, flat belts and forsmall and medium loads, V-belt drives are normallyemployed.8.5.1 Flat beltsThese are long centre drives with small slips. The slackside of the belt is kept on the top side to increase theangle of contact with the pulleys by sag on the top side.This is essential for an efficient transfer of load. Therecommended maximum power that can be transmittedby one belt of different cross-sectional areas is providedby the belt manufacturer and some ratings are given inTable 8.1. When selecting these drives, the followingparameters should be borne in mind:1 The ratio of the diameter of the pulleys should notexceed 6: 1 unless a jockey (idler ) pulley or similararrangement is made to press the top side to indirectlyincrease the arc of contact.2 Similarly, for shorter centre distances between thedrive and the driven pulleys, the arc of contact willdecrease. To ensure a good arc of contact, the centredistance C (Figure 8.10), should be kept as much aspossible, otherwise the provision of a jockey pulley,as noted above, will also be necessary. A higher arcof contact will ensure a better grip of the belt on thepulley and hence a smaller slip during transmissionof the load. A smaller slip would mean a highertransmission of load and vice versa.3 Vertical and right-angle drives should be avoided.4 Belts should not be tightened more than necessary,otherwise the drive and the driven shafts will comeunder torsion and excessive bending moment. Thebearings would also be subjected to excessivestresses.Specification of flat beltsThese belts are made of cotton duck (a cotton fabricused in making canvass and tents) with different mixesto provide them with a degree of hardness, as noted inTable 8.2.The duck is glued in thin layers (plies) with a rubbercompound and vulcanized (cured). The number of pliesused to make a belt and their quality defines the strengthof the belt (e.g. 3 ply x 32 or 3 ply x 34, etc.).Selection of flat belts for transmission of loadThe selection of flat belts is made along similar linesto that for V-belts (discussed later in more detail). Theload-transmission capacity of a flat belt can be definedbyw=SFP.Correction for arc of contact(8.1)whereP = load to be transmitted in kWW = maximum load-transmission capacity of the belt.By convention, this is provided by the belt manufacturerper 25 mm of belt width, at different speedsof the faster shaft (smaller pulley) for differentrecommended widths of belts, number of plies andtype and grade of duck, etc. We show Type I andType 11 in Table 8.1 for a general illustration. Theseratings may vary marginally from one manufacturerto another, depending upon their product mix andquality of curing.SF (Service factor) - as in Table 8.5.Correction for arc of contact - as in Table 8.7.In the following example we illustrate a brief procedureto select a flat belt for transmission of a load.Example 8.1Determine the width of a heavy-duty double-leather flat beltfor transmitting a load of 18.5 kW at 1480 r.p.m. from asquirrel cage motor to be switched directly on line. Diametersof pulleys are 250 mm on the motor side and 200 mrn on theload side. The centre distance between the pulley may beconsidered as 800 mm.P= 18.5 kWSF = 1.4 for heavy-duty Class 3, as shown in Table 8.5,presuming operations for 10 hours/day.Correction for arc of contact:Corresponding to D-d - 250 - 2oo (Figure 8.10)C 800= 0.0625The correction factor from Table 8.7 = 0.99. w= 18.5 x 1.4..0.99= 26.17 kWThe belt must be rated at least for this load.Speed of the faster pulley =1480 x 250200= 1850 r.p.m. (a)or K x 2oo60 x 1000= 19.36 m/s, which is quite low(< 30 m/s) hence acceptable.Consider a belt width of 200 mm to transmit this load.Therefore minimum belt rating per 25 mm = -25x 26.17200= 3.27 kW (b)For parameters (a) and (b) the possible belt sizes can be1 Four-ply 28 soft duck or four-ply 31 hard duck, accordingto Table 8.1 Type I, having a rating of 3.25 kW at 1800r.p.m. and 3.45 kW at 2000 r.p.m., or2 Three-ply 32 soft duck or three-ply 34 hard duck, accordingto Table 8.1 Type II, having a rating of 3.32 kW at 1800r.p.m.For a more judicious selection of belts, keeping economics inmind, it is advisable to seek an opinion from the manufacturer.

~ ~~ ~~ ~Transmission of load and suitability of bearings 8/205Table 8.1 Power ratings (kW/25 mm) for different widths of flat belts with 180" arc of contact on a smaller pulleyType 1Belt Speed of faster Smaller pulley diameter (mm)type shaft (cp.m.)100 125 160 200 250 315 400720 0.48 0.78 1.15 1.61 2.19 2.85 3.63960 0.62 1 .oo 1.45 2.05 2.77 3.55 4.381440 0.87 1.39 2.01 2.80 3.70 4.59100 0.09 0.14 0.20 0.29 0.39 0.53 0.67200 0.16 0.25 0.37 0.53 0.73 0.95 1.24300 0.23 0.37 0.53 0.75 1.03 1.37 1.77400 0.29 0.47 0.69 0.97 1.33 1.74 2.264 x 28 500 0.35 0.57 0.84 1.18 1.61 2.11 2.72and 600 0.41 0.67 0.98 1.38 1.88 2.46 3.154x31 700 0.47 0.77 1.12 1.58 2.14 2.78 3.55800 0.53 0.86 1.25 1.76 2.39 3.10 3.92900 0.59 0.95 1.38 1.94 2.62 3.40 4.261000 0.65 1.04 1.51 2.12 2.85 3.67 4.561200 0.75 1.20 1.75 2.44 3.26 4.15 5.031400 0.85 1.36 1.98 2.74 3.63 4.53 5.321600 0.95 1.51 2.18 3.01 3.95 4.811800 1.04 1.66 2.37 3.25 4.192000 3.45 4.382200 3.61 4.492400 3.73Type 11Belt Speed of faster Smaller pulley diameter (mm)rype sha8 (cp.m.)80 100 125 160 200 250 315720 0.44 0.61 0.85 1.21 1.62 2.09 2.61960 0.56 0.78 1 .os 1.55 2.06 2.64 3.261440 0.80 1.10 1.52 2.16 2.84 3.57 4.242880 1.88 2.52 3.39100 0.08 0.11 0.15 0.22 0.29 0.38 0.48200 0.14 0.20 0.28 0.40 0.53 0.69 0.87300 0.20 0.28 0.39 0.57 0.76 0.99 1.24400 0.26 0.36 0.51 0.73 0.98 1.27 1.59500 0.32 0.44 0.62 0.89 1.19 1.53 1.93600 0.37 0.52 0.72 1.04 1.39 1.79 2.253 x 32 700 0.43 0.59 0.83 1.18 1.58 2.04 2.55and 800 0.48 0.67 0.93 1.33 1.77 2.28 2.833 x 34 900 0.53 0.74 1.03 1.47 1.96 2.51 3.111000 0.58 0.81 1.12 1.60 2.13 2.73 3.361200 0.68 0.94 1.31 1.87 2.47 3.14 3.811400 0.78 1.08 1.49 2.11 2.78 3.50 4.181600 0.87 1.20 1.66 2.34 3.07 3.81 4.451800 1.32 1.82 2.56 3.32 4.072000 1.44 1.97 2.76 3.54 4.282200 1.55 2.11 2.93 3.73 4.422400 1.65 2.25 3.09 3.892600 1.75 2.37 3.23 4.002800 1.85 2.48 3.35 4.073000 2.58 3.443200 2.67 3.51Source IS 1370. See also IS0 22

8/206 Industrial Power Engineering and Applications HandbookTable 8.2-C-C>Dbut c d + DLarger pulleydia DFigure 8.10 Angle of contact with a flat-belt driveSpecification of flat beltsQuality Type of ducka Nominal weight(glm2)Soft 2832Hard 3134aAs in IS 5996Source IS 1370. See also IS0 228459509109708.5.2 V-BeltsThese are short centre drives unlike flat belt drives. Thebelt slip in such drives is negligible. The recommendedmaximum power that can be transmitted through suchbelts of different cross-sectional areas is provided by thebelt manufacturer. The normal cross-sections of V beltsin practice are given in Table 8.3. The cross-section of abelt depends upon the power to be transmitted and itsspeed. To select the appropriate section of the belt forthe required transfer of load refer to Figure 8.11 alsoprovided by the manufacturer. It is recommended thatTable 8.3numbersBelt sectionNominal cross-sections of V-belts and their codeArea of cross-sectionin mm2 W.T (approx.)Based on IS 2494-1: See also IS0 418422 x14\ I\ I32 x 19 40'38 x23(angle of V-belt)the pulleys be used with maximum possible diametersand distance between their centres be maintained as morethan the diameter of the larger pulley.Selection of V-BeltsThe load-transmitting capacities of a single V-belt, at180" arc of contact, are provided by the belt manufactureras standard selection data for the user for different areasof belt cross-sections and speed of the faster shaft. Wehave provided this data for a leading manufacturer, forbelts of section D, in Table 8.4, to illustrate the selectionof V-belts for the drive of Example 8.2.Figure 8.12 shows the arc of contact, i.e. the contactangle, the belt would make with the pulley. As for amotor a belt is also subject to unfavourable operatingconditions that require deratings depending upon theworking conditions, arc of contact and the length of thebelt selected as noted below:1 Service factor (SF)This will depend upon the type of drive, the torquerequirement and operations in hour per day etc. Thesubsequent deratings are the same for all manufacturersand we provide these in Table 8.5.2 Correction for length of beltThe longer the belt, the larger the load it can transmitand vice versa. These factors are also the same for allmanufacturers, as shown in Table 8.6.3 Correction for arc of contactThe standard ratings of belts are provided at 180" arcof contact. The smaller the diameter or the higher thespeed of the smaller pulley, the smaller will be thisangle and vice versa. The load-transmitting capacityof the belt diminishes with reduction of this angleand we show this factor in Table 8.7. A contact angleless than 120" would exert more centrifugal forces onthe motor shaft and the driving-end (DE) bearing. Ifthis is the case the shaft and the DE bearing, mayneed reinforcement, the provision of a jack shaft, oran additional support at the far end of the motor shaft.Figures 8.15 and 8.16 show the arrangement of a jackshaft and an additional support at the far end of theshaft respectively. Such a situation must, however, beavoided, as far as possible.In Example 8.2 we illustrate a step-by-step procedureto select the most appropriate size and lengthof belts and pulley sizes to transmit a particularload.Example 8.2Consider a reciprocating compressor operating in a processplant and using a motor of 110 kW, 980 r.p.m. The compressoris required to operate at 825 r.p.m. through V-belts. Theapproximate centre distance between the motor and thecompressor may be considered as 1 m.We can adopt the following procedure to select therecommended belt sizes:1 Determine the design power of transmissionDesign power = motor rating x SFAssuming the motor to be switched DOL and operating

~Transmission of load and suitability of bearings 8/207Figure 8.11 Selection of cross-section for V-beltsfor an average 14 hourdday, then the SF according toTable 8.5 = 1.3and design power = 11 0 x 1.3= 143 kWCalculate the speed ratio of the faster shaft to the slowershaft= -980= 1.19825Pitch dia. of larger pulleyThis is the same asPitch dia. of smaller pulleySelect the belt cross-sectionUse the selection curves provided by the manufacturer(which are almost the same for all manufacturers) (seeFigure 8.1 1). For a design power of 143 kW to be transmittedat 980 r.p.m., the recommended cross-section of beltsaccording to these curves is identified as D.Select the pulley diameterRefer to the manufacturer's catalogue for the recommendedpulley pitch and outside diameters. These data are providedin Table 8.8 for a leading manufacturer.whereL = belt pitch length (mm)C = centre distance between the two pulleys (mm).D = pitch diameter of the larger pulley (mm)d = pitch diameter of the smaller (faster) pulley (mm):. L = 2 x 1000 + 1.57(475 + 400) += 2000 + 1373.75 + 1.406= 3375.16 mm.(475- 400)'4 x I000We show the nominal inside length for various standardsections of belts in Table 8.9. According to the table thenearest standard belt available has a pitch length of 3330or 3736 mm. For a more closer length of belt, either changethe pulley size or select another make of belt or have itmade to order. But a length of 3330 seems reasonable,hence it is accepted.:. L=3330 mmThe centre distance will now be less than 1 m and can becalculated byConsider small pulley of d = 400 mm C = A + J ZThen the larger pulley D = 400 x 1.19where A = - n (D+ d)4 8= 476 mmThe nearest standard size available is 475 mm which is(D- d)'and 6=acceptable. 8Determine the belt pitch lengthThe pitch length of a belt is its circumferential length atthe pitch width of the belt. The pitch width of a belt isshown in Figure 8.12, and is provided by manufacturersas standard data. It can be calculated by(D- d)'L=2C+l,57(D+d)+-4c(These formulae are available in the product cataloguesof all leading manufacturers.)... A = 3330 - LL (475 + 400)4 8= 832.5 - 343.44(8.2) = 489.06

~Table 8.4Power ratings for section D V-belts, with 180" arc of contact with the smaller pulleySpeed offaster shaftPower rating for smaller pulley, with preferred pitch diameter ofAdditional power per belt, for speed ratio of(Kp.m.) 355 375 400 425 450 475 500 530 560 600 1.00 1.02 1.05 1.09 1.13 1.19 1.25 1.35 1.52 2.00mm mm mm mm mm mm mm mm nun mm to to to to to to to to to and1.01 1.04 1.08 1.12 1.18 1.24 1.34 1.51 1.99 overkW kW kW kW kW kW kW kW kW kW kW kW kW kW kW kW kW kW kW kW720 16.26 17.90 19.90 21.85 23.75 23.59 27.38 29.44 31.42 33.91 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25960 19.26 21.16 23.45 25.63 27.70 29.65 31.47 33.50 35.32 - 0.00 0.33 0.67 1.00 1.34 1.67 2.00 2.33 2.67 3.001440 21.22 23.03 - - - - - - - - 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50100200300400500600700800900m1100120013001400150016003.396.048.4110.5712.5514.3415.9617.3918.6219.6420.4320.9821.2721.2921.0320.463.706.619.2211.6113.8015.7917.5719.1420.4821.5722.4022.9623.2123.1522.764.087.3210.2412.9115.3517.5619.5421.2622.7223.8824.7425.2625.4225.214.458.0211.2414.1916.8719.3021.4623.3224.8726.0726.9127.3627.384.838.7212.2415.4518.3821.0123.3325.3126.9228.1428.9229.255.209.4213.2216.7019.8622.6825.1527.2228.8730.0730.7630.925.5710.1114.2017.9421.3224.3226.9129.0630.7331.8632.426.0210.9315.3719.4023.0326.2328.9631.1732.8133.826.4611.7516.5220.8524.7228.0930.9233.1634.7335.577.0412.8318.0422.7426.9130.4933.4135.6237.020.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.03 0.07 0.10 0.14 0.17 0.21 0.24 0.28 0.310.07 0.14 0.21 0.28 0.35 0.42 0.49 0.56 0.630.10 0.21 0.31 0.42 0.52 0.62 0.73 0.83 0.940.14 0.28 0.42 0.56 0.70 0.83 0.97 1.11 1.250.17 0.35 0.52 0.70 0.87 1.04 1.22 1.39 1.560.21 0.42 0.62 0.83 1.04 1.25 1.46 1.67 1.880.24 0.49 0.73 0.97 1.22 1.46 1.70 1.95 2.190.28 0.56 0.83 1.11 1.39 1.67 1.95 2.22 2.500.31 0.63 0.94 1.25 1.56 1.87 2.19 2.50 2.810.35 0.70 1.04 1.39 1.74 2.08 2.43 2.78 3.130.38 0.77 1.15 1.53 1.91 2.29 2.63 3.06 3.440.42 0.84 1.25 1.67 2.09 2.50 2.92 3.34 3.750.45 0.91 1.35 1.81 2.26 2.71 3.16 3.61 4.060.49 0.98 1.46 1.95 2.43 2.92 3.40 3.89 4.380.52 1.05 1.56 2.09 2.61 3.12 3.65 4.17 4.690.56 1.11 1.67 2.23 2.78 3.33 3.89 4.45 5.00CourtesyGoodyear

Transmission of load and suitability of bearings 8/209with an additional power-transmitting capacity for a speedratio of 1.19 = 1.74 kwlbelt.:. Total load-transmitting capacity = 25.62 kWlbeltDetermine correction factors(i) For the arc of contact shown in Table 8.7, which isthe same for all manufacturers. Corresponding to anarc of contact of - D- on the smaller pulleyCArc of.C\ I\ I?Angle\\ I/\ICross-sectionof a V-beltFigure 8.12of belt(475- 400)’and 6=8= 703.12Determining the belt details in a V-belt drive:. C = 489.06 + J489.06* - 703.12= 489.06 + 488.34= 977.40 mm, which is acceptable6 Determine the power rating per belt from the technicaldata provided by the manufacturer. This will vary fromone manufacturer to another, depending upon the qualityof rubber used. We have considered this data, as shownin Table 8.4, of a leading manufacturer.Corresponding to 1000 r.p.m. and d = 400 mm.Power rating of belt having section D = 23.88 kWlbelt,i.e. for 475-400 or 0.077977.40The correction factor is not less than 0.99.(ii) Determine the belt pitch length correction factor form(Table 8.6) which is same for all manufacturers.Corresponding to L = 3330 mm for the belt of sectionD, it is 0.87.Determine the number of belts for the total load to betransmitted.Corrected power capacity of each belt of sectionD = 25.62 x 0.99 x 0.87= 22.06 kW143:. Number of belts required = -22.06= 6.48or 7 beltsCounter-check the selection for the speed of the smallerpulley.Speed = a. d . N, . mmlminwhere N, = speed of the faster shaft = 980 r.p.m.d = pitch dia. of the faster (smaller) pulley= 400 mm.n x 400 x:.980Speed of faster pulley =60 x 1000= 20.51 mlswhich is less than 30 rnls, hence acceptable. Had it notbeen so, advice from the manufacturer on the selection ofbelts would have been necessary.Thus the specification of the belts and the pulleys will beBeltscross-section - Dpitch length L - 3330 mmNumber of belts 7Note To avoid uneven distribution of load on the beltswhen more than one belt is used the pitch length of thebelts must be identical, subject to permissible tolerances.For this purpose, it is advisable to use all belts of onemake only. As standard practice all belts are marked ontheir surfaces with their length and permissible tolerancefor easy identification.Pulleys(a) Smaller pulley on motor shaftPitch dia. = 400 mmNumber of grooves suitable for belts of section D = 7(b) Larger pulley on the compressor shaftPitch dia. = 475 mmNumber of grooves suitable for belts of section D = 7

8/21 0 Industrial Power Engineering and Applications HandbookTable 8.5Service Factors for flat and V-beltsTypes of loadsTypes of driving unitsLight startsHeavy starts direct on line'Soft' starters, star-delta, wound motorand A/T starters, or when the prime moveris fitted with centrifugal clutches ordelayed-action fluid couplingsClass Examples Hours of duty per day Hours of duty per dayIO and Over 10 Over 16 10 and Over 10 Over 16under to 16 under to 16Class 1(Light duty)Agitators (uniform density)Blowers, exhausts and fans 1 .o 1.1 1.2 1.1 1.2 1.3(up to 75 kW)Centrifugal compressors and pumpsBelt conveyors (uniformly loaded)Class 2(Medium duty)Agitators and mixers (variable density)Blowers, exhausts and fans (over 75 kW)Rotary compressors and pumps (otherthan centrifugal) 1.1 1.2 1.3Belt conveyors (not uniformly loaded)Generators and excitersLaundry machineryLine shaftsMachine toolsPrinting machinerySawmill and woodworking machineryScreens (rotary)Class 3(Heavy duty)Brick machineryBucket elevatorsCompressors and pumps (reciprocating) 1.2 1.3 1.4Conveyors (heavy duty)HoistsMills (hammer)PulverizersPunches, presses, shearsQuarry plantRubber machineryScreens (vibrating)Textile machinery)1.2 1.3 1.41.4 1.5 1.6Class 4(Extra heavy duty)Crushers 1.3 1.4 1.5 1.5 1.6 1.8For speed-increasing drives of speed ratio 1 .OO to 1.24: multiply service factor by 1 .OOspeed ratio 1.25 to 1.74: multiply service factor by 1.05speed ratio 1.75 to 2.49: multiply service factor by 1.1 1speed ratio 2.50 to 3.49: multiply service factor by 1.18speed ratio 3.50 and over: multiply service factor by 1.25Note1 The service factors do not apply to light duty drives.2 The use of a jockey (idler) pulley on the outside of the belt is not recommended.Special conditionsFor reversing drives, except where high torque is not present on starting, add 20% to these factors

Transmission of load and suitability of bearings 8/21 1Table 8.6Correction factors for belt pitch lengthCorrection factorBelt pitch lengthBelt cross-sectionsZ mm A mrn B rnm C mm D mrn EllUll0.80 6300.810.82 7000.830.84 7900.850.86 4058900.870.88 9900.890.90 47511000.910.925300.93 12500.940.956250.96 14300.970.98 70015500.99 16401 .oo 78017501.02 19401.031.041.05920205022001.06 23001.0710801.08 24801.09 25701.10 27001.111.12 29101.13 30801.14 32901.151.16 35401.171.181.191.201.219301560 27401000176011001210 1950313033301370 2190 373023401560 2490 40802720 46201760 280030801950 540033102180 35202300 61002500 4060684027002850 4600 762084103200 5380914036006100107004060686044304820 7600 122005000 1370053706070 152009100107004660504054206100685076509150995010710122301375015280168008.6 Checking the suitability ofbearings8.6.1 Forces acting on the DE motor bearingWhen transmitting the load to the driven equipment, themotor bearing at the driving end (DE) is normally subjectto two types of forces, radial and axial. The axial forcein a horizontal drive is normally zero. If it is not zerothis may be due to eccentricity in the transmitting line orany such reason that may subject the driving end bearingto an axial load in addition to a radial load. These forcesbecome severe when the load from the motor is beingtransmitted through belts. In such cases it is of the utmostimportance to check the suitability of the bearing andthe motor shaft to transmit the required load under suchconditions.We provide below a few guidelines for checking thesuitability of the bearing and the motor shaft, whenselecting a motor for these drives:Radial forcesThe radial forces acting on the bearing can be determinedby the belt pull and is expressed byP, =K. 973 . kW wkgD(8.3)N, .- 2

8/212 Industrial Power Engineering and Applications HandbookTable 8.7Arc of contact correction factors [flat and V-belts]Arc of contact on Correction factor, ;.e.__ D-d smaller pulley proportion oj 180"C(degrees)rating0.000.050.100.150.200.250.300.350.400.450.500.550.600.650.700.750.800.850.900.951 .OO1801771741711691661631601571541511481451421391361331301271231201 .OO0.990.990.980.970.970.960.950.940.930.930.920.910.900.890.880.870.860.850.830.8257.3( D - d)Note Arc of contact with smaller pulley = 180" -C1where C is the centre distance (mm)D is the pitch diameter of larger pulley (mm)Figured is the pitch diameter of smaller pulley (mm) Or0whereK =belt factor, 2 to 2.5 for V-belts and 2.5 to 3 forflat belts. Sometimes it may be higher (up to 4 to5). (The higher values must be considered whenthe distance between shafts is short or belt tensionis high).D = diameter of pulley on the motor (m)I =half of the pulley width (load distance) (m)W =weight of complete rotor and driven masses onthe motor shaft, such as the pulley and the belts(k&)'+' applicable when the belt pull is downwards[vertical drives, Figure 8.13(a)] and'-' applicable when the belt pull is upwards [verticaldrives, Figure 8.13(b)].'0' when the shaft is horizontal or the drive and thedriven equipment are in the same horizontal planeand in line with each other.The allowable pulley loads for standard motors aregiven in Table 8.10. These are recommended by a fewmanufacturers for their motors, but they are generallytrue for other motors also.The bending moment at the weakest section of theshaft, i.e. at the collar of the DE shaft (Figure 8.14).B.M. = P, .I =K. 973. kWDNr '7L.I mkgTable 8.8 Recommended standard pulley and outsidediametersOutside diameterBelt section A B C D EPitch mm mm mm mm mmdiamete? (mm)75 81.680 86.685 91.690 96.695 101.6100 106.6106 112.6112 118.6118 124.6125 131.6 133.4132 138.6 140.4140 146.6 148.4150 156.6 15X.4160 166.6 168.4I70 178.4I80 186.6 188.4190 198.4200 206.6 208.4 211.4212 220.4 223.4224 232.4 235.4236 244.4 247.4250 256.6 258.4 261.4265 276.4280 288.4 291.4315 321.6 323.4 326.4355375400425450475500530560630362.4 366.4 371.2391.2406.6 408.4 411.4 416.2441.2461.4 466.2491.2506.6 508.4 511.4 516.2546.2571.4 576.2636.6 638.4 641.4 646.2710 726.2800 806.6 808.4 811.4 816.2900 916.21000 1006.6 1008.4 1011.4 1016.21250 1258.4 1261.4 1266.21600 1616.22000 2016.2aThe limits of tolerances on the pitch diameter will be k 0.8%Courtesy Goodyear. See also IS 3142 and IS0 41830 Axial forces or thrust loadWhen these exist they can be expressed byPa (for horizontal motors) =axial force of pump,turbine or any otherload (kg)Pa (for vertical motors) = axial force as above +weight of complete rotorand driven masses on themotor shaft such as thepulley and the belts (kg.)

Table 8.9 Nominal inside lengths, pitch lengths L for all standard sizes of multiple V-beltsPitch lengthsTransmission of load and suitability of bearings 8/21 3Nominal cross- cross- cross- Cross- Crossinsidelength section section section section sectionA B C D E610 645 - - - -660 696 - -- -711 747 - -- -787823813848 -9148899259509329589659911016106710921001102610511102112810081034105911101136~~~1168 1204 1211 - --1219 1255 1262 1275 - -1295 1331 1339 1351 - -1372 1408 1415 1428 - -1397 1433 1440 1453 - -1422 1458 1466 1478 - -1473 1509 1516 1529 - -1524 1560 1567 1580 - -1800 1636 1643 1656 - -1626 1661 1669 1681 - -1651 1687 1694 1707 - -1727 1763 1770 1783 --1778 1814 1821 1834 - -1829 1865 1872 1885 - -1905 1941 1948 1961 --1981 2017 2024 2037 - -2032 2068 2075 2088 - -2057 2093 2101 2113 - -2159 2195 2202 2215 - -2286 2322 2329 2342 - -2438 2474 2482 2494 - -2464 2500 2507 2520 - -2540- 2583 -26672703 2710 27232845 2880 2888 2901 - -3048 3084 3091 3104 3127 -3150 - -3205 - -3251 3285 3294 3307 w 3404 - -3459 - -3658 3693 3701 3713 3736 -4013 - 4056 4069 4092 -4115 - 4158 4171 41944394 - 4437 4450 4473--4572 -4615 4628 4651 -4953 4996 5009 5032 -5334 -5377 5390 5413 54266045 - -6101 6124 61376807 - -6863 6886 68997569 - -7625 7648 76618331 - -8387 8410 84239093 - -9149 9172 918512141136651518916713Note All dimensions are in millimetres.Courtesy Goodyear1222013744152681679212233137571528116805

8/21 4 Industrial Power Engineering and Applications HandbookForceForce(a) W-adds(b) W-subtractsFigure 8.14Forces acting on the motor shaftFigure 8.13 Radial forces on the motor bearingsTable 8.10 Allowable pulley loads for standard motorsFrame size Poles Load Maximum allowable loaddistance 1Radial P, Axial Pa(mm) (k&) (k&)80 2 444 25 48 206 488 4890 2 424 31.5 50 256 518 67100 2 624 40 67 356 878 96112 2 904 50 99 406 998 99132 2 1264 62.5 154 506 1 xn160 2 2004 70 226 706 2668 286180 2 2144 80 262 806 3088 342200 2 2904 100 345 1006 4258 465225 2 2754 100 325 1006 4008 450Note The values given above refer to a bearing’s life of 20 000working hours. They are valid only for radial or axial loads.When both the loads are existing reference must be made to thebearing or motor manufacturer.Note Cylindrical roller bearings should normally not be usedfor such applications.When a bearing is subject to both radial and axialthrusts the equivalent dynamic bearing loading canbe expressed byP=X. F,+ Y. FawhereX = radial load factorY = axial load factorThe values of these factors are provided by the bearingmanufacturers in their catalogues, based on the ratiosF,IC, and F, IFr, where C, is the static load ratingin kg or N (based on a contact stress of 4200 MP,for ball bearings and 4000 MP, for roller bearings,also provided in these catalogues. MP, is the unit ofstress in Mega-Pascals.Note To assess the total force a bearing is likely to encountermore precisely, reference may be made to the bearingmanufacturers’ product catalogues. In severe load conditions,with an excessive bending moment on the shaft, the shaftstiffness and its suitability should also be checked. As shownin Table 8.10, motors for normal industrial use employ bearingswith almost 20 000 hours as safe running life. For more stringentduties or continuous drives as in the pulp and paper industry,a cement plant, refinery and petrochemical projects, thechemical industry and powerhouses which are 24-hour services,special bearings must be used with a safe running life of50 000 to 100000 hours.8.6.2 Load-carrying capacity and life expectancyof bearingsThe life of bearing would depend upon various factorssuch as:Type of driveMethod of load transmissionAccuracy of alignmentEnvironment of installationThe amount of radial and axial forces acting on thebearings orAny other factors discussed earlier.

Transmission of load and suitability of bearings 8/215No rotating mass can be balanced up to 0 micron, forobvious reasons. While rotating they are out of balance,which gives rise to rotating forces and adds to the radialload on the bearing. This indirectly affects the runninglife of the bearing, in addition to other factors notedabove. The selection of type and size of bearing, is thusgoverned by the speed of the machine, the type and weightof loads and the required life of the bearing which can beexpressed by106Ll0 = -(g)' hours (8.4)60. N,whereL,, = Rated life of bearings at 90% reliability, i.e. 90%of bearings produced by a manufacturer will exceedthis lifeC = basic dynamic load rating in kg or N (provided bythe bearing manufacturer). It is the load whichwill give a life of I million revolutionsP = equivalent dynamic bearing load in kg or Np = exponent of the life equation, which depends uponthe type of contact between the races and the rollingelements. It is recommended as 3 for ball bearingsand 10/3 for roller bearingsN, = speed of the machine in r.p.m.With this equation the life of the bearing can be determinedfor different load conditions and is predetermined forthe type of drive and service requirements. To select aproper bearing, therefore, the type of application and theloading ratio (ClP) should be carefully selected to ensurethe required minimum life. Bearing manufacturers' productcatalogues provide the working life of bearings fordifferent load factors and may be referred to for data onC, C, and other parameters.8.7 Suitability of rotors for pulleydrivesIn belt drives particularly, it is advisable that referencebe made to the motor manufacturer to determine thesuitability of the rotor shaft and the driving end (DE)bearing for transmission of the required load. A typicalformat of a questionnaire is also given below for providingthe manufacturer for load, belt and pulley data. The formatis suitable for all drives that may be subjected to excessiveforces on the shaft. These data will enable the manufacturerto determine the following parameters to check thesuitability of bearing and shaft strength and make suitablechanges if warranted:Load acting on the motor bearingsBending moment at the motor shaft due to pulley andloadPossible deflection of the shaftNote The shaft deflection should not be more than 1 1 % of the airgap between the stator and the rotor. For loads that exert more forceand torsional stress on the motor shaft and bearings than ispermissible, due to the larger width of pulleys which may shift theload farther away from the shaft collar, it is recommended that thepulley be mounted on a separate jack shaft, supported on two pedestalsas illustrated in Figure 8.15. Provided that the motor shaft can bemade longer, to support such a pulley and have sufficient strength,to take that load, one pedestal may also be adequate to support thefree end of the motor shaft as shown in Figure 8.16. The pulley isnow mounted between the shaft collar and the pedestal. A jackshaft or additional pedestal to support the motor shaft may also benecessary when the ratio of pulley diameters exceeds 6:l.In some cases, reinforcement of the shaft by increasing theshaft diameter, employing a better grade of steel and using a superiorgrade of bearings, may also meet the load requirement. The bearingbore, however, may pose a limitation in increasing the shaft diameterbeyond a certain point, say, beyond the diameter of the shaft collar.If a larger shaft diameter is required either a larger frame size of themotor may be employed, which may be uneconomical, or a jackshaft or pedestal may be used as noted above. Replacing standardbearings with larger bore bearings to use a shaft of greater diametermay not be possible in the same frame, due to pre-sized end shieldsand bearing housings which, for motors up to 250 kW, are normallycast and have fixed dimensionshoulding patterns.Questionnaire to determine the suitability of themotor shaft and the bearing for the requiredbelt driveFor critical loads and belt drives particularly, the user isadvised to seek the opinion of the motor manufacturer todetermine the mechanical suitability of the motor selected,its shaft and the DE bearing for the load to be transmitted,according to the drive system being adopted. The followingare some important parameters that may help to determinePedestal bearing-, -Figure 8.15 Arrangement of a jack shaftI Abearingseme7MotorShaft collarPulleyILong shaft7 PedestalbearingFigure 8.16 Arrangement of a long shaftBaseframe

8/21 6 Industrial Power Engineering and Applications Handbookthe required .suitability and should be sent to themanufacturer for their opinion:1 (a) 'Qpe of driven equipment(b) 'Ifipe of bearings provided in the driven equipment(c) Whether any thrust or radial load is falling onthe motor bearings from the driven equipment.2 Details of the belt drive (Figures 8.17 and 8.18)(a)of belt: flat or V-belts and their width ornumbers(b) Diameter of pulley on motor shaft D1 (mm)(c) Width of pulley on motor shaft W1 (mm)(d) Diameter of pulley on load side D2 (mm)(e) Width of pulley on load side W2 (mm)(f) Does the centre of the hub coiacide with thecentre of the rim of the pulley? If not, what is theeccentricity, E, in mm?(g) Distance between centres of pulleys, C (mm)(h) Weight of pulley on motor shaft (kg)(i) Magnitude of pulley force P (kgf or N)(j) Inclination pof the pulley system to the horizontalplane as shown in Figure 8.18.Driven equipmentIIDriven equipmentCentre of hub \ E /1Centreof rimMotorDriven equipmentFigure 8.17 Mounting of motor pulleyFigure 8.18 Disposition of driving and driven-end pulleysRelevant StandardsIEC Title IS BS IS0- FLAT BELT 137011993 351 22-Transmission belting friction and surface rubber belting- specificationsV BELTS 3142l1993 379011995 4183Pulleys - V grooved for endless V-belts sections Z, A, B, C, Dand E and endless wedge beltSections SPZ, SPA, SPB and SPC - Specifications- Endless V-belts for industrial purposes 2494-111994 3790/1995 4184- Industrial V-belt drives - calculation of power ratings 5292/1980Pulley and V ribbed belts for industrial applications-PH, PJ, PK, PL and PM profiles9982/1998Notes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

Transmission of load and suitability of bearings 8/217List of formulae usedSelection of flat beltsw=p.- SFCorrection for arc of contact(8.1 )P = the load to be transmitted (kW)W = maxiinum load transmitting capacity of the beltSF = service factorTo determine the pitch length of V-beltsRadial forces on bearingsL(8.3)K = belt factorD = diameter of pulley (m)W = Weight of complete rotor and driven masses on themotor shaft.Load-carrying capacity and life of bearingsL = belt pitch length (mni)C = centre distance between the two pulleys (mm)I) = pitch diameter ot the larger pulley (mm)il = pitch diameter of the smaller (fa5ter) pulley (mm)Lh = normal life of bearing in working hoursC = basic dynamic load rating (kg or N)P = equivalent dynamic bearing load (kg or N)p = exponent of the life equationN, = speed of the machine (r.p.m.)

9/21 9Winding Insulation -and itsMaintenance

Winding insulation and its maintenance 9/2219.1 Insulating materials and theirpropertiesThe common types of insulating materials in use forelectric motors are E and B for small motors and F formedium sized and large ones. General industrial practice,however, is to limit the temperature to class B limits,even if class F insulation is used.Insulation class A, previously in use, has beendiscontinued in view of its low working temperature.Motor frames have also been standardized with class Einsulation only by IEC recommendations, as in Table1.2, to harmonize the interchangeability of electric motors.This decision was taken because class E insulation offersa higher working temperature and a longer working life.These frames also ensure optimum utilization of activematerials such as copper and steel in a particular framesize. The classification of insulating materials is basedon their maximum continuous working temperature,established for 20 years of working life. The recommendedtemperature according to IEC 60034- 1 is, however, lessthan this, as shown in Table 9.1, to ensure an even longerlife. Where the ambient temperature is likely to be high,of the order of 60°C or so during operation, such asclose to a furnace, class F and H insulations are normallyused, as they have higher working temperatures andthermal resistivity.A brief description of the insulating materials in usefor different classes of insulation is given below to providean introduction to the types of materials being used inthe preparation of a particular class of insulation. Theactual ingredients may be an improvised version of thesematerials, in view of continuous research and developmentin this field, to search for still better and more suitablematerials.9.1.1 Insulation class AThis includes organic fibrous materials on a cellulosebase such as paper, pressboard, cotton, cotton cloth andnatural silk etc., impregnated with lacquers or immersedin an insulating liquid. The impregnation or immersionensures that the oxygen content of the air does not affectTable 9.1 Maxirnumlperrnissible working temperatures fordifferent insulating materialsClass of Maximum Permissible operating temperatureinsulation attainable as in IEC 60034-1 by the resistancetemperature as method"in IEC 60085Up to 5000 kW Above 5000 kW"C "C "CAb 105 100 100Eb 120 115 110B 130 120 120F 155 145 140H 180 165 165a Using the thermometer method this temperature will be less by10°C.These insulations are generally not used, for large motors, due totheir low operating temperature.the insulating properties or enhance the thermal ageingof the insulating material. Typical materials in this classare varnished cloth and oil-impregnated paper.9.1.2 Insulation class EThis includes wire enamels on a base of polyvinyl formal,polyurethane or epoxy resins as well as moulding powderplastics on phenol-formaldehyde and similar binders, withcellulose fillers, laminated plastics on paper and cottoncloth base, triacetate cellulose films, films and fibres ofpolyethylene terephthalate.9.1.3 Insulation class BThis includes inorganic materials such as mica, glassfibre and asbestos etc., impregnated or glued togetherwith varnishes or compositions comprising ordinaryorganic substances for heat resistance such as oil-modifiedsynthetic resins, bitumen, shellac and Bakelite.9.1.4 Insulation class FThis includes inorganic materials such as glass fibre andmica impregnated or glued together with epoxy,polyesterimide, polyurethane or other resins havingsuperior thermal stability.9.1.5 Insulation class HThis comprises composite materials on mica, glass fibreand asbestos bases, impregnated or glued together withsilicone resins or silicone elastomer. These materials mustnot contain any organic fibrous materials such as paperor cloth backing, which is covered under class B andeven F insulation systems.9.2 Ageing of insulationWith time, the insulation becomes brittle and shrinks,leading to cracks. The insulation at the point of cracksweakens gradually as surrounding pollutants find theirway through these cracks. The weakening of insulationwith time is called 'ageing'.The life of the insulation will also be affected by anexcessive operating temperature. It is halved for every1 1 "C rise in temperature over its rated value and occurswhen a machine is occasionally overloaded. Sometimesthe size of the machine may be only marginal when itwas initially chosen and with the passage of time, it maybe required to perform duties that are too arduous. Everytime the machine overheats, the insulation deteriorates,and this is called thermal ageing of insulation. Figure9.1 illustrates an approximate reduction in life expectancywith a rise in operating temperature.9.3 Practices of insulation systemsInsulation of steel laminations(i) For smaller ratings: by steam bluing the steelsurfaces on both sides.

9/222 Industrial Power Engineering and Applications Handbook-+5.5 +11.0 +I65 +22.0Operating temperature abovethe rated (B,,,) in "CEnd winding (over-hang)Figure 9.1 Reduction in life expectancy of a motor with a risein operating temperature(ii) For larger ratings: by phenol or synthetic resininsulation on one or both sides. Although insulationon one side is common, by ensuring that thelaminations are always punched and stacked inone direction only to avoid contact between thepunching burrs, insulation on both sides is alwayspreferred.Inisulation of the windingsDifferent manufacturers adopt different practices ofinsulating the coil or the windings. Practices generallyused may be one of the following.9.3.1 LT motorsWound statorBy simple impregnation in a recommended insulatingvarnish, normally synthetic or epoxy, followed bybaking (curing), in a temperature-controlled oven, at aspecified temperature for a specific period.For powerhouse insulation treatment the stator maybe dipped in varnish for a minimum two to three times,each dipping being followed by backing. Sometimesone immersion of the entire stator and two additionalimmersions of the overhangs followed by backing maybe sufficient.Formed wound machinesFor large motors, the practice is to wind the stator withformed coils (Figure 9.2). The coils are pre-formed andcured before insertion into the stator slots. They areinsulated with resin-rich glass and mica paper tapes. Theprocess of impregnation is therefore termed 'resin-rich'insulation. The completed formed wound stator is thenheated to remove trapped moisture and finally impregnatedin varnish class F or H as required, under vacuum andpressure. The stator is then cured in an oven as describedabove. The process of insulation and curing conforms topowerhouse insulation requirements. This practice facili-Figure 9.2View of a formed coiltates easy removal of an individual coil at site in case ofa damage and replacement with a spare coil. The usercan stock spare coils for such eventualities.9.3.2 HT motorsTo wind HT motors two methods are adopted:1 Formed wound machinesThis is a resin-rich system. The stator is wound withpre-formed coils on similar lines as noted above. Afterthorough testing on each coil for the polarization indexand dissipation factor (tan 4, as discussed in Sections9.5 and 9.6, and the impulse voltage withstand test,as discussed in Section 11.4.9, the individual coilsare completely cured and toughened before insertingthem into the slots. The rest of the process is as notedearlier.2 Vacuum pressure impregnation (VPI)For HT motors, the latest practice is to have the statorvacuum pressure impregnated (VPI) in insulating resinsas a standard procedure, not only to meet therequirements of 'powerhouse insulation' but to alsodevelop a more simplified insulating process, to cureand toughen the stator windings and to meet theseverities of all operating conditions a motor mayhave to encounter. As described later, this is termed aresin-poor insulating process because the insulatingtapes now have a low resin content as they are later tobe impregnated in resin. Performance and field dataof this insulating system have revealed excellent results,

Winding insulation and its maintenance 9/223surpassing those of the normal impregnating processand even the process of resin-rich formed coils. As aneconomy measure, the general practice of leadingmanufacturers is to adopt a resin-rich formed coilsystem for frame sizes smaller than 710 and resinpoorVPI for frame sizes 710 and higher. But it isalways recommended to adopt a resin-poor systemfor all HT motor windings, irrespective of frame sizes.In a pressure-vacuum impregnation system, sincethe whole stator iron bulk and the stator windingsform a solid mass, removal of one coil and itsreplacement is impossible, unlike in the previous case.But in view of the excellent properties of a postvacuumimpreg-nated insulating system, the chancesof any part of the stator winding developing anoperational defect are remote. In all probability nosuch localized damage would arise over the life spanof the motor. The windings may fail on account of afailure of the protective system to clear a fault orisolate the machine on a fault, but if the motor fails,the whole stator is scrap and a totally new stator hasto be requisitioned. The rating of such motors is slightlyless due to the reduced cooling effect.Bracing of the coil ends (overhangs)The coil ends must be rigidly supported and adequatelybraced with binding rings or tapes to prevent theirmovement and also absorb shocks and vibrations duringexcessive overloads, starting inrush currents (Ist), andvoltage surges.9.4 Procedure for vacuum pressureimpregnation (with particularreference to HT motors)In a formed coil system each individual coil is pre-formed,insulated and cured, and is made rigid before it is insertedinto the slots. The dielectric qualities of the coil insulationare monitored closely during the process of coil formationto ensure the required quality. For procedure andacceptance norms see Section 11.4.9.In the post-impregnated system, although the coils areformed as above, they are inserted into the slots whenthey are still in a flexible state. They are now easy tohandle and cause no damage to their own insulation orthe insulation of the slots while being inserted into theslots. The process up to the winding stage of the stator isthus faster and economical. The stator is then vacuumdried to remove trapped moisture, followed by immersionin a resin bath. It is kept immersed under vacuum so thatresin can fill the voids. The bath is then pressurized tocompress the resin so that it penetrates deeply into theslots, crevices and voids. Figure 9.3 illustrates the loweringof a stator’s pre-formed windings into a resin impregnatingtank. The stator is then cured in an oven under controlledconditions. The overall vacuum impregnating system maybe expensive in view of the equipment required to dipthe bulk of the stator into the impregnating resin to createa very high vacuum, but the excellent properties of postvacuumimpregnation may compensate this initial cost.Figure 9.3(a) A stator core during vacuum pressure impregnation (resin-poor insulation) (Courtesy: NGEF Ltd)

9/224 Industrial Power Engineering and Applications Handbookis to achieve a better life of the repaired motor, on the one hand,and facilitate an easy rewinding of the machine, on the other, asclass F is a thinner insulation which is easy to rewind at site.9.5 Maintenance of insulationFigure 9.3(b) Stator core after vacuum pressureimpregnation (VPI)9.5.1 Against insulation failureThe properties of an insulating material are greatly affectedby moisture, temperature, repeat overvoltages, andchemical vapour. Care must be taken to avoid theseharmful effects to achieve prolonged life of the machine.Whenever a motor is installed in a humid atmosphereand is switched on after a long shutdown, insulationresistance must be checked before energizing the motor.As a precaution, insulation resistance must be checkedbefore a restart after a long shutdown, even in temperateconditions. If the insulation level is found to be belowthe recommended level as shown in equation (9.1) itmust be made up as noted below.9.5.2 Making-up the insulation levelThis can be done by homogeneously warming and dryingthe windings in the following way:Figure 9.3(c)Exploded view of a VPI stator coreNote For HT motor windings, ordinary dipping and baking, asfor LT motors, is not recommended in view of the very high stressesto which the windings may be subject during switching or whileclearing a fault, or in the event of a system disturbance. See alsoSection 17.5. An ordinary cured motor winding may not be able towithstand these conditions.Repairing techniqueGenerally, no electrical fault would cause damage to thewindings except in the event of failure of the protectivesystem. Mechanical surface damage is, however, possibleduring transportation or installation or due to penetrationof foreign matter into the windings. It is possible torepair such damage at site without affecting the performanceor quality of the winding insulation. Theexcellent electrical, mechanical and thermal propertiesof the end turn windings localize such damage. Repair ispossible by slicing and cleaning the damaged part, andapplying to it the pre-impregnated insulating tapes andother resins. It is recommended to request the manufacturerto carry out such repairs.Note For motors that are insulated other than by vacuumimpregnation, it is recommended to rewind a damaged stator withclass F insulation, even if the original insulation was E or B. ThisA crude method can be to set up two lamps at bothends of the stator housing after removing the endshields. This can be done in smaller motors, say, upto ‘355’ frame sizes.By circulating roughly 50-60% of the rated currentunder a locked rotor condition, at a reduced voltage.To facilitate easy escape of the moisture to theatmosphere, the bearing covers may be removed.At important installations, built-in heaters are recommendedwhich can be switched ON when the motoris not in use. The rating of these heaters, known asspace heaters, is such that a continuous ‘ON’ heaterwill not have any harmful effects on the motorwindings, nor will they increase the winding temperaturebeyond the safe limits. The rule of thumb is toset the space heater rating so that it is able to raise theinside temperature of the motor by 8-10°C higherthan the ambient. Its main function is to keep themoisture condensation away from the windings bymaintaining a higher temperature inside the motorhousing.In certain cases, where the motor is too large and it isidle for a long period before it is installed and electricallyconnected (the space heaters are therefore OFF) eventhese heaters may not be sufficient to absorb moisturewhich might have condensed deep in the slots, unless, ofcourse, the heaters are kept ON for a considerably longperiod. In such cases, it is advisable to heat the motor bymethods 1 and 2 in addition to using the built-in heaters,until the insulation level of the windings reaches therequired level. Once the motor has been installed, thesespace heaters, when provided in the windings, are switchedON automatically as soon as the motor is idle, and thuseliminate deep moisture condensation.

Winding insulation and its maintenance 9/2259.5.3 Making-up of the insulation level in largemachines (1000 kW and above)IEEE 43 places special emphasis on determining theinsulation condition of such machines before energizingand even before conducting a high-voltage test. This canbe determined by the insulation test as noted below.Insulation resistance testInsulation resistance of the windings is a measure toassess the condition of insulation and its suitability forconducting a high-voltage test or for energizing themachine. A low reading may suggest damage to theinsulation, faulty drying or impregnation or absorptionof moisture. The insulation resistance may be measuredaccording to the procedure laid down in IEEE 43 betweenthe open windings and between windings and the frameby employing a direct-reading ohm meter (meggar). Theohm meter may be hand-operated or power-driven orequipped with its own d.c. source. 500 V d.c. is therecommended test voltage for an insulation resistancetest for all voltage ratings, as noted later. The recommendedminimum insulation resistance of the machine is obtainedby the following empirical formula:R, = kV+ 1 (9.1)whereR, = recommended minimum insulation resistance inMR (meg ohms) of the entire machine windings,at 40°C or 1 MR per 1000 V plus 1 MQ, andkV = rated machine voltage in kVThe winding insulation resistance to be used for comparisonwith the recommended minimum value (R,) isthe observed insulation resistance obtained by applyinga d.c. voltage to the entire windings for one minute,corrected to 40°C. In practice, motors having insulationresistance readings as high as ten to a hundred times theminimum recommended value R, are not uncommon.At site, when commissioning a new or an existingmotor after a long shutdown, it must have a minimuminsulation level according to equation (9.1). An 11 kVmotor, for instance, must have a minimum insulation of12 MQ. In normal practice, it is observed that when firstmeasured the resistance reading may show more thanthe minimum value and may mislead the operator, whilethe winding condition may not be adequate for a highvoltagetest or an actual operation. One must thereforeensure that the winding condition is suitable before themachine is put into operation. For this purpose, thepolarization index (PI), which is determined from theinsulation test data only as noted below, is a useful pointer.It must be evaluated at site while conducting the insulationtest then compared with the manufacturer’s referencedata for the machine to assess the condition of insulationat site and its suitability for operation. This is usually asite test, but to establish a reference record of the machine,it is also carried out at the works on the completed machineand test records furnished to the user.Polarization indexThis is the ratio of 10 minute to 1 minute insulationresistance readings. It will not be less than 1.5 for classA, 1.75 for class E and 2 for class B insulations for thewindings to be regarded as suitable for a high-voltagetest or actual operation. This index is based on thephenomenon that the insulation resistance of a windingrises with the duration of application of the test voltage.The rise is rapid when the voltage is first applied andthen it almost stabilizes with time, as noted in Figure9.4. For instance, after 30 seconds of the applied voltagethis resistance may be too small, of the order of only afraction of the final value. As a rough guide, the insulationresistance of a dry and normal winding may also continueto rise in the first few minutes after application of thetest voltage before it stabilizes. Normally it may take10-15 minutes to reach a steady state. If the windingsare wet or dirty the steady condition may be achievedwithin one or two minutes of application of test voltage.These test results are then compared with similar testdata obtained from the manufacturer on similar windingscarried out during manufacture. If the manufacturer’soriginal test results are available, the results obtained atsite can be quickly compared and the condition of theinsulation assessed easily and accurately. If the test facilityto obtain test results at 1 minute and 10 minutes is notavailable, the results may also be obtained for 15 secondsand 60 seconds and a graph plotted as shown in Figure9.5(a) to determine the polarization index.From these insulation tests several test data for 1 minuteand 10 minutes can be obtained during the process ofheating over a few hours and a graph plotted for timeversus insulation resistance. This graph is of great assistancein determining the state of the windings before energizing.Heating or cleaning-up of the windings should continueuntil the polarization index matches the reference value10008006001 4003$200024.$ 100F! 80C.s 609? 40I20-100.1 0.2 0.4 0.7 1.0 2.0 4.0 7.0 10.0Time (minutes)Figure 9.4 Variation of insulation resistance with time forclass B insulation

9/226 Industrial Power Engineering and Applications Handbook100I I I I II I I I I I0 20 40 60 80 100 120Time (hours)-Initial winding temperature 25°CFinal winding temperature - 75°CFigure 9.5(a) Typical values of 1- and 10-minute insulationresistances during the drying process of a class B insulatedwinding of a large machineT 501000.10.05--10 0 10 30 50 70 90 110Winding temperature "CFigure 9.5(b) Approximate temperature coefficient ofinsulation resistance of rotating machinesor reaches a minimum of 1.3 for all classes of windinginsulation. Only then can the windings be consideredsuitable for a high voltage-test or actual operation.Important aspects while measuring R, at site1 It will generally be seen that the insulation resistancewill first fall to a very low level before rising, thereason being that when heating begins, moisture inthe windings is redistributed in the stator windings. Itmay even reach the dry parts of the machine andindicate a very low value. After some time the insulationresistance will reach its minimum and stabilizeat this level, before it begins to rise again and reachits maximum. See also the curves of Figure 9.5(a).2 The insulation resistance thus measured at differentintervals during the process of heating up will representthe insulation resistance of the windings at thattemperature. Before plotting the curves, these testvalues must be corrected to one reference temperature,say, 40"C, to maintain coherence in the test results.3 To correct the value of insulation resistance at 40°Cthe following equation may be used:R40 = Kt40 ' Rt (9.2)whereR4, = insulation resistance in MR corrected to 40°C.R, = test insulation resistance in MR at t "C andKt40 = temperature coefficient of insulation resistanceat t "C.4 The temperature coefficient curve is given in Figure9.5(b). This is plotted on the assumption that theinsulation resistance doubles for each 18°C reductionin temperature (above the dew point).5 The test voltage should be d.c. and restricted to therated voltage of the windings, subject to a maximumof 5000 V d.c. for a 6.6 or 11 kV motor. Applicationof a higher voltage, particularly to a winding which isweak or wet, may cause a rupture of the insulation. Itis therefore recommended to carry out this test onlywith a 500V d.c. meggar. Accordingly, it is the practiceto conduct this test at the works also on a normalmachine, at 500 V d.c. only to provide consistency ofreference.6 The insulation resistance of the windings, R,, can becalculated byTest voltage in voltsR, = MR(9.3)Leakage current in ,uADepending upon the condition of the winding insulation,an increase in the test voltage may significantlyraise the leakage current and decrease the insulationresistance, R,. However, for machines in very goodcondition, the same insulation resistance will beobtained for any test voltage up to the rated voltage.The insulation resistance may hence be obtained witha low voltage initially and if the condition of thewindings permits, may be raised, but always less thanthe rated and in no case more than 5 kV, irrespectiveof the rated voltage.7 Before conducting the insulation test one should ensurethat there is no self-induced e.m.f. in the windings. Itis recommended that the windings be totally dischargedby grounding them through the frame of the machinebefore conducting the test.

Winding insulation and its maintenance 912278 Winding connections for insulation resistance testIt is recommended to test each phase to groundseparately with the other phases also grounded. Thisis because the insulation resistance of a completewinding to ground does not provide a check of theinsulation condition between the windings.It is observed that the insulation resistance of onephase with the other two phases grounded isapproximately twice that of the entire winding.Therefore, when the three phases are tested separately.the observed value of the resistance of eachphase is divided by two to obtain the actual insulationresistance.9.69.6.1Monitoring the quality ofinsulation of HT formed coilsduring manufacturingTheory of dielectric loss factor ordissipation factor (tan 6)Irrespective of the class of an insulation system and itsquality, it will have some leakage current through itsdielectric circuit on application of a high voltage. For allpractical purposes. therefore. we can consider an insulationsystem as an imperfect capacitor.During the process of insulation. impregnation of anindividual coil (resin-rich)* or the whole winding (resinpoor)..'somevoids in the insulation coating will alwaysexist. They cannot be elirninated, however good the processof insulation coating or impregnation. While the idealrequirement will be a completely void-free insulationcoatinghmpregnation, in practice this cannot be achieved.These voids caube internal discharges (corona effect)?,.Ii Res ir / - richThi\ i\ an insulating process for winding of HT motors with formedcoils. and employs clasi F insulating materials. The winding coilsare huilt individually. outside the \lots and are pre-impregnated andcured before they are inserted into the stator or rotor slots. Thisterm of coil making. however. possesses a poor heat transferability. troni coil to iron core :ind provides a poor bonding ofthe coil with the \lot. This may cause differential movement imidethe .;lot during a run. due to thermal effects. and a peeling of theinsulation, in addition to vibration and noise.Resin-poorThis is when the stator or the wound rotor is first wound with resinpoorformed coil\. and then vacuum impregnated. a\ a whole ma\\and cured.: C'or.ollcl qf/wThib I\ ;I discharfe that occurs due to ionlzation of the air in thein,mctlinte vicinity of21 conductor. It normally takes place on roundconductors or at curvatures. rough spots. protruding nuts and bolts.and occurs due to humidity at locations where there is a largeelectrwtatic flux. such as between two parallel running and currentcaqingconductors. The highcr thc potcntial hetween the conductor\.the more \e\erc this phenomenon will be. It is a purple glow aroundthe conductor. and is normally associated with a hissing noise andc;iuses lossc.;. We can sometime.; see this phenomenon on an HTpower pole and more so on a humid day. To minimize this effect.it is e\sentinl to keep the conductor surface clean, avoid sharpbends and curbatuw and aljo .jagged nut and bolt heads. Whereunavoidable. such surfkc.; may be co\ered (insulated) with a nonconductingtape or tube.or leakage current within the windings and leads to erosionin soft materials, or microscopic cracks in hard ones. Allthis may also lead to an eventual failure of thc insulation.The level of such dischargesAeakage currents shouldtherefore be restricted as much as possible, so that theycause the least harm to the insulation of the windings,during the machine's long years of opcration. For anillustration of the leakage current circuit. see Figures 9.6and 9.7.By measuring the capacitive and leakage currents, thephase angle (loss angle) 6 bctween them can bedetermined. The tangent of this angle. tan 6, will give anindication of the condition of insulation:In a capacitor. 6 = 90 - @ andCapacitor lossesran 6= --_i,Reactive output of the capacitor i,(Figure 9.7)A low tan 6 will mean a high degree of resin cure. Mostinsulation systems are composites of many materials. Inpractice, they almost always contain small voids. Considera coil side with a single void. The voltage distributionacross the insulation will be non-uniform. due to differentpermittivities of air and insulation. When a low voltageis applied, a proportion of this will appear across theITbc = Leakage lumped capacitancer = Leakage lumped resistanceI, = Leakage currentVg = Voltage to ground (phase voltage)Figure 9.6 Representation of leakage current in an HTinsulation systemFigure 9.7tan 6= L= LIC I,Phasor representation of the leakage current

9/228 Industrial Power Engineering and Applications Handbookvoid and the remaining voltage across the insulation(dielectric). When the applied voltage is increased, theair of the void at a certain value of the applied voltagewill break down, causing an internal discharge. The voidis now short-circuited and the full voltage will appearacross the insulation (dielectric). The tan 6 - voltagecurve at this point, where ionization begins, shows arapid change (Figure 9.8). In practice, an insulation system,whether of a coil, winding or a slot, will always containmany small voids, often located at different depths ofthe dielectric. The higher the number of voids, the steeperwill be the tan 6 V curve beyond a certain voltage level(Figure 9.8). The value of tan 6 at low voltage and therate of change of tan 6, i.e A tan 6, with an increase inthe applied test voltage, gives an indication of the conditionof the dielectric at higher test voltages, and also suggeststhe presence of moisture. Hence, A tan 6 is a measure ofvoids in the insulation system and indicates the qualityof curing. It is also a good method of monitoring thequality of insulation of HT formed coils during the courseof manufacture. This is also useful in analysing the ageingcondition of an insulation. The methods and norms ofacceptance limits are dealt with in IEC 60894. Foracceptance norms see Table 1 1.5.Voltage (V,) -Figure 9.8 Variation in tan 6 with the applied voltageFigure 9.9 A NEMA spray test being carried out on a statorwinding (Courtesy: BHEL)9.6.2 Wet test of resin-poor windingsMG- 1-20.48 recommends this test for large HT motorsthat are resin-poor. In this test the whole stator issubmerged in water, if possible, and a 10-minuteabsorption test is carried out. If this is not possible, thenthe windings are sprayed with water thoroughly, from allsides for 30 minutes (Figure 9.9). The water is mixedwith a wetting agent to reduce its surface tension. Duringthe last 10 minutes of the test the insulation resistance ismeasured at 500 V d.c, which should not be less than asindicated in equation (9.1). If it is acceptable, the windingsare then subjected to an a.c. high-voltage test at 1.15times the rated voltage for 1 minute while the windingsare still being sprayed. After the high-voltage test, the 1minute insulation resistance reading using a 500 V d.c.source is obtained. This should not be less than as specifiedin equation (9.1). This value will then form the referencedata for the site tests.Relevant StandardsIECTitleISBS60034- l/1996 Rotating electrical machinesRating and performance for rotating machines60071-1/1993 Insulation coordination. Definitions, principles and rules60085/ 198460894/1987Thermal evaluation and classification of electrical insulationSpecification for the insulation of bars and coils of highvoltage machines, including test methods472211992. 32511996 BS EN 60034-1/19952165 (Part-I) 1991 and BS EN 60071-1/19962165 (Part-11) 1991I27 1/1990 BS 2757/1994- BS 4999-144/1987Relevant US Standards ANSI/NEMA and IEEEANSUIEEE13 13.1/1996 Insulation Coordination. Definitions, Principles and RulesANSI/IEEE-I/1992 General principles for temperature limits in the rating of electric equipment and for the evaluation of electricalinstallationANSIlIEE 101/1987 Guide for the statistical analysis of Thermal Life Test dataNEMA/MG.1/1993 Motor and generators ratings, construction, testing and performance

Winding insulation and its maintenance 9/229:vote,\I In the tables of relevant Standards in this hook while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydilferent standards organizations. It is therefore advisable that for more authentic I-efei-cnces. readers \trould consult the relevantcrsanizations for the latest version of a standard.23Sonic of the BS or IS htandards mentioned against IEC may not be identical.The year noted against each standard may also refer to the year of its last amendment and not necessarily the )ear of publication.List of formulae usedMaking-up of insulation level in large machines:Insulation resistanceR,,, = kV + I (9.1 )R,,, = recommended minimum insulation resistance inMR of the entire machine windings, at 40°CkV = rated machine voltage in kV.RJo = insulation resistance in MQ corrected to 40°CR, = test insulation resistance in MQ at t "C andK,,,, = temperature coefficient of insulation resistance att "CTest voltage in voltsR1'l = Leakage current in PAMR(9.3)Further readingI. Anddo Cornpuneriti SPA. Italy. Pvo,qre.\.\ arid rln,c.lopnzenrr w d s it1 the large induction motor s tiitor vt.indirig inxrt/cirion.2. Hogg, W.K.. Miller, R., Rahach, G. and Ryder D.M. 'Therelationshipofpartial discharge amplitude distri hution with electricdamage at different levels of voltage and frequency'. IEEESjmposiurn oil Electrical Insularion. USA. June 1984.3. Lee-Neville., and Simons J.S., Handbook of Epo.k-) Rr.\tn\.McGraw-Hill, New York ( 1967).3. Meyer. H.. 'Post impregnatedmicalastic insulation for large motor\and generators.' Siemens Rn.irw. XLV April, No. 4 ( 1978).5. Moses,G.L.. Elec/riu/lrisu/u/ion, McGraw-HillNewYork( 1051 ).6. Moses, G.L., 'Alternating and direct voltage endurance studie\on mica insulation for electric machinery.' AIEE Tr~fn.ctrc.tio,l.s.70,763-769 ( I95 1 ),7. Renganian. S., Agrawal, M.D. and Nema R.S., 'Behaviour ofhigh voltage machine inwlation system in the preqence ofthermaland electrical stresses,'IEEE Trms. 017 €1.. E.I.. 20. No.1. Feh.(1985).8. Rhudy. R.G. ct al.. 'Impulse ioltage strength of AC rotatingmachines.' IEEE Tram.. 100, No. 8 ( 1 98 I ).9. Simons, J.S., 'Diagnosis 01' H.V. machine insulation.' /€€EProcredrng.5. 127. May (1980).

Installation andMaintenance ofElectric Motors10/231

Installation and maintenance of electric motors 10/233One can adopt a suitable procedure of installation,grouting, type of foundation and alignment etc., dependingupon the size of motor, the duty it has to perform and thelocation of the installation (such as hazardous or seismic,etc.). Here we discuss briefly only the important aspectsof installation and maintenance of electric motors.10.1 Installation of bearings andpu I ley sSpecial care needs be taken when mounting or removingthe pulley or the bearing from the motor shaft. Carelessnessin using a correct procedure or proper tooling may bedetrimental to the bearing’s life. It may even damage theend shield of the motor at the other end. Any hammerblows on the bearing, directly or indirectly, can causeirreparable damage to the bearing and the end shield atthe other end of the motor. In view of a bearing’s delicatenature, the following methods are recommended to carryout such tasks.INote The jig can also be motor operatedFigure 10.1 Mounting of the pulleyMounting of a bearing or a pulleyBearings up to medium sizes can be driven on the shaftseat with the aid of a tubular drift, supported on the innerrace of the bearing, then by hammering it gently with amallet, to transmit the blows at the other end to a rigidsupport. For large motors or pulleys, however, a fixtureas shown in Figure 10.1 can be used. The purpose of thisfixture is to grip the inner race of the bearing and causeno thrust on the balls or the rollers of the bearing. IS0286-1 recommends the tolerances and fits for pulley boresand these may be followed for an ideal fit. Table 10.1gives tolerances in the pulley bore for different shaftdiameters and Table 10.2 those in the bearing housing(end shield) bore diameter, where the bearing’s outerrace fits, as well as the motor shaft diameter, on whichthe inner race of the bearing is mounted. The bearing fitsare thus governed by the dimensional tolerancespermissible for the end shield bore diameter and thediameter of the motor shaft. These are called tight fit orshrink fit. It may be seen that any slip between the endshield bore and the outer race of the bearing or the diameterof the motor shaft and the inner race (bore) of the bearing,during transmission of load, may cause undue heating,vibrations and noise. This may also adversely, influence theefficiency of power transmission, and cause severe damageto the bearing inner and outer races, the shaft and thebearing housing (end shield) due to friction, abrasive wear,fretting, corrosion and cracks. All these effects musttherefore be eliminated by a proper fit in all mating parts.Dismounting of a bearing or a pulleyA claw-type puller, as shown in Figure 10.2, withadjustable jaws must be used when pulling out the bearingor the pulley from its seat. The claws are so set that theydo not bear against the outer ring of the bearing whileTable 10.1 Shaft diameter, its tolerance and pulley bore sizeShaft diameter(mm)ToleranceValue of tolerance Tolerance for pulley Bore sizebore(mm) (microns) (mm)9 JS6 LO045 + 15 9.01511141924283842485560657580f .0055 + 18 11.018JS6 + 18 14.018E }JS6E }K6m6m6f 0.0065+ 0.018+ 0.030+ 21+ 21+ 21+ 25+ 25+ 25+ 30+ 30+ 30+ 30+ 3019.02124.02128.02138.02542.02548.02555.03060.03065.03075.03080.03090 m6 + 0.035 + 35 90.035aThere is no lower limit of tolerance in holes.1 micron = 0.001 mm (lpm).Note: For larger sizes refer to IS0 286-1 or IEC 60072-2.

10/234 Industrial Power Engineering and Applications HandbookTable 10.2 Shaft and bearing housing diameters and theirtolerances, to provide a tight or shrink fit to the bearingsDeep groove hull hearingsCylindrical roller bearings(A) Shaft diameterTolerunce(A) Shaft diameterlhleranceUp to 18 mm j,Above 18-100 mm k,Above 100-160 mm m5(B) Housing boredicimeterUp to 40 mm kSAbove 40-160 mm ni5Above 160-200 mm tij(Bj Housing borediameterAll sizes h6 or j6 All sizes h6 Or j6Based on IS0 286- IPullFigure 10.3(a) Wrong method of dismounting the bearings,as the forces travel through the ballsNote The jig can also be motor operatedFigure 10.2 Dismounting of the pulleypulling out and thus exert no thrust on the balls or therollers of the bearing. Figure 10.3(a) and (b) illustratethe correct method.10.2 Important checks at the time ofcommissioningClean up the motor.Check all the components for their positioning andtightness.Check for pre-lubrication/grease and its monitoringattachment (provided in large machines).Check for free rotation of the rotor.Check for motor grounding.Check for winding insulation.Additional checks for large motorsCheck for proper connection and circulation of fluidor gas in the coolant circuit (Section 1.16).Check for satisfactory operation of auxiliary oil pumpand fan.Check all safety devices such as temperature sensorsin the windings and the bearings, PTC thermistors,vibration probes, space heaters and coolant circuit, fortheir correct fitting, wiring and functioning of the alarm,annunciation and tripping circuit of the protectiveswitchgear (Section 12.8).L Pull Pull 1Figure 10.3(b) Correct method of dismounting the bearingsCheck for satisfactory functioning of all gauges,indicators and recorders.Check for bearing insulation (dealt with separately inSection 10.4.5).Check for bearing housing grounding.Check for winding insulation by polarization index(Section 9.5) and dissipation factor, tan 6 (Section9.6)10.3 Maintenance of electric motorsand their checksOnly important aspects have been considered here:Protected-type motors, located in dusty environments,should be blown out periodically with clean and dryair. Replace the motor, if possible.Check the protective equipment for any worn parts orcontacts and their tightness.Check the condition of the cable insulation, its terminationandjointing. Motor connections should always be made

Installation and maintenance of electric motors 10/235through cable lugs to ensure a proper grip as shown inFigure 10.4. A poor cable termination will mean arcingand localized heat and may lead to joint failure.For oil-filled control equipment such as autotransformerstarters or oil circuit breakers (BOCBs or MOCBs,Chapter 19), insulating oil should be checkedperiodically for its insulating properties. Leadingmanufacturers of this equipment indicate the numberof switching operations under different conditions ofload and fault, after which the oil must be replacedand these must be followed.Each electric motor and connected control gear isgrounded separately at least two points. The groundresistance should be checked to ensure continuity ofground conductors. Refer to Chapter 21 for more detailson grounding requirements.Insulation resistance of the motor windings betweenphases and phase to ground should be checked andmade up in the event of deficiency, according to Section9.5.Slip-ring motors need a regular and meticulous checkof brushes, brush holder unit and slip-rings forcleanliness, accurate contacts, brush curvature, wearand tear of slip-rings and arcing (Figure 10.5). Thefollowing procedure may be followed:P- Sprin pressure to be maintained around 150-200g/cm for proper contact of the brushes with theslip-rings.- Cleanliness of slip-rings and the brushes.- Slip-rings, if roughened by arcing, must be cleanedwith fine glass paper, having a similar curvature tothe rings using a wooden block on which the glasspaper can be wrapped. Emery cloth should not beused.- When replacing with the new brushes, the newbrushes must be first ground to acquire a curvaturesimilar to that of the rings.- Brush lifting and short-circuiting devices can beemployed for motors required to run continuouslyfor a long period to minimize wear and tear of slipringsand brushes. However, when speed control isrequired or switching operations are frequent,continuous contact brush gear assembly must beemployed.10.4 Maintenance of bearingsGrease may leave skin effect on the races of the bearingsif the motor is stored idle for a long period. This maycause noise during operation and overheating of thebearings. After a long period of storage grease may alsodry and crack, and produce these effects. To detect this,bearing covers may be opened and the condition of thegrease and any skin effects checked. If such marks arevisible, the bearings must be taken out and washedthoroughly in petrol or benzene to which is added a fewdrops of oil, and then re-greased with a recommendedgrade and quantity of grease. Quantities of grease aboverecommended levels may cause heat the same way asquantities below recommended levels.Recommended brands and grades of lithium-based rustinhibitingbearing grease conforming to IS 7623 are givenbelow:Figure 10.4 Connection of cables in motor terminal boxthrough cable lugsFigure 10.5 Slip-ring assembly with brush lifting and shortcircuitinggear

101236 Industrial Power Engineering and Applications HandbookIMP2 and IMP3CastrolMP greases 2 and 3Lanthax EPl and 2 for large motors }Servogem-2 and 3Indian OilTitex HT - for large motors CorporationLithon-2HindustanLithoplex-2 for large motors PetroleumBalmerNLGI-2 and NLGI-3LawrieThese grades possess a high drop point, of the order of1 80"C, and a working temperature range from -30°C to+130"C.10.4.1 Cleaning bearingsFor cleaning bearings and bearing housings, beforeapplying the grease, only the following fluids should beused: benzone, white petrol and benzole (white petrol isthe most recommended).10.4.2 Relubricating intervalThe relubricating interval depends upon the runningconditions of the bearing, such as the bearing type, size,speed, operating temperature and also the type of greaseused. The re-greasing time is thus related to the serviceconditions of the grease and is represented bywhereRG = re-lubricating interval in hours of operationK = factor as noted belowfor ball bearings - 1for cylindrical roller bearings - 5for spherical roller bearings ortaper roller bearings - 10d = bearing bore diameter in (mm).(10.1)The relubricating period obtained from this expressionis valid for operations up to 70"C, when measured at theouter ring. This interval should be halved for every 15°Crise in temperature above 70"C, which is unlikely tooccur in an electric motor. For operating in areas that arehazardous or contaminated which may affect the life ofthe grease, the regreasing interval may be reduced,depending upon the installation's environment.A more generalized graph from a leading bearingmanufacturer is reproduced in Figure 10.6, and coverspractically all sizes and types of bearings to determinethe relubricating period.When in operation, the bearings should be regreasedthrough the grease nipples or grease injectors, if providedon the motor, after the recommended number of runninghours, as prescribed in Table 10.3. If grease nipples arenot provided, the motor may be temporarily stopped andregreased after removing the bearing end covers.10.4.3 Quantity of regreaseG, = 0.0050 . B grammes (10.2)whereG, = quantity of regrease in grammes. This does notinclude the grease initially placed in the coversand other spacesD = bearing outside diameter (mm)B = bearing width (mm).Table 10.3 provides a ready reference for the quantity ofgrease and its regreasing interval as prescribed by a leadingmotor manufacturer.10.4.4 General precautionsWhen long idle storage is foreseen, arrangements maybe made to rotate the rotor periodically by hand by 90"or so to shift the position of balls or rollers. This willprevent the grease from leaving any marks on the racesand avoid any static indentation on the bearings.After the recommended operating hours, as shown inTable 10.3, the bearings should be taken out and washedthoroughly as prescribed above and fresh grease applied.The bearing covers and the bearing housing should alsobe washed.Note1 For dust- and chemical-laden atmospheres as in cement plants,chemical industries and refineries and areas in alkaline andsaline environments reduce the regreasing intervals by roughly30%.2 For larger frames contact the manufacturer.3 The free space in the bearing should be completely filled whilethe free space in the bearing housing should be filled between30-50%.4 For more details refer to the product catalogues published bythe leading bearing manufacturers.10.4.5 Prevention from shaft currentsIn an electric motor the stator windings are in a circularform. Each phase is wound identically and spaced equallyto each other. Under ideal conditions the electric fieldproduced by these current-carrying conductors must bebalanced and neutralized, as in a three-core cable, producingno residual field. (For more details see Section28.8). But this is not always so. While some field inspace may always be present in all ratings, it assumescognisable strength in large LT and HT machines (2 2MW), using circular laminations and all machines usingsegmented stator punchings.* Such space fields inducean e.m.f. between the two shaft ends. This is not a directresult of proximity but may be due to asymmetry in themagnetic circuit because of harmonics present in thesystem, a non-ideal flux distribution, due to slightdisorientation or variation in the punching for the slotpositions, or an eccentric air gap between the stator andthe rotor. The phenomenon of disorientation in thepunchings for the slot positioning is independent of motorcurrent and influences large LT and HT machines equally.In HT motors even electrostatic effects between thewindings of the stator and the rotor due to stator voltage*Segmented stator punchings are made in two halves and are usedfor very large machines (> 1.5 MW).

Relubricating interval (RG)in operating hoursC b a1.5 -104 ---2 -103 -0 -6 -5 -4 -3 -2.5 -2 -1.5 -10' 9 a 7 -6 5 -4 -3 -2 -2.521.510406432.521.51037.554.543.532.521.510'32.521.51049a765432.521.510390765432Installation and maintenance of electric motors 10/2371.5 - 7.5 1.5-10' - 5 10'Speed (rpm)(Source: SKF,a = Radial ball bearingsb = Cylindrical roller bearingc = Spherical roller bearings, taper roller bearings, full complement cylindrical roller bearings, thrust ballbearings, cylindrical roller thrust bearingsFigure 10.6 Curves to determine relubricating periods for different types and sizes of bearings at different speedsmay also cause shaft currents as a result of capacitivecoupling.Both these effects cause circulating currents to flowthrough the motor frame, forming a complete electriccircuit via the bearings (Figure 10.7). These currents aredetrimental to the life of bearings due to sparking andconsequent pitting and heating effects. Such currents, ifnot prevented, may cause indentation (pitting) on theinner races of the bearings, as well as on their balls orrollers. The pitting may not only cause noise but mayalso loosen the motor shaft within the bearing, causingheating and chattering, so that a stage may arise whenthe rotor may start to rub the stator laminates and causedamage. These shaft currents, therefore, must be preventedwhile installing such motors and the following are somepreventive measures:1 One of the bearings should be insulated inside itshousing by providing a layer of thin insulation asindicated at location 3 in Figure 10.8 to preventcirculating current through the motor bearings.2 Insulation of both the bearings will not prevent thecirculating currents, as illustrated in Figure 10.7.Therefore a sheet of insulating material (rubber liningor Bakelite) may be provided between the motor'sfeet and the base frame, and insulated holding-downfasteners used for fixing the motor. This will preventthe eddy current or leakage current paths forming aclosed circuit through the bearings, as indicated atlocation 1 in Figure 10.8. One of the bearings, generallythe driving end, may also be grounded as illustrated,to prevent a build-up of electrostatic currents in therotor.3 Sometimes a coupling insulation as indicated at location2 in Figure 10.8 may also prevent the eddy currentsforming a closed circuit and generate the bearingcurrents.The insulation, as discussed above, must be protectedduring normal operation to avoid any damage to itotherwise the circuit may defeat the purpose of shaftinsulation. It is advisable to check the insulation periodi-

800010/238 Industrial Power Engineering and Applications HandbookTable 10.3 Quantity of grease and regreasing intervalsBearing boredia. d (mm)To be injected after running hoursSpeed zp.m + 750 100015003000Quantity of grease(g)Quantity of initialgrease(8)4550606570758000 ~40II6o8o80100909510080140188220110260120130320CourtesyM/s SiemensFigure 10.7 Even if both the bearings of the motor areinsulated, there will be a current path as showncally. Creepage of oil, water, moisture, dirt or metallicparticles may also short-circuit the insulation and defeatits purpose. Therefore, the insulating lining, whereverprovided, must be protected and kept clean.Summary1 Cognisable shaft currents may exist in large LT andHT motors of 2000 kW and above, using circularlaminations and all motors with segmented laminationsdue to the magnetic field caused by asymmetries.2 The problem of shaft currents may also be due todielectric leakages that may take cognisance in HTmotors of 2.4 kV and above. This can be prevented bygrounding one of the bearings to prevent the leakages.The bearing insulation is thus determined by themanufacturer while checking the shaft voltage at theworks. This forms a routine in-house test for all HTand large LT motors.To detect shaft currents, the normal procedure ofleading manufacturers is to measure the shaft voltageend to end, with a full voltage applied to the motorterminals. If this is 300-350 mV or more, it will indicatethat the bearings require insulation, as illustrated inFigure 10.8. On very large motors, using segmentedpunchings, shaft voltages even of the order of 1 to2 V have been noticed. These voltages are highlydetrimental to the life of bearings and are undesirable.As standard practice, all such motors are provided witha bearing insulation by the motor manufacturers. Theinsulation is generally provided between the bearingand the end shield at the non-driving end (NDE) ofthe motor, as illustrated at location 3 in Figure 10.8.10.4.6 Reasons for high bearing temperatureThis may be due toExcessive quantity of grease causing churningInadequate grease due to deterioration or leakageMisalignment, causing friction and excessive axialforces

Installation and maintenance of electric motors 10/239Bearing insulationDriven equipmentDriven Equipmentearing insulationInsulation at the foundationFigure 10.8Alternative arrangements to eliminate shaft currentsLoose fit of the bearing housing, causing both innerand outer races of the bearing to rotate inside the housingCorrosion or the presence of foreign matter in thebearing.Any of the above reasons may result in noise and anincrease in temperature and must be corrected. Criticalinstallations such as a refinery, a petrochemical plant, achemical plant or a petroleum pipeline may require specialprecautions and control to avert any excessive heating ofthe bearings, which may become fire hazards. For theseinstallations, bearing temperature detectors with a relayand alarm facility may also be installed in the controlcircuit of the switching device to give warning or trip themotor if the temperature of the bearing exceeds the presetsafe value.10.5 General problems in electricmotors and their remedyOnly the most general types of problems are discussedhere:Bearings make a churning noise or overheat:Check condition and level of grease as well as anyskin effects or watermarks on the races, balls orrollers. If there are watermarks, the bearings shouldbe replaced. Otherwise wash and regrease them, asexplained earlier.For creaking and harsh noises, check for misalignmentand belt tension.Motor not picking up speed:Check all phases for supply continuity.Check the voltage.Check the starter connections and contacts in allthree phases. Also check proper contacts and brushpressure in slip-ring motors. If these are satisfactoryand the motor still does not pick up, check thesuitability of the motor for the type of load andswitching method.4Fault in the switching device. Sometimes, theswitching device may not be functioning properly.To give an instance, one 150 h.p. squirrel cagemotor was selected for a centrifugal air compressor.The motor starting torque was adequately chosento start the compressor through an autotransformerstarter with a 40% tapping. The motor started butlocked up somewhere in the middle of the fullspeed and did not-accelerate further. The supplyvoltage and the transformer tapping were in order.A thorough check revealed that the voltage at themotor terminals was much less than 40%. A longcable drop from the A/T to the motor was suspectedbut when the voltage was measured at the A/Toutgoing terminals, it was almost the same as atthe motor terminals, thus eliminating the possibilityof a longer cable drop. In fact the A/T was faultyor not properly connected so that it was producingonly 25% voltage instead of 40% in the secondarycircuit thus seriously affecting the torque and themotor’s starting performance.Motor not taking up load. This may be due to incorrectstator connections. In general, motors of 3 h.p. andabove, except HT motors, are wound for deltaconnections and all the six terminals are located inthe terminal box to facilitate YlA starting. Theseterminals are connected in delta through metalliclinks (Figure 10.9). If the motor is inadvertentlyconnected in star, each motor phase will receiveonly 118 times the rated voltage. This will reducethe torque to one third of the rated torque (Section4.2) which may not be adequate to pick up the load.It may even damage the windings if the motor remainsenergized for a while due to excessive load current(Z’R losses).While connecting a delta-wound motor through a YlA starter, the metallic shorting links should beremoved. Otherwise the starter will have a dead shortcircuitat the motor terminal box and may burn thestarter, damage the motor terminal box and evenline cables.Motor takes longer to pick up. When conditions 3

10/240 Industrial Power Engineering and Applications Handbookf ai bi ci \(a) Terminal boxFigure 10.9 Terminals inside a terminal box(b) A connected windingand 4 above have been checked but this problemstill arises, then motor selection may not beappropriate for the type of load or the method ofswitching. For example, a ball mill at a thermalpower station required a 75 kW motor employing aY/A manual switching (an old installation). Themethod of switching was overlooked at the time ofmotor selection. The motor burnt during the startwhile the starter was still in the star position. Thestator was found to be completely charred and therotor damaged to the extent that its short-circuitingrings also melted due to excessive starting heat andsome molten metal was scattered over the statoroverhangs causing extensive damage. The reasonsfor such a failure can be as follows:The starter was manually operated and thus thechangeover time, from star to delta, largelydepended upon the skill of the operator, who mayhave been unaware of the implications of a longerstarting period in the star position. Moreover, onload, the starting heat in a Y/A starting is muchmore than on a DOL starting as discussed in Section2.7.3.Note A similar situation may be found in an automaticYlA starter when the setting of the timer, to changeoverfrom star to delta, is at a higher value to allow a longerstarting time to pick up speed. Then also the motor wouldmeet the same fate if the starting time exceeded the thermalwithstand time of the motor.The fuse rating, if used before the starter, mayhave been inadvertently high, inconsistent with thethermal withstand capacity of the motor. Thus thefuse did not blow which would have protected themotor from such severe and persistent overloading.The over-current relay selection or its setting maynot have been appropriate, otherwise the starterwould have tripped under such a faulty condition.Selection and over-current setting of the relayshould have been consistent with the motor’sthermal capacity.A ball mill requires a high starting torque (Section2.5 and Table 2.3) to accelerate its heavy masses.In the star position, the motor torque diminishesto one third the starting torque as on DOL. Themill would take a long time to accelerate and the78910ampere meter would show almost a constantstarting current. The operator, possibly also judgingthe speed of the ball mill, may have paused in thestar position before changing it over to delta butby then the motor had failed. Thus, it was a wrongapplication of the switching device for the type ofload.Motor trips and/or fuses blow during start0 Check selection of fuses for the type of load andswitching. Perhaps they are under-rated for thestarting inrush current and its duration. Forselection of fuse ratings see Section 12.10.4.0 Supply voltage may be low, causing the motor totake longer to start.Starter relay selection or relay setting may not bematching the starting requirements. The relaythermal characteristics must match motorcharacteristics. In heavy drives with prolongedstart times the relay may not operate if it is notproperly selected. For heavy loads, with largemoments of inertia and requiring a longer startingtime, a timer can be introduced into the relaycircuit, to bypass this until the motor acceleratesto a reasonable speed. Then the relay circuit isreactivated. The most common practice is toprovide a CT-operated thermal overload relay(Figure 13.53; see also Section 12.4.1). The CTshave a low saturation point. In the event of a highstarting current, they become saturated at two tothree times the full load current and bypass theexcessive starting kick through the relay and thusavoid a false trip during a prolonged start.Motor vibrates. A motor already tested at the manufacturer’sworks for vibration limits, as shown inTable 1 1.3, may still appear to have more vibrationsat site during operation. This may be due to thefollowing reasons:The foundation bed or the structure on which themotor is mounted is not rigid, is tilted or is uneven.Tighten foundation bolts and check for properalignment of the motor and the drive. Make thefoundation or structure as rigid as possible.0 Single phasing, in which case the motor may makea humming noise and cause vibrations due touneven flux distribution.Bearing-end play may also cause such vibrations.It is possible that the driven equipment to whichthe motor is coupled has a higher vibration levelthan the motor, resulting in quantum imbalanceand more vibrations than when the motor wastested. All attempts must be made, to bring thevibration level of the drive and the driven systemwithin the limits as prescribed in Table 1 1.3.The motor makes rumbling noise and the statorcurrent fluctuates. The rotor circuit may be brokenand should be repaired. If the motor also overheats,there may be an inter-turn fault or a short-circuitbetween the phases. Detect these and rectify ifpossible, otherwise rewind the motor.If heavy rain causes a flash flood at the site ofinstallation, the motor may be submerged in waterfor some time. A reasonably good motor, with good

insulation impregnation and baking, can be washedand dried, to work again. Dismantle the motor, takeout the rotor and bearings. Wash the stator and rotorwith clean water to remove all the mud and silt. Padit dry with cloth. Blow warm air over the stator andthe rotor and heat them gradually, adopting theprocedure in Section 9.5. Unless they show permanentwatermarks or rust or scratches, the bearings can alsobe washed dry and regreased as shown in Section 10.4.11 Cast iron body, feet or ribs etc. found broken orcracked during transit or otherwise. Replacement ofthe motor in such cases may not be practical.However, using the motor may not be advisable inview of a weaker foundation and insufficient cooling.In such cases the broken parts can be welded usingcast iron electrodes. Cracks, however, cannot beremedied. Unless the cracks are wide and may causeextensive damage during operation, the body maystill not require replacement. Minor cracks, however,which do not impair the motor's perfor-mance orcause development of further cracks, may becompromised.10.6 Winding temperaturemeasurement at siteSometimes the motor may appear to be runningoverheated. In fact it may not be so. The easiest way tomeasure the temperature at site is by a thermometer whichcan be conveniently inserted into the hole of the liftinghook. In very small motors where a lifting hook may nothave been provided, a small oil cavity can be drilled inbetween the top fins allowing the thermometer to beembedded there (Figure 10.10)..- IThermometerCavity with oilThermometerreading "CI IFigure 10.10 Measurement of winding temperature at site bythermometer method10.6.1 Temperature correctionWhen the temperature is measured by a thermometer asInstallation and maintenance of electric motors 10/241noted above it obviously does not measure the temperatureof the hottest spot inside the stator. The readings may, atbest, reflect the temperature of the stator housing insteadof the stator winding. As a rough estimate, we can takea temperature gradient of 30% between the surface andthe windings to obtain a near-realistic temperature of thewinding. If 8is the thermometer reading, then the windingtemperature may be around (W0.7) "C and it should beless than the safe working limits. For example, for classE insulation W0.7 I 105°C or 8 I 74"C, and for class Binsulation 8/0.7 I 110 "C, i.e. 8 I 77°C.This illustration is for general guidance when the motoris checked at site for its operating temperature. Thethermometer reading should not exceed the above figures.However, a few degrees above these figures may bepermissible, and this will depend upon the wall thicknessof the stator housing, the air duct between the housingand the stator core, the design of the cooling ribs and theeffectiveness of the cooling fan etc., which only themanufacturer can confirm.CorollaryA surface temperature as high as 70°C or more is obviouslya very high temperature and such a surface cannot beeasily touched. It is therefore natural for a human handto feel very hot when touching the surface of a runningmotor. But to derive a conclusion from this may bemisleading.10.7 Analysis of insulation failuresof an HT motor at a thermalpower stationA powerhouse (thermal) application is the most stringentapplication, as discussed in Section 7.19. Based on fielddata collected from various installations by differentagencies the general insulation failures observed may beattributed to the following.Electrical failures1 Failure at overhangsIn protected motors failure may be due to accumulationof fly ash at the overhangs (modern installationsuse only enclosed motors). Fly ash becomes anextremely good conductor when damp. The failurewill generally occur when a motor is switched onafter a prolonged shutdown.The motor may also fail due to system overvoltages,such as during a fast bus transfer (Section 7.19) ordue to voltage surges (Section 17.3).2 Failure of a coilCoil puncture is also a cause of insulation failure. Itmay be a result of poor impregnation at the manufacturingstage, or due to overvoltages, voltage surgesand ageing. In resin-poor insulations, where the wholestator and rotor, after impregnation, becomes a solidmass, the chances of an insulation failure are remote.In a formed coil (resin-rich) design, however, theremay be a differential expansion (thermal effect)

~~10/242 Industrial Power Engineering and Applications Handbookbetween the insulation, the copper conductor and theiron core during normal running. This may causeloosening of bondage between them, leading tovibrations and shrinkage of the insulation on cooling.This can result in cracking of the insulation, exposingit to the environmental pollution discussed later, andeventual failure.Prevention of insulation failuresWith the use of surge arresters and surge capacitors(Section 17.10)By monitoring the insulation condition of thewindings during maintenance, at least once a year,which can be carried out by measuring (a) thepolarization index (Section 9.5.3) and (b) thedielectric loss factor, tan 6 (Section 9.6) and makingup the insulation as in Section 9.5.2, when thecondition of the insulation is acceptable and onlyits level is less than permissible.A d.c. insulation resistance test or polarizationindex reveals only the surface condition of theinsulation and does not allow a realistic assessmentof internal condition. Loss tangent values are truereflections of the insulation condition to detectmoisture content, voids, cracks or general deterioration.The tan Gversus test voltage curve may bedrawn and compared with the original curve providedby the manufacturer, and inferences drawn regardingthe condition of the insulation. The different startingtan Gvalues will reveal the condition of the insulationin terms of amount of contamination, as noted inTable 10.4 (See IEE, Vol. 127, May 1980).Table 10.4 Conditions of insulation in terms of starting(tan 4Starting value of tan 6%Amount of contamination0-4 Low void content4-6 Clean6-10 Some dirt10-14 Dirt and moisture14-16 Gross contamination16-20Above 20Heavy deposit of oily dirtSevere oil and carbon contamination3 Flash-over in terminal boxThis may be due to fly ash, overvoltages or voltagesurges. For prevention see Section 7.18.Mechanical failuresRotor rubbing the statorSometimes, as a result of an unbalanced magneticfield, causing an air gap eccentricity or excessiveshaft deflection, the motor is not able to maintainthe small air gap between the rotor and the statorand this may lead to failure.It is also possible that after long running hours theballs or the rollers of the bearing have given way.As a consequence of misalignment in the coupling.PreventionEnsure that the supply voltage is balanced.Check bearings and air gap during maintenance, atleast once a year.Ensure an accurate alignment of load.Rotor stampings are loose or rotor bars are damagedMisalignment causes vibrations, which may eventuallylead to failure. The vibrations may also causecracks between the rotor bars and the end rings.Frequent starts and stops may also cause this becauseof excessive heat.PreventionCheck for accurate alignment.Check the rotor’s condition during the annualmaintenance.Environmental pollutionFailure may be caused by coal dust, fly ash andmoisture. Pollution may weaken the insulation,particularly of a protected type motor and result in afailure at some stage.PreventionBlow the surface clean with air at brief intervals.AgeingOvervoltages, voltage surges and overheating ofwindings over many years of operation may dry andshrink the insulation and develop cracks. Throughthese cracks, moisture and dust can penetrate anddestroy the insulating properties of the insulationresulting in an eventual failure of the insulation.Field experience has revealed that one of the majorcauses of failure of an HT motor is weak insulation,caused by environmental pollution and ageing.Relevant StandardsIEC Title IS BS IS060072- 11199160072-2/199060072-3/199460136/1986Dimensions and output series for rotating electrical machines. 1231/1991 BS 5000-10/1989 -Frame numbers 56 to 400 and Flange number 55 to 1080 BS 4999-141/1987Dimensions and output series for rotating electrical 1231/1991 BS 5000-1011989 -machines. Frame number 355 to 1000 and FlangeBS 4999-103/1987number 1 180 to 2360Dimensions and output series for rotating electrical 996/1991 BS 5000-1111989machines. Small built in motors. Flange numberBF 10 to BF 50Dimensions of brushes and brush holders for electrical 991911991, BS 4999-147/1988 -machinery 13466/1992

Installation and maintenance of electric motors 10/243IS0 system of limits and fits. Bases of tolerances, 919 (Part-YII) BS EN 20286- 286-111988deviations and fits 1993 1/1993Code of practice for installation and maintenance of rolling 3090/1996 - -bearingsSpecification for lithium based grease for industrial purposes 1623/1993 - -Code of practice for installation and maintenance of 900/1992 - -induction motorsDimensions of slide rails 2968/1991 (DIN 42923)Relevant US Standards ANSUNEMA and IEEEANSIIIEEE56/1992NEMA/MG2/1993NEMA/MG2/1989NEMA/MG10/1994NEMA/MG1111992ANSIIIEEE:432/1998Guide for insulation maintenance of large a.c. rotating machines (10 kVA and larger)Motors and generators ratings, construction, testing and performanceSafety Standards (enclosures) for construction and guide for selection, installation and use of rotating machinesEnergy management guide for selection and use of three phase motorsEnergy management guide for selection and use of single phase motorsGuide for insulation maintenance for rotating machines (5 h.p. to < 10,000 h.p).Notes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology andlor its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedRelubricating intervalD = bearing outside diameter (mm)B = bearing width (mm)10.1) Further readingRG = relubricating interval in hours of operationK = a factor depending on type of bearingd = bearing bore diameter (mm)Quantity of regreaseG, = 0.0050 x B grammes (g)G, = quantity of regrease (g)(10.2)1 Lundberg, G. and Palmgren, A,, ‘Dynamic capacity of rollerbearings, Acta Polytechnica, Mechanical Engineering SeriesI, Proceedings of the Royal Swedish Academy of EngineeringSciences, No. 3, 7 (1947).2 Lundberg, G. and Palmgren, A,, ‘Dynamic capacity of rollerbearing,’ Acta Polytechnica M.E., Series 2, RSAES, No. 4, 9(1952).3 Lundberg, G. and Palmgren, A., Load Ratings and FatigueLife for Roller Bearings, Std. 11-1978.4 Seameatson, R.W. (ed.), Motor Application andMaintenance Hand Book.

111245Philosophy ofQuality Systemsand Testing ofElectrical Mach i nes

Philosophy of quality systems and testing of electrical machines 111247SECTION 111 .I Philosophy of quality systems(a reference to IS0 9000)The quality of a product is the main consideration whenmaking a purchase. It may be a consumer durable, anindustrial product or a professional service. The qualityof a product also means a reasonable cost-efficient aftersalesservice and a consumer-friendly attitude on thepart of the producer. Farsighted societies and organizations.that conceived such a philosophy long ago evolvedmethods to practice this meticulously and con-sistently.This has helped them to upgrade their own methods andtechnologies to excel in their fields, and today they areahead of their competitors. They recognized thedisadvantages of poor quality as long ago as in 191 I , asmentioned in F.W. Taylor’s book Principles of ScientificMCln

11/248 Industrial Power Engineering and Applications HandbookIS0 9002 - Model for Quality Assurance in production,installation and servicing.Applicable to those who make their productson the basis of some proven designs, orwork as ancillaries for standard products.IS0 9003 - Model for Quality Assurance in FinalInspection and Test.Applicable to those who are engaged inthird-party inspection and testing.IS0 9004 - Quality Management and Quality SystemElements - Guidelines.Applicable to quality assurance for services,such as hotels and hospitals etc.All these standards have since been adopted by themember countries as their national standards, fully or inslightly modified forms, to suit their own requirementsand working conditions. These standards define and clarifythe quality norms and aim at in-house quality disciplines,to automatically and continually produce a product,provide a service or programme to the stipulated specifications,quality norms and customer needs. They guaranteea product or service with a minimum quality. Theenvisaged quality systems thus aim at a work culturethat pervades all those involved in different key activitiesor processes, to achieve the desired goal through carefullyevolved systems.Quality systemsThese deal broadly with the quality concepts as appliedto a product or services, such as design, programming ora process etc., practised by an organization, through bettercommunication and understanding. They can be achievedthroughQuality assurance schemes, andQuality check systemsThe above are implemented through Quality Management(QW.Quality Management stresses participation by allinvolved in the work system, and their commitment tofollow the systems meticulously and consistently.There is a need for interaction and assessment ofdifferent activities or processes to overcome any shortcomingsby improving or readjusting the system ofworking, operation or controls, to achieve a better workculture and an understanding and respect for all in thesystem. When followed, this will result in the following:Better work processesBetter production controlTimely remedial action to achieve the set goals ofquality and productivity.Reviewing and implementing the changing needs ofthe product in the face of tougher competition orchanged market conditions.To draft quality systems and to adhere tosuch disciplinesThis is a work culture. to be inculcated into the wholeworkforce and which must percolate from top managementto the shopfloor. The systems must be effective, wellcommunicated, understood thoroughly and adhered tostrictly. The mode of system will largely depend uponthe type of industry or services. Generally, the systemsare based on a commonsense technique, to achieve thedesired objectives, by dividing the total process relatedto designing, production, programming and other functionsinto many different key activities, identifying the likelyareas or processes where the work process may deviatefrom the set parameters or where it can be furtherimproved. All such key activities must be properly defined,documented, authorized and effectively enforced.The human element, seen in errors, ignorance, lack oftraining, lethargy, illiteracy or indiscipline, or indifference,must be monitored carefully. This may require either anadequately qualified and experienced person or properjob training. Indifference, for reasons other than the above,would be a matter for human resource development, wherea worker’s skills and habits may have to be adapted to fitinto the system.Ingredients of quality systems for amanufacturing unitaDefine the organizational structureDefine the responsibilitiesVendor selection and development, for the inputs thatgo into the production of the itemsChecks and controls over the inputsProduct designProcess, planning and developmentStage inspection, checks and controls, tests and afeedback system for corrective measures. The systemmust help to identify the potential quality areasRecordsDocumentationPackagingStorageTransportationReceipt at site andInstallation and operation.This is only a broad outline for formulating a system.Procedures not applicable may be deleted and those notcovered and are considered necessary may be includedin the above list.Infrastructure facilitiesTo implement the above, the basic facilities that must beavailable are:Human resources and specialized skillsFacility for trainingDesign facilities and engineering back-upManufacturing facilitiesTesting and quality check equipmentAny other facility considered necessary to implementthe quality systems in full.

Philosophy of quality systems and testing of electrical machines 11/249EmphasisModern management systems stress the need for continuousmonitoring of defined systems and procedures forenforcement and improvement of quality systems. Theseshould minimize immediately and eliminate ultimatelythe recurrence of similar problems through preventivemeasures.Auditing (a monitoring tool)As emphasized above. full monitoring of the proper implementationof quality activities is essential. To achievethis, a periodic audit and review of the working of thesystem is an essential element in any Quality ManagementSystem. Shortcomings, if detected, must be corrected assoon as possible and prevented in the future. Documentedreviews help in maintaining and improving managementsystems and techniques.EconomicsThis is an extremely important aspect of the wholeexercise. The basic purpose of all checks and controls isto reduce reworking, reprocessing, failures or rejectionsduring the work process with a view to produce a productof the required quality and hence minimize cost for betterfinancial returns. A high-cost input not commensuratewith the type of the product may defeat the basic purposeof adopting such a system.Product performance and feedbackThis is essential to improve the product by upgradingthe quality or modifying its design or other parametersthat may make it more acceptable and competitive in themarket. This is possiblc by creating a user-friendlyapproach to obtain reactions to the product’s performance.Customer quality checksFor the customer to ascertain the capability of a producerto deliver goods or services to the required standards itis essential for the producer to have all documents thatestablish his or her competence to deliver the goods orservices. according to the prescribed Quality Standards.These documents may coverAdequacy of the Quality Systems supported throughQuality Assurance Plans (QAPs). in design, process,installation and servicing.Capability to achieve the required quality, supportedby quality manualsControl over procurement of raw material and componentsthat compose inputsIn-house checks of various activitiesIn-house final inspection and testing.All the above documents may then be compiled andsent to prospective customers for comments. Anysuggestions may be incorporated into the originaldocuments to further improve in-house working. Thesedocuments will define the quality objectives of theproducing company.Guidelines to deviationsSometimes a compromise in the quality of the inputs(perhaps due to non-availability of a particular componentto the required specification) may be necessory to avoida production delay. Similar constraints in the design orprocess may also make it necessary to have a deviationin design, without undermining the quality requirements.The IS0 emphasizes that, for such deviations. writtenauthorization from a competent authority, and preferablyalso the written consent of the customer. is essentialbefore implementing such a change. This will be for aspecific period and for a specific number of items.Accreditation by I S0So far the IS0 has provided the necessary guidelines forquality systems, as noted above, and which a producercan adopt in a way that may suit best a product andprocess line. The rest is left to the government of a countryto appoint their own accrediting bodies for certificationof producers. These agencies are then authorized tomonitor, perform surveillance audits and issue certificatesto organizations, who wish to obtain these certificatesand conform to the predefined requirements of thesestandards. However, if a product is for global sales, anothercertification by a government body of that country mayalso be necessary, resulting in a multiplicity ofcertifications which. besides being cumbersome, mayalso become time consuming. It is therefore relevantthat the accrediting be necessarily. carried out at onepoint only by a central agency on behalf of ISO, andrecognized by all member countries. Several of thesebodies are now operating in various countries torecommend and accredit producers.Towards Total Quality Management (TQM)The above are broad guidelines along which producerscan formulate a quality system to suit their needs. Thereare no hard and fast rules for disciplines and systems -these arise from experience and continuous practice. Theymust, however, be made more effective so that theyinculcate respect and confidence in all, to adopt and followthem. No system, however good. can be imposed onanyone, unless they are willing to understand, acceptand respect it. This may require an appropriate trainingprogramme which may also he made a part of the worksystem. The overall aim of the above exercise is to achieve‘Total Customer Satisfaction’ through ‘Total QualityManagement’ (TQM). Total Quality Management is acomprehensive term for integrated activities that are putinto practice, with care and development of human resources,to provide complete satisfaction for a customer. It isa goal to be pursued with commitment and perseve-rance.In this era of globalization, anyone who wishes toremain in the race and to excel must make TQM anultimate objective and remember that ‘the pursuit ofexcellence is a never-ending journey’.

11/250 Industrial Power Engineering and Applications HandbookSECTION 211.2 Testing of electrical machinesThis section covers only the tests that are essential on acompleted motor, irrespective of the manufacturingprocedure and stage quality checks. If IS0 9000 guidelinesare assimilated, practised and enforced by a manufacturerso that a customer’s trust is obtained, a final pre-despatchinspection by the customer may not be necessary. Thecustomer, having gained confidence in the practices andQuality Assurance Systems of the manufacturer, mayissue an authorization to the manufacturer to despatchthe material under their ‘own inspection certificate’, ratherthan an inspection by the customer. We discuss belowthe test requirements procedure and the acceptance normsprescribed by various national and international standardsfor such machines and adopted by various manufacturers.To fulfil the above requirements the material inputsfor the motor, such as stampings, steel, enamelled copperwire, insulations and varnishes, bearings, enclosurematerials and hardware must be subjected to a series ofacceptance tests according to norms and standardspecifications. For example, enamelled copper wire usedfor windings must undergo tests at the initial inspectionstage before use. Stage inspection is carried out duringprocessing of the various motor components. Figure I 1.1identifies vital parts of a motor and the correspondingspecification standards, on which these components canbe checked and tested. Instruments and gauges used forinspection must be periodically calibrated and checkedfor dimensional accuracy.Purpose of testingThe purpose of testing an electrical motor is to ensure itscompliance with the norms of design, material inputsand manufacturing accuracy. It determines the mechanicalsoundness and electrical fitness of the machine for itselectrical and mechanical performance. Such tests determinethe following:1 Mechanical aspects: vibration and noisc level.2 Electrical aspects: guaranteed output and performance.3 Temperature rise at the guaranteed output to ascertainthe adequacy of the insulating material and life of themotor. If the temperature rise is more than permissiblefor the type of insulation used, it will deteriorate theinsulating properties and cause thermal ageing. As a*15:2269 -,Figure 11.1 IS specifications for various parts as used in an electric motor

Philosol 3hy of quality systems and testing of electrical machines 11/251rule of thumb, a temperature rise of 10% more thanthe rated may reduce the life ofthe insulation by 40-50%. See Section 9.2 for more details.4. Torque characteristics, i.e. starting torque, pull-outtorque and pull-in torque, in the entire speed range toensure that the motor will develop adequate torque atall speeds to meet the load requirement. See Section1.5 for more details.On an assembled motor the following visual checksare performed before it is run:1 That all the covers and canopies are fitted in theircorrect locations.2 Special-purpose motors such as increased safetymotors, flame-proof or explosion-proof motors mustbe checked for gaps, clearances and creepage distancesof all the mating parts forming flame paths. Theconstruction of these motors must follow IEC 60079as noted in the list of standards.3 The terminal box for its correct position4 The terminal box for proper connections, number ofterminals and their markings5 The correct direction of rotation of the cooling fans6 The heaters and thermistors when provided7 The clearance and creepage distances between theterminals and between each terminal and the ground.On a wound rotor motor, the following additional checksare also necessary after removing the slip-ring covers,prior to testing:1 The numbers. size and grade ofthe brushes and whetherthe brushes are bedded properly to the contours of theslip-rings2 Whether the brushes are free to move in the brushholder and are not slack3 The brush spring tension. Brush pressure, whenmeasured, using finger spp',ing balance, should bebetween 1 SO and 200 N/cm-.4 Whether the brushes are set concentrically on theslip-rings. and each individual brush holder is setapproximately 2-2.5 mm from the slip-ring surface.Slip-rings must be concentric and free from damageand blow holes.After the above checks the motor can be subjected totype and routine tests. For testing instruments, thefollowing class and grades are recommended.11.2.1 Electrical measurementsSelection of testing instrumentsThe indicating instruments used for electrical measurementsmust conform to IEC 6005 1 and have the followingaccuracies:0 For routine tests: class 1 .O or better, andFor type tests: not inferior to class 0.5.The current transformers and voltage transformers, whenused, must conform to IEC 60044 and IEC 60 186, asnoted in the list of standards. Instrument transformerswith the following accuracies must be used:For routine tests: Class 1 accuracyFor type tests: Class 0.5 accuracyQuality of test power suppZyVOltLlgeThe voltage must approach a sinusoidal waveform andshould be balanced. If at the time of conducting thetests the voltage is almost but not coinpletely balanced,arithmetical average of the phase voltage must be usedfor calculating the machine's performance.NoteI The voltage i\ considered to be virtually \inu\oidal if none ofthe instantaneow values of the wave differ from the instantaneousvalue of the same phase of the fundamental wave by more than5% of the amplitude of the latter (TEC 60033-1).2 A rystem of three-phase voltage is considered to he lirtuallybalanced if none of the negative hequence components eyceeds14 over long periods or 1.5% for short periods of a few minute?and zero sequence components exceed I% of the positivesequence components (IEC 60034-11. See also Section 12.2CurrentThe line current in each phase of the motor should bemeasured. If the line current is not exactly equal in allphases, the arithmetical average of the phase currentmust be used for calculating the machine's performance.PowerPower input to a three-phase machine may be measuredby two single-phase wattmeters, connected as in thetwo-wattmeter method. (Section 1 I .4.3). Alternativelya single polyphase wattmeter may be used.FrequencyThe frequency should be maintained ac close to thcrated frequency as possible. Any departure from therated frequency will affect the iosses and the efficiencyas shown in Section 1.6.2.11.2.2 Type testsType tests are conducted on the first machine of eachtype or design to determine the characteristics anddemonstrate its compliance with the relevant Standards.These tests provide a standard reference for any subsequentsimilar machine. The following are the type tests accordingto lEC 60034-1:1 Resistance measurement of all windings and auxiliarydevices (heaters and thermistors. etc.), when themachine is cold (at room temperature)2 (i) Temperature rise test at full load(ii) Resistance measurement of all the windings and

11/252 Industrial Power Engineering and Applications Handbook3456789101112temperature measurement by thermometer whenthe machine is hot.Load testOverspeed testSpeed-torque and speed-current curvesVibration measurements and noise level tests todetermine the machine’s mechanical performanceVerification of dielectric properties(i) On the completed machine with wound coils orformed coils: power frequency voltage withstandor HV test(ii) Process tests during the manufacture of formedcoils(a) Test for insulation resistance - discussed inSection 9.5.3(b) Test for dielectric loss factor or dissipation factortan Gfor rated voltages 5 kV and above(c) Test for impulse voltage withstand for ratedvoltages 2.4 kV and above.Note1 The impulse voltage test is meant specifically for formedcoils only. It serves no purpose for a wound machine,which is already impregnated and cured as a whole mass.2 The tests for insulation resistance and dielectric loss factorshould, however, be carried out on a completed machinealso with formed coils to establish reference data for fieldtests, as noted in Section 9.6. However, these tests on acompleted machine with formed coils do not monitor theprocess quality of insulation.No-load testLocked rotor testMeasurement of starting torque, pull-out torque andpull-in torqueOpen-circuit voltage ratio test for slip-ring motorsVerification of degree of protection.11.2.3 Routine testsRoutine tests are conducted on subsequent similarmachines. The purpose of a routine test is to ascertainthat the machine is assembled correctly and will be ableto withstand the appropriate high-voltage test, and willbe in sound working condition, both electrically andmechanically. As a minimum requirement these tests willconsist of1 Resistance measurement2 No-load test3 Verification of dielectric properties4 Insulation resistance test5 Measurement of open-circuit rotor volts for slip-ringmotors.11.2.4 Seismic disturbancesWe provide a brief account of such disturbances in Section14.6. This also deals with the recommended tests andtheir procedures to verify the suitability of criticalstructures, equipment and devices for locations that areprone to such disturbances.11.3 Procedure for testingIn the following we describe a brief procedure to conductvarious tests and measurements and computation of thetest results according to IEC 60034-1. (For more detailsof the testing procedure the reader should refer to thestandard.)11.3.1 Resistance measurementAt the beginning of the test the motor must be at ambienttemperature. In this condition the temperature andresistance of the windings should be recorded accurately.These values will be used later, with other test results, toevaluate the temperature rise and efficiency.Many types of winding connections are used, dependingupon the type of machine and the application for whichit is designed. The basic star and delta connections aremost common, but a combination of these two with parallelcircuits is also used, on multi-speed and dual-voltagemotors, as discussed in Section 6.1. The connectiondiagram of the motor, showing the connections of thewindings and the terminal arrangement, should also bechecked for correctness of the connections. Resistanceacross the line terminals of the windings should bemeasured, except for a star-connected motor, where phaseresistance is measured.Typical connections are shown in Fig. 11.2, whereR1 = e R, in star-connected windings, and2iR1 =3- xe- R in delta-connected windings.2 iThe following two methods are commonly used formeasurement of winding resistance.The drop of potential method or voltmeterammetermethodIn this method simultaneous readings of voltage at motorterminals and current are taken while using a d.c. sourceof supply and the resistance of the windings is calculated.Current must be restricted to 10% of the rated current ofthe windings. Errors introduced into the measurement,by the resistance of leads and contacts must be compensated.(a) Y-ConnectedRl(b) A-ConnectedFigure 11.2 Measurement of winding resistancei

~~Philosophy of quality systems and testing of electrical machines 11/253The bridge methodHere, the unknown resistance is compared with a knownresistance using a suitable bridge. Resistance above 1 Rcan be measured by Wheatstone bridge. Resistance lessthan 1 52 can be measured by a Kelvin double bridge,where the lead resistance must also be compensated.All precautions must be taken to obtain the temperaturesof the windings when measuring cold resistance. Thesetemperatures, when measuring the cold resistance, maybe obtained by a thermometer placed in contact with thewindings or by resistance temperature detectors, if theyare provided in the windings. The temperature of thesurrounding air will not be regarded as the temperatureof the windings, unless the motor has been idle for along period under similar atmospheric temperaturecondition. If the resistance of a copper conductor is knownat one temperature, it may be calculated for any othertemperature by using the following formula:(235 + t2)R,' = (235 + tl ) .R1whereR1 = resistance measured at tl"C, andR,' = resistance at t20CIf R1, R,' and tl are known, t2 can be calculated.In slip-ring motors the rotor winding resistance willbe measured at the point of connection of the rotor windingto the slip-rings, so that the slip-ring resistance iseliminated from the measurement of the true rotor windingresistance.11.3.2 Temperature riseThis test is intended to determine the temperature rise indifferent parts of the motor windings while running atrated conditions and the permissible temperature riselimits are specified in Table 11.1. While preparing forthe temperature rise test the motor should be shieldedfrom currents of air coming from adjacent pulleys, beltsand other components to avoid inaccurate results.Sufficient floor space should be left between the machinesto allow for free circulation of air. In normal conditions,a distance of 2 m is sufficient. The duration of thetemperature rise test is dependent on the type and ratingof the motor. For motors with continuous rating, the testshould be continued until thermal equilibrium has reached.Whenever possible, the temperature should be measuredboth while running and after the shutdown.1 While conducting the test the negative sequence voltageshould be less than 0.5% of the positive sequencecomponent. In terms of current, the negative sequencecurrent should not exceed 2.5% of the positive sequencecomponent of the system current. Measurement ofany one of the components will be adequate.2 For continuously rated machines, readings should betaken at intervals of one hour or less. For noncontinuouslyrated machines, readings should be takenat intervals consistent with the time rating of themachine. The temperature rise test should continueuntil there is a variation of 1°C or less between thetwo consecutive measurements of temperature.For motors with a short-time rating, the duration ofthe test should correspond to the specified short-timerating. At the end of the test the specified temperaturerise limits should not exceed. At the beginning of thetest, the temperature of the motor should be within5°C of that of the cooling air.In motors for periodic duty, the test should becontinued until thermal equilibrium has been reached.Unless otherwise agreed, the duration of one cycleshould be 10 minutes for the purpose of this test.Temperature measurements should be made at theend of a cycle to establish thermal equilibrium.Measurement of temperatureWhen thermal equilibrium is reached, the motor must bestopped as quickly as possible. Measurements must betaken both while the motor was running and after shutdown(wherever possible). No corrections for observedtemperatures are necessary if the stopping period doesnot exceed the values given in Table 11.2.Where successive measurements show increasingTable 11.1Permissible temperature rise for naturally (indirectly) cooled motorsMethod of measurementInsulation classA E B F Ha b U b a b U b a bResistance "C 60 60 75 - 80 80 105 100 125 125ETD "C 65 65 - - 90 85 110 105 130 130On the basis of IEC 60034-1a - < 50OO kWb - 2 5000 kWNotes1 The temperature rise limits are based above an ambient temperature of 40% For temperatures other than 40°C see Section 1.6.2(C).2 For exact details refer to IEC 60034-1.3 For other types of cooling methods refer to IEC 60034-1.

11/254 Industrial Power Engineering and Applications HandbookTable 11.2 Permissible stopping period of the motor aftershutdown when it will require no temperature correction0-50 kW 30 seconds51-200 kW 90 seconds201-S000kW120 secondsBeyond SO00 kWBy agreementAccording to IEC 60034-1temperatures after the shutdown, the highest value mustbe taken. When the rotor temperature is also required itmust be measured by recording the highest temperaturerecorded in the thermometers placed on the rotor barsand core, in squirrel cage motors, and on collector ringsin wound rotor motors. A thermometer should be insertedas soon as the rotating parts come to rest.Where the temperature can be measured only after themotor has stopped (as in temperature measurement bythe resistance method), a cooling curve is plotted, bydetermining the test points as rapidly as possible.Extrapolation of the cooling curve is carried out todetermine the temperature at the instant of shutdown.This may be achieved by plotting a curve with temperature/resistance readings as ordinates and time as the abscissausing semi-logarithmic graph for the resistance and alogarithmic scale for the time. This curve can be plottedon semi-logarithmic graph paper similar to that shownin Figure 9.5(b) to obtain a straight-line plot of resistanceversus time to help the correct extrapolation. The followingare the recommended methods to determine thetemperature rise:1 Resistance method - this is the most preferred methodfor motors up to 5000 kW.2 Embedded temperature detector (ETD) method-thismethod is used for stator windings of 5000 kW andabove as in IEC 60034-1.NoteFor motors 201-SO00 kW-both the resistance or the ETD methodmay be used.3 Thermometer method-this is recommended only whenboth resistance and ETD methods are not practicable.Resistance methodThis is the preferred method. The temperature of thewinding is determined by observing the increase inresistance of the winding with respect to the cold resistancemeasured.The resistance must be measured with extreme careand accuracy, since a small error in measuring theresistance will cause a much larger error in determiningthe temperature rise. When the temperature of the windingis to be determined by the resistance, the temperature ofthe winding before the test, measured either bythermometer or by ETD, may be considered as thecold temperature for the resistance measured. Themachine must be left cold for at least 12 to 24 hours,depending upon the size of the machine, to obtain astable reading.For copper windings, the temperature rise t2-t, maybe obtained from the ratio of the resistance bywheret, = temperature ["C) of cooling air or gas at the endof the testt2 = temperature ("C) of the winding at the end of thetestR;= resistance of the winding at the end of the testtl = temperature ("C) of the winding (cold) at themoment of the initial resistance measurement andR, = initial resistance of the winding (cold)ETD methodEmbedded temperature detectors are resistance temperaturedetectors (RTDs) or resistance thermometersor thermocouples, built within the machine duringmanufacture at points that are not accessible whenthe machine has been assembled. This method is generallyemployed for the likely hot spots of a machine suchas the slot portion and the overhangs of the statorwindings.At least six detectors are built within the machine,suitably distributed around the circumference and placedbetween the layers along the length of the core wherethe highest temperature is likely to occur. Each detectoris installed in intimate contact with the surface, whosetemperature is to be measured and in such a way that thedetector is effectively protected from contact with thecooling air. A detector embedded beneath the windinglayer inside the slot is of little consequence for it willdetect the temperature of the core and not of the winding.The location of the detectors must be as follows:For two coil sides per slot: When the winding has twocoil sides per slot, each detector must be located betweenthe insulated coil sides within the slot [see Figure 12.42).For more than two coil sides per slot: When the windinghas more than two coil sides per slot, each detectormust be located in a position between the insulatedcoil sides at which the highest temperature is likely tooccur.Since overhangs are vulnerable parts of a stator winding,detectors can also be placed within them (Figure 12.39).Note The embedded temperature detector method is inappropriatefor stator windings, which have only one coil side per slot, in suchcases the resistance method must be used with the same limits oftemperature rise. For checking the temperature of such a windingin service, an embedded detector at the bottom of the slot is of littleuse because it would give mainly the temperature of the iron core.A detector placed between the coil and the wedge will follow thetemperature of the winding much more closely and is, therefore,better for check tests, although the temperature there may also bea little less than the actual one.Thermometer methodThis method is applicable where neither the embeddedtemperature detector nor the resistance method is

Philosophy of quality systems and testing of electrical machines 11/255possible. The thermometer method is also recognized inthe following cases:When it is not practical to determine the temperaturerise by the resistance method, as in the case of lowresistance windings, especially when the resistance ofjoints and connections form a considerable percentageof the total resistance.Single-layer windings, revolving or stationary andWhen, for reasons of mass production, the thermometermethod alone is used. although the resistance methodwould be possible.In thi\ method. the temperature is determined bythcrniometers applied to the accessible surfaces of thewindings of the motor (see also Figure 10.10). The term'thermometer' may include mercury or alcohol bulbthermometers as well as embedded thermocouples andrcsistance thermometers, provided the latter are appliedto points accessible to the usual bulb thermometer also.When bulb thermometers are employed where there isa likelihood of a magnetic field, alcohol thermometersshould be preferred to mercury thermometers. as the latterare unreliable under such conditions.Measurement of cooling air or gas temperatureduring testThe cooling air temperature should be measured by severalthermometers placed at different points around and halfwayup the motor, at a distance of 1-2 m and protectedfrom heat radiation and draughts. The value to be adoptedlor the temperature of the cooling air or gas during a test.;hould be the mean of the readings of the thermometersplaced as mentioned above, taken at equal time intervalsduring the last quarter of the duration of the test.In order to avoid error, due to time lag in the temperaturemeasurement of large motors and variation in the coolingair or gas temperature all reasonable precautions mustbe taken to reduce these variations and errors as much aspossible.In cooling by means of forced ventilation, or wherethe machine has water-cooled air or gas coolers, thetemperature of the air or gas, where it enters the motormust be considered as the cooling air or gas temperature.For motors fitted with heat exchangers the temperaturereading of incoming air or gas or water should also betaken.Cooling air temperatureThe motor may be tested at any convenient value ofcooling medium temperature less than 40°C. But whateverthe value of this coolingmedium temperature, thepermissible rise of temperature during test should notexceed those shown in Table 1 1.1.In motors intended to operate under conditions in whichthc maximum cooling air tcmperature may exceed 4U°C,the temperature rise as given in the relevant specificationmust be reduced as follows:0 By 5°C if the temperature of the cooling air exceeds40°C by 5°C or lessBy 10°C if the temperature of the cooling air exceeds40°C by more than 5°C but not more than 10°C andBy agreement if the temperature of the cooling air ismore than 10°C above 40°C.Temperature rise of bearingsThe temperature rise of bearings may be measured asfollows:For sliding contact bearings (sleeve or thrust bearings)the temperature readings must be taken from as nearto the bearing surface as possible.For ball and roller type bearings the temperaturereadings must be taken at the stationary race.For oil lubricants it is customary to measure thetemperature in the oil reservoir.0 In forced lubricating systems, incoming and outgoingtemperature readings may be taken.NotrI The permissible temperature riw of the beat-ings will dependupon the type of lubricant used and the recommendations madeby the lubricant and hearing manufacturers.2 It may not be possible to measure the temperature riw of thchearings at the races in all case\. In such a case. it may bepermissible to measure the temperature Ihr ball and roller typebearing< at the stationary race or at the bearing cover.Temperature correctionI For motors specified for operation at an altitude higherthan 1000 m, but not in excess of 4000 m. the correctionmay be made as shown in Table 1.8.2 A temperature rise test may be carried out at anyconvenient cooling air temperature. When thetemperature of the cooling air during the test is lowerthan the stated site cooling air temperature by lessthan 3O"C, no correction should be made on accountof such a difference. When the temperature of thecooling air is lower by more than 30"C, the permissibletemperature rise on test should be the permissibletemperature rise under specified site conditions,reduced by a percentage numerically equal to onethirdof the difference between the specifiedtemperature of the cooling air at site and thetemperature of cooling air on test, where bothtemperatures are expressed in degrees Celsius.Example 11.1If the specified temperature of the cooling air at site is 48°Cand the temperature of the cooling air during test is 17°C (adifference of more than 30"C), the reduction in temperaturerise to take account of this difference will be:-== 48-17 10.33The permissible temperature rise on test will therefore be100 - 10.3% = 89.7% of the recommended temperature rise.These reductions apply to all the classes of insulation asindicated in Table 11 .I and the test is to be carried out at themanufacturer's works.

11/256 Industrial Power Engineering and Applications Handbook11.4 Load testTests on load are conducted to determine the performanceof the machine, such as its efficiency, power factor, speedand temperature rise etc. For all tests, a machine withload should be properly aligned and securely fastened.Load characteristics are obtained by taking readings athigh loads, followed by reading at lower loads. This isusually carried out at 125%, loo%, 75%, 50% and 25%of the full load values.Methods of loading1 Brake and pulley method (usually for very smallmotors). Considerable care needs be taken in theconstruction and use of the brake and pulley. Whenconducting this test conditions should be such that ascale pointer remains practically stationary at any givenload. Proper cooling, preferably water cooling, shouldbe provided for the pulley.2 Dynamometer method (for medium-sized motors, sayup to 500 hp). The output of an induction motormay be calculated by3 Calibrated machine. When brake and pulley ordynamometer methods are not possible. the test motormay be loaded onto a calibrated generator. Theefficiency curve of the generator must be available.When it is not possible to conduct any of the abovethree methods, the test motor may be loaded onto anuncalibrated generator or any other loading device.11.4.1 To determine efficiencyBy the input-output methodAs noted above, efficiency may be detemiined by adoptingany of the following three methods:Brake and pulleyDynamometer andCalibrated machine.By summation of lossesCalculation of efficiency is based on the readings obtainedafter the heat run test when the machine has achievedthermal stability. The losses will fall into following fourgroups:1 Core friction and windage loss (obtained from theno-load test)2 Stator copper loss (primary loss)3 Rotor copper loss (secondary loss)4 Stray load loss (hysteresis and eddy current core loss).The procedure to be followed is as follows:From the resistance measured across the stator lineterminals at the conclusion of the no-load test, calculatethe stator winding resistance R,.0 Calculate the no-load stator copper loss: i2. R, watts.Subtract the no-load stator copper loss from the statorinput power P,, at no-load, i.e. (Pnt- i2. RI). This isthe net no-load loss made up of core, friction, windageand stray losses.Calculate the total stator winding resistance:I?,!,,, ,= 1.5 (1.5 for A-connected, and 0.5 for Y-connectedwindings, Figure 11.2) x cold values of the resistance,measured across the stator line terminals, corrected tothe required temperature. The required temperaturevalues will be:For class A 75°CFor class E 90°CFor class B 95°CFor class F 115°CFor class H 130°C0 Calculate the stator copper loss on load: I,?. Rlhot.0 Add the net no-load loss and stator copper loss onload and subtract their sum from the stator input powermeasured on load. The remainder is the power inputto the rotor.0 Calculate the rotor copper loss (input to the rotor xpercentage slip on load)Subtract this rotor copper loss from the rotor input.Subtract the stray loss to give the motor output fornominal full load. The amount of stray loss is normallytaken as 0.5% of the nominal power output of themachine. The stray losses at other values of loads areobtained from:(required stator current)2x stray loss at full load(full load stator current)'After deducting the stray loss, the resultant kW givesthe machine output.kW outputEfficiency, 1) = kW inputX 100%As a check, add together the net no-load loss, thestator copper loss on load, the rotor copper loss onload and stray loss to give the total losses (total fixedloss plus load loss):The two must tally.11.4.2 Slip measurementkW inputFor the range of load for which the efficiency is determined,the measurement of slip is very important. Todetermine slip by subtracting from the synchronous speedthe value of speed, obtained through a tachometer is notrecommended. The slip must be directly measured byone of the following methods:1 Stroboscopic2 Slip-coil3 Magnetic needle4 Any other suitable method.

Philosophy of quality systems and testing of electrical machines 11/257Methods (2) and (3) are suitable for machines havingslip of not more than 5%.Stroboscopic methodOn me end of the motor $haft a single black radial lineis painted upon a white background. The slip is easilymeasured by counting the apparent backward rotationsof the black line over ;I given period of time.Slip-coil methodA suitable slip-coil. hating approximately 700 turn5 of1 mm diameter insulated wire. is pasjed axially overthe motor and its two ends are connected to a centreierogalvanometei-. When the motor is running, thegal\anometer pointer will oscillate. The number ofoscillations should be counted in one direction only,that is to the left or to the right, for a period of. say.20 seconds.The following formula will determine the percentagedip:whereS = slip (%)11 = number of oscillationst = time in seconds taken for II oscillations. and,j'= supply frequency in Hz.Magnetic needle methodIn this inethod ;I magnetic needle suspended on a sharppoint (so that it can rotate freely) is placed on the bodyof the inotor in the horizontal plane. The needle willoscillate and the number of oscillations should be countedfor ;I period of, say. 20 seconds. The percentage slip isthen calculated by the formula given above.11.4.3 Power Factor measurementThe Power Factor indy be mea\ured by one ot thelollowing three method\Watt to volt-ampere ratioTwo-wattmeterPower Factor meterWatt to volt-ampere ratio methodThe Power Factor is obtained by the ratio of the algebraicsum of wattmeter readings to volt-ampere readings. Fora three-phase system;Power Factor =Two- wattmeter methodWatts\'3 x line volts x line amperesOn a three-phase motor where the load is pulsating thePower Factor may be checked by the following formula.obtained from independent wattmeter readings:Power Factor = -_______ 1whereW, = the higher of the tu0 reading\ andW? = the Iowcr of the two reading\.If W2 gives a negative reading it should be considered asa minus quantity. From the above fomiula. graphs canbe plotted for Power Factor versus ( W2/\V',). W2/WI beingthe ratio of lower wattmeter reading to the higherwattmeter reading.If W, is negative, the ratio of ( W2/\Vl) should beconsidered as 21 ininus quantity. The falling curves 5houldbe designated 'for ( W2/WI)' and the ri5ing cur\c$ 'for(+W21WI)'. Similarly, ordinates on the left-hand sideshould be designated for (-W2/Wl) and on the right-handside for (+W2/Wi). See Figure 113.,Vote If two \~;iltic\ of the Power FactoI dctci-mined h> the \vat[ to\olt-ampere ratio md two-m attiiietci- tnethod\ do riot tall) for Jthree-phaw niotor. the tc\t may he repeated to climinate the error.Howevci-. \\ her-e the load i\ fluctuating. ii Poser Factoi- dctcrminedby a two-\battiiieter method \\ill he hisher h ~ tthat i determined bythe matt to volt-ampere ratio method. In thir case. the higher value\hould be taken a\ the cort-ect reading. The ilill'ercnce I\ due to theincluhioii of a pul\attng coinlwnent 01' cui-rent in volt-;impcrcs.whlch i\ ;I function of load rathct- than ol the riiotor it\clf. ThePower Factor determined h-on1 the ratio of atLnieter I-cadi ng i\ nutatfected by the presence ol ii pirlsating curt-ciit.Power Factor meter methodIn this method. a Power Factor meter is directly connectedin the circuit and ii direct reading i.; obtained at anyloading.11.4.4 Overspeed testAll niotors are designed to withstand 1.2 times themaximum rated speed. The test is simple and may becarried out by running the motor for 2 minutes at thehighei- specd. After the test. the motor must have nodeformation or any fault that may prevent it from operatingnormal 1 y.11.4.5 Test for speed-torque and speed-currentcurvesThe speed-torque characteristic is the relationship betweenthe torque and the speed. in the range from ~ero tosynchronous >peed. This relationship. when expressedas a curve. will include breakdown torque (pull-out torque).pull-up torque and starting torque. The speed-currentcharacteristic is the relationship of the current to thespeed.MethodsSpeed-torque and speed-current te\t\ may be carriedout by the following method\

11/258 Industrial Power Engineering and Applications HandbookX (minus)+0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 -0.8 -0.9 -1.00.6 1.10.5 1 .ot 0.4a 0.3ul80.2 0.70.1 0.6-0 0.50 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0x (plus)Figure 11.3 Power Factor versus ratio of watt meter readings0.90.8W2x= - = ratio of watt meter readingsWIof like polarity.-x=-!% w, = ratio of watt meter readings ofunlike polarity.DynamometerPony brakeRope and pulleyCalibrated machine.Readings of voltage, current and speed should be taken.The torque value is obtained directly by the dynamometer,pony brake and rope and pulley methods and indirectlyby the calibrated machine method. Speed-torque andspeed-current tests must be conducted at a rated voltageor as near to this as practical. When it is necessary toestablish values of current and torque at the rated voltage,based on tests made at reduced voltage, the current mayincrease by a ratio higher than the first power of thevoltage and the torque by a ratio higher than the squareof the voltage due to possible saturation of flux leakagepaths. See also Figure 27.2(b) and note under serial number9 Section 7.19 for more clarity. This relationship varieswith the design and, as a first approximation, is sometimestaken as the current varying directly with voltage andtorque with the square of the voltage.It is therefore necessary to take precautions during thetest to avoid a excessive temperature rise and consequentdamage to the windings. For wound rotor motors, speedtorqueand speed-current tests may be taken betweensynchronous speed and the speed at which the maximumtorque occurs.Pull-up torqueThe motor should be mounted with a suitable loadingarrangement and the rotor fully locked. The rated voltageat the rated frequency will then be applied to the motorterminals in the locked rotor condition. The loading onthe motor will then be reduced slowly so that the motorcan start and pick up speed. The value of pull-up torqueat which the rotor picks up speed and attains speedcorresponding to pull-out torque condition must be noted.Pull-out torqueThe motor should be mounted with a suitable loadingarrangement. The rated voltage, at the rated frequency,is then applied to the motor terminals at the no-loadcondition. The load on the motor may then be graduallyincreased, and the maximum load at which the motorstalls noted. The torque determined at this point is thepull-out torque.Note1 The motor should be immediately disconnected from the supplywhen it stalls.2 The motor should not be kept in the locked rotor condition formore than a few seconds to avoid damage to the windings.11.4.6 Vibration measurement testThe vibration limits have been classified into two groups:1 For shaft heights 56 mm and above, IS0 2373 (IEC60034- 14) has prescribed three categories of vibrationlevels in terms of vibration velocity, one for normaluse, N, and the other two for precision applications,i.e. reduced level R and special-purpose S. Whenrequired other than normal, these must be specifiedby the user to the manufacturer. Machines with ahigher degree of balance should be used only whenthis is essential. Such machines may be far tooexpensive to produce, and sometimes not commensuratewith the application.2 For shaft heights more than 400 mm, IEC 60034-14prescribes the vibration level, in terms of doubleamplitude vibration, which can also be derived fromthe velocity of vibration, using the following formula:a = 0.45 x %!&f

Philosophy of quality systems and testing of electrical machines 111259wherea = double amplitude of vibration displacement, peakto peak (mm)Vr,,,8. = r.m.s. value of velocity of vibration (mds)f = frequency of vibration, which is approximatelyequal to the supply frequency in Hz.Both these levels are indicated in Table 11.3. For moredetails and for conducting the vibration test, referencemay be made to IEC 60034-14.11.4.7 Measurement of noise levelWhen measuring the noise level, i.e. the limiting meansound power level in dB for airborne noise emitted by amachine, reference may be made to the following IECand IS0 publications:IEC 60034-9 - For recommended values of noise limitsin dB. A decibel (dB) is the unit of sound power leveland is derived from the unit of sound measurement (‘bel’),so called after the American inventor Alexander GrahamBell andldB=-bel110IS0 1680-1, Acoustics - Test code for the measurementof airborne noise emitted by rotating electrical machinery- Part 1: Engineering method for free-field conditionsover a reflecting plane.IS0 1680-2, Acoustics - Test code for the measurementof airborne noise emitted by rotating electrical machinery- Part 2: Survey method.IS0 3740, Acoustics - Determination of sound powerlevels of noise sources - Guidelines for the use of basicstandards and for the preparation of noise test codes.11.4.8 Verification of dielectric propertiesPower frequency withstand or HV testThis test is conducted only when it has been determinedthat the insulation resistance, measured as noted in Section9.5.3 is acceptable. The test should be performedimmediately after the temperature-rise test on thoseoccasions when the latter test is also to be carried out.This test reveals any weakness of the insulation orinsufficient clearance between coils or between windingand core. It consists of the application of a high voltage,as shown in Table 11.4, between the windings and theframe (or cores).Windings that are not under test and all other metalparts must be connected to the frame during the test. Thewindings under test should be completely assembled.The test voltage should be of power frequency and asnear to the sine waveform as possible. Component parts,such as space heaters, thermostats and resistance temperaturedetectors, which are connected to parts other thanthe power line, must be tested at twice their rated voltage,plus 1000 volts, with all other windings and componentsconnected together and then connected to the frame (core).Insulation breakdown during the application of a highvoltage should be considered as a dielectric failure. Thetest should commence at a voltage not more than onehalfof the full test voltage. The voltage should be increasedto the full value in not less than 10 seconds and thisvoltage will then be maintained for one minute. At theend of this period, the test voltage will be rapidlydiminished to one-third of its value before switchingoff.The test voltages for wound rotors, reversing and brakemotors are also indicated in Table 11.4. Repetition ofthis test is not recommended to avoid excessive stresseson the insulation. However, when this becomes necessarysuch as at site before commissioning, the test voltagemust be limited to only 80% of the actual test voltage.After the test, the insulation resistance must be checkedagain, to make sure that no damage has been caused tothe windings.Dielectric loss factor or dissipation factor tan SThis is a test to monitor the quality and dielectric behaviourof the insulating system of high-voltage machines, 5 kVTable 11.3 Limits of vibration levels when measured in a state of free suspensionVibration gradea Speed NP (r.p.m. j Maximum cm.s. value of the vibration velocity (mdsj for shajl heighf H (mm)56 S H I 132 132 < H 1225 225 < H I400 H > 400NRS>600 I3600 1.8> 600 5 1800 0.71> 1800 I 3600 1.12>600 6 1800 0.45> 18001 3600 0.712.81.121.80.711.123.51.82.81.12N.A1.82.81.81.8Based on IEC 60034-14“Vibration grade signifies the accuracy of rotor balancing, e.g. N = normal, R = reduced, S = special. This may be based on the type ofinstallation and the accuracy of the function the motor may have to perform.%or both 50 or 60 Hz systems.Note The level of vibration at site may be higher than that mentioned above, perhaps due to the foundation or the coupling of the load.This must be checked and adequate precautions taken to avoid excessive vibrations.

11/260 Industrial Power Engineering and Applications HandbookTable 11.4 Test voltages for conducting dielectric testsMotor size and raring Test voltage (xm.s.) Examp leStator windings(i) LT motors < 1 kW and ratedvoltage (V,) < 100 V(ii) All sizes and ratings of motorsRotor windings (for woundmotors)(i) Non-reversing motors(ii) Motors suitable for reversingduty or braking during runningby reversing the UC supplyIf high-voltage test needs berepated at site duringcommissioningCompletely rewound windingsPartially rewound windingsOverhauled machinesBased on IEC 60034-1500 v + 2vr1000 V+ 2V,(minimum 1500 V)(Same as col. 5 Table 11.6)IOOOV + 2 RV1000 V + 4 RVAt 80% of the aboveFull test voltage as for S Nos 1 and 275% of the test voltage as for S Nos 1 and 2lSV, minimum being lo00 V(a)(b)For 220 V motors, the test voltage= 1000 + 2 x220, Le. minimum 1500 VFor 11 kV motors, the test voltage= 1 + 2 x 11 = 23 kV.RV = rated open circuit standstill voltageacross the slip-ringsand above and ratings lo00 h.p. and above, during thecourse of manufacture of resin-rich formed coils. Fordetails on tan 6, see Section 9.6.1 and Figures 9.1 and9.8. This is mainly an in-house stage inspection for suchcoils. It is conducted on each individual coil, during thecourse of manufacture, to check for adequate insulationimpregnation and quality of insulation, before insertioninto the stator slots. The same process would apply tothe rotor slots of a wound rotor when the rotor opencircuit voltage is a minimum of 5 kV, which is rare. Thisis commonly known as the dielectric loss factor ordissipation factor of a motor winding coil and is thebasic measure of the condition of the insulation to ground.It also gives an idea of ageing or the general condition ofthe insulation. With the help of this data, the processingquality of the insulation of the coils can be easily monitoredas well as the condition of the insulation between theconductor laminations, the inter-turn insulation and theinsulation of the end windings (over-hangs) etc. Thisfactor is also useful in determining the insulation conditionof each slot of the stator or the rotor.To carry out this test on a wound machine (postimpregnated)would be pointless as the quality of theinsulation of such coils cannot be altered after they havebeen inserted into the slots. However, the test is carriedout on a completed identical machine to establish referencedata for field tests. Random sample testing is, however,possible with two identical coils placed in the slots. Thetest sample of slots can also be made. For more detailssee Section 9.6 and IEC 60894.Method of measurement and acceptance normsA Schering bridge* or an equivalent type of bridge isused to determine the values of loss factors tan 6 andA tan 6, i.e. the increase in tan 6values with the voltage.A graph is then plotted between the behaviour of tan 6with the applied voltage as shown in Figure 9.8. Thisgraph also provides basic reference data for field checksbefore the motor is energized.The loss tangent should be measured on the samplesat room temperature at voltages varying from 20% to*The basic principle of these methods is to charge a capacitor up tothe specified test voltage and then discharge it through the coilunder test.I 2 34 5tan 60.2,+ (tan 60.6, - tan 60.v,)100% samples 95% samples Remaining 5% samples30 x 2.5 x 1r3 3 x IO-’(A tan 6 - per step of 0.2VJ95% samples Remaining 5% samples5 x 10-3 6 xNote If more than 5% of the samples show test results in the range of columns 2 and 3, or between columns 4 and 5, the test can beregarded as satisfactory. Otherwise the test can be continued with an equal number of further samples, if necessary, even up to the totalnumber of bars or coil sides.

Philosophy of quality systems and testing of electrical machines 11/261100% of the rated voltage at intervals of 20%. The initialvalue of tan &b.zVr and the increment Atan 6, i.e.3 [tan 60.6~~ - tan 60~2~~1per measuring step shouldnot exceed the values indicated in Table 11.5 for therated voltages up to 11 kV.It can be seen that up to 1.1 times the rated voltage(VJ, the tan 6 value remains almost constant, and at1.2Vr it increases slightly. At higher voltages it increasessharply, and may become too high to cause a dischargesufficient to char the insulation if this voltage is allowedto exist for a longer period.11.4.9 Impulse! voltage withstand test of theinsulation system for rated voltages2.4 kV and above for machineswound with formed coilsAs discussed in Section 17.5 a machine may be subjectedto voltage surges due to external causes (lightning) orinternal causes (switching). By the power frequency HVtest or the dissipation factor, as in Section 11.4.8, orinsulation resistance tests as Section 9.5.3, the surgewithstand level of the insulating system cannot bedetermined. Hence, the need for an additional impulsevoltage withstand capability test of a coil for HV systems.The increasing application of vacuum- and gas-filled (SF6)circuit breakers and contactors for switching of HTmachines, has led to the need for a surge voltage withstandtest on the multi-turn coils of a machine to account forthe surges generated by a re-striking pheno-menon, suchas that caused by the closing or interrupting of contacts.Until a few years ago, there was no widely acceptedstandard for a voltage endurance test of the rotatingmachines. Different agencies had adopted differentpractices on different assumptions, pending a final decisionby the IEC working committee TC-2 of ZEC 60034-15.The committee submitted its report in 1988 and thefollowing test data, which are now universally adopted,are based on this report.The impulse test by means of a directly injected steepfrontedwave cannot be performed on a fully assembledmachine as the bulk of the voltage, if the front of thewave is less than Ips, will appear across the first fewturns only. Also, in the event of a failure, the wholewinding must be scrapped. The impulse test is thereforeperformed on completed individual resin-rich formed coilsonly after insertion into the slots.The impulse test is basically an in-house coil insulationwithstand test for surge voltages and forms a part of thetest requirement for HT machines with resin-rich formedcoils of 2.4 kV and above. Once the machine is assembled,such a test is unnecessary, as it may not be able to revealdeficiencies, if any, in the insulation of the coils deepinside the slots. Moreover, if a failure is noticed on theassembled machine, there is no option but to scrap thewhole winding.To assign the impulse levelIn Table 11.6 the values of column 2 relate to normaloperating conditions. Abnormal conditions may prevailwhen the machine is exposed to overhead lines, theType of impulse Winding switching impulse withstand level, when subjected Winding lightning impulseto switching impulses, with following wave frontswithstand level 1.2/50 psimpulse, phase to ground'FBetween 0.2 and 0.4 p 0.2 p (average kV (peak)value)Line voltage Normal design Special designlevel phase to level phase togroundgroundkV (peak) kV (peak)233a411.08.113.51218.216.226.92031.426.944.93249.013.833.856.33960.2Rated power frequencywithstand voltage (rm.s.)(as in Table 11.4 forstator windings) kVDesign criteria 1 3 p.u. kV 5 p.u. kV 1 0.65 (4Vr + 5)" kV I (4Vr + 5)b kV 1 (2V, + 1)'kVNoteV, = line voltage in kV (r.m.s.)vrp.u. = per unit voltage, which is the peak value of phase voltage in kV i.e. -43WC 60034-15 has prescribed an average impulse voltage, considering the average characteristics of the machine windings and theswitching conditions. One may therefore decide between columns 3 and 3a, depending upon the exposure of the machine to the internalswitching surges and surge protection devices if provided with the motor.bIEC 60034-15 recommends these values to be rounded off to the nearest whole number. The values noted above are not rounded off formore clarity.'See also Table 11.4.57.614.223.028.6

11/262 Industrial Power Engineering and Applications Handbookinterrupting device has multiple re-strikes during aswitching operation, or at locations expected to havesurges higher than normal due to transferences, such asat a power generating station. For all these conditions,higher values of impulse levels, as in column 3 of Table11.6, may be chosen or surge suppressors installed.Test recommendations1 The test coils should be finally processed and thenembedded into the slots.2 The number of sample coils must be two unlesschanged deliberately.3 All coils that are subjected to this test must fulfil thetest requirements. In the case of a failure, investigationsmust be carried out to establish the cause and thesame rectified before undertaking the insulating processon the next batch of coils.Test procedure1 For inter-turn insulationThe test should be performed by applying the voltagebetween the two terminals of the sample coils.The inter-turn test voltage must be generated bydamped oscillating discharge of capacitors or withthe help of a high-voltage generator to pulse a fastfronteddamped oscillating voltage wave. In orderto obtain an even distribution of the impulse betweenthe coil turns, the front time of the first voltagepeak must not be below 0.5 ,us. The resultantwaveform produced by the coil, is displayed on astorage oscilloscope, and can be compared withthe waveform of a known good coil. Figures 11.4and 11.5 show a clear difference in the waveformsbetween a good and a defective coil.The voltage peaks between the terminals of thesample coils should be 50% of the value given incolumn 4 of Table 11.6, i.e. '/2(4Vr + 5) kV.2 For main insulation (ground and inter-turn insulationwithstand test)Any of the following methods can be adopted:Power frequency voltage test- The HV test should be performed first by applyingr.m.s. value (2Vr + 1) kV for one minute, as shownin Table 11.4 between coil terminals and ground.Then the applied voltage should be increased at1 kV/s up to 2(2Vr + 1) kV, and then reducedinstantly at least at 1 kV/s to zero. If there is nofailure, the test will be considered as successful,fulfilling the requirement of column 4, of Table11.6. This procedure would give a test voltageequivalent to 2 . 4 . (2 V, + 1) kV, which is evenmore than the values of column 4, i.e. > (4V, + 5)kV.- A d.c. test is also permitted by IEC 60034-15.The d.c. voltage should be at least 1.7 times thepeak of the power frequenc routine test voltage,i.e. a minimum of 1.7 x % 72(2Vr + 1)kV.Impulse voltage testThe full test voltage of column 4 of Table 11.6, Le.a standard lightning impulse of 1.2/50 ps, should15t lo2 5Ql0,s oPD2 -58T-1 0-1 50 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8Time (milliseconds) +Figure 11.4 Voltage traces of a satisfactory coil (withoutinterturn fault)T520log oB .PD2 -102T-20 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8Time (milliseconds)-Figure 11.5 Voltage traces of an unsatisfactory coil (withinterturn fault)be applied between the coil terminals and the ground.The number of impulses should be 5, unless agreedotherwise by the manufacturer and the user.11.4.10 Principal considerations in framing thespecification for impulse voltagewithstand level and underlying the testprocedures1 When a steep-fronted voltage surge occurs betweenone terminal of the machine and the ground thecorresponding phase does not instantly (during therise time t, Figure 17.3) adjust to the same potentialat all the points on the curve. Hence, two types ofvoltages develop in the machine winding, i.e.Voltage between copper and ground, which can becalled the transverse voltage, and the voltage alongthe copper which can be called as longitudinalvoltage.While the transverse voltage stresses the main wallinsulation, the longitudinal voltage stresses the interturninsulation. The bulk of the components of both

Philosophy of quality systems and testing of electrical machines 11/263kinds normally appear on the first or the entrancecoil of the winding.2 In practice, voltage surges can be of various shapes(Figure 17.2) and may even be so steep as to have afront time as low as 0.2 ps or less. For the purpose ofthe impulse test, however, only a standard lightningimpulse, as defined by 1.2/50 (Section 17.6) can beconsidered.3 The impulse level in column 4 of Table 11.6 has beenso chosen that the machine winding will have asufficiently high level of insulation to fit into the systemof insulation coordination, as discussed in Section17.11 and Table 11.6.4 The test on a sample coil at 50% test voltage willindirectly represent the test on the whole machine, inthat the sample coil is tested under almost the sameconditions to which the whole machine would havebeen subjected when applied with the full test voltageof column 4 of Table 11.6.5 Quantum of impulse voltage. There is no agreedcalculation to determine the severity of impulse thatmust be applied to these two sample entrance coils asthis varies from one machine to another and otherfactors such as:- Rise time tl (Figure 17.3) of the voltage impulse- Length of the entrance coil, and- Number of turns.As discussed in Section 17.8 the bulk of the voltageof a fast-rising impulse wave applied to the wholewinding will appear across the entrance turns. Thismay vary from 40% to 90%, depending on the steepnessof the wave front. Report TC-2 of IEC 60034-15 hasrecommended a value of 50% as adequate to meetgeneral requirements. However, this value may befinally decided by the manufacturer of the machinein consultation with the end user, based on the surgegeneratingsource (interrupting device), its likely fronttime, the type of machine and its exposure to externalsurge-generating sources.subtracting the primary copper loss (ZitRl) at thetemperature of the test from the input watts. If the valuesof current and power are recorded, from about 130%normal voltage downwards, and a graph of power againstvoltage plotted, the core loss can be separated, fromfriction and windage losses.Interception with the zero voltage axis, which representsfriction and windage losses, may be found by plotting asecond graph with the square of the voltage as the abscissaand the watts as the ordinate (Figure 11.6),11.5.1 Locked rotor testThis test is conducted by supplying the stator windingswith the rotor in the locked condition. In slip-ring motors,the rotor windings are also short-circuited. The test iscarried out to determine the soundness of the rotor insquirrel cage motors, and to measure the starting current,power factor, starting torque and impedance. It also enablesus to draw a circle diagram, for single squirrel cage rotormotors and wound rotor motors. This test may be carriedout at a reduced voltage that will produce the rated currentof the motor. The locked rotor torque test is not to beperformed on a wound rotor motor. The starting torquein a wound motor has no relevance, as it can be varied asdesired. The locked rotor current test is carried out onboth squirrel cage and wound rotor motors. It should berecognized that testing induction motors in the lockedrotor condition involves unusual mechanical stresses anda high rate of heating. Therefore, it is necessary that:The direction of rotation be established prior to thistest.The mechanical method of locking the rotors must bestrong enough to prevent injury to nearby personnel ordamage to equipment.As the windings are heated rapidly, the test voltagemust be applied as quickly as possible. Care should betaken to ensure that the motor temperature does not11.5 No load testThe no-load test is a very informative method to determinethe no-load current, core* and pulsationt losses, frictionand windage losses, magnetizing current and the noloadpower factor. The test also reveals mechanicalimbalance, if any, performance of the bearings, vibrationand noise level of the motor.The motor is run on no load at a rated frequency andvoltage, until the watts input becomes constant (to ensurethat the correct value of friction loss is obtained). Readingsof line voltage, current, frequency and power input aretaken.The watts input is the sum of the friction and windagelosses, core loss and no-load primary loss (I:,&). Thesum of friction, windage and core losses is obtained by*Core loss is the magnetizing or hysteresis loss and represents theiron loss of the machine.'Pulsation loss is the harmonic loss of the machine.Friction andwindage losses 4Figure 11.6 No-load curves to separate out no-load losses

11/264 Industrial Power Engineering and Applications Handbookexceed the permissible value for a given class ofinsulation.The readings at any point should be taken within 6 secondsfor motors of output 7.5 kW and below and 10 secondsfor motors above 7.5 kW.Measurement of starting torque, pull-out torqueand pull-in torqueAny of the methods described in Section 11.4.5 may beadopted to measure the torques developed. The torqueshould be measured with the rotor in various positions,wherever possible. The minimum value should be takenas the starting torque. Readings of voltage, current,frequency and power input will also be taken. The startingtorque and starting current may be extrapolated whenthe test is carried out at a reduced voltage. For extrapolationof the test results at rated voltage the test must beperformed at least at three test voltages. At each testvoltage, readings of voltage, current, torque, frequencyand power input must be taken. The values of startingcurrent and starting torque may then be extrapolated fromthese curves. The effect of magnetic saturation is notconsidered in this test method.Alternate methodWhen the locked rotor torque cannot be measured by thedynamometer or other methods it may be accuratelydetermined as follows:974Torque = - x 0.9 x (input in kW - Stator 1; R1 loss)NsThe factor 0.9 accounts for a 10% reduction in the torqueas an arbitrary allowance for harmonic losses.When the torque is determined by the above method,the voltage during the test should be so adjusted that thelocked rotor current is approximately equal to the fullload current. After the locked rotor test, the resistance ofthe stator windings should be measured and may also beconsidered for calculating the 12R losses.11.5.2 Open-circuit voltage ratio test for slipringmotorsIn slip-ring motors normal voltage and frequency areapplied to the stator with the connection of the rotorbrush gear open-circuited. The phase voltages inducedin the rotor are then measured across the slip-rings.11.5.3 Verification of degree of protectionWe have defined the various types of enclosures adoptedby various manufacturers to suit different locations andenvironmental conditions in Tables 1.10 and 1.11. Herewe briefly discuss methods for testing these enclosuresto check their compliance with defined requirements.Table 11.7 Protection against contact with live or moving partsFirst characteristic Test requirementsnumber0123456No test is requiredThe test is carried out with a sphere of SO mm diameter. The sphere should not touch live or moving parts inside theenclosure.(a) Finger testIn LT motors, a standard test finger as shown in Figure 11.7 is used, connected by an incandescent lamp to one poleof a supply of at least 40 V, the other pole of the supply being connected to the parts intended to be live in normalservice. All parts must be connected electrically. The lamp should not glow when an attempt is made to touch the barelive parts or insufficiently insulated parts. Insufficiently insulated parts may be covered with a metal foil connectedto those parts that are live in normal service. Conducting parts covered with varnish or enamel only or protected byoxidation or by a similar process may be considered as insufficiently insulated. In HT motors, the clearance is verifiedwith the minimum clearance required to withstand the dielectric test as in Section 11.4.8.(b) Sphere testThe enclosure should not permit a ball of 12 mm diameter to enter the enclosure.The test is carried out with a steel wire of 2.5 mm diameter. The wire should not go through the enclosure.The test is carried out with a steel wire of 1 mm diameter. The wire should not go through the enclosure.The test is carried out by using an apparatus shown in Figure 11.8, consisting of a closed test chamber in which talcumpowder can pass through a sieve having square openings of 75 pm, and is held in suspension by an air current. Theamount of talcum powder is supplied at 2 kg per cubic metre size of the test chamber. The enclosure under test isplaced inside the chamber and is connected to a vacuum pump which maintains, inside the enclosure, a differentialpressure equivalent to not more than a 200 mm column of water.The test is stopped at the end of two hours if the volume of the air drawn in during this period is from 80 to 120 timesthe volume of air in the enclosure under test.If, with a vacuum equivalent to a 200 mm column of water, it is not possible to draw air 80 times the volume of theenclosure under test, the test must be continued until the value is attained. In no case must the test be carried out formore than 8 hours.The permissible amount of talcum powder penetrating the enclosure should not affect operation of the equipment.The test is similar as for number S but now no deposit of dust should be observed at the end of the test.

Philosophy of quality systems and testing of electrical machines 11/265These tests, however, do not cover special requirementsor environmental conditions such asExplosive areas orUnusual service conditions that may cause corrosion,fungi, etc.These requirements may require a special construction,a pressurizing arrangement or more sealings at the jointsto prevent entry of dust or exit of an arc taking placeinside the enclosure, or special treatment to the housingand larger clearances or creepage distances etc. For moredetails see Sections 7.14, 7.15, 7.16 and 7.17. Here wehave limited our discussions to the testing of electricalequipment as in Tables 1.10 and 1.11. For other compliancetests, there may be an agreement between the manufacturerand the user or a third-party agency, as noted in Section1 1.7, for certifying the use of equipment for hazardousareas.Nore We have discussed the test procedures and tolerances in generalterms. For more accurate test methods and tolerances see IEC 60529.Tests for the first number as in Table 1.10 andIEC 60034-5: protection against contact with liveor moving partsThese tests may be carried out as shown in Table 11.7.CylindricalRadius = 2@@SectionSectionA-A 0-0All dimensions in mrn18080Figure 11.7Standard test fingerAir flow meterGlass windowl-l-?7uCirculating pumpL- - 1Figure 11.8 Apparatus for the verification of protection against dust, according to IEC 60034-5 or 60947-1

11/266 Industrial Power Engineering and Applications HandbookTests for the second number as in Table 1.11:protection against ingress of waterThese tests may be carried out as shown in Table 11.8.11.6 Tolerances in test resultsThe test results so obtained will be subject to a toleranceas noted in Table 11.9, compared to data provided by themanufacturer based on their design data.11.7 Certification of motors used inhazardous locationsMotors intended for such locations need special attention.Since such installations may be highly prone to explosionand fire hazards, third-party agencies are generallyappointed by the government of a country to certify theuse of particular equipment or device at such locations.These agencies, ensure that the equipment is designedand manufactured in conformity with the requirementsof the relevant standards. It is mandatory on the part ofa manufacturer, supplying equipment for such locations,to first obtain certification for a product from such agenciesbefore it can be used for such installations. For instance,the Central Mining Research Institute (CMRI), Dhanbad,is the authorized agency in India which undertakesScrutiny of design and constructional detailsThorough testing of the machine.Based on the test certificates of CMRI, approval certificatesare issued by the relevant statutory authority.Table 11.8 Protection against ingress of waterSecond characteristicnumber0145Test requirementsNo test i< requiredThe test is carried out by the apparatus illustrated in Figure 11.9. Water is used and adjusted so that the dischargeis 3 to 5 mm of water per minute.The enclosure under test is placed for IO minutes in its normal operating position below the dripping apparatus,the base of which should be larger than the enclosure under test.After the test the amount of water which might have entered the interior of the enclosure should not interfere withsatisfactory operation of the equipment. No water should accumulate near the cable gland plate or enter thecables.The test equipment is the same as described for degree of protection I. But the enclosure under test is tilted upto an angle of +15" in respect of its normal operating position successively, in two planes at right angles (to coverall four sides). The total duration of the test will be IO minutes (2.5 minutes each side). The test results shouldbe the same as for degree of protection 1,The test is carried out by the apparatus illustrated in Figure lI.lO(a). It consists of an oscillating tube, formed intoa semi-circle, the radius of which is kept as small as possible, depending upon the dimensions of the enclosureunder test, but not more than I m. For larger surfaces, a hand sprayer, as illustrated in Figure Il.lO(b), may beused. During the test the moving shield is not removed from the spray nozzle. The water pressure is adjusted fora delivery of IO litreshin. The test duration should be 1 minute/m? of the surface area under test, but for not lessthan 5 minutes. The tube is oscillated to describe an angle of 60" from the vertical in either direction. The durationof one oscillation will be about 2 seconds. The water supply should be at least IO litres/min, at a pressure equalto a head of nearly 8 m of water (80 kN/m*).The enclosure under test is mounted in its normal position on a turntable, the axis of which will be vertical andheight variable, located near the centre of the semi-circle formed by the oscillating tube. The table is rotated tospray all parts of the enclosure equally. The enclosure should be kept under a spray of water for IO minutes.The test results should be the same as for degree of protection 1.The test is similar to that described for degree of protection 3 except that the oscillating tube will now oscillatethrough an angle of almost 180" with respect to the vertical in both directions and at a speed of 90" per second.The support for the equipment under test may be grid-shaped, so that no water is accumulated at the base. Theduration of the test will be 10 minutes.For larger surfaces the second method as noted in Figure I 1. IO@) may be adopted but the moving shield mustnow.be removed from the spray nozzle. The rest of the details remain the same as noted for number 3, when usinga hand sprayer.The test results should be the same as for degree of protection I.The test is carried out by washing down the test enclosures in every direction by means of a standard hose nozzleof 6.3 mm inside diameter, as illustrated in Figure 1 1.1 I, held at 3 m from the enclosure with a water pressureequal to a head of nearly 3 m of water (= 30 kN/m2), enough to give a delivery rate of 12.5 litreshin. Theduration of the test will be determined at 1 min/m' of the surface area under test, subject to a minimum of 3minutes.The test results should be the same as for degree of protection 1,

~ ~~ ~Philosophy of quality systems and testing of electrical machines 11/267Second characteristicnumberTest requirements6 The test procedure and test equipment is almost the same as for number 5 and generally as below:- Inside diameter of the test nozzle: 12.5 mm- Distance of the nozzle from the test enclosure: = 3 m- Water pressure: almost 10 m of water (100 kN/m2)- The above water pressure will give a delivery rate of 100 litredmin.Test duration: at 1 min/m* of the surface area under test, subject to a minimum of 3 minutes. After the test, therewill be no penetration of water inside the enclosure.7The test is carried out by completely immersing the test enclosure in water so that the head of water above thelowest portion of the enclosure is a minimum 1 m, while the highest portion is a minimum 150 mm. Durationof the test will be 30 minutes.After the test there must be no penetration of water inside the enclosure.This test may also be carried out in the following manner.The enclosure must be tested for one minute, with an inside air pressure equal to a head of about 1 m of water.No air should leak during the test. Air leakage may be detected by submerging the enclosure in water, with thejust covering the enclosure.8 The test procedure will depend upon the actual application.I1II1I Reservoir ILayer of sand and gravel to regulate flowof water, separated by metallic gauge andblotting paper.301Detail-A(Enlarged sectional view)20support wEquipmentunder testrnAs per IEC-60034-5 or 60947-1Note The support should be smaller than the equipment under testI ! k !'I ;* 31 2 0A 2 04View-A All dimensions in mmFlgure 11.9 Apparatus for the verification of protection against dripping water

11/268 Industrial Power Engineering and Applications HandbookEquipmentunder testCountenveightAll dimensions in mmFor the second numeralSpraying angle aSurface covered a3 4As per IEC-60034-5 or 60947-1t 60" - * 180"2 60" = ? 180"Figure 11 .lO(a) Apparatus for splashing waterViewed through 'A withshield removedFigure 11.10(b) Hand-held sprayerAll dimensions in mm

Philosophy of quality systems and testing of electrical machines 11/269A47iC'All dimensions in mmD = 6.3 mm for the tests of second numeral 5D = 12.3 mm for the tests of second numeral 6Figure 11.1 1 Standard test nozzle for hose testsTable 11.9 Permissible tolerances in performance figures.s. !VO Itrr1r TolrmnceI231567x9IOIII2Efficiency(a) By summation of losses:Motors up to SO kWMotors above 50 kW(hl By input-output testTotal loses applicable to inotors ahove SO kWPower FactorSlip a[ full load and at working temperature(a) Machines having output I kW or more1 b) Machines having output less than 1 kWBreakaway starting current (for squirrel cage motors)Locked inotor torquePull-out torquePull-up (pull-in) torqueVihrarionNoise levelLocked rotor current of squirrel cage motors with shortcircuitedrotor and with any jpecified starting apparatusMoment of inertia or stored energy constant, applicableto motors of fraine sires ahove 3 15-15% of (I00 - rJ)-10% of I100 - rJ)-15% of (100 - r J)+ 10% of total losses-1/6 of (I - cos 4). Subject to a minimum of 0.02 and maximum0.07.? 20% of the guaranteed slipk 30% of the guaranteed slip+ 20% of the guaranteed starting current (no lower limit)-15% to + 25% of the guaranteed torque (+ 3% inay be exceededby agreement)-10% of the guaranteed torque. except that after allowing for thistolerance, the torque will not be lesh than 1.6 or I .5 times the ratedtorque-15% of the guaranteed torque+IO% of guaranteed classification+3 dB (A) over guaranteed value+20% of the guaranteed current(no lower limit)+IO% of the guaranteed valueRelevant Standards60034- 1/1996 Rotating electrical machine\ BS EN 60014-1/1995Rating and pcrformanccTemperature rise measuremcnt of rotating electrical I280211 989machine\60034-2/1996 Method for determining losses and efficiency of rotating 40291199 I BS 4999-102/1987electrical machines60034-51199 I Rotaring electrical machincs 469 I /I 985 BS 4999- 1051 I988Claasification of degrees of protection provided byenclo\ure\ for rotating machinery (IP Code)60034-6/1991 Rotating electrical machines. Methods of cooling (IC Code) 6362/1995 BS EN 60034-6/1994

j ,~11/270 Industrial Power Engineering and Applications HandbookIEC Title ISBSIS060034-9/199760034- 12/198060034- I4/199660034-l5/199560034- 171199860038/199460044-1/199660044-2/199760051-1 to 960060- 1/198960060-2/199460068-2-6/199560068-2-7/ 198360068-2-47/198260072- 1/199160072-2/199060072-3/ 199460079-0/ 199860079- 1/199060079-2/198360079-7/199060079-10/199560079.1 1/199160079- I2/197860079- 14/199660079- I5/ I98760079- 1711996Rotating electrical machines 12065/1987Noise limitsRotating electrical machines. Starting performance of 8789/1996single speed three-phase cage induction motors forvoltages up to and including 660 VRotating electrical machines 12075/1991Mechanical vibration of certain machines with shaft height56 mm and higherRotating electrical machines 14222/1995Impulse voltage withstand levels of rotating ax. machineswith form wound stator coilsGuide for application of cage induction motors when fedfrom convertersIEC Standard voltagesSpecification for current transformers 2705General requirementsPart-111992Measuring current transformers Part-2/ 1992Protective Current TransformersPart-3/1992Protective current transformers, for special purpose Part-4/1992applicationsApplication guide for voltage transformers 4201/199 I,Specification for voltage transformers 4146/1991General requirements 3156-111992Direct acting indicating analogue electrical measuring 1248 - 1 to 9instruments and their accessoriesHigh voltage test techniques, for voltages higher than I kV 2071-1/1993General definitions and test requirementsHigh voltage test techniques. Measuring systems 1207 1 -3/l991207 1 -2/199 1Environmental testing. Test Fc. Vibration (sinusoidal)Test Ga. Acceleration, steady stateEnvironmental testing. Tests. Mounting of components, 9000,equipment and other articles for dynamic tests including 9001shock (Ea). bumb fEbl vibration (Fc and Fd) and steadystateacceleration (Cia) and guidanceDimensions and output series for rotating electrical 1231 / I 99 Imachines. Frame number 56 to 400 and Flange number55 to 1080Dimensions and output series for rotating electrical 1231/1991machines. Frame number 355 to 1000 and Flangenumber 1 I80 to 2360Dimensions and output series for rotating electrical 996/1991machines. Small built in motors. Flange number BF 10to BF 50Electrical apparatus for explosive gas atmospheres. 2148/1993General requirementsConstruction and verification test of flameproof2 l48/1993enclosures of electrical apparatusElectrical apparatus. Type ‘p’.7389/1991Electrical apparatus for explosive gas atmospheres. 638 111991Increased safety motors. General requirementsElectrical apparatus for explosive gas atmospheres. 5572/1994Classification of hazardous areasSpecifications for intrinsically safe electrical system ‘i’ 3682/ 1991General recommendationsElectrical apparatus for explosive gas atmospheres.Electrical installations in hazardous areas (other thanmines)Electrical apparatus with type of protection ‘N’Inspection and maintenance of electrical installations inhazardous areas (other than mines)5571/19919628/19888289/1991BS EN 60034-9/1998 -BS EN 60034-12/1996BS 4999-142/1987 2373BS EN 60034-15BS 1626/1993BS 762611993BS 772911995BS 89-1 to 9 -BS 923-1/1990 -BS EN 60060-21997 -BS EN 60068-2-6/1996 -BS EN 60068-2-7/1993 -BS EN 60068-2-47/1993 -BS 5000-10/1989BS 4999- I4 I / I 987BS 5000-10/1989BS 4999-103/1987BS 5000-11/1989BS 5501-111977,BS 5000-17/1986BS 4683-2/1993BS 5501-5/1997BS EN 50016/1996BS 5501-6/1977BS EN 60079-10/1996BS 5501-7/1977, 5501- -9/1982BS 5345.111989 -BS EN 60079-14/1997 -BS 6941/1988BS 5000-16/1997BS EN 60079-17/1997 -

~Philosophy of quality systems and testing of electrical machines 11/271IEC Title IS BS IS06007')- I Y/ I993 Repair and overhaul for apparatus used in explosiveBS IEC 60079-19 -atmospheres (other than mines or rxplosives)60079-20/ 1Y9660 I Xh/l9X7Data for flammable ga\es and vapours for electricalappdrat USMeasuring voltage transformers 3156-211992Protcctive voltage tranhformersCapacitor voltage transformers3156-3119923156-4/1992Rotating machine.;. Specifications for tests -Guide for testing insulation resistance of rotatingmachines, rated for I MW and abovcElectrical Relays.7X I6/199 IVibration, shock, bump and seismic tests on measuring 9000relays and protection equipment. Vibration tests (sinusoidal)Shock and pump tests ~Seismic tests -Degree 01' protection provided by enclosures (IP code)Guide for test procedures for the measurement of losstangent of coils and bars for machine windings?147/1962Meahurement and evaluation of vibration severity in-sitrr 11727/1990of large rotating machinesQuality management and quality assurance vocabularyQuality rystein-guidelines for selection and use ofI3999/1994I-I000/19XXstandards on quality systemsQunlity systcms. Model for qualit urance in design. 14001/1994production and servicing etc.Qualily systems. Modcl for quality assurancc in design. 14002/1994installation and scrvicingQuality systems. Model for qiiality assurance in final 1500311994inspection testGuidcliner on qudity managemcnt and quality system 13004/1991elementsTcst code Tor the measurement of airborne noise emittedby rotating electrical mdchincsEngineering method fix free field conditions over areflecting planeSurvcy methodDetermination of sound power levels of noise sourcesRelevant US Standards ANSI/NEMA and IEEEBS 772911995BS 762511993BS 1999-143/1987-BS EN 60255-21-1/1996BS EN 60255-21-2/1996BS EN 60255-2 I -3/1996BS EN 60529/1992BS 3999-144/1987BS 7851-111996 I ox 1 6 1 /I 995BS EN IS0 8402/1995 8102/1994BS EN IS0 YOO0.3 piins 9000/19X7BS EN IS0 9002/1994 9002/1991BS EN IS0 9003/1994 9003/1993BS bN IS0 9004/1994-1119949004/IYX7BS 745R-1/1991 1680-1/1986BS 7458-2/19911680-21 I986BS 4196 3740ANSMEEE-I 18/1992ANSIIIEEE. 112/1993ANSIIIEEE-43/1992ANSIIIEEE- I I 7/1992IEEE- I 1071 IO06ANSI/IEEE-i22/1998IEEE-7YY1995IEEE 4lYY.iANSVIEEE 133/19Y 1ANSVIEEE 134/1992ANSI/IEEE-275/1 99XNEMA-MG 111993NEMA-MG VI990Test code I'or resistance measurement.Test proccdurcs I'or poly-phase induction motors.Kecommcnded practice ror. testing insulation resistance of rotating machincrq.Test procedure for. evaluation of system of insulating materials for random wound a.c. electrical machines.Recorninended practices for thermal evaluation of sealed insulating systeinr for i1.c. electrical ni;ichincry ctnployingrandom wound hutor coils.Recommended practice for Thermal evaluation of sealed insulation systems for a.c. electrical machinery- employing formwound pre-insulated stator coils for machines rated 6.9 kV and below.Guidc for testing turn to turn insulation on form wound stator coils for a.c. rotating machincs.Rccommcndcd practicc for the evaluation of the impulse voltage capability of insulation systems for a.c, clectricalmachinery emplriying form wound statnr coils.Stundwd techniqucs for HV testing.Practice for insulation testing. Large ax. rotating machines with high voltage at very low frequency.Guide for functional evaluation of insulation system Inr large HT machines.Recommended practice for, thermal evaluation of insulation systems for a.c. electrical machinery employing fnrm woundpre-insulated stator coils. for machines rated 6.9 kV and below.Motors and generators rating. construction. testing and perfnrmance.Sound IcvcI prediction for installed rotating electrical machines.

11/272 Industrial Power Engineering and Applications HandbookNotes1 In the tables of relevant Standards in this hook while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydiffcrcnt standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.23Some of the BS or IS standards mentioned against IEC may not be identicalThe year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.Further reading1. Bartheld R.G., Motor Noise.2. DOC-ETDC 15(3059), Recommendation on insulation resistancetest for 3.3 kV motors, 1000 kW and above, June (1987).3. Essler, H. andMohanT.R., ‘Noise inelectrical machines,’SiemensCircuit, IV No. 3, April (1969).4. Hogg, W.K., Miller, R., Rabech, G. and Ryder, D.M., ‘Therelationship of partial discharge amplitude distribution withelectric damage at different levels of voltage and frequency’,IEEE Symposium on Electrical Insulation, USA, June (1 984).5. IEEE, Results of an investigation on the over voltages due to avacuum circuit breaker when switching an H. V motor, IEEE 85SM 370-2(1985).6. IS1 DOC ETDC 15 (2729), Indian standard Institute DraftDocuments forElectroTechnica1 Division Council, for limits ofvibration.7. Renganian, S., Agrawal, MD. and Nema, R.S. ‘Behaviour ofhigh voltage machine insulation system in the presence of thermalinsulation and electrical stresses’, IEEE Trans. on EIEI 20, No.I, Feb. (1985).8. Rhudy, R.G. et al., ‘Impulse voltage strength of AC rotatingmachines’ IEEE Trans. PAS, 100, No. 8(1981).9. Secco, M., Bressani, M. and Razza, E, Medium Motor andGenerator Plant, Ansaldo Componenti SPA, Italy. Progress anddevelopment trends in large induction motor stator windinginsulation.10. Simons, J.S., ‘Diagnosis of H.V. machine insulation’, IEEEProceedings, 127, May (1980).

Protection ofElectric Motors

Protection of electric motors 12127512.1 PurposeAn electric motor must be adequately protected againstall unfavourable operating conditions and internal orexternal faults. We have classified these conditions intothree categories to identify the most suitable protection:1 Unfavourable operating conditions2 Fault conditions3 System disturbances and switching surges (for HTmotors)12.2 Unfavourable operatingconditionsOperating conditions that may overload a machine andraise its temperature beyond permissible limits may becalled unfavourable. This overheating, however, will begradual (exponential), unlike rapid (adiabatic) heatingas caused during a locked rotor condition. The machinenow follows its own thermal curve and therefore aconvcntional thcrmal protcction dcvicc can be used toprotect it from such conditions. These conditions mayarise due to one or more of the following:(i) Otzerloading Due to excessive mechanical loading.(ii) Undervoltage Low voltage results in forced overloadingdue to higher slip losses and higher currentinput to sustain the same load requirement. Anunstable sub-distribution network, a number of smallLT loads on a long and already overloaded LTdistribution system, or inadequate cable sizes maycause an excessive fall in the receiving end voltage.See also Section 23.3, where we analyse the effectof low power factor on the terminal voltage.The effect of small voltage drops (Section 1.6.2)is taken care of by the standard overcurrent protectionused in the motor’s switching circuit. But for installationswhere the voltage available at the motorterminals may fall below the permissible level, say,below 90% of the rated voltage, overheating or evenstalling may occur if the voltage falls below 85%.In such a condition, depending upon the severity ofthe voltage fluctuation and the load requirement, aseparate undervoltage relay may also be used.The ‘no-volt’ coil of the contactor or theundervoltage trip coil of the breaker, used for themotor switchings, are designed to pick up at 85% ofthe rated voltage. These coils drop out at a voltagebetween 35% and 65% of the rated voltage and wouldnot protect the motor against undervoltages. In normalservice con-ditions the system voltage is not likelyto fall to such a low level, particularly during running.Thus, the protection will not prevent the closingof the contactor or the breaker when the supplyvoltage is *85% or more, nor will it trip the motoruntil the voltage falls to a low of *65% of the ratedvoltage. In both cases, therefore, separateundervoltage protection will be essential. Thisproblem, however, is a theoretical one, as an industrialpower system would seldom fluctuate so widely.Note Sometimes when there are perennial wide voltagefluctuations at certain locations/installations the manufacturersof the contactors on demand from users may design theirholding coils for even lower pick up and higher drop outvoltages than noted to save the feeders from unwanted trips,the user making extra capacity provision in the motor or gettingit redesigned for special voltages to surtain the wide voltagefluctuations.(iii) Reverse rotation This may occur due to a wrongphase sequence. While the motors are suitable foreither direction of rotation, the load may be suitablefor one direction only and hence the necessity forthis protection. A reverse rotation means a reverserotating field and is prevented by a negative phasesequence, i.e. a voltage unbalance or single-phasingprotection. Moreover, this protection is also of littlesignificance, as once the motor is commissionedwith the required direction of rotation, it is rare thatthe sequence of the power supply would reverse.(iv) Protection from harmonic effects Motors areinfluenced less with the presence of harmonics. Thisis due to the benign effects of harmonics on inductiveloads, on the one hand, and the motor providing nopath to the third harmonic quantities on the other, asit is normally connected in delta. In HT motors,however, which are normally star connected, theneutral may be left floating to provide no path to thethird harmonics.Higher harmonics increase the harmonic reactanceand have a dampening effect (Section 23.5.2(B)). Amotor circuit, LT or HT, possesses a high inductiveimpedance due to interconnecting cables and its owninductance, and provides a self-dampening effect tothe system’s harmonics. There is thus no need,generally, to provide protection against harmonicsspecifically, except for high no-load iron losses.If, however, high contents of harmonics exist, aswhen the machine is being fed through a static powerinverter (Section 6.13), they will produce magneticfields, rotating in space, proportional to the individualharmonic frequencies. These fields may be clockwiseor anti-clockwise, depending upon whether theharmonic is positive or negative. The fields producedifferent torques, which may also be clockwise oranti-clockwise. The net effect of all this is a pulsatingtorque. In a six-pulse thyristor circuit, for instance,the harmonic disorder is -5, +7, -1 I, +I 3, -17 and+I9 etc. giving rise to clockwise and anti-clockwisefields.To reduce the no-load iron losses caused by suchharmonics the machine core may be formed of thinnerlow-loss laminates (see also Section 1.6.2(A-iv)).When the machine has already been manufacturedand there is a need to suppress these harmonics,filter circuits may be employed along the linesdiscussed in Section 23.9.Excessive harmonics may also make the protectivedevices behave erratically or render them inoperative.Filter circuits would suppress the harmonics andeliminate these effects.Nore The protective devices which meature r.m.s. values ofcurrent also detect the harmonic contents and provide automaticprotection for the machine against harmonic disorder\. But in

12/276 Industrial Power Engineering and Applications Handbookinstallations that contain high harmonics and that may influencethe performance of machines operating on such a system, it isadvisable to suppress the harmonics rather than allow a machineto trip at higher harmonics. Drives generating high harmonics,such as static drives, are provided with harmonic suppressorsas normal practice.(v) Voltage unbalance (negative phase sequence) Thiscauses negative sequence components and resultsin excessive heating of the motor windings. Anunbalanced voltage may occur due to unevenlydistributed single-phase loads.A voltage unbalance may also be due to unequalphase impedances of the feeding HT line and maybe a result of unequal spacing between the horizontaland vertical formation of conductors or asymmetricalconductor spacings. These effects cause an unequalinduced magnetic field and hence an unequalimpedance of each conductor. The implication ofsuch a system can be studied by assuming it to becomposed of two balanced systems (Figure 20.1),one positive and the other negative (the negativesystem having the reverse rotating field and tendingto rotate the rotor in the reverse direction). Eachfield produces a balanced current system, the phasorsum of which will decide the actual current the motorwindings will draw. The main effect of a negativesequence current is thus to increase iron losses andreduce the output of the motor. In such a condition,as the current drawn by the motor increases, thetorque and the power developed reduce, and themotor operates at a higher slip. All these losses appearin the rotor circuit as slip losses. The rotor operatesat a higher slip and becomes relatively more heatedthan the stator and more vulnerable to damage.According to IEC 60034- 1, a negative sequencecomponent up to 1% of the positive sequencecomponent of the system voltage, over a long periodor 1.5% for a short period, not exceeding a fewminutes, and with the voltage of the zero sequencecomponent not exceeding 1 %, may be considered abalanced system.A motor is able to sustain a negative sequencevoltage up to 2% for short durations, while a sustainedunbalance may deteriorate its insulation and affectits life. For higher unbalances, a derating of themotor output must be applied according to IEC 60892or to Figure 12.1.A voltage unbalance may affect motor performancein the following ways:- It would cause overheating and the effective outputmav have to be reduced to avoid this. SeeFigure 12.1.- It would reduce T,,, T,,, and T, etc. by 0~ Vu2 ( Vu= unbalanced voltagd. An unbalance up to 1%will not affect motor performance below thepermissible limits. See also MG-1-14.34 and Figure12.2. The effect of unbalance on torque may,however, be considered insignificant. For example,even at an unbalance of 5%, the negative torquewill be = (0.05)* x T, or 0.25% of Tr, which isinsignificant.- Due to smaller output and torque, the slip wouldrise and add to the rotor's losses (mainly ironlosses) (see Figure 12.3). The voltage unbalancecan be calculated as follows:Voltage unbalanceMax. voltage variation from- the average voltage-Average voltagex 100% (12.1)Norr The negative sequence voltage is caused by an unbalancein the magnitude of voltages in the three phases, rather than inthe phase angle.Example 12.1Consider three line voltages 390 V, 400 V and 416 V. Then390+400+416the average voltage =3= 402 Vand the maximum variation from the average voltage:.= 416 - 402, Le. 14 Vunbalance = 14 x 100 = 3.48%4021toC.- cEQ OmeP -TrNegative torque-0.I I0 1 2 3 4 5% Voltage unbalanceFigure 12.1Derating in motor output due to voltage unbalanceFigure 12.2torqueEffect of negative sequence voltage on motor

~~~~~~~~Protection of electric motors 12/2773002751 250- 8 225&8 ,$go,-U8 1751561.’ 5:.148125-% Negafwe sequence current1000 10 20 i530 4o ’ 50 60 70@ Maximum hot - phase heating (positive and negativesequence currents in phase)@ Minimum hot - phase heating@ Average stator heating per phase@ Rotor heatingFigure 12.3systemHeating effect caused by an unbalanced voltageNote A supply system having a voltage unbalance of morethan 5% is not recommended for an industrial application,which may have a number of electric motors connected onit. Rural distribution, however, is an exception due toexcessive LT loads on the same network (Section 7.6), butsuch loads are mostly individual and not of the industrialtype.Stator currentsThe machine offers very low impedance to negativesequence voltages. As a result, the percentage increasein the stator current is almost the same as the startingcurrent on DOL switching. i.e. six to ten times the ratedcurrent.To understand this, consider the equivalent motor circuitdiagram of Figure 1.15 with a normal positive sequencevoltage Vr. The unbalanced voltage V, will produce anegative phase sequence and rotate the magnetizing fieldin the opposite direction at almost twice the supplyfrequency. The frequency of the negative phase sequencevoltage and current in the rotor circuit will thus become(2 - S) xfand the slip (2 - S). In such a condition, theequivalent motor circuit diagram (Figure 12.4(a)) willassume the impedance parameters. as shown in Figure12.4(b). Thus1. =vr~ ~ - _ _ ~ _ ~ ~ ~(1 2.2)Data at slip SFigure 12.4(a) Equivalent circuit diagram with a balancedsupply voltage (at slip S)/,V, = Component of unbalanced voltageFigure 12.4(b) Equivalent circuit diagram with an unbalancedsupply voltage (at slip = 2 - S)andI,, = .(12.3)I, = negative phase sequence current component(a) If the current during start (when S = 1) = I,,Then I,, in the former case,vrI,, = ___\(R, +RI)’ +(XI +SAX;)’Since (Xi + FFX; ) >> (R, + Ri )... v,= (Xi + ssx; )(b) Now consider the current I, during normal runningspeed, Le. when s = 0. Then in the latter case,I,, -- __- ~~,/[R, +Ri -Ri/2]’VUc:+(Xi +2ssXi)?Applying the same rule of approximation as above,since (X, + 2ssX? ) >> (R, + R;/2):. I, =and this relationship produces almost the same amountof current as during a start.CorollaryV”(Xi + 2 ssx; )Impedance of a motor during a normal running conditionto a negative phase sequence voltage will be almost the

12/278 Industrial Power Engineering and Applications Handbooksame as that of the impedance of a motor during a start,in a balanced supply system. Thus, the current effectduring normal running of a negative sequence voltagewill be the same as that of the starting current effect ona balanced supply system.The effect of voltage unbalance is, therefore, morepronounced than the percentage of unbalance itself. Forinstance, a voltage unbalance of 3% may cause a currentunbalance of 18-30%, which is detrimental to the life ofthe motor. The effective current in the stator windingswould depend upon the relative positions of the positiveand the negative sequence components. In one of thewindings they may be in phase, producing the maximumcurrent and the associated heating effect, and in the othertwo they may be 120" apart. In Figure 12.5 we havedrawn the maximum and the minimum effective currentswhich the stator windings may experience on an unbalancedsupply system.Stator maximum heatThis occurs when the positive and negative sequencecomponents fall in phase in which case the equivalentstator current will become- -I,, (max.) = I, + I,and the maximum heat generated, Heq(max.) ..(I,(Figure 12.3, curve 1)+ IJ2or H,,(max.) 0~ (I,? + I,' + 21, x I, ) (12.4)Example 12.2Consider a negative sequence component of 40% of therated current. Then the maximum heat generated as in equation(1 2.4)Heq (max.) = (1' + 0.4' + 2 x 1 x 0.4)ori.e. 1.96 or 196%= (1 + 0.16 + 0.8)Stator minimum heatThis occurs in the other two windings, which are 120"phase apart (Figure 12.5). HenceI,,(min.) = $I: + I: + 21, x I, cos 60"= J7-cand the minimum heat generated,H,, (min.) = ( I,' + 1,2 + I, x I,(Figure 12.3, curve 2) (12.5)Example 12.3The minimum heat generated in the above caseHeq (min.) = (1' + 0.4' + 1 x 0.4)or = (1 + 0.16 + 0.4)i.e. 1.56 or 156%Stator average heatIn practice the temperature attained by the stator windingswill be significantly below the maximum or even theminimum heats as determined above due to the heat sinkeffect. The heat will flow from the hotter phase to thecooler phasedarea (see curve 3 of Figure 12.3). But forprotection of the motor windings against negative sequencecomponents, the average heat curve is of no relevance,for it will take a considerable time for the heat to stabilizeat this curve, and much more than the thermal withstandcapacity of the winding most affected.It is therefore essential to provide adequate protectionfor the motor to disconnect it from the mains quicklybefore any damage is caused to the most affected winding.The protection is based on the maximum heat that maybe generated in the motor windings in the event of anegative sequence component in the system.Magnitudes of negative sequence componentsfor protectionThe additional heat generated by a negative sequencecomponent may vary from six to ten times the theoreticalheat produced by that amount of a positive sequencecomponent, as analysed above. It is therefore more relevantto consider a factor of 6 to represent the effect of such acomponent. The heat generated can be rewritten asHe, = (I,? +6Z,') (12.6)and the current on unbalance,Figure 12.5Equivalent stator currents during unbalanced voltageThe factor 6 is a design parameter for all future reference.In fact, this empirical factor has been established overmany years of experience and field data collected on thebehaviour and performance of a motor in such anunfavourable operating condition.

Protection of electric motors 12/279The heat derived from this equation may be less thanthe minimum heat (equation (12.5)) or even more thanthe maximurn heat (equation (1 2.4)) depending upon these\perity and the phase disposition of the negativecomponent with reference to the positive component.This can be illustrated by the following examples.Example 12.4Referring to Examples 12.2 and 12.3 above, the heat producedaccording to the empirical formula is as follows:Hes = (1' + 6 x 0.4')1.e. 1.96 or 196%which is the same as that derived in Example 12.2.i.e. 2.5 or 250%which is even more than the maximum heat obtained fromequation (12.4). Equation (12.6) is thus more appropriate fora protective device and reflects the effect of a negativesequence component in a motor winding more precisely.Rotor powerThe negative sequence voltage sets up a reverse rotatingfield and the slip of the rotor becomes '2 - S', comparedto the positive sequence slip S. The motor will thus operateunder the cumulative influence of these two slips, wherepower output P can be expressed by (see also Section2.3).CorollaryOne can thus easily obtain the significance of the factor 6 torepresent the status of the most affected winding of the motorin the event of a voltage unbalance resulting in a negativesequence current component. For more clarity, considerequations (12.4) and (12.6) to ascertain the similarity in boththese equations. Since both must represent the maximumheating effect:. I: + 6I: = I: + I,' + 2. I, x I,or 51,' = 21, x I"orI, = 215 x I,, Le. 0.4 I,Thus these two methods will yield the same result for a negativesequence component of 40%. If the negative sequence currentI, is lower than 40%, the heat produced from equation (12.6)will be lower than the minimum heat obtained from equation(12.5). In contrast, for a negative sequence component ofmore than 40%, the situation is likely to be reversed, sincethe heat produced as in equation (12.6) will be higher thanthe maximum heat produced by equation (12.4). See thefollowing example for more clarity.whereI,, = positive sequence current in the rotor circuit, andI,, = negative sequence current in the rotor circuitRotor heatThe unbalanced voltage will produce an additional rotorcurrent at nearly twice the supply frequency. For example.for a 2% slip, i.e. a slip of 1 Hz, the negative sequencestator current, due to an unhalanced supply voltage, willinduce a rotor current at a frequency of (2f- 1) = 99 Hzfor a 50 Hz system. These high-frequency currents willproduce significant skin effects in the rotor bars andcause high eddy current and hysteresis losses (Section1.6.2(A-iv)). Total rotor heat may be represented by=(I; + 3;") (12.9)(refer to curve 4 of Figure 12.3) and cumulative rotorcurrentExample 12.5Consider a negative sequence component of 15% and 50%respectively.(a) For a 15°/0 negative sequence component:From equation (12.5)He,(min.) = (1'+ 0.15' + I x 0.15)i.e. 1.1725 or 117.25%and from equation (12.6)He, = (1' + 6 x 0.15')i.e. 1.135 or 113.5%which is even less than the minimum heat obtained fromequation (12.5).(b) For a 50% negative sequence component:From equation (12.4)He,(max.) = (1' + 0.50' + 2 x 1 x 0.5)i.e. 2.25 or 225%and from equation (12.6)Heq (1' + 6 x 0.5')Example 12.6For example 12.2, the rotor heat at 40% stator negativesequence current= (1' + 3 x 0.4')i.e. 1.48 or 148%Refer to curve 4 of Figure 12.3.12.3 Fault conditionsThese are conditions in which overheating of the machinemay not trace back to its own thermal curves as in thefirst case. The temperature rise may now be adiabatic(linear) and not exponential and hence rapid. Now anormal thermal protection device may not be able torespond as in the previous case. Some conditions causingoverheating may not recessarily be fault conditions.Nevertheless, they may require fast tripping, and henceare classified in this category for more clarity. Suchconditions may be one or more of the following:

12/280 industrial Power Engineering and Applications Handbook1 A fault condition, such as a short-circuit between phases.2 A ground fault condition.3 A prolonged starting time.4 A stalling or locked rotor condition: An undervoltage,an excessive load torque or a mechanical jamming ofthe driven equipment may cause this. It may lock therotor during a start, due to an inadequate starting torque.Such a situation is known as stalling and results in anear-locked rotor condition (see Figure 12.6). Atreduced voltage start the motor will stall at speed N,and will not pick up beyond this. If the motor wasalready running and the motor torque takes the shapeof curve B due to a voltage fluctuation, the motormay not stall but may operate at a higher slip S2.Although, this is not a stalling condition it may causesevere overloading. At high slips, the current I, tracesback the starting current curve as on DOL as illustrated,and may assume a very high value.In certain cases, for example in large motors whereTpo is normally not very high, during a severe drop involtage Tpo may fall to a value less than the loadrequirement and result in a stalling or even a lockedrotor condition (see Figure 12.7).5 Frequent starts: It is not a fault condition, but rapidheating of motor's stator and rotor due to frequentstarts will be no less severe than a fault condition,hence considered in this category.6 Protection against single phasing.In all these conditions protection against 2, 3, 4 and 5 isgenerally applicable to large LT motors, say 100 h.p. andabove, and all HT motors. This protection is normallyprovided by a single-device motor protection relay,discussed in Section 12.5.Reduced voltagestarting or a lowvoltage conditionla = Current during stalling@ Motor torque@ Reduced voltage torque = Vzx/ \\Speed-Nrz N,lStalling -SI@ s,condition@ Load torque@ CurrentFigure 12.6 Stalling condition during start and runI,* tg0I athigherslip1,Single phasingKl-SpeedFigure 12.7 Stalling condition during runThis is a condition of a severe unbalance. Until the 1970sthis had been the most frequent cause of motor failureduring operation. About 80% of installed small andmedium-sized motors, say, up to 100 h.p. experiencedburning due to single phasing because of the absence ofadequate single-phasing protection. With the introductionof single-phasing protection in the 1970s, as a built-infeature with thermal overcurrent relays (OCRs), this causeof motor failure has significantly diminished in all thelater installations.The following are the possible causes of single phasing:Immobilization of one of the phases during operationby the melting of a faulty joint such as poor cableterminationBlowing of one of the fuses during a start or a runDefective contacts orA cable fault, immobilizing one of the phases.Effects of single phasing1 The Tpo of an induction motor, say, up to 100 h.p., isnormally more than 150-200% of T,. Therefore whenthe motor is operating at only one half to two thirdsof T, and experiences single phasing during the run, itmay still be running without stalling, although at ahigher slip. It will now be subject to a rapid burnoutwithout adequate protection. The motor will now drawmuch higher currents in the healthy phases, to supplementthe lost phase, as well as to compensate forthe higher slip losses. It may even stall if it wasoperating at more than 70% of T, at the time of singlephasing.2 If the motor is switched on, in a single phasingcondition, it will not rotate in the absence of a rotatingfield, similar to a single-phase motor without a startwinding.3 If the motor stalls during pick-up, it will come to astandstill as a result of a locked rotor. The motor will

Protection of electric motors 12/2814now be vulnerable to rapid burnout as the heat will belocalized and rapid in one part of the rotor and maydamage it without appreciably raising its overalltemperature.Since single phasing is a condition of severe unbalance,it causes varying proportions of dangerous currentsin the motor windings, as discussed below.Heat generated in a star-connected stator windingIn the event of single phasing, in star-connected windings,one of the phases of each positive and negative sequencecomponents will counter-balance each other, to producezero current in the open phase (Figure 12.8(a)). Themagnitude of these components in the two healthy phasesare equal, i.e. I, = I", and the maximum heat of the statoras in equation (1 2.4)=(I,? +I: +2xI, X I , )i.e = 41;/ Yk = 1,(a) Winding diagramand the maximum current Zeq = 21,The minimum heat as in equation (12.5)i.e=(I: +I: +I, XI,)=3I:and the minimum current Ieq12.8(b).Heat generated in a delta-connectedstator winding= &Ir. See FigureIn delta-connected windings we cannot derive the phasecurrents during single phasing by the above simplehypotheses. Now two of the phase windings are connectedin series, while the third forms the other path of theparallel circuit, as illustrated in Figure 12.9. The currentsin the immobilized phase as well as the healthy supplylines now vary disproportionately with load, the increasein the lone winding being more pronounced. A generalidea of the magnitude of phase and line currents soproduced is given in the curves in Figure 12.10 bymeasuring these currents by creating a single-phasingcondition.Heat generated in the rotorReferring to equation (12.9) the rotor heat can be expressedby0~ (I; + 31; )In single phasing, I, = Im, the rotor heat will be= (I; + 31;)or = 41;R Y BF1-3+veLineopenedOCR iLMotor winding in deltaI, = I,, lb = 0Figure 12.8 Stator and line currents on single phasing for astar-connected windingFigure 12.9 Stator and line currents on single phasing for adelta-connected winding

12/282 Industrial Power Engineering and Applications Handbook1% Motor current under healthy condition -Figure 12.10 Current magnitudes in different phases of adelta-connected stator winding and rotor during single phasingi.e. four times the normal heat of the rotor, and the rotorcurrent = 21,If single phasing occurs at 50% of the full load,the rotor current will be 100%. See curve 4 of Figure12.10.Conclusion1 Star-connected stators(a) The theoretical heats of the motor as derived earlierfor star-connected stators and rotors are almostthe same. But even if single phasing does nutlead to a stalled or a locked rotor condition, itmay cause the motor to operate at a much higherslip due to a lower torque, T,. The rotor is thereforesubjected to a faster temperature rise comparedto the stator due to excessive eddy current lossesand its smaller volume compared to that of thestator. In single phasing, therefore, although thestator and rotor current curves may appear similar,their relative heats will be substantially different.The rotor in such motors is thus more critical andmust be protected specifically.(b) Three-phase LT motors are built in delta connectionexcept for very small sizes, say, up to 1 or 2 h.p.,but all HT motors are generally wound in starconnection (see also Section 4.2.1). Generally, inLT motors the stator and in HT motors the rotorare more vulnerable to a fast burnout in the eventof single phasing.2 Delta-connected stators(a) In delta-connected windings, the lone winding X,(Figure 12.9) for motor loads 50% and abovecarries a current higher than the rated full loadcurrent and also higher than that of the rotor, andbecomes more vulnerable to damage comparedto the rotor. This difference is more significant atloads closer to the rated load (Figure 12.10).(b) A study of Figure 12.10 will suggest that in theevent of single phasing, protection should be suchthat it traverses the replica of the heating curve ofphase X .(c) It also suggests that the heat generated in the rotorcircuit, due to voltage unbalance or single phasing,is less than the maximum stator heat. The factorof 6 considered in equation (1 2.6) is adequate totake account of it.(d) For loads of less than 50% of the rated motorcurrent, this protection is not required as the currentin phase X will not exceed 100% of the full loadcurrent during a single phasing.CorollarySince an inter-turn fault also causes unbalance, it isprotected automatically when a negative sequenceprotection is provided depending upon its sensitivity andthe setting.3 Rotor circuit(a) In a rotor circuit the rotor current and the heatingeffects, due to single phasing, remain the samefor both star- or delta-connected stators due tothe same (2f- S) = 100 Hz rotor currents on a 50Hz system.(b) A stator thermal withstand curve cannot beconsidered a true reflection of the rotor thermalconditions. In a delta-connected motor (mostlyLT motors), the stator would heat-up more rapidlythan the rotor, and normally protection of the statormay also be regarded as protection for the rotor.But This is not so in the following cases:Prolonged starting timeStalling or a locked rotor conditionFrequent starts, andAll HT motors wound in star.In all the above conditions, the rotor would heatupmuch more rapidly than the stator due to itslow thermal time constant (r), and its smallervolume compared to that of the stator, on the onehand, and high-frequency eddy current losses athigh slips, due to the skin effect, on the other.True motor protection will therefore requireseparate protection of the rotor. Since it is notpossible to monitor the rotor’s temperature, itsprotection is provided through the stator only.Separate protection is therefore recommendedthrough the stator against these conditions for largeLT and all HT motors.

Protection of electric motors 12/283Note1 It is for this reason that rotors, as standard practice, aredesigned to withstand a much higher temperature of theorder of 400450°C in LT motors and 300-350°C in HTmotors, compared to a too-low temperature of the stator.This temperature is such that for almost all motor operatingconditions meticulous protection of the stator would alsoorotect the rotor. It is also observed that rotor failures aretherefore rare compared to stator failures.2 Nevertheless, whenever the rotor is more critical, despite ahigher rotor operating temperature, rotor thermal curvesare provided by the manufacturer for facilitating protectionfor the rotor also through the stator.3 The rotor design, its cooling system or the motor size itselfmay have to be changed substantially for motors to beinstalled in fire hazard environments to limit their temperaturerise in adverse operating conditions within safelimits (Table 7.6).12.412.4.1ProtectionProtection against unfavourable operatingconditionsProtective devices and their selectionWe will now discuss protective devices, and their selection,that will be essential to safeguard a motor against allunfavourable operating and fault conditions. A machinemay be provided with a modest or elaborate protectivescheme, depending upon its size, application and voltagerating. This enables savings on cost, where possible, andprovides a more elaborate protection where more safetyis necessary. Accordingly we have sub-divided theprotection as follows:1 Protection of small and medium-sized LT motors, upto 300 h.p.2 Protection of large LT motors, say, 300 h.p. and above.This is to be decided by the user, based on the loadrequirement and critical nature of the drive3 Protection of HT motorsSmall and medium-sized LT motorsProtection against overloadingThis can be achieved by an overcurrent relay. The basicrequirement of this relay is its selectivity and ability todiscriminate between normal and abnormal operatingconditions. Three types of such relays are in use: thermal,electromagnetic and static. Thermal relays are employedfor motors of up to medium size and electromagneticand static relays for large LT and all HT motors, asdiscussed in Section 12.5.The thermal relays in general use are of two types, i.e.Bimetallic, and0 Fusible alloy or eutectic alloyBecause of their spread between hot and coldcharacteristics, these relays allow a tripping time of lessthan the starting time when a hot motor stalls, so a separatestalling protection is normally not necessary. They detectthe r.m.s. value of the current and thus account for theeffects of harmonics, present in the current, drawn bythe motor. They also take into account the heating, dueto previous running of the motor as they are also heatedalong with the motor. This feature is known as thermalmemory. These relays thus possess tripping characteristicsalmost matching the thermal withstand capacity of themotor.Figure 12.11Thermal overcurrent relay (Courtesy: L&T)

12/284 industrial Power Engineering and Applications HandbookBi-metallic" thermal relays (Figure 12.11)These have three heaters in series with the circuit. Oneor more bi-metallic strips are mounted above these heaters,which act as latches for the tripping mechanism or togive an alarm signal if desired. The heaters may be heateddirectly for small motors or through current transformers(CTs) for medium-sized motors. Bending of the bimetallicstrips by heating, pushes a common trip bar in the directionof tripping to actuate a micro-switch to trip the relay orcontactor. The rate of heating determines the rate ofmovement and hence the tripping time, and provides aninverse time characteristic. The power consumption ofthe bimetal heating strips varies from 2 to 2.5 watts/phase, i.e. a total of nearly 7.5 watts.The latest practice of manufacturers is to introducea very sensitive differential system in the trippingmechanism to achieve protection even against singlephasingand severe voltage unbalances. In the relays withsingle-phasing protection a double-slide mechanism isprovided. Under single phasing or a severe voltageunbalance, the two slides of the relay undergo a differentialdeflection. One slide senses the movement of the bimetalthat has deflected to the maximum. while the othersenses the minimum. These slides are linked so that thecumulative effect of their movement actuates a microswitchto trip the relay. Figure 12.12 illustrates the trippingmechanism of an overcurrent-cum-single-phasing thermalrelay. Because of differential movement it possesses dualcharacteristics, as shown in Figures 12.13(a) and (b),one for an ordinary overcurrent protection during threephasenormal operation (Figure 12.13(a)) and the otherwith differential movement for overcurrent protectionduring a single phasing or severe unbalance (Figure12.13(b)). For instance, for a setting at the rated current(100% I,) under normal conditions the relay would stayinoperative (Figure 12.13(a)), while during a singlephasing it will actuate in about 200 seconds (Figure12.13(b)) and provide positive protection against singlephasing.@ During cold stateI@ During symmetrical 3 phase overloada Movement of lever duringan overloadingIII ; - RMicroswitchCharacteristics of a bi-metallic thermal relayThe thermal characteristics are almost the same as thoseof an induction motor. This makes them suitablc forprotecting a motor by making a judicious choice of theright range for the required duty. (See Figure 12.11 for atypical thermal overload relay and Figure 12.13 for itsthermal characteristics.) Ambient temperature com-*Any bi-metal combination, having large differences in theircoefficients of linear expansion, such as a bimetal of brass andsteel is used for such applications. One end of a strip is fixed andthe other is left free for natural movement. When heated, brassexpands more than the steel and bends towards the steel as shown,giving the desired movement to actuate a tripping lever.Fixed/ at 1 1 Movement on heatingone endSteelBrass@ Movement of lever during a singlephasing in phase RTripping action of the relayFigure 12.12 Tripping mechanism of an overcurrent-cumsingle-phasingthermal relaypensation is achieved through an additional strip in theoverload relay, which operates the tripping lever in theother direction than the main relay to achieve a differentialeffect and is so arranged that it is independent of themain relay.Operation of the relay may not necessarily start at thepreset value due to certain allowable tolerances. As inIEC 60947-4-1, the relay must not trip within two hoursat 105% of FLC but it must trip within the next two hourswhen the current rises to 120% of FLC. Also, it shouldtrip in two hours in the event of single phasing when theline current in the healthy phases is 115%, but it shouldnot trip in less than two hours during a healthy condition,

Protection of electric motors 12/285200 se(a) Tripping under a healthy condition.Figure 12.13 Typical characteristics of a thermal overcurrent relaywhen two of the phases carry loo%, while the third carries90% of FLC (a case of voltage unbalance). The curves ofFigure 12.10 illustrate the likely operating currents indifferent phases of a delta-wound motor on single phasingor voltage unbalance. A good thermal relay should beable to detect these operating conditions and provide therequired protection. The thermal curve of a relay is thusin the form of a band as shown in Figure 12.14.With the introduction of single-phasing detection andprotection feature in the conventional thermal relays thetripping current-time (Z2 versus t) characteristics of therelay traverse almost the same thermal curve as may beprevailing in the most vulnerable phase of the motorwinding during a single phasing, i.e. according to curve‘X’ of Figure 12.10. The characteristic curve of the relayis chosen so that it falls just below the motor thermalcurve and has an adequate band formation, somewhatsimilar to the curves of Figure 12.14.Relays for heavy dutySuch relays may be required for motors driving heavydutyloads with large inertias or for motors that employreduced voltage starting and require longer to accelerate.Consequently, a relay which can allow this prolongedstarting period without causing a trip during the startwill be desirable. CT-operated relays can be used forsuch duties. They comprise three saturated currenttransformers (CTs) associated with the ordinary bi-metalovercurrent relay (Figure 12.15). These saturated currenttransformers linearly transform the motor line or phasecurrents up to a maximum of twice the CT primary current.Above this ratio, the cores of the CTs become saturatedand prevent the secondary circuit reflecting the startingcurrent in the primary and thus prevent the relay fromtripping during a permissible prolonged start. For example,a CT of 150/5A will have a saturation at approximately300A, irrespective of the magnitude of the starting current.For schematics of such relays refer to Figure 13.54.Overcurrent setting of relaysThese can be adjusted by varying the contact traverse.The mechanism’s design is such that an increase or adifference in the line currents, due to voltage unbalanceor single phasing, drives the mechanism towards thetripping lever. These relays operate at 100% of theirsetting and are therefore set atRelay setting (% of FLC)(Operating current %) x I,- (typical)(CT ratio) x (Relay rating)

la286 Industrial Power Engineering and Applications Handbookcharacteristics as those of the motor, it is likely that the bi-metalmay cool faster than the motor and allow the motor to restart,without allowing adequate time to cool. In fact, the motor windingshave a considerably higher thermal time constant than the bimetalrelays and cool more slowly than the relays. Whereas the relay willpermit rapid repeated switchings, these may not be warranted. Handresetrelays are thus preferred, wherever possible, to give the operatoran opportunity to investigate the causes of tripping before a restart.tTP82(uE101,-Current (Amps.)Figure 12.14 Operating band of a thermal overcurrent relayNote All thermal relays are available in two types of trip contacts,self (auto) reset and hand (manual) reset. In self reset relays, aftera trip the relay resets automatically, as soon as the bimetallic elementcools down and regains its original shape. In a hand-reset relay,after every trip the operator has to reset the relay manually beforea restart of the motor. The latter type thus provides an opportunityto the operator to investigate the causes of a trip and correct these,if possible, before a restart. In the former case, the motor, withoutbeing subjected to an investigation, may restart on its own (whenit is so wired) as soon as the bimetal cools. Since it may not befeasible to achieve bimetallic elements having identical coolingEutectic alloy relaysThese relays also possess characteristics similar to thoseof a bimetallic relay and closely match the motor heatingand cooling curves. They are basically made of a lowmeltingeutectic alloy which has defined melting properties.The alloy, with specific proportions of constituentmetals such as tin, nickel and silver, can be made fordifferent but specific melting temperatures. This propertyof the alloy is used in detecting the motor’s operatingconditions.Such relays are in the form of a small tube insidewhich is a loosely fitted rotatable shaft, held by a verythin film of this alloy. The alloy senses the motortemperature through a heater connected in series withthe motor terminals and surrounding this tube. When themotor current exceeds the predetermined value, the alloymelts and enables the shaft to rotate and actuate thelever of the tripping mechanism.Such relays are satisfactory in performance and maybe adopted for all industrial controls. Some of thefeatures of these relays for use on motor starters aregiven below:1 The motor can be switched ON only after the alloysolidifies again and hence prevents the motor froman immediate switching after a trip, thereby givingthe motor adequate time to cool.2 An inadvertent or deliberate high setting of the relay(which is quite common to prevent frequent trippings)than recommended is not possible on these relays dueto their very narrow operating range. Prima facie, itmay appear to be a disadvantage with such a relay, asa number of alloy ‘strip sets’, with different operatingranges, may have to be stocked for every motor. Butthey provide more precise protection for the machine.Note The eutectic relays have an advantage by setting thepointer more closely in the field, based on the actual measurementof the load current.Figure 12.15 CT-operated thermal overcurrent relay((Courtesy: Siemens)3 Since the operation of the eutectic alloy relays dependsupon the magnitude of heating, which is a function ofcurrent and time, these relays also give an inversecurrent-time characteristics.4 Due to their very narrow operating range, these relayshave limitation in their application on drives withfluctuating loads such as cranes, hoists or pole changingmotors, or loads with intermittent duties with frequentvariations in the load current. In view of a generallywide voltage fluctuation on an LT distribution network,even a definite duty motor may have cognisable currentvariations, leading to unwanted trippings of themachine.

Protection of electric motors 12/37Note In view of the\e limitations, such relays did not findadequate acceptance under Indian conditions. As a result. theirmanufacture has been discontinued.Large LT motorsLarge motors call for a more judicious selection of relays.Unlike small motors, one cannot take for granted thatthe thermal characteristics of the relay will be the sameas that of the motor, and arbitrarily select any thermalrelay. To make use of the optimum capacity of a motorand to yet protect it from all possible unfavourableoperating conditions it is essential that the motor and therelay’s thermal characteristics are matched closely.Motors designed according to IEC 60034-1 are notmeant for continuous overload running unless specificallydesigned for this. They should be closely protected withthe available devices. On the one hand, the protectionshould be discriminating, to allow for starting currentsurge and yet detect an overloading, unbalance, shortcircuitor a ground fault before these cause damage tothe motor. On the other, it should ensure a full-loadoperation of the motor.A thermal relay cannot be set reliably to remaininoperative at 100% of the full load current and thenoperate instantly as soon as it exceeds this. A good thermalrelay can be set to operate between 1 10% and 115% ofthe /r, or even more if that is desirable, provided that thethermal capacity of the motor can permit this. To ascertainthis. availability of the motor thermal withstand curve isessential. Accordingly, the relay can be set for the optimumutilization of the motor by setting it forRelay setting (% of FLC)Nore It is also possible that in the operating region. beyond point‘A’. the curve of the thermal relay had fallen far below the motorthermal curve and had overprotected the motor. In other words. itwould have underutilized its capacity, in which caw, it will benecessary to call for a reselection of the thermal relay that wouldpermit optimum utilization of the machine and. if necessary. givingit support through an IDMT relay to cover the underprotected region.Such a combination of an OCR and an IDMT relay is dsfactoryfor detecting a system fault. overloading or a \tailing condition butit cannot guarantee total protection. This combination does nottrace a replica of the motor heating and cooling curve\. It cansimply detect the motor line currents and not the conditions thatmay prevail within the windings, such as thwc as caused by anunbalance or a single phasing. Nor can they accurately assess therotor’s heat caused by prolonged starting time or frequent starts.The,e relays, at best, can be employed with instantaneous definiteminimum time to inverse and very inverse /‘-t characteristics tomatch the machinc’s requirement as closely as po\sible. In view 01this. it will be worth while to have a single device protection againstoverload and stalling which may occur due to undervoltage.unbalance, single phasing or a ground fault. Such a protection ispossible through a single device niotor protection relay. discussedin Section 12.5.HT motorsThese call for a closer protection, which is possible througha single point motor protection relay (MPR). Since asingle MPR provides protections against unfavourableoperating as well as fault conditions. we discuss thisrelay separately in Section 12.5-Motor maximum operating current (%)1.1 or 1.15(typical)Additioriul protection through ci supplementunIDMT re la^Since thermal relays, with numerous characteristics andad,justable settings to match every individual motor, arenot feasible, the nearest characteristic relay available inthat range must be chosen. If it is considered necessaryto ensure adequate protection at each point of the motorcurve. this relay may be additionally supplemented throughan inverse definite minimum time (IDMT) relay, havinga definite time or inverse to very inverqe time characteristics,whichever may best suit the motor’s unprotectedregion on the thermal curve, as illustrated in Figure12.16. As can be observed, the closest relay chosen forthis motor does not protect it during a start due to ahigher tripping time than the motor thermal withstandtime (tr > r,,,), while during a run, beyond the operatingregion ‘A’, it lies closely below the motor curve asrequired. During a start, therefore, it has beensupplemented by an IDMT relay, whose startingcharacteristic lies closely below the motor thermalwithstand curve (rnl > t,r) and provides the required startingprotection. Hence with the use of these two relays, themotor can be fully protected.Motor thermalf,, < fm < 1,Figure 12.16 Supplementing a thermal relay with an IDMTrelay for complete motor protection

12/288 Industrial Power Engineering and Applications HandbookCommencementof short-circuitCurrent cut-offr level5fs = Pre-arcing timet, = Arcing timett = Total cut-off timeisc(peak) = Peak value of a fault currentis, = Peak cut-off current by fuseq \\\ c 4Symmetricalshort-circuitcurrentFigure 12.17 Cut-off feature of an HRC fuse12.4.2 Protection against fault conditionsProtection against short-circuits(use of HRC fuses)It is always desirable to protect a power circuit againstshort-circuits separately. HRC fuses are the immediateanswer to such a requirement for small and mediumsizedLT motors. They have a delayed action characteristicunder an overload and are instantaneous against a shortcircuitcondition and thus inherit the quality of discrimination.They reduce the electromagnetic and thermalstresses and enhance the withstand capacity of theprotected equipment for higher fault levels. Quite often,they are used on the receiving end side of a supply systemto enhance the short-circuit withstand capacity of all theequipment connected in the circuit and installed afterthe HRC fuses. Figures 12.17-12.19 illustrate how theHRC fuses, by quickly isolating the circuit on a faultwell below the prospective fault level, reduce the electromagneticand thermal stresses on the connected equipment.If the same fault had been cleared by an ordinary shortcircuitrelay it could reach its momentary peak value,which can rise to 2.2 times the r.m.s value of fault currentin LT and 2.5 times in HT circuits (Table 13.1 1) as it hadtaken more than one-half of a cycle to clear the fault.For instance, a fuse of 320A, of characteristics shown inFigure 12.18, will cut off a fault of 30 kA symmetrical,with a crest (prospective) value of around 63 kA at only15 kA, which is well below its prospective value.Selection and coordination of fuse ratingThe selection should be such as to provide properdiscrimination at the various levels of a multi-distributionnetwork. Our discussions generally take account of therecommendations in IEC 60947-4-1 regarding cooi-dinationbetween the short-circuit protection and the maincomponents such as switching devices [switch, breaker,20.1 o.21 2 4 6 8 10 20 io 40 6080Symmetrical prospective --tshort-circuit current I, (rms)(kA)Illustration:A 320A fuse will cut-off an lSc of 30 kA (peak valueup to 2.1 x 30 kA) at less than 15 kA (peak).Figure 12.18 Typical current cut-off characteristics for LT HRC fusesat prospective current up to 80 kAMCCB (moulded case circuit breaker) or MCB (miniaturecircuit breaker) and contactor] and the overload relay.These recommendations permit damage of componentson fault to varying degrees as noted below:Type 1: Under short-circuit conditions the contactoror the starter will cause no damage to the operator orthe installation but may not remain suitable for furtherservice without repairs or replacement of some of itsparts. In other words, contact welding of the contactoris allowed and burnout of OCR is acceptable. In eithercase replacement of components may be necessary.Type 2: On the other hand, Type 2 degree of protectionlimits the extent of damage in the case of a shortcircuit.Now under short-circuit conditions, the contactoror the starter will present no risk to the operator or theinstallation and will remain suitable for further service.In other words, no damage to the contactor or theOCR is permitted. It may, however, be interpreted thatcontact welding may be permissible to the extent thatthe contactor can be put to service once again after abrief period by separating the contacts with the help

Protection of electric motors 12/289-10 20 50 100 200 500 1000 2000 5000 10000 20 000 50 000Prospective current in amps (r.m.s)Figure 12.19 HRC fuse characteristicsof a tool such as a screw-driver but without calling forreplacement of any part.Type 2 coordination is more prevalent and commonlyused for all industrial applications. Below, we concentrateon coordination Type 2, permitting the least damage andlonger service life. This coordination can safely withstandnormal fluctuations in system parameters and operatingconditions during normal working. It is always advisableto verify the authenticity of the coordination in a laboratory.For procedure, to establish the type of co-ordination,refer to IEC 60298. To achieve the required precisecoordination we discuss a few typical cases below.Discrimination between fuse to fuseRefer to the normal distribution network of Figure 12.20.Selection of the fuse ratings must be made on the followingbasis:(a) Only fuses nearest to the fault should operate. Forinstance, for a fault at location C, the only fuses atlocation C should operate and not those at B or A.(b) To ensure the above, the total arc energy, Z2 . tt, of thelower fuses at C should be less than the pre-arcingenergy, I2 t, of the upper fuses at locations B or A.F f ,Figure 12.20 Coordination of fuse ratings and theircharacteristics

12/290 Industrial Power Engineering and Applications HandbookIn other words, the current-time (1’ - t) characteristicsof fuses at C should lie below that of fuses at B, andthat of fuses at B, should lie below the fuses at A.etc., throughout their operating range.Coordination of fuses with an overcurrent relav orany other overcurrent protective deviceThe selection of the fuses should be such that:(a) They do not operate during a start.(b) They do not operate against overloads as these aretaken care of by the overcurrent relay or any otherovercurrent protective device.(c) They should isolate the supply to the motor in theevent of a fault sufficiently quickly before the faultcauses damage to the connected equipment by burningand welding of contacts of the contactor or by causingpermanent damage or deformation to the bi-metalelements of the OCR. This is possible by ensuringthe let-through energy (1’ . R . t) of the fuses underfault conditions to be less than the correspondinglet-through energy of the OCR (Figure 12.21).Coordination of fuses with a breakerWhen the fault level of the system is more than thebreaking (rupturing) capacity of the associated breaker(usually MCB or MCCB) or when the fault-makingcapacity of the breaker falls short of the momentary peakvalue of the fault current of the system (Table 13.1 1) thatfuses can be used before the breaker to provide back-upprotection and supplement its breaking capacity, to makeit capable of performing switching operations successfullyon a fault. Coordination between the fuses and the builtinoverload and short-circuit releases of the breaker canbe carried out on the following basis:(a) Overloads on the system should be cleared by trippingthe breaker through the in-built releases only andnot through the fuses.(b) Short-circuit currents up to the breaking capacity ofthe breaker should also be cleared by tripping thebreaker through the in-built releases and not throughthe fuses.(c) Short-circuit currents in excess of the breakingcapacity of the breaker alone should be cleared bythe operation of the fuses.To achieve the above the characteristic of the fusesshould lie well above the characteristic of the overcurrentand short-circuit releases of the breaker for the lowerregion of currents, such that only the breaker operates.However, it should lie well below the characteristic ofthe breaker in the higher region of currents to ensure thatthe fuses operate sufficiently quickly and long beforethe in-built releases. The breaker is thus prevented fromoperating at currents that are in excess of its breakingcapacity. Figure 12.22 illustrates such a coordination,Note1 Back-up protection to supplement an interrupting device to makeupits rupturing capacity and make it capable of meeting thesystem fault level is a concept of LT systems only to sometimesreduce cost. Interrupting devices for an HT system normallypossess adequate rupturing capacity to meet system needs easily.Moreover, HRC fuses beyond certain voltages (>I I kV) andcurrent ratings (>IO00 A) are generally not used. Whereinterrupting devices have a limitation in their rupturing capacity(which may sometimes occur on an EHV system). the systemfault level can be altered accordingly. so that the availableinterrupting devices can still be employed (Section 13.4.1(5)).On an LT Fystem too bach-up protection ib nul recommendedfor an ACB or an OCB as they both would possess a trippingtime of more than a cycle (Table 19.1 ) compared to the curremlimiting properties of the HRC fuses. Current limiting propertiesof HRC fuses make them operate much faster (< cycle) thana breaker, during a fault condition. even when the breaker couldsafely clear the fault. In such cases, therefore, when breakersare required for a higher rupturing capacity than a standardbreaker possesses, then a breaker of a higher rating. which mayf tCurrent -On fault 1’Rt (fuses) < I’Rt (OCR)fsl - Safe thermal withstand time of motorduring a full voltage startFigure 12.21 Coordination of fuses with an OCR1st!IBreaking capacityCurrent --+of the breakerFault level ofthe systemOn faults exceeding /sc of breaker, /*Rt(fuses) c 1‘Rt (breaker)Figure 12.22 Coordination of fuses with a breaker

Protection of electric motors 12/291- 7have the required fault level, may be chosen with a lower settingof the OCR. This protection may, however, be applied in anMCB or an MCCB, when they possess a rupturing capacity lessthan required. Since the MCB and the MCCB can both becurrent limiting the characteristics of the fuses and the breakerscan be coordinated such that faults that are in excess of therupturing capacity of the breakers alone are handled by thefuses.In such cases it may often be possible to meet the requirementby selecting a higher frame size of MCB or MCCB, which maypossess a higher rupturing capacity also. If not, and to save oncost, one may provide HRC fuses for back-up protection.To make a proper selection of HRC fuses it is essential that thecurrent-time characteristic curves for the releases of the breakerand the fuses are available from their manufacturers.Coordination of fuses with a switch or a contactorSince both these devices possess a certain level of makingand breaking capacities, the same criteria will apply asfor the breaker noted earlier. Rating of fuses shall not bemore than the switch or contactor rating.Coordination of fuses with a transformerConsider a distribution HV/LV transformer. If the fusesare provided on both HV and LV sides, the fuses on theHV side must protect a fault within the transformer whilethe fuses on the LV side must clear overcurrent and faultconditions on the LV side. Thus, for a fault on the LVside, only the LV side fuses must operate and not the HVside, similar to the requirements discussed above.If the transformer is HV/HV, the same requirementmust prevail, i.e. for a fault on the downstream (secondaryside) only the downstream fuses must operate and notthe fuses on the upstream (primary side).,Nore It is. however, possible to eliminate the use of HRC fuses inLT hl~tems at least. with the availability of more advanced technologyin an MCCB or an MPCB (motor protection circuit breaker). SeeSection 12.1 I for D fuse-free system.12.4.3 Protection against stalling andlocked rotorMotors which do not possess a sufficient gap betweentheir hot withstand and starting curves generally call forsuch a protection. Large LT motors and all HT motorsare recommended to have a separate protection againstsuch a condition. A locked rotor protection relay basicallyis an overcurrent relay, having an adjustable definite timedelay to trip the motor when it exceeds its permissiblestarting time, but before the safe stall withstand time.This feature is available in a motor protection relaydiscussed later. Where, however, such a relay (MPR) isnot provided, a high-set IDMT overcurrent relay can bechosen to match the upper range of the motor thermalwithstand curve (Figure 12.16).12.4.4 Protection against voltage unbalance ornegative phase sequenceSuch a condition also generates negative sequencecomponents. For small and medium-sized motors, say,up to 100 h.p., no separate protection for such a conditionis normally essential, when the overcurrent relay possessesa built-in single-phasing protection or the tripping circuitis provided with a single-phasing preventor. If a separateprotection for this is considered desirahle, one may usea negative phase sequence relay like the one shown inFigure 12.23(b) or similar static or PLC-based relays.For larger motors, however, one should employ a relaylike a motor protection relay. which covers in one unitall the protection as described in Section 12.5.Note One should employ only current \enring relays ar far a\possible for such applications as a negative sequence current has asevere effect on motor windings due to a much lower negatiLeqequence impedance of the motor (Section 1?.2(v)) than ;Icorresponding negative sequence voltage.Such a relay is connected on the supply side a\ shown in Figure12.23(a). The arrangement is such that unless the relay closes, themotor switching circuit will not energize and the motor will notstart. The contact closes only when the supply voltage i\ normaland the phase sequence positive. Even for undervoltage conditions.the torque developed by the relay may not be adequate to close thecircuit. Such relays are, therefore, effective against(i) Negative sequence voltages when torque developed bq it\coil is negative.(ii) Voltages far too low to produce an adequate torque to closcthe relay. This is possible during a start only. as during a runthe relay contacts are already established and the coil doernot detect a fall in the voltage.(iii) Single phasing during a start. a% this will also produce a negativetorque to close the relay.12.4.5 Protection against single phasing (SPP)An ordinary thermal relay senses only the line currentsand is not suitable for detecting a single-phasing condition.Referring to the curves of Figure 12.10, an ordinary relayset at 110% of FLC will not trip in the event of a singlephasing when the motor is operating underloaded. say,at only 60% or less of the rated current, while in the lonephase X , the current would be as high as 135% of FLC.For single-phasing protection, ‘single-phasing preventors’are available. Although it is essential to protecteven a small motor against single phasing. there is littlepoint in employing these preventors unless they formpart of the basic starter (OCR with built-in SPP feature)as an economic consideration.As discussed above, most of the leading manufacturersof switchgear components produce thermal overcurrentrelays with a built-in feature of single-phasing prevention.The use of a separate single-phasing preventor device isthus not necessary up to medium-sized motors, wherethis feature is available in the relays. For large-motorsand for critical installations, a separate unit for singlephaseprotection may be required for prompt singlephasingalarm and/or trip. Interestingly, where star deltaswitching is employed, the overload relays, which arenow connected in phases of the motor windings,automatically sense an abnormal condition in any 01- allthe phases, and provide a single-phasing protection. Seethe power circuit diagram in Figure 13.56 for more clarity.A separate single-phasing protection device is availablein two versions:1 Voltage sensing, and2 Current sensing

la292 Industrial Power Engineering and Applications HandbookR Y BT T Tsequence voltage---Static negative sequence relay (Courtesy: English Electric)INFigure 12.23(a) Schematic for a negative phase sequence(NPS) relayVoltage-sensing preventors have limitations and are notreliable since they offer protection up to the sensingterminals only. Protection beyond these terminals up tothe motor terminals is not possible. In the event of aphase failure beyond this point, the voltage-sensingequipment will not detect this. (See the schematic ofFigure 12.24.) Current-sensing preventors are thereforerecommended for more reliable detection of a faultanywhere within the system up to the motor terminals.Moreover, voltage-sensing preventors may act erratically,when the motor is generating high back e.m.f. on singlephasing, and also when the power factor improvementcapacitors are connected across the motor terminals. Inthis case high back e.m.f. can be produced across thevoltage-sensing relays, which can make its operationuncertain.The current-sensing solid-state type relays consist ofa filter circuit to sense the negative sequence current.The output of this filter is proportional to the negativesequence component of the current. The output is fed toa sensor, which detects the level of negative sequencecomponent of current and trips the starter circuit whenthis level exceeds the set limit. (See Figure 12.25.)Normally such preventors are designed up to 30 A formotors up to 20 h.p. For larger motors, the output currentcan be sensed through CTs of 5 A secondary (Figure12.26). 5A secondary is chosen to make detectioneasy.At a preset value, the relay operates and opens thePLC or micro-processor-based negative sequence relay(Courtesy: Alstom)Figure 12.23(b)control circuit to trip the starter unit. To avoid nuisancetripping, due to surges and momentary line disturbances,a time delay of 4 to 7 seconds is normally introducedinto the tripping scheme.12.4.6 Protection against voltage surges(for systems 2.4 kV and above)Voltage surges may be of two types, external or internal(Section 17.5). A motor will require protection againstboth for absolute safety. For external surges, lightningarresters are provided as standard practice at the receivingend as illustrated in Figure 12.27, to protect the electricalinstallation as a whole. This lightning arrester will limitline surges due to external causes, within safe limits asin Table 13.2 for series I or Tables 14.1 and 14.2 forseries I1 voltage systems. In all likelihood it will alsoprotect the main insulation of the rotating machines. Theinsulation level of a motor is much less than other electricalequipment such as transformers and switchgears connected

Protection of electric motors 12/293R Y B NR Y B NwSingle phasing-. ~--- . .c -~130 & On OffSP tripE, I1i Lwgure 12.24 rower ana control scneme tor a voltage-operated singlephasingpreventerF1-3 sNSinglephasingpreventorSingle phasingOIC & On OffSP tripFigure 12.26 Power and control scheme for a current-operatedsingle-phasing preventor (use of CTs for higher ratings)For internal causes, when considered necessary,particularly when multiple reflections are expected atthe motor terminals due to the long lengths of cablebetween the starter and the motor, an additional surgearrester may be provided at the motor terminals or asnear to it as possible, in association with a surge capacitor,as shown in Figure 12.27. This will account for theprotective distance (Section 18.6.2) and also limit themagnitude and steepness of the switching surges to protectthe turn insulation. This subject is dealt with in greaterdetail in Sections 17.7 and 17.8.Single phasing* SP trip6-co --'1 I 13Figure 12.25 Power and control scheme for a current-operatedsingle-phasing preventor (direct sensing up to = 30 A)on the system. The line side lightning arrester is selectedto protect the basic equipment only, but a motor normallyinstalled away from the arrester will be subject to onlyan attenuated lightning surge by the time it reaches themotor, and hence would be safe.12.5 Single-device motor protectionrelaysThese relays are programmed to provide all possibleprotection necessary for a machine in one unit such asovercurrent, unbalance and single phasing, locked rotor,prolonged starting time and multiple starts, short-circuitand ground faults etc. They take account of the heatingeffect caused by negative sequence components, whichmay arise due to system unbalance, a fault in the motorwindings or a single-phasing condition. They are alsotemperature compensated and can be set to follow closelythe changes occurring in the operating temperature ofthe machine, so that the machine will trip only when itheats up beyond the permissible limits. It thus ensuresthe optimum utilization of its capacity. Moreover, theovershoot of the thermal element is kept low (of theorder of 2% or so) and the relay can be set between thestarting I* - t and the 'hot' I* - t characteristics. withoutthe risk of operation on a hot start. They thus possess a

p12/294 Industrial Power Engineering and Applications HandbookOverhead lineLine side lightning- arrester to limitthe amplitude ofTransformer external surgesd b((1 -I \=Main switch gear busTo protect the,turn insulationSurge capacitoruto tame theamplitudesteepness- 'GFigure 12.27 Typical scheme for surge protection of a rotatingmachinedefinite advantage over a conventional thermal overcurrentrelay. They incorporate circuits which can be set to tracethe motor's internal conditions during operation and henceprovide a thermal replica protection to the machine throughsignal, blinker and alarm facilities, available with them.These relays are recommended for large LT motors (say,300 h.p. and above) depending upon their applicationand all HT motors, where more precise and accurateprotection is rather essential, besides requiring theoptimum capacity utilization of the machine. A goodrelay will normally incorporate the following features:1 It measures r.m.s. values to take account of harmonicspresent in the supply system. Ir,,,s, =4 If + 1: + If + . . . where I,, I3,I, are the differentharmonic components.2 It stays immune under permissible operatingconditions.3 It gives an alarm or an indication of a likelyunfavourable operating condition well in advance.4 It trips quickly on a fault condition and relieves themotor from the prolonged stresses of the fault, whichmay causc cxcessive thermal and electrodynamicstresses.5 It simulates the motor's cooling-down condition, forat least 30-60 minutes, during a temporary powerfailure.6 It monitors and displays the starting conditions suchas time of start and number of starts etc.7 It measures the values of VI, I, and temperature, 8etc., and displays/monitors the cause of the last trip.8 Trip indication will memorize the operatingconditions of tripping with causes of the trip untilthe fault is acknowledged.9 All trips with causes and starting parameters areprogrammed for an accurate diagnosis of the causesof a trip. It helps to identify remedial action necessaryin the operating conditions of the load for a healthyfunctioning of the machine.IO With these relays, there is no need to use HRC fusesor thermal OCRs. The power circuit is thus devoidof any heat generation in these components andprovides energy conservation.These relays are available in various versions such as,Electromagnetic: These are quickly outdated but wediscuss these relays briefly below to give an idea ofthe basic operating principles of such relays. The sameprinciple of application is then transformed into a staticor microprocessor-based relaySolid state: based on discrete ICs andSolid-state microprocessor based: these are moresensitive and accurate. They can be made digital to beconnected to a computer for remote monitoring andcontrol of the process that the motor is driving.12.5.1 Electromagnetic relaysIn this case the relay is in the form of a bridge circuit andthermal detection is achieved through various methodsother than a bi-metallic heater element discussed below.Heat sink methodThis is achieved through a heat sink thermistor, Th,. Thethermistor has two heaters HI and H2 (one heated by thepositive and the other by the negative sequence component).They form one arm of a sensitive Wheatstone bridge.A temperature-compensated thermistor, Th2 forms theother arm of the bridge (Figure 12.28). The heat sinkthermistor is heated directly or indirectly by the linecurrents. The heating changes its resistance, which isused to provide a signal by the relay, for either trippingor an alarm. Heater, H2, of the negative filter network isso designed that it produces six times the heat that wouldbe produced by heater HI for the same amount of current.Heaters HI and H2, thus detect a heating effect equivalentto (Z,? + 61: ), which would be the maximum heatproduced in the stator or the rotor during an unbalanceor a single phasing, as discussed in Sections 12.2(v) and12.3. They also overcome the deficiency of theconventional thermal overcurrent or IDMT* relay byproviding longer periods on overloads yet tripping quicklyon short-circuits. These inherent features of heat sinkrelays make them ideal for motor protection and areused by various manufacturers of motor protection relays.These relays also possess a property to discriminate*IDMT: instantaneous definite minimum time

Protection of electric motors 121295Instantaneousunbalance unitC7Instantaneousovercurrent unitR1:l IsolatingI T"THIH,Heater energized by positive filterHeater energized by negative filterTh, Main thermistorTh, Ambient temperature compensated thermistorR, R, R, etc To give the bridge linear resistanceitemp characteristicBallastresistorAuxiliaryD.C supplyFigure 12.28A typical scheme of a heat sink circuitbetween start and stalling conditions due to their low'overshoot' and can therefore detect a prolonged startingor stalling condition without a trip.Operation of heat sink thermistor relayThis relay will operate whenever there is an unbalancein the bridge circuit due to a change in the resistance ofthe main thermistor. because of the combined heating ofheaters H, and H,, as a result of overload, voltageunbalance, inter-phase fault or a single phasing. Theoperating time will depend upon the rate of heating, i.e.the amount of overload or unbalance, thus giving aninverse time characteristic.The following protections may be possible:1 Overload2 High set instantaneous overcurrent through the positivesequence network. An initial delay of a few cycles isintroduced to avoid a trip during a start, whereas itwill trip instantly on a phase fault, cable fault or ashort-circuit.3 Instantaneous unbalance current.4 Prolonged starting and stalling protection. When thestarting current of the motor does not fall within apredefined starting time or up to the thermal withstandtime of the machine, the relay will trip dependingupon the setting of its stalling detection unit.5 Instantaneous ground fault by providing a zerosequence relay in the residual circuit of the CTs.Therefore a motor protection relay, which may bea 'normal time' or a 'long time', will be able to providethrough one unit a comprehensive motor protection,generally requiring no more protection, except for asurge protection, which may be needed for an HTmotor.12.5.2 Solid-state relaysThere are no moving parts in these relays (Figure1 2.29(a)).They employ static technology, Le. transistorized integratedcircuits to achieve a thermal replica protection.To detect the presence of a negative sequence component,filter circuits are used to separate positive and negativecomponents I, and I, of the input current in each phase.These currents are then used to produce voltages proportionalto these currents. These voltages are then fed to asquaring circuit to give a heating effect in the motorwindings proportional to I,' and I t. Figure 12.30(a) isa typical schematic diagram of such a relay and a briefoperating description is given below.The three line currents through the secondary of thethree CTs, as shown in Figure 12.30(b), are fed to asequence filter network which separates the positive andnegative sequence components of the line currents drawnby the motor. These currents are then fed to separate

12/296 Industrial Power Engineering and Applications HandbookSolid state motor protection relayRelay withdrawn from its chassisFigure 12.29(a) (Courtesy: English Electric)potentiometers, at different settings, through the instantaneousoperating elements /, and I,. The potentiometersprovide two output voltages V, and Vu, corresponding tothe positive and negative sequence current componentsrespectively. Ve and Vu are fed into the squaring circuitsto give kl . Vj and k, . V , effects. These values arethen added to give an output voltage effect ofkl . Vy' + 6k2 . V,' to the integrator. A feedback circuitacross the integrator causes the output voltage from theintegrator circuit to rise exponentially from zero to avoltage which is equivalent to about 105% (typical) ofthe relay setting current. The output voltage from theintegrator is fed to a level detector, which energizes anoutput unit when the set voltage is reached. This operateselectrically separate contacts that can be used to trip orgive an alarm to the motor power circuit.Figure 12.29(b) PLC or micro-processor-based motor protectionrelay (Courtesy: Alstom)12.5.3 Microprocessor-based relaysA static as discussed above is capab1e Of providingmore functions and more accurate operations and can

Protection of electric motors 12/297also memorize historical data and monitor a process moreclosely. This relay can also be provided with a microprocessor.A microprocessor-based relay consists of PCboards, a processor board and other electronic circuits.R Y EA typical relay is shown in Figure 12.29(b). These relayscan also be made digital to be connected to a centralcontrol system for close monitoring and control of aprocess. Now they can have much wider application,such as better communication and information feedbackfacilities, to optimize a process and maximize productivity.For more details refer to Section 13.2.3 or contact themanufacturers.The following are the normal protections that may beavailable in both-a solid-state or a microprocessor-basedrelay, making them a single-device protection:Figure 12.30(a)Power diagram1 Prolonged starting protection. If the temperature riseof the machine is more than 50% of the permitted rise(0,) during the first start, the relay will lock out toallow a pause and prevent a consecutive or a quickrestart until the machine cools sufficiently and thetotal temperature rise 0, or f?,, does not exceed e,(equations (3.2) and (3.4)) during the next start. Thestarting time is fed to the memory of the relay tomonitor the total starting time.Likely settings - 8 versus tripping time, or time ofstart versus tripping time-whichever occurs first.Likely features - an advance alarm and an indicationbefore a trip.2 Stalling or locked rotor protection. This is also detectedby the prolonged starting time as well as overheatingof the machine. It is possible that the machine wasalready under operation and hot when it had stalled.Under such a condition, the rotor operates at a highfrequency and is more vulnerable to damage. Since itis not possible to create a replica of the rotor, separatekNotea- Sequence filterb- time delay (= 60 ms(-3 cycles for ac- I, time delay ] 50 Hz system))d- Squaring circuite- Squaring circuitf - Summation circuitg- integratorh- Level detector and outputi - Power supply1 - Output (Th)k- Ig InstantaneousFigure 12.30(b) Typical schematic illustrating the squaring technique

12/298 Industrial Power Engineering and Applications Handbook.? is mI'LI ITime-One cycle i3 jDuring an overload condition, the relay maybe set to give alarm if normal conditionsdo not restore by the end of the cycle.__--02 Irm"N2 s! u?eel$ 041e,, = e,-0 tl t2 5 t4TimeNote The thermal data are marked for loads I,, and /12Figure 12.31motorII2Heating and cooling curves of an intermittent dutyprotection, other than a prolonged starting time, istherefore essential to monitor both 8 and starting timeand hence this protection.Likely settings - 8 versus tripping time or time ofstart versus tripping time. A problem may arise inproviding an accurate time setting when tSt is largedue to high-inertia drives or reduced switching voltages,when it may approach the safe stall time.Nofe In both the above cases, which are almost similar, so faras the switching heats of the stator or the rotor are concerned,the overcurrent protection (noted at serial no. 4) is redundant,as its time constant is much higher (of the order of severalminutes) compared to the temperature rise, particnlarly of therotor, which is linear and much more rapid under such conditions.Therefore, such protection saves the machine from excessivethermal stresses.3 Repeat start protection (Figure 12.31). The relay nowdetects both the summated starting times and temperaturerises 8, and 61, of the windings. It also detectsthe cooling time effect between the two starts if itexisted, to protect the machine against excessivestresses by a lock-out feature against repeated startswhen the summated starting time exceeds the presettime. The total time of start is set according to thethermal capacity of the machine.Likely settings - summated starting time versus trippingtime, or 8 versus tripping time.4 Overcurrent protection. To provide a thermal replicaprotection, the relay is set according to motor's heatingand cooling (I' - t) curves supplied by the motormanufacturer. If these curves are not available, theycan be established with the help of motor heating andcooling time constants, as in equations (3.2) and (3.4).A brief procedure to establish the motor thermal curveswhen they are not available is explained in Section3.6.5 System unbalance protection. As discussed earlier,an asymmetry in the supply voltage causesHigh 12R losses in the stator, andEddy current losses in the rotor.Thus, besides voltage unbalance it can also detect aninter-turn fault, which leads to a current unbalance. Asmall amount of unbalance is already detected by thethermal element of the overcurrent protection but asevere unbalance, such as during a single phasing,would require quicker protection and hence, thisprotection. The relay may be set for an I, of around3% or so.Likely setting: I, versus tripping time.6 Short-circuit protection. To provide an instantaneoustripping on a short-circuit delay of, say, one or twocycles may be introduced into the tripping circuit tobypass any transient currents and avoid an unwantedtrip.Likely setting: I,, versus tripping time.7 Ground leakage current protection. A separate groundfault protection should normally not be required inview of the protection already available against anunbalance. But since the setting of the unbalancedelement may not be sensitive enough to detect a smallground leakage current, a separate ground leakageprotection is recommended. This can be achieved bydetecting the ground leakage current, I,, through theground circuit or the residual current (Sections 21.2.2and 2 1.4).Likely setting: Zg versus tripping time.The setting may be kept on the higher side to avoidnuisance tripping due to:(a) Difference in CT outputs which may also causeunbalanced currents and initiate operation of therelay, particularly during a start when the currentis high.(b) Flow of zero sequence cable capacitive leakagecurrents during an external ground fault, and(c) Momentary ground transient currents whiledischarging a switching surge through a surgearrester if this has been provided at the motorterminals.8 Lockout or blocking feature. Such a feature is necessaryfor complete safety of the machine after every trip forany of the reasons discussed above. It is imperative

~~~~Protection of electric motors 12/299that the relay blocks and prevents the next switching,unless the fault is acknowledged, the reason for thetrip is analysed and normal operating conditions arerestored. The settings of the relay for all such unfavourableoperating or fault conditions are made dependingupon the functions and the setting ranges availablewith the relay. Since all the protections are based onr.m.s. values, a sensitive relay will also detect theharmonic quantities present in the system and providemore accurate protection for the machine.It may often be asked why separate settings arenecessary for different types of protection when it ispossible to set the relay for a replica protection. It is truethat the relay will monitor thermally what is occurringwithin the machine as long as there are only normal tounfavourable operating conditions. But this is not so,when a fault condition occurs. It is possible that on afault the temperature rise is not consistent with the assumedthermal replica due to high time constants. For instance,for a fault of a few seconds the temperature rise of thewhole machine will almost be negligible (equation (3.2)).This would delay the trip, whereas the heat may belocalized and very high at the affected parts and mayescape undetected as well as electrodynamic forces, whichmay also cause damage to the machine. Similarly, on asingle phasing, monitoring of the stator temperature aloneis not sufficient, as the rotor would heat up much faster(in Y-connected stators) due to double-frequency eddycurrent heating and less weight compared to the stator.Similarly, it may take much longer to trip under a stalledcondition. A severe unbalance, as may be caused by aninternal fault, may also result in heavy negative phasesequence rotor currents and require protection similar toshort-circuit protection. A short-circuit protection willnot detect a single phasing. Hence the necessity to provideseparate protection for different operating and faultconditions to achieve optimum utilization of the machine,with the least risk of damage. The relay must discriminatebetween an unfavourable operating condition and a faultcondition. While the former may permit a delay tripping,the latter will need a more discrete and quick tripping tosave the machine from more severe damage.Underload protectionIt is also possible to provide this protection in such relays.This will provide very vital system process information.A sudden drop in load may be the result of a fallout ofthe load due to disengagement of the coupling, breakageof a belt or a tool, etc. It can therefore help monitor thesystem process line more accurately.12.6 Summary of total motorprotectionIn view of the effect of various unfavourable operatingconditions on a motor’s performance, one should bemeticulous when selecting the most appropriate protectionto safeguard a motor under the most unfavourable operat-ing conditions. See also Section 3.6 where we haveprovided a brief procedure to reproduce the motor thermalcurves 8 versus I and vice versa, for the relay to providea replica protection to a motor and to have closer settingsof its various protective features. The following is a briefsummary of our discussions so far.As discussed in Section 12.2 and explained in Figure12.3, an ordinary overcurrent relay will detect betweencurves 1 and 2, as against the average heating curve 3and is thus oversensitive to insignificant amounts ofunbalance, not actually causing harm to the motor. Forinstance, an unbalance in voltage of, say. 3% may causean unbalance in the current (i.e. an apparent overloading)of, say, 18% in one particular phase and may be detectedby an ordinary hi-metallic thermal overload relay, whilethe stator does not heat up accordingly. Even the rotor,which is more sensitive to an unbalance is heated corres-ponding to a current of ,I; + 3(0.181, )? or 1.047/,,i.e. an overloading of nearly 5%. Similarly, during singlcphasing the relay should be able to detect twice the fullload current in a star-connected winding, or as shown incurve X of Figurel2.10, for a delta-connected winding.and not corresponding to the stator currcnt of 43 . I, fora star-connected winding and curve RIY of Figure 12. IO.for a delta-connected winding. An ordinary relay willtake longer to operate corresponding to the line current,whereas some of the internal circuits will be subject tomuch higher currents during this period.For meticulous protection, therefore, it is advisable touse a motor protection relay for large LT and all HTmotors. The following motor details and working conditionsare essential to know before making a properselection of a protective scheme:1 Type of motor - squirrel cage or slip-ring2 Rating - kW (CMR)3 Rated voltage and current4 Type of starting5 Starting current versus time characteristics6 Locked rotor current and corresponding ‘hot’ thermalwithstand time7 Motor thermal withstand characteristic curves8 Number of starts or reversals, if required, and theirfrequencyY Ambient temperature and10 Maximum fault currentHaving discussed the effect of the above parameterson the motor’s performance. we will now illustrate byway of an example a general case to broadly suggest aprocedure that can be followed to select the protectivescheme for a motor. For more detailed selection of themotor protection relay, reference may be made to therelay manufacturer.Example 12.7For the purpose of protection, consider the squirrel cagemotor of Example 7.1 for one hot start for which this motor issuitable.Step I: Assume the motor starting current versus timecharacteristics to be shown in Figure 12 32, and divide the

121300 Industrial Power Engineering and Applications Handbookt=-',":,"::::qrKEiq- 2o- Speed (rpm) N,Figure 12.32 Determining the duration of starting current= 980accelerating torque curve into three sections, A, B and C asshown, to ascertain the magnitude of the starting current andacceleration time for the different sections.Section A(a) I, = 750% of I,:. la"=(b)or/b = 730% of I,= 740% Of 1,& = 82% of T,- 450 x 974980= 367 rnkgand starting time t,A =o,821866 x 333.3375 x 3674.52 secondsAssume the safe time to be 5.2 seconds (considering a safetyfactor of nearly 15%)Section E(a) lb = 730% of I,1, = 630% of I,:. I,, == 682% of I,(b) T, = 48%or 215 mkgand starting time t,E = 7.71 seconds.Consider the safe time as 8.8 seconds.Section C(a) I, = 630% of I,v-1, = 100% of lr:. I,= ~= 450% Of I,(b)& = 80% or= 358 mkg.and starting time t,C = 4.36 seconds.Consider the safe time as 5.0 seconds.Plot the starting current and thermal curve of the motor as inFigure 12.33.Considering TJ = 94% and p.f. = 0.85, thenI, =450Amp. (Max.) at 80% of 3.3 kVx 2.64 x 0.85 x 0.94= 123.2 A

Protection of electric motors 12/301instantaneoussetting 850%TWO consecutive/hot startsI, 10038 40Time (seconds) +Figure 12.33 Plotting of starting current and thermal curvesUse a CT of ratio 15011A.Note If a power capacitor is connected after the relay, sayacross the motor terminals, to improve the p.f. of the machine,then only the corrected value of the current must be used.For example, if a capacitor bank of 130 kVAr is used for anindividual correction of the machine thenwhere,Sinceand130I, = I~'3 x 2.64= 28.43 A/c = capacitor currentat 0.85 p.f. = 123.2 AI, (active) = 123.2 x 0 85= 104.72 A/r (reactive) = 123.2 x 0.527= 64.9 A:_ Net reactive current = 64.9 - 28.43= 36.47 Aand I, (corrected) that the relay will detect= b!(104.72' + 36.47')= 110.9 AAll further calculations must be based on this current.104.72Asin @ = 0 527 164.9AStep II: Selection of HRC fusesI, = 123.2 AI,, = 750% of 1,Le. 924 Afs = 19.0 secondsConsider four equally spread starts/hour. From the selectionchart in Figure 12.34, select the characteristics B-6 anddetermine the fuse rating as 350 A on the ordinatecorresponding to a starting current of 924 A on the abscissa.Step 111: Selection of bimetallic overcurrent relayConsider the available setting range as(a) Overcurrent: 80-125%(b) Instantaneous unit: 800-1 400%The thermal curve of the motor does not show any significantoverload capacity and therefore the relay must be set asclose to the full-load current as possible, say at a setting of11 0%. ThenOvercurrent relay tapping = l.' 123.2 (relay rating: 1 A)1.05 x 150= 86% Of 1 ANote The factor 1.05 (typical) is the pick-up current of therelay.Say the relay is set at 85%. Then it will operate at 85/86 x 110,i.e. at 108.7% of I,, which should be appropriate and theinstantaneous setting= 1.087 x 750 = 815.25%say, 815%This setting is only a calculation. The exact thermal curves ofthe motor and the relay should be available to closely matchtheir characteristics at every point. An ideal relay in this casewould be one which, without tripping, will permit twoconsecutive hot starts, i.e. its characteristic should fie above

12/302 Industrial Power Engineering and Applications HandbookFigure 12.34-Motor starting current (A)Fuse selection chart for 6.6 kV system for a motor with run-up time not exceeding 60 secondsthe motor starting curve and close below the thermal curveas shown in Figure 12.33.Static thermal relay (discrete ICs or microprocessor based)For medium and large motors, 300 h.p. and above, this typeof protective relay should be preferred to achieve optimumutilization of the motor's capacity. Consider the available settingranges in the vicinity of(a) Thermal overload unit: 70-130% of the CT rated current.(b) Instantaneous (I,) unit: 600-1200% of the thermal unitsetting.(c) Instantaneous (1") unit: 200-600% of the thermal unitsetting.(d) Ground fault (I,) unit: 20% of the rated current(e) Stalling protection unit: 150400% of the CT rated current,2.5-25 secondsSettingsThe overload unit setting: as worked out above, at 85A,i.e. for 108.7% of I,.An instantaneous setting of 750% of the relay settingshould be appropriate, which can protect currentsexceeding7.5 x 1.087 x 123.2 Aor 1004 AUnbalanced setting. If set at loo%, can operateunbalanced currents to the extent of110' = 100' + 6 I,' (from equation (12.6))or I, = P O Z 61°02= 18.7%, i.e. a voltage unbalance of nearly 3%Stalling protection unit setting. The current unit is to beset at 200-300%, and the time delay unit a little abovethe starting time but less than the safe withstand stalltime.(v) Single-phasing setting. If set at loo%, can operate singlephasingcurrents to the extent of110' = I,' +6/,'or I, = 41.58%Note: The above exercise is merely an approach to theselection of the most appropriate relay and its protectivesettings for a particular machine. The exact selection of therelay and its setting will depend upon the type of relay, itssensitivity and protective features supplied.12.7 Motor protection bytherm istorsA thermistor is a thermally sensitive, semiconductorsolid-state device, which can only sense and not monitor(cannot read) the temperature of a sensitive part ofequipment where it is located. It can operate preciselyand consistently at the preset value. The response time islow and is of the order of 5-10 seconds. Since it is onlya temperature sensor, it does not indicate the temperatureof the windings or where it is located but only its presetcondition.This is a later introduction in the sensing of temperaturecompared to the more conventional types of temperaturedevices available in an ernbcdded temperature detector(ETD), such as a thermocouple or a resistance temperaturedetector (RTD) described below. Thermistors can be oneof the following types:(i) NTC - having a negative temperature coefficient,and(ii) PTC - having a positive temperature coefficient.The resistance-temperature characteristics of both thesetypes are shown in Figures 12.35 and 12.36. One can

Protection of electric motors 12/303-Response temp "CFigure 12.35 Characteristic of an NTC thermistor160/0.363,aSwitching point(Curie point) \Figure 12.36 Characteristic of a PTC thermistornote an exponential variation in case of an NTC thermistor.Thermistor resistance decreases with an increase intemperature whereas in a PTC thermistor the resistanceremains constant up to a critical temperature and thenundergoes a very steep and instantaneous change at apredefined temperature, known as the Curie point. It isthis feature of a sudden change in the resistance of aPTC thermistor that makes it suitable for detecting andforecasting the motor's winding temperature at the mostvulnerable hot spots when embedded in the windings.However, they can be installed only in the stator windings,as it is not possible to pick up the rotor signals whenprovided on the rotor side. Also to take out leads fromthe rotor circuit would mean providing slip-rings on therotor shaft, which is not recommended.The resistance values of a PTC thermistor are givenas:(a) The resistance values in the range of -20°C to (Tp-20)"C should not exceed 250 0, with all values ofthe measuring d.c. voltages up to 2.5 V.(b) At (Tp - 5)"C - not more than 550 !2 with a d.c.voltage of not more than 2.5 V.(c) At (T,, + 3°C - not less than 1330 R with a d.c.voltage of not more than 2.5 V.(d) At (Tp + 15)"C - not less than 4000 R with a d.c.voltage of not more than 7.5 V.where Tp. is the tripping reference temperature or Curiepoint as indicated in Table 12.1.A typical characteristic curve of a PTC thermistor havinga Curie point of 120°C is shown in Figure 12.37.

12/304 Industrial Power Engineering and Applications HandbookTable 12.1 Curie points for a few PTC thermistorsWinding temperature (“C)Insulation class __f E B FTemperature referenceSteady overload condition 155 165 190Stalled condition 215 225 250Recommended reference temperature(Curie point) for thermistors (“C)Drop-off (tripping) Tp 130 140 160Warning Tp 110 120 140@ Curie point for ideal curve@ Curie point for typical curve-Temperature (“C)Figure 12.37 Typical characteristics of a PTC thermistorThe current through the thermistor circuit reducesdrastically and instantly as soon as the critical temperatureis attained as the resistance rises manifold. This featureis utilized to actuate the protective relay used in thetripping circuit, to protect the motor from overheating.Figure 12.38, illustrates a typical PTC thermistor protectivecircuit. It is this feature that has made PTC thermistorsmore useful and adaptable universally, compared to theNTC type. It is interesting to note that in a PTC thermistor,the switching point at which the resistance rises suddenlycan be adjusted, and the device can be designed for anytemperature to suit a particular application or class ofinsulation. They are normally available in the range of110, 120, 130, 140 and 150°C and can withstand atemperature up to 20O0C, such as required during impregnationand curing of the stator windings. Depending uponthe reference temperature (Curie point), the thermistorsmay be made of oxides of cobalt, manganese, nickel,barium and titanium. For motor protection, they are chosenaccording to the class of insulation used in the winding.For example, for a class E motor the switching point canbe chosen at 120°C and for class B at 130°C (refer toTable 9.1). These are tripping temperatures. For a prewarningby an alarm or annunciation they can be set ata slightly lower temperature so that an audible or visualindication is available before the motor trips to give anopportunity to the operator to modify the operatingconditions, if possible, to save an avoidable trip.Since the thermistor circuit will trip the protective circuitas soon as the thermistor current reduces drastically, itprovides an inherent feature to trip the protective circuiteven when the thermistor circuit becomes damaged oropen-circuited accidentally, extending a feature of ‘failto safety’.A thermistor is very small and can be easily placedinside the stator overhangs, bearings or similar locations,wherever a control over the critical temperature is desired.It is not provided in the rotor circuit (particularly squirrelcage rotors), as noted earlier. This device is embedded inthe windings before impregnation, for obvious reasons.For exact temperature monitoring, the thermistor is alwayskept in contact with the winding wire. The number ofthermistors will depend upon the number of statorwindings and the specific requirement of warning ortripping or both. Likely locations where a thermistorcan be placed in a motor are illustrated in Figure 12.39(a).Such a device can actually predict the heating-upconditions of a motor winding, at their most vulnerablelocations. It does not only provide total motor protectionbut also ensure its optimum capacity utilization. Theconventional methods of a motor’s protection throughan overload relay, a single-phasing preventor, a reversepower relay or negative sequence voltage protection alldetect the likely heating-up conditions of the motorwindings under actual operating conditions, whereas athermistor can sense the actual winding heating-upcondition. A thermistor may prove highly advantageousfor the protection of motors that are operating on a powersystem that contains many harmonics, and the actualheating of the motor windings may be more than theapparent heating, due to distortions in the sinusoidalwaveform (Section 23.8). A thermistor detects thissituation easily by sensing the actual heat. It is therefore,possible to employ such a single device for motorprotection to make protection simple, compact, muchmore economical and even more accurate. It also extendsan opportunity to an optimum utilization of the motor’scapacity. The only likely shortcoming to the operator orthe working personnel is the total absence of an indicationof the actual fault or the unfavourable operating conditions.The cause of a trip is now only guesswork, which is notdesirable, and hence the necessity for an elaborateprotective scheme, discussed above. But thermistors arevery useful for predicting unexpected hot spots in a motorduring actual running, which other devices may not beable to do. They are therefore employed extensively in

Protection of electric motors la305R-Y-B-N-sw3G-II M5 5PTCthermistorsTripping unit"DIwith an accuracy of less than 1 "C and a response time ofthe order of 0.2 second. Pure metals possess almost alinear temperature coefficient of resistance p, over a widerange and this characteristic is used in monitoring thetemperature of a particular object. In pure platinum+ pe) szR, = ~ ~ ( 1where R, = resistance at temperature 8°CRo = resistance at 0°C = 100 Q8 = end temperature "Cp = 3.85 x per "C at 0-100°CThe exact resistance variations of a P, - 100, RTD overa range of temperatures are given in Table 12.2, and notvery different from those calculated by the above equationand drawn in Figure 12.41.ApplicationAll the above sensing and monitoring devices are basicallysupplements to overload and single-phasing protection.They are worthwhile for critical installations that requiremore accurate sensing or monitoring of the operatingtemperatures of the different vital components or likelyhot spots. They also eliminate any chance of tripping forall operating conditions that can be controlled externally.A few applications may be speed variation, affectingcooling at lower speeds, frequently varying loads, frequentstarts, stops, plugging and reversals etc. In all these cases,the other protections may not detect the heating conditionsof the vital components of a machine as accurately as atemperature detector.Figure 12.38 A PTC thermistor protective circuitlarge and critical motors as temperature sensors. Theyare also being used as thermistor relays to trip a motorby converting the resistance change into a current changeof the control circuit. Figure 12.40 shows such a relay.12.7.1 Embedded temperature detectors (ETDs)They are basically temperature monitoring devices andcan indicate the temperature on a temperature scale. Theymay be one of the following types.ThermocouplesThese are bimetal elements, consisting of a bimetaljunction which produces a small voltage proportional tothe temperature at the junction. They are thus able todetect the winding temperature conditions when embeddedinside the motor windings. For more details refer to DIN43710.Resistance temperature detectors (RTDs)These are normally of pure platinum wound on a ceramicor glass former and encapsulated in a ceramic or glassshell, having an operating range of -250°C to +750"C,12.8 Monitoring of a motor's actualoperating conditionsTo warn of an unfavourable operating condition by theuse of an audio-visual alarm or trip or both, schemes canbe introduced in the control circuit by means of atemperature detector or other devices to monitor any orall of the following internal conditions of a motor:Motor winding temperatureMotor or driven equipment bearing temperature0 Coolant circuit inlet water pressure and temperatureMoisture condensation in the windingsMotor speedVibration levelSafe stall timeRotor temperatureAny other similar condition, interlocking with otherfeeders or sequential controls etc.To detect the above, sensing and monitoring devicescan be installed in the motor. Some of these are built-induring the manufacturing stage and others are fitted atsite during installation.Several types of sensing and monitoring devices areavailable that are embedded at a motor's vulnerablelocations and the hot spots at the manufacturing stage

12/306 Industrial Power Engineering and Applications Handbook1 Temperature monitorin the windings.@ PTC thermistors in theend windings.@ Temperature sensors inthe bearings.@ Vibration probes@ Pulse transmitters@ Speed responsiveswitchesTachogenerator (TG)Figure 12.39(a) General safety devices and their locations in a motorn I u or rnerrno-couplesterminals for monitoringand protection-IThermistor terminalsfor alarm or triDI"Thermistors and RTDs orthermocouples areembedded in all the threephases of the statorwindingsFigure 12.39(b) Wiring diagramand final assembly of the motor by the motor manufacturer.All these accessories are normally optional and must berequisitioned to the motor manufacturer at the time ofthe initial indent. These devices can be of the followingtypes.12.8.1 Motor winding temperature detection(by PTC thermistors and RTDs)HT motors specifically, and large LT motors generally,are recommended to have six such detectors, two in eachphase to sense or monitor the temperature of likely hotFigure 12.40 Thermistor motor protection relay (Courtesy: L&T)spots and provide an audio-visual alarm, annunciationand protection signals. For a stator winding, the preferredlocations for a PTC thermistor and an RTD may beconsidered as below:

Protection of electric motors la307Table 12.2 Characteristics of a Pt - 100 RTDTemperatureeocResistancea050100110115120125130135140145150100119.40138.50142.28144.18146.06147.94149.82151.70153.57155.45157.32RTD or -,thermo-coupleConductor oftop coil sideConductor of/-bottom coil sideSource IEC 60751160150t 140c: -130C2.Yi? 1201101000 25 50 75 100 125 150Temp (0)OC-Figure 12.41 Characteristic of a Pt -100 resistance temperaturedetector (RTD)PTC thermistors These are fragile and require usablespace in the slots. They are normally fitted in theoverhangs of the stator windings, as shown in Figure12.39(a). A sudden problem with the motor, causingoverheating, is instantly detected by an audio-visualalarm or a trip. They are preferred for smaller motors.For larger motors, protection through monitoring ispreferred to sensing only. Monitoring is possible throughRTDs or thermocouples.RTDs or thermocouples These are normally embeddedin the stator windings as illustrated in Figures 12.42and 12.39(a). The winding temperature can now bemonitored continually and a temperature replica ofthe machine obtained at any time. Figure 12.39(b) showsFigure 12.42 Location of a RTD or thermo-couple in a motorwindingtheir simple wiring. They are generally preferred tothermistors for large LT and all HT motors.12.8.2 Bearing temperature detection(by PTC thermistors or RTDs)Motors are also recommended to have one bearingtemperature detector in each bearing. This can be fittedwithin the threaded walls of the bearing that reach up tothe bottom of the bearing shell, i.e. close to where theheat is produced. Each detector may have two sets ofcontacts, each having ‘2-NO’ contacts, rated for 5 A at240 V a.c. and 0.5 A at 220 V d.c. One set can be set ata lower value to provide an audio-visual alarm and theother at a higher value to trip the motor.12.8.3 Coolant circuit water pressure andtemperature (moisture) detectionWater-cooled motors, type CACW (cooling type ICW37 A 81 or ICW 37A 91) (Section 1.16, Table 1.12)should be fitted with moisture detectors to provide anaudio-visual alarm in the event of a leakage in the watercircuit or a higher coolant temperature.12.8.4 Detection of moisture condensation in thewindings (by space heaters)Motors generally above 37.5 kW located in a humidatmosphere or required to be stored idle for long periodsmay be provided with one or two and even more spaceheaters, depending upon the size of the motor, suitablefor 240 V, 1-$I a.c. supply, to maintain the motor’s internaltemperature slightly above the dew point to preventmoisture condensation or deterioration of the insulationduring a shutdown. The heaters are located inside themotor at the lower end of the stator so that they areeasily accessible and their removal and replacementpresents no problem. The rating of total heating powermay vary from about 100 watts to 3500 watts, dependingupon the size of the motor. For motors up to 400 kW, one

12/308 Industrial Power Engineering and Applications Handbookor two space heaters, totalling about 100-250 watts, maybe adequate, whereas, for a 10 000 kW motor there maybe as many as four to six space heaters, totalling about3500 watts.12.8.5 Vibration probesThese may be used to detect the hunting of the machineand to provide an audio-visual alarm for unusual runningor for vibrations.12.8.6 Use of speed switch or tachogeneratorThis is a speed-sensitive device and is employed to monitorthe starting time, t,, in normal motors or the heating-uptime, tE, in increased-safety motors (Section 7.13.2). Thisis installed to detect a stalling condition for criticalinstallations where a false tripping due to an overprotectionor a malfunction of the locked rotor protection relay isnot desirable, or where the starting time, t,, may exceedthe heating-up time, tE. The speed may be set at about10-30% of N,, depending upon the speed-torque curvesof the motor and the load. The purpose is to detect aspeed, N, (Figure 12.6), where a locked rotor conditionmay occur during a normal speed-up period. This situationmay arise due to a voltage fluctuation, a sudden excessivetorque or a load demand, etc. At the preset speed, theswitch contacts open to energize a timer. If the motor isnot able to pick up within the motor thermal withstandtime, the timer will operate and actuate the tripping circuitin conjunction with the locked rotor protection. When amotor is statically controlled, this device is used for speedcorrection through the feedback control system (Section6.6).All these devices and their likely locations are indicatedin Figure 12.39(a). The sizes of control cables to connectthese devices are indicated in Table 12.3.12.9 Switchgears for motorprotection12.9.1 LT motorsThe general arrangement of a motor’s protectionswitchgear (MCC*) is shown in Figure 12.43(b). Thistakes into account the protection of the main equipmentsuch as motors and cables and makes provision forisolation of the outgoing circuits from the incoming supply.The single line diagram for this switchgear is shown inFigure 12.43(a). The overcurrent relays are selected, asfar as possible, with thermal characteristics to those ofthe motors. The overload setting should preferably bewithin 25-75% of the relay range.For small motors with a number of brands and varyingthermal characteristics the above may not be practical.Moreover, to arrange the thermal curves for each relayand motor and then match them individually for closerprotection may also not be practical. The practice adopted* MCC - motor control centre. For details refer to Chapter 13.Table 12.3 Recommended sizes of cables for various sensingand monitoring devices(A) LT motorsNumber of cables and sizes(i) PTC, thermistors or embedded 12 x 2.5 mm2, 65011100 Vresistance temperature detectors(RTDs): for 6 sets of copperconductor cables(2 cores for each set)(ii) PTC, thermistors or resistance 4 x 2.5 mm2, 650/1100 Vbearing temperature detectors(RTDs): for 2 sets(iii) Moisture detectors:2 x 2.5 mm2, 650/1100 VFor each detector(iv) Space heaters:2 x 2.5 mm2, 650/1100 VFor each space heater(v) Speed switch:2 x 2.5 mm2, 650/1100 VFor each switch(B) HT motorsGenerally the same as for LT motors, except that therecommended size of cables will be 6 mm2Note Wherever a cable lead connecting the above devices (suchas for RTDs) has to pass through a magnetic field, it may be screenedwith tinned copper-braided wires to nullify the effect of stray fields.The field may distort the readings.universally, therefore, is to select the most appropriaterelay, within the required range, from among the rangesand brands available in the market for different ratingsof motors. Table 12.4 suggests the recommended relayranges, sizes of switches, fuses and cables for differentsizes of motors. This selection is only for general,guidance. For a more detailed approach, refer to the notesto the table.12.9.2 HT motorsWith the availability of 3.3 and 6.6 kV vacuum contactorsthe control of HT motors up to 6.6 kV systems has nowbecome easier and economical, compact and even morereliable. For 11 kV systems, vacuum as well as SF6(Sulphur hbxafluoride) breakers can be used. The HTmotor’s switching and protection through a vacuumcontactor provides a replica of an LT system. The earlierpractice of using an HT OCB, MOCB, or an air blastcircuit breaker for the interruption of an HT circuit isnow a concept of the past.The two comparatively new type of breakers, vacuumand SF6 are exceptions and have gained favour in viewof their reliability and durability. For details on thesebreakers, see Sections 19.5.5 and 19.5.6, which also dealwith their switching behaviour and phenomenon of arcreignition. Figures 12.44 and 12.45 illustrate typical powerand control circuit diagrams respectively, for a 6.6 kVbreaker-operated motor starter.The latest practice for up to 6.6 kV motors is to use anHT load break, fault-making isolator in conjunction withthe appropriate type and size of HRC fuses, a vacuumcontactor and a motor protection relay. The contactors

IIIIProtection of electric motors 12/309also provide a three-phase simultaneous make and breakof contacts through their ‘no volt’ coil. Figures 12.46and 12.47 illustrate typical power and control circuitdiagrams, respectively, for a 6.6 kV vacuum contactoroperatedmotor starter.- Bus barsContactorOvercurrentrelayII Motor controlcentre/I /’1’HT isolatorsThe isolator should be a load-break and fault-make typeof switching device. Normally, it is provided with anintegral grounding switch, such that in the isolated positionit will automatically short-circuit and ground all the threephasesof the outgoing links. This is a basic safetyrequirement for a gang-operated isolator, mounted onoutdoor poles and operated manually from the groundby a disconnecting lever. For an indoor switchgear,CableIsolator\MotorP.B.Note1. Switch-fuse unit or a circuit breaker.2. Isolator near the motor is also recommended, when the motoris away or not visible from the switching station (it is a safetyrequirement)3. A stay put type stop push button may also be mounted nearthe motor for maintenance safety-GFigure 12.43(a) Power line diagram of an MCCIsolatinglinkI ShortinaI links -1 - 1 7 ) 16111181191Pre trip alarmStallingalarm/tripRemote thermal-Alarm Trip Control supplyisolating contact contact contactlinkFigure 12.43(b) MCC for single line diagram of Figure 12.43(a)Figure 12.44 A simple power circuit diagram for a breakeroperatedHT motor starter

~ ~~~LT12/310 Industrial Power Engineering and Applications HandbookTable 12.4 Selection table for switches, fuses, relays and cables for different sizes of LT motorskW HP Approx. IJd3 Switch rating HRC fuses for Over current relay range** Size ofA1 conductor cablesFLC at backup protection ~415 V (I,) DOL For For Y/A - ForDOL For Y/A F o T F F o r TA A A A A Astarter starter0.18 0.25 0.7 0.45 16 16 4 - 0.5-1/0.6-1 -1.5 - --0.25 0.33 0.9 0.52 16 16 4 - O.S-l/n.h-l 1.5 - -0.37 0.5 1.1 0.64 16 16 4 - 1-2/1-1.6 -1.5 - -0.55 0.75 1.5 0.87 16 16 6 - 1-2/1.5-2.5 -1.5 - -0.75~.1.0 2.0 1.15 16 16 6 - 1-2/1.5-2.5 -1.5 - -1.1 1.5 2.6 1.50 16 16 6 - 1-5-Y2.54 -1.5 - -1.5 2 3.7 2.14 16 16 IO - 2-4/2.s4 -1.5 - -2.2 3 4.9 2.84 16 16 16 6 3-6/4-63 1.5-3/2.54 1.5 1.5 1.53.7 5 7.8 4.50 16 16 16 10 6-12/6-10 3-6/4-62 1.5 1.5 1.55.5 7.5 11.3 6.50 25 16 25 16 6-1219-14 4816-1 0 2.5 1.5 1.57.5 IO 15.5 8.95 25 25 25 25 10-16/13-21 6-1 2/6-10 4 2.5 1.59.3 12.5 19.0 11.00 63 25 35 25 12-24113-21 6-1 2/9-14 6 4 2.5II 15 22.0 12.70 63 25 50 25 12-24/20-32 6-1 219- I4 6 4 2.515 20 29 16.80 63 63 63 35 16-32/20-32 12-24/13-2 I 10 6 418.5 25 35 20.20 63 63 63 50 244512842 12-24/20-32 I6 10 622 30 40 23.20 63 63 63 50 24-45/3045 16-32/20-32 16 16 630 40 54 31.20 100 63 100 63 33-63145-70 24-45/2842 25 25 1037 50 66 38.20 125 100 125 80 50-90160-100 2445/2842 3.5 25 1645 60 80 46.20 250 100 160 100 50-90/60-100 32-63/45-70 50 35 2555 75 100 57.70 250 250 200 125 70-110/90-150 32-63145-70 70 70 2575 100 133 76.80 250 250 250 160 90-135/90-150 50-90/60-100 95 95 5090 125 165 95.50 250 250 250 200 140-170/120-200 70-110/90-150 150 150 70110 150 196 113.00 400 250 355 200 180-300 90-135/90-150 185 185 95125 170 220 127.00 400 250 400 250 180-300 120-155/90-150 240 240 120160 220 285 165.00 630 400 500 320 2nn-400 120-200 400 400 150180 245 315 182.00 630 400 500 320 280400 120-200 400 400 185200 270 350 205.00 630 400 630 355 280-400 180-300 2x 185 2x 185 240220 300 385 222.00 630 400 630 400 350-500 180-300 2 x240 2 x 240 240*The recommended ratings are for general guidance only. For more accurate selection of fuses, it is advisable to check these ratings by comparingtheir I* - t characteristics with those of the selected OCR (see Section 12.4. I).**The overcurrent relay range may vary slightly from one manufacturer to another.Notes1 It is recommended that relays beyond 100 A be CT operated for better accuracy and stability.2 For drives with longer starting times, relays suitable for heavy-duty starting, i.e. CT operated, must be employed (Section 12.4.1).3 In YlA starting, the relay is connected in phases. Its rating is therefore considered as I,/& i.e. 58% of I,. With drives having longer start times,this relay may trip during a start. Provided that the motor thermal rating permits a current up to 21, (approx.) in the star position, for the durationof the start the OCR may be connected on the line side with a range corresponding to 1, This is times more than the range in the phaaea andwould take nearly (&)’, Le. three times longer to trip than when it is used in the phases.4 The sizes of cables considered above refer to normal operating conditions and a distance of around 10/15 metres between the starter and themotor.5 (i) For slip-ring motors, the selection of components may be made Corresponding to a Y/A switching.(ii) With the short-circuiting and brush-lifting device, the cable on the rotor side, when rated for only half the rated rotor current, will be adequate,as it has to carry the current during start only.however, it is not so essential, since the switchgear itselfis adequately protected for a ground fault. Generally, allmanufacturers provide such a built-in feature as standardpractice, irrespective of its application. For a capacitorbank also, if installed in the circuit, such an isolating andgrounding switch is essential to ground the capacitorbanks on a switch-OFF. Figure 12.48 shows a typicalHT isolator.HT HRC fusesAs discussed in Section 17.7, during a start HT motorsmay encounter severe switching surges on the system.The starting inrush current on such a system can thereforehave a momentary very high peak, similar to and evenmore severe than that shown in Figure 4.5, dependingupon the type of interrupter being used and the surgeprotection scheme, if provided. Unlike LT motors, oneshould therefore take cognisance of the switching surgephenomenon, in addition to the thermal rating demand,to sustain the repeat start surges while selecting the fuserating rur the switching of an HT motor.The following points may be borne in mind whileselecting the correct rating of the fuses:

Protection of electric motors 12/31 1+ swTI32 /C86I$3fOIC trip*MPR: Motor protection relaySi 4positionFigure 12.45 Typical schematic and control wiring diagram for power circuit of Figure 12.44Starting time of the motorNumber of consecutive startsNumber of equally spaced starts per hourConsider the rating of the fuse at about twice the ratedcurrent for a DOL starting and about 130% for anA/T starting. The rating will depend upon the I* - tcharacteristics of the HRC fuses and must ensure thatthey remain intact while handling short-durationexcessive current during normal operation, such asduring a start, and clear a fault condition quickly.Current-Time (I2 - t) characteristicsThe typical current-time characteristics for 50-630 A,6.6 kV fuses are shown in Figure 12.49. The cut-offcurrent characteristics are given in Figure 12.50.Method of selectionThe manufacturers of HRC fuses provide a selectionchart to help a user to select the proper type and size offuses. For the fuses covered in Figures 12.49 and 12.50we have also provided a selection chart (Figure 12.5 I),for motors having a starting time of not more than 15seconds. Similar charts from other manufacturers canalso be obtained for different ratings of fuses and differentstarting times of motors.Vacuum contactorsNormally a vacuum contactor (Figure 12.52) is preferredto an ordinary air break contactor for the following reasons:Since the contacts make and break in vacuum there isno oxidation of the contacts. The contacts thus do notdeteriorate and provide a longer working life.Dielectric strength of high vacuum is approximatelyeight times that of air (Figure 12.53).The arc energy dissipated in vacuum for a giveninterrupting current is one tenth of that in oil (Figure12.54).It calls for less maintenance.Very high contact life, say, over 1.5 million operations,which is more than enough for lifetime operation.High operating rate suitability, say 1000 or moreoperations per hour. This makes them suitable even

12/31 2 industrial Power Engineering and Applications Handbooklsolatosw6.6 kV 3 Ph 50 HzR Y BcT4, - Auxiliary supplyfail alarmlinkStallingiRemote thermalTIAlarm Trip ControlIsolating contact contact supplylinkcontactFigure 12.46 A simple power diagram for a 6.6 kV motor controlwith an isolator, contactor and a motor protection relayfor applications requiring frequent switchings, brakingsor reversals etc., and gives them a distinct advantageover the conventional type of contactors.For more details on vacuum interrupters see Section 19.5.6and Table 19.1.12.1 0 Selection of main componentsThe selection of main components such as switches andcontactors is made on the basis of theirContinuous current rating (CMR): to carry the circuitcurrent continuously.Thermal capacity: to perform the required switchingduties and sustain the fault conditions, at least up tothe cut-off time of the short-circuit protective device,say, the HRC fuses.Electrical and mechanical life: which is defined bytheir AC duty. As in IEC 60947-4-1 for contactors andIEC 60947-3 for switches. It is noted in Table 12.5.NoteI The type of duty defines the capacity of a switch or a contactorby the value of the current and the p.f. of the associated circuit,it can make or break on fault. For values of currents and p.f.for different duties, refer to the relevant standards as notedabove.2 There are a few more utilization categories. For details refer toIEC 60947-4- 1 .In fact, the same contactor or switch can perform differentduties at different thermal ratings and have correspondingelectrical* lives.For instance, an AC3 duty contactor when performingthe duty of a resistive load can carry a higher load and itsnormal rating can be overrated. Similarly, when performingthe duty of AC4, it can carry a lower load and will requirederating. Accordingly, its electrical life will also beaffected. As standard practice, such ratings are prescribedby the manufacturers in their product catalogues. Basedon the above and the discussions so far, we have providedin Table 12.4 the recommended ratings of switches, fuses,relays and sizes of aluminium cables for different motorratings. These recommendations will generally provideprotection consistent with coordination type 2 as in IEC60947-4-1. From this table a quick selection of thecomponents can be made for any rating of LT motors.For HT motors, the protection is specific and must bedetermined on a case-by-case basis. The components soselected and backed-up with overload and short-circuitprotections will ensure thatThey will make and break, without damage, all currentsfalling even outside the protected zone of a thermalovercurrent relay or the built-in overcurrent (o/c) andshort circuit (sk) releases of a breaker, but within theprotected region of the HRC fuses, as illustrated bythe hatched portion, of the overcurrent and short-circuit,I* - t curves (Figure 12.55).They will withstand the let-through energy (Z2 . t) andthe pcak let-through current of the relays/releases orthe fuses, when making or breaking the circuit on fault,without damage or welding of the interrupting contacts.The basis of selection of these components is brieflydescribed below.12.10.1 Switches and contactorsThese are selected so that they will sustain without damageto its contacts or to any other part the motor switching*Mechanical life is independent of current and does not depend onload, duty or application.

Protection of electric motors 12/31 3-240V ACmtrol sup(:sw\CFICF2CFICF;LI--Lock relay OUT--_)IIIIIII-'e1h4-* To motor forwinding andperature etc.DC switching and (protection controlcircuit-I-rol. supplyhealthyThermistorSpacerelay circuitheatercircuitAC control circuit forauxiliariesFigure 12.47 Typical schematic and control wiring diagram for power circuit of Figure 12.46inrush current (ZSJ and possess a thermal capacity ofmore than the fuse let-through energy (Figures 12.56and 12.57).12.10.2 Breakers (ACBs or MCCBs) (Figure 12.58)While a fuse-free system is gradually gaining preference,for all ratings, as discussed later, it is recommended thata breaker (ACB or MCCB) be employed for large motorsof at least 300 h.p. and above to ensure better protectionfor the motor.Figure 12.48 HT isolating switch (Courtesy: Siemens)12.10.3 Overcurrent relay (OCR)It is necessary to coordinate the HRC fuses with theovercurrent relay to ensure that during a fault the relay iscapable of withstanding the let-through energy of thefuses without damage (Figures 12.11 and 12.15).

12/314 Industrial Power Engineering and Applications Handbook2010t1-Symmetrical prospective current (rms) AmperesFigure 12.49 Typical time-current characteristics of 6.6 kV HRC fuses12.10.4 HRC fusesThe rating of the fuses often decide the rating of theabove components. They are readily available up to 1000A to switch all sizes of motors recommended on an LTsystem. The characteristics of the fuses must match withthe thermal withstand Z2 - t characteristic of the equipmentor the circuit it has to protect, as discussed earlier. Broadlyspeaking, they must prevent the switch or the contactorfrom breaking currents beyond their thermal capacityand prevent contact welding. The minimum rating of thefuses is chosen as 1.6 to 2 times the full-load current ofthe motor for a DOL start and about 1.25 to 1 .SO for aYlA start to ensure that the fuses stay intact withoutinterrupting the motor during a start. One particular brandof HRC fuses has been considered here in making theselection. The ratings may slightly vary with the otherbrands of fuses, depending upon their I’ - t characteristics(Figure 12.59).1 .o1 .o 5 10 T 150 10030 43.8Prospective current +kA (rms symmetrical)lllustra tionA 100 A fuse will cut-off an I,, of 30 kA (peak value up to2.5 x 30 kA) at less than 6 kA peakFigure 12.50 Typical current cut-off characteristics of 6.6 kVHRC fuses at prospective currents up to 43.8 kA12.10.5 Selection of cablesThis is also based on the full-load current. A cablegenerally has a much higher thermal capacity than thecut-off time of the HRC fuses of a corresponding rating.Also refer to Z2 - t curves for cables in Chapter 16,Appendix 1, Figures A16.4(a) and (b) and cable selectioncriterion as in Figure AI 6.5. It is, however, advisable tocheck the fuse let-through energy, which should not bemore than the short-time rating of the cable. The followingmay also be kept in mind while selecting the cables:

Protection of electric motors 12/315tFigure 12.51 Fuse selection chart for a 6.6 kV system for motors with run-up time not exceeding 15 seconds2 1a51 - Auxillary switch2 -Closing solenoid3 - On/off indicator4 - Mechanical off-push button5 -Operation counter6 - Mechanical latching device7 -Trip coil8 - Metallic base frameFront view1 - Upper contact terminal2 - Epoxy cast armature assembly3 -Vacuum interrupter4 - Lower contact terminal5 - Epoxy resin moulded bodyRear viewFigure 12.52 Vacuum contactor (Courtesy: Jyoti Ltd.)

12/31 6 Industrial Power Engineering and Applications Handbook250 I I I I I I6005001 400sEA 300ga,L! 200T-0 5 10 15 20Contact gap (mm)Figure 12.53 Dielectric strength of various media as a functionof contact gap1000-10 20 300Symmetrical breaking current (kA)Figure 12.54 Comparison of arc energy produced by medium-voltagecircuit breakersL, JCurrent- t tBreaking capacityof the breakeri Fault level ofthe systemOn faults exceeding /sc of breaker, /*Rt (fuses) < /*Rt (breaker)Figure 12.55Coordination of fuses with a breakerFigure 12.56Switch disconnector fuse unit (Courtesy: L&T)Whether the installation of the cable is in air, a duct orin ground. This will determine the type of cable required,i.e. armoured or unarmoured.Ambient temperature.Ist and its duration.Number of power cables running together and theirconfiguration. For more details refer to Chapter 16,Appendix 1. The cooling of the cables is affected bythe number of cables and their formation. This detailis also provided in Appendix 1. For more details consultthe cable manufacturer.Length of the cable from the starter to the motor. Thiswill help to determine the voltage drop from the starterto the motor terminals during a start. It must be limitedto only 2-3% of the rated voltage because the incomingreceiving point voltage itself may already be less thanthe rated. This is illustrated in FigureA16.3. When thecumulative effect of all such drops exceeds 6%, it may

Protection of electric motors 12/317Table 12.5 Duty cycles for a contactor or a switchSerial no. AC duty ApplicationFor contactorsFor switches1 AC 1 AC21Nearly resistive switching such as heaters, resistance furnaces and lighting loads etc.2 AC2 AC22 High-resistance and low-inductance switching such as a slip-ring motor switching3 AC3 AC23 High-inductive switching such as the switching of squirrel cage motors and inductors.Occasional inching and plugging operations, such as during start-up period4 AC4* -* Stringent inductive switching, such as the switching of a squirrel cage motor withinching and plugging operations (Section 6.20.1(B))5 AC6b AC23 High capacitive switching such as capacitor banks*Applicable only for contactors. A switch is neither required nor suitable to perform a duty such as inching or plugging.tend to destabilize the distribution system and influenceother feeders connected on the same system.A heavy drive, requiring a prolonged starting time,may require larger fuses. In which case the cable sizemay also be increased accordingly.In some cases, where the nearest rating of the fuseitself is too high for the rated current, a larger cable isrecommended. The thermal (Z2 - t) characteristics of allsuch components will vary from one manufacturer toanother and may not be readily available with a designor a field engineer, while making the selection. Themanufacturers of such components therefore as standardpractice, perform this coordination for their products andmake such data readily available for the user to make aquick selection. It may be noted that OCR and fuses atleast, of different brands, will require a new coordination.12.1 1 Fuse-free systemFigure 12.57 Air break contactor (Courtesy: L&T)Fuses are prone to cause a single phasing by not operatingall three of them simultaneously. They may also requirea longer downtime to replace. Therefore, the new concept.-Air circuit breaker (ACB) (Courtesy Siemens)Figure 12.58Moulded case circuit breaker (MCCB)(Courtesy: GE Power Controls)

la318 Industrial Power Engineering and Applications Handbook1000100t 108P0.$1.?9gFigure 12.59 HRC fuses (Courtesy: L&T)On a fault, they would also trip long before the prospectivepeak of the fault current, and have characteristics similarto Figure 12.17 and limit the let-through energy (Z2Rt) ofthe fault. Although a costlier proposition, they eliminatethe above deficiencies of a fuse and ensure a three-phaseinterruption, on the one hand and a prompt reclosing,after a trip on fault, on the other. The contactor, however,may still be essential in the circuit to permit frequentswitching operations. A separate OCR may also beessential to closely match the thermal characteristics ofthe motor, as the characteristics of the releases of theMCBs or MCCBs may underprotect the motor. Such anexercise may also be time consuming initially, as themanufacturers of MCBs and MCCBs may, at best, providetheir characteristics and coordination with their brand ofcomponents. The coordination of MCBs and MCCBswith components of other brands may have to be doneindividually, by the design engineer, more so initially,until all in the field become acquainted with thisphilosophy and characteristics of all brands of MCBsand MCCBs.12.11.1 Motor protection circuit breakers (MPCBs)A few manufacturers have provided an improvised versionof the above in the form of a motor protection circuitbreaker (MPCB). This can match closely the thermalcharacteristics of a motor. In this case one MPCB will besufficient to replace the HRC fuses and the thermal relayand a separate OCR may not be necessary. Since therange of MPCBs is limited presently, so are the motorratings that can be protected. With the use of MPCBs noindividual component matching will be necessary as inthe above case. Figure 12.60 illustrates a typicalcharacteristic of an MPCB. Notice that the basic differencebetween an ordinary OCR and an MPCB is the withstandcapability of MPCB on a fault. An MPCB is now able totake care of fault currents and switching transient currentsand isolate the circuit much more quickly on an actualfault. An ordinary relay cannot withstand fault currentsand requires backup protection through HRC fuses.The characteristic shown also suggests that the MPCBscan withstand current transients up to 100 times the ratedcurrent (or the current setting) and hence are capable ofswitching a capacitor and protecting them againstoverloads.0t/,I = 71, LIIranse"1 UP to 2 x 1stIr Current "'- (= 14/,-201,) during anopen transient conditionFigure 12.60 Coordination of MPCB characteristics with themotor characteristics (eliminating the use of HRC fuses)that is gaining preference is a fuse-free system, particularlyon an LT system. It is possible to achieve such a systemthrough miniature and moulded case circuit breakers(MCBs and MCCBs) designed especially to match theZ2 - t Characteristics of a motor. They are also designedto be fast acting and current limiting, like HRC fuses.12.11.2 Component ratingsIn both the above cases, besides saving on the cost of theswitch and fuses, one can also economize on the cost ofother equipment, such as main power cables, contactorand the internal wiring of the starter (particularly fromthe busbars to the MCB, MCCB or MPCB). The ratingof all such equipment can now be lower and commensuratewith that of the MCB, MCCB or MPCB and, hence, canbe very close to the full load current of the motor andthus economize on cost. In conventional fuse protection,their rating was governed by the rating of the fuses, andthe fuses had to be of higher rating than the rated currentof the motor to remain immune from momentary transientconditions and also to allow for a minimum fusing time

Protection of electric motors 12/319Relevant StandardsIEC Title IS BS60034-1/199660034-1 1/197860072-11199160072-2/199060072-3/ 199460298/199060470/197460644/1979607511198360892/198760947/199960947-1/199960947-3/199860947-4- 11199660947-5-1/1997Rotating electrical machinesRating and performance for rotating machinesRotating electrical machines. Built-in thermal protection, rules forprotection of rotating electrical machinesDimensions and output series for rotating electrical machines. Framenumber 56 to 400 and Flange number 55 to 1080Dimensions and output series for rotating electrical machines. Framenumber 355 to 1000 and Flange number 1180 to 2360Dimensions and output series for rotating electrical machines. Smallbuilt-in motors. Flange number BF 10 to BF 50A.C. metal-enclosed switchgear and controlgear for rated voltages above1 kV and up to and including 52 kVA.C. contactors for voltages above 1 kV and up to and including 12 kVSpecification for high-voltage fuse-links for motor circuit applicationsIndustrial platinum resistance thermometer sensorsEffects of unbalanced voltages on the performance of 39 cage inductionmotorsSpecification for low-voltage switchgear and controlgearLow-voltage switchgear and controlgear. General rules and testrequirementsSwitches and SFUSIFSUSContactors and motor starters. Electromechanical contactors and motorstartersControl circuit devices and switching elements - electromechanicalcontrol circuit devices4722/1992 BS EN:60034-1/1995- BS: 4999-1 11/1987123 111991 BS: 5000- 10/1989BS: 4999-141/19871231/1991 BS: 5000-10/1989BS: 4999-103/1987996/1991 BS: 5000-1 1119893427/1991 BS EN: 60298/19969046/1992 BS: 775-2/1984- BS EN: 60644/1993- BS EN: 60751/1996- BS EN: 6094713947-111993 BS EN: 60947-11199813947-3/1993 BS EN:60947-3/199213947-4-1/1993 BS ENz60947-4-11199213947-5-1/1993 BS EN : 60947-5-1/1992Relevant US Standards ANSUNEMA and IEEEANSI C.37.42/1989 Standard on distribution cut out and fuse linksANSI C.37.47/1992 Specifications for distribution fuse disconnecting switches, fuse supports and current limiting fusesNEMAMG 111993 Motors and generators ratings, costruction, testing and performanceNEMA-AB111993 Moulded case circuit breakers and moulded case switches up to 1000 V a.c. Rating more than 5000 ANEMA-AB3/1996 Moulded case circuit breakers and their application, up to 1000 V a.c. Rating 5000 A and moreNEMAIFU-111986 Low-voltage cartridge fusesNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.during the switching operation of the motors.List of formulae usedVoltage unbalance (negative phase sequence)Voltage unbalanceMax. voltage variation from- the average voltagex 100%Average voltageStator current on an unbalanced voltage(12.1)I, = vr (12.2)R, +R; +-xR; +[X, +SXSSX;]~S l2and

12/320 Industrial Power Engineering and Applications HandbookCumulative rotor current(12.10)I, = negative phase sequence current componentVu = unbalanced voltageR; = rotor resistance referred to statorX; = rotor reactance referred to statorStator heat on an unbalanced voltage(a) Maximum theoretical heatHe, (max.) = (If + I,’ + 21, x I, )(b) Minimum theoretical heatHe, (min.) = (I: + I: + I, x I, )Actual heat generatedHe, = (I: + 61,’)Actual current on unbalanceI,, = JrnRotor power during an unbalance(12.3)(1 2.4)(12.5)(12.6)(1 2.7)(12.8)I,, = positive sequence current in the rotor circuit, and12- = negative sequence current in the rotor circuitTotal rotor heat0~ (I; + 3I:” )I,, = negative sequence rotor current(12.9)Further readingI Beeman, D., Industrial Power Systems Handbook, McGraw-Hill, New York (1955).2 Dommer, R. and Rotter, N.W., ‘Temperature sensors for thermalover-load protection of machines’, Siemens Circuit, XVII, No.4, October (1982).3 Ghosh, S.N., ‘Thermistor protection for electric motors’, SiemensCircuit, XI, No. 3, July (1976).4 Kaufmann, R.H. and Halberg, M.H., System over- voltages,causes und protective measures.5 Kolfertz, G., ‘Full thermal protection with PTC thermistors ofthree-phase squirrel cage motors’ , Siemens Review, 32, No. 12( 1965).6 Lythall, R.T., AC Motor Control (on earth fault protection andthermistor protection).7 Ramaswamy, R., ‘Vacuum circuit breakers’, Siemens Circuit,XXIII, April (1988).Source materialUnbalanced Voltages and Single Phasing Protection,M/s Minikc Controls Private Ltd.High Voltage HRC Fuses, Publication No. MFG/47, The EnglishElectric Co. of India Ltd.Surge Suppressors-Bulletin No. T-109, Jyoti Ltd.Thermistor Motor Protection Relay Catalogue-SP 50125/5482,Larsen and Toubro.Motor Protection Relay, English Electric Co. of India I.td, CatalogueRef. PR. 05:101:A/6/85.Negative Phase Sequence Relay, M/s English Electric Co. of IndiaLtd, Catalogue Ref. PR:01:306:A.Thermal Bimetallic Overload Relays, Bhartiya Cuttler Hammer.Product Catalogue C-305 MC 305/ANA3/6/85.

Ai321AppendixRules of Thumb forEvery-day Use

Appendix A1323Power requirements for pumpshp =Alternatively.USGPM H, p4000 qIGPM Hb p3300 qwhereUS GPM = discharge in US gallons per minuteI GPM (UK) = discharge in imperial gallons per minuteH, = head (in feet)= suction head + static delivery head +frictional head (105s) in pipes and fittings+ velocity headFor determining the frictional head, refer to friction lossin pipes. bends, elbows and reducers and valves asprovided in Tables A. 1 and A.2:p = specific gravity of liquid in g/cm3q = unit efficiency of pumpLPS H, . por h.p. =75 ' qLPS = discharge in litres per secondH,, = head (in metres)Friction loss in pipesTable A. 1 provides. for a particular rate of discharge inGPM. the friction loss in pipes for every 100 feet ofstraight pipe length, reasonably smooth and free fromincrustation.Friction loss in bends, reducers, elbows etc. are providedin Table A.2 in equivalent pipe length.To determine the size of pipeThe economics would depend upon the smoother flowof fluid without excessive friction loss. A smaller sectionof pipe may not only require a higher h.p. for the samesuction and lifting head due to greater frictional losses,but may also cause the pipe to deteriorate quickly as aresult of the additional load on its surface. Losses due tobends and valves should also be added in the total frictionloss.ExampleConsider a discharge of, say, 20 LPS against a total suctionand delivery head of 150 m through a mains 1000 m long.Considering an average of 25 bends, elbows, tees and reducerfittings in the total length of pipe, then from Table A.1.Total equivalent length of pipe (assuming that every fittinghas an average 5 m of equivalent pipe length, to account forfriction) = 1000 + 25 x 5= 1125 mA 125 mm pipe has a friction loss of nearly 33 m per 1000 mand a 150 rnm size of pipe, 13 m per 1000 m.:. Total frictional head for a 125 mrn pipe = E x 33= 37.125 mand for a 150 rnm pipe -- - x 131000= 14.625 rnA friction loss of 37.125 m in a total length of 1000 m is quitehigh and will require a larger motor. Therefore, a 150 mrnmain pipeline will offer a better and more economical designcompared to a 125 mrn pipeline such as the reduced cost ofthe prime mover and lower power consumption during thelife of pumping system, in addition to a longer life span of a150 mm pipe compared to a 125 mm pipe.Power requirements for lifts(i)For linear motion drivesWhere the weight of the cage and half of thepassengers load is balanced by the counter weightP=0.746 x W x V2.75 x qwhereP = kW requiredW = passengers load in kgV = speed of lift in m/sq = unit efficiency of the drive.(ii) For rotary motion drivesUsing equation ( 1.8).p=- T.N974 ' qwhereP = kW requiredT = load torque in mkgN = speed of drive in r.p.m.q = unit efficiency of the drivePower requirements for fans(A.2)(A.3)p=- LPS p ' (A.4)75. qwhereP = kW requiredLPS = quantity of air in litres per second.p = back pressure of air at the outlct in metres of acolumn of water.Note These are actual power requirements for various applications.Add 10-154 to these figures to select the size of the motor toaccount for unforeseen lowx during transmi\\ion from motor toload and other frictional losscs. Too large a motor will give a poorpower factor and a poor efficiency. while too sinall a six will runoverloaded (Section 1 .E). These concideration\ must he kcpt inmind while selecting the motor rating.

~ ~ ~~ ~~ ~~~ ~~~~~~~~~~~ ~~~~~~~~~~~~~~N324 Industrial Power Engineering and Applications HandbookImportant formulaeMoment of inertiaGD2 = 4 . g. (m?)= 4 wr2whereW = m . g (mass x gravity)r = radius of gyrationTemperatureConversion from Fahrenheit (OF) to Celsius ("C)-- F-32 - C(A.6)9 5whereF = "Fc = "CAbsolute zero = kelvin (1 K)1 K = -273.15"CTable A.6Some useful units1 Pa (Pascal) = 1 N/m21 bar= IO5 N/m'= 1 atm.11 dtm (atmoqphere) = I kgf/cm2= 735 6 mm of mercury= 101325 x 105N/m21 bar =lo5 pascal = 0 9870 atm1I01325 x 10' N/m'1 torr =760- 1- 760 atm1Unit oj"PressurePressurePreqrurePreswreAbsolute zero is the theoretical temperature, at whichthe atoms and molecules of a substance have the leastpossible energy. This possibly is the lowest attainabletemperature.Conversion tableLengths1 inch = 25.4 mm1 foot = 30.48 cm1 mile = 1.609 km1 cm = 0.3937 inch1 metre = 39.37 inches= 3.28 feet1 km = 3280 feet= 0.621 mileAreas1 in2 = 6.45 16 cm21 ft2 = 0.0929 y21 cm2 = 0.155 in-1 m2 = 10.8 ft21 m2 = 1.196 yd2Volumes/weights*I Imperial gallon (UK) = 4.546 litres1 N = 0 101972 kgtForce*1 US gallon = 3.79 litres1 pint = 0.568 litre1 kg f = 9 807 NForce1 litre = 0.22 Imperial gallons1 Ib = 0.453 kg1 kg tm = 9 807 N m Energyltorque1 kg = 2.204 Ib1 MT = 0.984 ton1 kg f/m2 = 9 807 N/m2Force~ ~ ~ ~ ~ _ _ _ 1 T = _ 1.016 MT _ _lJ=1Nm=OI0187kg fmEnergyhorque1 litre per second = 13.2 gallons per minute1 ft' of water = 6.23 gallons1 W = 1 Us Power1 rn3 of water =220 gallons = 35.31 ft31 ft3/s = 22,500 gallons/h(GPH)1 Wb=lV.s Flux= 375 gallons/minute (GPM)1 atmosphere = 30" of mercury1 T = 1 Wb/m2 Flux density= 14.7 lb/inz1 Hz = 1 Hz (s-I) FrequencyTorqu egorce1 cm.kg = 0.866 inch pounds1 m.kg = 7.23 foot poundsWork done1 kW = 1.36 metric h.p. (PS)= 1.341 h.p.1 h.p. = 0.746 kW= 1.014 PS (metric h.p.)1 PS (mhp) = 0.736 kW= 0.986 h.p."Unless specified all conversions made earlier relate to Imperialgallons only.

Table A.l(Contd)Pipe diameter in millimerresDisch. 40 50 60 70 80 I00 125 150 200 250 300 350 400 450 500 600 700 800 900 1000 1200VS15 480 185 39 12 4.6 2.1 1 .o

~~~~Table A.l(Contd)Pipe diameter in millimetresDisch. 40 50 60us70 80 I00 125 I50 200 250 300 350 400 450 500 600 700 800 900 1000 12009.0 150 75 229.517083 2510 190 94 2812 270 130 4014 370 180 5516 470 230 7218 600300 9020 720 370 11022 860 450 13524520 16026 620 18728 720 22030 800 25035 330__4045045__505 6070055 85060 100026532097 22 6.5 2.4 1.1 0.55 0.28115 25 7.5 2.8 1.3 0.65 0.35 0.206570__370420140 29 9.0 3.4 1.5 0.75 0.40 0.23163 35 10.9 4.0 1.8 0.88 0.47 0.2775480 185 39 12 4.6 2.1 1.0 0.53 0.30

3.3~~~~Table A.l(Confd)Pipe diameter in millimetresDisch. 40 50 60 70 80 100 125 150 200US250 300 350 400 450 500 600 700 800 900 1000 120014 5.2 2.4 1.1215 5.8 2.7 1.317 6.5 2.9 1.45i: I I 7.2 ~ 1.68.0 3.7 1.80.60.70.770.850.9532 1 11.2 1 5.3 I 2.6 1.442 I 15.5 I 7.0 1 3.5 1.85I I I 4.5 2.510.54 0.22I 1 I 5.7 3.2 1.8 I 0.68 I 0.30 I 0.14 1 I3.8 2.2 I 0.83 I 0.36 I 0.18 I I4.62.7 1.0 0.43 0.22 0.125.63.2 1.2 0.52 0.25 0.156.5 3.7 1 1.4 I 0.62 I 0.3 I 0.17 I7.3 4.2 I 1.65 I 0.70 1 0.34 I 0.19 I8.4 5.0 I 1.85 1 0.8 I 0.4 I 0.22 111.515.06.6 2.6 1.1 0.52 0.288.5 3.3 1.5 0.7 0.38 0.2219.0450 I 1 I l l l l l 11.0 4.2 1.85 0.8 0.48 0.2821.514.0 5.2 2.3 1.1 0.6 0.3528 16.5 I 6.2 1 2.8 I 1.3 I 0.72 I 0.4233 20 7.3 3.3 1.6 0.85 0.50 0.1940I I I I I0.224753 0.760.250.28

Table A.2Friction in fittings in equivalent of pipe length (ft)

Appendix N329Table A.3Table of conversions of gallons per minute into litres per second'GPM us GPM WS GPM ws GPM ws5101213.215202530323540455055606570758085901001051101151201251301351401451501551601651700.3790.7580.9101 .o1.141.521.892.282.52.653.153.413.794.164.554.925.35.686.056.456.827.587.958.338.79.19.459.8510.210.611.011.3511.7512.112.512.917518018519019520021022023024025026027028029030031032033034035036037038039040041042043044045046047048049050013.2513.6514.014.414.815.1615.916.917.418.918.219.720.421.222.022.723.424.225.025.726.427.228.028.729.430.231.031.732.433.234.034.735.436.237.037.95105205305405505605705805906006106206306406506606706806907007107207307407507607707 8079080081082083084085038.639.440.240.941.642.443.243.944.745.546.247.047.748.449.250.050.751.452.253.053.754.455.256.056.757.558.359.159.860.561.362.162.863.564.48608708808909009109209309409509609709809901000105011001150120012501300135014001450150016001650170017501800185019001950200065.266.066.767.468.268.969.770.471.271.972.773.474.275.075.779.083.387.091.094.598.5102.0106.0110.0113.5121.0125.0129.0132.0136.5140.0144.0148.0151.5Imperial gallons (UK)Table A.4 Head of water in feet and equivalent pressure inpounds per square inchHead lb/in2 Head Ib/in2 Head lb/in2(ft) (ft) ( ft)1234567891015202530354045500.430.871.301.732.172.603.033.403.904.336.508.6610.8312.9915.1617.3219.4921.6555606570758085909510011012013014015016017018023.8225.9928.1530.3232.4834.6536.8138.9841.1443.3147.6451.9756.3060.6364.9669.2973.6377.96190200225250275300325400500600700800900lo0082.2986.6297.45108.27119.10129.93140.75173.24216.55259.85303.16346.47389.78433.09Table A.5 Pressure in pounds per square inch andequivalent head of water in feetlb/in2 Head lb/in2 Head lb/in2 Head(ft) (ft) (ft)123456789101520253035402.314.626.939.2411.5413.8516.1618.4720.7823.0934.6346.1857.7269.2780.8192.364550556065707580859095100110120130140103.90115.45126.99138.54150.08161.63173.17184.72196.26207.81219.35230.90253.98277.07300.16323.25150160170180190200225250275300325350375400500346.34369.43392.52415.61438.90461.78519.51577.24643.03692.69750.41808.13865.89922.581154.48

PART IISwitchgearAssemblies andCaptive PowerGeneration

Switchgear andControlgearAssemblies13.4 Design parameters and service conditions for aswitchgear assembly 13/34213.4 1 Design parameters 13/34213.4.2 Service conditions 13/362135Deciding the ratings of current-carrying equipment,devices and components 13/36413.5.1 Awgning a short-time rating 13/364g a bus system 13/368Constructional feature5 of a huService conditions 1 Y370Cumplying with devgn paraembIies 13/372YnYVLllLUl LL'LIU.V.7 ,,I '4 Y.U." ""I '.IRequirements other than constructnterlocking of feederc to prevent phgear assemblies 13/383a switchgear or a controlgear assembly 1313.10 Power circuits and control scheme diagrams 13/38713.10.1 Interlocking and control scheme for a typicair-conditioning plant 1338713.10.2 Diffeient types of starters and instrumentsng procedure of switchgear and controlgeaies and treatment of effluenttment of non-ferrous components

Switchgear and controlgear assemblies 13133513.1 ApplicationThese assemblies are fitted with switching devices(breakers, switches, fuse switches and contactors etc.)and control and measuring instruments, indicating,regulating and protective devices etc. to transform theassemblies into composite units, called control centresto perform a number of functions in the field of distributionand control of electrical power. Some of these functionsmay be one or more of the following:To control, regulate and protect a generator and itsauxiliaries in a power station.To control, regulate and protect the conversion, whennecessary, from one voltage to another, in a generatingstation or a switchyard for the purpose of furthertransmission or distribution of power.Transmission of power.Distribution of power.The basic idea of adopting to such a control system isto broadly accomplish the following in normal operation:1 To have ease of operation and control a group of loador control points from one common location.2 To monitor system operations for better coordinationbetween the various feeders and rapid control of thefeeders.3 To provide a sequential operation when requiredbetween the various feeders or to have an electricalinterlocking scheme between them. For a general idearefer to Figure 13.5 1, illustrating a typical sequentialscheme showing electrical interlocking between thevarious feeders for an air-conditioning plant.Thus the basic purpose of a centralized power orauxiliary control system is to achieve in service:Ease of operation and controlMore flexibilityEase of testing the electrical installationEase of checking the control scheme, if any, on noloadbefore commencing the process.More safetyMore reliability13.2 Types of assembliesDepending upon their application, a switchgear or acontrolgear assembly can be one of the following types:1 Open Type This type of assembly is without anenclosure, as used in an outdoor switchyard or as mountedon a pole, such as a gang-operated switch.2 Metal enclosed type This type of assembly iscompletely enclosed on all sides by sheet metal exceptfor the operating handles, knobs, instruments andinspection windows. The more conventional of them inuse may be classified as follows.13.2.1 Power control and distributionPower control centres (PCCs) (Figure 13.1)These may receive power from one or more sources ofsupplies and distribute them to different load centres,which may be a motor control centre (MCC) or adistribution board (DB), as illustrated in Figures 13.2PanelnumbersFigure 13.1A typical cubicle-type power control centre (Courtesy: ECS)

13/336 Industrial Power Engineering and Applications HandbookControl bus and wire way chamber(shrouded from main bus) 1 Gravity operatedIndications & instrumentsrmounted on an auxiliary doorGroundingj_-fRollermounts onguide rails/+Trolleys racked outFront view of drawn out trolleysRear view of drawn out trolleysFigure 13.2Details of a draw-out motor control centre (MCC) (Courtesy: ECS)and 13.3 respectively. They therefore comprise largerratings of feeders compared to the feeders in an MCC ora DB.Motor control centres (MCCs) (Figure 13.2)These receive power from the PCC and feed it to a numberof load points, the majority of them being motors operatingon an electrical installation or a process line.When there is only one process line and one MCCalone is adequate to control the entire process, it is possibleto combine the PCC and the MCC into one unit to saveon space and cost. The assembly may now be called aPMCC (power-cum-motor control centre).Distribution boards (DBs) (Figures 13.3 and 13.4)These are comparatively smaller assemblies and distributepower to the utilities of an installation, which can be anindustrial, a commercial or a residential complex. Theutilities may be one or more of the following essentialservices:Lighting loadsCooling and heating loadsWater supplies (pump sets)Firefighting (pump sets)Lifts and escalatorsWelding sockets etc. for future maintenance of theinstallation.A DB becomes larger when it serves a residential colony,a multi-storey building or a shopping complex where themain loads are of utilities only.13.2.2 Auxiliary controls and monitoring of aprocessSuch control assemblies are designed to programme thecontrol of a process line and also to monitor itsperformance through audiovisual indications or a tripcommand. They receive control cables carrying thedifferent process signals from the MCC, drives and theprocess control devices (such as temperature, pressure,flow and speed-measuring devices) to actuate the requiredaudiovisual indications or a trip command when theprocess-operating parameters fluctuate by more than thepredefined limits. They operate at control voltage anddo not handle any power device. Depending upon theapplication, they can be termed a control panel (CP), amimic panel or a relay and control panel. Figures 13.5and 13.6 illustrate a few such assemblies. (Also refer toSection 16.8 and Figure 16.11). All the above assembliesare floor mounted and free standing, except those that

Switchgear and controlgear assemblies 13/337Figure 13.4 A non-compartmentalized distribution board-term i n a I sFigure 13.3 A typical cubicle-type fully compartmentalizedpower distribution board (Courtesy: ECS)are too small and which can be mounted on a wall orfixed in a recess.These assemblies may be further classified into singleordouble-front assemblies. In double-front, they may bewithout a maintenance gallery as shown in Figure 13.7,or in a duplex design, with a maintenance gallery betweenthe front and rear rows of panels to provide easy accessat the back of the feeders, as shown in Figure 13.8.In the following text, although we have tried to coverthe types of switchgear assemblies mentioned above,more details have been provided for assemblies that relateto a power-generating station, an industry, or installationswhere use of an electric drive is more common and mayrequire more care. The design of components and devicesmounted in a switchgear or a controlgear assembly isbeyond the scope of this book. The different types of HTinterrupting devices, particularly breakers, being moreintricate are, however, discussed in Chapter 19.13.2.3 PLC-based control panels*Conventional method to activate controls has been through*PLC - Programmable logic controller (a registered trade mark ofAllen Bradley Co. Inc., USA).Figure 13.5 Atypical mimic-cum-process control panel (Courtesy:ECS)

131338 Industrial Power Engineering and Applications HandbGokControlShrouding7 ShroudingbusesFront IFrontFigure 13.6 A non-compartmentalized outdoor-type control panel(Courtesy: ECS)auxiliary relays (contactors). The sensing instrumentsthat are located at the various strategic locations of aprocess line or on the machine feeding the line sense theactual operating conditions continuously and give a signalthrough its auxiliary relay in the event of an error beyondpermissible limits.In a process industry (e.g. chemical, fertilizer, cement,paper, textile, tyre or petrochemicals) such process controlsare numerous, and to closely monitor and control eachthrough the conventional method is arduous and mayimpose many constraints. They also require cumbersome,excessive and complex wiring and a very large controlpanel, as shown in Figure 13.6, for a small process line.There will also be limitations in terms of accuracy,response time and control besides maintenance problems.Thus the method is not suited for critical applications. Arelay may have to activate another before it activates thecorrection and regulates the machine feeding the processline, hence adding to more delay. A contactor operation,particularly, introduces an element of time delay (5 to 10ms each relay) in executing correction because of itsinherent closing and interrupting times. Also there maybe a number of these operating one after another in tandembefore correcting the process. The time delay may affectthe quality of process and may not be warranted. Withadvances in solid-state technology control and feedbackdevices can now be in the form of logical controllers,which respond instantly and can be programmed in adiscrete way, to suit any process needs. A controller isextremely small and can perform the duties of hundredsof such relays and requires only a small space forFigure 13.7 Side view of a double front panel with a commonhorizontal bus, but separate vertical bus bars (not visible) (Courtesy:ECS)mounting. It is wired internally and eliminates all externalwiring. There are no moving contacts and hence nophenomenon of operating time, contact arcing, or wearand tear of contacts.A logical controller is a solid-state digitally operatedcontrol device which can be programmed to follow a setof instructions such as logic sequencing, timing, countingand arithmetic analysis through digital or analogue inputand output signals to respond to the different processrequirements. These may be regulating the speed of themotor to perform inching duties as when a loader craneor an elevator is required to adjust its hoisting heightprecisely at different loading stations or to adjust themovement of a moving platform of a machine tool. Figure6.49 illustrates a typical process line. It can also performdata handling and diagnostic activities by storing dataand messages and also give signals to activate, adjust orstop a process. A logical controller can thus replace arelay-based circuit more efficiently and also offers manymore operational facilities and ease of handling than arelay, and all this instantly, without causing a time delay.It contains three main sections:

I Control busesMain HorizontalBusbarsI IJc‘ESwitchgear and controlgear assemblies 13/339or any other parameter. Or it can be employed ininverter logistics, when it is being used for the controlof the drive itself. Figure 13.9 illustrates a basic logicalscheme.3 Input and output interface This is the interfacebetween the controlling devices and the processor.The input/output (YO) unit receives signals from theinput devices and transmits output action signals tothe controlling devices.NoteA separate unit is used for programming and editing (e.g. a handheldprogrammer or a computer). For on-line editing, keyboardsare used. For on-line monitoring, a PLC is interfaced with a computerand special software.A logical controller is thus a more effective method ofreplacing the auxiliary relays and is capable of performingmany more functions instantly.A process requiring accurate and instant speed controlsmust adopt static motor controls, described in Section6.9, and their control schemes must be activatedthrough programmable logic controllers (PLCs) discussedabove.Figure 13.8 A typical general arrangement of a double frontpanel with a walkway gallery (Courtesy: ECS)1 Central processing unit (CPU) This is in the formof a micro controller and can be called the brain ofthe PLC. It computes and analyses the various datafed into it. It acts like a comparator and makes decisionson the corrective action necessary to fulfil processneeds according to the instructions received from theprogram stored in the memory and generates the outputcommands.2 Memory unit This is the unit that stores the dataand the messages and the diagnostic information. Itstores all the data that defines the process to help theCPU act logically and also stores diagnostic information.It is a part of a computer that containsarithmetical and logical controls and internal memoryand programming devices. The unit receives inputs(temperature, pressure, speed or any information whichmay form a part of the process) from the system. Itthen compares it with the reference data, which isalready programmed into its memory module, analysesit and then sends a corrective signal to the processline devices, controlling temperature, pressure, speedShielding of signalsIt is important to prevent the control signals from becomingcorrupted by electromagnetic interference caused by thepower circuits, particularly those carrying the motorcurrents from the machine to the power-cum-control panelhousing the drive and the PLC. During each switchingsequence, the motor will draw switching currents anddevelop switching voltages. Even small electrostaticinterference may lead to malfunctioning of the drive.Shielded (screened) control cables are therefore recommendedfor this purpose for the control panels’ internalwiring, particularly between the incoming terminals andthe logic controllers (PLCs), cables carrying the TGanalogue or pulse encoder digital corrective signals fromthe field and all sensor and analogue circuits.The shield is generally a layer of copper wires in theform of a spirally wrapped screen around the controlcable. It is grounded at the panel. It is also recommendedto avoid a parallel running of power and signal controlcoaxial cables. Also, for each signal a separate two-corecontrol cable must be used, since common return ofdifferent analogue signals is not recommended. A simplegrounding scheme for the shielded control cables isillustrated in Figure 13.10.13.3 Conventional designs ofswitchgear assemblies (alsoreferred to as switchboards)Depending upon their application, these may have oneof the following construction:Fixed type1 Industrial type as shown in Figure 13.11, or

13/340 Industrial Power Engineering and Applications Handbooki[7?Jtmonitoringoutput(corrective ,-jsignals)-1I -i InputsII --c --LnIln;;r'CPU Memory I/Ointerface. . . . . . . . . . . . . . . . . . . .KeyboardProgrammable controller@ CPU or micro-controller. It is a computing and analysing unit@ Memory module which stores the reference data diagnosticsand the operating programme of the process line@ Input/output interface:It is a relaying unit to receive actual parameters from the fieldand relay out corrective signals@ A sort of electronic switch to protect the controller from control voltagespikes. They may occur because of very high control voltage (in therange of 24 to 220V d.c.) compared to the very low voltage circuit of themicroprocessor, generally in the range of 5V d.c. or lessFigure 13.9 A simple representation of a programmable controllermoduleLight sensitive base, whichbreaks down and startsconducting when the LED glowsLED . I ---_- L-.?-@ Input I y:signal ILOptical isolation> --moduleLight f emissionVAn optical isolation scheme(One for each signal)Screen of cable connected to panel groundScreened cable1 MCCcableDane12Cubicle type which may be(a) In a non-compartmentalized construction as shownin Figure 13.4, or(b) A compartmentalized or modular construction asshown in Figures 13.1 and 13.3.1,-Gencoder = GNote It is recommended to provide separate grounding for electroniccircuitsFigure 13.10 A simple grounding scheme for the shielded controlcablesDraw-out1 Semi-draw-out type, or2 Fully draw-out type, as shown in Figures 13.2 and13.12.13.3.1 Fixed-type constructionIn a fixed construction, all the feeders in the switchboard,feeding the various load points, are securely mounted inthe assembly and rigidly connected to the main bus. Inthe event of a fault in one feeder on the bus side, ashutdown of the entire switchboard may be required. Aprocess industry or critical loads can ill afford such anarrangement. However, since this is the most cost-effectiveswitchboard, it is also the most common type and is usedextensively. It also suih all applications, except a processindustry or critical loads, which may not be able to afforda total shutdown or prolonged downtime in the event ofa fault. In such cases a draw-out type switchboard willprove to be a better choice as discussed below. Afixed-type construction may further be classified asfollows.

BushbarchamberCable boxSwitchgear and controlgear assemblies 13/341will be contained and localized only to the faulty feeder,without spreading to the nearby feeders.Cable boxFigure 13.11 A typical industrial type power distribution boardIndustrial-type constructionIn this construction there is a common bus that runshorizontally and is mounted on vertical floor structures.The feeders are mounted above and below this busbarchamber, as shown in Figure 13.11. Since there are onlytwo feeders in a vertical plane, these switchboards occupya sizeable floor space, but they are rugged and easy tohandle. They are good for very hard use such asconstruction power - i.e. the temporary power requiredduring the construction period of a project - and have toweather severe climatic and dusty conditions. It is possibleto construct them in a cast iron enclosure making themsuitable for extremely humid and chemically aggressiveareas and also for areas that are fire-prone. The use ofsuch assemblies is now on the decline, due to theavailability of better cubicle designs.Cubicle-type constructionThis is in the form of a sheet metal housing, compact indesign and elegant in appearance. The feeders are nowmounted one above the other up to a permissible heightat which the operator can easily operate. It thus makesan optimum utilization of the vertical space and saves onfloor area. They can be further classified as follows:Non-compartmentalized type In this type a groupof feeders are housed in one enclosure, and attendingone would mean an exposure to the others (Figure13.4).Compartmentalized type In this type each feeder ishoused in a separate compartment (module) of its ownand attending one would limit the exposure only tothat unit (Figures 13.1 and 13.3). In this constructiona fault, particularly of the nature of a short-circuit,13.3.2 Draw-out constructionIn this construction each feeder is mounted on a separatewithdrawable chassis as shown in Figures 13.2 and 13.12.In the event of regular maintenance or repairs, they canbe swiftly racked-out or racked-in to their modules withoutdisconnecting the incoming or outgoing power connectionsor the control terminals*. The modules of identical typescan also be easily interchanged and defective modulesreplaced by spare modules in the event of a fault.Downtime is now low. This arrangement is thereforerecommended for all critical installations that require anunintermpted power supply and cannot downtime duringoperation. It is most suited for installations such as powerstations, refineries, petrochemical plants, fertilizers, andsimilar process plants. Similarly, hospitals, airports,railways, etc. are also such critical areas that mayexperience chaos due to a disruption of utilities unlessthe normal supply is restored swiftly. In such places thistype of construction is more appropriate. A draw-outassembly can be designed only in a cubicle constructionand is totally compartmentalized. The two types of drawoutconstructions noted above can be broadly describedas follows.Semi-draw-out typeIn this design the incoming and outgoing power contactsare of the draw-out type, but the control terminals arethe plug-in type. (For plug-in type terminals see Figure13.13 (a)). The plug-in contacts of the control terminalassembly are wired and left loose in the moving trolley,and are engaged manually with the fixed contacts mountedon the frame, after the trolley is racked-in and seated inits place. Similarly, the control terminals are to bedisengaged manually first, when the trolley is to be drawnout.Such a construction is cumbersome and requires utmostcaution to ensure that the terminals are properly disengagedbefore the trolley is racked-out. Otherwise it may pullthe wires and snap the connections and result in a majorrepair. It is also possible that, due to human error, theoperator may slip to engage the terminals at the firstattempt and may have to do it at a second attempt, addingto the downtime, while energizing or replacing a faultytrolley, eventually defeating the purpose of a draw-outsystem.Fully draw-out typeIn this construction the control terminals are of the slidingtype (Figures 13.13(b) and (c)). The moving contacts aremounted on the trolley while the fixed matching contactsare mounted on the panel frame. These contacts engageor disengage automatically when the trolley is racked-inor racked-out of the module respectively. This type of*Without disconnecting the control terminals. This is possible onlyin fully draw-out type construction.

13/342 Industrial Power Engineering and Applications HandbookLight &P0 topsIChassis inisolatedposition-Padlocking of switch...-Figure 13.12(a) Part view of a typical fully drawout-typemotor control centre (Courtesy: ECS)construction eliminates human error and reduces rackingtime. The trolley can now be replaced swiftly with theleast downtime.Figure 13.12(b)Door interlock defeat facilityPadlocking and defeat interlocking facilities13.4 Design parameters and serviceconditions for a switchgearassembly13.4.1 Design parametersA switchgear or a controlgear assembly will be designedto fulfil the following design parameters:Rating1 Rated voltage2 Rated frequency3 Rated insulation level4 Rated continuous current rating and permissibletemperature rise5 Rated short-time current rating or fault level of a system(breaking current for an interrupting device)6 Duration of fault7 Rated momentary peak value of the fault current(making current for an interrupting device)A switchgear assembly may be assigned the followingratings:1 Rated voltage This should be chosen from Series Ior Series I1 as shown in Table 13.1. Nominal systemvoltage is the normal voltage at which an equipmentmay usually have to perform. The maximum systemvoltage is the highest voltage level which the supplysystem may reach temporarily during operation andfor which the equipment is designed.2 Rated frequency 50 or 60 Hz (refer to Table 13.1).3 Rated insulation level This will consist ofPower frequency voltage withstand level, accordingto Section 14.3.3 andImpulse voltage withstand level for assemblieshaving a system voltage of 2.4 kV and above,according to Section 14.3.4.Assigning the impulse level to an HT switchgearassemblyAs discussed in Sections 17.5 and 23.5.1, an electricalnetwork may be exposed to different voltages surgeswhich may be internal or external. The extent of exposureof the connected equipment would determine its level ofinsulation. IEC 6007 1-1 has recommended the desired

Switchgear and controlgear assemblies 13/343Female portionMale portion Terminals disengaged Terminals engagedFigure 13.13(a) Typical plug-in-type terminals1IAuxiliary contactsr3 41. Trolley2. Auxiliary contacts (sliding type)3. Outgoing power contacts (female)4. Outgoing power contacts (male)5. Insulator6. Incoming power contacts (female)Figure 13.13(b) Rear view of a withdrawable chassis illustratingpower and auxiliary contact detailsPower contactsFigure 13.1 3(c) Drawout power and auxiliary (control) contactslevel of insulation for different system voltages and theextent of their exposure, as shown in Tables 13.2 and13.3, for the HT switchgears installed in a distributionnetwork. The recommended insulation levels will takeaccount of the following aspects.Indoor installationsFor installations that are electrically non-exposed, internalsurges and overvoltages may be caused by one or moreof the following:(a) At surge frequencySwitching surges, as may be caused during theswitching operation of an inductive circuit (Section17.7.2) or a capacitive circuit (Section 23.5.1)Grounding overvoltages, as may be caused by asudden ground fault on a phase in an isolated neutralsystem (Section 20.2.1)(b) At power frequencyMomentary overvoltages due to a sudden loadrejection, which may overspeed the generator anddevelop higher voltages orSustained overvoltagesAll such installations are assigned the impulse voltagesas in list I of Tables 13.2 and 13.3.Outdoor installationsInstallations that are electrically exposed to lightningare assigned the impulse voltages as in lists I1 or 111 ofTable 13.2 or list I1 of Table 13.3, depending upon theextent of their exposure to lightning and the method oftheir neutral grounding.

~~ ~~ 52~ 145~13/344 Industrial Power Engineering and Applications HandbookTable 13.1 Rated voltages and frequencies for metal enclosedbus systems also applicable for switchgear assembliesSeries INominal Rated maximum Nominal Rated muximumsystem voltage system voltage system voltage system voltagekVkV0.415* 0.440.6* 0.663.3 3.66.6 7.211 1215 17.522 2433 3644 5266 72.5Series II(ANSI C37-20C)kVkV0.48 0.5080.60 n.m4.16 4.767.2 8.2513.8 15.014.4 15.523.0 25.834.5 38.069.0 72.5- -For higher voltages -refer to Tables 13.2 and 13.3 for series 1 andIEC 60694 for series 11.Frequency: 50 or 60 HzI Frequency: 60 HzBased on IEC 60694, BS 159 and ANSI C37-2OC.NotesSeries I: Based on the current practices adopted by India, Europeand several other countries. *These voltages have nowbeen revised from 0.415 kV to 0.4 kV and from 0.6 kVto 0.69 kV as in IEC 60038.Series 11: Based on the current practices adopted in the USA andCanada.List I1 When the system is not directly connected to anoverhead line, such as when connected through atransformer.2 When the neutral of the system is grounded solidly orthrough a small impedance (Section 20.4).3 When the system is provided with a surge protection.List I1 or List 1111 When the system is connected to an overhead line.2 When the neutral of the system is not solidly grounded.3 When no surge protection is provided.4 When the installation demands a high degree ofsecurity, such as in a generating station.NoteIS 8084, however, has opted for list I1 for the busbar systems for allapplications.Although the rated values are based on experience and fielddata collected, sometimes these values may be exceeded in actualoperation, due to changed service or weather conditions. It is,therefore, recommended to take cognisance of this fact and providea surge protection to contain the surge voltages within the ratedvalues of Tables 13.2 and 13.3. It is recommended that actual sitetests be conducted on similar installations to ascertain the likelylevel of voltage surges that may occur during operation. To be onthe safe side, it is recommended for outdoor installations, particularly,to contain the severity of the surge voltages, with the help ofsurge arresters, within the maximum assigned impulse voltages(BIL) of the equipment, considered in Tables 13.2 and 13.3, lessthe protective margins, for range I and range I1 voltage systemsrespectively.Table 13.2 Standard insulation (impulse) Levels for Series 1 range I voltage systems (1 kV < V,,, 2 245 kV)voltage' v,,,kV(c m.s.)system voltage V,,,kV(ctn.s.)Standard one-minute power frequencywithstand voltageKV*(r.m.s.)Standard lightning impulse (1.2/50~)withstand voltage phase to ground**kV* (peak)List I List I1 List IIlList I List II List IIl3.36.611153.67.21217.520 40 -40 60 -60 75 -75 95 -22334466-110132154220Notes2436, 72.5100123I701 245As in IEC 60071-1 and IEC 60694150 185 -185 230 -230 275 -275 325 -360 39s 4609s 125 -145 I70 -250 - -325 - -380 450 -450 550 -550 650 -650 750 -850 950 1050* More than one lightning impulse insulation level is indicative of the extent of exposure of equipment to lightning surges. Correspondinglythe power frequency withstand voltage can also he more than one.** Lightning stress between the phases is not more than the lightning stress between a phase and the ground. For more details refer to IEC6007 1-1.'IEC 60071-1 has specified only V,,, values. V,,,,, is indicated based on the practices adopted by various countries.

Table 13.3 Standard insulation (impulse) levels for range II voltage systems (V, > 245 kV)Switchgear and controlgear assemblies 13/345NominalsystemvoltageHighest voltagefor equipmentvmStandard one-minutepower frequencywithstand voltageStandard switching impulse (250/2500 ps)withstand voltageStandard lightningimpulse (1.2bOp)withstand voltagephase to ground’kV(r.m.s.)‘ kV(r.m.s.) kV(r.m.s.)’Phase to ground Ratio to the phase Phase-tokV(peak)4 to ground peak phasevalue as in kV (peak)’IEC 60071-1kV (peak)4275300380750 1.50850 1.50 1275________~330362450850 1.50 1275 I 1050950 1.50 14251175400420520950 1.50 14251050 1.50 1575130014255506208008301425 1.70 2420 I 2100Notes As in IEC 60071-1 and IEC 60694‘IEC 60071-1 has specified only V,,, values. Vr,,,s, is indicated based on the practices adopted by various countries.’Lightning stress between the phases is not more than the lightning stress between a phase and the ground. For more details refer to IEC:60071- 1.’These values are applicable for- Type tests - phase to ground- Routine tests - phase to ground and phase to phase.4More than one lightning impulse insulation level is indicative of the extent of exposure of an equipment to lightning surges.’These values aremeancfor type tests only.4 Rated continuous current ratings andpermissible temperature riseThe current ratings should generally be selected fromseries R-10 of IEC 60059 which comprises the followingnumbers;1, 1.25, 1.6, 2, 2.5, 3.15, 4, 5, 6.3 and 8 etc. and theirmultiples in 10e.g. for the number 1.25 the standard ratings may be;1.25,12.5, 125, 1250 or 12500A etc. IS 8084 has suggestedthe following ratings in amperes based on the aboveseries:100,250,400,630, 800, 1250, 1600, 2000, 3150,4000, 5000,6300, 8000, 10000, 12500 and 15000 etc.The temperature rise limits of the various parts of aswitchgear assembly during continuous operation at therated current must conform to the values in Tables 14.5and 14.6.Diversity factorThe connected load of an electrical installation is generallymore than its actual maximum demand at any time. Thereason is the liberal provisions made for loads that are tobe used occasionally - spare feeders, maintenance powersockets and some reserve capacity available with mostof the outgoing feeders. Moreover, some of the feedersmay be operating underloaded. To design a switchgearassembly, particularly for its incoming feeder and thebusbar ratings, based on the connected load would beneither economical nor prudent, and the protective schemetoo would under-protect such a system.Based on experience, IEC 60439-1 has suggested amultiplying factor, known as the ‘diversity factor’ asnoted in Table 13.4. This factor depends upon the numberof outgoing circuits and is defined by the ratio of themaximum loading on a particular bus section, at anytime, to the arithmetic sum of the rated currents of all theoutgoing feeders on that section (Figure 13.14). Thisfactor also helps to determine the most appropriate andeconomical ratings of the main equipment, such as thetransformer, cables or the bus system, associatedswitchgear and the protective devices (Example 13.1).The values of Table 13.4 are based on a generalassumption, and may vary from one installation to anotherdepending upon the system design, the derating factorsconsidered and any other provisions made while choosing

13/346 Industrial Power Engineering and Applications HandbookTable 13.4Conventional values of diversity factorSection -1 2 3 4Number of main outlets Diversis factor 630A 630A 800A2 to 3 0.84 to 5 0.76 to 9 0.6IO and above 0.5As in IEC 60439-1l/C 4000AACB630A630A630A400A630A400Athe rating of a feeder (outlet), as noted above. It is thereforerecommended that this factor be specified by the user,depending upon likely capacity utilization, to help theswitchgear assembly manufacturer to design a moreeconomical busbar system. In the absence of this, thefactors as indicated in Table 13.4 may be applied.Example 13.1Consider the power distribution system of Figure 13.14, havingthe following feeder details:I/C feeder, 4000A11 No. 800 A = 800 AO/G feeders 7 Nos. 630 A = 4410 A and,4 NOS. 400 A = 1600 A:. Total connected feeder load = 6810 Aand diversity factor for 12 Nos.feeders as in Table 13.4= 0.5:. Maximum loading on theincoming feeder or the mainbusbars at any time = 6810 x 0.5= 3405 AAccordingly we have selected the rating of the incomingfeeder as 4000 A. 4000 A being the next standard rating after3150A. The maximum loading on each vertical section isworked out in Figure 13.14. These ratings of vertical busbarsare when the arrangement of busbars is to individually feedeach vertical row. If one common set of busbars is feedingmore than one vertical section, the rating of busbars can befurther economized. But one must take cognisance that toomany tapings from one section of the bus may weaken thebus system.If two sections are joined together to have a common verticalbus system, say, Sections 3 and 4, then the rating of thecommon bus will be:Total connected load = 2060 + 2230= 4290 ADiversity factor for 8 numbers of feeders = 0.6:. Maximum ratingor say= 4290 x 0.6= 2574 A= 2500 Aas against 1600 A + 1800, i.e. 3400 A, worked out in Figure13.14, when both the sections were fed from individual busbars.5 Rated short-time current rating orfault level of a systemTo establish the fault level of a systemThe fault level of an electrical network is the capacity ofTotal connected feederloadTotal connected loadDiversity factor as perTable 13 4Maximum loading atany timeRecommended rating ofbusbar630A2520A08201 6A2000A400A2060A681 OA0.81648A1600AFigure 13.14 Illustration of diversity factor400A2230A-0.81784A1800Athe source of supply to feed the faulty circuit, and isrepresented in kVA or MVA. Consider the simpletransmission and distribution network of Figure 23.1,which is redrawn in Figure 13.15 for more clarity. Thisillustrates the impedances of the network at various points,and their role in the event of a fault. The impedance ofa circuit is built through the self-impedances of thewindings of the various machines in the circuit, generators,transformers and motors etc., and the impedances of theconnecting transmission and distribution lines and theassociated cables.To increase the impedance of the network, a seriesresistor or reactor is sometimes used to contain the faultlevel of a system within a desirable limit. This may berequired to make the selection of the interrupting deviceeasy, and from the available range, without an extra costfor a new design as well as an economical selection ofthe interconnecting conductors and cables. Such a situationmay arise on HV >66 kV or EHV >132. kV transmissionnetworks, when they are being fed by two or more powersources, which may raise the fault level of the system toan unacceptable level. The cost of the interrupting devicefor such a fault level may become disproportionatelyhigh, and sometimes even pose a problem in availability.Ground fault current is controlled by a method similarto that discussed in Section 20.4.2. The electricityauthorities of a country generally provide the preferredfault levels, depending upon the availability of theinterrupting devices. They also suggest the likely generationof overvoltages in a faulty circuit and the healthyphases on a ground fault as a result of groundingconditions, as guidelines to the system designers to designa transmission or a distribution network for various voltagesystems. The guidelines may also recommend the

ImpedanceSection- z,Switchgear and controlgear assemblies 13/347Zl -31 z2 -32 z3 zt3 4 zt4 z5- - -F! %, ,LT loads-----&$ I I I11 kV to24 kVyard132, 220 or400 kV 33 kV@11, 6.6 400/440 Vor 3.3 kV0-v-Generating station Primary Receiving Secondary Main receiving L.T. distributiontransmission station transmission and H.T. distributionstation@Generator.@Isolated phasebus system@ EHV generatortransformer@ EHV transmission line@ H.T. secondarytransmission transformer@ Secondary transmission line@ H.T. distribution transformer@ H.T. distribution network@ L.T. distribution transformer@) L.T. distribution network(i) Fault at ‘A’ is calculated by impedance Ztl(ii) Fault at ‘B is calculated by impedance 213(iii) Fault at ‘C’ is calculated by impedance Zt4Note The actual fault at any point will be much lower than calculated with the above impedances Ztl, &,Zt3 and 214 because other impedancesfrom the source of supply (Transformers in the above case) up to the point of fault, are not considered while designing a system.Figure 13.15 Typical layout of a typical transmission and distribution network and significance of circuit impedances at various pointsmaximum loading of a line and the maximum number offeeding lines that may be connected to a common grid,to limit the fault level of the system within the desirablelimit. Some typical values are noted below:Nominal system Limiting fault No. of additionalvoltage level feeding lines765 kV 2500 MVA Nil400 kV 1000 MVA 5220 kV 320 MVA 3132 kV 150 MVA 2For more details refer to Section 24.8. It is possible thatin the course of time more generating stations may beinstalled to meet the rising demand for power. Theirfeeding lines too will be added to the existing grid toaugment its capacity. This would also enhance the faultlevel of the existing system. To ensure that the prescribedfault level is not exceeded, a detailed network analysismay be carried out to determine the minimum possibleimpedance of the grid, at various vulnerable locations,to establish the likely revised fault level. If it is felt thatit may exceed the prescribed limit, current-limiting seriesreactors may be provided at suitable locations to yetcontain the fault level within the prescribed limits. Forcurrent limiting reactors refer to Chapter 27.With the availability of more advanced interrupters infuture, it will be possible to upgrade the present guidelinesand permit connections of more feeding lines on anexisting grid without having to resort to a series reactor.Below, we analyse the likely fault levels of a systemunder different circuits and fault conditions for an easyunderstanding of the subject. It is a prerequisite to decidethe level of fault, to select and design the right type ofequipment, devices and components and the protectivescheme for a particular network.A power circuit is basically an R-L circuit. In the eventof a fault, the system voltage (V, sin 4) may occursomewhere between V = 0 and V = V, on its voltagewave. This will cause a shift in the zero axis of the faultcurrent, Zsc, and give rise to a d.c. component. The faultcurrent will generally assume an asymmetrical waveformas illustrated in Figure 13.27.The magnitudes of symmetrical and non-symmetricalfault currents, under different conditions of fault andconfigurations of faulty circuits, can be determined fromTable 13.5, where Z1 = Positive phase sequenceimpedance, measured under symmetrical load conditions.The following values may be considered:

~ ~~~13/348 Industrial Power Engineering and Applications HandbookTable 13.5 RMS values of fault currents under different conditions of fault in a power system-Sr. no. rype of fault Configuration of faulty circuitFault currentsSymmetricalI1 Unsymmetrical1 3-Phase(i) Star connected with isolated neutralI;rf- V I kc2Phase to phasekc(ii) Delta connectedkc.(i) Star connected with isolated neutralt- "--l(ii) Delta connectedkc3Phase or phases tcgroundi(i) Star connected, solidly groundedI- "I(i*VIz, + Z? + Z"I(ii) Star connected, impedance groundedI- V I[3*xJ-x 1J5 z, + z2 + Z" + 3x, **1Notes* Refer to Section 20.6.2** When the neutral is impedance grounded, three times its impedance must be added to Z,, in view that this impedance would fall in serieswith each phase.1 It is equal to the phase impedance of the overheadlines or cables. For low current systems, LT or HT,this impedance is nearly equal to the resistance of thecircuit, as R>>XL. Due very low XL, the impedanceremains nearly the same even on a fault, as a result ofvery little change of flux, except the skin effect, whichis moderate for moderate currents. Refer to the datasheets for sizes of cables and conductors provided inChapter 16, Appendix 1, which shows that R >>XLfor smaller ratings. For overhcad lines, refer to Tables24.l(a) and (b).However, as the current rises, the situation changes

Switchgear and controlgear assemblies 13/349gradually as a result of the proximity effect, whichadds to the leakage flux of the circuit and diminishesits reactance and the impedance. The decrease inimpedance contributes to limiting the fault level ofthe system. Refer to Table 30.7, which shows thegradual decrease in X, with the current.2 It is equal to the short-circuit impedance of transformersand motors. Now the machine undergoes a quickchange of flux, due to a change in the applied voltage(it changes two peaks in one half of a cycle, Section1.2.1). The impedance during a fault is thereforedifferent from that during normal running. As standardpractice, the fault impedance is provided in p.u. bythe machine manufacturers. The content of R is nowfar too low, compared to X, (Rc

13/350 Industrial Power Engineering and Applications Handbookimpedance of the source of supply alone and selectingthe equipment and devices for the worst-case operatingconditions. The highest impedance is considered whensetting the protective relays to make them actuate evenon the smallest fault (other than of a transitory nature).Inferences from Table 13.6Rotating machinesThese have Z, < Z1 andZo

Switchgear and controlgear assemblies 13/351This is a simple calculation to determine the maximumsymmetrical fault level of a system, to select the type ofequipment, devices and bus system etc But to decide on arealistic protective scheme, the asymmetrical value of thefault current must be estimated by including all the likelyimDedances of the circuitTransformer-ITransformer-2Fault levels of an LT systemTable 13.7 suggests some standard ratings of thedistribution transformers, their full load current on theLV side and the corresponding fault level, consideringan average impedance according to IEC 60076- 1.Referring to transformers of 2.500 kVA and above, thefault current, assuming an impedance of 6.2.5%, will workout to more than 50 kA. The LT protective systems, asnormally designed and manufactured, are suitable for afault level up to SO kA or so (also see Table 13.7 and thenote). It is, therefore, recommended that the size of thetransformers for the LT distribution may be chosen, asfar as possible, up to 2000 kVA. For large industrial andother LT loads, the practice is to choose more than onetransformer of smaller ratings, to cater for the requiredload. rather than one transformer of larger capacity. Inthe same context, when more than one transformer areusccl to feed a large industrial load, distribution of theload must also be done in as many sections as the numberof transformers. The parallel running of any two of thetransformers must be avoided, even when they are suitablefor ;i parallel operation and the source of supply is alsothe same. Otherwise the impedances of all the transformersthat may be running in parallel would also fall in parallel.It will reduce the effective impedance of the faulty circuit,and multiply the system fault level. Use of current limitingreactors on LT systems is not a recommended practice,for reasons of relatively much higher losses of the reactor,compared to the power handled by it. A few typical. hutgenerally adopted, distribution networks are illustratedin Figures 13.16 and 13.17.Another factor that may discourage the use of higherratings of transformers is the need to run a number ofcables in parallel, to cater for such high currents. It mayprove to be cumbersome, besides requiring a very highderating of the cables. Even the design of such largeratings of bus systems and their high eddy and magnetizingspace currents would call for highly skilled engineering.A single large transformer will also provide less flexibilityto thc system. A fault in this transformer or its shutdownfor a normal maintenance will result in the shutdown ofthe whole installation. However, a single large transformerhas its own merits also. The user may evaluate whichsystcm would suit him the best and decide the ratingaccordingly.Fault level of a system under cumulative infiuenceof two power sourcesShort-time luting of' the tap-offiIt is possible that in certain installations as a result ofsystem needs it may not be possible to limit the faultlevel of the system within a desirable limit. A powergcneratingstation. connected to an external grid, toP _y__OiG feedersOiG feedersCautionThe configuration represents scheme of twoincomings and one bus coupler. To avoidparallel operation. only two of these mustbe 'ON' at a timeFigure 13.16 Interlocking requirementsTransformer-1 Transformer-2 Transformer-3P:;irri:P P POiG feeders O/G feeders OiG feedersCautionThe configuration represents scheme of three incomings and two buscouplers. To avoid parallel operation, only three of these five must be'ON' at a time.Figure 13.17 Interlocking requirements

~ transformer13/352 Industrial Power Engineering and Applications Handbookaugment the capacity of the power network is one example.Figure 13.21 (described later) illustrates a typical electricallayout of a power-generating station where the generatorG is connected to an external grid. The generator voltageis normally much lower than the voltage of the externalgrid. It is therefore connected to the grid through a stepupgenerator transformer GT. The fault level of the gridon the generator side is thus governed by the impedanceof the GT. In the event of a fault on this section, say atthe bus section connecting G with GT, one side of it willbe fed by the G and the other by the GT. Thus, no part ofthis section would carry a fault current of more than onesource at a time. The main interconnecting bus systembetween G and GT may therefore be designed for a faultlevel, whichever is the higher of the two sources.Through the tap-offs of the bus, the unit auxiliarytransformers (UATs) are connected to feed the stationauxiliary services. For more clarity we have taken outthe portion of the tap-offs from Figure 13.2 1 and redrawnit in Figure 13.18 to illustrate the above system and itsinterconnections. The tap-offs are now subject to thecumulative influence of the two supply sources. In theevent of a fault on this section, both the sources wouldfeed the same and the fault current through the tap-offswould add up. The tap-offs should thus be designed forthe cumulative effect of both fault levels. For the sake ofan easy reference, Table 13.8 suggests a few typical valuesof fault currents, worked out on the basis of data consideredfor the G and GT. One such example is also worked outin Example 13.3.Impulse level of the tup-offsThese sections should also possess a higher level ofinsulation, as prescribed in Table 32.1 A.Example 13.3Consider a generator G of 250 MVA, 16.5 kV having a faultlevel of 61 kA and the generator transformer GT of 31 5 MVA,16.5 kV1400 kV having an impedance of 14.5 i 10% (Table13.8, Figure 13.18):. full load current of GT,I, = 315 kA& x 16.5and fault current from equation (13.5)315 1Is, =& x 16.5 0.145 X 0.9where Z, is considered with the lower tolerance, i.e. 14.5%-lo%, to obtain the maximum level of fault current to designthe system.:. I,, = 84 kABut the GT is connected to a power grid and the grid to atransmission system. If we consider the fault level of thetransmission system as 40 kA, as in Table 13.10, then themaximum fault that can occur on the LV side of the GT will begoverned by the fault level of the transmission system(40 kA) and not the GT (84 kA).Accordingly, the fault level for which the tap-offs should bedesigned,= 61 + 40= 101 kAand momentary current= 101 x 2.5 (Table 13.11)= 252.5 kAOne-minute power frequency withstand voltage from Table32.1A.(i) For main interconnecting bus between G and GT - 50 kVGenerator(ii) For the tap-offs - 55 kV16.5 kV G250 MVA (&NoteFor guidance on short-circuit calculations refer to IEC 60909Generatortransformer16.5 kV1400 kV315 MVATo feed the external grid(the grid is connected totransmission system, havinga fault level of 40 kA)flsc = 40 kA lsc = 61 kA1 (UAT)Figure 13.18 Fault level of tap-offs of a bus system undercumulative influence of two power sourcesFault level of a generator circuitThe condition of a fault on a generator is similar to afault in an R-L series circuit, as noted in Table 13.5,except for the effect of arr,iature reaction and variationin the field current. The cunent waveform of a generatorunder a fault condition, therefore, becomes modified asa result of armature reaction. Since the flux crossing theair gap during a short-circuit in the first few cycles islarge, the reactance is the least and short-circuit currentthe highest. This current is termed the sub-transientsymmetrical fault current (Issc). The impedance at theinstant of the short-circuit that determines this current istermed the sub-transient reactance XT . The Y being toolittle, it is normally ignored. Issc exists in the system forjust three or four cycles, depending upon its rate of decay,which is a function of Issc/Ir.The design parametersmust ensure that its peak value including the d.c.component (a x IsSc ) does not exceed 15 times thepeak rated current (15 x fi. I, or 21 . I, as in IEC60034-1). It corresponds to the current of the first peak.

Table 13.8 Typical parameters of a power generating station and its likely fault levelsSwitchgear and controlgear assemblies 13/3531 GeneratorGenerator size (rated MVAx p.f.)Approximate sub-transientunit reactance,RatingGenerator fault level,Isc = (*)[a x vr x Is, 1Rated p.f.Stator nominal voltage, V,Stator current (CMR)a, I,Frequency, fMWX,"(%)MVAkAMVAkVAHZ6015.47525.564870.81139365067.517.784.426.214770.810.516395011015.9137.545.398650.81172175012013.2141.258.9410720.8510.5778050200/2 1016.824753.8714700.8515.7530505025016.929460.8917400.8516.50 2905050017.258894.1734250.85216 200502 Generator transformer (GT)RatingMVAImpedanceZp%Generator side fault level,considering the lowertolerance of impedance, kAbVector groupSPe7512.5f 1098512.5 f 10435 42Staddelta t---14012.5 f 10%65(Ynd 11)-14012.5f 10468Three-pha25014+ 10%73two windin31514.5f 10%3450015* 10%I223ne-phase.wo-winding,3 Nos 200MVA each3 Generator isolated phase bus ductRated current (CMR)a A400050008000800010000I2 500I8 000One second short-timerating (as for GT)~kA3542656873$4I224 Tap-offs of the generator bus ductApproximate one-second kAshort-time rating (rounded ofOb(it is the summation ofthe fault levels of the generatorand the generator bus duct: refer toFigure 13.18 and a typical calculationin Example 13.3)61 68 110 127 127 145 1165 Unit auxiliary transformer (UAT)No. of transformers111 or 2I or 2lRatingImpedanceMVAZp%12.57.5167.525 for 1and12.5 for 210 for 25MVA 7.5 foi12.5 MVA51.5 for 1md16 for 210 for 31.5VIVA 7.5 foiI6 MVAl of 25)r1. of 31.5LO for 25HVA 10 for11.5 MVAVector group typec Dynl or Ddo >< 3-phase, 2 winding >aCMR - Continuous maximum ratingbThe GT HV side is connected to a power grid and the grid receives power from many sources and can have a fault level higher than thatof G or the GT. On a fault anywhere in the main length or the tap-offs of the IPB between G and the GT (Figures 13.21 and 13.18respectively) both generator and the grid will feed the fault (grid through the GT). The fault level of the IPB straight length may thereforebe considered as the fault level of the grid limited upto the fault level of the GT (impedance of GT limiting the fault level). Accordinglywe have considered the fault level of the straight length of the IPB as the fault level of the GT. It is, however, only for illustration, exactfault level shall depend upon the system parameters, also refer to Example 13.3 clarifying this.Source NTPC and BHEL.

13/354 Industrial Power Engineering and Applications HandbookThese reactances are measured by creating a fault,similar to the method discussed in Section 14.3.6. Theonly difference now is that the fault is created in any ofthe phases at an instant, when the applied voltage in thatphase is at its peak, i.e. at V,, so that the d.c. componentof the short-circuit current is zero and the waveform issymmetrical about its axis, as shown in Figure 13.19,where,(i)(ii)CL - defines the sub-transient state andand the magnitude of the first peak can be definedbyI, = I /Z.~ssC(rm*)BL - defines the transient state andThe short-time current, Is, (1 or 3 seconds), of thesystem to which this machine would be feeding isdefined by It,,,,... I,, = Jz . I,(iii) AA' - defines the steady-state condition andandX: = sub-transient state symmetrical reactance -WphaseThe maximum value of this reactance willdepend upon the voltage regulation grade and isSub-transientstate (16, = 0)-I- Duration of faultTransient stateSymmetrical fault current at V,,,, when /dc = 0Steadystate(bc = 0)Figure 13.19 Short-circuit oscillogram of a generatorcspecified in BS 4999-140, for different grades.X; = transient state symmetrical reactance - Q/phase.Since the interrupter will take at least three orfour cycles to operate from the instant of faultinitiation, it is this transient reactance that ismore relevant for the purpose of short-circuitcalculations.X, = Steady-state symmetrical (synchronous)reactance - Wphase.r = series resistance of the generator, being toosmall, 0.5-10% of the sub-transient reactance,i.e. r

Table 13.9 Fault currents (rms.) in a generator circuit under different fault conditionsSwitchgear and controlgear assemblies 13/355Sr. Type of fault Configuration of Fault currentsno. faulty circuit Sub-transient Transient Steady stater(T1K.L Jr xa6 143 x4Vl 1fi Xd(ii) Delta connected2 Phase to phase (i) Star connected(ii) Delta connectedJ 5 . x . 1 Js.K.1Jr x; +x2 fi xd+x23 Phase or phases (i) Solidly groundedto groundt--v--l(ii) Impedance groundedNotes1 The above are r.m.s. values. For peak values multiply them by &.2 The first peak with full asymmetry, i.e. the making current IMr will be represented by 2 (P x I,,,) (where P = factor of asymmetry asin Table 13.11).3 x2 and xo are negative and zero phase sequence reactances of the generator respectively, in n/phase.4 Since the ground fault currents in generators can be higher (Section 20.10.1) than the sub-transient state current, special care need betaken while grounding a generator to limit the ground fault current. Section 20.10.1 covers this aspect also.5 The relays and the breaker will operate only during the transient state, hence the significance of transient state values to set the currentand the time of the protective and isolating devices.6 In certain cases, where a long delay may be necessary for the protective scheme to operate, it may be desirable to use the maximumsteady-state short-circuit current 42 . I,, for a more appropriate setting, rather than the maximum transient current 42 . I,, as by then thefault current will also fall to a near steady-state value, Is,~r,,,,s,).x; = 0.04 Wphase = 1391 Ax; =0.02 Wphasex, = 0.015 Wphasex, = 0.007 WphaseI, =1000 x 1000& x 415and base reactanceXb =41 5& x 1391= 0.17 Wphase

13/356 Industrial Power Engineering and Applications Handbooki7 IM (momentary peakmaking current)-state Transient state state(ld, = max.)I(bc = 0)- Duration of fault -:. the corresponding unit reactances for the above per phasereactances will beX, = 1 .O or 100%X;, = 0.235 or 23.5%Xg = 0.117 or 11.7%X, = 0.088 or 8.8%X, = 0.04 or 4.0%{ 'X = 2b land the symmetrical fault currents will be(i)Sub-transient current415Isscrms =43 x 0.02= 11.98 kAand the magnitude of the first peak, including the d.c.component;I, = & x 11.98= 16.94 kA(ii) Transient current415It,,, =fi x 0.04= 5.99 kAand short-time current, I,, = & x 5.99(iii) The factor of asymmetry= 8.47 kA- Magnitude of the first peak current (lM)Maximum transient current 2&(or Is,-)---16.948.47= 2, which is the same as given in Table 13.11-(iv) Steady-state short-circuit current415Istr,, =fi x 0.17= 1.41 kAwhich is almost equal to the full load current (/J. Thesmall difference is due to d.c. component.Selection of basesA power system is connected to a number of power supplymachines that determine the fault level of that system(e.g. generators and transformers). The impedances ofall such equipment and the impedances of the interconnectingcables and overhead lines etc. are theparameters that limit the fault level of the system. Forease of calculation, when determining the fault level ofsuch a system it is essential to consider any one majorcomponent as the base and convert the relevant parametersof the other equipment to that base, for a quickercalculation, to establish the required fault level. Belowwe provide a few common formulae for the calculationof faults on a p.u. basis. For more details refer to atextbook in the references.Actual quantityp.u. = (13.1)Chosen basep.u. = per unit quantity of V, I or Z1 (fault impedance)(e.g. for a base impedance of 10 R, 5 R will be 5/10, Le.0.5 p.u. or 50%)3'3. v. IkVA = ~ (Vis in volts)1000v 1and Z, = ?. -t'3 I- v' - kVA x 1000If Z,, is the p.u. impedance thenActual impedance (Z,)z, = Chosen base impedance (Z,)wherez-Base V2- Base kVA x 1000Z, .(base kVA) x 1000:. z, = etc.Base V2If Zpl = p.u. impedance at the original base, andZp2 = p.u. impedance at the new base,thenV,'kVA,(1 3.2)(13.3)(I 3.4)

(13.6)~~~Switchgear and controlgear assemblies 13/357where suffixes 1 and 2 refer to the original and the newbase values respectively.Fault current I,, =I,].e.. rated currentshort-circuit unit impedance(13.5)and fault MVA =baseMVA etc.~15/24kV*116.63.30.415-127.23.60.44LPFault levels of HT systemsWe illustrate a typical powerhouse generation andtransmission system layout in Figure 13.21, and reproducein Table 13.10 the typical fault levels of differenttransmission and distribution networks in practice fordifferent voltage systems.We also provide a brief reference to a protective scheme,usually adopted in a large power-generating station as inSection 16.8.2.6 Duration of faultWe have mentioned two systems, 1 second and 3 second.A choice of any of them would depend upon the locationand the application of the equipment and criticality ofthe installation. Generally speaking, it is only the onesecondsystem that is in practice. The three-second systemmay sometimes be used for low fault level networks,where a. I,,, would fall within the capability of theavailable interrupting devices and at reasonable cost.7 Rated momentary peak value of the fault currentA fault current on a power system is normally asymmetricalas discussed next, and is composed of asymmetrical a.c. component ISC(r,m,s,) and an asymmetricalsub-transient d.c. component Id, (Figure 14.5). The forcesarising out of Zsc are referred to as electromagnetic andTable 13.10 Typical fault levels for an integrated transmissionand distribution networkNominalsystemvoltagekV(r.m.s.)76540022013233Highestsystem voltagekV (r.m.s.)800420245Symmetricalinterruptingcurrentrating (r.m.s.)kA Is,40403 1.51402513 125-*-40404043/50Minimummomentarycurrent peak fordynamic ratingkA(PWIO010079110062.5177.562.5-*-IO010010090.01105*Since this represents the generator voltage, therefore the faultlevel will be governed by the generator and the generator transformeras indicated in Table 13.8.those by asymmetry, i.e. Id, as dynamic. The cumulativeeffect of this is electrodynamic and is quite significant.It requires adequate care while designing a current-canyingor supporting system. Switching devices and other currentcarryingcomponents and their mounting structures, suchas the busbar systems in a PCC, MCC or a bus duct etc.must withstand such stresses during a short-circuit. It istherefore of vital importance to take account of thisasymmetry and to determine this to form an importantdesign parameter for switching devices and all currentcarryingsystems.The peak value of a fault current will depend upon thecontent of the d.c. component. The d.c. component willdepend upon the p.f. of the faulty circuit and the instantat which the short-circuit commences on the current wave.(Refer to Figure 13.27, illustrating the variation inasymmetry with the p.f. of the faulty circuit. For ease ofapplication, it is represented as a certain multiple of ther.m.s. value of the symmetrical fault current Isc.)The content of asymmetry may decay quickly andmay exist in the system for just three or four cyclesfrom the commencement of the fault, depending uponthe time constant, z, of the circuit. The time constant, 7,is the measurement of the rate of decay of the d.c.component, and is the ratio of the system reactance, L,to the system resistance R, i.e. L/R. A large L/R willindicate a high time constant and a slow rate of decayand vice-versa, as illustrated in Figure 13.22. Theasymmetry is therefore measured by the peak of thefirst major loop of the fault current (which may occur inany of the phases, as it has occurred in phase Y in theoscillogram shown in Figure 13.23). The subsequentloops will be smaller and less severe and thus thesignificance of the first loop. This is referred to as themomentary peak value of the short-circuit current forthe most severe fault conditions, such as at extremelylow p.f.s (RlX, being very low) when the recovery voltagemay be the maximum and the fault current the highest.These values are given in Table 13.11 in the form oflikely multiplying factors for different symmetrical r.m.s.values of fault currents, Zsc, according to IEC 60439-1for LT and IEC 60694 for HT systems and are drawn inFigure 13.22. These values are almost the same asprovided by ANSI-C-37120C as well as in Table 28.1.The exact values of this factor may be estimated bycreating a short-circuit condition and obtaining anoscillogram. For details refer to Figure 13.23.NotesThe rated momentary peak value of the fault current, lM, willrelate to the dynamic rating of an equipment. It is also knownas the making current of a switching device and defines itscapability to make on fault.The peak value of the asymmetry is considered to determinethe electrodynamic stresses to design the mechanical systemand the supporting structure for the current-canying components.A breaker will not trip instantly when a fault occurs, but onlyafter a few cycles, depending upon the actuating time of theprotective relays and the breaker’s own operating time. It willtherefore generally trip only during the transient state of thefault. The breaking capacity of an interrupting device, unlikeits making capacity, is therefore defined by the peak value ofthe transient state fault current, Le. by Is, (Table 13.9).Conventionally it is termed the r.m.s. value of the fault current.

13/358 Industrial Power Engineering and Applications Handbook400 kV transmission line132 kV transmission lineTo 400 kVtransmissionlineI-ICT-1- do-il’To 132 kVtransmissior7400 kVtransfer400 kVmain +bus-2400 kVmain +bus-1Match line 1From generatortransformerIFrom stationtransformer(Contd.)Figure 13.21 A typical powerhouse generation and transmission system, also illustrating power distribution to unit and stationauxiliary services

Switchgear and controlgear assemblies 13/359(Contd.) To switch yard From previous page To switch yardArMatch lineGeneratortransformer'I132kV/33kV or 6.6 kVStation transformertransformer-1transformer-2H.T. breaker3kV/1 1 kV/6.6kVI IH.T. breakerStationbus jducti-----------.iTie-connectionTie-connectionIIH.T. breaker33 kV/11 kVor 6.6 kVUnitH.T. board/Unit L.T.transformerUnit L.T. boardUnit L.T.transformeriiStation L.T.transformerC._ 0-tjloads,Unit sideStation sideLegend:- Three phase twowinding transformer- Three phase two windingtransformer withON-load tap changer& - Inter connecting tietransformer (ICT)(z - Circuit breaker- Isolator1 - Isolator($ - MotorloadFigure 13.21 (Contd.)

13/360 Industrial Power Engineering and Applications HandbookU0.2 0.4 0.6 0.8 1.0cos @ (tuxL) -c3.02.52.0 t6$1.5 $(II%1.0 8Figure 13.22 Approximate time of decay of the d.c. componentand the factor of asymmetry as a function of system P.F. duringa fault4 When designing a current-carrying system it is the r.m.s. valueof the fault current, Isc, that is relevant to determine the thermalstresses (= I&) during a fault, to choose the correct materialand size of the current-carrying components. (The duration ofasymmetry is too short to cause any significant heating of thecurrent-carrying components.)0.53Table 13.11 Multiplying factors to obtain the momentary peak(maximum r.m.s. or dynamic) values of the short-circuitcurrents including the sub-transient d.c. component atdifferent power factors (R/XL)Prospective short-circuit Factor of asymmetry to Cos Qor short-time current;Isc kA(xm.s.) (symmetricalbreaking current)obtain the peak shortcircuit or makingcurrent I, WXL)(a) For LT systemsup to 5 1.5Above 5 to 10 1.7Above 10 to 20 2.0Above 20 to 50 2.1More than 50 2.2(b) For HT systems Min. 2.50.70.50.30.250.2-Note For CTs this multiplying factor has been specified as 2.5 forall voltage systems, as in IEC 60044-1. See also Section 15.7 formetering and protection current transformers.Example 13.5For a 50 kA(r.m.s.) fault level on an LT system, the momentarypeak value of the fault current, C, will be = 50 x 2.1 = 105 kA.8 Causes of asymmetryAcurrent wave propagating symmetrically about its zeroaxis, i.e. when the envelopes of the peaks of the currentwave are symmetrical about its zero axis, is termedsymmetrical (Figure 13.24) and a wave unable to maintainthis symmetry is termed asymmetrical (Figure 13.25).The p.f. during a short-circuit as noted in Section 13.4.1is quite low, and is normally of the order of 0.1. Thecurrent will now lag the voltage by nearly 84" (Figure(58 completed cycles)1.16 Sec.Figure 13.23 Oscillograms of an actual short-circuit test carried out on a power distribution panel (Courtesy: ECS)

Switchgear and controlgear assemblies 13/36117.1 I). To analyse the shape of a current wave on ashort-circuit, consider the following conditions that mayoccur at the instant of the fault:When the short-circuit occurs at a current zero, Le.,when the applied voltage is almost at its peak, thevoltage and current waves will follow Figure 13.19,the current lagging the voltage by almost 84". Thecurrent will now be almost symmetrical.When the short-circuit occurs at a voltage zero thecurrent will also commence at zero. This is an unusualsituation when both the voltage and the current wavescommence at zero and yet cannot propagate in phasewith each other, in view of the current lagging thevoltage by almost 84". This situation is resolved by ashift in the zero axis of the current wave by almost84", as illustrated in Figure 13.26. Now it is able tofulfil its above condition again. The current will nowbe fully asymmetrical.Let us consider a more realistic situation, when theshort-circuit may occur somewhere between the abovetwo conditions.Supposing the current and the voltage waves both havesome value on their respective wave forms at the instantof short-circuit. The current will again tend to becomesomewhat asymmetrical but not fully. The content ofasymmetry will depend upon the instant at which the shortcircuitcondition occurs on the current wave and the p.f.of the faulty circuit (Figure 13.27). The higher the recoveryvoltage at the instant of fault, the lower will be theasymmetry (at V,, the d.c. component will be zero) andvice versa (at Vo, the d.c. component will be the maximum).It is observed that there may be asymmetry in thesystem as long as the short-circuit condition lasts, asillustrated in Figure 13.20, i.e. up to the opening of theinterrupting device. (For opening times of interrupters,refer to Table 19.1.) But the content of the asymmetrymay be quite feeble after three or four cycles. However,if the short-circuit condition still prevails. such as whenEnvelopes of peaks-. , ,EnveloDes ofaxis\Envelopes of peaks' Envelopes ofpeaksFigure 13.24 A symmetrical waveformFigure 13.25An asymmetrical waveformShift in thezero axisShort circuit/-current waveShift in thezero axisShort circuitcurrent wave073IHealthy--Shorl circuitcondition I conditionVoltage waveHealthy - ~ Short circuitcondition condition Voltage waveFigure 13.26 Illustrating a shift of nearly 90" (drawn at 90" forease of illustration) in the current wave at nearly 0.1 PF, fulfillingthe condition of current lagging the voltage by nearly 90" yetrising together at zero voltage (minimum applied voltage)Figure 13.27 Approximate illustration of a short-circuit conditionoccurring when both voltage and current waves are not at theirnatural zeros. Current shifting its zero axis from A, A2 to B, B, torise from zero again at the instant of short-circuit

Derutitig0.95131362 Industrial Power Engineering and Applications Handbookconducting a short-circuit test up to the desired durationof 1 or 3 seconds, the short-circuit current, althoughtheoretically asymmetrical until the test period, may beregarded as symmetrical (having reached its steady state)after three or four cycles. The content of asymmetry isignored after three or four cycles for all calculations andpractical purposes. In fact, a d.c. component less than50% that of the peak symmetrical component of the fault,current Is,, at any instant during the course of shortcircuitcondition may be ignored. In other words, therelevance of the asymmetry may be considered only upto the first peak, as the immediate subsequent peaks mayalso be less than 50% of the peak value of Is, (Idc < 0.5ZM) at that instant.The generation of an asymmetrical current on an a.c.system, leads to the inference that a short-circuit conditionwill give rise to a d.c. component due to a shift in itszero axis. During the sub-transient state the value of theasymmetrical current will be the phasor sum of thesymmetrical Is, and the asymmetrical Idc, currentcomponents. For details refer to Section 14.3.6.13.4.2 Service conditionsAmbient temperature, altitude and atmospheric conditionsat the place of installation of electrical equipment areconsidered to be the service conditions for the equipmentto operate and perform its duties. All electrical equipmentis designed for specific service conditions and variationsmay influence its performance. Below we analyse theinfluence of such non-standard service conditions on theperformance of equipment and the required safeguardsto achieve its required performance.Ambient temperatureThe rating of an indoor or outdoor switchgear assemblyis referred to at an ambient temperature of 40°C. For ahigher ambient temperature the rating of the assemblywill be reduced in the same proportion as for the busbarsystems and as shown in Table 28.3.AltitudeFor assemblies using the surrounding air as theinsulation and cooling mediumThe standard altitude for an LT switchgear assembly is2000 m and for HT 1000 m, as in ANSI-C-37.20C andIEC 60439-1. At altitudes higher than this, the normalincrease in temperature will become greater and thedielectric strength less for all assemblies using air as theinsulating and cooling medium. For applications at higheraltitudes, therefore, the derating factors as noted in Table13.12 must be applied to obtain the reduced level ofdielectric strength, i.e. one-minute power frequencyvoltage withstand, impulse voltage withstand and thecontinuous current rating etc. to which level the switchgearassembly will now be rendered to. To achieve the originallevel of dielectric strength, the insulation system of theswitchgear assembly will have to be improved by thesame degree as the derating. This can be achieved byTable 13.12 Derating for higher altitudes for metal enclosedswitchgear assemblies and bus systemsAltitude (mi ~ factorsLT systems HT systems 1 Dielectric strength Continuous currentrutirig~2000 1000 1 .0 1 .OO2600 1500 0.993900 3000 0.80 0.96Notes I Intermediate values may be obtained by interpolation.2 No derating i, applicable in case of insulated switchgear(GIS) assemblies.Source ANSI C-37.20.2 and C-37.23increasing the clearances and the creepage distances toground and between phases as discussed in Section 28.5.2and Tables 28.4 and 28.5.To achieve the original value of the continuous current,the size of the current-carrying components may also beincreased to carry the higher amount of current inproportion to the derating.Atmospheric conditionsA normal enclosure is meant for a reasonably cleanatmosphere and a relative humidity not more than SO%for LT and 95% for HT indoor enclosures. Where theatmosphere is laden with fumes or steam, saline or oilvapours, heat and humidity, excessive dust and water orcontaminated with explosive and fire hazardous gases,vapours or volatile liquids (Section 7.11) a specialenclosure with a higher degree of protection is requiredas in IEC 60529 or IEC 60079-14. For non-hazardousareas, the enclosure can be generally one of those discussedin Tables 1.10 and 1. I 1, and when required can be providedwith special treatment to the metallic surfaces. Forhazardous areas, however, special enclosures will beessential as discussed in Section 7.11.In outdoor type switchgear or controlgear assembliesthe normal practice is to provide a double door in thefront to house the front panel and protect the door knobs,meters, lights, pushbuttons, reset knobs or other accessoriesmounted on the door and thus prevenl water ordust leaking through joints, knockouts and fitments etc.It is also recommended to have a canopy on the top ofthe enclosure to protect the panel from direct rain. Figures13.6 and 13.28 illustrate this type of construction.Flame- or explosion-proof assemblies (Type Ex. d)For hazardous areas flameproof enclosures alone arerecommended, except in areas with moderate intensityof contamination and where such assemblies are locatedaway from the affected area and in a separate wellventilatedroom, when pressurized enclosures may alsobe safe. The reason for this precaution is that frequentarcing takes place within the enclosure on each switchingof a contactor, switch, breaker or an OCR etc. and alsoduring operation of power and auxiliary contactors.The classification of gases, vapour and volatile liquidsaccording to their ignition temperatures has been givenin Table 7.4. The basic requirements of these enclosures,

Switchgear and controlgear assemblies 131363With front doors openatmosphere. As a switch, a contactor or a breaker producesan arc during a switching operation, an explosion mayoccur within the enclosure on a re-switching. It is alsopossible that when the enclosure door is opened to check,test or replace a component, contaminants may haveentered the enclosure. It is also likely that on a re-closingof the feeder door, the closure is not perfect due to humanerror and contaminants have leaked into the enclosure.All this may lead to an explosion on a re-switching.Since it is not practical to manufacture a flameproofenclosure due to its size and bulk and the number ofknockouts and openings on the doors for switches,metering, indicators, and pushbuttons (PBs) etc., it iscommon practice to locate these assemblies some distancefrom the affected area in a separate well-ventilated room.Depending upon the location and intensity of contamination,it may be permissible to meet the requirement byusing a pressurized enclosure by maintaining a positivepressure inside the enclosure similar to that for motors(Section 7.13.3). When there are many switchgearassemblies, the room itself can be pressurized, which issafer and easier. Small enclosures, however, such as aPB station, switch or a switch fuse unit or an individualstarter unit etc., which can be easily made of MS platesor cast iron, as discussed in Section 7.13, can be mountedin the hazardous area while the main MCC can be installedin the control room, away from the contaminated areaand from where the process can be monitored.NoteAvoid soldering of cable and wire terminal joints at such installations,which may loosen with temperature and time.Excessive vibrations and seismic effectsInstallations prone to heavy vibrations and shocks suchas on the mounting platform near a large power-generatingunit, on a ship, locomotive and similar installations mayrequire provisions to absorb shocks and dampen vibrationsto prevent loosening of components and wire connectionsin a switchgear or a controlgear assembly. Some economicalmethods to achieve this can be by providing:With front doors closedFigure 13.28 A typical compartmentalized outdoor-type panel(Courtesy: ECS)which remain the same as for motors conforming to IEC60079-1, are also discussed in Chapter 7. Some of thevital requirements that a flameproof controlgear orswitchgear assembly must fulfil are, however, as follows:1 To prevent an electric arc or flame from the enclosureescaping to the atmosphere.2 To withstand, without distortion, an explosion, withinthe enclosure.An internal explosion is normally a consequence ofabsorption during a switch-on after a delayed shutdown.When an enclosure is switched OFF after normal running,the internal air cools and absorbs contaminants from theAnti-vibration pads at the foundation0 A rubber pad between the bottom of the assembly andits base orMounting pads below the base frame.But shocks of a seismic nature are different and may bemore violent. They may require a special steel structure,forming the switchgear or the controlgear assemblies, tosustain the shocks without damaging or looseningcomponents and wires mounted inside the assemblies. Itmay be possible to achieve this by using thicker sheetsfor the enclosures and providing additional reinforcement,wherever necessary, and exercising special care inselecting and fixing the hardware. The degree ofreinforcement and other provisions to make the wholeassembly and inside mounts suitable for a particularintensity of seismic effects will depend upon the probableseismic conditions in that area. This subject is dealt inwith more detail in Section 14.6.

13/364 Industrial Power Engineering and Applications HandbookNoteOne may observe that all the above special service conditions aregenerally the same as for the rotating machines discussed in Section1.6.13.5 Deciding the ratings of currentcarryingequipment, devicesand componentsThe rating of current-carrying equipment (switchgearassemblies, such as for the main bus system), devices(breakers, switches and contactors) and components(connecting links and wiring etc.) is defined by twoparameters:Thermal rating or continuous current rating Thistakes account of heat losses ( = I; . t, t being the timeto reach thermal equilibrium). as generated during along continuous operation under rated conditions. Inbreakers, switches and contactors, it will also definetheir duty cycle, i.e. the type of duty they may have toperform (Section 12.10 and Table 12.5).To the basic current requirement is applied thederating factors for various service conditions, as notedin Section 13.4.2. The equipment, devices andcomponents may then be chosen to be as close (nearesthigher) to this rating as possible from the availablestandard ratings. Based on these ratings, the minimumcross-sectional areas of the other current-carrying partsused in the circuit. such as interconnecting links andthe cables-are calculated.Short-time rating This will defineThe electrical rating on a fault or short-term high currentthermal effects, expressed by a I& . tsc , i.e., short-circuitcurrent (Isc) and its duration (tSC- 1 or 3 seconds).The mechanical endurance of the current-carrying partsof all the equipment, bus system, devices andcomponents, used in a particular circuit as well as theload-bearing members and supports on which they aremounted. The electrical parts of a device (breakersand switches, etc.) are the responsibility of thecomponent manufacturers. The manufacturer of theswitchgear assembly is responsible for the busbarsystems, metallic links and wires.The mechanical requirements of the load-bearingmembers, supports, busbars and metallic links in theincoming circuit will depend upon the electrodynamicforces arising out of the first major peak of the faultcurrent, as discussed in Section 28.4.2. This is not,however, applicable to devices that are current limitingas well as the circuits and its components that areprotected by it. These devices isolate the faulty circuitlong before the fault current reaches its first highestpeak. Yet the short-time rating for the incoming circuit(as the outgoing circuits may not experience thesefaults), should be selected from the standard shorttimeratings only, as defined in Tables 13.7 and 13.10.For this short-time rating is then calculated the minimumcross-sectional area of the current-carrying components.To determine the cross-sectional area for this shorttimerating and decide the mounting structure, supportsand hardware etc. for the busbars, a brief procedure isdescribed in Example 28.12.Both ratings, thermal and short-time, will define thecross-sectional area of the current-carrying components.The more severe of the two will then prevail. In thefollowing text we explain the procedure to assign theshort-time current rating to equipment, a device or acomponent.13.5.1 Assigning a short-time ratingThe bus system of a switchgear assembly, its interconnectinglinks and wires are the protected type components,whereas an interrupter (breaker, switch or a fuse) maybe a protecting or protected type, depending upon theirapplication and location in the circuit. A contactor andan OCR are therefore protected devices in the samecontext. for they provide no short-time protection. Aprotecting device may become protected when it is alsoprovided with a back-up protection.A breaker, usually an MCCB or an MCB on an LTsystem, can be provided with backup HRC fuses toenhance their short-time rating. This may be done when theavailable MCCBs or MCBs possess a lower short-timerating than the fault level of the circuit they are requiredto protect, and make them suitable for the fault level ofthe circuit. But this is not a preferred practice and isseldom used. As a rule of thumb, the device that isprotecting must be suitable to withstand electrically andendure mechanically the system fault current for a durationof one or three seconds, according to the system design.The duration, however, is no criterion for a currentlimiting type protecting device, and a protected equipment,device or component can have a short-time ratingcommensurate with the tripping characteristics of theprotecting interrupter. Accordingly these two types oftripping characteristics are explained below.Delay trippingThe breakers (OCBs, MOCBs, ABCBs, SF6 and VCBs)for HT and ACBs and MCCBs* for LT systems are devicesthat have a tripping time of more than a cycle on fault.The tripping time can vary from 1 cycle (20 ms) to 5-6cycles (100-1 20 ms for 50 Hz systems), depending uponthe type of interrupter as discussed in Chapter 19 andnoted in Table 19. I .This time allows the fault current to reach its peak andtherefore all the equipment, devices and componentsprotected by such a device must be suitable for the fullfault level of the system. While the tripping time is usuallyin milliseconds the duration of fault, tSC. is consideredas one or three seconds. The longer duration than necessaryis to account for the various time lapses that may occur*MCCBs(i) When they are normal duty as considered here.(ii) They are also available in the current limiting type. Whencurrent limiting. they will not fall into this category.

Switchgear and controlgear assemblies 13/365before the actual tripping. When the tripping is electricalit may involve a trip coil and a motor mechanism adding10 the mechanical tripping time, actuating time of relays,minimum time of tripping of interrupter itself (Table19. I and some safety margins. While the actual trippingtime may still be quite low, it is customary to design asystem for one or three seconds and for which is designedall the equipment. devices and components, protectedby such a device.The value of short-time rating . t,') of the systemmay now exceed, the thermal rating of some of theequipment, devices and components, i.e. Ithermal rating. This condition may be more Ismaller ratings, particularly 600 A and less (such as forbusbars), than larger onec. At higher currents, the naturalthermal rating itself, due to the higher cross-sectionalarea. will become higher than the required short-timerating. The short-time rating will remain the same for aparticular fault level of a system, irrespective of the currentrating of the circuit. For more details refer to Section28.4.1. Whenever the short-time rating exceeds the thermalrating. a larger area of cross-section of the main busbarsystem and the other current-carrying components willbecome necessary.Figure 13.29 illustrates a simple distribution systemand location of the main buses, devices and componentsto define the current ratings of all such devices andcomponents under different operating conditions. Theideal current ratings of these components are given inTable 13.13 for an easy illustration.The fault currents also develop electrodynamic forces,F,,,. as in equation (28.4) due to the sub-transient d.c.component. These forces play an important role in themechanical design of the interrupting device, the loadbearingand mounting structures for the interrupter andthe bus system, and the hardware used in a switchgearassembly. All such mechanical parts, supports andhardware should be adequate to withstand such forceswhen they arise. A procedure to arrive at the ideal size ofthe current-carrying components, mounting structure, typeof supports and hardware etc. is discussed in detail inExample 28.12.If the short-time rating of the interrupting device ishigher than the fault level of the system, which is thecase with modern interrupting devices, the fault level ofthe system alone will prevail for the busbars, componentsand hardware. For example. for a system fault level of50 kA, if the interrupter used is of 65 kA short-timerating, the bus system and all associated componentswill be designed for SO kA only.Current limiting typeExamples are HRC fuses (both LT and HT) and MCCBsand MCBs (LT only). which are available with currentlimiting features and are in extensive use. The trippingtime of these devices is extremely low and much lessthan one half of a cycle of the current wave. They thereforedo not allow the fault current to rise to its prospectivepeak. The protected devices and components can thus beselected based on the let-out energy of such devices onfault. which is extremely low, than the fault level of thesystcm. IfI,, = system fault levelt,, = pre-arcing time of the HRC fuses or tripping timeof MCCBs or MCBs, which would be much lessthan S ms (less than one quarter of a cycle for a50Hz system; see Figure 12.18 for more clarity)then the let-out energy of the current limiter,I,: ' t,,If I, is the equivalent I -second fault currentthen I,'or. 1 = 13 . t,,I, = I,, vFiSince tSc will be too low (< 5 ms for a 50 Hz system)I, will be much less than 7% of I,, in all situations. Toassign a short-time rating to the protected devices andcomponents in such cases is therefore of little relevance.As noted above, current is the cause of heat, for which isassigned the thermal duty of a current-carrying device.component or part. Also note the following:Current limiting devices need not be protected. sincethey are already very fast acting and, hence, selfprotected.But they are also rated for the same fault level forwhich the system is designed as they are connecteddirectly to the system. This is a safety requirement.Similarly, in a draw-out switchgear assembly. the I/Cand O/G power contacts of a module and its mounts(insulators and supports) being already protected maybe suitable only for the thermal rating of their feeders.When such devices are used at more than one locationin a circuit, their ratings must be meticulouslycoordinated to ensure isolation of the faulty circuitalone. Refer to Section 12.4.2 and Figure 12.1 1 formore clarity.A current-carrying device or component in adistribution network may be subjected to varyingdegree of electrodynamic stresses, depending uponits location with reference to the network. Referringto Figure 13.29, the circuits away from the source ofsupply are subjected to lower stresses than the circuitsnearer to it. Accordingly, the coordination is donebetween the protective devices, used in the upper andlower streams of circuits, to ensure that only the faultycircuit is isolated, rather than isolating other circuitsin the upper stream (Section 12.4.1).Notes on Table 13.13An isolator, such as at locations C, and C2 in Figure13.29, may sometimes be used to isolate the circuitsit is feeding, say, for maintenance or repair\. Thisisolator is simply a switch and provides no protectionto the circuits. For a fault on the outgoing side, theindividual outgoing feeders must have their ownprotection. For a severe fault elsewhere in the system,there must be a protective feeder closeby, in the upperstream.Role of an OCR: this is only an overcurrent protectiondevice and does not provide short-circuit protection.

13/366 Industrial Power Engineering and Applications Handbook50 Hz systemFault level50kA, 1 sec________-___I 1-[Not a recommended practicewhen high rupturing capacityinterrupters are available.]. _ . _ . -. _ .impedance)I.l~ < 8.5kAi< 50kA, 1 sec.1 sec.-All OIGfeeders< 50kAAll OIGfeeders

Components< Interrupting devices > 1 -Equipment and components AApplicationWhen used as a protective device(as incomer)When used as a protected device'as outgoing)'When used as anisolator (as incomer)As protected partsParameters(1)Tripping time> 1 cycle(2)Tripping time< half cycle(3)Tripping time Tripping time> 1 cycle< 1/2 cycle(4) (5)Interrupter tripping time > 1 Interrupter tripping timecycle< half cycle(7) (8)Type of devicesOCBs", ACBs,MCCBs in LT andMOCBS~ABCBsb, SF, andVCBs in HTMCCBs,MCBs in LT and HRCfuses in both LTand HTAs in column 2 As in column 3 Switching devicelocations C, and Cz, asFhown in Figure 13.29Main bus, its feeding tap-offsand interconnecting links orcables etc. and all outgoinglinks, interconnecting devicesthat are also connected on themain bus such as shown atsection F, Figure 13.29.All outgoingconnections by solidlinks or cables1, Short-time ratingor symmetricalfault levelI,,I,,, if connected on the main bus, otherwise< Isc depending upon the circuit impedanceDepending upon theprotective feeder on theupstream. Refer toFigure 13.292 Duration of fault 1 or 3 secondsd

13/368 Industrial Power Engineering and Applications Handbookin Section 13.4.2 to arrive at the most appropriatesizes of components, bus sections, etc.13.6 Designing a bus systemWe discuss in detail in Chapter 28, the procedure todesign a bus system, including its mounting and supportingstructure and hardware for a required fault level.13.6.1 Constructional features of a bus system(i) Busbars and wirewaysIn the cubicle construction of a switchgear assembly thebusbar chamber is normally located at the top of theassembly and runs through the length of it. It is usuallysuitable for extension, through fish joints at either end,if required at a later date. For installations having topcable entry, the busbar chamber may also be located atthe bottom of the assembly or the depth of the panelincreased, with an additional shroud between the topbusbar chamber and cable chamber. From these mainbusbars are tapped the vertical buses for each verticalpanel. Manufacturers may adopt different practices forhorizontal and vertical busbar arrangements to economizeon their cost of production. We illustrate the most commontypes of busbar arrangements.A separate control wireway may also run through thesame busbar chamber, with suitable segregation orshrouding between the main bus and the control bus.This arrangement can be seen in Figures 13.2 and 13.7.The control bus system may be required for one or moreauxiliary supplies for the following auxiliary services.Motor winding heating up to 30 kW: control bus voltage24 V a.c.Motor space heaters above 30 kW: control bus voltagegenerallyv,43A.C. control supply: control bus voltage generallyllOV orNoteThe interpanel control wiring for interlocking between feeders,space heaters and panel illumination will also run through thiswireway or control bus chamber.(ii) Busbar mounting configurationsManufacturers may adopt different practices to mountthe main and auxiliary busbars, depending upon the size,rating and fault level of the system. Some of the recommendedand more common of these are illustrated inFigure 13.30(a)-(d) and discussed briefly below,I NElevation(a)SectionalviewFigure 13.30 Possible arrangements for busbar mountingsystemsArrangement (a)Busbars are mounted one below the other, horizontallybut in a vertical disposition. The cooling is better andrequires less derating. The short-circuit withstand capacityis high due to high sectional modulus but occupies morevertical space. This configuration is also adopted by somemanufacturers.Arrangement (b)This is similar to (a) above but each busbar now is mountedhorizontally. Due to obstruction in heat dissipation, thisarrangement requires a higher derating. It is also proneto collecting dust and provides a habitable surface forlizards and rodents etc. Therefore this is not a recommendedconfiguration.Arrangement (c)This is similar to (a) except that now they are in thesame plane and are not one below the other. Althoughheat dissipation would be slightly better than (b), thistoo is not a recommended configuration.Arrangement (d)All busbars are now in one plane and in a verticaldisposition. This is the most appropriate and mostcommonly adopted configuration. With such an arrangementany rating is possible. For higher ratings, the CopperDevelopment Association (UK) have recommended manymore configurations of busbar arrangements with a viewto have a better utilization of the metal up to its optimumcapacity. For more details refer to Section 28.7.2(iii) andFigure 28.14.(iii) Busbar mounting systemsTo obtain a strong busbar mounting system, suitable towithsland the electrodynamic forces arising out of a systemfault, modern practice is to make use of thermosettingplastics, such as DMC (Dough Moulding Compounds)

Switchgear and controlgear assemblies 13/369and SMC (Sheet Moulding Compounds) for the busbarmounting supports. These compounds are suitable forcompression, transfer mouldings and injection mouldings.They are basically fibre- or glass-reinforced thermosettingplastics (FRP or GRP) and possess good physical andthermal stability, high mechanical strength and excellentelectrical properties as shown in Table 13.14.Moulding compoundsWith the advent of these compounds in the 1960s, thehitherto more conventional insulating materials, such asphenol formaldehyde (popularly known as Bakelite) andwood (veneered impregnated) have been almost replacedby them. These compounds offer better electromechanicalproperties than conventional materials. Below we describethe basic mix and properties of these two basic compounds,for a brief reference.DMC (Dough Moulding Compound)This is also known as Bulk Moulding Compound (BMC).It is blended through a mix of unsaturated polyester resin,crosslinking monomer, catalyst, mineral fillers and shortlengthfibrous reinforcement materials such as choppedglass fibre, usually in lengths of 6-25 mm. They are allmixed in different proportions to obtain the requiredelectromechanical properties. The mix is processed andcured for a specific time, under a prescribed pressureand temperature, to obtain the DMC.SMC (Sheet Moulding Compound)This is a material produced from the impregnation ofglass fibre-mat (fibreglass, which is in the form of drysheet, is commonly known as chopped stranded mat(CSM)) or rovings, with a liquid and unsaturated polyesterresin, which thickens chemically to a dry sheet form.The total mix is sandwiched between polyethylene filmsand then roller-pressed to impregnate and consolidate it.The chemical thickening enables the material to be handledafter the polyethylene film has been removed beforemoulding.SMC is used where its superior strength and impactresistance over DMC are more important. The improvedproperties, particularly its strength, over DMC is a resultof reduced degradation of the glass and the ability to uselonger fibre. In DMC, this is usually 6-25 mm, while inSMC it is about 25-50 mm.The compounds so formed have excellent thermalstability and are self-extinguishing and even completelyfire-retardant. Their properties are given in Table 13.14.A few common types of insulators and supports are shownin Figure 13.31.(iv) Making busbar connectionsAluminum, being a highly oxidizing and malleable metal,requires utmost precautions when making a connectionor a joint. The joint may be a fish joint for connectingtwo straight sections of a bus, tap-offs or even thethimbling of cables on the aluminium extended links. Toavoid a rapid formation of non-conducting oxide film onthe surface of the metal, the surface must be treatedproperly before making the joint. To avoid this one maytake the following precautions:1 Clean the surface with a wire brush to loosen theoxide film and then wipe it off with a soft cloth. Theuse of a wire brush serves a dual purpose; first, scrapingand removing the oxide film, and secondly, providingthe surface with a moderate knurling (roughness),which helps to make a better surface-to-surface contactand, in turn, a better joint.2 Apply a contact grease with the following properties:0 To be chemically neutralTo have a nil or negligible electrical resistanceandTable 13.14Properties of thermosetting plasticsProperties Unit DMC SMC Relevant standardsfor test methodsPhysical properties1 Specific gravity2 Shrinkage3 Water absorption4 Operating temperatureMechanical properties1 Minimum tensile strength2 Minimum compressive strength3 Minimum cross-breaking strength orflexural strength (bending strength)4 Impact strengthElectrical properties1 Dielectric strength (min.)2 Tracking indexgdccmdmm%%"Ckgf/cm2kgflcm2kgflcm2kgf cdcm2kVlmmV1.85-1.95002-.0030.2-0.30.2 (m.)140-150250-5001200-1800700-120030-4010-1410001.8,00150.150.15 (m)140-150BS:278250&900 BS:27821600-2000 BS:27821400-1 800 BS:278260 BS:278210-14 BS:27821000 BS:5901Note These values are only indicative and may vary with the quality of mix and process of curing etc., and differ from one manufacturerto another. For exact values, contact the manufacturer.

13/370 Industrial Power Engineering and Applications Handbook(c) Any other chemically neutral grease, whichwill have no electrical resistance and whichcan withstand a minimum tracking temperatureof 200°C.3 The joints must be tightened with a torque wrench.For the recommended torque values refer to Table29.1.NotePetroleum jelly is not recommended due to its low trackingtemperature. The minimum tracking temperature is recommendedto be 200°C, the same as for the busbars during a fault. Also referto Section 28.4.1.Figure 13.31(a) Insulators to hold busbars in flat configuration(Courtesy: J.K. Plastics).Figure 13.31(b) Insulators to hold busbars in vertical configuration(Courtesy: Vinayak Gorp.)Figure 13.31 (c) Conical insulators (Courtesy: J.K. Plastics)To have a minimum tracking temperature (atwhich the grease may start conducting) of 200°C.Apply this grease swiftly after the surface cleaningto avoid a fresh oxidation. The followingare a few types of greases:(a) Servogem 2 (multipurpose) from the IndianOil Corporation(b) Multipurpose grease H from HindustanPetroleum Ltd orFastenersOnly high tensile (HT) fasteners must be used for busbarjointing and their interconnections or links not only totake care of the fault level but to also maintain therecommended contact pressure over a long period ofoperation as noted in Table 29.1. An ordinary fastenermay not be able to withstand or sustain this torque forlong. Similarly, the busbar supports, which are mountedon only two or three fasteners, should also be fitted withthese fasteners.Electroplating of HT fastenersHT fasteners in normal manufacturing are blackphosphated and then lubricated. They are not required tobe electroplated, as they do not rust unless the phosphatecoat itself is damaged. Such fasteners when used forelectrical purposes, such as for mounting and jointing ofbusbars and their supports, are not generally exposed tooutdoor conditions. The phosphate coating thus remainsintact and an electroplating (zinc or tin passivation) isnot required.Moreover, electroplating of HT fasteners may posethe problem of hydrogen embrittlement, which can causecracks on their surfaces. The HT fasteners are alreadyheat-treated and have a high content of carbon. Whenthey are electroplated, whatever hydrogen they may emitduring acid pickling is trapped on the surface, as it formsa strong bonding with the carbon. The C-H bondingrenders the surface brittle. Rapid removal of hydrogentherefore becomes essential to save the hardware fromsurface cracks. It is possible to do this by tempering thehardware at a low temperature, say, 100-120°C for about30 minutes, before transferring them to the electroplatingbath.If such fasteners are stove-enamelled (which is normallynot done), the trapped hydrogen is removed automaticallywhile being stoved. Since the fasteners are used only forthe assembly of switchgear or busbar systems, they areused at room temperature only. Therefore, if they areelectroplated, they must be tampered, which is timeconsuming and adds to the cost of production. Moreoverthis has no technical advantage. The HT hardware musttherefore be used as they are. When it is absolutelynecessary to electroplate them, tampering will be essential.13.6.2 Service conditionsThese are the same as discussed in Section 13.4.2. For

Switchgear and controlgear assemblies 131371busbar systems, however, a more elaborate exercise wouldbe necessary in high rating systems, 1600 A and above,due to skin and proximity effects as discussed in Section28.6.3.13.6.3 Complying with design parametersRated voltage and frequencyThe switchgear assembly, its components and the bussystem must be designed for the rated voltage andfrequency.Rated insulation levelTo comply with the rated insulation level, all the currentcarryingcomponents forming part of the assembly shouldhave clearances and creepage distances according to theirrelevant standards whereas busbars and busbar connectionsmust have the distances noted below.Clearance and creepage distances forair-insulated busbarsThe clearances and creepage distances should be maintainedas shown in Tables 28.4 and 28.5. These valuescan be reduced when:(a) A barrier of insulation is provided between theconducting parts, such that the clearances andcreepages now achieved are no less than as specifiedin Tables 28.4 and 28.5.The current-carrying conductors are covered withan insulation suitable for withstanding a one-minutepower frequency voltage as in Tables 13.2 and 14.3for series I and 14.1 and 14.2 for series I1 voltagesystems.Any provision or arrangement that can withstandthe one-minute power frequency voltage as in thesetables at a lesser clearance or creepage distancesthan specified in Tables 28.4 and 28.5, such as byproviding extra insulation wherever necessary. Toobtain clearances for open-type outdoor and neutralgrounding switchgears, refer to BS 7354.13.7 Designing a switchgearassembly13.7.1 Rated continuous current rating andpermissible temperature riseThe rating of current-canying devices and componentsshould be selected according to the continuous currentthey have to carry and the duty they have to perform.Deratings, depending upon the service conditions, shouldalso be applied when deciding their continuous currentTable 13.15 Current rating and technical data for 1100 V, single-core flexible, PVC insulated copper conductor cables forcontrol and power wiringCross-sectional Equivalent diameter Nominal thickness Nominal overall Maximum resistance Current rating d.c. orarea of copper conductors of insulation diameter at 20°C single phase a.c. at30°C ambientmm2 mm mm mm nlkm A0.5 0.94 0.6 2.3 37.10 40.75 1.20 0.6 2.55 24.70 71 .o 1.34 0.6 2.70 18.50 111.5 1.6052.5 2.14.0 2.616.0 3.210.0 4.616.0 5.90.60.70.80.81 .o1.42.95 12.70 143.65 7.60 194.35 4.7 1 265.65 3.10 467.15 1.884 648.95 1.138 8525.0 7.635.0 8.750.0 10.61.4 10.65 0.6845 1121.4 11.75 0.5227 1381.6 14.05 0.3538 172As in IEC 60540Courtesy Finolex Cables Ltd.Notes1 Consider an average derating of 67% for power cables when operating at an ambient temperature of 50°C.2 Consider an average derating of 0.8 for a number of power cables bunched together, generally not more than six at a time.3 For control cables, derating may not be material in view of very low control currents. Whenever required, the following averagederatings may be consideredup to 10 cables 0.7020 cables 0.6030 cables 0.5040 cables 0.40

+13/372 Industrial Power Engineering and Applications Handbookrating. The sizes selected must then be counter checkedfor their mechanical endurance to sustain the faultconditions of the system or the circuits on which theyare connected, depending upon the protective schemeadopted and its time of isolation on fault (Le. 1 s, 3 s orcurrent limiting) as discussed above.The ratings and sizes of main components and cablescan be selected from manufacturers' catalogues. But cablesrequired for the switchgear internal control and powerwirings, being typical of all, are normally identified bytheir cross-sectional area rather than the current ratings.We have therefore provided the technical data and currentratings for the most common sizes of such cables for aready reference in Table 13.15.Horizontalbusbarchamber13.7.2 Design considerations for switchgearassembliesBelow we discuss briefly the constructional requirementsand general manufacturing practices for cubicle-typeswitchgear and controlgear assemblies, and the electricaland the mechanical design considerations to comply withthe above design parameters and service conditions.Constructional requirements and generalmanufacturing practicesThickness of sheet steelLoad-bearing members and frame Two to threemm (14, 12 or 10 SWG) depending upon the size ofstructure and weight of the components to be mounted.Covers and partitions From 1.6 to 2 mm (16 or 14SWG). Larger size of doors, doors having a number ofrelays, instruments and other devices. Also doors formimic control panels etc., required to be mounted witha number of instruments, relays or indicating devicesand carrying their load and wiring weight, should bemade of thicker gauges and/or stiffeners must beprovided at the back of the door for strength and toavoid shaking and buckling of doors.Base frame Three to four mm (10 to 8 SWG) MSSheet or MS channel of section ISMC-75 (75 mmwide) or ISMC-100 (100 mm wide) depending uponthe size and weight of the assembly as shown in Figure13.32 and 13.48.Gland plate Three to four mm of MS or non-magneticmaterial, depending upon the number, sizes and typeof cables (single core or multicore) it has to carry(Figure 13.33).NoteAll the three phases (R, Y and B) of a single-core or a multicorecable must pass through a common opening in the gland plate.When this is not possible, such as when using single-core poweicables,and each core is required to pass through a separate glandto hold it securely in place, the gland plates, through which thesecables will pass, must be made of a non-magnetic material(aluminium, SMC/DMC or Bakelite etc). This is an importantrequirement to eliminate electromagnetic induction in all surfacesthat have magnetic properties due to the proximity effect (Section28.8) caused by each phase. In a three-core cable, the field inducedI1' +Front elevation+ +IView of base frame (plan)(All dimensions in rnm)IFoundation/ holesFigure 13.32 Typical module sizesBottom clearance foreasy operation ofmodule handle andpush buttons etc.by each phase, being in a circular form, is ncutralized, due to phasetranuposition, (Section 28.8.4). In individual single core cables eachcore produces its own field, which is not neutralized and createsmagnetic currents. causing eddy current and hysteresis losses inthe gland plate if it is made of MS. This may cause excessive heatin the gland plate and result in insulation failure of the cables. Itmay also lead to a short-circuit condition.CorollaryIt would bc interesting to note that to eliminate S U C a ~ phenomenonin large metal-enclosed current-carrying systems a segregated phasebus system is in fact preferred, to shield the magnetic influence ofone phase on the other (Section 28.2.2). In a segregated system theconductor on each phase is enclosed by metallic barriers, similar tothe cable by the gland plate. But a gland plate is totally differentfrom a segregated system. The thickness of gland plate of only3-4 mm, provides no shielding for the field produced by each core

Switchgear and controlgear assemblies 13/373Before the cablingAfter the cablingFigure 13.33 Arrangement of gland platesof cable running through the gland plate. In a segregated system,the enclosure runs the full length of the conductors and providestotal shielding for the induced fields. Although the rating of cablesmay not be high for the purpose of the proximity effect, the closevicinity of its phases, when not in a circular form, may causeenough mutual induction between them, which can heat the glandplate beyond safe limits. MS plates have been seen to melt as aresult of this effect leading to a severe fault. Hence the necessityfor providing the gland plate of a non-magnetic material to eliminatethe main heating constituents, hysteresis and eddy current losses.For details on the proximity effect and magnetic shielding, refer to(Sections 28.8 and 27.3.2).Other requirements1 Provision of a segregation between the adjacentfeeders in a modular design.2 Wherever a circuit breaker is to be housed, this shouldbe in a separate compartment.3 Shrouds and shutters are essential to cover all liveparts in a feeder module that may be exposed to theoperator when the feeder door is opened (see Figure13.3). This is a safety requirement for the operatorattending the feeder. There may be two types ofdoors for the accessibility to the live parts:(a) Doors which may not be opened for carryingout day-to-day operation or maintenance. Suchdoors may be almost the fixed type, as for abusbar chamber or the rear panels of a frontoperatedassembly. They may be bolted to theframe so that, when required, they can be openedonly with the use of prescribed tools. This is asafety requirement for access to the live partsonly by authorized persons. To provide an extrashroud between the door and the live buses/parts in such cases is not mandatory.(b) Doors which may be opened often to carry outday-to-day operation and maintenance. Theyshould be the removable type and openedmanually. In these cases, all parts that may stillbe live, even after the switching device has beenturned OW, before opening the door must beprovided with a shroud or a shutter (this is amandatory requirement as discussed later underinterlocking schemes). Refer to Figure 13.3showing such an arrangement.

13/374 Industrial Power Engineering and Applications HandbookNoteFor instrument modules, relay and control modules or control panelsor all power modules, where an interlock with the door is notpossible or is not provided, a proper shroud or shutter must beprovided on all exposed live parts rated above 240 V.To provide a folded and extensible construction toallow for ease of alteration and extension of assembliesat site in future, if required.To have a modular construction with a wide choiceof module sizes for optimum utilization of the usablearea in each vertical panel, which is normally1800 mm as illustrated in Figure 13.32. The generalpractice is to have the module sizes in the ratio of1/6 (300 mm), 1/4 (450 mm), 1/3 (600 mm) and1/2 (900 mm), etc. Some manufacturers, however,supply 1/8 (225 mm) and 1/9 (200 mm) size ofmodules when the sizes and number of componentsfor a module are less and can be accommodated insuch a small module size. For critical installations,however, such as for a refinery or a petrochemicalplant or for the essential services of a powergeneratingstation or installations that are in humidconditions or are contaminated, it is advisable tohave a module size no smaller than 1/6 (300 mm)with a view to providing more space within the feederto achieve larger clearances between the live partsand to lessen the chances of a fault during normaloperation, besides providing extra working space.6 The operating height of each operable mount, i.e.the centreline of the breaker or the switch handleand pushbuttons or reset buttons, including the resetprobe of a protective relay on the feeder doors, isrecommended to be no higher than 1900 mm fromground level. This is an operational requirement forease of operation.7 The height (we may consider it from the bottomline) of the indicating instruments such as a voltmeterand an ammeter, which the operator may have toread often, is also recommended to be no higherthan 2000 mm or less than 300 mm from groundlevel.8 The terminals provided to receive cables to makeexternal connections, may preferably be located noless than 200 mm (at the terminal centreline) fromthe gland plate.9 For ease of maintenance, all the busbars, horizontalor vertical, control or auxiliary, should be easilyapproachable. Figure 13.34 shows a typical rear viewof an assembly with the main horizontal and verticalPress fit SMC/DMC shrouds on the bus bar jointsPVC shrinkablesleeve on busbars3 pin socket forlightingBarrier on the terminalsto protect from ahuman contact andfalling tools-Cable alleysTranslucent shroon the terminalsFigure 13.34 Rear view of a typical MCC showing space heaters at the bottom of each cable alley and shrouding of the live parts

Switchgear and controlgear assemblies 13/375IO111213buses and Figure 13.2, illustrates the front of thesame assembly with two sets of single-phase controlbuses.Also the cable alley should be easily accessible tomake cable connections, and facilitate easy maintenanceand regular checks, as illustrated in Figures13.2 and 13.34.Shrouds are recommended in the front and on thetop of the terminals of each feeder to provideprotection to the operator from live parts. They willalso prevent the tools falling inadvertently from anupper module onto the live terminals of the lowermodule. Look closely at Figure 13.34 for thesefeatures, where in the front is provided a typicaltranslucent shroud to enable a check of the terminals,without opening the shroud. On top is providedanother shroud to prevent the terminals from fallingtools. If the shroud is of polycarbonate (acrylic hasa low temperature index), it should be suitable towithstand a temperature of up to 200°C withoutdeformation. This temperature may be reached duringa fault at the terminals.For safety reasons, the busbar chamber and the cablealley should be separate and shrouded from eachother.Where wires or conductors may pass through a metalsheet, a rubber grommet, bushing or other mechanicalprotection should be provided to prevent the wiresfrom insulation damage.Mechanical interlocksProvisions must be made for safety features such aspadlocking and door interlocking arrangements. Doorinterlocking is required to ensure that the feeder doordoes not open when the feeder is live.Similarly, it must not be possible to switch the feederON when feeder door is open.A defeat mechanism to bypass the door interlock mayalso be necessary for the purpose of testing.Padlocking arrangements may be required to lock thefeeder in the OFF position when the machine isundergoing a shutdown or repairs.Refer to Figure 13.12 showing these features.Protection from electric shocks (grounding system)1 Main grounding The provision of a groundingarrangement is mandatory through the length of thepanel. It may be of aluminium, galvanized iron (GI)or copper. (See also Section 22.4.)2 Grounding of each feeder The most effective systemis to ground each feeder with the main ground bus atone point at least. It is important to note that eachfeeder is grounded automatically through the metallicsupports of the assembly, on which are bolted all theswitchgear components (the whole assembly is alreadygrounded). For an ideal condition, an additionalgrounding of the components should normally not berequired but this grounding may not be foolproof dueto the painted frame on which the switchgear componentsare mounted. It is possible that the componentsmay not make a perfect ground contact through thebody of the switchgear frame and it is thereforerecommended that each comDonent is seuaratelvgrounded. For a cubicle design, a separate groundbus of a smaller section than the main ground busmay be run through each vertical section and connectedFigure 13.35 View of a module of a fully drawn-out MCC showing grounding arrangement through a continuous runningground bus

13/376 Industrial Power Engineering and Applications Handbookto the main ground bus and each individual feedermust be bolted through this. Figure 13.35 illustrates thisarrangement through one module of a vertical panel.3 Door grounding Similarly, Ihe door is also a part ofthe main frame and is automatically grounded throughthe mounting hinges and the door closing knobsllatchesetc. But a separate door ground wire connecting theframe is now also recommended. Where, however,there is no door wiring, no additional door groundingis essential.NoteFor more details on grounding and the grounding practices, refer toChapters 21 and 22.13.7.3 Essential features of a draw-out MCC1 All power and controlgears are mounted on withdrawablechassis (Figure 13.36).2 The chassis moves on low-friction rolling mountsor guide rails (Figures 13.2 and 13.36).3 Guide rails are telescopic and are necessary to ensuresafe and aligned movement of the trolley whileracking it in or out of its module to avoid misalignmentof the moving contacts. A misalignment maycause an inadvertent contact of the draw-out contactswith the adjacent fixed contacts of the other phases,which are mounted on the live vertical bus (Figure13.35) and may cause a flashover and a short-circuit.4 The chassis for both fully draw-out and semi-drawoutMCCs is fitted with self-aligning plug-in-typehigh-pressure contacts for incoming and outgoingpower connections. The control terminals for controlconnections are the manually connected, plug-in typein semi-draw-out-type MCC, as shown in Figure13.13(a) and spring-loaded sliding-type in a fullydraw-out MCC, as shown in Figures 13.13(b) and(c).5 It is preferable to provide a cranking or a push-indevice with a latching arrangement to lock the trolleyin both service and isolated positions, as shown inFigure 13.36.Test position(front slot)Details at ‘D’Latching deviceFigure 13.36 Front view of a typical withdrawable chassis (trolley)

Switchgear and controlgear assemblies 13/3776 For a positive ground connection and to providetotal safety for the operator, a spring-loaded scrapingtypegrounding assembly may be provided on eachtrolley so that the moving part can slide on the fixedgrounding strip fitted in the assembly, and makebefore and break after the power contacts haveengaged or disengaged respectively. Figures 13.2and 13.35 illustrate such an arrangement. Thegrounding contact may be of brass and silver plated(a good practice). It may be fitted on the tray ofeach module and permanently connected to theground bus running vertically through each verticalpanel and connected to the main horizontal groundbus.A positive grounding arrangement will meanautomatic grounding of the trolley as soon as it ismounted on the tray for racking-in and then remainingin permanent contact with the ground contact whenthe trolley is fully racked into its service position.7 In a draw-out MCC, it is recommended that shroudsare provided at the main incoming power contacts,from where the vertical bus is tapped, to feed thedraw-out module. Manufacturers may adopt differentpractices to achieve this. The most common is aDMC/SMC (Bakelite, which was more commonearlier, is now being discarded, being hygroscopicand inflammable) gravity-operated drop-down shutter,which lifts automatically while the trolley is beingracked in and slides down as the trolley is beingracked out. Refer to Figure 13.35, showing a feederwith the shroud lifted. Figure 13.2 also illustrates afew feeders with shrouds fitted. One of the feedersis shown with a lifted shroud.8 Indicating instruments, lamps, pushbuttons, selectorswitches and reset knobs etc. are mounted on thetrolley on a hinged auxiliary door, as shown in Figures13.2 and 13.12.The main outer door on the frame may have eitheran opening to seat the auxiliary door of the trolleyon it or telescopic knock-outs for all such door mountsto provide a peep-through type of aperture. On itare mounted the light and pushbutton tops. Figures13.2 and 13.12 illustrate this type of arrangement.The latter alternative provides a better arrangement,for it ensures a greater degree of protection. It hasthe most significant advantage when the trolley isremoved from its module and the outer door canstill be securely shut on the module to provide totalprotection for the empty module from dust, verminand rodents as well as inadvertent human contactwith the live incoming terminals. In the other design,the outer door has a large knock-out to seat theauxiliary door. which may remain open when thetrolley is removed for repairs or replacement, andexpose the interiors.9 Low contact resistance is desirable at all currentcarryingcontacts, such as the busbar joints, betweenbusbars and the incoming fixed power contacts,between incoming fixed power contacts and incomingmoving power contacts on the trolley and betweenthe outgoing moving and fixed power contacts etc.This is to ensure proper surface-to-surface contactand eliminate arcing between them. Otherwise itmay develop hot spots and result in corrosion of thecontacts in the normal operation. It may also lead toan eventual failure of the joint/contact.N(JtcThe main incoming male contacts are generally made of copper orbrass and are either bolted or clamped on the vertical bus. Since thebus is generally of aluminium. the contacts may form a bimetallicjoint with the bu\bars and cause corrosion and pitting of the metal.This may result in a failure of the joint in due courTe. To minimizemetal oxidation and bimetallic corrosion, the contacts must be silverplated.If the main incoming male contacts are made of aluminiumalloy, which ic normally a composition of aluminium-magnesiumand silicon, they must be provided with a coat of bronre. copperand tin to give it an adequate mechanical hardness and reriqtance tocorrosion. For more details refer to Section 29.2.5.1011121314151617Each trolley is recommended to possess three distinctpositions of movement, as shown in Figure 13.36.i.e. ‘service’, ‘test’ and ‘isolation’:Service: this is the position of the trolley when it isfully inserted into its housing (module) and the powerand control contacts are fully made.Test: This is the position of the trolley when thepower contacts are isolated but the control circuit isstill connected, because it is lapped directly fromthe auxiliary bus. This condition is essential tofacilitate testing of control circuits with functionalinterlocks, without energizing the connected load.Isolation: This is the position of the trolley whenthe power and the control circuits are both isolated.Depending upon the site requirements, sometimesthe control circuit may be required to be still energizedfor some test requirements.The feeder door should not close unless the trolleyis racked-in, up to the test position at least.The trolley should not permit its withdrawal whenits switch is in the ON position.While racking-out the trolley, it should not be possibleto completely withdraw it unless it has reached the‘isolation’ position. In the isolation position theremust be a ‘holding-on’ latch arrangement, to preventan abrupt fall of the trolley from its module. Figure13.36 illustrates this feature.Interchangeability of a module with another moduleof the same type and size is the basic requirement ofsuch panels.Provision of an extra interlocking facility (peg andhole system) may also be essential to preventinterchangeability between two similar trolleys whenthe circuits or the functions of the two otherwiseidentical trolleys are different, and it is undesirableto interchange these trolleys.Withdrawal of a draw-out circuit breakcr or awithdrawable switch or contactor will not be possibleunless they are in an OFF position.Operation of withdrawable equipment, such as abreaker or components on a withdrawable chassis,will not be possible unless it is in service, test. isolatedor totally removed positions.

13/378 Industrial Power Engineering and Applications Handbook13.7.4 Requirements other than constructionalfeaturesMechanical and electrical interlocksAn industrial load having a connected loadrequirement of more than 2000 kVA may normallycall for more than one feeding transformer forlimiting the fault level of the system, as discussedin Section 13.4.1(5). It may also have a standbyemergency source of supply. The two feedingtransformers, although they may be identicalelectrically and suitable for parallel operation,are not supposed to run in parallel with a view tolimiting the fault level. The emergency source,as a result of different electrical parameters, isnot run in parallel with any of the two incomingsources. To achieve the required safety by avoidinga parallel operation, it is essential to provide amechanical or an electrical interlock or bothbetween all the incoming feeders. Schemes toachieve the required safety interlocks are describedin Section 13.7.5.When there are more than one sources of supply,it is recommended to distribute the loads also inas many sections as the incomers, and provide atie-circuit between every two sections, to obtainmore flexibility. Now fault on one section orsource of supply will not result in the loss ofpower to the entire system. Figures 13.16 and13.17 illustrate this type of distribution.transformers must be provided withcurrent limiting fuses at both ends.Control wiring Wiring from supervisory or annunciatordevices to the terminal blocks may be carriedout with smaller wires, as may be recommended forsuch devices. However, they should run throughseparate wire bunches, and not through the bunchesof control wires for easy identification and to remainunaffected by heat of control wires.For easy identification and prompt maintenance itis mandatory to segregate all control wires whenthey are carrying more than one control supply(e.g. at different voltages and both a.c. and d.c.),and run them in separate bunches. The controlwires must also be of different colours for differentcontrol supplies. The colour codes have beenstandardized for different control sumlies (refer toIEC 60445).Space heaters with temperature control Theseare recommended with a view to eliminate condensationof moisture, particularly when the switchgearis idle and the atmosphere is humid. The space heatersare normally rated for Vr/fi, 40 W, single phase.They are located appropriately such as in the cablealley, and are switched ON when the switchgear islikely to be idle for a long period. The number ofspace heaters will depend upon the size and type ofthe switchgear. For a cubicle-type panel, it isrecommended that at least one space heater beprovided in each vertical panel. They should bemounted at the lower portion of the panel for betterheat Circulation through natural heat Convection.or straight through jointsFigure 13.37(a) Pressfit SMC/DMC shrouds for busbar jointsFigure 13.37(b) SMCIDMC shrouds for bus joints and tap-offs

Switchgear and controlgear assemblies 13/379Figure 13.34 illustrates a likely location for such aheater. To prevent condensation of moisture theyare recommended to reach a temperature rise of only5-1 0°C above the ambient temperature, inside thehousing and are controlled automatically through apre-set thermostat.A three-pin socket, rated for Vr18, 5 A, may also beprovided for panel lighting or hand lamp (Figure13.34).Panel numbering on acrylic sheet or aluminiumanodized plates may be fixed on the front and therear of each vertical panel for quick identificationof each panel section (Figure 13.1).For safety to personnel during maintenance and toprotect the live system from lizards and rodents thebusbars may be covered with PVC tape or heatshrinkablePVC sleeve. The joints and the tap-offscan be protected through SMCDMC shrouds, asshown in Figure 13.37. The PVC taping, however,is not recommended for it may suffer cuts or tearoff during working on the busbars and become looseor worn with time. PVC sleeving is a morerecommended practice but sleeving may imposelimitations. It is possible that one may not be able toprovide a true skin-fit sleeve through the length ofthe busbars, which may affect its cooling. At certainplaces, it may have air bubbles from where it willprovide a reduced heat dissipation.For higher rating systems, say 2500 A and above,sleeving is normally not used. Instead, a non-metallic,semi-glossy black paint may be provided to makethe bus conductors act like a black body and dissipatemore heat. This will also add to the current-carryingFigure keys910capacity of the busbar system. Painting is not ameasure of safety but a technique to enhance thecurrent carrying-capacity of the busbar system. Forsafety during maintenance, some other form ofshrouding can be provided, such as by providingSMC/DMC shrouds at all such places where thelive bus may be exposed to the operator attending tomaintenance work.For precautions in making joints, refer to Section 29.2.Painting. A thorough surface treatment of the sheetmetaland a good painted surface are prerequisitesfor equipment to provide long years of operation.For the benefit of those in the field of manufacturingof such assemblies, we have provided a brief procedurefor the sheet treatment and surface painting ofthese assemblies in the Appendix.The above are the more obvious constructional, designand safety features for a switchgear or a controlgearassembly. For more details and additional requirementsrefer to IEC 60439-1 for LT, IEC 60298 and IEC 60694for HT and ANSI-C-37/20C, common for LT and HTswitchgear and controlgear assemblies.13.7.5 Interlocking of feeders to prevent paralleloperationMechanical interlocking schemeUse of castle locksDifferent figure locks such as m, andaare used with a common master key(Figure 13.38). The master key can unlock all locksbut will be locked with them, J-Bl orlock that it unlocks. To remove the key, the lock must belocked first. Then only the key can be used for the otherlocks and thus achieve the required interlocking. Thelock holds the lever of the closing mechanism of theinterrupter and prevents it from closing. The number oflocks will be the same as the number of interruptingdevices, but the keys will be less and will be for only asmany interrupting devices as are permitted to be switchedat a time (generally equal to the number of supply sources).Two sources of supplies (Figure 13.39)The two incomers (IK), fed from two different sources,can be fitted with two locks IA-1 and , andone master key m. This key will allow only oneincomer to be switched at a time.Figure locksFigure 13.38 Castle figure locks and keysTwo sources of supplies and a bus coupler (Figure13.40)The two incomers (IIC) and the coupler can be fittedwith three locks, m, and m, and two

131380 Industrial Power Engineering and Applications Handbook:i2locks :ElI/C - 2( (52ii)P5 locks :pLJD3 keys :IA-lr2I- 1"Figure 13.39 Two sources of supply(52iii)nmc13 locks :IA-lI-BIFigure 13.40 Two sources of supply and a bus couplerkeys 14-1 and m:locks IA-1 and FlUI-B1for the incomers and AB for the coupler. The keyIA] can operate UC-1 or the coupler and keycan operate I/C-2 or the coupler. Thus, only two of thethree interrupting devices, I/C-1 and coupler, I/C-2 andcoupler, or I/C-1 and I/C-2 can be switched at a time toachieve the required interlocking.Three sources of supplies and two bus couplers(Figure 13.41)The three incomers and two couplers can be fitted with fivelocks, m, PI, m, andand three keys VI, I/,The interlocking is achieved as follows:Key 1 A- 1 -can allow switching of I/C-1 or coupler1- can allow switching of I/C-2, coupler1 or coupler 2and 1 C - I can allow switching of I/C-3 or coupler 2Figure 13.41 Three sources of supply and two bus couplersFigures 13.39-1 3.41 are mechanical interlocking schemesThus, only three of thc five interrupting devices can beswitched at a time as required without causing a paralleloperation between any of the two incomers.For a larger number of supply sources, each havingtwo interrupting dcvices, one as incomer and the otheras coupler, two more locks, i.e.and Fiand one key, Le. D may be added for each extrasource and so on.L l lNoteThe mechanical interlocking xheme is generally required formanually operated interrupting devices not fitted with electricallyoperated tripping mechanisms such as an undervoltage (U/V) or ashunt trip (S/T) release. A switch or a switch fuse unit (SFU orFSU) is a device that cannot be provided with an electrically operatedtripping mechanism. Sometimes even manually operated breakersor MCCBs which can be fitted with an U N or S/T are required tohave a mechanical interlocking scheme, although this is not apreferred method when an electrical interlocking scheme is possible.This aspect is discusscd later.When the breaker (including an MCCB) is provided with anelectrical closing mechanism through a motor or a solenoid,mechanical interlocking is not recommended as mechanicalinterlocking will make electrical closing redundant, for obviousreasons.Electrical interlocking scheme (when theinterrupters are manually operated)The preferred way to achieve interlocking between morethan one source of supplies is through electrical schemesand )I. only, wherever possible. They are foolproof and can alsobe operated remotely. Mechanical schemes are generallyfor smaller installations where, as a result of smallerratings or cost considerations, a breaker is not used andthat imposes a limitation on adopting an electricalinterlocking scheme.The electrical interlocking should preferably beprovided through shunt trip releases. It must have aseparate a.c. or d.c. source of control supply, such thatthe operation of the scheme ib independent of the mainsource of supply. For the same reason, interlocking throughundervoltage (UN) releases is not recommended as its

Switchgear and controlgear assemblies 13/381operation would be dependent on the condition of supply.An U N release is generally fitted to trip the interruptingdevice when the supply ceases and not for any othercontrol schemes.The schemes are logical and simple and have beendrawn, to ensure that no two supply sources can ever beswitched in parallel. The scheme prevents to switch aninterrupter, that may cause it to operate in parallel withanother, unless the first source is opened first. This isillustrated in the following schemes;Two sources of supplies (Figure 13.42)The NO (normally open) contact of I/C-1 (52i) is wiredin the trip circuit of I/C-2 (52 ii) and vice versa. As soonas an interrupter is closed, the tripping circuit of theother gets ready to trip.NoteThe function of a shunt trip coil is to trip an interrupter. As soon asits coil is energized, it releases the closing lever of the interrupterand trips it. The coil is rated for a short time, since it is in the circuitfor a very short period only when it is required to trip the interrupter.To ensure that it does not continue to remain energized after carryingout its function, the interrupter’s ‘NO’ contacts are also wired inseries with the coil as shown. The coil becomes de-energized assoon as the interrupter trips. Normally two NO contacts of theinterrupter are wired in series, as standard practice, by the interrupter(breaker or MCCB) manufacturers to share the arc energy on a tripand enhance contact life, particularly when the control supply isd.c. which is normally the case.Two sources of supplies and a bus coupler(Figure 13.43)The logic is the same as above. In the trip circuit of eachinterrupter is wired the NO contacts of the other twointerrupters. Obviously only two of the three interrupterscan be switched at a time.Three sources of supplies and two bus couplers(Figure 13.44)The scheme is now more complicated but the followinglogical approach will make it simple:(a) When YC-1 (52i) is required to be switched(i) Interlocking with I/C-2 (52ii): I/C-2 (52ii) andB/C- 1 (52 iv) should not be in a closed positionat one time.(ii) Interlocking with VC-3 (52iii): I/C-3 (52iii),B/C-1 (52iv) and B/C-2 (52v) should not allbe in a closed position at one time. Refer tothe control scheme for I/C-1 (52i)., I/C - 2(52ii). ---..(52i)*A, I/C - 1IIC - 2 l/;G)l f I/C - 1(52ii)(52i)1BIC(52iii)(52iii)B/c(52i)*I1(52i)*(52i)*III1 1 II/C - 1 I/C - 2(52i)(52ii)Legends detailsSw - 1 - Control supply ON/OFF switchF1 - F2 - Control fusesST,, ST2 - Shunt trip coils of breakersI/C - 1, 2 or 52(i, ii) - Incoming sources of supplies*Note Normally 2NOs of the interrupters are wired in seriesto share the arc energy and enhance the contact life.Figure 13.42 Electrical interlocking scheme for manuallyoperated breakers for two sources of suppliesL___--IIC - 1 BIC I/C - 2(52i) (52iii) (52ii)Legends detailsSw - 1 - Control supply ONIOFF switchF1 - F2 - Control fusesST1, ST, ST3 - Shunt trip coils of breakersI/C - 1, 2 or 52(i, ii) - Incoming sources of suppliesBIC - 1 or 52(iii) - Bus coupler*Note Normally 2NOs of the interrupters are wired in seriesto share the arc energy and enhance the contact life.Figure 13.43 Electrical interlocking scheme for manually operatedbreakers for two sources of supplies and a bus coupler

13/382 Industrial Power Engineering and Applications HandbookSw-1F1IIIIIIIIIIIII-% 11Qa 1'eLC I'I18 IIB/C - 2 I/C-3 B/C - 22 II I IIIIIIIII I I I I I III I I I I I IIIIII I I I I IIIF2Legends detailsSw - 1 - Control supply ON/OFF switchF, - F, - Control fusesST,, ST?, ST,, ST4 and ST5 - Shunt trip coils of breakersl/C-l,2,3 or 52(i,ii,iii) - Incoming sources of suppliesB/C-l,2 or 52(iv,v) - Bus couplers.I/C- 1 BE-1 I/C-2 B/C - 2 I/C-3(52i) (52iv) (52ii) (534 (52iii)*Note Normally 2NOs of the interrupters are wired in series to share the arc energy and enhance the contact life.Figure 13.44 Electrical interlocking scheme for manually operated breakers for three sources of supplies and two bus coupler.(b) When YC-2 (52ii) is required to be switched(i) Interlocking with I/C-1 (52i): I/C-1 (52i) andB/C-1 (52iv) should not both be in a closedposition at one time.(ii) Interlocking with YC-3 (52iii): VC-3 (52iii)and B/C-2 (52v) should not both be in a closedposition at one time. Refer to the control schemefor I/C-2 (52ii).(c) When YC-3 (52iii) is required to be switched(i) Interlocking with YC-2 (52ii): VC-2 (52ii) andB/C-2 (52v), should not both be in a closedposition at one time.(ii) Interlocking with I/C-1 (52i): YC-1 (52i), B/C-1 (52iv) and B/C-2 (52v) should not all be ina closed position at one time. Refer to the controlscheme for YC-3 (52iii).(d) When B/C-1 (52iv) is required to be switched(i) Interlocking with VC-1 (52i) and I/C-2 (52ii):These should not be in a closed position at onetime.(ii) Interlocking with I/C-1 (52i) and YC-3 (52iii):YC-1 (52i), B/C-2 (52v) and VC-3 (52iii) shouldnot all be in a closed position at one time. Referto the control scheme for B/C-1 (52iv).(e) When B/C-2 is required to be switched(i) Interlocking with VC-2 (52ii) and YC-3 (52iii):Both should not be in a closed position at onetime.(ii) Interlocking with YC-1 (52i) and YC-3 (52iii):YC-1 (52i), B/C-I(52iv) and VC-3 (52iii) shouldnot all be in a closed position at a time. Referto the control scheme for B/C-2 (52v).Electrical interlocking scheme when theinterrupters are also electrically operatedMotor-operated interrupting devices are employed whenthe system requires remote-controlled power switching,as for an auto-reclosing scheme. The electrical interlockingschemes remain generally the same, as discussed earlier,but with an additional circuit for the motor spring chargingmechanism and the closing coil of the interrupter. Briefdetails of the electrical closing features are as follows.Spring charging motor mechanismThe purpose of the motor is to charge the closing springthat closes the interrupter, independently of the speedand operation of the motor or of the operator when closedmanually. (The interrupter can be closed manually or

fSwitchgear and controlgear assemblies 13/383electrically when the spring is fully charged.) As soon asthe spring is discharged, the motor recharges it automaticallyto prepare it for the next operation. The motormay be fed through the same source as for the maincontrol scheme or another source, depending upon thesystem design. But the source must be reliable andindependent of the main power supply as far as possible(such as through a battery).NoteSmall interrupters and MCCBs particularly may be electricallyoperated through a solenoid valve.Limit switch (LS)The spring charging mechanism is fitted with a limitswitch, having generally 2N0 and 2NC change-overcontacts. The 2NC contacts are wired in series with themotor to allow it to charge the spring mechanismimmediately on discharge of the spring, to prepare it forthe next operation. As soon as the spring is fully charged,the NC contacts change over to NO and cut off the supplyto the motor terminals (Figure 13.45).The 2N0 contacts of the limit switch are wired inseries with the closing coil of the interrupter (Figure13.45). As soon as the spring is fully charged these contactschange over to NC and the closing coil circuit gets readyto close.Closing coil (CC)The charged closing spring may be released manually orelectrically by energizing a closing coil as shown in Figure13.45.Breaker control switch (CS)To close or trip the interrupter locally or remotely abreaker control switch is also wired with the closing andthe shunt trip coils of the interrupter, as shown in Figure13.45. The switch is a spring return type to ensure that itresumes its original (neutral) position as soon as it hascarried out its job of closing or opening the interrupter.With the above features in mind, the control logic for thevarious interlocking schemes becomes simple as shownin Figures 13.45-13.47:Two sources of supplies: Figure 13.45.Two sources of supplies and a bus coupler: Figure13.46.Three sources of supplies and two bus couplers: Figure13.47.13.8 HT switchgear assembliesUse of an LT switchgear assembly is more frequent thansw-1!f-+ T7'I I IContactI/C-2(52ii)I/C-2(52ii)I"{ ...........-.............h N ...............IK-I(52i) 1 -BContact arrangement of CSLE?nd details:-1IIIIII (52i) IIII1 I cc-I jI--_--_III__-_ST-1 ~t----CC-I ,2, ----- - Closing coils ofinterruptersCS-1,2, ------ Breaker controlswitches (springreturn to neutral)M1,2, ----- - Spring chargingmotorsLS - Limit switches withspring chargingmechanismVC-I(52i)I/C-2(52ii)Figure 13.45 Control scheme for two electrically operated sources of supplies

13/384 Industrial Power Engineering and Applications HandbookIIC-I(52i)(52iii)II(52ii) IIII(52ii) III. .IIC-I BIC IIC-2(52i) (52iii) (52ii)Figure 13.46 Control scheme for two electrically operated sources of supplies and a bus couplerthat of an HT switchgear assembly. Moreover, HTassemblies above 33 kV are generally the outdoor typeand installed in the open. The discussions above havetherefore laid greater emphasis on the features of an LTswitchgear assembly. The features for an HT assemblyup to 33 kV are not very different except for thickerenclosure and heavier load-bearing members, to carrylarger HT equipment (interrupting devices, CTs, PTs,insulators etc.). The clearances and creepage distanceswill also be greater, according to the system voltage(Section 28.5.2). For more details refer to IEC 60298,IEC 60694 and ANSI-C-37/20C.13.9 General guidelines duringinstallation and maintenance ofa switchgear or a controlgearassembly1 Installation and fixing The assembly should beinstalled in a room that is well ventilated (except foroutdoor assemblies), on a rigid concrete or steelfoundation to eliminate vibrations. A simple and morecommon practice of fastening on a separate steelfoundation frame is illustrated in Figure 13.48. Theassembly may also be fixed on a concrete foundationdirectly, but this will be found to be more cumbersomeand time-consuming.2 While putting into serviceOne may follow field tests as discussed in Section14.5 before putting the switchgear or the controlgearassembly into service.If the insulation resistance is observed to be lowthe interior of the switchgear or the controlgearassembly should be dried out to attain the requiredvalue before energizing it. The procedure to drythe moisture is similar to that for motors and isdiscussed in Section 9.5.2. In this case during theheating-up period the insulation resistance may firstdrop and reach its minimum, stabilize at that leveland then start to rise gradually. Continue the processuntil it reaches its required level as shown in Table14.7.In HT switchgear assemblies provide a surge arresteror a lightning arrester, wherever necessary. Alsorefer to Section 17.11.It is advisable that all the protective devices providedin the assembly be set to their minimum values toensure fast tripping. They can be adjusted for propersettings when putting the assembly into service.3 Maintenance of busbars and busbar connectionsAll the joints, particularly power joints at the mainbusbars, tap-offs and cable ends, should be checked

Switchgear and controlgear assemblies 1313851(Contd.)BIC- 1 IIC- 2(52iv) (52ii)BIC - 2IIC- 3(52iii)IIC-(52i)IIC- ;(52ii)IIC- 1 BlC- 1IIC - BIC- 2I A, LS !I(52ii)IIld- 1(52i)BIG-1(52iv)L"IIC - 2(52ii)B(52ii), IIC-2(52ii)IIC-I BIC - 1IIC-3(52iii)BIC-2(534- * "IIC-2 BIC - 2 IIC-3(continued)(52ii)Figure 13.47 Control scheme for three electrically operated sources of supplies and two bus couplers

13/386 Industrial Power Engineering and Applications Handbookperiodically for any pitting or loosening. Such acheck is recommended at every six months ofoperation and must be carried out meticulously,particularly for aluminium busbar joints and cableterminations. Aluminium is a highly malleable andductile metal and under high temperatures andpressures has a tendency to run out and loosen itsgrip. At locations that are critical, contaminated orhumid, or that are subject to vibrations, the periodof maintenance checks may be reduced based onexperience. A logbook can also be maintained tomonitor the variance in important parameters andto take preventive measures during operation.Draw-out components- In a draw-out MCC all current-carrying componentsshould be periodically checked for theirsilver-plating, proper contact area, spring pressureand tightness of joints (this procedure can followthe manufacturer’s maintenance schedule) or atleast during the six-monthly or yearly maintenancecheck ups of the busbars and the busbar joints asnoted above. The contacts or their springs maybe replaced when the contacts are worn out orthe silver-plating is withered. Silver preventsoxidation of copper contacts and also eliminatesbimetallic corrosion.NoteTarnishing (blackening) of contacts is a common characteristic ofsilver-plated contacts. It is the formation of silver oxide filmwhich is a good conductor of heat and electricity is not a cause forconcern.- Clean and lubricate all the main incoming andoutgoing power contacts as well as the auxiliarysliding contacts at least once a year. Use of neutralgrease as noted in Section 13.6.l(iv) is recommended.A properly greased contact will also helpto avoid a flashover.- For cleaning of contacts, use white petrol or carbontetrachloride or perchloro-ethylene. Never usesandpaper for it may damage the silver-coatingand also render the surface uneven which maycause arcing and pitting during operation. Whilecleaning, care must be taken that the solvent doesnot reach the insulating components.Bottom frame ofthe PCC or MCCFoundation orgrouting frame(i) Panel fixing by welding(ii) Panel fixing by boltingDetails of fastening at ‘D,’ (typical)ADetails of fastening at ‘D2’ (typical)Switchgearassembly -(rear)---/----Switchgear- assembly(front)Foundation frameof the panel \Figure 13.48(a) Illustration of a typical installation of high-rating switchgear assembly on a cable trench

Switchgear and controlgear assemblies 13/387and quick-responding PLC and microproces4or-basedcontrols, as discussed in Section 13.2. The logistics forPLCs or microprocessors are also the 5ame as forelectromagnetic controls.Switchgearassembly ----(rear)Cable trenchSwitchgearassembly(front)Figure 13.48(b) A typical installation of a low-rating switchgearassembly on a cable trench- The cranking screw and the guide rails on whichthc trolley slides must also be coated with ordinarygrease to provide a smoother operation and toprevent corrosion.- Check the grounding contacts periodically fortheir positive ground connections.For the sake of brevity, this subject is not dealt in withgreat detail here. Refer to IEC 60694 and BS 6423.4 Discharging of a power capacitor Whenever powercapacitors are installed in a switchgear assembly andare not discharged automatically on a switch OFF,through its own interrupting device, these must bedischarged manually by grounding its terminals heforeits feeder devices and components are physicallyhandled13.10 Power circuits and controlscheme diagramsFor a ready reference to the readers, we provide powerand control scheme diagrams, usually required in dayto-dayuse, while wiring an MCC or a control panel, ormaintenance at site.All the control circuits are based on conventionalelectrical and electromagnetic (auxiliary contactors andtimers) controls. The latest trend for large or complicatedcontrols. however. is to have more compact, accurate13.10.1 Interlocking and control scheme for atypical air-conditioning plantFor the application of individual scheme\, as illustratedabove and an easy understanding of these schcmes. weconsider below a conventional type of air-conditioningplant for its various controls. interlocks and operatingrequirements.This type of air-conditioning plant may have thefollowing three closed circuits:I Refrigerant circuit: Figure 13.49(aj is a flow diagramfor the refrigerants. The refrigerant used prevxtly ischlorotloro carbon (CFC- 1 1 ~ 12, I 13. 1 I4 or 1 IS) butuse of this gas is being gradually discontinued (by2000 latest 2005) as this causes ozone layer depletionand global warming. It shall be gradually replaced byhydrochloro fluro carbons (HC FC-22, 123. 141 and142). But this too is not totally environment friendlyand shall be replaced by 2040 (latest) by hydro fluorocarbons (HFC-I 34a) which will be more safe. However,research is on to invent yet better blend\ ofrefrigerantswhich may be quite environment friendly. [For moreinformation on refrigerants refer to UNEP IE/PAC(United Nations Environment Programme. Industryand Environment Programme Centre. USA].2 Condenser water circuit: Figure 13.49(bj is a flowdiagram for the condcnser water.3 Chilled water circuit: Figure I3.49(c) is a flow diagramfor the chilled water.The power circuit single-line diagram is shown in Figure13.50. The following are the control\ and protectionsthat may be generally required for wch a plant.CompressorControl of the compressor is achieved by engaging therequired number of cylinders. In. say. a 16-cylindercompressor if we engage only four cylinders;. thecompressor will run at 25% capacity. and if we engageeight cylinders, the compressor will run at 50%. capacity.Electrically operated solenoid valves are pro\;icled forcapacity control. Energy can be conserved by using staticcontrols, as discussed in Section 6.15.For protection and temperature control of thecompressor the following safety devices may be provided:Water flow switchesHigh/low pressure cut-outsLube oil pressure switch andSafety thermostatAir-handling unit (AHU)To control the room temperature flow of chikd water iscontrolled in AHU, Figure 13.49(c). A thermostat sensesthe room condition and activates a motorised wlenoidvalve in AHU coil which in turn ad,justs the flow of

13/388 Industrial Power Engineering and Applications HandbookWater 12-14"CCondenserliquidCompressedhot-gasCondenserChilledwaterPumpCompressor 11Super heated gasCold condenserwaterChillerExpansionvalveCondenserI -.Air 21-24°CwaterFigure 13.49(a) Refrigerant circuit Figure 13.49(b) Condenser water Figure 13.49(c) Chilled water circuit(typical) circuit (typical) (typical)Air handlingunitFigure 13.50 Single line power diagram for a typical air conditioning plantchilled water to the required level. There are other methodsalso in practice such as by pass of certain amount of airthrough coil and variable speed fan drive (for energyconservation) (section 6.15).Humidity controlTo control the humidity in a conditioned space humidifiersreheaters are provided Humidifiers to increaseand reheaters to decrease the humidity.Electrical interlocks between the drivesThe following interlocks are generally provided betweenthe various drives:The condenser water pump will not start unless thecooling tower fan is already running.The chilled water pump will not start unless thecondenser water pump is also running.The compressor will not start unless the chilled waterand condenser water pumps are also running.The control and protection scheme diagram for all theabove requirements is given in Figure 13.5 1. It is presumedthat all the drives are provided with direct on-lineswitching.13.10.2 Different types of starters andinstruments wiringTable 13.16 provides the list of symbols and abbreviationsused.Figure 13.52 - Simple DOL starter

T'DIcizwater flowswitchQ --4*---Cooling waterflow switch---e,-*---HPswitchu: --4&--QLPswitch3 --4*---Oil pressureswitch---e,*---i4u s5u0-i7104'-0- N-dCrankcaseheater(COMP-1)mvContd. next pageIthermostatN

13/390 industrial Power Engineering and Applications HandbookFigure 13.51 Contd.N:llll_0 - 3 sw 0-q sw 0-qD~COMP(Contd. below)COMPhlvvCapacity controlsolenoid valvesRefrigerationsolenoid valve-nVCompressor motor (COMP)sw)line LiquidStart(local)- CHPDXHF!h2CHPA c cOn Off CHWP CHWJ\v /V\On OffLiquid lineVsolenoid valveHeater feeder (HF)Humidifier pump (HP)Figure 13.51(Contd.)

Switchgear and controlgear assemblies 13/391Table 13.16SymbolList of symbols and abbreviations usedAbbreviation Description Symbol Abbreviation DescriptionRYBN3@ and neutral power supplyTHThermostatswHeavy-duty switchD/ODrawout terminalFPower fuseTerminalCFControl fuseo/cThermal overcurrent relayFSUFuse switch unithIndicating light(colour of lenses, R-red, G-green,A-amber, W-white)MCBlcoupledcircuit breakerAAmmeterASSAmmeter selector switchIsolatorCCdPower contactor(M - main, D - delta,S - star)Coil of main contactorCoil of auxiliary contactorVvssVoltmeterVoltmeter selector switchTCoil of timerCTCurrent transformerNormally open auxiliary contactVoltage or control transformerNormally closed auxiliary contactTrPower TransformerRResistanceOCROvercurrent relayLReactanceTDCTDOTime delay contactscsCurrent shunts/swStart PBStop PBSelector switch(L/R - locallremote, AIM- auto1manual)Breaker control switch (springreturn to neutral)(C - closed, N - neutral, T - trip)Limit switch with spring chargingmechanismDoor switchPushbutton (colour of knob,R - red, G - green, Y - yellow)0IISHMGGSpace heaterSurge arresterMotorGeneratorCableGrounding

13/392 Industrial Power Engineering and Applications HandbookR Y B NswmIrLLCmIr+Figure 13.52 Simple DOL starterO K On OfftripFigure 13.54 DOL starter with CT-operated over-current relay(i) For a high-rating motor or(ii) For a heavy-duty motor (irrespective of rating)O V D h lRunninginterlockICI.LItripFigure 13.53 DOL starter with provision for remote controland starting and/or running interlocks3Figure 13.53 - DOL starter with provision for remotecontrol and starting and/or runninginterlocksFigure 13.54 - DOL starter with CT-operatedovercurrent relay(i) For a high-rating motor or(ii) For a heavy-duty motor (irrespectiveof rating)Figure 13.55 - Reversing DOL starterFigure 13.56 - Star-delta starterFigure 13.57 - Auto transformer starterFigure 13.58 - Primary resistance starterFigure 13.59 - Dual-speed starterFigure 13.60 - Three stage stator-rotor starterFigure 13.61 - Schematic for panel space heater andinternal illuminationFigure 13.62 - Schematic for instrument wiringNoteAll control schemes shown with auxiliary contactors and timerscan be easily replaced with PLCs and microprocessor-based controls.Refer to Section 13.2.3 for more details.

Switchgear and controlgear assemblies 131393R Y B N7-c12tC122h4ItriD on onfFigure 13.55Reversing DOL starterR Y E Ni.Db DS0 0MM0triphl!h2h3-Motor windings in AFigure 13.56Star-delta starter

13/394 Industrial Power Engineering and Applications HandbookCFNL 1OCRb'\-

Switchgear and controlgear assemblies 13/395CFa a blb c c g1 1 -Lc1r)hllI~> cIC On 1trip trip slow fast(slow) (fast)c2c11h3-if8 pole (A) 4 pole double star(’f ‘f)(slow)(fast)Figure 13.59 Dual-speed startel

13/396 Industrial Power Engineering and Applications HandbookR Y B Nd3d:CCOCR !;.cOCRinter'ock p'Cd2 d:/d314 15 16 I2__ -;ontrol Res. Res. Motor MotorOn In Cut-off Off On11-OICtRlPI* Contactors connectedin A to reduce their size.Figure 13.60 Typical three-stage stator-rotor starter

Switchgear and controlgear assemblies 13/397R Y B NF31Under voltagerelavra22ovac&e2T1bl -Thermostat cut off contactS15 - Space heater switchS16 - Panel door switchell - Space heatere2 - Interior illuminationFigure 13.61illuminationSchematic for panel space heater and internalTo alarmcircuitTo tripcircuit1 1 1 1Figure 13.62 Schematic for instruments’ wiring

13/398 Industrial Power Engineering and Applications HandbookRelevant StandardsIEC Title IS BS IS060034-11199660038/19946005 1 - 111 9976005911 9996007 1 - 11199360071-2/199660076- 11199360079- 1/199060079- l4/199660112/197960044- M99660186/19876025560258/196860265- M99860265-2/198860282- 11199860298/199460439- 11199660439-2/199160439-3/199360439-4/199560445/1988605 17/1990605 291 1989606 17-2/1996606941199660947- 111 99960947-3/199860947-4- 1119966090911988-Rotating electrical machinesIEC standard voltages, rating and performanceDirect acting indicating analogue electrical measuringinstruments and their accessoriesDefinitions and general requirementsStandard current ratings (based on Renald SeriesR-10 of IS0 3)Insulation coordination. Definitions, principles and rulesInsulation coordinationApplication guidePower transformers - generalConstruction and verification test of flameproofenclosures of electrical apparatusElectrical installations in hazardous areas (other thanmines)Method of test for determining the comparative and theproof tracking indices of solid insulating materialsunder moist conditionsCurrent transformers, specification and applicationVoltage transformers, specification and applicationElectrical relaysDirect acting recording electrical measuringIinstrument and their accessoriesSwitches and switch isolators for voltages abovelO0OVGeneral and definitionsSpecification for high voltage switches. High voltageswitches for rated voltages of 52 kV and aboveHigh voltage fuses. Current limiting fuses exceeding1 kVA.C. metal enclosed switchgear and controlgear forrated voltages above 1kV and up to and including 52 kVSpecification for LV switchgear and controlgearassemblies. Specification for type-tested and partiallytype-tested assembliesParticular requirements for busbar trunking systemsParticular requirements for low-voltage switchgear andcontrolgear assemblies intended to be installed inplaces where unskilled persons have access to theiruse. Distribution boardsLow voltage switchgear and controlgear assemblies.Particular requirements for assemblies for constructionsitesIdentification for equipment terminals and ofterminations of certain designated conductors,including general rules for an alphanumeric systemGas insulated metal-enclosed switchgear for ratedvoltages of 72.5 kV and aboveDegree of protection provided by enclosure (I.P. code)Graphical symbols for diagrams - Symbol elements,qualifying symbols and other symbols having generalapplicationCommon specifications for high voltage switchgear andcontrolgear standardsLow voltage switchgear and controlgear. General rulesand test requirementsLow voltage switchgear and controlgear. Switches, disconnectors,switch-disconnectors and fuse-combinationunitsLow voltage switchgear and controlgear. Contactors forvoltages not exceeding 1000 V a.c. or 1200 V d.c.Guide for short-circuit calculations in a 3-44 systemCode of practice for selection, installation andmaintenance of switchgear and controlgear up to 1 kV.General4722/1992/325/19961248-1 to 9-2165-1119912165-2/19912026- 1/19912 148/19935571/19912705-1 to 4/19923156-1 to 4/19923231-1 to 3,3842-1 to 12-9920-1 to 4/19929385-1 to 53427/199 18623-1/19938623-2/19938623-3/199311353/19912 147/1962120323427/199 113947- 11199313947-3/199313947-4- 11199313234101 18-1/1991BSEN 6003441995BS 89-1/1990BSEN 60071-11996BSEN 60071-11997BSEN 6007641997BS 4683-2/1993550 1 -5/1997BSEN 60079-14/1997BS 5901/1980BS 7626/1993BS 7729/1994BS 7625/1993BS EN 60255BS 90/1993BSEN 602651/1998BSEN 60265-/I994BSEN 60282/1996BSEN 60298-1/1996BSEN 60439-1/1994BSEN 60439-2/1993BSEN 60439-3/1991BSEN 60439-4/1991BS 5559/1991BSEN 6051711997BSEN 60529/1992BSEN 60617-2/1996BSEN 60694/1997BSEN 60947- 1/1998BSEN 60947-3/1992BSEN 60947 - 4-111992BS 7639/1993-

Switchgear and controlgear assemblies 13/399IEC Title IS BS IS0SelectionInstallationMaintenanceSpecification for control transformers for switchgearand controlgear for voltages not exceeding lOOOV a.c.Specification for high voltage busbars and busbarconnectionsMechanical properties of fasteners.10118-2/1991101 18-3/199 1101 18-4/19911202 1 - 19878084/19921367-2011 996-BS 6423/1993BS 15911992BSEN 20898-711995Torsional test and minimum torques for bolts andscrews with nominal diameters 1 mm to 10 mmSpecification for spring washers for generalengineering and automobile purposes. Metric seriesCode of practice for design of high-voltage openterminalstationsSpecification for electrical cable soldering socketsCode of practice for maintenance of electricalswitchgear and controlgear for voltages above 1 kVand up to 36 kVCode of practice for maintenance of electricalswitchgear for voltages above 36 kVSpecification for voltage regulation and paralleloperation of a.c. synchronous generatorsMethods of testing plastics3063/1994-655411991-Relevant US Standards ANSIA'EMA and IEEEBS 4464/1990BS 735411990BS 91/1988BS 6626/1993BS 686711993BS 4999-140/1987BS 2782ANSVIEEE-C37.2O.U 1993ANSYIEEE-C37.20.2/1994ANSI/IEEE-C37.20.3/1993ANSI/IEEE-C-37.23/1992ANSI C.37.47/1992NEMA L1111983ANSVIEEE 24111991ANSIlIEEE 2421199 1ANSYIEEE141/1993ANSUIEEE 131211993ANSI/IEEE 13 13.1 /I 996ANSUIEEE C37.1611997NEMA/ICS-2/1993NEMAIICS-2.3/1990NEMA/ICS- 111993NEMA/ICS-3/1993NEMA/ICS-6/ 1996ANSYIEEE C37.21/1998ANSI/C84.1/1995NEMA WC-57/1990NEMA SG-6NEMAPB 1/1990NEMA/PB 1.1/1991NEM A/PB2/1995Metal enclosed low voltage power circuit breaker switchgearMetal clad and station type cubicle switchgearMetal enclosed interrupter switchgearGuide for calculating losses in isolated phase busSpecifications for distribution fuse disconnecting switches, fuse supports and current limiting fusesIndustrial laminated thermo setting productsRecommended practice for electric power systems in commercial buildings (IEEE Grey Book)Recommended practice for protection and coordination of industrial and commercial power systems(IEEE Buff Book)Recommended practice for electric power distribution for industrial plants (IEEE Red Book)Voltage ratings for a.c. electrical systems and equipment above 230 kVInsulation coordination, definitions, principles and rulesPreferred ratings, related requirements and application recommendations for LV power circuit breakers anda.c. power circuit protectorsIndustrial control and systems, controllers, contactors and overload relays, rated not more than 2000 V ax. or750 V d.c.Instructions for the handling, installation, operation and maintenance of MCCsIndustrial controls and systems. General requirementsIndustrial control and systems. Factory built assembliesIndustrial control and systems. EnclosuresStandard for control switchboardsElectric power systems and equipment - Voltage ratings (60 Hz )Standard for control cablesPower switching equipmentPanel boardsGeneral instructions for proper installation, operation and maintenance of panel boards, rated 600 V or lessDead front distribution switchboardsNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

13/400 Industrial Power Engineering and Applications HandbookAppendix: Painting procedure ofswitchgear and controlgearassemblies and treatment ofeffluentA13.1 IntroductionPainting of all metallic surfaces of a switchgear or acontrolgear assembly is an essential requirement to provideit with an aesthetic appearance, on the one hand, and toprevent it from rust and corrosion, on the other. Paintingserves these purposes by providing the machine with ahard and longer-lasting metallic surface. We describebriefly, the basic procedure to paint and test paintedsurfaces. In the discussion, we have laid more emphasison MS sheet-metal surfaces as these are more typical.The painting procedure for other metal surfaces,although similar, the process of pre-treatment for castiron components or non-ferrous metals, such as aluminiumand copper, may need more care. The process of pretreatmentin such cases may vary slightly than for MS,as noted below. Such surfaces may require a change inthe type of chemicals, their concentration and durationof treatment. The final surface preparation and paintingprocedure, however, will remain the same for all.The total painting procedure, may be divided into therollowing operations:1 Sheet pre-treatment (phosphate coating)2 Preparing the surface3 Applying the final paint4 Curing the paint5 Testing the painted surfaces6 Providing a peelable coating compound, if necessaryand7 Effluent treatment and discharge of waste water.A13.2 Sheet pre-treatment (phosphate coating)In Table A13.1 we describe the most common practicesbeing adopted to pre-treat and phosphate ferrous andnon-ferrous surfaces, before applying paint:Degreasing and cleaningDegreasing is a process to remove oil, grease, dirt andswarf (file dust) ctc. from a surface.Types of cleaners (degreasing agents)There are several types of cleaners available for thispurpose, for example:1 Alkaline cleaners (caustic based) These are causticsoda based and are suitable for ferrous metals only.They are more effective in removing greases ofvegetable oils, rather than mineral (petroleum) oils,as they do not saponify the mineral oils.2 Neutral cleaners (non-caustic based) These areethylene oxide condensates, and easily emulsify themineral oils and greases. They are more useful forsheet-metal components, which contain no leadcompound lubricants (as used for deep-drawingoperations), and are also suitable for non-ferrousTable Ai3.1Process of sheet pre-treatment (phosphate coating)hor ferrous metalsPre-treatment processHeavily scaled andheavily rusted surfaces(hot-rolled sheets)Heavily scaled, butmildly rusted-surfaces(hot-rolled sheets)Mildly scaled andmildly rusted sufaces(cold-rolled sheets)For non-ferrousmetals1 Degreasing and cleaning1 Water rinsing3 Descaling or acid pickling4 Water rinsing5 De-rusting"JJJJJJJJJ-JJ6 Water rinsing7 Zinc phosphatingb8 Water rinsing9. PassivationhJ~JJJSecond water rinsing orneutralizing isrecommended, inHCI picklingJJJ.~ ~J"~~JJJJJJNo. of tanksNine-tank methodEight-tank methodSeven-tank methodFive-tank methodahstead of de-rusting, a pickling process may also he sufficient, depending upon the surface condition of the sheets.bAftcr the pickling process, if the phosphating bath contains traces of sulphate (SO,) or chloride (CI) salts, the phosphated surface maybecome highly hygroscopic, and may absorb atmospheric moisture through even a very well painted surface, and show rusting with passageof time (depending upon the atmospheric conditions at the place of installation). To avoid this and to achieve a long life for all paintedsurfaces, it is recommended that all these salts are first neutralized. This is possible with the use of de-mineralized (DM) water, at least forthe make-up baths of phosphating and passivation. Where water is heavy and contains mineral salts, a small DM unit can he installed. Itis an inexpensive procedure for the quality of phosphate coating that would be achieved. This will also enhance the working life of thesebath solutions, and economize on their consumption. De-mineralization means removal of all sulphate (SO,) and chloride (Cl) salts.Note: The transfer of jobs from one tank to another may he done by an overhead travelling hoist to handle bulky and heavy objects.

Switchgear and controlgear assemblies 13/401metals. These cleaners emulsify faster than other typesand are more suitable for heavy production lines.Emulsion cleaners These are emulsified chlorinatedsolvents and are kerosene based, suitable for mineraloils (petroleum and heavy petroleum greases) anddeep-drawn components, using lead compounds aslubricants. They are also suitable for non-ferrousmetals.Solvent cleaners These are tri-chloro ethylene (TCE)and are highly evaporating cleaners, possessing toxicproperties. Their application is the same as for neutralcleaners.Concentration, bath temperature and dipping timeThese are indicated in Table A13.2.Tank materialA mild-steel (MS) tank with wall thickness of 4-6 mm,with a heating arrangement and a thermostat temperaturecontrol, will be ideal for this purpose. Even when a coldprocess is adopted, heating will be imperative duringwinter or in cold climates, where the bath temperaturewould be less than 40°C. A protective lining inside thetank is generally not essential, as the chemical does notattack the metal. An anti-corrosive paint, such as epoxyor polyurethane may, however, be provided to enhancethe life of the tank. The thermostat and heater may be ofstainless steel or ordinary water heaters could be used.The tank may be made slanting ( *inch in an 8-feet lengthis adequate) to have an overflow system to remove theoil and grease scum during the degreasing process.Checking the bath concentrationThe concentration of bath solution in caustic-basedcleaners can be checked by a simple titration method asnoted below, while for the remaining types of cleaners,a visual check of the degreased surfaces will be sufficient.The titration method is as follows:e Pipette out 10 ml of bath solution and add 90 ml ofdistilled water (total 100 ml).Table A13.2Concentration, bath temperature and dipping timeType of Concentration Bath temperature Dipping timecleanerAlkaline 3-5% by weight 90-95°C 10-15 mincleaners or volume of bathNeutral 3-5% by (i) 50-70°C 3-5 mincleaners weight or (ii) Also possible 10-15 minvolume of bath at room temp.(4045°C)Emulsion No dilution 5040°C 3-5 mincleanersSolvent No dilution 80°C 3-5 mincleanersNote The exact values will depend upon the surface condition ofthe material and the field experience and skill of the operator. Formore accurate details consult the chemical manufacturer.Remove 10 ml of this solution and add a few (three tofive) drops of phenol phthalein and shake.Titrate this against N/lO-HCl solution until the colourchanges to pink.A burette reading will indicate the actual pointage ofthe bath compared to the standard (approximately3-6) for a concentration of 3-5%. Obtain thestandard pointage from the chemical manufacturer.For each one point of lower strength, add 1 litre of chemicalper 100 litre of bath.NoteN/lO-ready-made laboratory chemicals are easily available.Precautions to be observed1 When making a fresh bath, stir the chemical well ina separate container, preferably in hot water, beforemixing it into the tank.2 Skim off oil and grease scum from the surface of thebath every day or more frequently, depending uponthe amount of work being handled.3 When using alkali base cleaners, protect the eyes andskin from direct contact with the chemical.Water rinsingTo rinse, wash the surface in clear, continuous runningwater to remove all traces of degreasing agent from thesurface. Tko/three dips at room temperature are sufficient.An MS tank with a wall thickness of 3-4 mm isadequate, otherwise the thickness can be similar to thatfor the degreasing tank but without the heatingarrangement. It may also be coated with an anti-corrosivepaint to enhance its life.Carry out the rinsing before the chemical dries on thesurface.Descaling or acid picklingThis is a process to remove heavy black scale and rustfrom the surface. Hot-rolled sheets that may have suchscale formation need only be acid pickled. Cold-rolledsheets, which may carry no such scales, need not be acidpickled. Depending upon the type of surface, one of thefollowing methods may be adopted.For heavily scaled and heavily rusted surj5acesAcid pickling This can be done under the followingoperating conditions, either with sulphuric acid(H2S04), or hydrochloric acid (HC1). H2S04 releasesa lot of fumes and is ineffective under cold conditions.It forms iron sulphate, which forms a hard deposit atthe bottom of the tank and is difficult to remove (seetable on next page).Tank material A mild-steel (MS) tank, with a wallthickness of 4-6 mm, and having an acid-proof liningof FRP (fibre-reinforced plastic), rubber or PVC willbe suitable for this purpose. A heating arrangement,even if a cold process is adopted, will be ideal for

13/402 Industrial Power Engineering and Applications HandbookParameters H2s04 HClConcentration 25% 30-50%(by weight)Bath temperature 60 ? 5OC 4W5"CApproximate time 15-20 min. 15-20 min.of picklingInhibitor: since the acid 0.25-0.5% 0.25-0.5%attacks the metal and by weight by weightcauses pitting, aninhibitor must be addedNotes1 Concentration of acid and time of pickling will depend uponthe condition of the surface and the bath temperature.2 One or two dips during the course of pickling will give betterresults and enhance the efficiency of the process, as it willquickly remove loose scales.3 To reduce metal attack and fumes, use an acid inhibitor duringacid pickling. A 0.01% concentration is recommended.winters and cold climates. The heaters and thermostatmust have a stainless steel body.3 Checking bath concentration Acid content can bechecked with the help of a pH paper. If this indicatesmore than 3.5, add more acid to make it less than thisreading.4 PrecautionsAt lower concentration or lower bath temperature,the acid will attack slowly on scale and rust andwill take longer to pickle. Therefore, monitorconcentration and bath temperature.H2S04 reacts with water and generates heat. Whenusing this chemical, pour it slowly into the water.Acid fumes, being heavy, will vaporize over thetanks. They must be vented quickly to the atmospherethrough an exhaust on the pickling tank otherwisethe completed phosphate surfaces may be adverselyaffected. Apparently, the surfaces may not showrusting immediately, but may develop it while theequipment is in operation.5 Water rinsing To rinse, wash the surface in clear,continuously running water to remove all traces ofacid from the surface. Two or three dips at roomtemperature will be sufficient. The details of tankwill be similar to those for the acid tank but withoutany heating arrangement.6 De-rusting This is a process to remove rust from thesurface. The procedure of de-rusting is given in TableA13.3, column 2. The de-rusting chemical isphosphoric acid based, and does not contaminate thephosphating tank.7 Tank material A mild-steel (MS) tank with a wallthickness of 4-6 mm having a heating arrangementand a thermostat temperature control will be required.Since the phosphoric acid-based, rust solvent iscorrosion resistant, no tank lining is necessary. Theheaters and thermostat may be of stainless steel orlead-covered for better durability.8 Water rinsing To rinse, wash the surface in clearwater an MS tank with a thickness of 3-4 mm isadequate. It may, however, be provided with corrosionresistantpaint to extend its life.Heavily scaled but mildly rusted surfacesThe process is similar to the above, except that de-rustingand water rinsing tanks can now be eliminated if H2S04is used for the pickling. But when HCl is used, the secondwater rinsing tank should be retained, to remove all tracesof HCl thoroughly before it enters the phosphating tank.Otherwise the trapped traces of HC1 (chloride contents)will contaminate the phosphating bath and adversely affectthe phosphate coating. This may be shown by rusting,not immediately but in the course of time. It may benoted that traces of chloride are not removed so easilyand hence the need for a second rinsing.In place of simple water rinsing, it is more appropriateto add a neutralizing agent such as hexa-amine or sodiumnitrite (NaN02) to make a bath of 34% concentration.In the bath, the job may be dipped for two or threeminutes at room temperature to neutralize all the trappedtraces of acid. This method also helps to accelerate theprocess of phosphating when the job is transferred to thephosphating tank. The concentration of bath solution canbe checked along similar lines to those for the toner usedin the phosphating tank.The tank will be similar to the first rinsing tank. It isrecommended that the rinsing be done at an elevatedtemperature of, say, 60-70°C, even when HC1 picklingis carried out in cold conditions to agitate the air andeasily remove all traces of chloride.The bath water may be checked for any acid traceswith the help of a pH paper. This should give a readingabove 6, preferably around 7.NoteWhen de-rusting is adopted after HCl pickling, the second waterrinsing, as recommended above, is not essential as the traces ofchloride, if any, will be neutralized in the de-rusting tank, which isa phosphoric acid-based rust solvent.Mildly scaled and mildly rusted sui$acesNow the process of acid pickling may be eliminated ifdesired. Instead, only the de-rusting process can be used,as indicated in column 2 of Table A13.3. Alternativelyacid pickling may be carried out as before, but at a lowerconcentration and temperature, as noted in column 1.Since one cannot always be certain of the quality ofsheet surfaces it is advisable to follow the process ofacid pickling.Sand blastingThe scale can also be removed by shot blasting usingabrasive grits such as dry sand, less than 1 mm $. Thismethod is more suited for components not suited to thedip method and cast iron components, in which the acidmay become trapped in the porous surfaces. For sheetmetalcomponents and complicated shapes and crevices,the dip method alone is recommended.PhosphatingProcessThis is a process to provide a fine coat of zinc phosphateor zinc calcium phosphate on ferrous and non-ferrous

Table A133 Pickling and de-rusting process in mildly scaled and mildly rusted surfacesSwitchgear and controlgear assemblies I 3/403DescriptionConcentrationTemperatureApprox. time of picklin;and de-rustingInhibitorChecking theconcentrationWater rinsingPickling-cum-de-rusting tank(Hot-rolled or cold-rolledsheets with mild scaling and rust)110-15%60 f 5°C25-30%4045°C5-10 min. 10-15 Min.0.25-0.5% 0.2545%by weightby weightBy pH paper: should not be more than 3.5Only one water rinsing, with a few extra dips,is adequateDe-rusting tank (phosphoricacid-based rust solvent)2Heating type solventHot-rolled sheets withmild scalingCold type solventCold-rolled sheets15-20% 15-20%60-70°C 4045°C(M. 80°C)10-15 min. 3-5 min.Not requiredaSee footnoteAs noted in steps 1-8a The phosphoric acid-based solvent used for de-rusting can be checked as for degreasing but titration will now be carried out against anN/10 NaOH solution until the colour changes to green. A burette reading will indicate the actual pointage of the bath compared to thestandard (almost 4 for a concentration of 5% and 16 for a concentration of 20%). Obtain the standard pointage from the chemicalmanufacturer. For each one point lower strength, add 1.2 e of solvent per 100 I of bath.surfaces. It is a highly corrosion-resistant bonding toprotect the surfaces from corrosion and rust. The rustmay creep under the painted or scratched surfaces andcrevices. This is a phosphate base chemical and can beapplied cold or hot. However, the cold process is notrecommended, as it may give a coat of about 3-5 g/m2whereas for the equipment being discussed the coat mustbe above 5 g/m2.Concentration:Hot process - 3-5% by volume of bathCold process - 10-15% by volume of bathTonel; accelerator or oxidizing agentThis is alkaline in nature, say, of sodium nitrite (NaNO,)base and may be added to accelerate the process. At avery high temperature, however, above 70"C, it becomesineffective. The bath temperature must therefore be keptbelow this.Concentration:250-3OOg/lOOO 1 of bath volume.Where the phosphate coating is required to be more than5 g/m2 an extra hot process is used, as noted later, whenthe use of toner (accelerator) becomes redundant, as it isineffective above 70°C.Bath temperature and approximate time of dipping1 For the accelerated process:For a normal coating:Hot process: 60-70°C for 2-5 minutes will providea phosphate coating of up to 5 g/m2.Cold process: 40-45°C for 20-25 minutes willprovide a phosphate coating of up to 3-3.5 g/m2.2 For the unaccelerated process:For a heavy coating:Hot process: 80-90°C for 5-7 minutes will provide aphosphate coating of more than 5 g/m2.MaterialAn MS tank with a wall thickness of 3-4 mm, having aheating arrangement and a thermostat temperature controlwill be required. No protective lining is necessary as thephosphate coating itself is protective.Checking the concentration of the bathCarry this out as for tank no. 1 (see above):Pipette out 10 ml of bath solution.Add a few drops of methyl orange indicator and shake.Titrate it against an N/10 NaOH solution until thecolour changes to yellow.Note the addition of NaOH, which will indicate freeacidity.To the same solution add a few drops of phenolphthalein, and titrate it against N/10 NaOH until apink colour appears, which will indicate the total acidityof the bath. This is approximately 35 to 37 for aconcentration of 5% for a hot process and 60 to 64 fora concentration of 10% for a cold process. Obtain thestandard total acidity of the hot or cold processchemicals from the manufacturer.For each one-point lower strength, add 125 to 150 mlchemical per 100 e of bath. Consult the manufacturerfor the exact details.

13/404 Industrial Power Engineering and Applications HandbookChecking the toner0 By starch iodide paper (original colour white)Dip a piece of this paper into the bath solution for afew seconds.The change of colour of the paper will indicate thecondition of the toner, Le.White - no tonerPale mauve (light blue) - insufficient tonerBlue or dark blue - correct quantity of tonerBlack - excess toner.Precautions1 The phosphated surface must be transferred to thewater rinsing tank without delay.2 To protect the surface from contamination by foreignmatter, it should not be touched, wetted or subject tocondensation.3 The sludge of the phosphate bath that settles at thebottom, must be cleaned as frequently as possible.The clear solution from the surface can be siphonedinto an empty rinsing tank. After cleaning the tank,the clear solution can be poured back into the tank.4 The phosphate solution is acidic. Continuous humancontact or splashing of bath solution must be avoided,and hands or skin washed clean with a dilute solutionof 1-2% of ammonium bicarbonate.Water rinsingTo rinse, wash the surface in clear, continuous runningwater to remove all traces of soluble salts which maycause blistering on the surface. The tank can be similarto the phosphating tank. It may, however, be coated withan anti-corrosive paint to extend its life.PassivationThis is the final neutralizing rinse after the pre-treatmentto obtain a better corrosion resistance. The phosphatedsurfaces are treated with chromic acid-based or acidifiedsodium dichromate solutions which are not affected bymoisture and thus protect the phosphate coating.ConcentrationHot process - 125-150 g/1000 1 of bath volumeCold process ~ 250-500 g/lOOO 1 of bath volumeA very high content of this acidic solution may dissolvethe phosphate coating.Bath temperature and approximate dipping timeHot process - 60-70°C for 30-45 secondsThe hot process is generally not recommended as itmay dissolve the phosphate coatingCold process - 4045°C for 60 seconds or so.Tank for passivationThis is similar to that for phosphating. It may, however,be coated with an anti-corrosive paint to extend its life.Checking the bath concentrationThis can be done by checking the pH value of the bathsolution.Use a universal pH paper. The pH value should bebetween 2 and 3.If the value exceeds 3, add more chemical.If it is less than 2, drain a part of the bath and replenishit with fresh water.DqingIt is essential to dry passivated surfaces promptly to protectthem from moisture and atmospheric contamination. Thedrying may be carried out by blowing compressed air,which is easier and more economical, or by placing inthe same oven as for the paint. Special care need betaken with hidden surfaces, such as in corners, bendsand crevices, to ensure that there is no trapped moisture.SealingThe phosphate coating itself is not protective unless sealedwith a protective coating of primer. Sealing is thereforecarried out by applying a coat of primer within 12 hoursof phosphating, if the atmosphere is dry, or immediatelyif it is humid. Otherwise the atmospheric humidity mayreact with the surface and form a film of rust (Le. ferricoxide (Fe,O,)).Nores1 The chemical concentrations, bath temperature and process timesnoted above are only indicative and for general reference. Theymay vary with the type of chemicals, the manufacturer and thecondition of the surface to be treated. Details may be obtainedfrom the chemical manufacturer to formulate the internal sheettreatment process guidelines.2 It is strongly recommended to check the concentration of allthe baths every day before commencing work. A passivationsolution particularly, must be changed frequently, rather thanadding more chemical to the same bath, depending upon theamount of work every day.A13.3 Pre-treatment of non-ferrous components1 Degreasing - with neutral or non-causticbased chemical, otherwisesame as for ferrous metals.2 Pickling and de-rusting - Not necessary, as there isno scale formation or ruston the non-ferrous surfaces.3 The rest of the process is almost the same as for ferrouscomponents.A13.4 Size of tanksThese should be suitable to accommodate the size andvolume of a switchgear or a controlgear assembly beingmanufactured by the unit. The size noted below shouldbe adequate to meet most needs:Length - 3 mWidth - 1 mDepth - 1.2 m

Switchgear and controlgear assemblies 13/405The size, however. should be commensurate with thesize of the assemblies and the scale of work.AutomationThe entire sheet pre-treatment process described abovecan also be made automatic as noted below:Set each thermostat at the required temperature wheneverheaters are provided.Define the process time of each operation and set thehoist to dip. lift and carry the job to the next tank etc.A13.5 Procedure for liquid paintingMaking the surfaceWithin 12 hours of surface pre-treatment the surface mustbe realed through a coat of primer as described in TableA13.1. After the primer coat, the surface must be airdried or stoved. The stoving method is always preferred,being faster and neater, compared to an air-drying process,which takes longer to set, and the painted surfaces maycollect suspended dust particles from the atmosphere.After the primer is set. the surface may be applied, ifrequired, with a very thin coat of putty to fill in any pinholes or other irregularities. The putty is also air dried orstoved and then rubbed gently with emery paper andwashed with water to obtain a smooth plane surface,ready to be coated with the final paint. In fact, puttyfilling is not recommended because it is wasteful. It alsoadversely affects the strength of the paint film andincreases porosity on the surface, and should be avoided;IS far as possible. It may not be necessary when coldrolledsheets are used for fabrication.A13.6 Applying the final coat of paintAfter the surface has been prepared, the final coat ofpaint is applied. The brief procedure for painting is almostthe same as for the primcr and described in Table AI 3.4.It is recommended to apply the primer or paint insidea spray booth. which would offer the following advantages:Conventional method using a spray booth (wetmethod) This traps the primer and the paint fumes,after routing them through a curtain of water, and thenexhausts them into the atmosphere. This procedure,therefore, causes no environmental pollulion. Withinthe plant area it also protects the operator and othersfrom a health hazard. It also protects machinery installednearby from paint fumes and also the plant from a firehazard. The waste water, after treatment andneutralization, is discharged into the drains. To achievethis, the spray booth is provided with blowers on the tophaving its suction through a trough of water. (Refer toFigure A 13.1, illustrating this arrangement.) It serves adual purpose: first. it creates a draught of air within thebooth to help eliminate all the paint fumes and second,it produces acurtain ofwater todissolve all oversprayedpaint. The dissolved paint can then be collected at thebottom in a trough and disposed of. and the troughcontaminatedwater can be drained out after neutralizationand effluent treatment. Figure A 13. l illustratesa typical layout of a medium-sized paint shop.Electrostatic method This is also a wet method likethe conventional process, except that the paint is nowelectrostatically charged, similar to the powder paintin a dry method as discussed later. The paint, beinghighly charged electrostatically, is wrapped around theobject automatically.Liquid paintThese paints are resin based and the paints required forsheet-metal surfaces are generally alkyd-based resins.For general industrial applications. any of the followingtypes of enamel paints may be used,Air dryingAir drying-cum-stovingStovingFor special applications, however, such as for normallyhumid areas, and contaminated or chemically aggressivelocations, epoxy paints ai-e considered to be moreappropriate. They provide a protective coating which isresistant to chemical fumes, corrosion and temperature.Chlorinated rubber paints, which also fall into the samecategory of prolective paints, may also be used for theseareas but, not being temperature resistant. are not preferredto epoxy paints.Preparation of paint, its viscosity, solvent, thicknessof one coat, air pressure, curing temperature and time ofcuring will remain the same as for the primer (TableA13.4).Important notes on Table A13.4 and procedure forpaintingIThickness of coat:The recommended thickness of the total coat (primerplus paint) will depend upon the site conditions.For a normally clean environment, a coat of up to50 microns is considered adequate. For a dusty orhumid location requiring constant servicing andcleaning, a thicker coat, say, up to 70-80 microns.is considered to be adequate.A thickness of up to SO microns is possible throughone coat of primer and paint. To obtain a greaterthickness an additional coat of paint. rather thanprimer, may be applied after almost curing the firstcoat. A thickness of primer of more than 30 micronsis riot considered satisfactory as it may diminish itsadhesive properties. To obtain a thickness of up to100 microns, for instance, each coat (one of primerand two of paint) may be around 30-35 microns.Whenever a second coat of paint is required forbetter adhesion of paint, it is better to rub the paintedsurface of the first coat with a finer emery paper

~~~ 304013/406 Industrial Power Engineering and Applications HandbookTable A13.4Priming the treated surfacesDescription Air drying Air drying-cum-stoving Stoving1 I 31 Purpose (i) To provide protection to the treated surfaces by sealing(ii) To provide an adhesive surface for the final paint(iii) To smoothc thc surface23Types of primersPreparation(i) For synthetic paints, zinc-chromate primers are recommended, which contain zinc chromatepigments and are highly corrosion resistant(ii) For epoxy paints, epoxy primers must be used(i) Remove the skin from the surface of the primer, if any, and stir well the contents of the drumto make it a homogenous mixture(ii) Filter the mixture through a cambric cloth4 To adjust the viscosity by a 21-25 seconds 21-25 seconds 21-25 secondsFord Cup Viscometer No. 4(as for synthetic and epoxy primers)~-5 Solvent (thinner):(i) For synthetic primersGeneral purposeStovingStoving(ii) For epoxy primersEPOXYEPOXYEPOXY6 Thickness of one coat:(i) Synthetic primers25-30 microns25-30 microns25-30 microns(ii) Epoxy primers35-50 microns35-50 microns35-50 microns-8 Curing temperature:(i) For synthetic primers(ii) For epoxy primers(a) Air drying(b) Epoxy ester9 Curing time(i) For synthetic primers(ii) For epoxy primers(a) Air dryingRoom temperatureaRoom temperaturea-2 hours surface dry,12-16 hours hard dryT~ _120- 130"C(Maximum140°C)Not recommendedb140-1 50°C20-30 minutes2 hours surface dry, (i) 12-16 hours at room temperature -12-16 hours hard dry (ii) 20-30 minutes at 60-80°C(b) Epoxy ester-minutes10 Surface fillers (putty)' Air drying Air drying or stoving Stoving< (synthetic or epoxy, as the case may be) 3I1 Final surface making~.(i) Room temperature" or(ii) 100-120°C(i) Room temperature" or(ii) 60-80°C(i) 12-16 hours at room temperature(ii) 20-30 minutes at 100-120°CRub the putty, if applied, with water and slightly coarse emery paper No. 180-220. It is better to rubthe surface, even when no putty is applied to provide the surface with a knurling effect, to have abetter adhesive surface for the paint.a Room temperature is considered to be around 4045°C. In winter or for cold climates, where the room temperature may be less than this,suitable heating arrangements must be provided to obtain the desired results.Air-drying epoxy paints are not required to be stoved due to chemical reaction at about 12O"C, which may affect its hardness. To speedup the curing, it may be stoved at a maximum temperature of, say, 60-80°C as noted in column 2.'As discussed earlier, putty filling is generally not advisable unless the surface is poor, uneven or has pin holes. It is advisable to use primersurfacer, which provides a thicker coat and helps to fill such surfaces. The epoxy primer is thicker and may do this job in most cases.(No. 300 to 400) and then to wash and dry it beforeapplying the second coat.For protective coatings, such as with epoxy paints,the minimum thickness of paint (primer plus paint)should be around 70-80 microns, which is alsopossible through one coat each of primer and paint.2 If different paint shades are required on outer andinner surfaces paint any one side first, cure it almostcompletely and then apply the second shade on theother side. Even when there is a wrapping on thesecond side, can be easily wiped and cleaned, withoutaffecting the first.3 The temperature and time of curing, as indicated inTable A13.4, are indicative and for general guidanceonly. They may vary with the type and quality ofpaint and effectiveness of the furnace. For exact details,consult the paint manufacturer. The operator may alsovary the given parameters slightly, based on his ownexperience and the end results.4 Curing time will also depend upon the thickness ofcoat and the shape of the workpiece.5 Air-drying paints are easily available as noted abovebut they may not suit a regular production line as:They would require a longer drying time (slow

Switchgear and controlgear assemblies 13/407AircompressorPollution freepaint fumesthrough watercurtainA/ 4 4 4/ ! III1-loExhaustfan11Spray boothExhaust fanExhaust fanInfrared radiation orclosed cycle convectioncuring furnaceWater i- BlowersBaffles to collect. paint and forman air curtainGang-way1Acid alkaliproof flooring/ ;_____________IIIOverheadI Monorail hoistIIIIStartingIGa31Exhaust fan I 4 Exhaust fantIExhaust ducting overthe pickling tankpointScrubberto dissolveacid fumesExhaustfan mNote:* Not very efficient therefore not popular.For flow diagram of the fume treatmentsystem refer to Figure AI 3.1 0PicklingtankDuctingArrangement of exhaust ducting(Detail-X-X')Figure A13.1 A liquid painting system illustrating typical layout of a medium-size paint shop showing a seven-tankpre-treatment process

13/408 Industrial Power Engineering and Applications Handbookprocess), which a regular production line can illafford.They would require a larger storage area, whichshould be dustproof but well ventilated to providefor sufficient air circulation.During the drying time, the workpiece may collectdust from suspended dust particles in the atmosphere.The air dried surfaces may not be as neat and hardas the stoved surfaces.6 Stove curing is a rapid method and is obtained bybaking the paint for a specific time in a furnace at aspecific temperature. The temperature and time willdepend upon the type of paint being used and itsthickness, the shape of the workpiece and theeffectiveness of the furnace. (Refer to Table A13.4.)The oven may be electric or oil-fired convection type,with an arrangement to circulate the hot air aroundthe workpiece. The heaters or the furnace may beinstalled at the bottom of the enclosure to cause thehot air to circulate by natural or forced convection.The heat consumed is high in such cases, as the wholefurnace and its parts are heated first, and only thencan it heat the workpiece. The heating-up time willthus depend upon the weight and size of the job andalso the size and effectiveness of the furnace, but thejob is not influenced by any external factors, such asair draught or atmospheric dust.An infra-red (IR) bulb-type oven, where the heatingis caused by radiation, was earlier considered a moreeffective and energy-saving method compared to theconvection type due to direct heating of the paint.There was no heat loss to heat the body of the furnaceor the workpiece itself. The paint is baked at the surfaceonly, without thoroughly heating the workpiece. Suchfurnaces may not be airtight, as they are made flexibleto adjust the workpiece at the most effective distancefrom the bulbs to obtain the most effective heatradiation. Being adjustable they can also accommodateany size of job but they may be influenced by externalfactors, such as draughts and suspended dust particleswhich may collect on the surface of the workpiece.To save on heat loss and protect the job from atmosphericdirt, the radiation-type oven may be closedfrom all sides, as much as is practical. When closed,it will also conserve heat and reduce baking time.However, the influence of surrounding conditions andthe draughts are found to be great deterrents to takingfull advantage of radiation efficiency. The latest trend,therefore, is to use a closed-chamber, convection-typefurnace (Figure A13.2). The use of glass-wool andthermocoal as interior insulation now makes it possibleto require only a moderate heat for the interior of thefurnace and to prevent any heat loss through thefurnace's body. The furnace is now totally dust-freeand such furnaces are thus highly energy efficient forcuring paint.Figure A13.2DoorClosed-chamber convection-type furnace1 Flexibility and adhesion test This is conducted tocheck the flexibility of the coat. It can be carried outon a conical mandrel or on a folding apparatus asillustrated in Figure A13.3 (IS0 3205 and 3270) bybending the test piece on it. The surface to be testedis kept on the outer side. The piece is bent through180" (almost double folded) and examined for anycracks in the film. No cracks should develop.2 Stripping or hardness test This is carried out tocheck the hardness of the painted surface and can beperformed by a scratch hardness tester (IS0 3205 and3270). A weighted tungsten-tipped needle is fixed atthe far end of the test piece through a weight of 1 kg.The needle is then drawn at 30-40 m/s through itscoated surface and the weight is increased up to 4 kg.-A13.7 Testing of the painted surfacesThe following tests are generally recommended for testingthe painted surfaces:Figure A13.3Apparatus for determining flexibility and adhesion

Switchgear and controlgear assemblies 13/409After the test, the scratch so marked should not showup the bare metal surface. Figure A13.4 illustrates thetest arrangement.A simpler method would be using scratching knivesand making small squares in the surface (minimum100) around 1 mm2 for primer and 2 mm2 for thefinal surface and applying adhesive cellophane tape25.4 mm wide, with an adhesion strength of 40 k2.8 g/mm then pulling it off suddenly. Not more than5% of the squares must peel off.3 Thickness of coat This can be checked by a magneticcoating thickness tester gauge (Figure A13.5)4 Gloss Gloss meters may be used to measure thespecular gloss of the paint.5 Shade This may be checked visually with the helpof a standard shade card.6 Corrosion-resistance test This can be done withthe help of a salt spray test. The test piece is suspendedin a salt spray chamber (FigureA13.6) for seven daysin 100% relative humidity (IS 101 and IS 11864).After the test, the surface should have no signs ofdeterioration or corrosion.7 Acid resistance This can be checked by using N/10(H2S04) solution. When a few drops are spilt on thetest piece, or when the test piece is dipped for almosthalf an hour in the solution. It should develop nocorrosion spots on the surface.test pieceCQ30-40 m/sFigure A13.4wScratch hardness testing apparatusFigure A1 3.5 Thickness tester gaugeIIBarrierSalt solutionFigure A13.6Salt fumesSalt spray chamber8 Alkali resistance This can be checked by N/10(NaOH) solution. When a few drops are spilt on thetest piece, or the test piece is dipped for almost halfan hour in the solution, it should develop no corrosionspots on the surface.A13.8 Peelable coatingThe finished outer surfaces of the assemblies may becoated with a peelable coating compound which can beeasily sprayed and air dried. The coat forms a translucentpeelable film, suitable for protecting the finished surfacesduring assembly, transportation and installation fromscratches, oil marks, grease and dirt etc. The film can beneatly stripped after the equipment is finally installed.Approximate spraying data are:Viscosity of peelable compound by Ford Cup No. 4 =- 150-180 secondsAir pressure - 3-4 kg/cm2 (40-50 Ib/in2)Film thickness - 20-30 microns (one coat isenough)Drying time - Surface dry 20-30 minutes andhard dry in about four hoursA13.9 Electrostatic technique of powderpaintingIn the previous section we discussed a more conventionaltype of painting process. A rather new (early 1960s) andmore advanced technology is found in the electrostaticprocess of a powder coating system. This uses no liquidsand no primer coat and can save on paint consumptionby up to 50% compared to the conventional liquid paintmethod, due to an almost closed-loop painting process,incorporating a paint recovery and recycle system. Itallows no paint fumes to the atmosphere, and causes noenvironmental pollution or fire hazard at the workplace.The technique is thus judged to be highly economical,besides being environment-friendly. It also ensures anabsolutely uniform and perfect painted surface due to a

13/410 Industrial Power Engineering and Applications Handbookuniform electrostatic field. Since the field is the same atall points of the metal surface, it attracts the chargedpowder paint particles with equal force and allows onlyas much paint at the surface as is actually required andfor which the spray gun is set (Le. the electrostatic field).Principle of electrostatic techniqueSimilar electric charges repel and opposite electric chargesattract each other. This is the principle on which theprocess of electrostatic spray painting is based. The paintparticles are charged with a strong electrostatic field,which is created at the tip of the electrostatic spray gun,as a result of high potential difference between the tip ofthe gun and the object. The gun is maintained at a highnegative potential with the help of a power pack unit,generating adjustable d.c. output, in the range of 0-100kV, negative (Figures A13.7(a) and (b)). The variablepotential difference is necessary to adjust the distanceElectrostatic gun?Figure A13.7(a) Electrostatic spray gun with the power packunit (Courtesy: Statfied)between the gun and the object and also maintain thepaint thickness. Grounding may be performed by thehooks holding the workpiece. The highly charged paintis released through the gun at high pressure and is attractedby the oppositely charged object, to which it uniformlyadheres. Both sides of the object may be painted at thesame time.Due to the diminishing strength of the electrostaticfield, which will depend upon the thickness of the paintalready coated, the spraying process provides a uniformcoating on the entire surface. The process may even bemade automatic through tracer guns, which may be setto move on a set pattern, like robots. For more detailscontact the manufacturers. This system may be moreadvantageous for industries that have jobs of a repetitivenature and in large quantities such as cars and homeappliances.Powder paintsPowder paints are dry coating materials and are in theform of dry polymer powders. They are electrostaticallycharged and applied with force to the surface to be painted.On heating, they polymerize and form a tough film. Asurface so produced will generally be better than theconventional wet lacquer paints, for obvious reasons.There are no solvents or thinners involved. The powderis made of resin hardeners, crosslinking components,pigments, extenders and additives etc. The blend is mixedin an extruder at a temperature of lOO"C, cooled andgranulated and sieved to separate out grits. This processcauses no blisters on the surface. It is a single-coat process.In one pass a thickness of up to 60-70 microns can beobtained. The oversprayed powder can be recovered andre-used.Curing of paintThe paint is then cured in an oven. The curing time willdepend upon the thickness of coat, shape of the workpieceand type of oven and its effectiveness. Generally, a coatof up to 60-70 microns at a stoving temperature of 180-200°C (depending upon the type of powder) should takearound 10-12 minutes to cure. Contact the manufacturersfor exact details.Precautions to be observedThe storage and processing of powder paints must becarried out under well-controlled conditions, and preferablyaway from fire hazardous areas. The powderon mixing with air becomes inflammable and can causean explosion. Powder paints should be stored at about25°C.Figure A1 3.7(b)Electrostatic gunEquipment requiredElectrostatic gun, with built-in protection to cut offpower when the gun is too close to the groundedworkpiece.HT variable d.c. generator (0-100 kV negative) froma single-phase LT power source. The power requiredis around 60 W only.

Switchgear and controlgear assemblies 13/41 1Power output and velocity controller to monitor controlover quantity of powder and thickness of coat.Spray booth with paint recovery system.Oven.Figure A13.8 shows a simple layout of a powder coatingsystem with a powder recovery arrangement and FigureA13.9 its straight-line layout.Limitations of powder coatingIn an electrostatic technique, as a result of electrostatict t tExhaust fanExhaust fanExhaust faned radiation orremove moistureEffluentEffluentqTHTb drzageExhaust ducting overthe pickling tankScrubberto dissolveacid fumes*Not very efficient therefore not popular.Note For flow diagram of the fume treatment system refer to Figure A13.10-Arranaement of exhaust ductingdetail-X-X'Figure A13.8 A dry painting system illustrating typical layout of a medium-sized paint shop for powder painting showing a seven-tankpre-treatment process

13/412 Industrial Power Engineering and Applications HandbookBlower boxwith motorConnection \ w A-\\ ,e,,\\ISeven tanks for Oven for drying Power pack unit Spray booth Powder recovery Oven for curingpretreatment [A low temperature (= 0-1 00 kV equipment (160-200°C)(= 40-60°C) furnace] d.c. negative)with electrostaticpower outfitFigure A13.9Single-line painting process for a powder systemcharge, there is a wrapping effect and the paint alsodeposits at the edges of the other side of the workpiece.It is therefore advisable that the same shade be appliedon both sides of the workpiece. While it is possible toapply different shades, this would be cumbersome andmore time consuming. A lighter shade inside and a darkershade outside is sometimes required for a switchgear ora controlgear assembly. Although possible, this is noteasily performed by this process. When two shades arerequired a more common practice is to paint one side(usually the outside) first with a powder coating and theother side may then be coated with the liquid paint.This is satisfactory only for a totally folded construction.At welded joints, powder deposition will be less asa result of flux and carbon deposition. These depositsweaken the electrostatic field at such points and reducethe powder coating. However, in practice, it may notmatter for it will hardly affect the coating by 5-10 micronsor so, which may be immaterial in total service life. Themost obvious disadvantage, however, is encountered whenthe painted surface is damaged or dented duringtransportation or installation etc., and is required to berepaired outdoors such as cars or switchgear assembliesat sites. Whenever the surface is damaged, the paint cracks,and it is difficult to be removed or touched up. Althoughit is possible to peel off the affected area with the help ofspecial solvents, available from the manufacturers of suchpaints, the process of repairs outdoors is neither convenientnor provides a reliable finish. It is better to carry outrepairs only at the plant.These are a few disadvantages that have been deterrentsin promoting this technique in all fields, the automotivesegment particularly. Nevertheless, there is a willingnessto adopt this technique, wherever possible, in view of itsother inherent advantages, discussed earlier, particularlythat of being environment-friendly. It is being used forall products where damage during transit or installationis rare such as household appliances, refrigerators, airconditioners, washing machines, etc. In the field ofswitchgear enclosures, this technique is also being graduallyadopted by most manufacturers, as once they areinstalled, there is little chance of damage.A13.10 EMuent treatment and fume controlThe process related to the paint shop does not impose asignificant pollution load on the environment comparedto many other industrial activities. It is, however, essentialthat all possible aspects of environmental pollution bywastewater, environmental hydrology, environmentalhydraulics and pneumatics, air, solid waste, noise andhazardous wastes etc. are reviewed to control any kindof pollution within the prescribed limits. Otherwisesubsequent tragedies, if caused by environmental negligencein the industrial processes, may lead to theformation of stricter environmental laws.It is mandatory for a good paint shop to controlpolluting fumes and treat wastewater before it is dischargedinto the drains. To do this, effluent treatment processesmust be carried out to prevent pollution of the environmentand contamination of ponds, rivers or farmlands, intowhich the wastewater is discharged.A paint shop may pollute the surroundings in twoways: first, by acid and paint fumes that contaminate theinside of the plant and also pollute the outside environment;second, by the discharge of the pre-treatment tanksand oversprayed paint that pollute the drainage systemor where it is finally discharged.NoteThe same process will hold good for treating effluents from anelectroplating shop should it also exist with a pain1 bhop. The effluentsfrom an electroplating shop are strong wastes and need specialconsideration, as noted in the subsequent text and considered inFigure A13.12.Physio-chemical analysis of effluent wasteIn order to design a treatment system for industrial

Switchgear and controlgear assemblies 13/41 3effluents, it is essential to determine the physio-chemicalcharacteristics of the effluents. For this:Composite samples can be collected over an 8- or 24-hour period at half-hour or one-hour intervals fromthe regular effluent discharges such as from the pretreatmenttanks. The volume of each sample will dependupon the variation in the discharge flow rate.Grab samples can be collected from periodic dischargessuch as from the spray booth where the fresh water isreplenished only periodically.Direct samples can be obtained from the pre-treatmenttanks, when a new bath is made and the old one isdrained at an interval of two or three days or so,depending upon the amount of work. The samples socollected should be analysed in a recognized testlaboratory to ascertain the constituents of the effluents.A typical analysis report of the samples collectedfrom a medium-sized paint shop handling about 5 MT/day of sheet-metal work, is provided in Table A13.5.These results are only of a final discharge, and mayvary from plant to plant, depending upon the amountof work and the process of pre-treatment employed.Although the values obtained are not alarming, theydo require effluent treatment, as the untreated effluentdoes not conform to the limits prescribed as standardtolerance level, to comply with the environmentallegislation. Using environmental engineering practicesone can obtain the desired standard levels in allparameters as in Table A1 3.6 or as may be prescribedby the local civic authorities of a particular area.TableA13.5 also indicates the test results of the samplesof the final discharge when the effluents are treated alongsimilar lines as discussed later. The test results are wellwithin the tolerable limits. The recommended tolerancelevels are provided in Table A13.6. Any constituentexceeding the prescribed limits must be properly treatedbefore final discharge.The recommended tolerance levels, as noted in TableA13.6, are applicable to an industrial area where thefinal discharge of the treated effluent is let through apublic sewer. These levels may vary, depending uponTable A13.5 Average test results of effluent discharge (tankwash and spray booth mix of a medium-sized paint shoounder consideration) before and after treatmentTable A1 3.6SI:no.123456I891011121314151617181920SourceRecommended tolerance levelsParameterLevel of BOD(biochemical oxygendemand)Level of COD (chemicaloxygen demand)Presence of alkali tracesPresence of acid tracesTotal suspended solidsOil and greaseDissolved phosphates(as P)Chlorides (as CI)Sulphates (as SO4)Cyanides (as CN)Total chromium (as Cr)Hexavalent chromium(Cr)Zinc (as Zn)Iron (as Fe)Total heavy metalsPhenolic compounds(as C6H50H)Lead (as Pb)Copper (as Cu)Nickel (as Ni)Bio-assay TestRecommended tolerancelevel (for the disposal oftreated effluent into inlandsur$ace waters)30 mg/l250 mg/lMaximum up to 9.0, asmeasured by pH valueNot less than 5.5 asmeasured by pH valueNotes:(a) Clean water has a pHvalue of I to 7.5(b) 1 mg/l = parts permillion (ppm)(c) pH: the logarithm ofthe reciprocal of thehydronium ionconcentration.(Hydronium ionconcentration of asolution varies on thedegrees of alkalinityand acidity)100 mgll10 mg/l5 mgll600 mg/l(drinking water containsaround 400-500 mg/l)1000 mg/l0.2 mgll2 mg/l0.1 mg/l5 mg/l3 mg/l7 mgn1 mg/l0.1 mgll2.0 mgll2.0 mg/l90% survival 96 hoursMinistry of Environment and Forests, CPCBSI:no.PollutantBODCODpH valueTotal dissolvedsolidsTotal suspendedsolidsOil and greasePhosphate as PChloride as CIZinc as ZnTest resultsBeforetreatmentAftertreatment58 mg/l 25 mg/l604 mgfl 242 mg/l8.25 6.84208 mg/l 1950 mg/l546 mg/l 25 mg/l9 mgn 4 mg/l62.2 mg/l Nil1985 mgll 5 IO mg/l3.5 mg/l NilTolerancelevel30 mgll250 mgll5.5-9.0-100 mgllIO mg/l5 mgll600 mg/l5 mgllthe location of the plant. For sensitive or residential areas,for instance, the levels may be more conservative, decidedby the local civic authorities. The effluent treatment willdepend upon these values.Effluent treatment and discharge of waste waterThe appropriate method for effective effluent treatmentis to segregate the strong wastes from the wastewaterand treat them separately, as shown in Figure A13.12.The strong wastes consist of discharges from electroplatingplant, spray booth and spent passivation liquor etc.Accordingly all the pollutants noted in Table A13.5 canbe treated in the following manner:

13/414 Industrial Power Engineering and Applications Handbook1 Pre-treatment area Acid fumes can be controlledby providing exhaust ducts (lip suction) on the acidpickling tank with a scrubber in the exhaust pipe asshown in Figure A13.1 to absorb the acid fumes fromthe tank and its surroundings. This then passes themthrough the scrubber, to have them dissolved in sodawater and then discharged to the atmosphere. Theacid fumes with soda water become converted intoHzSOJ or HCl, depending upon the acid used forpickling. The saturated soda water collected at thebottom of the scrubber unit is then transferred to theeffluent treatment plant for further treatment. A typicalflow diagram illustrating the fume treatment systemis shown in Figure AI 3.10.The wash water and the spent acid from all the pretreatmenttanks is also transferred to the effluenttreatment plant for further treatment. Spent passivationliquor from the passivation tank is a strong waste andit may be provided with a separate pipeline to theeffluent treatment plant, as shown in Figure A13.12.2 Painting area (applicable to liquid paints)A spray booth can be installed as discussed aboveand provided with a running-water curtain systemand an exhaust arrangement to absorb the oversprayedpaint and its fumes. The paint fumes arecirculated through the curtain of water and theharmless fumes are then discharged to the atmosphere.The wash water mixed with paint which is a strongwaste can be transferred (periodically, if the watercurtain system is a closed cycle rather than acontinuous running water system) to the effluenttreatment plant for further treatment, as shown inFigure A13.12.3 Ventilation The paint shop must be well ventilatedand provided with exhaust fans to circulate fresh airwithin the shop. Roughly 20 to 30 air discharges perhour are considered adequate. Wet-painted surfaces,however, must be protected from suspended dustparticles in the atmosphere and this can be done bySoda water inletk5-29uClean airClean air4Fumes7Scrubber unitI TiFlow control 3Saturated water toeffluent treatment(pH value 9-10.5)Figure A13.10 Typical flow diagram for fume treatment system

Switchgear and controlgear assemblies 13/41 5promptly drying in an oven. Air-drying paint thereforemust be avoided as far as possible.Wastewater treatmentRapid industrialization has created many problems forthe treatment and disposal of industrial wastes which aremainly responsible for the pollution of rivers. ponds andfarmlands, when they are discharged directly, withoutproper treatment.The effluents from a paint shop are slightly alkalinewith moderate to weak BOD concentration. They mayalso contain traces of heavy metals such as zinc andchromium which, if not trcated. can create environmentaland health hazards by harming crops, subsoil drinkingwater and aquatic tlora and fauna. These effluents mustbe meticulously treated and neutralized before dischargingthem into the drains. The method of treatment is simpleand can be performed along the following lines.Figure A 13.1 I illustrates a flow diagram of a simpleeftluent treatment plant. Chemicals are fed into thetreatment tanks for neutralizing. coagulation, flocculationand detoxification. Sufficient air incorporation by meansot surface or diffused aeration is carried out to replacethe displaced oxygen in the waste waters. Figure AI 3.12zhows the layout and hydraulic profile of a typicaleftluent treatment plant which can be installed for thetreatment of effluents generated from a medium-siLedpaint shop. In the figure we have assumed about S MT/day of sheet metal work being handled by the shop. Theactual sizes and numbers of tanks will depend upon theamount of discharge and type of effluents. In FigureA 13.12 the sizes given should be adequate for the size ofpaint shop we have considered. The process in the flowdiagram would be capable of handling all qualities ofMS sheets. The brief treatment process is describedbelow.For level of biological chemical demand (BOD)BOD is a measure of organic matter present in wastewaterand can be degraded with the help of biomass. The levelof BOD, if it exceeds the permissible limits of 30 mgll,will render the wastewater septic and unsafe for use. Ahigher level of BOD. when detected, can be degraded(neutralized) biologically with the help of biomass (microorganism).The biomass consumes and degrades theorganic matter present in the effluent.Since the discharges of effluents from such a paintshop will generally he moderate the required BOD levelis automatically maintained without any further treatment.Where required, a moderate additional oxygen supply,for the growth of biomass, can be made available throughaeration of the wastewater. This can be carried out bypassing oil-free compressed air through the effluent. Thepercolation of effluent dissolves additional oxygen fromthe atmosphere and helps the growth of bacteria. Theoxygen requirement would be about 2 kg/kg of BODremoved.In cases of still higher levels of BOD an additionalsupply of biomass may become essential. and this can beeasily obtained from cowdung or municipal waste. Tosupplement biomass growth, nutrients such ac urea anddi-ammonium phosphate may be added.The aeration system adopted should be capable ofconverting the organic waste into more stable inorganicform or to cellular mass. The effluent is aerated at allpossible stages as shown in Figure A13.12. The flowfrom reaction and aeration tank No. S after primary andchemical treatment is passed through the primary aedimentationtank No. 6, where most of the soluble and colloidalorganic materials are metabolized by n diverse group ofbacteria to carbon dioxide and water. At the same time,a sizeable fraction ofthe effluents converted into ;I cellularmass settle by gravity sedimentation.Toxic waste (strong waste fromelectroplating shop, spray boothand passivation liquor)Wash water from pre-treatment areaWash water from scrubbetP9Chemical dosing tanks9___Treated waterY \ *__c_IAgitatorReaction tankSedimentation tankSludge drying bedsFigure A13.11Flow diagram - effluent-treatment plant

13/41 6 Industrial Power Engineering and Applications HandbookInlet-1 Chemical Agitatordosing tanks+ I q5:OlL 1500 -I 30000100 m0 mToxic --c30004b A'Chemicaldosing tanksPlan3rr "To reactiontank@oil & grease trap for toxic waste@ Reaction tank for toxic waste@ 011 & grease trap for pretreatmentwaste water@Collection or equalization andareration tank0 Reaction and aeration tank forphosphatic and acidic wastes@)Clarification or primarysedimentation tank@ Polishing tank (secondarysedimentation or storage tank)@Sludge drying bedsChemicaldosing tanksSection Ad'Clear water. for recycle orInlet-IToxic waste (strong waste from electroplatingshop, spray booth and spent passivation liquor)Inlet-2Wash water from pretreatment area andscrubber.To avoid overflow of tanks during rainy season,it is advisable to keep them covered.All dimensions are in 'mm' and just for illustration.Section B Bpe 0 1 metre 3 metre(suggestive)Section CC'Figure A13.12Layout of a typical effluent-treatment plant

Switchgear and controlgear assemblies 13/41 7The sedimentation and secondary settling tanks Nosh and 7 are designed to provide sufficient retention timefor settling of biomass. The settled biomass is recycledto aeration tank No. 5 to maintain the desired mixedliquor suspended solids (MLSS). The sludge settled intank No. 6 can be drained to the sludge drying beds No.8, where it can be dried, scrapped and disposed on land.The filtrate from the sludge drying beds is taken back tosedimentation tank No. 6. The overflow from thesecondary sedimentation tank No. 7 is treated water. freefrom effluents and can be discharged into the drains orfarmland or can also be recycled.For level of chemical oxygen demand (COD)COD provides a meawre of the oxygen equivalent tothat portion of the organic matter in a sample that issusceptible to oxidation by a strong chemical oxidant. Ahigher level of COD can be degraded (neutralized)chemically. The level of COD, however, automaticallyreduces with the level of BOD and no further treatmenti\ normally necessary.To neutralize the alkali and acid effluentsThe alkalinity or acidity in wastewater is usually determinedbb titration of' the wastewater with a standardalkali or acid to the transition point of methyl orange(pH range 3.1 to 4.4). Almost every effluent which isfree from caustic alkalinity (pH less than 8.3) requiresaddition of acid to bring it to the methyl orange transitionpoint. If the wastewater has a pH greater than 8.3, atitration with acid using a phenolphthalein indicator shouldbe made and the result recorded as alkalinity to thephenolphthalein.If the liquid i\ acid to methyl orange, alkalinity isabsent and the acidity must be determined. When methylorange is used as an indicator for titration, the resultmay be recorded as mineral acid acidity and for a feworganic acids. which give acidic solutions, the result maybe recorded as acid to methyl orange. In some cases itmay be desirable to determine total acidity by titrationwith alkali using phenolphthalein as an indicator, in whichcase the result is recorded as total acidity. This wouldinclude acidity due to carbon dioxide. as well as that dueto tnineral and organic acids.The alkalinity and acidity is generally controlled, basedon the amounts and also by maintaining the pH value ofthe effluent discharge between 5.5. and 9.0, the idealbeing around 7.5. as far as possible. The pH can bechecked with the help of a pH meter.Acidic effluents can be treated with sodium carbonate(Na2C03) or calcium hydroxide (Ca(0H):). i.c. limewater. The basic (alkaline) effluents can be treated withany mineral acid, preferably sulphuric acid (H2S04), untilthe required pH value is obtained.For phosphating and passivating effluents(phosphate as P)These are also acidic in nature, and are taken care ofautomatically when the pH value of the effluents ismaintained at the required level. If the passivation iscarried out by using chromic acid, chromium must betreated separately by using sodium tnetabisulphite.To remove solidsBy using sedimentation (settling) tank\. the suspendedsolids can be removed as they settle at the bottom. Propercoagulants and llocculants such as lime and alum areutilized to convert the amount of dissolved matter intosuspended solids.The numbers and sizes of tanks will depend upon theamount of discharge and the type of effluent\. For anormal paint shop, the size should be such that it is ableto allow enough retention time for the suspended solidsto settle at the bottom. To avoid turbulence in the tanks.the arrangement of inlet and discharge pipes can be asillustrated in Figure 13.12. From the polishing tank No.7 clear water, which will meet the desired standards. canbe discharged into the drains.The scum of oil and grease, if any. collected at thesurface of the tanks may be scooped out and destroyedwith the waste paints. The sludge (solids) that settle atthe bottom can be collected from the underflou of thetanks or pumped out.For zinc and other heavy metalsThe treatment for zinc can be carried out with normalcoagulants until it has formed any complex (chemicalcompound). Other heavy metals and cyanide constituents.however. will require special attention.Disposal of waste primer, paint, oil and greasescumThese areconsidered to be highly hazardous (\trong) wastes.The oversprayed paint that collects at the bottom of thespray booth, and on its baffles, as well as the scum of oiland grease may be collected and stored in drums and sealed.It is then disposed of at a recognized dumping ground forindustrial wastes or destroyed in an incineration plant.

~~13/41 8 Industrial Power Engineering and Applications HandbookRelevant StandardsIECTitleIS BS IS0Methods of test for ready mixed paints and 101-1 to9 BS 3900 3205, 3210enamelsMethod for specifying phosphate conversion 3618/1991 BS 3189/1996 9717/1990coatings for metals 6005/1991Apparatus for salt spray test 11864/1991 - -Specification for electrostatic hand held spraying -BS EN -equipment for non flammable material for5OO59/1991painting and finishingRelevant Standards of American Society for Testing MaterialsASTM DI654/-79a/1984ASTM B117/1990Standard method for evaluation of painted or coatedspecimens subjected to corrosive environmentsStandard test method of salt spray (FOG) testingNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publicationList of formulae usedSelection of basesActual quantityPer unit quantity, p.u. = (13.1)Chosen baseActual impedancez- v2 n (13.2)I - W A x 1000If Zp is the p.u. impedance thenz, =Z, . (base kVA) x 1000Base V 2= p.u. impedance at the original base, andZp2 = p.u. impedance at the new base(1 3.3)(1 3.4)Fault currentI .Isc =Rated current1.e.Z, Short-circuit unit impedance (1 3-51Fault MVA =Base MVA etc.ZPFurther reading(1 3.6)1 Beeman, D. (ed.), Industrial Power Systems Handbook, McGraw-Hill, New York.2 Coelho, F. ‘Modern MCC designs’, Siemens Circuit, XXII, July,No. 3 (1987).3 Lythall, R.T., Group motor control boards, Electrical Review30 June (1961).4 Lythall, R.T., The J & P Switchgear Book, Butterworths.5 Lythall, R.T., Sitnplijied Short-circuit Culculations, BelmosPeebles Ltd.

Testing of MetalenclosedSwitchgearAssemblies

Testing of metal-enclosed switchgear assemblies 14142114.1 Philosophy of quality systemsThis has been covered in Section 11.114.1.1 Quality assuranceTo fulfil quality requirements, the material inputs goinginto the making of assemblies such as MS sheets (fortheir surface finish, thickness and bending properties),hardware (for their correct sizes, quality of threads andtensile strength etc.), busbars and cables (for their areaof cross-section, conductivity and quality of insulationin the case of cables), insulators and supports (for theirsizes and quality of SMC, DMC or any other materialused), and all other switchgear components such asbreakers, contactors, switches, fuses, CTs, PTs and metersetc. (for their ratings, duty class, class of accuracy(wherever applicable) and specifications etc.) must beproperly checked and recorded, according to themanufacturers; internal quality checks and formats.This should be carried out before they are used inassemblies to eliminate any inconsistency in a materialor component.Similar stage inspections are necessary during the courseof manufacturing to assure quality at every stage andeliminate any mistakes in construction, assembly orworkmanship during the course of manufacturing. Thus,a product of desired specification and quality is ensured.14.1.2 Purpose of testingThe purpose of testing a switchgear or a controlgearassembly is to assure its compliance with the designparameters, material inputs and manufacturing consistency.Here we discuss this for a switchgear assemblyonly. For tests on a controlgear assembly, the relevanttests may be chosen from those prescribed for a switchgearassembly.14.2 Recommended testsThe following are the recommended tests according toIEC 60439- 1 for LT, IEC 60694, IEC 60298 for HT andANSI C-37/20C for LT and HT switchgear assembliesthat may be carried out on a completed switchgearassembly.14.2.1 Type testsType tests are conducted on the first enclosure of eachvoltage, current rating and fault level to demonstratecompliance with electrical and constructional designparameters. The tests provide a standard reference forany subsequent enclosure with similar ratings andconstructional details. The following tests may beconducted to demonstrate verification of the following:I Insulation resistance or measurement of leakagecurrent, both before and after the dielectric test2 Dielectric properties:Power frequency voltage withstand or HV test onpower as well as control and protective circuitsImpulse voltage withstand test for system voltages2.4 kV and above3 Temperature rise limits (or rated continuous currentcapacity)4 Short-circuit strength5 Momentary peak or dynamic current6 Protective circuits7 Clearance and creepage distances8 Degree of protection:Enclosure testWeatherproof test9 Mechanical operation.14.2.2 Routine testsRoutine tests are conducted on each completed assembly,irrespective of voltage, current, fault level and constructionaldetails and whether it has already undergone thetype tests. The following steps will form routine tests:1 Checking for any inadvertent human error duringassembly.This check may be carried out both at the works,after final assembly, and at the site after the installationbefore conducting routine tests or energizing theassembly. The checks can be carried out at a lowervoltage to detect any shorting links or spanners leftinadvertently on live terminals, weak insulation orinsufficient clearance or creepage distances betweenthe phases or phase to ground conductors. In the caseof a defect the same must be rectified immediately.2 Inspection of the switchgear assembly to check thefollowing:General arrangement and appearanceMechanical operation of all movable parts andinterlocksRandom checking of bolted connections for properjointing and tighteningGenerally all important requirements as discussedin Section 13.4 that may have been required in thespecifications of the switchgear or controlgearassembly.3 Inspection and verification of electrical wiring:For power connectionsFor controls, metering and sequential operations ifanyFor protective circuitsFor grounding of instrument transformersFor checking the insulation of the control wiringFor a polarity test of CTs.4 Verification of insulation resistance or measurementof the leakage current, both before and after thedielectric test.5 Verification of dielectric properties, limited to powerfrequency voltage withstand or HV test on power aswell as control and protective circuits.14.2.3 Seismic disturbancesIn Section 14.6 we provide a brief account of these

14/422 Industrial Power Engineering and Applications Handbookdisturbances as well as the recommended tests and theirprocedures to verify the suitability of critical structures,equipment and devices for locations that are earthquakeprone.14.2.4 Field testsGenerally the following tests may be carried out at siteafter installation and before energizing an assembly:1 Checking for any human error as described in Section14.2.2.2 Visual inspection of the switchgear assembly3 Inspection of electrical wiring (see Section 14.2.2)4 Verification of insulation resistance or measurementof the leakage current, both before and after thedielectric test, when an HV test is being conducted atsite.5 Verification of dielectric properties, limited to powerfrequency voltage withstand test. This test is neithermandatory nor recommended, but may be required ifa modification is carried out in the switchgear assemblyat site.14.3 Procedure for type testsBelow we outline the procedure for conducting the abovetests at the manufacturer's works.14.3.1 Electrical measurementsSelection of testing instrumentsThe instruments used for electrical measurements mustconform to IEC 60051-1. Instruments with the followingaccuracies must be used:For routine tests: class 1.0 accuracy or better, andfor type tests: not below class 0.5 accuracy.The current transformers and potential transformers mustconform to IEC 60044-1 and IEC 60044-2 respectively.Instrument transformers with the following accuraciesmust be used.For routine tests: Class 1 accuracyFor type tests: Class 0.5 accuracy14.3.2 Verification of insulation resistanceThis test is covered under field tests, Section 14.5(4).14.3.3 Verification of dielectric propertiesPower frequency voltage withstand or HV testThe test voltage must be as close to a sine wave aspracticable and frequency as noted in column 3 of Tables14.1 and 14.2, for series I1 and Table 13.2 for series Ivoltage systems, and applied for one minute as follows:1 If relays, instruments and other auxiliary apparatus,have a voltage other than the main voltage, they shouldbe disconnected from the main circuit, while conductingthis test and tested separately, according to theirvoltage rating as shown in Table 14.3.2 All the CT (current transformer) secondaries must beshorted and grounded while conducting the test onthe main circuit. This condition is redundant whileconducting test on the auxiliary circuit.3 PT (potential transformer) windings must be disconnectedby removing the control fuses from its both sides.4 The test voltage at the moment of application shouldnot exceed 50% of the rated test value and should beNominal system Rated maximum One-minute power frequency voltage withstandvoltagesystem voltage at a frequency not less than the rated 60 Hz(phase to ground)I 2 3kV (r.m.s) kV (r.m.s.) kV (r.m.s.) kV (r.m.s.)Standard lightning impulse(1.2/50 ,us) voltagewithstand (phase to ground)4kV beak)List I List I1 List I List I1 List I List I14.167.2013.8023.034.5-69.04.76 19 19 - - 60 608.25 26 35 24 30 75 9515.0 35 50 30 45 95 11025.8 50 70 45 60 125 15038.0 70 95 60 80 150 20048.3 120 120 100 100 250 25072.5 160 160 140 140 350 350

~~~ ~Testing of metal-enclosed switchgear assemblies 14/423Table 14.2 For series II voltage systemsInsulation levels, power frequency and impulse withstand voltages for metal-enclosed switchgear assembliesNominal systemvoltageI 2Rated maximum, system voltageOne-minute power frequencyvoltage withstand at afrequency not less than therated 60 Hz (phase to ground)3Standard lightning impulse(1.2/50 p) voltage withstand(phase to ground)4One-minute d.c. voltagewithstand (phase toground)5kV (r.m.s.)kV (r.m.s)0.480.64.161.2a13.814.4123.034.569.01 15.5b c b C b c- - -0.508 2.2 3.1 -0.635 2.2 - - - 3.1 -4.16 19.0 19 60 60 21 218.2515.0::::72.536.0 26.036.0-36.050.0- 60.080.0-80.0160.095 7595-95110- 125150 150- 35050 3150IOaANSI C-37/20C specifies this rating only for metal-clad switchgear assemblies and not for metal-enclosed bus systems.bWith breaker assemblies‘With isolator assembliesBased on ANSI C-37/20CNotes1 (a) The procedure for a d.c. test is same as for a.c. Due to variable voltage distributions when conducting d.c. tests, the ANSI Standardrecommends that the matter may be referred to the manufacturer for system voltages 25.8 kV and above.(b) For a power frequency voltage withstand test, the d.c. test is generally not recommended on a.c. equipment, unless only d.c. testvoltage is available at the place of testing. The d.c. test values, as provided above, are therefore for such cases only and areequivalent to power frequency a.c. voltage withstand test values.2 Power frequency tests after erection at site, if required, may be conducted at 75% of the values indicated above. Also refer to field tests(Section 14.5, Table 14.8).Table 14.3 For series I voltage systems. Dielectric test voltagesfor control and auxiliary circuits and also LT power circuitsNominal auxiliary/powervoltageOne-minute power frequencyvoltage withstand at anyfrequency between 45 and65 Hz (between phases andground)Control and auxiliary circuits1 Where the rated insulation 1000 V (r.m.s.)voltage does not exceed 60 V2 Where the insulation voltage (2 V, + 1000) volts with aexceeds 60 Vminimum of 1500 V (r.m.s.).Power circuits3 For voltages between 300and 660 VAs in IEC 60439-1(i) For fixed as well as draw-out assembliesThe trolley is in the service position in draw-out assemblies(Section 13.3.2).1 Between phase to phase and each phase to groundwith the switching devices in the closed position2 Between phase to phase and each phase to groundwith the switching devices in the open position. Thistest is to be conducted on both line and load sides ofthe switching device.3 Between the line and the load terminals of eachphase, the main switching devices being in the openposition.2500 V (rms) (ii) For draw-out assembliesThe trolley is in test position and the main switchingdevice in the closed position,raised gradually but rapidly and then maintained atthat level for one minute, reduced rapidly up to 50%of its value and then disconnected.5 During the test, one pole of the testing transformershould be connected to ground and the frame of theassembly.1 Between phase to phase and each phase to ground onboth line and load fixed terminals.2 Between line and load fixed terminals of each phase.ANSI-C37/20C recommends this test to be conductedat a value 10% higher than specified in Table 14.2 forboth power frequency and impulse voltage withstandtests.

14/424 Industrial Power Engineering and Applications Handbook(iii) Additional test requirements for outdoor HTassembliesThe following are a few additional test requirements foran outdoor HT assembly, recommended by IEC 60694for rated voltages up to 100 kV.Wet test While an indoor type switchgear asscmblyrequires only a dry power frequency voltage withstandtest, the outdoor type switchgear assembly alsoneeds a wet test under wet conditions to check theexternal insulation. For the test procedure refer to IEC60060- 1.Artificial pollution test The purpose of this test is toprovide information on the behaviour of the externalinsulation while operating in pollutcd conditions. Thetest may be performed only if thought necessary,depending upon the degree of contamination at theplace of installation. For the test procedure refer toIEC 60060- I .Test results Any disruptive discharge or electricalbreakdown during the application of high voltage shouldbe considered as a dielectric failure.14.3.4 Impulse voltage withstand test (for systemvoltages 2.4 kV and above)The impulse voltage is applied as shown in Tables 14.1and 14.2 for series I1 or Table 13.2 for series I voltagesystems, with a full wave standard lightning impulse1.2/50 p (Section 17.6.1) a front time equal to or lessthan 1.2 ,LE, and the crest value equal to or more than therated full wave impulse withstand voltage. According toANSI C-37/20C, three positive and three negativeimpulses must be applied without causing damage orflashover. Should a flashover occur in only one or anygroup of three consecutive tests, three more tests areallowed to be conducted. If the equipment passes thesecond group of three consecutive tests, it should beconsidered as acceptable. The flashover which occurredearlier may be considered as random and irrelevant.14.3.5 Verification of temperature rise limits (orrated continuous current capacity)The test must be carried out indoors, reasonably freefrom draughts. Thermocouples should be used to measurethe temperature, and the ambient temperature can bemeasured by thermocouples or thermometers.Measurement of temperature of enclosure,insulation and current-carrying partsThe thermocouples or RTDs should be located such asto measure the hottest spot, even if this means drilling ahole in the current-carrying parts.ProcedureI The test may be conducted on a completed assemblyof a switchgear or a controlgear or a part of it,whichever may be regarded as a complete section.The purpose is to achieve a near-service condition.To do this in a multi-section switchgear or controlgearassembly it is advisable to test at least three verticalsections joined together and measuring the temperaturerise on the middle section. This is to restrict the extraheat dissipation, through the sides, except natural heattransfer, and also to simulate the influence of heattransfer to this section through other sections.2 The test must be conducted at the rated* current, at afrequency with a tolerance of +2% and -5% of therated frequency and the voltage in a sinusoidalwaveform, as much as practicable. See also Section11.3.2. The test is carried out until the temperaturereaches almost a stable state, i.e. when the variationdoes not exceed 1°C per hour. To shorten the testduration, the current may be enhanced during the initialperiod to reach a fast, stable state.3 Since it may not be practical to create the actualoperating conditions at the place of testing, normalpractice is to simulate these conditions on the followingbasis.The heat generated by a current-carrying componentor conductor is its watt loss and is expressed by I’R,where I is the current and R the resistance of thecircuit under consideration. The watt loss of eachcurrent-carrying component installed in the testassembly is estimated and added to arrive at theapproximate watt loss during the actual operation.Based on this loss is calculated of the total heatersrequired. These heaters are then suitably located inthe test assembly to represent all the incoming andoutgoing feeders, their power cables and any othercurrent-carrying component.Sources generating heat1 Power circuitsInterrupting devices - switches, breakers, MCCBs,Measurement of ambient temperafureThe ambient temperature should be measured duringthe last quarter of the test by at least three thermometersor thermocouples placed equally around the switchgearassembly, at almost the centre level and at about 1 metrefrom the body of the enclosure. The ambient temperatureto be considered must be the average of these readingsand should be within 10-40°C. To ensure that theambient temperature is unaffected by magnetic field,alcohol thermometers must be used and not mercurythermometers.

~Testing of metal-enclosed switchgear assemblies 14/425Coils of the timersControl fusesCoils of the measuring instruments (A, kWh and kWmeters etc.)Wattage of the indicating lightsVA burden of the instrument and control transformersand Control terminals etc.The VA burden and the corresponding p.f. of all suchcomponents are provided by their manufacturers. VAcos$ is the content of watt loss. For more details seeSection 15.6.l(iii).3 Power connections and control wiring The losswithin such components is measured by their resistance,which, in the case of cables, is a function of their sizeand length. The loss in the external power cables iscalculated similarly, parts of which run inside theassembly to connect the various feeders, by measuringtheir average length inside the assembly.Calculating the resistance of each current-carryingcomponent separately is a very cumbersome andlengthy procedure, in addition to being not veryaccurate due to the large number of approximations.Some of the joints and components may still havebeen omitted from these calculations. The easier andmore often recommended procedure is to measurethe resistance between the extreme ends of each feederin its ON condition by an Ohm-meter. This resistancewill also include the contact resistance of each terminaland joint.With the rating of each feeder and the resistance soobtained, the 12R loss of each feeder can be calculatedand totalled. This is the loss at the ambient temperatureat the test place. It may be corrected to the operatingtemperature of the assembly as shown in Table 14.5.For sample calculations, we have considered it to be90°C. Table 14.4 provides a step-by-step procedurefor estimation. The total watt loss of the assembly sodetermined is an estimate for the required heatersthat may be installed inside the assembly, to achievealmost a true replica of the watt loss, as during actualoperation. These heaters are located within theassembly under test to circulate heat uniformly to allparts to reach a rapid thermal equilibrium. To provideheaters for individual feeders is very cumbersomeand serves no purpose. The operating conditions aresimulated similarly in the adjacent panel sections andheaters are provided there also. If the bus rating ofthese sections is different from the rating of the sectionunder test, then the heating effect of these busbarsshould also be estimated and the rating of the heatersaltered to account for this.Main busbars (horizontal and vertical)During the test the main busbars are fed at the ratedcurrent, for which the switchgear assembly is designed.They are heated naturally and therefore no resistance ofthe main bus need be measured. The busbars are shortedat one end and the current is fed from the other througha variable-current injection set at a reduced voltage of3-10 V, or enough to achieve the rated current. Thearrangement saves on power requirement and consum-Table 14.4 Computation of heat losses for temperature rise test (single-line diagram, Figure 14.1)- -S,:no.ComponentNo. offeedersFeederrating, (I)AComponentresistance per poleR mUpoleWatt loss (PR) at roomtemperature(32 "C) W, wattsWatt loss at operatingtemperature of 90°CW, watts(A)123Power circuits55 kW DOL feeder,Comprising;1 No. Switch - 250A3 Nos. Fuses - 200A1 No. Relay - 90-50A5.5 k W DOL feeders,each comprising1 No. Switch - 63A3 Nos. Fuses - 25A1 No. Contactor - 16A1 No. Relay - 9 - 14ASFU feeder comprising1 No. Switch - 63A3 Nos. Fuses - 32A1211101232Circuit resistancebetween I/C andOIG powerterminals coveringall components andpower wiring ormetallic links andtheir contactresistances,including endterminations. Bymeasurement= 0.33As above = 0.52As above = 0.461 x 3 x 1 io2 x 0.33 x= 11.982 x 3 x 122 x 0.52 x 10-~= 0.451 x 3 x 32'x 0.46 x= 1.41Since W - R andRgo = R32 [I + -20 (90- 32)]:. W, at 9OoC,= W, . [I + "20(90 - 32)]= 11.98 [l + 3.93 x 10-3(58)]= 14.71(xz0 for copper fromTable 30.1= 3.93 x per "c)0.45 [l + 3.93 x= 0.55(58)]1.41[1 + 3.93 x (58)]= 1.73Total loss in powercircuits= 16.99 W(Contd)

114/426 Industrial Power Engineering and Applications HandbookTable 14.4 (Contd.)(B) Control circuitsI Component 1 55 kW feeders 5.5 kW feedersI QQVAa Watt loss VA" Watt loss1 63/32A SFUfeedersContactor coilCTAmmeterON lightAuxiliary contactor85 at 0.3 p.f.7.Sb5b7b15 at 0.35 p.f.25.57.55.07.05.2515 at 0.35 p.f.5b5b7b-5.25557NILLoss at 40°CWd0Total50.25 Total22.25(I) loss at 90°C w,,, = w,,,[ 1 + wz0 (90 - 40)1= 50.25 [I + 3.93 x IO-' (so)]= 60.12 for copper)2 Control cables, 2.5 mm2(copper flexible).Cable resistance at 20°C(Table 13.15) RkmControl circuit current:. Watt-loss at 20°C (For 1A) ~@ 10 dfeedel7.6less than IAx 7.6 = 0.0761000Vgo = 22.25[1+ 3.93 x 10-'(50)1= 26.62or 2 feeders = 53.24@ 5 m/ feeder7.6less than IA5 x 7.6 = 0.038 per feeder1000NIL(11) Watt loss at operatingtemperature (90°C)wgo = w20 [I 4- =20(90 - 20)]1 x 0.076 [I + 3.93 x 10-'(70)]= 0.0972 x 0.038 [I + 3.93 x 10-3(70)]= 2 x 0.048 = 0.096Loss in control circuits (I + 11) I 60.22 53.34 ITotal loss in control circuits = 113.56 W (B)If the loss in the control cables is small, this can be ignored for a quicker estimation of losses.(C) Power cables (aluminium)(mm2)Average length from cablegland to the powerterminals (m)Cable resistance at 20°C(Table 13.15) Wkm:.Watt loss at 20°CWatt loss at, operatingtemperature (90°C)Wg" = W2" [ 1 + -20 (90 - 20)xz0 for aluminium fromTable 30.1= 4.03 x lo-' per OC3 x 95 3x41 .5 2.02.50.321 x 3 x 110' x 1.5 x 0.321000= 17.4217.42 (1 + 4.03 x IO-' x 70)= 22.337.5422x3~12' x-x7.541000= 13.0313.03(1 + 4.03 x IO-' x 70)= 16.73'1, x 25I .21 x 3 x 32' x 2.5 x 1.21000= 9.229.22(1 + 4.03 x x 70)= 11.82Total loss in power cables= 50.85Total watt losses A + B + C= 181.40 W:.Heaters required for 180 W, which may be arranged in the sizes of 3 of SOW each andin Figure 14.2.of 30 W (or as convenient) and located as showna From manufacturers' catalogues at 4OoCWe may consider these at unity p.f.

Testing of metal-enclosed switchgear assemblies 14/427Table 14.5 Temperature rise limits: for buses, bus connections and other parts of a switchgear assemblyType of bus connectionLimit of hottest spot temperature rise Limit of hottest spot totalabove an ambient of 4OoCtemperatureOC "C(A) 1 For busbars and busbar connections ofaluminium or copper2 For busbars and busbar connections ofaluminium or copper silver plated orequivalent3 Terminals for external insulated cables50657090105110(B) For parts exposed to contact by a human body1 Parts handled by operator(i) of metal 15"(ii) of insulation 252 External surfaces, covers(i) of metal30b(ii) of insulation 4055657080Note: For details of temperature rise of various parts and materials of an HT switching device refer to IEC 60694.Based on IEC 60439-1 and 2aPart~ that are not frequently handled, may be allowed a higher temperature rise.bParts that are exposed but need not be touched during a normal operation, may have higher temperature rise by 25°C for metal surfacesand 15°C for insulating surfaces.ption. If the ratings of the main bus and the sectional bus(vertical bus feeding a group of feeders) are different, asin large switchgear assemblies, then two separate currentsources may be used, one to feed the main bus and theother the sectional bus. The sectional bus can now bedetached from the main bus as shown in Figure 14.2,and applied with the appropriate diversity factor, as shownin Table 13.4, to simulate the test condition to obtainalmost the operating condition. For the purpose ofillustration, we consider the single-line diagram of a powerdistribution circuit shown in Figure 14.1. The generalarrangement of its switchgear assembly is illustrated inFigure 14.2.Locate the RTDs at the likely hot spots, as at the jointsof the busbars. Figure 14.2 illustrates the likely locationsof the RTDs. The test may be carried out as noted earlierand temperature readings tabulated at 30-minute or 1-hour intervals, whichever is more appropriate. Thetemperature rise, estimated with the highest temperaturerecorded by any of the RTDs, would refer to the ambientFigure 14.1 Single-line diagram for an assembly under a heat run test

14/428 Industrial Power Engineering and Applications HandbookRemove the links andMainshort the verticalView-Abusbars bus separately IShorting/ linksMain busbarsShorting links-1cwView-AVertical sectional bus 200A[(sum of all the vertical feedersof this row) x (diversity factor).But considered as min 200Al\Variable current injection \-v-' \:;-'set for the main bus Dummy panel Panel under test Dummy panel0-2000 A, 3-10 VVariable current injectionset for the vertical bus0-IOOOA, 3-10 VNote: Dummy panels are also heated simultaneously,the same way, as the panel under testLocation of RTD'S0 Location of heatersFigure 14.2 Temperature rise test on a switchgear assemblytemperature of the test location and may be corrected tothe desired ambient temperature at which the assemblyis likely to operate. The temperature thus obtained is thetemperature rise that the assembly would attain duringcontinuous operation. The corrected temperature rise mustfall within the limits prescribed in Table 14.5.If8, = temperature rise estimated at the test location0, = ambient temperature at the test locationO,, = highest temperature recorded by any of theRTDs8, = ambient temperature at the place of installationThen8, = O,, - 8,and corrected temperature rise = 8, + (8, - 8,)A successful test will ensure:1 For insulating materials: the highest temperature risewill not exceed the hot spot temperature rise asrecommended in Table 14.6.2 For busbars and busbar connections: the highesttemperature rise will not exceed the hot spot temperaturerise as recommended in Table 14.5.Table 14.6 Temperature limits for insulating materials as usedin a switchgear assemblyClass f3finsulutingmateriulYAEBFHLimit of hottest spottemperature rise ubovean ambient of 40 "C50608090I15140Limit of hottest spottotal temperuture"C90100120130155180

Testing of metal-enclosed switchgear assemblies 14/4293 Other parts of the switchgear assembly and the auxiliarycomponents, for which limits have been specified,will not exceed the hot spot temperature rise, asrecommended in Table 14.5.Genercrl notes on testing procedureThe main bus through its entire length is fed with itsrated current, while in operation it would carry adiminishing value after every feeder or a sectional bus.The sectional bus is fed similarly.If a control bus is also used add for its heat loss. Athird current source may be required if a temperaturerise in this bus is also desired.Keep the control circuits energized if possible, to furthersave on calculations and to obtain more accurate results.In the sample calculations as shown in Table 14.4 weconsider this in a de-energized condition for the sakeof more clarity.Each feeder is considered at its optimum rating, basedon the current rating of the motor or the rating of thepower fuses in a SFU or FSU feeder while the currentmay be much less in actual operation.If the temperature rise, as determined above. exceedspermissible limits it will be desirable to provide extralouvres, a forced cooling arrangement, larger busbars ora change in their configuration whichever is moreconvenient and easy to implement.Q "The test conditions as noted above may over-estimatethe rise in temperature during actual operation. Somelatitude may therefore be considered while analysing thefinal results if the temperature rise thus estimated exceedsthe prescribed limits only marginally.14.3.6 Verification of short-circuit strengthThis test is conducted to verify the suitability of theequipment to withstand a prospective short-circuit currentthat may develop on a fault. It may also be termed thesteady state symmetrical fault current ISc or the shorttime(withstand current) rating of the equipment. Whenthe equipment is an interrupting device, it is referred toas its symmetrical breaking current.It is permissible to test just one panel of a multi panelassemblyso long as the construction of other panelsis similar and busbar arrangement and supports are thesame. The value of the prospective short-circuit currentmay be determined from a calibrated oscillogram. Thetest current in any phase should not vary by more than10% of the average in the three phases and must beapplied for a predetermined time of 1 or 3 seconds. Unlessspecified otherwise, this should be considered as to be 1second.The oscillogram must reveal continuity of the currentduring the test period. The frequency of the test circuit0 Short circuit generator@ Circuit breaker@ H.T. transformer@ Master breakerTo isolate the circuit after the test is over and to also interrupt the test inbetween, in case the test piece fails. The breaker must possess aninstantaneous capacity of more than the test current and the short-circuitMVA of the feeding generator. To achieve the desired voltage it must besuitable to perform the duties of repeated short-circuit tests.@ Switch (to perform the making and breaking duties during the test)@ Source side reactance 'L (to control the magnitude of the test current)@ Source side resistance 'R' (to control the p.f. as per the test requirements)@ High current step down transformer@ Test object@ Current shunt@ R-C voltage dividers@ Recording instruments:EM0 : Electro-magnetic oscillograph, to record power frequency quantitiessuch as short circuit average and peak currents in each phase,voltage across each phase during and after the test, generatorvoltage and time duration of test, as recorded in Figure 14.4LCRO : Cathode ray oscillograph, to record voltages of atransient nature. For instance, re-striking voltages(TRVs), whose frequency of oscillations is beyondthe response range of EMO.Figure 14.3General arrangement of a power circuit to conduct a short-circuit test

14/430 industrial Power Engineering and Applications Handbookcan have a tolerance of up to 25% of the rated frequencyfor LT and 10% for HT assemblies. Figure 14.3 illustratesa general arrangement for such a test.Inference from the oscillogramFrom the oscillogram, shown in Figure 14.4 one caneasily determine the average r.m.s. value of the shortcircuitcurrent, I , its duration and the momentary peakcurrent. For easy evaluation, this oscillogram has beendivided into ten equal parts (1 to 10) and is redrawn inFigure 14.5 for more clarity. The short-circuit commencesat point D1 and concludes at point A2, AIA2 being theoriginal zero axis. At the instant of short-circuit, the zeroaxis shifts to BlA2. DIBl is the initial d.c. componentthat decays to zero at A, at the conclusion of the test.Z, 11, ..., Ilo etc. are the r.m.s. values of the a.c.components of the asymmetrical fault current at instantsI, 2, ... , 10 as indicated. They diminish gradually andreach their steady-state condition about the original axis,A1A2, in about three or four cycles of the short-circuitcondition (Section 13.4.1(8)).The values of Z, I,, ..., Ilo can be calculated from thed.c. components and the r.m.s. values of the symmctricala.c. components, lac,, lac,, ..., Iaclo at the instants of 1, 2,..., 10 at which are referred the values I,, I,, ..., Ilo.Say, for Io, if Zaco, is the r.m.s. value of the symmetricalcomponent of the a.c. fault current and ZdcO thecorresponding d.c. component on a BlA2 curve thenI, = Jm (since zac0 and ZdcO are almost 90" apart).The values of ZI, I,, ..., I,, can be determined alongsimilar lines. The curve CI C2 defines the asymmetricalaverage fault current ZaV.If I,, is the average r.m.s. value of the asymmetricalshort-circuit current, then by using the Simpson formula,I, can be calculated by using.i ,hr. 3 (58 completed 1.16 sec. cycles)I,,, = -[I: +4(1:+ I: +I: +I: +I;)+ 2(I:+ I: + I,' +I:)+ If01This is also known as the asymmetrical breaking currentand tends to become the symmetrical r.m.s. value of thefault current I,, aftcr almost four cycles from the instantof fault initiation, as discussed in Section 13.4.1(8).For more clarity and a better understanding of theoscillogram and also to determine Iaco and Ideo moreaccurately, a few cycles of the first section of the oscillogramare shown in Figure 14.6. The d.c. component isassumed to decay quickly and approach zero by the instantB2, i.e. within the first section of the test oscillogram.The asymmetrical fault current envelope C,C4 will alsoapproach an almost steady state about its original axisAIAz by B2. 0, and O2 are considered arbitrary instantsof current zeros on the asymmetrical current wave.If and Z& are the peak symmetrical a.c. componentsof the fault current at these instants as noted inFigure 14.6 and I&, and Z&, are the corresponding d.c.components thenwill represent the symmetrical r.m.s. short-circuit current(or symmetrical breaking current of an interrupting device)at the instants O1 and O2 respectively, and39.6 kA(,rnS)UBFigure 14.4 Oscillograms of an actual short-circuit test carried out on a power distribution panel (Courtesy: ECS)

Testing of metal-enclosed switchgear assemblies 14/431Momentary peak or making current (6), Table 13.11Asymmetrical peak current envelopesAsymmetrical average fault current, lavIfiJrSub-transient stateDuration of faultNote The curves C, C2 and C3 C4 are considered symmetrical after the first few cycles say 4 to 5. For better clarity, these 4 cycles areredrawn in an enlarged form in Figure 14.6Figure 14.5Determining the average r.m.s. value of the short-time current /=from the oscillogram obtained during a short-circuit testwill represent the asymmetrical r.m.s. short-circuit current(or the asymmetrical breaking current of an interruptingdevice) at the instants O1.sustain the asymmetry for the period up to its opening. Itmay also be referred to as the interrupting duty of thedevice.Notes1 The d.c. component, Idc, at any instant should be a minimum of50% that of the corresponding peak value of the a.c. componentof the symmetrical fault current I,,,. lac,, ..., laclo, it. at anyinstant, during the period of the short-circuit condition, l,, shouldbe 2 0.5 . 42 . Iac. Otherwise the asymmetry may be ignored,being insignificant.2 The peak value of the asymmetrical fault current determinedfor the first maximum peak may be considered as the momentarypeak value of the fault current 1 ~ .3 The oscillogram also reveals the following vital informationfor an interrupting device, if used in the circuit, to make orbreak on fault;Symmetrical breaking currentThis is the steady-state symmetrical fault current, whichthe faulty circuit may almost achieve in about three orfour cycles from commencement of the short-circuitcondition at point DI (Figure 14.5) and which theinterrupting device should be able to break successfully.R.M.S. asymmetrical breaking currentThis is the r.m.s. value of the asymmetrical fault currentIo, I,, ..., Zlo that the faulty circuit may generate, i.e.Making currentThis is the same as the momentary peak value of thefault current INI and defines the capability of theinterrupting device to make on fault.Short-circuit generator (Figure 14.3)Since this equipment is short-circuited repeatedly it isspecially braced to withstand repeated voltage transientsand is mounted on a resilient base to minimize themechanical shocks transmitted to the base. If the generatoris motor driven it is disconnected just before creating theshort-circuit condition, otherwise the generator may haveto feed the motor and be subject to stress. The stator haslow reactance to give maximum short-circuit output andhas two windings per phase as illustrated in Figure 14.7.These are arranged so that they can be connected inseries or parallel, in star or delta etc., to provide fourbasic three-phase voltage systems as follows:Stator connections Nominal voltage1 Parallel delta6.35 kV2 Parallel star ax6.35= 11.OkVI,, = d m 3 Series delta 12.7 kVIt will determine the ability of the interrupting device to4 Series star x 12.7 = 22 kV

14/432 Industrial Power Engineering and Applications Handbook4JFigure 14.6 Illustration of asymmetry during a short-circuitTesting of the main circuits (with short-circuitprotective devices)For a switching device (which has not been previouslytested for a short-circuit test). This should be closedand held in the normal service position. The test voltage(that would generate the required level of fault current)may be applied on one set of terminals, the otherterminals being shorted. The test may be continueduntil the short-circuit device operates to clear the fault,but in no case for less than 10 cycles. In LT assembliesthe point where the short-circuit is created should be2 f 0.4 m from the nearest point of supply.For a switching device having no protection (e.g.an Isolator). The required test current may be appliedfor the necessary duration (1 or 3 seconds) and thedynamic and thermal strengths should be verified.For the main busbars- In LT assemblies, when the test is conducted onbusbars, the length of busbars should be minimum2 m. If it is less than this, short-circuit may becreated at the ends of the busbars.- If the busbars consist of more than one section incross-section, or different distances between thesupports or the busbars, the test may be conductedseparately on each section.Test results---A successful test should reveal no undue deformation.Slight deformation of busbars is acceptableprovided that the clearance and the creepagedistances, as given in Tables 28.4 and 28.5, aremaintained. The insulation of the conductors andthe mounting supports should show no sign ofdeterioration. The degree of protection will notbe impaired.For withdrawable parts, such as a draw-out breakeror a draw-out chassis, proper movement of themovable parts and making of the contacts shouldbe ensured. To verify this requirement, the chassismay be moved in and out for at least 50 times.Clearance and creepage distances must bemaintained in the service, test and isolated positions.(i) Parallel delta(iii) Series deltaI 1 I 14.3.7 Verification of momentary peak or(ii) Parallel stardynamic current(iv) Series starFigure 14.7 Arrangements of windings in a test generatorThis test is carried out to verify the mechanical fitness ofthe buses, their interconnections, other current-carryingparts and the mounting structure to withstand theelectrodynamic forces developed during a fault. It ismeasured by the first major peak (IM) of the oscillogramas discussed above and is obtained during the course ofthe short-time rating test (Section 14.3.6). The valueobtained will not be less than those specified in Table13.11. For more details see Section. 13.4.1(8). The testprocedure and the test current are generally the same asfor the short-time rating test, except that when the test isbeing conducted exclusively to determine the momentarypeak current, the duration must not be less than 0.3 second,Le. 15 cycles for a 50 Hz system as in IEC 60694. Refemngto the oscillogram of Figure 14.5, the momentary pear

Testing of metal-enclosed switchgear assemblies 141433value of the fault current of the first major loop, aftercommencement of the fault condition at instant D1, isindicated as ZM. Referring to the original oscillogram,(Figure 14.4), it occurs in phase ZExample 14.1For more clarity we have reproduced in Figure 14.3 an actualtest circuit and in Figure 14.4, the oscillograms of the testresults of a short-circuit test successfully carried out on anLT power distribution panel (Figure 14.8) for a system faultlevel of 50 kAfor 1 second, at CPRl (Central Power ResearchInstitute). From a study of these oscillograms (Figure 14.4),we can infer the following test results:1 No. of completed test cycles = 5858:. duration of test for a 50 Hz system = - = 1.16 second502 Test current IR = 39.6 kA/y = 47.2 kA/B = 41.2 kAHighest of the above exists in phase Y at 47.2 kA:. equivalent current rating of the equipment under testfor1 second = 47.2. J116= 47.2 x 1.077= 50.836 kA3 The maximum peak current also appears in phase Yandmeasures at 110.6 kA at the first loop of the current wave.This loop is 11 0.6/50, i.e. 2.21 times the test current andsatisfies the requirement of Table 13.11.Figure 14.8test at CPRlA motor control centre (MCC) after short-circuitAnalysis of the test resultsAs discussed above we establish two basic parametersfrom a short-time withstand test, i.e.1 Thermal capability of the equipment under test, Le.I,, and its duration2 Mechanical compatibility through the peak makingcurrent, ZM, as in Table 13.1 1.Before creating a fault condition, to obtain the requiredZsc the impedance of the test circuit is adjusted so thatthe required fault current is obtained in all the phases oncreating a short-circuit. To provide the required thermaleffect (Z: . t), the duration of test, t, is then adjustedaccor-dingly. The relevant standards therefore stipulatethat the test current may be higher or lower than requiredand can be compensated by adjusting its duration, t.However, due to the minor variations in the phaseimpedances, all the phases may not be subjected toidentical severity of faults. For instance, in the abovetest each phase has recorded a different fault current. Toevaluate the fault level from these test data, the generalpractice has been to consider the phase that has recordedthe highest fault current as the base, which may occur inany of the phases. In the above test, it has occurred inphaseY. For this fault current, the test duration is adjustedto achieve the required severity of fault in terms of thermaleffect (502 x 1 in the above case).Some userslconsultants, however, are of the opinionthat by this method the other phases are not subjected tothe same severity. Accordingly, they prefer to considerthe phase that is subjected to the least fault current as thebase. Accordingly, the test duration should be adjustedfor this phase. In the above case, the minimum severityhas occurred in phase R, with only 39.6 kA. Accordingto this philosophy the test duration should be enhanced to502 or 1.59 s(39.6)2as against 1.16 s in the above test.Even then it is essential, that the peak making current,ZM, of the required magnitude is achieved during the test.This is one parameter that cannot be established by

14/434 Industrial Power Engineering and Applications Handbookhypothesis. It is therefore imperative that the minimumpeak current according to the multiplying factor, shownin Table 13.1 1 , is obtained during the course of the testitself, for example a minimum of 2.1 x 50, i.e. 105 kA inthe above test. The multiplying factor will correspond tothe specified fault current, (50 kA) and not the test current,e.g. 47 kA in the above test. If during the course of thetest, the peak making current is less than this, the testwill be considered invalid.14.3.8 Verification of the protective circuitsAll protective circuits must be checked for continuityand the operational and sequential requirements, if any,in addition to the following:Checking for the grounding of instrument transformersby means of a low-voltage source (1 0 V or so) usinga bell, buzzer or a light.0 Control wiring insulation test, as in Table 14.3.Polarity test: to check the connections through thepotential transformer to ensure that the connectionsbetween the transformer and the meters or relays havea correct relative polarity. Otherwise the meters wouldshow erratic readings, while the relays would transmitwrong signals. This test may also be conducted with alow-voltage source of 10 V by observing the deflectionof the instruments.14.3.9 Verification of clearance and creepagedistancesThe clearances may be verified as in Table 28.4, whereasfor creepages Table 28.5 may be followed.14.3.10 Verification of degree of protectionEnclosure testThe types and degrees of enclosure protection are generallythe same as defined for motors in Section 1.15, TablesI. 10 and 1.1 1. The testing requirements and methods ofcarrying out such tests are also almost the same as formotors, and as discussed in Section 11.5.3.Weatherproof testThis test is applicable to all outdoor metal-enclosedswitchgear and controlgear assemblies, as in IEC 60298,IEC 60694 and ANSI C-37/20C. The enclosure to betested should be complete in all respects including itsmounts, bushings (for HT switchgear assemblies, 1 kVand above) and wiring. One or more vertical units can betested simultaneously as may be convenient, but not morethan 3 m panel width can be tested at a time. For amultiple unit switchboard, however, at least two verticalunits should be tested together to check the joints betweenthe units.ProcedureAll surfaces or the enclosure must be tested uniformlyfor 5 minutes each. The water will be impinged on thesurface of the enclosure from a distance of not morethan 3 m from all sides through a square shaped nozzleof a capacity of 30 Omin *lo% at a pressure of 46N/cm' +IO% and a spray angle of 60-80". (See Figure14.9). The rate at which the water is impinged on thesurface under test should be almost 5 mdminute. Standardnozzle designs are also available which can ensure thedesired quantity of water at a pressure of 46 N/cm2 *10%. Kefer to IEC 60298.For a uniform spray on the entire test surface morenozzles may be employed. Normally, surfaces of up to awidth of 3 m may be tested at a time. For larger widthsthe test may be conducted in two steps. Normally, onlyone vertical surface is tested at a time. Besides the verticalsections, the test will also be conducted on:The roof surface, from nozzles located at a suitableheight0 The floor outside the enclosure for a distance up to1 m in front of the switchgear assembly, the assemblybeing in its normal position.Figure 14.9 illustrates the above requirements.Test resultsThe test may be considered successful ifNo water droplets can be observed on the insulationof the main and auxiliary circuitsThere are no water droplets on the electrical componentsor mechanism of the equipment0 There is no significant accumulation of water on anypart of the structure or other non-insulating parts. Thisrequirement is to minimize corrosion.Platform ~GLzz=K-rFD= 2.5 to 3.0 rnI D I Minimum heiaht above floor level IFigure 14.9 Arrangement for weatherproofing test3

~ ~~~ ~~~~Testing of metal-enclosed switchgear assemblies 14143514.3.11 Verification of mechanical operation (LTand HT)This test is conducted to establish the satisfactoryfunctioning of mechanical parts, such as switching devicesand their interlocks, shutter assembly, draw-out mechanismsand interchangeability between identical drawoutmodules. A brief procedure to test these features is asfollows.Switching devicesThese should be operated SO times and removable partsinserted and withdrawn 25 times each.Mechanical interlocksThe interlocks should be set in the intended position toprevent operation of the switching device and insertionor withdrawal of the removable parts. Fifty attempts mustbe made to operate the switching device and removableparts inserted and withdrawn 25 times each.The test may be considered successful if the operationsof the interrupting mechanism, the interlocks andother mechanical features after the test are found satisfactory.14.4 Procedure for routine testsThe tests against step numbers 1, 2 and 3 in Scction14.2.2 are of a general nature and no test procedure isprescribed. The rest are similar to the tests covered undertype tests. The procedures of tests and requirements oftest results will remain the same as discusscd carlicr.14.5 Procedure for field testsThe tests against step numbers 1, 2 and 3 are of generalnature (see Section 14.2.4) and no test procedure is laiddown for them4. Verification of insulation resistance ormeasurement of the leakage currentThe purpose of this test is to check for proper insulationof all the insulated live parts and components and toensure protection of a human body against electrical shockswhile the equipment is energized and is in operation.The test detects weak insulation, if any, and this must berectified before putting the equipment into service.This test should be conducted both before and afterthe HV test if this test is to be camed out. The test beforethe HV test ensure the quality of insulation and the testafter the HV test checks that this has not deterioratedafter the HV test. If the insulation resistance is found tobe lower each circuit and component must be checkedseparately to identify the weak area and corrective stepstaken to improve the resistance to the required level. Thetest can be conducted in two ways.Insulation resistance methodThe insulation resistance can be checked with the helpof an appropriate megger (MQ-meter), as recommendedby IS 101 18-3, and shown in Table 14.7. According tothis, for an LT system, an insulation resistance of 1 MRwith a 1 kV megger for a completed switchboard,irrespective of the number or outgoing circuits, isconsidered to be safe. With this insulation resistance, theswitchboard can be put to an HV test or actual use. Thevalues of insulation resistances for higher system voltagesand the recornmended rating of megger are indicated inTable 14.7.IEC 60439- 1 recommends the minimum insulationresistance for an LT system to be 1 kR/V per circuitrderred to the rated voltage to the ground.Leakage current methodThe leakage current is measured as in IEC 60298. Thismethod is generally applicable to an HT system byapplying the full rated voltage between the insulatingsurface, say, between a phase and the ground. The leakagecurrent thus measured Fhould not exceed 0.5 mA.IllustrationReferring to Table 14.7, the recommended insulation valuewill ensure the following leakage current for an LT system:System voltage - say, 415 VInsulation resistance = 1 MR = lo6 (2:.415Maximum leakage current = -A10= 0.415 mAIf the system had been 660 V, this current would exceed0.5 mA, which is permissible for an LT system.For an 11 kV system, the recommended insulation resistanceaccording to the same table = 100 MR = 100 x lo6 !Jii xi03 A:. Maximum leakage current = 100 x 106and for 33 kV = 0.33 mA= 0.11 mATable 14.7 Insulation resistance for different voltage systemsSwtem iolrage Minimum inrrr/utroii Tipe of mrggerre,ic.tanceMRkV(dc J-~ ~1 Auxiliary and Lontrol I for one or inore Mdnudl. 0 5circuit\ (dll secondary circuitswiring circuits)2 For a completed 1 Mdnud. I Oswitchboard of up to1000 V with a numberof outgoing circuits_ _'i Above 1000 V and I 00 Motorired.up to and including min 2533 kV4 Above 33 kV IO00 Motorized.min 25~- ~A? in IS I01 18-3

14/436 Industrial Power Engineering and Applications HandbookThe values thus shown in Table 14.7 take cognisance ofthe recommendations of IEC 60298 to maintain the leakagecurrent at less than 0.5 mA for all HT systems.The dielectric test may be limited to a power frequencyvoltage withstand test. This test is considered neithernecessary nor advisable at site if already conducted atthe manufacturer’s works unless there is a majormodification or repairs at site in the existing switchgearassembly. Repeated application of a high voltage maydegrade the properties of the insulation system used. Ifsuch a test becomes necessary at site, it may be carriedout at a reduced test voltage of 85% for LT systems as inIEC 60439-1 and at 80% of the test values for HT systemsas prescribed in note 2 of Table 32.1 (a) as in IEC 60298,for series I voltage systems or at 75% as prescribed innote 2 of Table 14.2 for series I1 voltage systems, as inANSI C 37.20C.When the required test voltage is not available at site,the reduced voltage power frequency withstand test maybe carried out at still lower test voltages, depending uponthe voltage availability at site. Then the duration of thetest must be increased as shown in Table 14.8, and IS10118-3 and BS 159.In this case also verification of insulation resistanceor the measurement of leakage current will be carriedout before and after the HV test.Table 14.8Rated test voltage %10083.5157066.66051.7Duration for dielectric test at reduced test voltagesTest duration in minutes14.6 An introduction to earthquakeengineering and testingmethodsThe consequences of an earthquake on life and propertyhave caused increasing concern among scientists,engineers and educational institutions. All are becomingmore conscious about preventive measures for buildingsand critical installations against possible earthquakes tomitigate, if not eliminate, their devastating effects. It isbecoming common practice to conduct seismic studieson a region where a large and/or critical project is to belocated before commencing work and locating the moreimportant buildings in seismically safer areas. Buildings,other structures and important machines are designed towithstand earthquakes. Analytical methods and laboratorytest facilities have also been developed to demonstratethe suitability of structures to withstand scismic cvents.Additional information on newer aspects on thebehaviour of the earth during an earthquake has beenobtained on a regular basis. This has been possible throughseismographs and accelerographs installed at the variousstrategic locations throughout the world (see Section14.6.2). Availability of new information on aspects onthe behaviour of the earth’s body is making it possible toupdate and improve the above test methods and facilities.With such information, it is also hoped that our scientistsmay soon be able to predict an earthquake, its intensityand location well in advance.The likely ‘responses’ of buildings, structures andinstallations to such events are now available. These helpus to study earthquakes more closely and their effectsand enable us to take more authentic and appropriatepreventive measures at the design stage. We offer anintroduction to this subject with a view to make thestudents and the engineers more conversant with andaware of these geological phenomena and to be moreconcerned about safety for life and property. A fewexamples may be houses, buildings, hospitals, industrialplants, power generating and distributing systems, dams,bridges and handling of hazardous materials. These andsimilar systems should be given constructional and designconsiderations to make them reasonably safe against sucheffects. In further discussions we consider only thesecondary systems that are supported on the primarysystem and consist mainly of the electrical and mechanicalmachines, devices and components. The primary systems,which include houses, buildings and main structures,(columns, beams, trusses, floors, walls), dams and bridgesetc., fall within the purview of civil and structuralengineering and are not discussed here.Our present discussions relate only to the laboratorytesting of safety-related secondary systems, as areemployed in critical areas such as areas of emergencypower supply and reactor power control supply etc. ofa nuclear power plant (NPP) according to IEEE 344and IEC 60980. There are other codes also but IEEE 344is referred to more commonly. Basically, all such codesare meant for an NPP but they can be applied to othercritical applications or installations that are prone toearthquakes.It should be noted in subsequent discussion that thedesign of structures and foundations for machines andequipment play a vital role in absorbing or magnifyingseismic effects. A proper design consideration in suchareas at the initial stage can save the primary (and, in allprobability the secondary) systems from such effects toa large extent. Testing may be essential only for thecritical equipment, such as being used in the criticalareas of an NPP or similar, more critical installations.14.6.1 Seismic disturbancesRandom vibrations, such as those caused by an earthquake,cause shocks and ground movements and are termedseismic disturbances. Shocks and turbulence caused bya heavy sea, landslides and volcanic eruptions are alsoexamples of shocks that may cause vibrations and resultin tremors, not necessarily earthquakes. Nevertheless,they may requir-e design considerations similar to thosefor an earthquake, depending upon the application (e.g.naval applications, hydro projects, dams and bridges),

Testing of metal-enclosed switchgear assemblies 14/437and their vulnerability to such effects. Our presentdiscussions relate to shocks and vibrations caused by anearthquakes and laboratory testing of equipment mountedon primary systems against such effects, particularly thoserequired for critical areas of an NPP.Causes of seismic disturbancesScientists suggest different theories for the causes of anearthquake. One of many such theories is the ElasticRebound Theory. This suggests that with the evolutionof the earth, several tectonic processes have been takingplace within it. These processes have caused severedeformations in the crust and have resulted in the formationof ocean basins and mountains. These impose elasticstrains on the earth’s crust. These strains build up withthe passage of time, and eventually they overcome theresilience of the earth’s crust and result in its rupture.The rebound of the ruptured crust causes an earthquake.Geologists and seismologists have explained this theorymore comprehen-sively through the Plate Tectonic Theorywhich can be briefly explained as follows.The outer shell of the earth, consisting of the uppermantle and the crust (Figure 14.10), is formed ofa numberof rigid plates. These plates are 20 in number and areshown in Figure 14.1 1. Of these, six or seven are majorplates, as can be seen in the map. The edges of theseplates define their boundaries and the arrows indicatethe direction of their movement. These plates containthe continents, oceans and mountains. They almost floaton the partially molten rock and metal of the mantle.The outer shell, known as the lithosphere, is about 70 to150 km thick. It has already moved great distances belowthe earth’s surface, ever since the earth was formed andis believed to be in slow and continuous motion all thetime. The plates slide on the molten mantle and moveabout IO to 100 mm a year in the direction shown by thearrows. The movement of plates is believed to be thecause of continental drifts, the formation of ocean basinsand mountains and also the consequent earthquakes andvolcanic eruptions.The movement of these plates carries with it continents,ocean basins and mountains. Scientists believe thatconvection currents are generated as a result of greatheat within the earth, as illustrated in Figure 14. IO. Belowthe crust, the hot rocks and metal in liquid form rise tothe crust, cool and sink into the mantle causing a turbulencethrough heat convection. The hot rocks become hardenedat the surface of the mantle and push the crust which ispart of the hug plates that are afloat the mantle. Thismovement of plates can cause the following:I When the plates move away from each other, the moltenrock from the mantle fills the gaps between them toform ocean basins.2 While moving away from one plate, they will be movingcloser to another and may collide. One plate may pileup over the other and form mountains.3 If the plates pull down, they would sink into the mantleand melt to form ocean basins. Some of the moltenrock of these plates may travel to the earth’s surfacethrough the crevice so formed due to heat convectionand cause a volcano.Rigid Plates (= 20 Nos.)floating on the mantle inSection ofthe earthbodyABCDETemperature of thedeepest parts of thecrust can go up tothe lithosphere 870°CMolten mantlehot rock and- = 6490 km -- =I2980 km--- --J1TemperatureName of the Approx.section thickness Outer pari Inner partInner coreOuter coreMolten coreLithosphereCrustup to870°C 2200°C70 km to870°C -150 kmIto 40 km(under thecontinents)Figure 14.10 Construction of the earth4 When the plates slide past each other, they may causestresses at the edges of the crust. The stresses maybuild up and at some stage exceed the rcsilicncc of theearth’s crust and cause a fault, i.e., cause the crust torupture and shift. When this occurs, it causes anearthquake in the form of violent motion of the earth’ssurface and/or large sea waves. Major earthquakes occurbecause of this phenomenon.Magnitude and quantum of energy releasedThe magnitude of shocks and vibrations caused by anearthquake is the measure of energy released (E) at thefocal point in the form of seismic waves. It is measuredon a Richter scale. An American seismologist called

- Eurasian date14/438 Industrial Power Engineering and Applications Handbook,%Antarctic date /1 LFigure 14.11 Map showing tectonic plates and their boundaries. The arrows indicate the direction of their movements(Courtesy: World Book Encyclopaedia)Charles Richter suggested that the magnitude of anearthquake can be expressed byM = log AwhereM = magnitude of the earthquakeA = maximum amplitude, as recorded by the WoodAnderson seismograph in microns at a distance of100 km from the epicentre.Since the distance of the instrument from theepicentre will usually not be exactly 100 km, adistance correction must be applied to obtain themagnitude of the earthquake, defined as,M = log A - log A,. Distance correction curvesbetween cpiccntral distance of the seismographversus log A, (which are also sometimes referredto as attenuation curves) are used for this purpose.One standard attenuation curve is shown in Figure14.12(a).This definition of the magnitude of earthquake is usedfor the records of Wood Anderson type torsionseismograph. This has a dampening equal to 80% of thecritical natural, period of 0.8 second and a magnificationof 2800.The value of M is determined from seismograph recordsat different locations and a mean value is obtained todefine the magnitude of the earthquake. The minimumvalue of M which may cause appreciable damage isconsidered to be about 5. The extent of damage causedby a higher M will depend upon the depth of focus, thedistance and the soil stratification. Generally, an earthquakecan have a focus varying from 5 to 150 km fromthe earth's surface. It is generally seen that an earthquakeof M = 5 may be felt up to a radius of 150 km and cancause substantial damage within a radius of up to 8 kmwhile an earthquake of M = 7 may be felt up to 400 kmand can cause damage up to a radius of 80 km. An M =8 may be felt up to 800 km and can cause damage up toa radius of 250 km. At Koyna (India), for instance, anearthquake having M = 6.5 was felt up to a radius of 400km and caused destruction up to 60 km or so*. Theenergy thus released is considerable and can be gaugedby its magnitude as shown in Table 14.9.To obtain an idea of the energy that may be released andthe destruction that it can cause, one may compare it withthe energy of 8 x lo2" ergs released during the atomicexplosion at Hiroshima, Japan, in 1945. This is equivalentto an earthquake of M = 6.33. The extent of destructionmay be equivalent to an explosion of 10 such bombs if Mis 7.0 and many times more at yet higher magnitudes.t,82._6g4.EaT018L 3C.z0i?8202E.YQ1-01 10 100 1000Distance in kmNote Amplitude recorded by Wood AndersonSeismograph 'A is in mm.Figure 14.12(a) Distance correction curve for determining themagnitude of an earthquake*The recent earthquake (2001) with its epicentre at Bhuj (India) was felt upto > 1000 km and caused destruction upto > 400 km. It was measuredas M = 8.1 and lasted for about 45-50 seconds.

Testing of metal-enclosed switchgear assemblies 14/439Table 14.9 Likely energy released at different magnitudes ofan earthquakeM 5.0 6.0 6.5 7.0 7.5 8.0E(1020 ergsa) 0.08 2.5 14.1 80 446 2500"An erg is the unit of work in the cgs system and is equal to theenerg required by a force of 1 dyne to move an object 1 cm (1 ergJoule).= 10- Y .The intensity of an earthquake is a subjective assessmentof its effects on the primary systems and inhabitants insurrounding areas and is measured on the Mercalli scale.As noted above, this decreases with distance from theepicentre while the magnitude remains the same. Fordetails refer to DD ENV 1998. Generally, the magnitudeand intensity of an earthquake at a location are interrelated.The energy so released propagates in the form of wavesand travels through the stratification of the earth's crustin all directions, longitudinal (X axis), transverse (Y axis)and vertical (Z axis) at the same instant, with varyingmagnitudes subjecting to vibrations whatever stands inits way on the earth's surface, such as buildings andtrees etc. These waves are recorded in the form of irregularbroad bands, Le., multi-frequency waveforms, composedof many sine waves of different frequencies, similar to aharmonic waveform, discussed in Section 23.5.2 and asshown in Figure 23.7. It is known as the time history ofan earthquake. The normal practice to describe the motionof an earthquake for a particular location, based on seismicstudies and data obtained for and from nearby areas, isin the form of a response spectrum (RS), as discussed later.Duration of an earthquakeAn earthquake may last for 4-6 seconds only for M =5.5 or less and for over 40 seconds for M > 1.5. Thegreater the magnitude, longer will be the duration. Anearthquake of M 2 6, for instance, may last for 15-30seconds and produce a maximum horizontal groundacceleration of the order of 0.1 g to 0.6 g (98 cds2 to590 cm/s2) and higher, inflicting maximum damage inthe first 5-10 seconds only, and a frequency band between1 and 33 Hz (IEEE 344).Figure 14.12(b) represents an actual time history(accelerogram) of the earthquake that occurred in Chamoli,India, on 29 March 1999. It had apeak ground acceleration ofnearly 0.15 g and a predominant frequency of about 2 Hz.It is also accepted that after such an event, the rupturedearth surfaces may try to settle down again. It is possiblethat during the course of such a realignment there maystill remain pockets of energy between the two platesuntil they finally settle. These may develop into releasesof stresses once again, leading to occasional tremors orearthquakes even for several days after a major earthquakeor volcanic eruption. The earthquakes in Turkey areexamples where two equally devastating earthquakesoccurred between September and November 1999.14.6.2 Recording an earthquakeThis is carried out with the help of an instrument knownas a seismograph which amplifies and records smallmovements of the earth's surface and helps to identifythe epicentre and focal depth and determines the magnitudeof an earthquake.More than a thousand earthquakes with a magnitude ofat least 2 (corresponding ground acceleration, < 0.01 gto 0.02 g) occur daily. But earthquakes less than M = 5are considered minor, as they are generally harmless.Earthquakes of the same magnitude may cause varyingamounts of damage at different locations, depending uponthe soil stratification and design considerations of theprimary systems. Conventional seismograph are notsuitable for recording major ground movements and gooff-scale (and sometimes are even damaged) when severeearthquakes occur in their vicinity. For recording significantground movements, strong motion accelerographsare used. Records of accelerographs are called accelerograms.They are basic requirements for seismic analysis,to design earthquake-resistant structures and buildingsand for other engineering applications. Large numbersof accelerograph stations have been established invulnerable locations throughout the world to recordseismic waves for further research and to take preventivemeasures by improving design practices. Geologists andscientists make use of these data to determine themagnitude of an earthquake and analyse the sourcemechanism to further their quest to predict an earthquakemore accurately, while design engineers use them fordeveloping earthquake-resistant systems, structures andequipment etc. So far these records have provedinsufficient to provide required forecasts about anearthquake's location, time of occurrence and magnitude.Nevertheless, with continued efforts in this direction, itis hoped that one day it will be possible to predict anearthquake more accurately.Response spectrum (RS)A response spectrum (RS) is analytically determined bycalculating the peak response (also called the spectralresponse) of a linear single degree of freedom systemwith different natural periods and dampening for a givenacceleration time history of ground movements recordedduring an earthquake. It forms a part of seismic studiescarried out for a particular area, and provides informationabout far responses of different types of structures duringan earthquake. It takes cognisance of the fact that anearthquake can be expressed indirectly in the form of aresponse spectrum. This spectrum is the peak responseof a linear single degree of freedom system on theoccurrence of an earthquake, as a function of its naturalfrequency (periods) for different dampening. They are inthe shape of a curve, natural frequency or natural periodversus peak amplitude of vibrations of the system, asillustrated in Figure 14.13. They have a broad band asnoted later and can be expressed in any of the followingforms for different dampening:1 Ground displacement response spectrum in terms ofspectral displacement versus natural period of oscillations(Figure 14.14).2 Ground velocity response spectrum in terms of spectralvelocity versus natural period of oscillations (Figure14.15).

14/440 Industrial Power Engineering and Applications HandbookN200,(153.73 cm/se?j I-200 I0 5Peak ground acceleration = 0.15 gPredominant frequency* = 2 Hz-10 1520 25Time in seconds'This can be determined by drawing a Fourier spectrum, whichwould identify the fundamental frequencies that build up themulti-frequency spectrum.(a) Acceleration-8" I0-5 10 15Time in seconds(b) Velocity20 251 lo--10 I0 5 10 15 20 25Time in seconds(c) DisplacementFigure 14.12(b) Time history of earthquake at Charnoli, India, which occurred in March 19993 Ground acceleration response spectrum in terms ofspectral acceleration, g, versus natural period of oscillations(Figure 14.16).These spectra represent the nature of peak displacement/velocity/forces and their magnitudes that may generatein a vibrating system of different damping levels andperiods on the occurrence of an earthquake. They formthe basis of equipment design and their seismic testingand are provided by the user to the equipment manufacturer.In the above instance, the response spectra shown inFigures 14.14-14.16 are for 5% damping. The dampinglevels described here are generally in the range 1-10%and frequency in the range of 1-33 Hz for systems atground level and 0.5-10 Hz for floors above the ground

-Testing of metal-enclosed switchgear assemblies 1414411 801501 1'207 0.902C2-$ 060a!:70.30-c0 000.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 2.60Period in seconds (T)Figure 14.13Typical broad band floor response spectra showing the floor acceleration at different damping levels403530A25--.$20..C u5 155361050Damping - 5%EL Centro - Sd(USA)Chamoli (India)I\I-w'I'\,', JBiF04 08 12 16 20 24 28Period in seconds (T) ~Figure 14.14 Displacement response spectra of the Koyna(India), Chamoli (India) and EL Centro (USA) earthquakeslevel. due to filtration, discussed later. It is, however,observed that most of the vibrating bodies fall in afrequency range of 2-15 Hz.Any of the three RS is adequate to derive a timehistory of an earthquake to simulate test conditions ina laboratory. This, however, being a complex subject,assistance must be obtained from experts in the fieldfor constructing an RS for laboratory testing, preparinga mathematical model for analytical assessment ordeciding the level of ground movement and frequencyof excitation etc. To assist those in the field, theInternational Association of Earthquake Engineering(IAEE) has prepared a world list, coding the variouscountries on the following bases:Seismic zonesSeismic coefficientAcceleration of ground motion (.;)Velocity of ground motion (X), andDisplacement of ground (x).The above factors must be taken into consideration whileconstructing an RS.Floor response spectrum (FRS)Some equipment will be on the ground, and some onfloors above the ground in which case floor movementmust be considered rather than ground movement. Floormovement is different from ground movement becauseof structural behaviour, characteristics and filtration ofground frequencies. Filtration of ground frequencies maylead to resonance and quasi-resonance conditions andmagnify the floor movements. compared to groundmovements.Required response spectrum (RRS)This is the response spectrum, constructed for aparticular location, for a future earthquake. It is basedon seismic studies conducted for that region andpast seismic records of and around that region, ifavailable. It forms the basic parameters for the designand testing of an object.

14/442 Industrial Power Engineering and Applications HandbookFigure 14.15 Velocity response spectra of the Koyna (India), Charnoli (India) and EL Centro (USA) earthquakesThe RRS is defined for different levels of damping,as shown in Figure 14.13. The RRS that is moreappropriate for the test object is chosen. The dampinglevel will be assessed as noted earlier, otherwise5% damping may be considered.Artificial broadening of the spectral accelerationsat the peaks of RRS: If we refer to an RS such asthat shown in Figure 14.16 we will notice a fewpeaks which, in fact, represent the resonance or quasiresonanceconditions. The RS may drop sharplyimmediately before or after such peaks. It is possiblethat during an earthquake, the corresponding peaksmay occur shortly before or after the peaks consideredin the RS as it may not be possible to estimate thefrequency of the structure so accurately, at the timeof constructing the RS. Since the test conditions1.80-0.4 0.8 1.2 1.6 2.0 2.4 2.8Period in seconds (T)Figure 14.16 Acceleration response spectra of the Charnoli(India), Koyna (India), and EL Centro (USA) earthquakes1.50t l'*O0.90C.2 c??0.60090.30L0.000.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40Period in seconds (T) --+Figure 14.17 Broadening of the spectral peaks of RRS

Testing of metal-enclosed switchgear assemblies 14/443Response spectrum0 03 sec 1 C--. , Penodic (sec) f = - 1 --1 HzzpA Zone 1f--33 Hzf, frequency (Hz)ZPA = Peak ground or floor acceleration as recorded by the(Zero period time history. The rest is a response record.acceleration)Figure 14.18 A ground or floor response spectrumwill only trace back the RRS, it is possible that theobject is not sufficiently loaded for such periodsand may fail during an earthquake while the testmay not be able to detect it and the object maywccessfully withstand the test. To overcome suchan uncertainty, it is normal practice to artificiallybroaden the spectral acceleration in the peak regionsof the RRS by k 15% T or so (T being periods ofpeaks). The broadened spectral peaks are illustratedin Figure 14.17.Zero period acceleration (ZPA)The maximum ground or floor acceleration, as a resultof an earthquake, can be obtained from a given RRS. Itcorresponds to acceleration at high frequency, i.e. morethan 33 Hz. This is illustrated in Figure 14.18, andrepresents the peak ground or floor acceleration of atime history of an earthquake, from which the RRS isdeveloped. During a test, the peak acceleration of theshake table motion (ZPA) should be at least 10% greaterthan the ZPA of RRS, according to IEEE 344, to accountfor any likely severity in the event of an earthquake.14.6.3 Constructing the RRSSeismic analysis is carried out for all important engineeringstructures such as dams, bridges and nuclear power plants.For regions where these are to be located the likelyexpectations of an earthquake as well as the extent of itsmagnitude must be assessed on the basis of the seismichistory and the earthquake records of the region (Figures14.12 to Figure 14.16). Based on these and other factorssuch as soil stratification, site dependent response spectraare determined. These are the RRS for equipment mountedon the ground. However, for equipment mounted on thefloors of buildings, the floor spectra are determined. Todo this from ground movements the buildingktructure isanalysed and the response time history at various floorsis determined. From these time histories the FRS fordifferent floors is established. The floor response spectrum(FRS) so obtained is used as the required responsespectrum (RRS) for floor-mounted equipment (secondarysystems). The test conditions are developed to simulatefloor spectra. They must also be regarded as the basis ofthe design response spectra for all critical equipmentand devices.The following are the main parameters that must beconsidered to arrive at the most appropriate responsespectra:1 Magnitude of the earthquake, hypocentral distanceand soil stratification.2 Based on above peak value of ground acceleration,duration and frequency range.An RRS is normally constructed for several levels ofcritical dampings as illustrated in Figure 14.13. The mostappropriate of these is then chosen for the purpose oftesting. Any of the above response spectra can bedeveloped into a time history of the earthquake. similarto that in Figure 14. I2(b).HypocentreThis defines the focus, i.e. the point of source within theearth’s body, from where the stored energy is released. Itcauses an earthquake and travels outwards in the form ofseismic waves to the earth’s surface.EpicentreThis identifies the part of the earth’s surface directlyabove the hypocentre and where it produces the mostsevere ground movements. Away from the epicentre, theacceleration and the intensity of ground movementsdiminish.Soil stratification (rocky, alluvial orsedimented etc.)From their focal point to the earth’s surface seismic wavestravel through the earth’s crust and the soil. Thestratification of soil, Le. the earth’s layers above the crust.plays an important role, as the intensity and frequenciesof an earthquake, as felt on the earth’s surface, will dependupon the type of soil strata.It is observed that as a result of damping of soil, thesoil may absorb some or most of the energy producedduring an earthquake, depending upon the thickness andtype of strata. Hence this may help to diminish. to agreat extent, ground vibrations, i.e. ground acceleration,velocity and displacement. Further studies on the subjecthave revealed the following:Bedrock Ground displacement in bedrock is less andhence there is no or only a small settlement of a structure

14/444 Industrial Power Engineering and Applications Handbookbuilt on this rock. But since the seismic forces actdirectly on the structure, there is no damping of theseforces or filtration of frequencies. The structure restingon such rock therefore should be adequate to absorband sustain all the energy of an earthquake. Rock,however, forms a solid part of the earth’s crust andprovides a stable foundation for a building or a structure.It is least affected during a seismic event, as there isvery little settlement. But in many places, the rockmay be deep below the earth’s surface and it may notbe practical or economical to build the foundations onsuch rock. The universal practice, generally, is to restthe foundations on shallow soil layers only (Figure14.19).Small or moderate thickness of soil Where there issome soil, ground displacement will be greater andseismic waves will pass through the soil. There maybe some settlement of the structure due to soilcompaction. While the structure will now be less subjectto seismic forces, this may prove to be a worse case,as in addition to the structure being subject to almostthe full intensity of the earthquake, there may also besettlement of the soil, which may result in settlementof the structure and cause it to collapse or developcracks.Building// structure \-00n0cFoundation*Alluvial soil formed of a number of layers of non-uniformnon-homogeneous soil of different stratificationsFigure 14.19 A typical stratification of soilDepth offoundationReasonable depth of soil When the soil is deeperthere may be considerable settlement before seismicwaves reach a structure. This soil consolidation maycause a substantial differential settlement of the structureand damage it. Although the intensity of the shock andground movements will now be less damage may besevere as a result of settlement rather than the intensityof the earthquake, as most of the energy will be absorbedby the soil. At an increasing distance of the structureor object from the focal point of the earthquake, groundmovements will diminish.Greater depth of soil when there is a deep layer ofsoil, the intensity of the earthquake will reduce. Thegreater the distance from the focal point, the smallerwill be ground movements. In such cases it is seenthat the settlement of the soil below the structure maybe negligible as it would have already settled by thetime the shock reached the surface. and hence damageto the structure would be reduced.Soil does not provide as solid a base as rock. Thestrength of a foundation built on soil and its ability towithstand an earthquake will therefore depend uponthe quality and depth of soils which may be formed ofa number of soil layers of different stratifications anddepths. Sandy soil or soil with sedimentary deposits,for instance, will have less strength and will provide aweaker base, as such soils may settle more during aground movement.14.6.4 Theory of testing a system for seismiceffectsA study of seismic effects on a structure, equipment ordevice will reveal its worthiness to withstand an earthquakewithout appreciable damage and perform satisfactorilyduring and after sudden shocks and vibrations. It is possibleto study their performance through prescribed seismicwithstand tests. Where a test is not possible, due to thesize and/or weight of the object, performance can beassessed through mathematical analysis. Seismic testingis a complex subject. To provide the full details here isneither possible nor the purpose of this text. We havecovered this subject only broadly to provide an introductionto the applicability of earthquake engineering tomore constructive use structures, particularly to take safetymeasures in the initial stages when commencing a newproject. For those in this field and who are seeking moredetailsklarifications on the subject, references have beenprovided at the end of this chapter. Whatever minimuminformation is considered necessary to familiarize anengineer with this subject are provided below. Nationaland international specifications on rotating machines,switchgears and switchgear and controlgear assembliesand bus systems as discussed in Chapters 11, 14 and 32,respectively, do not normally require such tests. Theybecome vital when such equipment is installed in a nuclearpower plant and where, by virtue of its failure ormalfunctioning during or after such a disturbance, theymay cause a process destabilization. Such a destabilizationmay jeopardize the safety and integrity of the main plant,and result in an accident or radioactive radiation beyondcritical limits. The radiation may cause a catastrophe to

Testing of metal-enclosed switchgear assemblies 141445the inhabitants in the vicinity. Such tests are advisableeven for machines, devices and components that are tobe installed in other critical areas, such as a refinery, apetrochemical project. handling and filling areas ofinflammable liquid, gas or vapour where also as a resultof failure of such machines system process or controlmay be jeopardized and cause serious accidents, andresulting in heavy loss of life and property. The seismicworthiness relate more to the primary system than to thesecondary. For a secondary system, it applies only to\afety-related equipment or devices installed in criticalareas as noted above. The suitability of primary systemsis verified through analytical means only, as laboratorytest for such systems are not practicable.Hydro projects, dams, bridges. naval equipment andany installations that are prone to continuous shocks andvibrations also require their primary and secondarysystems to have a better design and operational ability towithstand seismic effects or other ground/surface vibrations.No specific tests are presently prescribed for suchapplications. But response spectra can be established evenfor such locations and the primary and secondary systemsannlysed mathematically or laboratory tested.We define below some common terms in earthquakeengineering to clarify test requirements and methods:- -Ground acceleration This is the time history ofground acceleration as a result of an earthquake, whereinultiple frequency excitation predominates (Figure14.12(b). A ground response spectrum (GRS) can bederived from this history.Floor acceleration This is the time history ofacceleration of a particular floor or structure causedby a given ground acceleration (Figure 14.16). It mayhave an amplified narrow band spectrum due tostructural filtration, where single frequency excitationand resonance may predominate. depending upon thedynamic characteristics of the structure. A floor responseyectrum (FRS), as shown in Figure 14.18, can bederived from this history. Consideration of GRS orFRS will depend upon the location of the object underte\t.Broad band This means multiple frequencies ofground movements. During an earthquake these assumemulti-frequency characteristics, which are representedby broad band response spectra (Figure 14.13). When,however. such a response is transmitted to secondarysystems and objects mounted on floors, it becomes anarrow band, due to floor and structural filtration, andthe amplitude of vibration is magnified. Themagnification will depend upon the natural frequencyand damping of the secondary systems and the objects.As normal practice, all systems and objects, mountedon the ground or a floor, must undergo multi-frequencytests. The shake table is excited to achieve a movementthat represents a broad band waveform which willinclude all frequencies in the range of 1-33 Hz.Natural frequency When an object is mounted insitu (as in normal operation) and given an initial externaldisplacement or velocity in any direction and thenreleased, the body will oscillate about its initial positionin a sinusoidal waveform as illustrated in Figure 14.20.The amplitudes of oscillations will depend upon weight,stiffness and configuration. The record of theseoscillations is known as free vibration record. The rateof oscillations will determine the natural frequency ofthe object. Figure 14.20 shows one such free vibrationrecord.Damping is the characteristic of a vibrating systemwhich defines how fast the amplitudes of a freely vibratingsystem will decay. The greater the damping of a system,the faster the amplitude will decay and vice-versa. Themagnification of vibrations of a system, as a result ofground movements, will depend upon its natural frequencyand level of damping.Generally, all systems are flexible to some extent, excepta few that may be completely rigid. A flexible systemcan be represented as shown in Figure 14.21, where‘resistance’ represents the restoring force developed withinthe system, when applied with a force to displace it fromits original axis X-X‘. It will try to regain its originalshape and position and vibrate about its axis until itattenuates due to damping. Vibrations are caused as thesystem (which may be any object) returns to its originalposition and overshoots the original axis X-X’ to the otherside. Thus vibrations of the system about its axis,commence until they attenuate. The ‘dashpot’ representsthe resilient characteristic of the body that would try todamp the oscillations thus developed and attenuate it.The following mathematical expressions derive the propertiesof a system when excited by an external forceF,,,:F([) = mx” + cx’ + kxwhereF,,, = force applied to the object as a function of timex = displacement of the objectx‘= velocity of the mass attained when affected byvibrationsx’’ = acceleration of the mass attainccl during the courseof restoring force.ni = mass of the objectc = coefficient of viscous dampingK = coefficient of restoring force (or stiffness of thefoundation)This equation can also be rewritten as32 xJf+ L-.x’+ --k .xtn rn mIf w, is the natural angular frequency of vibration of anobject in rad/s, Le. 2~;fheing the natural frequency ofvibration in Hz, thenIk0, = -V’; = 2nfSince the natural time for one full vibration (one cyclc),1T= 7

14/446 Industrial Power Engineering and Applications Handbook-1.0004J= 0.208)-Figure 14.20 A typical free vibration record (sine wave) illustrating natural frequency of vibration and level of damping of an object,- System responsePeak response(spectral response)Ground motionPeak acceleration (ZPA)Figure 14.21 A single degree of freedom systemmx” kx cx’(Inertia (Restoring (Dampingforce) force) force)Free body diagram of massand 77 =2 f i77 is the fraction of the critical damping constant, then-=2q,- C %GrnFand - = X” + 2 . q . W, . x’ + W: . xinThis is an important equation that defines the behaviourof a vibrating body under different conditions of appliedforce or motion F(t). From this it can be inferred that theresponse or movement of object ‘x’ will depend upon qand cu, . q is termed the fraction of critical damping andw,, the angular natural frequency of the system. Withthe help of these equations, the response characteristicsof an object to a force F(t, can be determined.Damping characteristicsAn object can acquire the following four types of dampingcharacteristics:

Testing of metal-enclosed switchgear assemblies 14/447Undamped systems (Figure 14.22(a))Where there is no damping force, such as friction orair resistance, the object will continue to oscillate aboutits initial position for ever (but this does not happen asit is natural). Nowc=o:. q = 0 and.ql = .Y, = x1 etc.Underdamped systems (Figure 14.22(b))Most systems fall in this category. As a result of therestoring force, the object returns to its original position,overshoots its original axis and goes to the other sideand hence the oscillations commence. For mathematicalconvenience, it is generally assumed that damping isviscous in nature, which means that the amplitude ofa vibrating body will decay exponentially, i.e..Y,, .Y, .Y?--- - -.Y, .\‘2 .Y,andI-y Ior q = - log, -2n A 2Original -0positionQ(T) t-(a) Undamped free vibrationsDampi;-(b) Underdamped free vibrationsetc= x, = xiq=oxo> xh>x,> x; etc1>q > 0It-(c) Critically and overdamped system (no oscillations)Figure 14.22 Different damping levels of a free vibrating systemFor underdamped systems, 0 < q < 13 Critically damped systems (Figure 14.22(c))The object may just reach its original position. By thetime it does so, it loses all its restoring force due todamping and does not overshoot. Such systems do notoscillate. For critically damped systemsq= 14 Overdamped systems (Figure 14.22(c))The damping strength of the object is such that it mayabsorb most of its restoring force before it reaches itsoriginal position. There are no oscillations. Foroverdamped systems, q > 1.Resonance leading to negative dampingWhen the natural frequency of a system, or objectmn.( Un = J$)coincides with that of the ground orfloor: W, where ;esonance may occur due to a groundmovement, this will cause magnification of vibrationamplitudes, which will rise in successive cycles anddestabilize the system or object and render it highlyvulnerable to failure. The maximum magnification occursat a ground or floor frequency, w, slightly less than thenatural frequency, w,. The magnification of vibrations isa function of its dampening characteristics, as noted earlier.Figure 14.23 illustrates the influence of the naturalfrequency of the system or object on the magnificationof its vibration amplitudes, as a result of a groundmovement (F,,,), having a frequency, W, for differentlevels of system or equipment damping.Similarly, a build-up of amplitudes and consequenthigher swings (oscillations) are sometimes noticed intransmission line conductors, tall poles or suspensionbridges due to strong winds at critical speeds. This is acondition of negative damping, when the amplitudesmagnify and must be avoided at the design stage. Thiscan be done by selecting suitable lengths of transmissionlines between two poles, the height of poles or the lengthsof suspension bridges or by carrying out constructionalchanges in the system to achieve a natural frequencysafe to avoid a possible resonance.14.6.5 Test response spectrum (TRS)This is a response spectrum obtained during a test in alaboratory while exciting the shake table with groundmovements as in the RRS. The test object is mounted onthe shake table. The test object should respond normallyduring such movements. The test conditions (i.e. TRS)should closely overlap the required seismic conditions(i.e. RRS) of Figure 14.25.14.6.6 Test requirementsTo test large to very large objects such as primary systems,where a laboratory test is not practical, seismic checksare established through analytical means. Similarly, forvery large and heavy secondary systems such as turbines,alternators or transformers and reactors also their

14/448 Industrial Power Engineering and Applications Handbookanalysed during and after the tests. Where tests in all thethree axes are not possible, two axis tests (two orthogonalhorizontal and one vertical) are also permissible, whichcan also simulate the three axes conditions. For moredetails refer to IEEE 344 or IEC 60980.+ IW

2it is designed. For the purpose of design, the peakground movements should not be considered to beless than 0.1 g (IEE 344). The object should functionnormally during and after the earthquake withoutmalfunctioning or affecting the safety and the integrityof the nuclear plant throughout its operating life. Allthe equipment installed in a nuclear power plant mustundergo this level of testing.Safe shutdown earthquake (SSE) test This is definedby the most severe earthquake that could occur,producing the maximum vibratory ground movementsat the place of installation. Safety-related machines,devices and components should remain functionalduring an earthquake of this magnitude and maintainthe safety and integrity of the plant until a safeTesting of metal-enclosed switchgear assemblies 14/449shutdown. For the purpose of design, the peak groundmovements should not be considered to be less than0.5 g (IEEE 344). In other words, equipmentconforming to this test is capable of safely shuttingdown the whole plant and maintaining it in the eventof an earthquake of this intensity. It will also be ableto perform its duties when normal conditions arerestored. Such equipment is subjected to at least fiveOBE tests before applying the SSE test. The logicbehind such a stipulation is a statistical study ofearthquakes which suggested a higher probability ofmoderate-intensity earthquakes and a lower probabilityfor the most severe earthquakes. A small earthquakemay occur on more than one occasion for which thetest object must be suitable, whereas the most severeearthquake may occur only once in lifetime.Figure 14.24(a)tests at UoRPanels mounted on a shake table during seismic14.6.7 Test equipmentThis is a shake table, dynamically balanced, consistingof a platform on which the test object is mounted (Figure14.24(a)). Special arrangements may be required toaccommodate very large objects, such as rotatingmachines, motors, generators turbines and transformers,except extremely large machines, where it may beimpractical to establish such laboratory test facilities.UoR, India, possesses a shake table of 3.5 m x 3.5 mwith a payload capacity of up to 20 tonnes. It is providedwith electrohydraulic actuators with feedback controlssuitable to develop the required one horizontal and onevertical movement of the test table up to 3 g (ZPA). Thetest 'is usually conducted for a time history compatiblewith the RRS between a frequency range of 0.2-SO Hz.The TRS is calculated conventionally at 74 frequencypoints, as recommended by the United States NuclearRegulatory Commission (USNRC). The time duration,considered at about 20 seconds, will be adequate for theattack and decay periods of the shake table. Of this, theduration of a strong movement is considered for at least15 seconds, as illustrated in Figure 14.24(b). Time ofI 100 sarnpleskec III0.00 2.00 3.00 4.00-6.00 8.00 10.0 12.0 14.0 16.0 18.0 20.04 Rise time Strong motion time (>15 sec.) -C-~ecay time 4Time (Seconds)Note The TRS will envelop the RRS by minimum 10%Figure 14.24(b)Typical time history (RRS) of shake table movements during a laboratory test

14/450 Industrial Power Engineering and Applications Handbookrise, t,, is adjusted according to the shape of the RRS.The data is acquired at 100 samples per second, enoughto meet the requirements of accurate recording offrequency in earthquake movements.14.6.8 Test procedureSimulating an RRS in a test laboratoryNormally the user provides the nature of a probableearthquake in the form of RRS, i.e. accelerationcharacteristic curves, period versus spectral acceleration,such as those in Figure 14.18. The first objective is togenerate a signal which should be able to produce a timehistory, on a shake table, whose response spectra matchthose of the RRS.The RRS is simulated on the shake table, generallywithout the test object to protect it from repeated shocks.This is done in all three directions, X, Y and 2, by shakingthe table in the respective direction and adjusting itsmovement according to the relevant RRS. The maximumacceleration of the table is kept higher than the spectralacceleration corresponding to the ZPA of the relevantRRS, subject to a minimum of 0.1 g for an OBE and0.5 g for an SSE test, as discussed above. The movementof the table is recorded as ‘time versus g’ (Figure 14.24(b)).If the three directional facility is not available, the twoorthogonal RRS of the horizontal directions X-X’ andY-Y’ can be superimposed on each other to obtain anequivalent horizontal RRS analytically as shown in Figure14.26. This is a permissible procedure equivalenthorizontal RRS is then translated into the shake system.With the recording of the three RRS or two, as the casemay be, the table is now actuated in the respectivedirection, according to the simulated RRS and the responsespectra of the history of table acceleration obtained, whichcan be termed TRS and compared with the original RRS(Figure 14.25). This is plotted for 74 frequency points,0.2-50 Hz, as prescribed by the USNRC, most of whichare concentrated in the range 1-15 Hz, as noted earlier.The test response spectra should overlap the RRS by atleast 10% at most of the 74 frequency points, to be onthe safer side, as prescribed by TEEE 344 and illustratedin Figure 14.25. The table is actuated to obtain anacceleration equal to at least the ZPA of the respectiveRRS as noted above. Shortcomings in the TRS, if any,are compensated by readjusting the control signals to thetable actuators and the actual tests are carried out on thetest object as follows.1 OBE testPassive test The test object is mounted on the shaketable and subjected to both horizontal (X and Y)movements or one cumulative orthogonal horizontaland one vertical ground movement simultaneously.Accelerometers are mounted on the shake table tomeasure its movements and also on the test object atits most vulnerable points. These points may beidentified by the manufacturer or the user to monitorthe behaviour of the object in such locations.The test is conducted for nearly 20 seconds, ofwhich at least 15 seconds are covered for the strongmovement (Figure 14.24(b)) the remaining 5 seconds1 .ogC.cF?P 8 0.6g070.49;;:;\. . ... . ......... .. .0.2910.0 rns 100.0 ms 1.0 sNatural period (seconds) +10.0 sFigure 14.25 Comparison of the test conditions (TRS) with the required conditions (RRS). This is established in all three directions

+Testing of metal-enclosed switchgear assemblies 141451Cumulative RRS of twoorthogonal horizontal RRS'SYTest objectturned by 90'YRRS in X-X directionXYPosition-I ,'\Shake table/ Position-2Test obiectturned by 90tt.03 sec 1 sec33 Hz Period t (sec) ~1 Hzf- f, Frequency (Hz)Figure 14.26 Analytical construction of an equivalent horizontalRRSbeing for the table pickup and decay periods. Figure14.25 illustrates one test result, indicating the TRSexceeding the RRS by at least 10% at most of thefrequency points. The behaviour of the object isassessed for its structural and mechanical worthinessand the test is repeated in the following four positions:I Equipment axis X-X' parallel to the X-X' axis ofthe shake tablc (Figurc 14.27 position 1).2 Equipment turned by 90" from the first(position 2).3 Equipment axis X-X'at45" to the line of horizontalactuator of thc shakc tablc (position 3).4 Equipment turned by 90" from position 3(position 4).Objects symmetrical through their length and widthand by the distribution of thcir weight about the verticalaxis can be tested in position 1 only. The rest can betested in all four positions.Testing under energized conditions Testing equipmentunder energized conditions should bc pcrformcdonly after tests under passive conditions have beenconducted successfully. The same test in four differentpositions, under energized conditions, should beconducted and the behaviour and performance of theequipment assessed.2 SSE testIn this test equipment in a passive state is tested first inany of the four positions noted above for an OBE test.When successful, it is tested for one SSE test. If thesetests are successful, then the following tests may beconducted to complete seismic testing.Five numbers - OBE tests one after the other.I I I IPosition-3Position-4Figure 14.27 Mounting of test object and applying forceOne number - SSE test under passive state or energizedconditions as desired and the behaviour and performanceol the equiprnent assessed.Test results After each test, the test object i s subjectedto visual and routine checks. For electrical equipment,component and devices the test requirements willgenerally remain those discussed in Chapters I 1. I4and 32 for rotating machines, switchgear and controlgearassemblies and bus systems respectively, unless thereis a specific requirement for an additional test. If so,this must be specified by the user. In a switchgear ora control assembly in particular. the contacts of allbreakers, contactors, and moving and fixed power andcontrol contacts of a draw-out assembly should bemonitored during the course of seismic tests to ensurethat there is no chattering (bouncing). or breaking andmaking of contacts that may destabilize the plantfunctions or throw it out of control. The indicatingand measuring instrumentation and devices must givecorrect indications and readings. Equipment having toperform more complex duties, such as a process control,flow control, or sequence control, or when the controlpanel is also controlling and monitoring many otherdrives or devices, must be monitored more closelyduring seismic tests to detect malfunctioning of anycomponent that may jeopardize the safety and integrityof the plant. The actual operating conditions aresimulated as far as is practical, before performing such

14/452 Industrial Power Engineering and Applications Handbooktests, to represent the behaviour of all componentsand devices during the test and their repercussions onthe process they are controlling.14.6.9 Preventive measuresIn general, this discussion relates more to primary systems,which include civil foundation and structures, on whichthe whole building rests. Correct civil foundations, structures,columns, beams and trusses are major componentsin mitigating the effects of an earthquake. All these mustbe capable of withstanding the shocks and vibrations ofan earthquake, according to the response spectraconstructed for that location. It may be noted that athigher floor levels, the building tends to act like a vibratingfilter and may transmit to the object frequencies close tothe natural frequency of the secondary structure. In otherwords, the multi-frequency band of the ground movementsmay reduce to a narrow frequency band, almostdominating the natural frequency of the secondary systemand may become a potential cause of a likely resonancewith the structure. Objects on upper floors may thus besubject to higher accelerations, sometimes many timesmore than ground accelerations. Hence the necessity toavoid a resonance at the design stage itself. It is possibleto do this, by keeping the fundamental frequency of thefloor or structure, where the secondary systems are to bemounted, away from the predominant frequency that mayfilter out from the ground level.More precautions in mounting the electrical or mechanicalmachine, component and device on each floor orstructure wouldcontain themagnification of seismic effectsto a large extent. Critical machines and components thenneed be designed with such a natural frequency that theydo not resonate with the filtered-out predominant frequencyof the structure or foundation. This is possible by suitablymodifying the method of mounting or making smallalterations in the construction of the object. It may benoted that a resonance may take time to build up and anearthquake may not last that long. The occurrence of aresonance, therefore, may not be as significant as thequasi-resonance (i.e. initial stages while the resonance isstill building up) and this is reflected by the jagged peaksof the RS. The latest practice in the field of civil andstructural engineering is to build a resilient ground floorrather than arigid one, which can absorb the most vibrationsof a seismic event and filter out to the upper floors onlythe frail motions. For more details, refer to the workavailable on the subject, as mentioned in the references.By exercising these checks, most problems can beaverted at the installation stage. The remainder can beavoided in the design of mechanical systems/constructionof a machine rather than its electrical design, even in thecase of an electrical machine, to ensure their requiredbehaviour during an earthquake.The following may be a few such design areas thatmay be considered, in improving the performance of amachine during an earthquake:Bearings and bearing housings in a rotating machine,in view of the very small gap between the stator andthe rotor.Rigidity of doors, bus and wiring systems etc. in aswitchgear or controlgear assembly.Limiting the use of brittle metals such as ceramic orporcelain. Although, avoiding the use of such materialsin some cases may not be practical, as in the case oflightning arresters, bushings and insulators used in anHT switchgear, instrument and power transformers andreactors.It is mandatory to employ only static or microprocessorbasedrelays, components and devices whereverpossible.Extra care needs be taken on the quality of fastenersand the method of jointing and bolting.Resilient but rigid foundations such as by providingspring mounts or rubber pads for machines on thefloor or for components and devices mounted on themachine so that they are able to absorb the vibrations,caused by resonance and quasi-resonance effects, dueto filtered out narrow band ground movements. Thestiffness of the foundation (coefficient of the restoringforce, k) may be chosen such that it would make thenatural frequency of the equipmentmuch less than the ground (exciting) frequency, W. Itis recommended to keep > 2, to achieve an effectivePvibration isolation.The foundation and machine body/structure mustprevent internal resonance, which is possible by slightlyaltering their designs.The critical items may then be checked for the requiredresponse spectra (RRS) as discussed above.Relevant StandardsIEC Title IS BS60044- 1/1996 Specifications for current transformersGeneral requirementsMeasuring current transformers60044-2/1997 Application Guide for voltage transformersSpecification for voltage transformersGeneral requirements60044-411980 Method for measuring partial discharges in instrumenttransformers2705 BS 7626/1993Part 111992Part 2/19924201/ 199 I BS 7625/19934 l46/l 99 1 BS 7629/199511322/1990 BS 6184/1992

Testing of metal-enclosed switchgear assemblies 14/45360044-6/1992601 86/19956005 1 - 11199760060- 111 98960060-2/199460298/199460439-1/199260439-2/198760694/199660898/199560980/1989Protective current transformersRequirements for transient performanceCapacitive voltage transformersGeneral requirements for measurement andprotection (all voltages)Direct acting indicating analogue electrical measuringinstruments and their accessoriesDefinitions and general requirementsHigh voltage test techniques - General definitions andtest requirementHigh voltage test techniques. Measuring systemsA.C. metal enclosed switchgear and controlgear forrated voltages above 1kV and up to and including 52kVLow voltage switchgear and controlgear assembliesType-tested and partially type-tested assembliesParticular requirements for busbar trunking systems (busways)Common specifications for high voltage switchgear andcontrolgear standardsElectrical accessories: circuit breakers for over currentprotection for household and similar installationsRecommended practices for seismic qualification ofelectrical equipment of the safety system for nucleargenerating stationsCode of practice for selection, installation and maintenanceof switchgear and controlgear up to 1 kV. InstallationCriteria for earthquake resistant design of structuresSpecification for high voltage busbars and busbarconnections2705-3/ 19922705-4/ 19923156-1 to 4/19921248-1 to 9207 1- 1/19932071-2/19912071-3/19913427/19918623-1/19938623-2/19933427/19918828/199610118-3/19911893/19918084/1992BS 7626/1993BS 7625/1993BS 7729/1995BS 89-1/1990BS 923-111990BSEN 60060-2/1997BSEN 60298/1996BSEN 60439-111994BSEN 60439-2/1993BSEN 60694/1997BSEN 60898/1991DDENV 1998 (1-5)11996BS 159/1992IEEE 4/1995ANSI C.37.51/1989ANSI C.37.55/1989ANSI C37.57/1990ANSI/IEEE 37.81/1989ANSVIEEE-308/ 1992ANSIAEEE-323/199 1ANSI/IEEE-336/199 1ANSIAEEE-344/1993ANSI/IEEE-649/1992ANSVIEEE-693/199 1NEMA 250/1991NEMAlICS 6/1996Relevant US standards ANSUNEMA and IEEEStandard techniques for HV testingSwitchgear metal enclosed low voltage a.c. power circuit breaker switchgear assemblies-conformanceproceduresMetal clad switchgear assemblies. Conformance test proceduresMetal enclosed interrupter switchgear assemblies. Conformance test procedures.Guide for seismic qualification of class IE metal enclosed power switchgear assembliesCriteria for class IE power system for nuclear power generating stationQualifying class 1 E equipment for nuclear power generating stationStandard Installation, Inspection and testing requirements for power instrumentation and control equipment atnuclear facilitesRecommended practices for seismic qualification of class 1E equipment for nuclear power generating stations (NPGS)For qualifying class IE motor control centres for nuclear power generating stations (NPGS)Seismic design of substation equipmentEnclosure for electrical equipment up to IOOOV. Test criteria and design testsIndustrial controls and systems. EnclosurestestNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

14/454 Industrial Power Engineering and Applications HandbookFurther reading1 Agrawal, P.N., Engineering Seismology, Oxford and IBHPublishing Co. Pvt. Ltd, New Delhi, India.2 Basu, S., Kumar, A. and Lawama, K.K., ‘System identificationof a circuit breaker pole’. Proceedings of Tenth Symposium onEarthquake Engineering, Roorkee, pp. 97949 1 (1994).3 Basu, S., Prajapati, G.I., Bandopadhyay, S., Kumar A. andArya, A.S., ‘Design and development of a shake table facility’,Proceedings of the Eighth Symposium on EarthquakeEngineering Vol. I. pp. 1 to 8 (1986).4 Bressani, M., Bobig, P. and Secco, M., ‘A support experimentalprogram for the qualification of safety-related medium-voltageinduction motors for nuclear power generating stations.Presented at the International Conference on the Evolutionand Modem Aspects of Induction Machines Torino, July (1986).5 Gass, Smith, and Wilson, Plate Tectonics in Understanding theEarth, Oxburgh, E.R. (ed.), pp. 263-285 (1973).6 Krishna, J., Chandrasekaran, A.R. and Chandra, B., Elementsof Earthquake Engineering, South Asian Publishers (P) Ltd,New Delhi, India.7 Kumar, A. and Basu, S. ‘Isolation in a shake table system’,Workshop Proceedings on Base Isolation, New Delhi, June,(1989).8 Majumdar, S., ‘Seismic qualification testing of switchboards’,Siemens Circuit, 2/91 Vol. XXVl April (1991) and 1/92 Vol.XXVI January (1992).9 Pankaj, Basu, S. and Kumar, A,, ‘Seismic qualification ofequipment - a need for greater understanding,’ Proceedings ofNational Conference on Role of Continuing EngineeringEducation in Industrial Restructuring, University of Roorkee.February 1995, pp 157-162.IO Shun Zo Okamoto. Introduction to Earthquake Engineering,University of Tokyo Press, Japan (1973).I1 Symposium on Earthquake Effects on Plant and Equipment(Vols I and II), Organized by BHEL energy system groupHyderabad, India, and Indian Society of Earthquake Technology,Roorkee, India (1984).

Instrument andControlTransformers :Application andSelection151455

Instrument and control transformers: application and selection 15145715.1 introductionTransformers are used in an auxiliary circuit, linked to apower circuit, to indicate, measure and control its voltagesand currents. They find application in a switchgear or acontrolgear assembly and a switchyard. It would beimpracticable to produce indicating and measuringinstruments or protective devices to operate at high tovery high voltages or currents. The universal practice,therefore, is to transform the high voltages, say, 415 Vand above. and currents above 50 A to reasonably lowvalues, as discussed later, for these applications. Indicatingand measuring instruments and protective devices aredesigned for these reduced values. The transformers usedto transform voltages are known as voltage transformers*and those to transform currents as current transformers.Below we discuss their classifications, basic requirementsand design parameters.15.2 Types of transformer15.2.1 Voltage transformers (VTs)The\e may be classified as follows:I Instrument voltage tramformers(i) Conventional two-winding, electromagneticvoltage transformers(ii) Residual voltage transformers (RVTs) and(iii) Capacitor voltage transformers (CVTs). Thesemay be used for metering or protection, withvery little difference between the two as notedlater.2 Control transformers15.2.2 Current transformers (CTs)These may be classified as:1 Instrument current transformers(i) Measuring current transformers(ii) Protection current transformers and(iii) Special-purpose current transformers, class 'PS'.2 Interposing current transformers3 Summation current transformers4 Core balance current transformers (CBCTs)15.3 Common features of a voltageand a current transformer15.3.1 Design parameters (service conditionsand likely deratings)The\e are similar to parameters for a switchgear assemblyas discussed in Section 13.4. Since they are directly* Potential transformer (PT) is not the appropriate word to identifyan instrument voltage transformer.associated with the same power system and interruptingdevices as a switchgear assembly, they should generallymeet the requirements for a switchgear assembly. exceptfor small variations in the test requirements. For moredetails refer to the following publications:1 For voltage transformersIEC 60044-2 and IEC 60186 (for two-windingtransformers such as CVTs)2 For current transformersIEC 60044- 1 and IEC 60044-6SECTION I: VOLTAGETRANSFORMERS15.4 General specifications anddesign considerations forvoltage transformers (VTs)These transformers develop a voltage on the secondary.substantially proportional to the voltage on the primary(there being no knee point saturation. as is sometimesrequired in CTs (Section 15.6.l(viii)).15.4.1 Instrument voltage transformers1 Rated primun voltageThis will generally be the nominal sy\tem voltage, exceptfor transformers connected between a phase and the groundor between the neutral and the ground, when the primaryvoltage will be considered as I/? 3 times the nominalsystems voltage (V,)2 Rated secondary voltageInEurope and Asian nations this is generally 110 or 1 101y;3 V, (63.5 V) for phase-to-phase or phase-to-groundauxiliary circuits respectively. In @e USA and Canadathese voltages are 120 oy_ 120/ % 3 V for distributionsystems and 1 IS or 1 IS/ %'3 V for transmission systems.3 Rated frequeizcxThis may be 50 or 60 Hz as the system may require. Thepermissible variation may be considered as +2% formeasuring as well as protection VTs. These limits arebased on the recommended variations applicable for aswitchgear assembly (IEC 60439- 1 ) or an electric motor(Section 1.6.2).4 Insulation systemsThese transformers may be PVC taped. thermoplastic(polypropylene) moulded, fibreglass taped. polyester resincast or epoxy resin cast depending upon the system voltageand the surroundings. HT indoor transformers, for instance,are generally polyester or epoxy resin cast, and areeconomical with good dielectric properties. They areresistant to humid, chemically contaminated and hazardousareas. Outdoor HT transformers, however. may be epoxy

15/458 Industrial Power Engineering and Applications Handbookresin cast, oil or SF, insulated and oil or SF6 cooled.Epoxy insulation provides better mechanical and constructionalqualities. They are resistant to humid, contaminatedand corrosive atmospheres and are suitable for all HTsystems. They are mechanically strong and can bear shocksand impacts.5 Creepage distancesFor outdoor installations the recommended minimumcreep distances for all types of voltage or currenttransformers are given in Table 15. I , according to IEC60044-1 or IEC 60044-2.6 TappingsTappings are generally not necessary, as a transformer isdesigned for a particular voltage system. If and whensuch a need arises (as in a control transformer (Section15.4.5)) they can be provided on the primary side of thetransformer.7 Rated outputThe standard ratings, at 0.8 p.f. lagging, may be 10, 15,25, 30, 50, 75, 100, 150, 200, 300, 400 or 500 VAor asthe auxiliary circuit may demand. The procedure todetermine the total VA burden of a circuit is described inSection 15.4.5. Typical values of VA burden for a fewinstruments are given in Table 15.2 from data providedby the manufacturers.8 Rated burdenThis is the maximum burden the transformer may have toTable 15.1 Recommended values of minimum creepagedistances for a VT or a CTPollution levelLight 16Medium 20Heavy 25Very Heavy 31Table 15.2InstrumentsMinimum creepage distance between phaseand ground mm per kV(rm.s.)(phase to phase)Typical values of VA burdensI Voltmeter 52 Voltage coil of a watt-meter or a powerfactor meter 53 Voltage coil of a frequency meter 7.54 Voltage coil of a kWh or a kVAr meter5 Recording voltmeters7.556 Voltage coils of recording watt-meters andpower factor meters 7.57 Voltage coil of a synchroscope 15Maximum burdenf W ""These VA burdens are for moving iron instruments. For electronicmeters these values would be of the order of 0.1 to 0.5 VA.feed at a time. The preferred values will follow series R-10ofISO-3 (IEC60059) andasnotedinSection 13.4.1 (4).9 Short-time ratingThis is not material in voltage transformers, as neitherthe voltage measuring instruments nor the protective relayswill carry any inrush current during a switching operationor a fault. No short-time rating is thus assigned to suchtransformers.The electromagnetic unit, however, as used in a residualVT (Section 15.4.3) or a capacitor VT (Section 15.4.4)should be suitable for carrying the heavy discharge orinrush currents during a capacitor discharge or switchingrespectively.10 Accuracy classThe accuracy of a VT depends upon its leakage reactanceand the winding resistance. It determines the voltageand the phase errors of a transformer and varies with theVA on the secondary side. With the use of better corematerial (for permeability) (Section 1.9) and better heatdissipation, one can limit the excitation current and reducethe error. A belter core lamination can reduce the coresize and improve heat dissipation.Measuring voltage transformers Standard accuracyclass may be one of 0.1, 0.2, 0.5, 1 or 3. The recommendedclass of accuracy will depend upon the typeof metering and generally as noted in Table 15.3.Protection voltage transformers Generally, a measuringvoltage transformer may also be used for thepurpose of protection. A protection transformer, however,is assigned an accuracy class of 3 or 6, which definesthe highest permissible percentage voltage error at anyvoltage between 5% of the rated voltage up to thevoltage obtained by multiplying the rated voltage bythe rated voltage factor of 1.2, 1 .5 or 1.9. And whenthe secondary has a burden between 25% and 100% ofthe rated burden at a p.f. of0.8 lagging. This accuracyclass is followed by a letter 'P' such as 3P and 6P etc.The voltage and phase displacement errors should notexceed the values noted in Table 15.6.Table 15.3 Recommended class of accuracy for differenttypes of metersApplication1 Precision testing or as a standard 0.1VT for the testing of other VTs2 Meters of precision grade 0.53 Commercial and Industrial meters 1 .oClass of accuracy4 Precision Industrial meters (Indicating 0.2 or 0.5instruments, recorders and electronicintegrating meters)5 General industrial measurements 1 or 3(Indicating instruments and recorders)6 Purposes where the ratio is of less 3importanceNote To choose a higher class of accuracy than necessary is notdesirable.

Instrument and control transformers: application and selection 15/459Note.$I A low Loltage of 596, at which the transformer is required tomaintain its accuracy limit, is of great significance. A protectiontransformer is required to operate under a fault condition, duringwhich the primary voltage may dip to a value as low as 5% ofthe rated voltage.2 It is possible to have two windings in the secondary circuit ofa VT when it is required to perform the functions of bothmeasurement and protection.I1 Rated voltage factorThis is the multiplying factor which, when applied to therated primary voltage, will determine the maximumvoltage at which the transformer will comply with thethermal requirements for a specified time as well as withthe relevant accuracy requirements. This factor carries agreater significance, particularly on a fault, when healthyphases may experience an overvoltage and the protectionVTs may all the more be required to accurately sensethis and activate the protective circuit. Such a situationmay arise on a ground fault on an isolated neutral systemor a high impedance grounded system (Section 20.6).Table 15.4, following IEC 60044-2, suggests therecommended voltage factors and their permissibledurations for different grounding conditions.12 Circuit diagramTo illustrate the important features of a VT, let us analyseits equivalent circuit diagram. Refer to a simple diagramas in Figure 15.1 which is drawn along similar lines tothose for a motor (Section 1.10, Figure 1.15). For easeof analysis, the ratio of primary and secondary turns hasbeen considered as 1 : I . Then from the circuit diagram,the following can be derived:and this is drawn in the form of a phasor diagram (Figure15.2). The phase displacement between phasors andVi is the phase displacement error '6' as discussed later.13 Voltage error or ratio errorThis is the error in the transformed secondary voltage asgenerally caused by the excitation current I,, and as shownin Figure 15. 1. It is the variation in the actual transformationratio from the rated and is expressed byK, V,* - V,*Voltage error = x 100%whereK, = rated transformation ratioVI = actual primary voltage (r.m.s.)V, = actual secondary voltage (r.m.s.)NO I e*Only the r.rn.s. values and not the phasor quantities are consideredto define the voltage error. The phase error is defined separately.Together they form the composite error. Refer to Table 15.5 forrneawring and Table 15.6 for protection VTs.VIVI - Primary voltageV,' - Primary induced emfVi - Secondary induced emf referred to the primary side(V, being secondary induced emf not shown)In( - Excitation or no-load current/, - Loss component of current supplying the hysteresis and eddycurrent losses to the voltage transformer core (it is the activecomponent)/, - Magnetizing component producing the flux '0' (it is the reactivecomponent)R, - Primary circuit resistanceRg - Secondary circuit resistance referred to the primary sideX, - Primary circuit reactanceX; - Secondary circuit reactance referred to the primary sideZ, - Primary circuit impedanceZg - Secondary circuit impedance referred to the primary sideZ- Load (burden) impedanceFigure 15.1 Equivalent circuit diagram for a voltage transformer14 Phase displacement errol; 6This is the difference in phase between the primary andthe secondary voltage phasors (6). The direction of thephasors are so chosen, that the angle is zero for a perfecttransformation. Refer to the phasor diagram, Figure 15.2,and Table 15.5 for measuring and Table 15.6 for protectionVTs .15 Limits of voltage and phase displacement errorsAt rated frequency, these should not exceed the valuesgiven in Table 15.5 for measuring VTs. at any voltagebetween 80% and 120% of the rated voltage and aburden of 25-100% of the rated burden at a p.f. 0.8lagging.For protection VTs these should not exceed the valuesgiven in Table 15.6 at any voltage between 5% of therated, up to the voltages obtained by multiplying therated voltage by the rated voltage factor as in Table15.4, and a burden between 25% and 100% of therated load at a p.f. 0.8 lagging. At voltages lower than5% of the rated, the limits of error may increasedisproportionately and become up to twice the specifiederrors at about 2% of the rated voltage. the limits ofVA burden and p.f. remaining the same.15.4.2 Electromagnetic voltage TransformersThese are single-, double- or three-phase wound-typetrans-formers with windings on both primary andsecondary sides (Figures 15.3(a) and (b)).

~~ ~ ~ ~-~15/460 Industrial Power Engineering and Applications HandbookVl6 = Phasor displacement errorNote The phasor diagram is drawn taking the applied voltage VI as the reference phasor. It can also be drawntaking the primary induced emf V,' as the reference. The logic to the diagram and the subsequent results shallhowever, remain the same.Figure 15.2Phasor diagram for a voltage transformerTable 15.4 Rated voltage factorsSr: no. Rated voltage factor Rated time Method .f system grounding Method of primary connections1 1.2 Continuous All types of system grounding (i) Between lines or(ii) Between transformerstar point and ground2 1.2 Continuous An effectively grounded system Between line and ground3 1.5 30 seconds An effectively grounded system Between line and ground4 1.2 Continuous An ineffectively grounded system Between line and ground5 1.9 30 seconds An ineffectively grounded system Between line and ground6 1.2 Continuous (i) An isolated neutral system or Between line and ground(ii) A resonant grounded system7 1.9 8 hours A resonant grounded system Between line and ground-. ~~Table 15.5 Recommended limits of voltage and phasedisplacement errors, applicable for all types of measuring VTs(only electromagnetic and capacitor VTs).(A residual VT is basically a protection VT)Class of accuracy0.1 0.10.2 0.20.51 .Q0.51 .o3.0 3.0As in IEC-60044-26 voltage (ratio)error faPhase displacement (6)_+ minutey-~ ~5102040Not specified"These errors are valid only when the voltage is between 80% and120%, burden 25-1006 of the rated burden and p.f., 0.8 lagging.ApplicationThey are used for both measuring and protection purposes.As a measuring VT, they are used to feed a voltmeter,kW, kWh or a kVAr meter, a power factor, frequencymeter or a synchroscope. As a protection VT they areused to feed a protective circuit, incorporating voltagesensing protection relays. To save on cost and mountingspace, they may also be wound for one common primaryand two secondary windings, one for metering and theother for protection. For markings, see Section 15.10.1(2)and Figure 15.35.15.4.3 Residual voltage transformers (RVTs)When the primary of a three-phase two-windingtransformer, having its secondary wound for a threephaseopen delta, is connected across an unbalanced supplysystem, a residual voltage across the open delta will appear.This is the principle on which this transformer is based(Figure 15.4(a)). As discussed in Section 21.2.2, andillustrated in Figure 21.7, the phasor sum of all the threeline to ground voltages in a three-phase balanced systemis zero, i.e.v, + v, + v, = 0When this balance is disturbed, due to either an unbalancein the loads or due to a ground fault, a residual or zerophase sequence voltage in the neutral circuit will appear.When one of the phases in the secondary of a threephasetransformer is open circuited and a three-phasesupply is applied to its primary windings, there will appear

Instrument and control transformers: application and selection 15/461Transformers with HT fuseFigure 15.3(a) Typical indoor epoxy resin cast instrument voltage transformers up to 11 kV (Courtesy: Kappa Electricals)Table 15.6 Recommended limits of voltage and phasedisplacement errors, applicable for all types of protection VTs(electromagnetic, capacitor and residual VTs)Class of accuracy % voltage (ratio) Phase displacement (6)error?fminutes3P6P3.0 1206.0 240ta(b) 11 kV three-phase indoor(a) 33 kV single-phase outdoorFigure 15.3(b) Typical HT instrument voltage transformers9As per IEC-60044-2aThese errors are valid only when the-voltage is between 5% to‘rated voltage factor x loo%’, burden 25-100% of the rated burden,and p.f., 0.8 lagging.At voltages lower than 5%, the limits of error may increase. Theybecome up to twice the specified errors at about 2% of the ratedvoltage, the limits of VA burden and p.f. remaining the same.NoteThe choice of class 3P or 6P will depend upon the application andthe protection scheme of the system. The following may be consideredas a rule of thumb when making this choice.a residual or zero phase sequence voltage across theopen terminals at the secondary. This represents theresidual or the zero phase sequence voltage, whatevermay exist in the main supply system. This voltage willbe zero when the main primary system is balanced andhealthy.Important parameters1 Residual voltage The residual voltage appearingacross the secondary windings will be three times thezero sequence voltage, if it is found in the primary(i) Class 3PThis class may be selected for protective devices that operateon the basis of phase relationship between the voltage and thecurrent phasors, such as in a directional overcurrent protection,reverse power or directional distance protection.(ii) Class 6PThis class may be selected for protective devices where theiroperation does not depend upon the phase relationship betweenthe voltage and the current phasors, such as for an overvoltage,overcurrent or an undervoltage protection. For instance, aresidual VT should have this accuracy class.(iii) When a residual VT is employed for capacitor discharges itrequires no accuracy class.

15/462 Industrial Power Engineering and Applications HandbookR Y B1 1 1=GHealthy systemIn this case, all the three phases would be balanced andthe residual voltage, Ve, will be zero. The three voltagephasors in the open delta windings will be as illustratedin Figure 15.5(a). The phasor sum of these phasors iszero. Therefore V, = 0.Figure 15.4(a)voltageConnections of a residual voltage transformer5Ground fault on one phaseSystem neutral groundedConsider a ground fault on phase R. The voltage acrossthis phase will become zero and the phasor diagram willbe as illustrated in Figure 15.5(b). The other two phasorswill remain the same as in a healthy system and add togive the residual voltage V,, i.e.v, = JVT' + v; + 2.V,VT cos 120"where VT is the phase voltage across the secondarywindings=2/2v:-vT2=vT= 3 x zero phase sequence voltage drop.The voltage across open delta is thus the same as the fallin the voltage of the faulty phase. It will lead the currentcaused by the ground circuit impedance.Figure 15.4(b)A five-limb transformer to carry unbalanced fluxwindings. This is due to an open magnetic circuit in thesecondary open delta winding having no return paththrough the third magnetic limb. This phenomenon doesnot exist when three single-phase transformers are used,as each transformer core winding will form a closedmagnetic circuit of its own.In a normal three-limb transformer the resultant flux,on a ground fault, of the two healthy lines limb willreturn through the transformer limb of the grounded line,inducing a heavy short-circuit current in that winding.This will induce a voltage which will be reflected in thecorresponding secondary winding, and the voltage acrossthe open terminals of the delta will not be a true residual.This situation is overcome by providing a low reluctancereturn path, suitable for carrying the maximum value ofunbalanced flux without saturation. This is achieved bythe use of a five-limb transformer (Figure 15.4(b)). Thetwo additional outer limb are left unwound.2 Residual voltages under different operating conditionsTo extend the ease of application of this device,consider the following circuit conditions to determinethe quantum of residual voltages:Healthy system: System neutral, grounded or ungrounded.Ground fault on one phasc- System neutral grounded- System neutral isolatedSystem neutral isolatedWhen the system neutral is isolated, the voltage acrossthe faulty phase R will be the same as the ground potentialand the ground potential will become equal to the phasevoltage VT as illustrated in Figure 15.5(c). The voltageacross the healthy phases will become 8VT, i.e. &times more than the normal phase voltage. The phasorsV, and V, will thus be 60" apart than 120" and 180"from the primary..'. v, =J(&~T)'+(&~T)*+= J3v; + 3v; + 3v;= 3vTi.e. three times the healthy phase voltage.2~&v~.&v~~COS60"Important requirementsGrounding Based on the above, it is essential thatthe primary windings of the transformer have a groundedneutral, without which no zero sequence exciting currentwill flow through the primary windings. Although theopen delta will develop some voltage on an unbalancein the primary, it will only be the third harmoniccomponent, as would be contained by the primarywindings' magnetic flux and not the zero sequencecomponent.Voltage factor Since this transformer may have toperform under severe fault conditions, it should besuitable for sustaining system switching surges as wellas surges developed on a fault. A voltage factor ashigh as 1.9 (Table 15.4) is generally prescribed forthese transformers.

Instrument and control transformers: application and selection 15/463jRY B YTertiarywindingV&windingTertiarywindinga- -yvTmN'R"CV,=Ov, = VT ISecondary windingIC'vs Y Y - BSecondarywinding vs1 R" I y" 1 8"1R"'Figure 15.5(a) An RVT in a healthy system Figure 15.5(b) An RVT under ground fault on a 34four-wiregrounded neutral system= &vpTertiarywindinga' a"1_ ve = 3vT cN'I I 1Secondary [ 11; 113 mwinding 7= 3v,V'3.VT€3"'{3. vsF- - - - -3,Figure 15.5(c) Ground fault on aY3-44 three-wire delta or 34, four-N \'3 vs wire ungrounded star system

15/464 Industrial Power Engineering and Applications HandbookShort-time duty When this transformer is requiredto discharge a charged capacitor bank it should becapable of withstanding heavy inrush discharge currents(see below for its application). The following may beconsidered when designing a transformer for dischargingpurposes:1 The size of the capacitor banks, their voltage andthe impedance of the capacitor circuit.2 Rate of discharge of the trapped charge.3 The temperature of the primary windings after thedischarge.4 The magnitude of electrodynamic forces on theprimary windings, which may be developed by thedischarge currents.Applications They may be used to carry out thefollowing functions:1 To detect a ground fault or operate a directionalground fault relay (Section 21.7.4).2 To operate a neutral displacement relay (Figures26.4 and 26.9).3 To detect an unbalance in a three-phase normallybalanced capacitor bank (Section 26.1.1(8)).4 To discharge a charged capacitor bank over a veryshort period, particularly when a fully chargedcapacitor is interrupted. See also Section 25.7.5 To discharge an interrupted HT circuit before a reclosing.An HT system, say, a transmission line ora cable when interrupted, develops high transientvoltages as discussed in Sections 23.5.1 and 20.1.Unless these transients are damped to a reasonablylow level so that they are not able to endanger theconnected devices on an automatic reclosing, thedevices installed in the system may become damageddue to the resulting switching transients. The normalpractice to deal with this is to damp the transientsthrough an electromagnetic transformer such as this.The transformer, however, may have to be designedfor such a duty to sustain the electrostatic stressesarising from such transients, the discharge time andimpedance of the interrupting circuit up to thetransformer.15.4.4 Capacitor voltage transformers (CVTs)This type of voltage transformer is normally meant for ahigh to an ultra-high voltage system, say, 110 kV andabove. While a conventional wound-type (electromagnetic)voltage transformer will always be the firstchoice, it may become costlier and highly uneconomicalat such voltages.The size and therefore the cost of a conventionallywound voltage transformer will be almost proportionalto the system voltage for which it is wound. As a costconsideration, therefore, a more economical alternativeis found in a Capacitor Voltage Transformer (CVT) (Figure15.6(a)).A CVT consists of a capacitor divider unit in which aprimary capacitor C, and a secondary capacitor C2 areconnected in parallel between the line and the ground(Figure 15.6(b)). A tapping at point A is provided at anintermediate voltage VI, usually around 12 to 24 kV,which is reasonably low compared to the high systemvoltage. This helps restrict the phase error, on the onehand, and facilitates an economical intermediate woundtransformer T,, on the other, to perform the same duty asa normal wound voltage transformer. The purpose ofline capacitors is thus to step down the high to very highsystem voltages to an economically low value. Throughthe tapping point A is connected a conventional and aless expensive wound-type intermediate or auxiliaryvoltage transformer T,, rated for the intermediate voltageVI in association with a reactor L (Figure 15.6(b)). Theuse of the reactor is to almost offset the heavy capacitivevoltage component. If possible, the reactor and thetransformer may be combined in one unit to make iteasier to operate. The secondary of the transformer israted for the required standard voltage, say, 110/

Instrument and control transformers: application and selection 15/465Primary voltageterminalEHV lineCapacitordivider unitIntermediate 4 L1voltage terminal 5 IT hCIc,LSAZR1s1-1s22s1-2s23S1-352TrLine voltageIntermediate voltagec1= ve.Primary line capacitanceSecondary line capacitanceVariable tuning inductanceSurge suppressorLoad (burden) impedanceDamping resistor to preventferro-resonance effectsSecondary tappingsConventionally woundintermediate VT(electromagnetic unit, EMU)v,/d3Figure 15.6(a) Capacitorvoltage transformer (CVT)rated voltage 36-420 kV andabove (Courtesy: ABB)i 1'G331GRSecondaryvoltageSecondaryvoltageFigure 15.6(b) Schematic diagram of a basic capacitor voltage transformer (CVT)NoteFerroresonance: This phenomenon may occur in an isolated neutralsystem employing a CVT, similar to an RVT (Section 20.2.1). Thecore of the non-linear electromagnetic unit (EMU) may saturatemomentarily during a ground fault or even during a healthy operation,under certain circuit conditions. For instance, low-frequencytransients or a fault on the secondary side may cause momentarysaturation of the magnetic core of the EMU, which may, in turn,resonate with the ground capacitive reactances Xcs and give rise tosub-harmonic oscillations. These may be detrimental to the insulationof the EMU of the CVT as well as the terminal instrument anddevices connected to it. These oscillations must be damped as faras possible. This can be achieved by inserting a damping resistanceR in the EMU circuit, as illustrated in Figure 15.6(b).ApplicationA CVT may be used to carry Out the functions:1 To measure as well as protect a high-voltage system,generally 110 kV and above. To save on cost andmounting space, the electromagnetic unit may bewound for two secondary windings, one for meteringand the other for protection.

15/466 Industrial Power Engineering and Applications HandbookV1/d3ZRDuring resonance, when:X, = X, the circuit would behave like a normal transformerV, : Voltage drop across the line capacitorsV, : Voltage drop across the inductanceRl : Primary resistance representing losses across '@ and 'L' andthe intermediate voltage transformer (EMU)R;: Secondary resistance of the intermediate VT referred to theprimary side./A : Loss componentIrn : Magnetizing componentZ: Load (burden) impedanceR : Damping resistor to prevent ferro-resonance effectsFigure 15.7 Equivalent circuit diagram of a CVT2 To feed the synchronizing equipment.3 As a coupling unit for carrier signals (Section 23.5.2and Figure 23.9(b)).4 To damp the transient voltages on the primary side.For markings refer to Section 15.10.1 and Figure 15.35.15.4.5 Control transformersRefer to Figures 15.8 and 15.9. These transformers arequite different from a measuring or a protectiontransformer, particularly in terms of accuracy and shorttimeVA ratings. They are installed to feed power to thecontrol or the auxiliary deviceslcomponents of atsFigure 15.9 A typical outdoor type oil-filled 11 kV controltransformerswitchgear or a controlgear assembly not supposed to beconnected directly to the main supply. These transformersdo not require a high accuracy and can be specified bythe following parameters:Rated primary voltage The normal practice for anHT system is to provide a separate LT feeder for theauxiliary supplies. The primary voltage will be thenormal system voltage, V,., when the transformer isconnected line to line or Vr1a when connected lineto neutral.Rated secondary voltage This is 24,48, 11 0, 220,230, 240 or 250 volts, or according to the practice ofa country. Tappings, if required, can be provided onthe primary side.Rated burden This is the maximum burden thetransformer may have to feed at a time. The preferredratings will follow series R-10 of ISO-3 (Section13.4.1(4)).Short-time VA burden This accounts for thebnFigure 15.8 Typical single-phase and three-phase control transformers

Instrument and control transformers: application and selection 15/4675maximum switching inrush VA burden of the variousauxiliary devices connected in the switching circuitsuch as contactors, timers and indicating lights. Unlessspecified, the short-time VA burden of the transformerwill be a minimum of eight times its rating at 0.5 p.f.lagging. It can be expressed in terms of VA versuscos 4 and drawn in the form of an inrush curve, foreasy selection of a transformer rating (Figure 15. IO).Voltage regulation In view of heavy currents duringthe switching of an auxiliary circuit, the reactanceand the resistance drops of these transformers shouldbe designed to be low to ensure a high degree ofregulation during a switching operation. Regulationof up to 6% for control transformers rated for 1.0kVA and above and up to 10% for smaller ratings isconsidered ideal (NEMA Standard suggests thesevalues as 5%).For brevity, only the more relevant aspects arediscussed here. For more details, refer to IEC 60044-2 and IEC 601 86, for instrument voltage transformersand IEC 60076-3 for control transformers.ApplicationThese may be used to feed the solenoid or the motor ofan interrupting device (such as an electrically operatedbreaker), indicating lights and circuits, auxiliary contactorsor relays, electrical or electronic timers, hooters or buzzers,and all such auxiliary components and devices mountedon a controlgear or a switchgear assembly requiring aspecified control voltage.Procedure to determine the VA rating ofa control circuitThe total VA level of a control or an auxiliary circuit isthe phasor sum of the VA burdens of each individualcomponent and device connected in the circuit, andconsuming power. It is advisable to add the VA burdensvectorially rather than algebraically. Since a controltransformer may feed more auxiliary components anddevices consuming power compared to an instrumentVT, the VA rating of such transformers is generally higherthan of an instrument, metering or protection VT.Algebraic summation will lead to a higher VArequirement than necessary. The transformer should notbe too small or too large to achieve better regulation inaddition to cost. From Figure 15.11 the following maybe derived:orVAr=W+VArVAT = ,/W2 + VAr’where VAT = Total VA burdenVA = VA burden of individual componentandw= w, + w, + ...VAr = VAr, + VAr2 +. , .W,, W,, VAr, and VAr, are the active and reactivecomponents respectively of the VA burden of a device ata p.f. 4, and &.The following may be ascertained when selecting therating of a control transformer:Maximum hold-on (continuous) VA burden and thecorresponding p.f. of all the devices likely to be inservice at a time.Pick-up VA or short-time VA: An electromagnetic devicesuch as a contactor or a timer carries a high inrushcurrent, also known as ‘sealed amperes’, during aswitching operation and it is associated with a highmomentary pick-up VA burden on the circuit and thefeeding control transformer. The effect of the maximummomentary pick-up VA burden and the correspondinginflow p.f. of all the components likely to bc switchedat a time must be calculated.Maximum lead burden of the connecting wires underthe above conditions.The control transformer to be selected may have arating nearest to the maximum hold-on VA burden socalculated and must be suitable to feed the required inrushcurrent at the p.f. so calculated without affecting itsregulation. So long as these two points fall below theinrush curve of the control transformer, its regulation-0 01 02 03 04 05 06 07 08 09 10Cos I$ (of control circuit)Figure 15.10 Inrush characteristics of a control transformerFigure 15.11Phasor representation of a Load (VA burden).

~~15/468 Industrial Power Engineering and Applications Handbookwill be maintained within the prescribed limits. Figure15.10 illustrates this requirement and Example 15.1demonstrates the procedure to determine the required VAof a control transformer.Example 15.1Consider the control scheme of an auto-control capacitorpanel as shown in Figure 23.37. The scheme shows thecontrol voltage as being tapped from the main bus. But forour purpose, we have considered it through a controltransformer 41 511 10 V.The following data have been assumed:System: 415 V, three-phase, four wire.Control voltage: 11 0 V a.c.Control wire: 2.5 mm2 (resistance of wire = 7.6 Q/lOOO m, asin Table 13.15).Approximate length of wire for each feeder up to the powerfactor correction relay (PFCR): 35 m.(1) A study of the control schemeComponent Total quantity nos Maximum hold-on occurs whenall the six steps of the PFCRare ONMaximum inrush occurs whenfive steps ofthe PFCR are ON and the sixth isswitched ONMain contactor 125 A 6Auxiliary contactor 6 A 2On indicating light 8PFCR 161 (auto or manual)6 (auto or manual)1 (6 steps)-Hold on(2) Approximate VA burden and cos f$ for each component, as available from the manufacturers’ catalogues5155Inrush1-11-Componenr VA cos @ W = VA cos @ VAr = VA sin @125 A contactorHold-on 65Inrush 9006A contactorHold-on 15Inrush 115Indicating lightHold-on 7Inrush 7PFCR (each step)aH o 1 d - o n 50.3 I0.420.330.60I1120.15 61.79378 8174.95 14.1669 92775--aPFCRs are available in both static and electromagnetic versions. Their VA levels therefore vary significantly due to inbuilt switchingrelays, LEDs (light emitting diodes) and p.f. meter etc. For illustration we have Considered an average VA burden of 5VA at unity p.f. foreach step. For static relays, this may be too low(3) Computing the maximum hold-on (steady state) and inrush burden values and their cos f$Maximum hold-on values 1 Maximum pick-up (inrush) valuesComponentQtyHold-on for five steps already ON1Inrush for the sixth stepTotal Total Qty Total Total Qty , Total , TotalW VA r W VA r I W VArMain contactor 6Auxiliary contactor 1Indicating light 6PFCR (steps) 66 x 20.15 6 x61.79 5 5 x 20.15 5 x 61.79 1 378 817= 120.90 = 370.74 = 100.75 = 308.954.95 14.16 I 4.95 14.16 - - -6~7=42 - 5 5x7=35 - 1 7 -6X5=30 - 5 5X5=25 - 1 5 -Total197.85 384.90 165.70 (a) , 323.11(b) ~1 390(cj 1 817(d):. VA = J197.852 + 384.90’= 433 without considering theburden for wire leads:. Total inrush, W = 555.7 ( a + c)and VAr = 1140.11 (b + d):. VA= 4555.7’ + 1140.11’= 1268 without consideringthe burden for wire leads

~WCorrrrol c.irc.ciit currmfIL = 43311 IO = 3.94Aand lead burden = IL’ . Kwhere R is the resistance of the connecting wires at theoperating temperature (90°C, as in Table 14.5)= 676X 35 X -[I + 3.93 x IO-’ (90-20)1I000(for details refer to Table 14.4)= 2.035 R:.Lead burden = 3.94’ x 2.035= 31.59 W:. Total maximum steady-state hold-on burdenI+’= 197.85 + 31.59 = 229.44And VAr = 384.9:.Maximum VA = 4229.44’ + 384.9’= 448. 1229.44at a steadq-\tate cos d = -448. I0= 0.51(2 ~iIIl1~ IInstrument and control transformers: application and selection 15/469Control circuit currentI, = 1268/110 = 11.53Aand lead burden = 1 I .53’ x 2.035= 270.53 W:. Total maximum inrush burden= 555.7 + 270.53= 826.23and VAr = l14n.1 I:. Maximum short-time VA = d826.23’ + 1140.1 I’= 1408.0at an inrush (short-time) cos 6 = 826.231408.0= 0.587Rating of control transformerSelect a continuous rating = 500 VAat a COS 4 = 0.51and short-time rating = 1500 VAat a coyb = 0.5X7The actual values a5 worked out above must fall belowthe inrush curve of the selected control transformer of500 VA, as illustrated in Figure 15.12.15.4.3 Summary of specifications of a VTIn Table 15.7 we list the data that a user must provide toa manufacturer to design a VT for a particular application.Some of the data chosen are arbitrary to define the specifications.15.5 Precautions to be observedwhile installing a voltagetransformerI Since a VT forms an inductive circuit, it generatesheavy switching current surges which should be takeninto account when deciding on protective fubeb. Afuse with an appropriately high rating should be chosento avoid a blow-up during switching.2 As a result of the generally high rating of protectivefuses they provide no adequate protection against aninter-turn fault. For critical installations, and for HTVTs particularly, a separate protection may be providedfor inter-turn faults.3 When an HT-VT develops an inter-turn fault on theHV side, there is no appreciable rise in the primarycircuit current and which may not be detected. But, it-00 587cos @(of control circuit)Figure 15.12 Checking the suitability of the 500 VA controltransformer for the required duty for Example 15 1may cause local discharge and heating up of the interturns,leading to dangerous fault currents and ionizationof oil in an oil-filled VT. It is advisable to provide aBucholtz relay to protect oil-filled VTs by detectingthe presence of gas in the cvcnt of a fault.4 For lower voltages (c 33 kV), any fault on the VTwill be detected by the protective devices installed inthe main circuit.5 Temperature detectors may also be provided in thewindings of large VTs as are provided in a motor(Section 12.8).

~~~~15/470 Industrial Power Engineering and Applications HandbookTable 15.7 Summary of specifications of VTsSr: no. SpecifcationsMeasuring VTs Protection VTs Control transformers--1 System voltageAs in Table 13.1 (I/& V for line to neutral transformers)2 Insulation level (peak) Generally as in IEC 60044-2 or Tables 13.2, and 14.3 for series I and Tables 14.1 and 14.2 forseries I1 voltage systems345Class of insulation say,FrequencyNominal voltage ratioE50 or 60 HzB50 or 60 HzE50 or 60 Hz-6I8910~11-12-~Output (VA) say,Short-time output (VAburden)and corresponding p.f.Class of accuracyGrounding conditionsThe rated voltage factor andthe corresponding rated timeService conditionse.g. 6.6 kV/l IO V for two phase or three phase transformers and 1 times this for line to neutraltransformers 36500 500 1500- - Say, 8 times the rated VA at0.2 p.f.0.1, 0.2, 0.5, 1 or 33P or 6PNot applicableWhether an isolated neutral system, an effectively grounded system or a non-effectively groundedneutral systemNot applicableIndoors or outdoorsWhether the system iselectrically exposedAmbient temperatureAltitude, if above 1000 mHumidityAny other important featuresDepends upon the system Depends upon the system faultgrounding. Refer to Table 15.4 conditions and generally as 1.9Table 15.4Indoors or outdoorsIndoors or outdoors0 Whether the system is Whether the system iselectrically exposedelectrically exposedAmbient temperature Ambient temperature0 Altitude, if above 1000 m Altitude, if above 1000 m0 Humidity HumidityAny other important features Any other importantfeaturesMarking of VTs (a) 6.6 kV/lIO V, 500 VA, 6.6 kV/110 V, 500 VA, 6.6 kV/48 V, 1500 VA andclass I" class 3P short-time VA: eight times therated VA at 0.2 p.f.System voltage and insulation level, class of insulation and frequency etc. for all types ofVTsaWherever two separated secondary windings are provided, say, one for measuring and the other for protection, the markings will indicateall such details as are marked against (a) for each secondary winding.For more details and voltages higher than 66 kV, refer to IEC 60044-2.SECTION II: CURRENTTRANSFORMERS15.6 Current transformers (CTs)These may be one of the following types:Ring (Figures 15.13(a)-(d))Bar primary (Figures 15.14(a) and (b) and 15.15(a)and (b))Wound primary (Figures 15.15(a) and 15.16)In ring-type CTs the primary current-carrying conductoris passed through the ring and the ring forms the secondarywinding (Figure 15. 13). Generally, all LT CTs are producedthus, except for very small ratings of the primary current,say, up to 50 A, when it becomes imperative to designthem in the form of a wound primary due to the designconstraints discussed in Section 15.6.5(v).For special applications, such as for a motor protectionovercurrent release (Figure 12.15) where the use of aring-type CT may appear crude and the connectionscumbersome, a bar primary CT may be used, as shownin Figure 15.14. HT CTs, as a matter of necessity and tomaintain correct clearances and dielectric strength betweenthe primary current-carrying conductor and the secondarywindings, are made in the form of a bar primary or woundprimary only, depending upon the primary current rating(Figures 15.15 and 15.16). Inawoundprimary, theprimaryis also wound the same way as the secondary.Types of insulationAn LT CT may be insulated in the following ways,depending upon their location and application.Tape insulated (Figure 15.13(c)) For normalapplication and generally clean atmospheric conditions.Epoxy resin cast (Figures 15.13(a) and 15.14) To

Instrument and control transformers: application and selection 15/471(a) LT epoxy resin cast(b) LT fibreglass taped (c) taped LT PVC (d) Ground leakage relay with CBCTFigure 15.13Ring type CTs for measuring or protection (Courtesy: Prayog Electricals)provide greater mechanical strength and a betterinsulation system. They are more suitable for humid,contaminated and corrosive atmospheres and for allHT systems. They are mechanically strong and canbear shocks and impacts.NoteAll HT CTs are normally manufactured with epoxy resin cast.Fibreglass tape (Figure 15.13(b)).compact.PolypropyleneTo make it more15.6.1 General specifications and designconsiderations for current transformersThe rated frequency, insulation systems and therequirement of creepage distances will generally remainthe same as for a voltage transformer (Section 15.4.1).For the remaining parameters, the following may benoted.(i) Rated secondary currentThis will be 1,2 or 5 A, 1 and 5 A being more common.All measuring and protection devices are alsomanufactured for these ratings only as standard practice.A 1 A secondary is not recommended in higher ratiosdue to increased induced voltage on the secondary sideduring an accidental open circuit on load. It may damagethe inter-turn insulation or cause a flashover, besidesbeing dangerous to nearby components or a human body,if in contact. This we have illustrated in Example 15.8,under section 15.9.(ii) Rated outputThe standardVAvalues may be one of2.5,5,7.5, 10, 15and 30 generally, depending upon the application, althougha value beyond 30 VA is also acceptable.(iii) Rated burdenThis is the value of the impedance of the secondary

15/472 Industrial Power Engineering and Applications Handbook(a) Single-phase HT CT11 kVCT 33 kV CTFigure 15.15(a) Typical outdoor-type oil-filled bar primary orwound primary HV CTs (Courtesy: Kappa Electricals)(b) Three-phase LT CTFigure 15.14 Typical indoor-type bar primary epoxy resincast CTs (Courtesy: Kappa Electricals)circuit (the impedance of all the devices connected to it),expressed in ohms and power-factor or volt-amperes, atthe rated secondary current. The CTs will be selectednearest to the computed total VA burden in the circuit. ACT with a higher VA burden than connected will have aslightly higher error besides size. A slightly less VA ratingthan the connected may normally be permissible, subjectto confirmation by the manufacturer. Typical valuesof VA burdens at the rated current for the devices thata measuring CT may have to usually feed are asfollows:Instruments and measuring devices*Moving iron ammetersRecording ammetersCurrent coils of watt-metersRecording watt-meterskWh and kVAr metersThermal demand ammetersThermal maximum demand ammetersPower factor (p.f.) meters1.5 to 5 VA2 to 10 VA5 VA5 VA5 VA3 VA4 to 8 VA5 VAFigure 15.1 5(b) 400 kV bar primary outdoor current transformer(Courtesy: BHEL)*These values are typically for moving iron instruments and devices.For electronic instruments and devices they would be of the orderof 0.1 to 0.5 VA.

istrument and control transformers: application and selection 15/473VA = 52 x [l + 3.93 x 10" (90-20)]1000= 2.42 VA (for details refer to Table 14.4)(a) Single phase HT CTComputing the VA burden1 The VA values of some of the devices used in thecircuit may be available at a different current ratingfrom the actual rated secondary current (1 or 5 A)chosen for the CT circuit. To compute the VA burdenof a circuit when selecting the correct VA level of aCT, the VA values of all the devices not correspondingto the rated current of the circuit must be first convertedto the rated current and only then added. This isessential because the VA level of a CT varies in asquare proportion of the current passing through it,Le., VA = 12. As a result, at lower operating currentsits VA capacity to feed a circuit would also decreasesharply while the VA requirement of the instrumentsor the relays connected in the circuit will remain thesame. It is therefore important that the VA level of theCT is raised in the same inverse square proportion ofthe current to maintain at least the same level of VAto make it suitable to activate the measuring orprotective devices connected in the circuit, i.e.(b) HT epoxy resin cast(c) HT epoxy resin castwhere VA, and VA2 are the VA levels of a circuit atcurrents I, and Z2 respectively.(d) LT epoxy resin cast(e) LT tape woundFigure 15.16 Typical indoor-type wound primary CTs formeasuring or protection (Courtesy: Kappa Electricals)Protective devices In view of the large variety of thesedevices such as static or electromagnetic, they may beobtained from catalogues or their manufacturers.Copperflexible leads (wires) The approximate resistancesof such conductors at 20°C are provided in Table 13.15.They can be estimated at the operating temperature (9OoC,as in Table 14.5 or as desired).VA burden = 12Re.g. the VA burden of a CT having a rated secondarycurrent of 5 A with the length of the 2.5 mm2 connectingleads as 10 m.Example 15. 2Consider a 5 A secondary CT circuit connected to the followingdevices:Device I = 0.3 VA at 1 ADevice II = 5 VA at 5 ADevice 111 = 7.5 VA at 5 AThen the total burden at 5 A will be= 0.3 x (5/1)' + 5 + 7.5= 7.5 + 5 + 7.5= 20 VATherefore, one should select a 20 VA CT.Similarly, if this value was required at the 1 A secondary, thenthe total burden would be= 0.3 + 5 x (1/5)'+ 7.5 (1/5)'= 0.3 + 0.2 + 0.3= 0.8 VAIn this case one can select a 2.5 or 5 VA CT.2 The current element of a relay is wound for a widerange of current settings in terms of the rated secondarycurrent of the CT, such as 1040% for a ground faultprotection, 50-200% for an overcurrent and 300400%for a short-circuit protection. At lower current settings,while the VA requirement for the operation of therelay will remain the same, the VA capacity of the CTwill decrease in a square proportion of the current. ACT of a correspondingly higher VA level wouldtherefore be necessary to obtain the reduced VA level,

15/474 Industrial Power Engineering and Applications Handbookat least sufficient to operate the relay. At a 40% setting,for instance, the CT must have a VA of (U0.4 I)2 or6.25 times the VA of the relay and at a setting of 20%,(U0.2 I)’ or 25 times of the relay. Therefore when therelay setting is low this must be borne in mind and aCT of a higher VA burden be chosen. Such aconsideration, however, is more pertinent in the caseof electro-magnetic relays that have a high VA levelthan in electrostatic (electronic) relays that have anear negligible VA level at only around 0.005 VA.Where three CTs for unrestricted or four CTs forrestricted ground fault or combined OK and G/Fprotections are employed in the protective circuit, theVA burden of the relay is shared by all the CTs in paralleland a normal VA CT may generally suffice. Such is thecase in most of the protective schemes discussed inSections 21.6 and 15.6.6(1), except for those employingonly one CT to detect a ground fault condition, such asfor a generator protection with a solidly grounded neutral(Figure 21.12).(iv) Circuit diagramThis can be drawn along similar lines to those for a VT(Section 15.4.1(12)). Refer to the simple diagram inFigure 15.17, from which we can derive the following:E= 7;- Land---I,, = I, 4- I;Primary sideASecondary sideA., R; X; Zle;. r .t t te, - Primary induced emf4 - Secondary induced emfe; - Secondary terminal voltagefor bar primary e, = e,R, - Primary circuit resistanceR; - Secondary winding resistance referred to the primary sideX, - Primary circuit reactanceX; - Secondary winding reactance referred to the primary sideZ- Load (burden) impedance/,( - Excitation or no load current/& - Loss component supplying the hysteresis and eddy currentlosses to the CT core (it is the active component)/m - Magnetizing component producing the flux ‘4’ (it is thereactive component)1; - Secondary current referred to the primary sideFigure 15.17Equivalent circuit diagram of a current transformer6 = Phase displacement (phase error) between i, and I;.Figure 15.18 Phasor diagram of a CTand from this is drawn the phasor diagram (Figure 15.18).-The phasor difference between zand I,, i.e.results in a composite error I;. The phase displacementbetween and 7; by an angle ‘6’ is known as the phaseerror. The current error will be important in the accurateoperation of an overcurrent relay and the phase error inthe operation of a phase sensitive relay. The compositeerror will be significant in the operation of a differentialrelay.(v) Current error or ratio errorThe error in the secondary current from the rated causedby the excitation current I,, or the variation in the actualtransformation ratio is expressed by:(K, xr; -1;)Current error = x 100% (1’2 = K, . 12I;(K, being the rated transformation ratio.))Refer to Table 15.8 for measuring and Table 15.9 forprotection CTs.(vi) Phase errorAs noted above, this is the phase displacement betweenthe primary and the secondary current phasors. Angle 6in Figure 15.18 is generally expressed in minutes. For aperfect transformer, the direction of phasors is chosen sothat this displacement is zero. Refer to Table 15.8 formeasuring and Tablc 15.9 for protection CTs.(vii) Composite errorRefer to the phasor diagram in Figure 15.18 and Table15.8 for measuring and Table 15.9 for protection CTs.This error can also be expressed by*Only the r.m.s. values and not the phasor quantities are consideredto define the current error. The phase error is defined separately.Together they form the composite error.

~~Accuracyclass2 % Current (ratio) erroP at % of rated primary current + Phase displacement angle 6 (Figure 15.18)in minutes at % of rated primary current% rated. I, 5 20 50 100 1200.1 0.4 0.2 NA 0.1 0.10.2 0.75 0.35 NA 0.2 0.20.5 1.5 0.75 NA 0.5 0.51 .o3.05.05 20 100 12015 8 5 530 15 10 1090 45 30 303.0 1.5 NA 1 .o 1 .o 180 90 60 60- - 3 - 3 Not specified- - 5 - 5 Not specifiedComposite error = - loo il/T joT (K; i2 - i,)* dt %I1whereK,, = rated transformation ratioI, = actual primary current (r.m.s.)I2 = actual secondary current (r.m.s.)i, = instantaneous value of the primary currenti2 = instantaneous value of the secondary currentT = duration of one cycle= 1/50 s or 20 ms for a 50 Hz system.(viii) Knee point voltageThis is the point on the magnetic curve of the laminatedcore of the CT at which the saturation of the core willstart. It is defined as the point where an increase of 10%in the secondary voltage will increase the excitation currentI,,, by 50% (Figure 15.19). Beyond this point, a verylarge amount of primary current would be required tofurther magnetize the core, thus limiting the secondaryoutput to a required level.(ix) Instrument security factor (SF)This is the ratio of instrument limit primary current tothe rated primary current. Consequently a high SF willmean a high transformation of the primary current andcan damage instruments connected to its secondary. Formeasuring instruments therefore it is kept low, as it isrequired to measure only the normal current and not thefault current.15.6.2 Measuring current transformersThese are employed for the measurement of power circuitcurrents through an ammeter, kW, kWh or KVAr andpower factor meter, or similar instruments requiring acurrent measurement. They must have a specified accuracyclass as in IEC 60044-1 and the secondary currentsubstantially proportional to the primary within a workingrange of about 5-120% of its primary rated current. TheyTable 15.9 Limits of error for protection CTsAccuracy Current error at Phase displacement Composite errorclass rated primary angle 6 (Figure at ratedcurrent 15.18 at rated accuracy limitprimary current primary current% minutes %5P +1 + 60 510P *3 - 1015P +5 - 15As in IEC 60044-1are required to commence their saturation beyond 120%of the primary rated current and saturate fully by 500%,as a system is not warranted to operate on an overload orshort-circuit, and will be interrupted through its protectivedevices. Thus a low knee-point voltage or a low saturationlevel is needed to protect the connected instruments fromfault currents (overcurrent factor) on the primary side.For example, in a measuring CT of 1000/5 A, thesecondary current will be in direct proportion to theprimary current from about 50 to 1200 A and the corewill start saturating beyond 1200 A.Overcurrent factor for instrumentsAs in IEC 6005 1, the measuring instruments are requiredto have an overcurrent factor of not more than 120% fortwo hours for instruments of all accuracy classes, 200%for 0.5 second for class 0.5 or less, and 1000% for 0.5second for class 1 accuracy and above. Overcurrents ordurations longer than this may damage the instruments.Accuracy classThis defines the maximum permissible current error atthe rated current for a particular accuracy class. Thestandard accuracy classes for the measuring CTs may beone of 0.1, 0.2, 0.5, 1, 3 and 5. The limits of error inmagnitude of the secondary current and the phase error,as discussed in Section 15.6.1 and shown in Figure 15.18,

15/476 Industrial Power Engineering and Applications Handbook-- IEa, DlPC.-.?.- Iu:Excitation current (I,,,) -Figure 15.19 Knee point of the excitation characteristic of acurrent transformer+ VA, +Main CT Interposing CT(VAM) ( 4)(a) Schematic diagrammust be as in Table 15.8, according to IEC 60044-1,when the secondary burden is a minimum 25% of itsrated'burden for CTs up to class 1 and 50% for CTs ofclasses 3 and 5.The recommended class of accuracy will depend uponthe type of application and is generally as notedbelow:ApplicationClass ofaccuracy1 Precision testing or laboratory testing CTs 0.12 Laboratory and test work in conjunction with high 0.2precision indicating instruments, integratingmeters and also for the testing of industrial CTs3 Precision industrial meters (indicating instruments 0.5and recorders)4 Commercial and industrial metering 0.5 or 15 Use with indicating and graphic watt-meters 1 or 3and ammeters6 Purposes where the ratio is of less importance 3 or 515.6.3 Interposing current transformersThese are auxiliary CTs, and are sometimes necessary toalter the value of the secondary of the main CTs. Theyhelp to reduce the saturation level and hence the overloadingof the main CTs, particularly during an overloador a fault condition. They are used especially where theinstruments to which they are connected are sensitive tooverloads. They have to be of wound primary type. Sothat the main CTs are not overburdened they have a VAload that is as low as possible. Figure 15.20 illustratesthe application of such CTs and their selection is madeon the following basis:VAI, = VAc + VAI + VALwhereVA, = VA of the main CTsI+Main CT(b) Equivalent control circuit diagramFigure 15.20 Use of interposing CTsVAc = circuit losses between the main and the interposingCTs at the primary rated currentVAI = VA of the interposing CTs at the primary ratedcurrentVAL = VA of the load (instrument) connected on thesecondary of the interposing CTs, including theconnecting leads.15.6.4 Summation current transformersThese are required to sum the currents in a number ofcircuits at a time through the measuring CTs provided ineach such circuit. The circuits may represent differentfeeders connected on the same bus of a power system(Figure 15.21(a)), or of two or more different powersystems (Figure 15.2 1 (b)). A precondition for summationof currents on different power systems is that all circuitsmust be operating on the same frequency and must relateto the same phase. The p.f. may be different.Each phase of these circuits is provided with anappropriate main CT, the secondary of which is connectedto the primary of the summation CT. Summation ispossible of many circuits through one summation CTalone per phase. The primary of summation CTs can bedesigned to accommodate up to ten power circuits easily.If more feeders are likely to be added it is possible toleave space for these on the same summation CT.The summated current is the sum of all the CT secondaqcurrents of the different circuits. The rating of theinstrument connected on the secondary of the summationCT should be commensurate with the summated current.The error of measurement is now high, as the errors ofall individual CTs will also add up vectorially.

, (Instrument and control transformers: application and selection 15/477Any main CT that is underloaded will also add to theerror in the measurement. Similarly, if provision is madein the primary of the summation CT to accommodatefuture circuits but is not being utilized it must be leftopen. otherwise it will also add to the error. The impedanceof the shorting terminals will add to the impedance ofthe circuit and will increase the total error.As the currents of each circuit are summed by thesummation CT. the VA burden of each main CT is alsoborne by the summation CT in addition to its own. TheVA level of the summation CT, including its own, isshared proportionately by all the main CTs in the ratio oftheir primary currents. Referring to the three differentcircuits of Figure 15.2 1 (b), having the ratings as shownin Table 15.10, the rating of the summation CT can bechosen as 340011 A. If we choose a VA level of this CTas 25 VA, making no provision for the future, then theVA burden shared by each main CT will be as calculatedin the last column, ignoring the losses in the connectingleads. Based on this, the VA burden of each main CT canbe decided.15.6.5 Protection current transformersThese are employed to detect a fault, rather than measuringthe current of a power system or the connected equipment.There is a fundamental difference in the requirement ofa measuring and a protective transformer in terms ofaccuracy. saturation level and VA burden. Unlike aR Y B ,I'Main CTsPI\\Summation CT(a) Measuring the sum load of three circuits on phase B.B.measuring CT, a protection CT will have a high saturationlevel to allow the high primary current to transformsubstantially to the secondary as may be required,depending upon the current setting of the protective ortripping relays. For protection CTs, therefore, the accuracyclass is of little relevance up to the primary rated current,but a true reflection in the secondary is more importantof a fault condition in the primary.CorollaryBoth requirements of measuring and protection cannotbe met through one transformer generally. Thus two setsof transformers are required for a power circuit associatedwith a protection scheme, one for measurement and theother for protection.(i) Accuracy limit primary currentThis is the highest limit of the primary current that canbe transformed to the secondary, substantially proportional,complying with the requirement of the composite error(Section 15.6.1). For example. a protection CT 2000/SArepresented as 5P10 means that a primary current up toten times the rated (Le. up to 2000 x 10 A) will induce aproportional secondary current. The factor 10 is knownas the accuracy limit factor as noted below.(ii) Accuracy limit factor (ALF)This is the ratio of the rated accuracy limit primary currentto the rated primary current. For example, in the abovecase it is2000 x 10 -2000The standard prescribed factors can be one of 5. IO, 15.20 and 30.(iii) Accuracy classThis defines the maximum permissible composite errorat the rated accuracy limit primary current, followed byletter P for protection. The standard prescribed accuracyMain CTsRlYlBir"""""""lBurdenT""""""""i a-BurdenBurden(b) Measuring the sum load of two circuits connected on different supply sources.Figure 15.21 Application of summation CTsR2y282SummationcTs

15/478 Industrial Power Engineering and Applications HandbookTable 15.10Circuit whose Current rating Main CT VA burden sharedcurrent is being ratio by the mainsummedCTsCircuit 1Circuit 2Circuit 31000 A800 A1600 A1000/1 A *25 'Oo03400= 7.0 VA800/1 A *253400= 6.0 VA1600/1 A *25 l6Oo3400= 12.0 VATotal load Ratio of *VA of= 3400 A summation summationCTs = 3400/1 CTs = 25classes may be one of 5P, 10P and 15P. A protection CTis designated by accuracy class, followed by accuracylimit factor, such as 5P10. The current error, phase error6 (Figure 15.18) and the composite error with the ratedburden in the circuit are as in Table 15.9, according toIEC 60044-1. It should be chosen depending upon theprotective device and its accuracy requirement todiscriminate. Closer discrimination will require moreaccurate CTs.NoteAn accuracy class beyond 1OP is generally not recommended.(iv) Output and accuracy limit factorsThe capabilities of a protection CT are determined by theprimary inputs of a CT such as the primary ampere turnsAT (primary current x primary number of turns), coredimensions and the quality of laminations. All this isroughly proportional to the product of the rated output(VA) and the rated accuracy limit factor of the CT. Fornormal use, the product of the VA burden and the accuracylimit factor of a protection CT should not exceed 150,otherwise it may require an unduly large and moreexpensive CT. For example, for a 10 VA. CT, the accuracylimit factor should not exceed 15 and vice versa. Theburden and the accuracy limit factor are thus interrelated.A decrease in burden will automatically increase itsaccuracy limit factor and vice versa. In a ring or barprimary CT, which has only one turn in the primary, theampere turns are limited by the primary current only, thuslimiting the accuracy and burden of such CTs. This is onereason why CTs of up to 5OAare generally manufacturedin a wound primary design, with a few turns on theprimary side to obtain a reasonably high value of VAburden and accuracy. For larger products than 150, it isadvisable to use more than one protection CT, or use lowsecondary current CTs, i.e. 1 A instead of 5 A.NoteFor similar reasons, a measuring CT of up to 50 A primary currentis recommended to be produced in a wound design.Example 15.3Consider a protective scheme having a total VA burden of 15and requiring an accuracy limit factor (ALF) of 20::. VA x ALF = 15 x 20 = 300which is too large to design a CT adequately. In such casesit is advisable to consider two sets of CTs, one for thoserelays that are set high and operate at high to very highcurrents (short-circuit protection relays) and the second forall other relays that are required to operate on moderateoverloads. For example, consider one set of CTs for shortcircuitprotection havingVA = 5and ALF=20i.e. a 5P20 CT having a product of VA x ALF of not more than150 and the other set for all the remaining protections having,say,VA = 10and ALF = 5i.e. a 10P5 CT having a product of VA x ALF of much lessthan 150.NoteFor high set protective schemes, where to operate theprotective relays, the primary fault currents are likely to beextremely high, as in the above case. Here it is advisable toconsider a higher primary current than the rated for theprotection CTs and thus indirectly reduce the ALF and theproduct of VA x ALF. In some cases, by doing so, even oneset of CTs may meet the protective scheme requirement.Example 15.4Consider a system being fed through a transformer of 1500kVA, 11/0.433 kV, having a rated LV current of 2000 A. Theprotection CT ratio on the LV side for the high set relay maybe considered as 4000/5 A (depending upon the setting ofthe relay) rather than a conventional 2000/5 A, thus reducingthe ALF of the previous example from 20 to 10. Now only oneset of 15 P10 CTs will suffice, to feed the total protectivescheme and have a VA x ALF of not more than 150.Other considerations when selecting aprotection CT1 The accuracy limit factor (ALF) will depend upon thehighest setting of thc protective device. For a 5 to 10times setting of the high set relay, the ALF will be aminimum of 10.2 A higher ALF than necessary will serve no usefulpurpose.3 It has been found that, except high set relays, all otherrelays may not require the ALF to be more than 5. Insuch cases it is worth while to use two sets of protectionCTs, one exclusively for high set relays, requiring ahigh accuracy limit factor (ALF), and the other, witha lower ALF, for the remaining relays. Otherwise choosea higher primary current than rated, if possible, andindirectly reduce the ALF as illustrated in Example15.4 and meet the requirement with just one set ofprotection CTs.

15.6.6 Special-purpose current transformers,type 'PS'These are protection CTs for special applications such asbiased differential protection, restricted ground faultprotection and distance protection schemes, where it isnot possible to easily identify the class of accuracy, theaccuracy limit factor and the rated burden of the CTs andwhere a full primary fault current is required to betransformed to the secondary without saturation, toaccurately monitor the level of fault and/or unbalance.The type of application and the relay being used determinethe knee point voltage. The knee point voltage andthe excitation current of the CTs now form the basicdesign parameters for such CTs. They are classified asclass 'PS' CTs and can be identified by the followingcharacteristics:Instrument and control transformers: application and selection 15/479CTR = lplI\Rated test winding current0 Nominal turn ratio (the error must not exceed k 0.25%)Knee point voltage (kpv) at the maximum secondaryturns,v, 2 2v*,where Vk = knee point voltage andV, = maximum voltage developed across therelay circuit by the other group of CTsduring a severe most through fault.Maximum magnetizing (excitation) current at thevoltage setting (VfJ of the relay or at half the kneepoint e.m.f. to be 5 30 mA for 1 A CTs for most highimpedanceschemes. The manufacturers select a properiron core to limit this to help reduce the effective relaycurrent setting and improve its sensitivity. Magnetizingcharacteristics, V, versus I,,,, (V, being the CT secondaryvoltage under rated conditions), as shown in FigureIS. 19, are provided by the manufacturer to facilitaterelay setting.Maximum resistance of the secondary winding correctedto 90°C or the maximum operating temperatureconsidered. In fact, it should be substituted by theactual operating temperature.We discuss below a high-impedance differentialprotection scheme to provide a detailed procedure tosclect PS Class CTs.1 High-impedance differential protection schemeThe scheme primarily detects an inter-turn fault, a groundfault or a phase fault. It can thus protect a bus systemand windings of critical machines such as generators,transformers and reactors in addition to a ground fault.The differential system is a circulating current systembetween the two winding terminals of the equipment oreach section of a multi-section bus system being protectedas illustrated in Figure 15.22. The scheme is based onKirchhoff's law, which defines that the phasor sum ofthe currents entering a node is zero, i.e._ - _ -II + I. + 1 3 + I? = 0L 2 -uGf,, F2 - 2 sets of identical class PS CTsRelay - High impedance three element differential protection relayWp - Windings of a power equipment or section of a power systemto be protectedFigure 15.22 A circulating current scheme to provide a phaseand a ground fault differential protectionas illustrated in Figure 15.23.Applying this law to a three-phase, three wire system.iR + i, + iB= 0and to a three-phase four-wire systemI, + iu + I , + I , = oWhen a three-phase four-wire system feeds non-linearor single-phase loads this balance is upset and theunbalanced current flows through the neutral. The samerelationship can be expressed as- - - -I R + IY + 10 = In.....I, + I , + I , + I, = 0Figure 15.23 Kirchhoffs law - sum of currents entering a nodeis zero

whcrc15/480 Industrial Power Engineering and Applications HandbookSimilarly, the balance is disturbed in the differentialscheme on a fault of any type and a spill current, whichis the difference between the currents drawn by the twosets of CTs, flows through the relay. Since the schemefunctions on the principle of balance of currents, it isimperative that the two sets of CT parameters, such astheir ratio, secondary resistance and the magnetizingcurrent, should be identical, except for the permissibletolerances as discussed in Section 15.10.2. The secondarylead resistances, from the CTs to the relay terminals,should also be the same, otherwise, spill currents mayflow through the relay, even under healthy condition andcause an unwanted trip, or require a higher minimumsetting of the relay. A higher setting of the relay mayjeopardize its sensitivity to detect minor faults. Since itis not practical to produce all CTs to be identical, smallspill currents under healthy condition are likely and theminimum relay setting, IEt, must account for this. Belowwe consider three different cases to explain the principleof circulating currents, along with the procedure, to selectthe CTs and carry out the relay setting.Equivalent circuit diagram and selection ofclass PS CTsRefer to the control circuit diagram of Figure 15.24,drawn for the scheme in Figure 15.22. It is drawn on asingle-phase basis for ease of illustration, whereIfl, Ir, = If = CT secondary currentsI,,, lm2 = I, = CTs’ excitation currents (these willLlvary with the CT secondary voltage;refer to Figure 15.19)I,,, I,. = I, = circulating currentsIre = spill or differential current throughthe relayVfl, Vfr = V, = CT secondary voltages under ratedconditions (these relays are definedby both the current and the voltagesettings)R, = resistance of the relay coil= VA~ VA is the burden of the13relay. This may be specified in termsof its current rating 1 A or 5 A orsetting current Ist. Considering thisto be 1 VArelay at a setting of 0.05 A,R,=-- -400Q(0.05)’I,, = relay settingR12 = R, = maximum resistance of theconnecting leads from the CTterminals to the relay terminals. Forcalculating this, for an estimatedlength and size, refer to cable datain Table 13.15XCT,, XCT2 = XCT = equivalent excitation reactances ofthe CT secondary windings. In ringtypeCTs, they are generally verylow and can be ignored for ease ofderivationRCTI, RCTZ = RCT = equivalent resistances of the CTsecondary windingscL2(1) Healthy condition, I,, = lCz - I,, = 0DifferentialrelayFigure 15.24 Equivalent control circuit diagram for a differentialground fault protection scheme of Figure 15.22The two current through the relay are in opposite directionsthereforeIre = IC2 - Strictly speaking, these areIC,all phasor quantities but= - Im2) - (If,- I~z) only their magnitudes areUnder healthy conditions considered for ease ofillustration, as quantitiesIf1 = If2of similar parameters suchas If,, If2 and I,,,,, I, falland I,, =Irn2almost in phase with each:. I,, = 0other.Hence, in a healthy condition there will be no spill currentthrough the relay and it will stay inoperative.Through-fault conditionRefer to Figure 15.25(a). On a fault occurring outsidethe protected zone, all the CTs that fall in parallel will

Instrument and control transformers: application and selection 15/481i-III_--*---,'$ G (Neutral of the winding grounded)F,, F2 - 2 sets of identical class PS CTs.Wp - Windings of a power equipment or section of a powersystem to be protectedFault location 1:The relay stays inoperative for a fault occurring outsidethe protected zone even if it is within the CTs zonefault location 2:It falls outside the CTs zone.(a) Principle of operationL2Small spill current through the relayI,, = IC2 ~ci(bj Control circuit diagramFigure 15.25 A through-fault condition outside the protected zone in a differential schemeshare the fault almost equally, depending upon the locationof the fault and the impedance of each CT circuit up tothe point of fault. The balance of the CTs secondarycurrents is therefore disturbed, but only marginally, asthe polarities of the two sets of CTs also fall in oppositionand neutralize most of the unbalanced current (Ic2- I,,)through the relay. The small spill currents may be takencare of by the minimum setting of the relay to avoid atrip in such a condition. Hence, the relay may remaininoperative on a moderate fault, as illustrated in Figure15.25(b).But this may not always be true, as it is possible thatone or more CTr in the faulty circuit may saturate partiallyor fully on a severe through-fault and create a short circuit(V, = 0) across the magnetizing circuits of all the CTsthat are saturated. Refer to Figures 15.26(a) and (b). TheCTs' resistances, however, will fall across the relay circuit.Assuming that the other sets of CTs in the circuit remainfunctional, this would cause a severe imbalance and resultin a heavy unbalanced current through the relay and anunwanted trip. Under such a condition.a severe through-fault.i, = maximum fault current through the secondaryof the CTs, on a severe through-fault. This maycorrespond to the fault level of the machine orthe system being protected, depending upon themachine or the system impedances that mayfall in the faulty circuit.The protection must be designed to remain inoperativein such a fictitious fault condition. This condition willalso determine the stability limit of the protection schemeand can be considered as the minimum voltage setting ofthe relay. In fact, this setting will have a sufficient safetymargin, as the knee point voltage, V,, of the CTs isconsidered quite high, of the order of V, 2 2Vft on theone hand, and the saturation of the CTs is possible onlyunder extreme conditions, on the other. Hence the levelof Vft developed by the CTs may not be as high as thoughtand when the relay is set at this voltage it will providesufficient stability.vt, = i,, (Ret + R,) IVorrwhereVi, = maximum voltage that may develop across therelay circuit by the other groups of CTs duringIt is advisable to choose the CTs with low \econdary current. say.at 1 A, to permit a lower relay setting for the voltage and thecurrent trip coil\. The reduced voltage across the relay will al\oimprove the stability level of the protection wheme.

15/482 Industrial Power Engineering and Applications Handbookn T SensitivityYT ItitRelayThis is the ability of the scheme to detect the weakestinternal fault.StabilityThis can be defined by the most severe external fault atwhich the scheme will remain inoperative. It should alsoremain inoperative in healthy conditions. That is it shouldbe immune to the momentary voltage or current transientsand normal harmonic contents in the circulating current.Series LC-filter circuits are generally provided with therelay coil to suppress the harmonics and to detect thefault current more precisely.L,(a) Power circuitUse of stabilizing resistanceIt is possible that the voltage Vf, may become sufficientlyhigh to cause a spill current on a through-fault higherthan the relay pick-up current, I,[. To ensure that no spillcurrent higher than the relay setting, Is,, will flow throughthe relay circuit under a through-fault condition, theimpedance of the relay circuit is raised substantially. Itcan be obtained by using a stabilizing resistance, R,,,such that the differential circuit will act like a highimpedancepath for this spill current, compared to thevery low magnetizing impedance of the saturated CT.This resistance is shown in Figure 15.26(b). It will allowa current of less than the relay pickup current, Is,. Tofulfil this condition, the impedance of the differentialcircuit must be a minimum to ensureshort circuitedThe normal practice is to choose R,, based on the settingvoltage required. In the above equation, Vft is the minimumvoltage required across the relay branch (R, + Rst) forpushing a current equal to I,, to ensure that the relaystays immune on a through-fault. During an internalfault,the fault current is much more than I,,, and henceit is easy to detect. The equation also implies that R,, ischosen high to limit the relay current during a throughfault(assuming that one of the CTs is fully saturated) toless than its pickup current. Solving the above equationfor R,,,L2Vn = & (RCT + ROI,, = Fault level of the equipment under protection(b) Control schemeFigure 15.26 Power circuit and control scheme during a verysevere external fault conditionSince the additional resistance will stabilize theprotective scheme during a maximum through-fault conditionwithout raising the relay setting, Is,, it is appropriatelytermed the stabilizing resistance. Figure 15.27 shows anarrangement in a relay circuit and for the purpose ofillustration, it is shown separately in various controlcircuits (Figures 15.24, 15.25(b) and Figure 25.26(b)).As standard practice, this resistance is supplied withthe relay by the relay manufacturer. It is of variable type,to suit system conditions and the actual fault level. The*This is relay-specific. The manufacturer may specify VAcorresponding to its rated current of 1 A or 5 A or setting current I??

I,,,RIYIBINI Stabilizingresistors(1450 R eachfor example 15.6)Non-linearresistorsInstrument and control transformers: application and selection 151483Figure 15.27 Three-element high-impedance circulatingcurrent relay schememaximum value of the stabilizing resistance to be suppliedwill depend upon the type of protection (ground or phaseor both) and the relay setting. Generally, it may varyfrom SO to IS00 Q.Fault within the protected zoneRefer to Figure 15.28(a). The balance of the two sets ofCTs is disturbed again. The CTs now have the samepolarity and currents as the two sets add up to cause ahigh-imbalance spill current through the relay. Referringto Figure 15.28(b),1I,, = I,, + 12 = (1\P These are all phasor- I,,,?)quantities, but considered+ (I,,, linear, for ease of~ 1illustration and without= 142 + I,,, - 21,n much errorThe additive characteristic of the scheme now has highstability and prevents the relay from operating on moderateexternal faults, while it is sensitive to small spill currentsfor all internal ground and phase faults, including windingfaults.2 Current setting of the relayThe relay has voltage as well as current settings. Theformer defines the stability limit against through-faults,as discussed above, while the latter determines thesensitivity of the protected zone.If IPf = minimum fault current through the primary(chosen on the basis of the rated full-load currentof the machine or the system being protected)required to trip the relay. It may be termed theminimum primary operating current (POC) ofthe scheme. Ipf, In terms of the secondary= n x I,fIZ = turn ratio of the CTsI,, = corresponding to the V,, to account for the mostsevere through-faultI,, = relay current setting, i.e. minimum spill currentrequired to operate the relayThen- - -I,, = I,, + I,,Since on a fault the p.f. is low (Section 13.4.1(5)) allHigh impedance differential protectionrelay. It operates for the fault occurringwithin the protection zoneFault current through ground for faulton phase B(a) Power circuitFIL-/IL2 4, = IC2 + 41(b) Control schemeFigure 15.28 Fault within the protected zoneDifferentialrelay

15/484 Industrial Power Engineering and Applications Handbookthese quantities may be considered in phase with eachother, with little error,:. Isf = I, -k IstIf there are N number of CTs connected in parallel, themagnetizing current will flow through all of them. In aGF protection scheme all the three CTs of all the feedersbeing protected together will fall in parallel, while incase of a combined GF and phase fault protection scheme,only one third of these CTs will fall in parallel. The CTin the faulty circuit must be able to draw enough currentto feed the magnetizing losses of all the CTs falling inparallel and the relay pickup current, ISt. The sensitivityof the differential scheme can therefore be expressedmore appropriately asNon-linearresistorStabilizingresistorRelay(N being the number of CTs falling in parallel) and interms of the primarySince it determines the sensitivity level of the protectionscheme, it must be kept as low as possible to detect evena small fault. To achieve a high degree of sensitivity it istherefore essentialTo have the CTs with a low I,To keep the number of CTs in parallel as small aspossible, suggesting protection of individual feeders,rather than many feeders together, particularly whenthe equipment is critical and requires a higher level ofsensitivity for adequate protection.As the relay will have only one current setting for alltypes of faults, it is recommended to keep it around 20-40% of the rated current of the machine or the systembeing protected. This setting will be sufficient to meetthe CT’s magnetizing current requirements and also tripthe relay.For a ground fault scheme, it is recommended toconsider a still lower setting to ensure effective detectionof the ground fault current and rapid disconnection ofthe machine or the bus system being protected. A lowersetting may be desirable as the actual ground fault currentmay already be larger than is being detected by the relaydue to a higher impedance of the ground loop than assumedpreviously.As a rule and as recommended by IEC 60255-6, thePOC may be chosen within 30% of the minimum estimatedground fault current. When the scheme is required todetect only a ground fault, a single-pole relay is connectedbetween all the CTs’ shorted ends (Figure 15.29). All theCTs now fall in parallel.When the scheme is required to detect the ground faultas well as the phase faults, a triple-pole relay is used,each pole of which is connected between the shortedterminals of the two same phase CTs and the neutralformed by shorting the other terminals of all the CTs, asshown in Figure 15.22. The setting of all the poles iskept the same. In other words, the sensitivity level remainsthe same for all types of faults.L- GRelay- High impedance single element ground faultdifferential protection relayFigure 15.29 Scheme for only ground fault differential protectionIn the case of overcurrent and ground fault protectionit becomes much higher than in a single-pole relay. Nowthe requirement of the minimum primary operating current,Ipf (equation (lS.l)), which is a measure of sensitivity, isgreatly reduced. The CT on the faulty phase has to feedonly one third of the CTs that fall in parallel of eachrelay coil rather than all the CTs, that fell in parallel inground fault protection using only a single-pole relay.The CTs are designed for the worst conditions of fault,even when the scheme is designed to detect only a groundfault. This may be a phase to phase and ground fault,causing a severe unbalance. The iron core of such CTsmust therefore possess near-linear magnetizingcharacteristics, to the extent of the fault level of themachine or the system being protected. This is to achievea near-replica of the magnitude of the fault in thesecondary, which may be 15 to 20 times or more of therated current. In generators, it can increase to 21. I, (Section13.4.1(5)). For the CTs, a saturation level sufficient totransform the maximum primary fault condition to thesecondary is therefore considered mandatory to ensurethat the CTs do not saturate during the most severe faultcondition, and render the tripping scheme erratic. Thisalso ensures better stability of the relay, particularly duringsevere most through-fault conditions (outside the CTs’detection zone) such as a bus fault, as illustrated in Figure15.30. It is normal practice to define the secondary voltageof the CTs by its knee point voltage (kpv), V,. Thisvoltage will depend upon the type of relay, its VA burdenand the required stability of the system. It is commonpractice to make this at least twice the relay setting voltageon the most severe through-fault, Le. V, 2 2V,.

Instrument and control transformers: application and selection 15/485high voltages across the CTs and the relay, particularlyduring internal faults, when the CTs have the same polarityand the spill currents are additive. As in IEC 60255-6, itmust be limited within 3 kV across the relay circuit toprotect the CTs and the relay. An approximate formulato determine the likely peak voltage across the relaycircuit is given byF, F3F2Through faults which may be much higher than atF2 but outside the CTs zone- Internal fault being fed by two sources althoughlimited by the equipment impedanceFigure 15.30 An internal fault being fed by more than one sourceThe most severe fault is the capacity of the machine orthe system being protected to feed the fault, and isdetermined by its fault level as indicated in Tables 13.7and 13.10. To consider a higher fault level than this,such as of the main power supply, is of little relevance asit would fall outside the detection zone of the CTs andwould serve no useful purpose except to further improvethe stability level of the protective scheme.Applying this scheme to system protection, where thenumber of circuits and hence the number of CTs arehigh. will mean a high POC (equation (15.1)). A highPOC may not be desirable, as it may underprotect thesystem. In such cases, it is advisable to divide the systeminto more than one circuit and apply the schemeindividually to all such circuits (Example 15.6).3 Suppressing system harmonicsSuch relays are normally instantaneous, highly sensitiveand operate at low spill currents. Since they detect theresidual current of the system, the current may containthird-harmonic components (Section 23.6) and operatethe highly sensitive relay in a healthy condition. To avoidoperation of the relay under such conditions, it is a normalpractice to supply the relay coil with a tuned filter, i.e. aseries L-C circuit to filter out the third-harmoniccomponents. The capacitance of the filter circuit mayalso tame a steep rising TRV (Section 17.10.3) during amomentary transient condition and protect the relay.4 Limiting the peak voltageAs this is a high-impedance scheme, it can result in verywhere Vp = peak voltage across the relay andV,,, = theoretical rnaxi~nurn CT secondary voltageacross the relay circuit at the maximuminternal fault current. (The maximum internalfault current is the level of fault of themachine or the system under protection.)This must also take into account any othersupply sources that may also feed the fault.such as more than one supply bus, as shownin Section 13.4.1(5) and Figure 13.18, andillustrated in Figure 15.30. If the cumulativefault current is I, then the maximum CTsecondary voltage will beV,, = lscc x impedance of the relay circuit.This can be limited by using a non-linear resistance calledMetrosil" across the relay, as shown in Figure 15.27. Ifvoltage reaches a dangerous level, this resistance willprovide a low-resistance parallel path to the current andlimit the voltage across the relay to about 1 kV. Thecurrent I through the non-linear resistance is given byV,, = K x Ip (K and p are constants)All these values are provided by the relay supplier whenthis resistance becomes necessary.5 I Selecting class PS CTsGround fault protection of a machine and setting of therelay. The following example illustrates the procedure toselect class PS CTs for a typical G/F scheme. In practice,this scheme would be more appropriate for phase andground fault protections, as illustrated in Figure 15.22.Example 15.5Consider a generator, 10 MVA, 3.3 kV, for ground faultprotection having a sub-transient reactance x; = 12 f 10%(Figure 15.29).Grounding method:Overload capacity:Relay type:Rating:VA:solidly grounded150% for 30 seconds (as in IEC60034-1 )differential1A1, at the setting current, lSt*This is a brand name given by the manufacturer of the non-linearresistor. a GEC group company in the UK. General Electric. USAcall it Thyrite, and similar names have been given to it hy differentmanufacturers. Basically, it is a Sic non-linear resistance to providethe desired overvoltage protection. Refer to Section 18.1. I for moredetails.

28515/486 Industrial Power Engineering and Applications Handbook10 x 103I, = 7%I3 x 3.3= 1750 AThe fault level of the system,I,, = 1750 x -100(equation (13.5))10.8= 16.20 kA(Assume a lower value of x; (12 - 1.2 = 10.8%) to be on thesafe side.)Consider CTs with a ratio of 2000/1 A and having R,, = 7 Rand the magnetizing chracteristics as in Figure 15.31. Considera lead resistance from CT terminals to the relay to be 0.5 Rper lead.:. Total lead resistance, R, = 2 x 0.5Fault current in terms of the secondary,is, = 162001x - = 8.1 A2000= 1 R (presuming this to be atthe operating temperature)1 Relay voltage setting (stability limit)V,, = isc(RCT + R,) (considering the resistances at the operatingtemperature)= 8.1 (7 + 1)= 64.8 Vsay, 65 V or nearest higher setting available on the relay.:. Minimum kpv, V, = 2 x 65= 130 VKnee point2 Relay current settingConsidering a ground circuit resistance of, say, 2 R:3.3 x:. I, =100045 x 2where /g is the ground fault current= 952.6 A (say 950 A)Let us consider a setting of, say, 30% of Ig:... /,f = 0.3 x 950= 285 AReferring to Figure 15.28(a) the number of CTs that will fallin parallel,N=6and Im corresponds to the relay voltage setting of 65 V fromthe curve of Figure 15.31 = 15 mA.From equation (15.1)285 = 2000 (6 x 0.015 + ISt):. I*t = ~ - 6 x 0.0152000= 0.1425 - 0.09= 0.0525 ATherefore the relay can be set between 5-7.5% of 1 A.3 Stabilizing resistanceTotal desired relay circuit impedance- 65 - 1238R0.0525Relay resistanceR -VA- = Lr -I:; (0.0525)'= 363 R:. Required stabilizing resistanceR,t =I238 - 363 = 875 R/ III4 Peak voltage across the relay circuitv,= 2 & J m (1 5.2)where V,, = is, x R, (considering that there are no other feedsto the generator internal fault from othersources)= 8.1 x 1238= 10,027.8 V:. V, = 2&,,/130(10,027.8 -130)15 I, 1.5/,Excitation current I, (mA) +(Ampere-turns converfed into Amps.)RCT = 7 ClCore material - CRGO silicon steelFigure 15.31 Assumed magnetizing characteristic of 2000/1A class PS CTs= 3208 Vwhich is a marginal case. It is, however, advisable to providea non-linear resistance.5 Specification for class PS CTsCTR = 2000/1

Instrument and control transformers: application and selection 15/487R Y BCircuit breakercharacteristicsUnder healthy conditionFigure 15.32 Phase and ground fault differential protection scheme for a transformer and feeder bus protectionQuantity = 6 numbersVk 2 130 VI,,, = maximum 15 mA at a Vk/2 of 65 V.The CT manufacturer must provide the user with themagnetizing characteristics of the CTs, i.e. I,,, versus V,.I1 Protection of a feeder circuitExample 15.6Consider a power distribution system as shown in Figure15 32, where a transformer of 50 MVA, 33/11 kV, having afault level of 750 MVA, is feeding a bus connected to sixfeeders of different ratings All the CTs for a combined phaseand ground fault may be connected in parallel as illustratedThe CTs on the primary side of the transformer will be similarto those on the outgoing feeders, except for the insulationsystem and the turn ratio, to provide identical secondarycurrent and magnetizing characteristics, as on the secondaryside of the transformer The relay may be set for a slightlyhigher value to account for the slight error introduced and theconsequent spill currents to avoid an unwanted trip50 x lo3/ - -~1.732xT1= 2625 AConsider a reasonably low value of Ipf, say, 25% of I,, toachieve a high level of sensitivity and still feed the I,,, to allthe1 = 7 CTs3... I,, = 0.25 x 2625= 656.25 AAssume the CTs on the secondary side of the transformer tobe 3000/1 and on the primary to be3000 11 or 1000 A1 33 1and their magnetizing characteristics as in Figure 15.33.

15/488 Industrial Power Engineering and Applications HandbookI Knee ooint I-0 20 I, 1.51,Excitation current lm (mA)(Ampere-turns converted into Amps.)Figure 15.33 Assumed magnetizing characteristic of 3000/ 1 Aclass PS CTsFor all the CTs let R,, = 10 R and lead resistanceRi = 0.75 x 2 = 1.5 RThe fault level of the systemIs, =7501.732 x 11= 39 (say 40 kA)and in terms of the secondaryis, = 40 x - lo3 x l3000= 13.3 A1 Relay voltage setting (stability limit)V,, = 13.3 (10 + 1.5) (considering the resistance at the operatingtemperature)= 152.95 Vsay, 155 V or nearest higher setting available on the relay.:. Minimum kpv, V, = 2 x 1552 Relay current setting.= 310 V (Figure 15.33)Ipf = 656.25 AI,,, at 155 V = 20 mA from the curve of Figure 15.33.. 656.25 = 3000 (7 x 0.02 + Ist)or ISt = 656.25 -7 x 0.023000= 0.0788ATherefore, the relay can be set, say, at 10% of 1 A.The scheme is suitable to detect both a ground fault anda phase fault.3 Stabilizing resistanceRelay circuit impedance, R, = - 1550.1= 1550 R1Relay resistance, Rr =s= 100 R(for the same relay as in the earlier example):. Required stabilizing resistance,R,t = 1550 - 100 = 1450 R4 Peak voltage across the relay circuitVk = 310 VV, = is, R,= 13.3 x 1550= 20,615 V:. V, = 27/1,/310(20,615 - 310)= 7095 Vwhich is more than 3 kV. Hence, a non-linear resistance willbe necessary across the relay branch and must be orderedfrom the manufacturer with the relay.5 Specification for class PS CTsCTs : 33 kVCTR = 1000/1 ACTs : 11 kVCTR = 3000/1 AVk 2 310 VQty 3 numbersQty 21 numbersI, = as low as possible, but not more than 20 mA at 155V. The CT manufacturer must provide the magnetizingcharacteristics.Notes1 In the above case the incoming feeder would trip, evenwhen the fault occurs in any of the outgoing feeders, whichmay not be desirable. It is therefore recommended thatthis scheme be applied to individual feeders, so that incase of a fault, only the faulty feeder is isolated rather thanthe whole system.2 With the relay one should also order from the manufacturer(a) Stabilizing resistance of 1450 R and(b) A non-linear resistance to discharge the excess inducede.m.f., across the relay circuit.Both these resistances are illustrated in Figure 15.27.15.6.7 Core-balanced current transformers(CBCTs)These are protection CTs and are used for ground leakageprotection. They are also a form of summation CTs, wherethe phasor sum of the three phase currents is measured.The phasor difference, if any, is the measure of a groundleakage in the circuit. They are discussed in Section 21.5.15.7 Short-time rating and effect ofmomentary peak or dynamiccurrentsThe normal practice of users when selecting a measuringor a protection CT has been to specify only the current

1.5Instrument and control transformers: application and selection 15/489ratio and the likely maximum VA burden it may have tofeed as well as the class of accuracy for metering andaccuracy limit factor.Fault level is normally not mentioned, nor is it requestedby the CT manufacturer. Generally, it should be sufficientto meet the likely fault level of the system and its durationin most cases, particularly on an LT system. For criticalinstallations, large feeders and all HT systems, however,it is recommended to check the suitability of the CTs forthe system fault level and its duration.A short-circuit on a system will cause overheating asa result of the short-time current, Isc, and its duration of1 or 3 seconds, according to the system requirementsand its protective scheme. It will also developelectrodynamic forces (equation (28.4)) as a result of themomentary first peak of the fault current (in CTs it is 2.5times the short-time current, Isc, as in IEC 60044-1; seealso Table 13.11). These forces may result in electricalas well as mechanical damage to the windings of a CTdepending upon the number of turns in the primarywinding and the configuration of the coil. For bar primaryCTs, having only one turn in the primary, such forces arethe least, hence the statement above. With a lower classof accuracy, a low VA level, and a lower accuracy limitfactor (for protection CTs) a CT can easily be built to bemechanically rugged. But higher requirements of suchparameters may necessitate a bulky CT, disproportionatein size and cumbersome to install.In most applications, a bar primary CT is generallyused and a normal CT may be suitable. But for too smallratings, where the use of a wound primary CT isimperative, short-circuit effects must be considered, exceptthe CTs for an LT system, where the fault level for suchsmall ratings may be very low and may not matter (Section13.4.1(5)). For applications on an HT system, where awound primary CT is imperative, choice of a CT fromstandard wound primary CTs may still be possible, meetingthe minimum requirements of class of accuracy, VA burdenand short-time rating. IEC 60044-1 indicates for measuringand protection CTs the maximum short-time factors (STF)that can be obtained economically for a normal woundprimary CT whereShort-time factor (STF) =Rated short-time currentRated primary current---Is,I,Some STFs for more important CTs are reproduced inTable 15.11.Example 15.7Consider a 1.5 MVA 33/11 kV transformer having a fault levelof 750 MVA. The STF can be calculated as below with acircuit of 11 kV:and I,= ~Isc =750~a x 11= 40 kAx 1000a x 1179ATable 15.11 Maximum short-time factors obtainableeconomically corresponding to rated output, accuracy class,accuracy limit factor and rated short-time for wound primarycurrent transformersAccuracy Rated output STF obtainable, corresponding to theclass VA rated short times up to(A) Measuring CTs0.5 2.551015301 2.5510153 302.55101530(B) Protection CTs5P 10 2.551015305P15 2.551015305P20 2.5510153010P5 2.55101530lOPl0 2.5510153010P20 2.55101530:. STF=40 x 100079= 5060.5 s 1.0 s 3.0 s11007505003152001100100067550021511001000675500215550375225150-32521515085-325200100--1000750425375150600425215200-32521512585-115525350215125775700415350200115100475350200400275150100-25020010060-25012575--700525300215100425300200125-2252007560-450300200150154504002752001104504002152001102251509060-1351106035-1351540--4003001751506025017511015-1251105035Consider a CT with a ratio of 100/5 A and select a barprimary CT. If a bar primary is not practicable, then foran STF of almost 500, we can choose a wound-typemeasuring CT from Table 15.11 with an accuracy class-

15/490 Industrial Power Engineering and Applications Handbookof 0.5 and above and a corresponding VA burden of 5 fora short-time current of 1 second. If these parameters arenot suitable use the measuring CT with a higher-ratedprimary current to meet the requirement.Similarly, for a protection CT from Table 15.11 choosean accuracy class of 10P5 with a VA burden of 5 for asecond short-time current. If this does not meet the need,the protection CT may also have to be selected with ahigher-rated primary current.15.8 Summary of specificationsof a CTIn Table 15.12 we listed the data, that a user must provideto a manufacturer to design a CT for a particularapplication. Some of the data chosen are arbitrary todefine the specifications.15.9 Precautions to be observedwhen connecting a CTIt is mandatory to ground the secondary circuit of theCTs (in a balanced 3$ system, the current through theneutral will be zero; see Section 21.2.2, Figure 21.7).It is required to eliminate the error due to accumulationof electrostatic charge on the instruments that mayinfluence the readings. All the CTs in a circuit mustbe grounded at one point only otherwise circulatingcurrents may raise the potential of the circuit, whichis dangerous and may damage instruments or give theoperator a shock or even trip other relays connectedin the circuit.One should not allow the CT secondary to be opencircuited when it is energized, for it may inducedangerously high voltages. This phenomenon isexplained in Example 15.8.Example 15.8To determine the terminal voltage of a CT during an accidentalopen circuit under an energized condition consider a meteringCT connected across a few instruments. Refer to the followingfigure based on Figure 15.17, showing an equivalent CTcircuit referred to its secondary side.We have assumed the following parameters,Connected Load = 7.5 MVASystem Voltage = 11 kVBurden of all instruments connected across theCT = 7.5 VALead resistance = 2 x 0.5 R = 10Rated current, I, =7.5 x 106fix 11 x 103= 393.7 AConsider a CT ratio of = 40015 A11 io3Load impedance Z, =45 x 393.7= 16.13 51Z, referred to the secondary side = 16.1 3 (400/5)2= 103.23 kRFigure 15.34 Shorting of all unused CT terminals in a CTsecondary circuit using a selector switchFor ease of analysis we have ignored (without much error)the CT's own resistance and reactance.Total resistance of the instruments under rated condition= 7.5/5'= 0.3 R11 400q= -x-& 5= 508.1 kV(a) Under energized condition when the CT's secondaryis a closed circuit, voltage developed across the relay,e; = , e* x 0.3circuit impedance508.1 x lo3x 0.3- (103.23 x lo3 + 1 + 0.3)= 1.48 V(b) Under energized condition when the CT's secondaryis accidentally open circuited, the current will haveonly the magnetizing path and the voltage inducedacross the CT open terminals will be the same asacross the magnetizing circuit. Under this situationthe magnetizing circuit shall carry the same currentas caused by the primary current, which is very high.:. Voltage developed across the CT open terminals,= 508.1 x lo3x Ze(103.23 x lo3 + Ze)[where Z, = Impedance of excitation circuit,]1 1 1Z, - 170 +J60For simplicity, considering the approximate impedance of theexcitation circuit (without much error) as J60 R.. = 508.1 x lo3 x J60(103.23 x lo3 + J60)= J295 Vwhich is approximately 200 times that of voltage under normalcondition and hence highly detrimental for the insulation ofthe CT, the connecting leads and the human contact etc.Depending upon the system loading at the instant of CTcircuit interruption, it is possible that the primary current isenough to cause a saturation of the CT core. When so, it is

Instrument and control transformers: application and selection 15/491I**0.3 R4 One should select a lower secondary current, say, 1 ACT, for installations requiring long lengths of connectingleads, such as for remote measurement of currentor other quantities. It is advisable to limit the extraVA burden on the CTs, on account of such leads.SECTION 111: TESTING OFINSTRUMENT AND CONTROLTRANSFORMERSExcitation circuit*Typical parameters of the CT**Resistance of instruments of 7 5VA(a) Under energized and closed circuit condition- u- - -Z, = 103 23 kQ 1.0R393 7A* Approx. impedance of the excitation circuit.(b) Under energized but open circuit condition.CT circuit referred to the Secondary side.likely that the induced voltage across the CT open terminalsmay give a further momentary kick upto 242 times the voltagecalculated above, as the current and hence the voltage shallundergo a rapid change from one peak to the other withinone half of a cycle, Section 1.2.1.3 Provision must be made to short-circuit all the CTsecondary terminals not in use (for example, in threeenergized measuring CTs, connected to a commonammeter through a selector switch, when either noneor any one of the CTs only may be connected to theammeter at a time, the other CTs remaining out ofcircuit). In such cases, except for the CT in use, theremaining CTs should be shorted. The selector switchesare therefore designed so that all the CT terminals notin use are shorted automatically through the switch,even during a change-over from one CT to another. Atypical circuit diagram of the switch is shown in Figure15.34, which illustrates the fulfilment of this requirement.It may be observed that in the OFF position, allthe CT secondaries are shorted and when any one ofthem is in circuit, the remaining two are shorted.All such switches must be the ‘make before breaktype’,so that the CT terminals are shorted beforebeing connected to the load (ammeter) during thechangeover.15.1 0 Test requirementsThe following tests are recommended on a finished voltagecurrent transformer:Type tests These are conducted on a finished voltageor current transformer, one of each design and type,to verify their compliance with the design data andrelevant Standards.Routine tests These are conducted on each finishedvoltage or current transformer to verify their suitabilityfor the required duty.Field testsSpecial tests Any tests that are not covered aboveand are considered necessary by the user may be agreedupon between the manufacturer and the user.15.10.1 Voltage transformers1 Type tests These will cover the following tests:(i) Temperature rise test(ii) Verification of dielectric properties on the primarywindings; To check the insulation level, as inTables 13.2 and 14.3 for series IandTables 14.1and 14.2 for series I1 voltage systems.(a)Power frequency voltage withstand or HVtest.(b) Impulse voltage withstand or lightningimpulse test for system voltages 2.4 kV andabove.Since a VT is associated with a switchgear,either with its assembly or with theswitchyard. the above two tests are almostthe same asthoseforthe switchgear assemblyand as discussed in Chapter 14. The testrequirements and procedure are also similar.(iii) Wet test for outdoor type transformers Theoutdoor VTs are also tested for dielectric propertiesunder wet conditions. The procedure to createthe wet conditions and to carry out the test arespecified in IEC 60060-2. In wet conditions, theVT has the same test voltages as specified above.(iv) Verification of accuracy The test results obtainedmust comply with the values of Tables 15.5 and15.6 for a measuring and a protection transformerrespectively. For brevity, we have limited ourdiscussions as above. For more details, exact testvalues and test procedure refer to IEC 60044-2.

15/492 Industrial Power Engineering and Applications HandbookTable 15.12 Summary of specifications of a CTSr. no. Specifications Measuring CTs Protection CTs Special-purpose protectionCTs type ‘PS’1 System voltage As in Table 13.12 Insulation level (peak)As in IEC 60044-1 or Tables 13.2 and 14.3 for Series I and Tables 14.1 and 14.2 for Series I1voltage systems3 Class of insulation E B B4 Frequency 50 or 60Hz 50 or 60 Hz SO or 60 Hz5 Nominal current ratio 600/5 A 2000/5 A 2000/5 A6 VA burden < 2.5, 5, 7.5, 15 or 30 >7 Class of accuracy 0.1, 0.2, 0.5,1, 3 or 5 (SP, IOP or 15P)a -8 Accuracy limit factor (ALF) - (5, 10, 15)b -9 Short-time current I,, and its 25 kA for 1 second SO kA for 1 second -duration10 Dynamic current Minimum 2.5 times I,, Minimum 2.5 times I,, -(in the above case 62.5 kA) (in the above case 125 kA)1112131415Nominal turns ratioLimiting secondary‘resistance at 90°C (a)Knee point voltage (V)Excitation current at kneepoint voltage (or at any otherrequired voltage or both) (A)Service conditions1/400Indoors or outdoors Indoor or out-door Indoors or outdoorsAmbient temperature Ambient temperature Ambient temperatureAltitude, if above 2000 m Altitude, if above 2000 m Altitude, if abovefor LT and abovefor LT and above2000 m for LT and1000 mm for HT 1000 mm for HTabove 1000 mm for HTHumidity Humidity HumidityAny other important Any other important Any other importantrequirement requirement requirement39500.0516Making of CTs600/5 Ad 2000/5 A 6.6 kV, 2000 A, 1/400IO VA, class 115 VA, class 5P10 950 x 0.05 R3.System voltage and insulation level, class of insulation, frequency, short-time rating anddynamic current rating etc. for all types of CTsNotes“The class of accuracy for protection CTs is recommended to be not more than IOP as far as possible.bProduct of VA ahd ALF not to exceed 150.‘The limiting secondary resistance is required to determine the secondary limiting e.m.f. which is = (FS) x rated secondary current x VAx resistance of secondary windings at 90°C or the highest operating temperature as in Table 14.5, whereFS = Instrument security factorRated instrument limit orimarv currentRated primary currentdWherever two separate secondary windings are provided, say, one for measuring and the other for protection, the markings shall indicateall such details that are marked against (a) for each secondary winding.2 Routine tests These will cover the following tests:(i) Verification of terminal marking (refer to Table15.7 and Figure 15.35, illustrating types oftransformer connections).(ii) Power frequency withstand test on the primarywindings. This must be conducted Only On theelectromagnetic unit of a VT. For example, whentesting a capacitor VT it must be conducted onlyon the secondary circuit, i.e. the electromagnetictransformer. The test values and test procedurewill remain the same as discussed above.NoteA repeat power frequency test, if considered necessary, must beperformed at 80% of the prescribed test voltage. See also Section14.5.

Instrument and control transformers: application and selection 151493(iii) Power frequency withstand test on the secondarywindings. This must also be conducted only onthe electromagnetic unit of a VT, as noted above,at 2 kV for 1 minute, similar to the control andauxiliary circuit dielectric test (Table 14.3).(iv) Verification of accuracy: as under type tests above.Field tests Power frequency withstand test on theprimary windings (for un-grounded VTs). The valueof the test voltage and the test procedure, is almostthe same as that for a switchgear assembly (Section14.5).Additional tests on a capacitor VT The testsdiscussed above refer generally to the electromagneticunit only. To test the whole VT, the following testsare recommended. For the test procedure and resultsrefer to IEC 601886.Type testsITests on capacitors(a) Self-resonating frequency test - applicable onlyto carrier coupling capacitors.(b) Power frequency wet withstand test on outdoorcapacitors.(c) D.C. discharge test.(d) Impulse voltage withstand test.(e) Partial discharge (ionization) test.2 Temperature rise test.3 Impulse voltage withstand test.4 Ferro-resonance test.5 Transient response test.6 Verification of accuracy.Routine tests1 Tests on capacitors(a) Capacitance and tangent of the loss angle (tan 6)(b) Power frequency dry withstand test(c) Sealing test2 Verification of terminal marking3 Power frequency withstand test on the secondary circuit4 Verification of accuracy15.10.2 Current transformers1 Type tests These will cover the following tests:(i) Short-time current (Isc) test(ii) Momentary peak or dynamic current test. (Thismust be conducted at a minimum of 2.5 /sc)RT TYorNR' Y'or N(1) Single phase VTr71R' 1Y'or IN'r72R' 2Y'or 2N(2) Single phase VT with twosecondary windingsmr7R' N R' N'(5) Single phase residual VTi i P T'1'1-1R Y B Nt t t tY Y JB'T NB B t i2"B A Y(3) 3-phase VT(4) 3-phase VT with twosecondary windings(6) 3-phase residual VTY"Figure 15.35 Single and three-phase VTs with one and two windings in the secondary

15/494 Industrial Power Engineering and Applications Handbook(iii) Temperature rise test(iv) Verification of dielectric properties on the primarywindings to check the insulation level as in Tables13.2 and 14.3 for series I and Tables 14.1 and14.2 for series I1 voltage systems.(a) Power frequency voltage withstand or HVtest.(b) Impulse voltage withstand or lightning impulsetest for system voltages 2.4 kV and above.Since a CT is associated with a switchgear, eitherwith its assembly or the switchyard, the abovefour tests are almost the same as those for aswitchgear assembly and as discussed in Chapter14. The test requirements and procedure are alsosimilar.(v) Wet test for outdoor type transformer: This testis similar to that for a VT and as discussed inSection 15.10.1(1).(vi) Verification of accuracy: the test results obtainedmust comply with the values of Tables 15.8 and15.9 for a measuring and a protection CTrespectively.2 Routine Tests These will cover the following tests:(i) (a) Verification of terminal markingRefer to Figure 15.36, illustrating types oftransformer connections.(b) To check the polarity of a C T It is imperativethat the terminals of a CT are wired withcorrect polarity, with reference to the primaryin each phase. A reversal in any phase willlead to incorrect meter readings, in meteringCTs and erratic signals to the protective relaysin protection CTs. Although CTs are markedwith polarities by their manufacturers, suchas PI P2 for primary and SI S2 for secondary(Figure 15.36) it is possible, that by humanerror at the time of fitting the CTs, care is nottaken to maintain the same polarity in all thethree phases, or their connections are madeinadvertently, without ascertaining theircorrect polarity. It is also possible that on areconnection, such as at site, whilereassembling the modules of a switchgear ora controlgear assembly, such an omission ismade. It is therefore advisable that the polarityof the CTs be ascertained at site beforecommissioning the equipment, such as aswitchgear or a controlgear assembly or aswitchyard utilizing a few CTs.D.C. voltage test to ascertain the polarity: Asimple procedure to ascertain this is indicated inFigure 15.37. A low reading d.c. voltmeter isconnected across the CT secondary windings anda battery of 6-12 V through a switch across theprimary. On closing the switch, the meter needlewill give a momentary flicker. If the polarity iscorrect, the flicker will be positive on connectionand negative on disconnection. For HT CTs,mounted on transformer bushings, it isrecommended to short-circuit the maintransformer secondary (LV) windings to reducethe overall impedance of the transformer to1 1Sl s2(1) Single ratio CT (2) CT with an intermediatetapping on the secondary windingC. C9s1 s2 1s1 Is, 2% 2s2(3) CT with a primary windingin 2 sections which may beconnected in series orparallelFigure 15.36(4) CT with two secondaty windings,each with its own magnetic core.A CT wound in different combinationsachieve an appreciable deflection of the voltmeterneedle.(ii) Power frequency withstand test on secondarywindingsThe secondary windings should be capable ofwithstanding a rated power frequency, shortdurationwithstand voltage of 3 kV for 1 minute.(iii) Power frequency withstand test betweensectionsThis test is applicable when the CT's primaryand secondary windings have two or moresections. Then the section in between will becapable of withstanding a similar voltage as notedin item (ii) above.(iv) Inter-turn overvoltage testThis test is performed to check the suitability ofthe inter-turn insulation to withstand the high-0PrimaryFigure 15.37at sitePlVoltmeterorAmmeterpzSwitchSecondaryterminalsl~l~l~+D.C. source(6V-12V)Circuit to check the polarity of a bar primary CT

Instrument and control transformers: application and selection 151495Relevant StandardsIECTitleIS BS IS060034- 1/199660044-41198060051 1 to 960059/199960060- 1/198960060-21199460076-31198060044- 11199660044-6/ 199260044-2/199760186/199560255-61198860439- 111992Rotating electrical machinesRating and performanceMethod for measuring partial discharges ininstrument transformersDirect acting indicating analogue electricalmeasuring instruments and their accessoriesStandard current ratings (based on Renaldseries R-10 of ISO-3)High voltage testing techniques. Generaldefinitions and test requirementsHigh voltage test techniques. MeasuringsystemsPower transformers. Specification forinsulation levels and dielectric testsSpecification for current transformersGeneral requirementsMeasuring current transformersProtective current transformersRequirements for transient performanceApplication guide for voltage transformersSpecification for voltage transformersGeneral requirementsCapacitive voltage transformersGeneral requirements for measurement andprotection (all voltages)Electric relays, requirementsLow voltage switchgear and controlgearassemblies. Type-tested and partially typetestedassembliesSummation current transformersApplication guide for capacitor voltagetransformersSpecification for control transformers forswitchgear and controlgear for voltages notexceeding 1000 V4722/1992325/199611322l19901248 - 1 to 9207 1- 1/19932071-3/19912026-1 and 3/19912705part- 1/1992part-2/19922705 part-3/19922705 part-4119924201119914 146/19913 156.1119923 156- 1/19923156-2/19923156-3119923156-4/19923231 and 3842-8623-1/19936949119735547/199112021/1987Relevant US Standards ANSUNEMA and IEEEBS EN 60034-1/1995BS 6184/1992BS 89-1 to 9BS 923-1/1990BS EN 60060-2/1995BS 171-311993BS 7626/1993BS 762611993BS 772911995BS 7625/1993BS 772911995BS EN 60255-6/1955BS EN 60439- -1/1994IEEE 4/1995ANSYIEEEC57.13/1987ANSInEEEC.57.13.21199 1Standard techniques for HV testing.Instrument transformers (CTs and VTs)-Requirements.Conformance test procedure for instrument transformers (CTs and electromagnetic VTs)Notes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

15/496 Industrial Power Engineering and Applications Handbook(VIvoltage developed in the secondary circuit in theevent of an accidental secondary open circuit onload. The inter-turn insulation of the windingsshould be capable of withstanding an inter-turnovervoltage of 4.5 kV peak across the completesecondary winding. The test may be conductedby keeping the secondary winding open circuitedand applying a primary current less than or equalto the rated primary current for 1 minute, sufficientto produce a voltage at the secondary terminalsequal to 4.5 kV (peak).Power frequency withstand test on primarywindingsThis test is same as for item (iv) (a), under typetests.NoteA repeat power frequency test, if considered necessary, must beperformed at 80% of the prescribed test voltage. See also Section14.5.3(vi) Partial discharge measurement(vii) Verification of accuracyThis test is the same as under type test item (vi)Special testsThe following additional tests may be conducted whenconsidered necessary:(i) Chopped lightning impulse test. Refer to IEC60044- 1(ii) Measurement of the dielectric dissipation factor(tan a, applicable to only liquid immersed primarywindings, rated for 110 kV and above.NoteFor lower voltage systems, say, 2.5 to 10 kV, measurement ofdielectric loss factor tan 6, along similar lines, to those recommendedfor motors and discussed in Section 9.6.1 is advisable as a processtest to monitor the quality of an HV insulation system, during thecourse of manufacture and using the Fame value for future referencewhen checking the quality of insulation when energizing or duringa field test.List of formulae usedDifferential ground fault protectionCurrent setting of the relay,Zpf = n (N x Z, + Is,) (15.1)Ipf = minimum fault current through the primary rcquiredto trip the relayn = turn ratio of the CTsZ , = magnetizing current of the CTs corresponding tothe V,Z,, = relay current settingN = no. of CTs falling in parallelTo limit the peak voltagev, = 2Jz ,'V,(V, - v,, (15.2)Vp = peak voltage across the relayV, = theoretical maximum CT secondary voltage acrossthe relay circuit at the maximum internal faultcurrentV, = knee point voltageFurther readingPmtecrive Relays and Application Guide, GEC Measurement, GeneralElectric Co. Ltd, Stafford, Uk.

161497Captive(Emergency) PowerGenerationucr set 1613V4

Captive (emergency) power generation 16/49916.1 IntroductionIt is common practice to provide a standby emergencysource of supply at all important installations such aslarge factories, railways, airports and other essentialservices. This is usually achieved with the use of a captivediesel generator (DG) set (Figure 16.1). Here we brieflydiscuss these machines, their characteristics and selectionfor a required application. We also consider schemesthat are commonly used to start a DG set and run itindividually or insparallel with an existing source ofsupply, which may be another DG set or an infinite bus.16.2 DG setThis comprises the following partsEngineThis is the main prime mover (PM) for the generator andmay be a gas, petrol or diesel engine, depending uponthe availability of fuel. In the discussions below, weemphasize a diesel engine, being used more commonlyfor captive power generation.The control of power output of a generator is obtainedthrough this PM only. It has a drooping characteristic onload, as shown in Figure 16.2. These characteristics areused to control the fuel supply to the engine through aspeed-regulating governor, which controls the poweroutput of the generator.The difference in the speed of the engine at no loadand full load is termed the speed droop, and is expressedas a percentage of the no-load speed, i.e.Speed droop or speed regulation,AN= x 100%whereAN = speed droop or speed regulation (%)No = speed of the engine at no load (r.p.m.)N, = synchronous speed or speed of the engine at fullload (r.p.m.)The droop is maintained at around 3-5% by the leadingmanufacturers. The lower the droop, the better will bethe performance of the engine on load. SinceNs -fAf-ANand therefore, the smaller will be the fluctuation in thefrequency of the generated power. But for parallelFigure 16.1Diesel generator set (Courtesy: Kirloskar Electric)

16/500 Industrial Power Engineering and Applications Handbook32750 -P8of the engine0-025 50 75 100%Load --+Figure 16.2 Typical speed-load characteristics of a 1500r.p.m. engine with 4% droopoperation of the generator, a higher droop will meanbetter load sharing. Refer to Examples 16.2 and 16.3.GovernorThis senses the speed of the machine and performsextremely fast and accurate adjustments in the fuel supplyto the PM. In turn it regulates the speed and the output ofthe PM within predefined limits, depending upon thedroop of the PM. The governor may be a mechanical(manual), hydraulic or electronic (automatic) device.The governor can be set to make the machine run at aconstant speed, even on load variations, with extremelyquick and almost instantaneous speed control, and thusmaintain a near-zero AN. In a parallel operation they canalso control load sharing automatically and accurately.Power grids, receiving power from different sources, areextremely susceptible to frequency variations. Even asmall Afof the order of 0.5 Hz, may cause the system totrip. A fast-actuating governor with low response time(as low as 0.5 second) can overcome such a situation byquickly regulating the speed of the PM.GeneratorThese may be of two types:1 Rotating armature These have a rotating armatureand a static field excitation system. The output fromthe armature is taken through the sliprings.2 Static armature or brushless alternators Thesehave a rotating field excitation system and are nowused more commonly compared to the conventionaltypes noted above, particularly in the medium andlarge ratings. Both generators are self-excited and havea self-regulated excitation system. For the mainparameters and general operating conditions, refer toBS 4999-140.16.3 Operating parametersThe following are some important operating parameters:16.3.1 Residual voltage for self-excitationThe armature of the machine will normally have a residualvoltage of around 8 V (for LT machines) across theterminals when running at the synchronous speed. If not,as when the generator is operated after a long shutdown,a d.c. voltage of 12 V can be applied through a batteryfor a few seconds to obtain the required residual voltage.16.3.2 Operating PFSmall generators such as those used for captive powergeneration are seldom used as synchronous motors orsynchronous condensers. To save on the cost of machinestheir field system is generally designed for 0.8 p.f. lagging,unless designed for another application for a differentp.f. The generator output is also defined at 0.8 p.f. laggingand rated in kVA. The 0.8 p.f. so selected is in consonancewith the average p.f. at which a power system would beoperating generally. The maximum kW rating of themachine is therefore defined by kVA x p.f. The operatingp.f. plays a vital role in the selection of the machine. Itis desirable that the load which the generator may haveto feed has a p.f. of at least 0.8 lagging and more, but notbeyond unity. The machine may not perform well at p.f.slower than it is designed for, as well as in the leadingmode, becauseAt lower p.f.s the field system is required to beoverexcited, which may cause excessive heating ofthe field windings.At lower p.f.s the machine will deliver even less thanthe theoretical output (< &. V’ I. cos 6) due to higher12R losses, which will remain the same while the activecomponent ( I. cos 4) will be reduced correspondingto the lower p.f. See also Section 23.3.In the leading mode the field system will be ineffective.When this is required the manufacturer must beconsulted. In the leading mode when the machine issuitable to operate such, the voltage will improve andthe machine will operate in the underexcited mode.While the field winding will now be less stressed, theleading p.f. is not healthy for the machine and theequipment connected on it because(a) The capacitive mode will cause an overvoltageacross the machine windings during a switchingoperation (Section 23.5.1) which may damage themparticularly the end turns.(b) In the leading mode the harmonics, when presentin the system, will magnify and further distort thevoltage and current waveforms. The windings ofthe machine are therefore more stressed due tosuch spurious overvoltages (vh). For vh, see Section23.5.2(A) and equation (23.1).Thus, in the leading mode the machine tends to becomeunstable. It is therefore mandatory to operate the machinewell within its stability region, i.e. between 0.8 p.f. laggingand unity, unless it is also designed for a leading mode.Every machine has its own operating parameters as shownin Figure 24.9. To obtain its best performance, it must beoperated within these parameters.

~ ReactiveCaptive (emergency) power generation 161501Generally, machines up to 1000 kVA are designed forthese parameters. Generators used for hydropowergeneration and operating on smaller heads may alsosometimes be required to operate in a leading mode ab asynchronous condenser to improve the p.f. of the system.This may happen when the water head in the reservoirfalls below its minimum required level and is not capableof generating the required minimum power. When usedin these conditions the field system has to be designedfor both lagging and leading modes.Region '6'---+16.3.3 A generator as a synchronousmotor or a condenserWhen a generator is designed for a leading p.f. (in theunderexcitation mode) it can operate as both a synchronousmotor and a synchronous condenser. The machine is nowself-starting and does not require a prime mover.As a synchronous motor The machine is run primarilyto drive a mechanical load and is operated at thesynchronous speed and at unity p.f. The efficiency isnow better than that of an induction motor. Except inassisting the system by consuming power at unity p.f.,it does not help the system to improve its p.f.As a synchronous condenser The machine is operatedwithout any mechanical loading. It is now used primarilyto supply leading reactive power to improve the systemp.f. The machine can now be operated with continuouslyvariable leading reactive power, with the help of anautomatic voltage regulator (AVR) and a quadraturedroop control (QDC), and can even improve a varyingp.f. of the system caused by fluctuating loads. Themachine can be made to operate up to 0.1 p.f. leading(as a capacitor) without affecting its stability. The activepower, however small, is delivered at 0.1 p.f. and isconsumed to feed its own no-load losses.NoteWhen de\ired, the machine can also be designed as an inductor tosupply lagging reactive kVAr, with the help of AVR and QDC. Itwill serve little purpose. if used both as a motor and a condenser.Refer to Figure 16.3. illustrating the trajcclury of the current phasor,I,. For the current I, at a p.f. cos I$ (leading),- Active power output per phase = V.fi cos I$. andpower output per phase = V.li sin I$We can observe that at higher leading p.f.s (region Ain Figure 16.3), while the active power will rise, thereactive power will be too low to contribute effectivelytowards the p.f. improvement of the system. This is alsotrue at lower leading p.f.s (region B). While the reactivepower will now rise, the active power will be too low, toperform any mechanical duty, the more so when part ofit will be consumed to feed its own no-load losses. Theusual practice, therefore, is to use it either as a synchronousmotor (region A) or as a synchronous condenser (regionB) at a time. Extra kVA can be built in to the machinepartly 10 irnprove the system p.f. and partly to performthe required mechanical duty. The following examplewill illustrate this.Low reactive ,/ 1,power, Low activeI,Figure 16.3condenserUsing a machine either as a motor or as aExample 16.1Consider a process plant having a connected load of 15 000kW and a running load of 12 500 h.p. at almost 0.65 p.f.lagging. Let a few large induction motors aggregating 2000h.p. be replaced by as many oversized synchronous machines,with the purpose of improving the system p.f. in addition toperforming the motor's duties.Considering an average efficiency of the induction motors asq = 92% andp.f. = 0.9 laggingEquivalent kW = 2ooo 0.7460.92= 1622 kWandkVA rating at 0.9 p.f.1622= - 0.9= 1802 kVAReactive kVAr = 1802 sin cos-' 0.9= 1802 x 0.4359= 785 kVAr (lagging)For ease of illustration, all these parameters have beendrawn in the phasor diagram (Figure 16.4). To select therating of the synchronous condensers, consider their averageefficiency,q = 95% andp.f. = 0.3 leading1622:. Total kVA rating =0.95 x 0.3= 5691 kVA. Say, 5700 kVAand reactive kVAr = 5700 . sin cos-' 0.3= 5700 x 0.95= 5415 kVAr (leading)

16/502 Industrial Power Engineering and Applications Handbookmr* lnm'2Figure 16.4 Improving system PF with the use of asynchronous condenser2 kc90 d r-T 5415kVAr1 2 785kVAr-ADraw the phasor diagram for the actual load also at 12 500h.p. at 0.65 p.f... Active load = 12 500 x 0.746= 9325 kWand kVA =9325~0.65= 14 346 kVAand reactive load, kVAr = 14 346 sin cos-' 0.65= 14 346 x 0.76= 10 903 kVAr (lagging)After replacing these large induction motors with as manyoversized synchronous motors, while the active load at 9325kW remains the same, the reactive load of induction motorsat 785 kVAr will be eliminated and instead a leading reactiveload of 5415 kVAr will be added. The net compensationtherefore will be= 785 + 5415= 6200 kVArleaving an uncompensated reactive load= 10 903 - 6200= 4703 kVArAnd an improved loading = 49325' + 4703'as against 14 346 kVA= 10 444 kVA9325and improved p.f. = -10 444= 0.89 as against 0.65It is, however, recommended for better control and machineutilization that when the load's demand is for constant-speedoperation, this must be met through separate synchronousmotors at unity p.f. and the p.f. must be improved separatelythrough synchronous condensers with variable field excitation.If the synchronous condensers are employed only to improvethe system p.f. from 0.65 to, say, 0.9 lagging, then the ratingof the machines can be determined as follows:Total active load = 9325 kW:. kVA at 0.9 p.f. lagging9325= - 0.9= 10 361 kVAAnd reactive kVAr = 10 361 sin cos-' 0.9= 10 361 x 0.436= 451 7 kVAr:. Compensation is required for= 10903-4517= 6386 kVArThen the kVA of the synchronous condensers, operating at0.1 p.f. leading and having an efficiency of 98%- 6386 (Figure 16.5)0.98 sin cos-'0.1- 63860.98 x 0.995= 6549 Say, 6550 kVAThe AVR may be designed for variable duty, for automaticcontrol of the reactive power, to the required level throughfeedback control systems. The machines will now operateonly as synchronous condensers without performing anymechanical duty.16.3.4 Field systemAutomatic voltage regulator (AVR)This device controls the generator and maintains a steadystatearmature voltage automatically within the predefinedlimits. It also serves to control the reactive kVAr loadingduring a parallel operation or when the machine is beingused as a synchronous condenser for reactive powercompensation through a quadrature droop control (QDC)as noted below.Quadrature droop control (QDC)This is a scheme introduced in the AVR circuit to adjustthe reactive power (kVAr) of a machine during a paralleloperation or when it is being used as a synchronouscondenser. It prevents a reactive circulating current, I,,

Captive (emergency) power generation 16/503through the armature windings when the two machinesare operating in parallel (Figure 16.21), or controls thereactive component within the required limits whenoperating on an infinite bus (Figure 16.26(a)). The limitof such circulating currents is defined to be within 5% ofthe rated current of the machine. The QDC circuit isillustrated in Figure 16.6. The basic purpose of the circuitis to detect the content of kVAr being fed by the machinewhen operating in parallel. The AVR, in turn, adjusts thefield excitation to vary the operating p.f. of the machineto control the kVAr to the required level.If the machine operates at p.f.s lower than 0.8, theexcitation requirement of the machine would be high.This is a case of overexcitation and may damage thefield system. For such operations, the machine wouldrequire a double derating, depending upon the p.f. atwhich it has to operate, one for the lower p.f., due to thereduced active component of the current (I cos 4) andthe second because of higher excitation demand. In suchcases the manufacturers may be consulted. A correctivestep, however, would be to improve the system p.f. byinstalling a few capacitor banks to achieve a system p.f.between 0.8 and 1.0.CorollaryAt low p.f.s the generator operates at a low level ofexcitations (armature reaction demagnetizing). During afault, therefore, when the p.f. of the circuit falls it willalso cause a fall in the excitation level and in turn in theterminal voltage. A low voltage, however, would reducethe severity of the fault.Similarly, when the machine is required to operate atleading p.f.s, the field system has to be redesigned, asthe normal field system, which is designed for laggingp.f.s, will be ineffective, as discussed below.16.4 Theory of operationThe reference voltage Vyh is obtained from any of theFigure 16.5 Improving system PF with the use of asynchronous condenserIQDCr---------~~ControlItransformerI-VnvR = Resultant voltage to AVRR = Burden resistance in AVRFigure 16.6 A normal quadrature droop circuit (QDC)two phases of the armature windings through a controltransformer, T,. The current reference, I,, is obtainedthrough a metering CT provided in the third phase andwired to the burden resistance, R, of the AVR. Based onthis, the AVR takes corrective steps by altering the fieldexcitation of the machine to adjust the load (I,) p.f. sothat the reactive load (kVAr) supplied by the machinewill fall within the pre-set value during a parallel operation.The kVAr being supplied by the generator will influencethe setting of the AVR in the following ways:Unity PF (Figure 16.7(i)) The reference current i,produces a voltage V, across the burden resistance R,which adds to the reference voltage, VYb, at right angles.It causes a very small change in the AVR terminalvoltage, V,.0.8 PF lagging (Figure 16.7(ii)) At 0.8 p.f. this isonly marginally more than the above.Zero PF lagging (Figure 16.7(iii)) Now the marginis substantial as the reference voltage VYb and V, addlinearly, being in phase opposition (a case ofoverexcitation). The resultant voltage at the AVR willbe high and current lagging, causing the field excitationto rise substantially and adjusting this to the requiredvalue to make the machine share the desired kVArloading. The QDC in the AVR circuit thus helps maintaina p.f. balance and also the kVAr loading by varying itsexcitation.Zero PF leading (Figure 16.7(iv)) V, now subtractsfrom VYb (a case of underexcitation). The resultantvoltage at the AVR reduces but the armature reactionis magnetizing as the current is leading. The AVRtherefore, is redundant as it has no control over thegenerated voltage and leads to instability. Thegenerators, designed for a lagging p.f. operation,therefore, are not suitable for leading p.f.s. Whereversuch a need arises, the generator field system has to bedesigned for a leading p.f.I

16/504 Industrial Power Engineering and Applications HandbookY'Yb(i) p.f. - unityVAvR - Marginally morei"7IVAVRd- I / jcos@=o(iii) p.f. - ZeroVAvR - Maximum(ii) p.f. - 0.8 laggingVAvR - Slightly moreI/Y(iv) p.f. - LeadingVAvR - DiminishesFigure 16.7 Phasor diagrams for QDC, referred to thesecondary side of the control transformer16.5 Guidelines on the selection ofa DG set@= 0.8Several factors are important in the performance of agenerator, and not the service conditions alone, asdiscussed for motors, in Section 1.6. In addition to serviceconditions, the operating power factor plays a significantrole in the selection of a DG set, as noted above. Thefollowing p.f. conditions may occur in practice, dependingupon the type of loads connected on the system. Refer toFigure 16.8.which is normally selected for only 0.8 x kVA (kW = p.f.x kVA). At higher p.f.s, while the active component OAof the revised current OA" remains the same as at 0.8p.f., the generator will draw a lower current at I& andoperate underloaded, while the engine will be fully loadedand the DG set will deliver its optimum rated output. Fora resistive load of 100 kW, for instance, operating atunity p.f., a generator rated for 100 kVA is not suitablebecause the kW rating of the engine for a 100 kVA setwill be only 0.8 x 100 i.e. 80 kW.:. The DG set (at least the engine) for such a load mustbe selected for 100 x 1, i.e. 100 kW or 125 kVA.16.5.3 PF leadingA generator is normally designed for a lagging p.f., andis not suitable to operate at leading p.f.s. At leading pfs,the armature reaction becomes magnetizing, and the fieldsystem loses its control over the terminal voltage. Duringa leading p.f., therefore, the terminal voltage rises rapidly,as the field system becomes redundant, rendering themachine unstable. The field current reduces as the leadingp.f reduces. When the leading p.f. reduces further, sayfrom OC to OC' (Figure 16.8), the field current (excitation)reaches almost zero, a condition known as self-excitation.This is the critical point of the AVR when it will losealmost all its control. Any further reduction in thc p.f.will cause overexcitation, causing the voltage to risesteadily without any control. Leading p.f.s thus causeinstability and rapid voltage rises.The above situation is, however, found when themachine is run singly. When it is operated in parallelwith another source, the field excitation will not influenceTrajectory for maximumgenerator current, OA'16.5.1 PF lagging and less than 0.8The active component of current OB reduces and sodoes the kW rating of the generator. When the generatoroperates fully loaded, the engine operates underloadedbut the reactive component BB' rises, requiring a higherlevel of excitation (a higher field current), which maycause damage to the field system, which is designed for0.8 p.f. lagging. It would thus require a derating of thecomplete machine indirectly. Generally, the derating maybe of the order of 15% for a p.f. of 0.6 and 8% for a p.f.of 0.7. For exact deratings consult the manufacturer.16.5.2 PF between 0.8 and unityThe maximum kW rating of the DG set remains at 0.8 xkVA, even when the p.f. of the system it is feeding risesto unity. This is due to a limitation in the engine capacity,OA - Maximum active component of generator = O.S/,AA - Maximum reactive component of generator = 0.6/,(if O A = /,)Figure 16.8Effect of PF on the output of a DG set

Captive (emergency) power generation 16/505the output voltage. as it will adjust only its operating p.f.(Section 16.9.1 B-2 and D-2). A difference in excitationbetween the two machines when operating in parallelwill cause a circulating current I, (Figure 16.15). It willadd to one machine and subtract from the other, dependingupon which machine is relatively overexcited, comparedto the other. The QDC, as discussed in Section 16.3.4,will play the required role to limit the reactive loading ofthe two machines within permissible levels, when such asituation arises.16.6 Types of loadsThe p.f. varies with the type of load. Here we discuss thelikely loads, their behaviour and precautions that mustbe taken when selecting a DG set for various types ofloads.16.6.1 Linear loadsSuch as motors (not really but can be considered sobeing balanced loads) heating loads, capacitor andincandescent lighting loads etc. Of the linear loads, thefollowing will require special consideration.a surge suppression device in the rectifier assembly toprotect against such voltage surges.16.6.2 Non-linear loadsLoads of rectifier, thyristor, UPS (uninterrupted powersupply) and battery chargers etc.Fluorescent tube lights.All such loads that have non-sinusoidal waveforms.All these loads cause a distortion in their pure sinewave current waveforms. These distortions are termed‘harmonics’ and such loads ‘non-linear loads’. Thyristorand rectifier loads fall into this category and affect thesinusoidal waveform of the generator voltage and distortit. Typical voltage distortions in the output supply ofa machine as a consequence of current distortions,caused by such non-linear loads are shown in Figure16.9 for a particular type of generator whose windingshave a pitch factor of 2/3 (which suppresses thirdharmonic quantities). The magnitude of distortions mayvary with other machines having different windingparameters. The distortion causes:Motor loadsThese are highly inductive and cause heavy inrush currentsduring a switching operation, depending upon the typeof starting being adopted. The generator being a highimpedancemachine, application of such high currentloads will cause a heavy voltage dip, which can be up to1 O-lS% of the rated voltage, until the AVR acts to restorethe pre-set value of the terminal voltage. Althoughrestoration of the rated voltage is rapid due to a lowresponse time of the AVR (of the order of 0.5-1 second)the generator is required to supply a much higher current,at least for this duration. Consideration must be givenfor all such loads on the system and their switching currentsat the time of selecting the generator rating. A generatoris generally suitable to carry a momentary but infrequentcurrent inrush up to 2 to 2.5 times its rated current forhardly 10 seconds.Resistive loadsWhen the p.f. of the load is more than 0.8, select anenginc of a higher rating, as noted earlier.Capacitive loadsThese are used for improving the p.f. of the system.When they are installed in the circuit, they must beswitched OFF with the loads, to avoid a leading p.f. Butit must be ensured that the p.f. of the system does not fallbelow 0.8, otherwise it would overload the generator aswell as its field system.During a switching operation, the capacitive reactanceof capacitors and inductive reactance of the generatormay resonate and cause a voltage surge in the fieldwinding, as a result of transformer coupling between thearmature and the field. This may damage the bridge diodes.The practice of the leading manufacturers is to provide-U ’25 50 75 100Courtesy: Crornpton Greaves% Non linear loadWinding parameters;Pitch factor - 2/3Sub-transient reactance, x: - 12%Figure 16.9 Voltage distortion caused by different typesof thyristor drives (with the use of IGBTs, however, thedistortion would be greatly reduced)

16/506 Industrial Power Engineering and Applications Handbook1 Excessive heating of magnetic cores, as a result ofharmonic frequencies due to hysteresis and eddycurrent losses (equations (1.12) and (1.1 3)).2 Overloading of neutral conductors due to thirdharmonic currents flowing through the neutral. Thesize of the neutral must be increased by 150-200%of the normal size in such cases, depending uponthe severity of the harmonics.3 Higher voltage stresses, V, (equation (23.1)) maylead to dielectric breakdown.4 Ageing of insulation of the generator and otherequipment connected to such a system.5. Noise and resonance problems (Section 23.5.2(C))in the electrical distribution and communicationnetworks.6 Problems during a parallel operation due to distortedquantities of currents and frequencies. Nonsinusoidalcurrent will cause a non-linear impedance(mostly reactive) voltage drop on load in thegenerator windings. The terminal voltage will alsocontain harmonic quantities. It is possible that evena fast-responding AVR may not be able to maintainthe required sinusoidal voltage waveform in thewhole cycle and cause circulating currents. Generatormanufacturers recommend that not more than 40%of such loads be connected on the system at a time,when the remaining loads are linear. To suppressthe third harmonics, use of YlA or A/Y transformersis common to feed such loads and filter out thethird harmonics. For more information refer toSection 23.6.1.For all such loads, the machines will require a derating.For appropriate selection it is therefore essential first todetermine the content and magnitude of harmonics andthen to consult the manufacturer for the selection of themachine. To determine the magnitude of harmoniccontents. refer to Section 23.7.16.6.3 Special loadsLoads of welding sets, which have intermittent duty cycles,are mostly single phasc, cause low p.f. and stress thegenerator windings intermittently. Unbalanced loads, suchas single phase loads, distributed unevenly cause currentunbalance and low p.f.All loads that may overstress the gcnerator windingsfall in this category and call for special consideration.The manufacturer must be consulted for the right choiceof machine.There are no ready-made formulae to deal with such asituation, except experience. It will largely depend uponthe skill of the engineer to make the right choice ofmachine to meet the requirements in consultation withthe manufacturer and also observing certain disciplinesas noted above, such as maintaining the p.f. between 0.8lagging and unity, suppressing harmonics as far as ispractical, and maintaining a balance of loads. It is alsodesirable that neutrals are not interconnected when morethan one machines is operating in parallel to eliminatecirculation of third harmonic currents.16.7 Starting of a DG set16.7.1 Through an auto-mains failure (AMF)schemeThis is a common scheme to bring a standby DG set online automatically on the failure of the main source ofsupply with the help of a battery backed-up ignitionscheme. On the failure of the main supply, generally threeignition pulses are given to the engine to auto-start themachine. If the engine fails to start at three attempts, afurther pulse is blocked. To start, the engine is madeautomatic. The generator excitation or the field currentis also pre-set, which adjusts the generator voltage to therequired level automatically through the AVR. Provisionis also made to start it manually in case the AMF schemefails and to also facilitate routine testing and give it moreflexibility. The scheme may be briefly described as follows(see the control and scheme drawing in (Figure 16.10).Control supplyA battery backed-up d.c. source of control supply isprovided for the AMF panel and engine ignition. Thecontrol scheme, as illustrated, generally consists of a220 or 240 V a.c. source of supply, with a transformerrectifier unit, to provide a 24 or 48 V d.c. control voltage,to charge the battery as required and a battery back-up ofsuitable capacity.Selection schemeAuto or manual selection (switch SW2)A switch SW2 is provided for the selection of the engineto be started in auto or manual modes:Auto mode In auto mode the engine startsautomatically on failure of the a.c. bus voltage, V,,through an undervoltage or bus voltage relay (Relaycode 27) (Figure 16.10). The relay is provided on thegenerator control panel with time delay contacts. Timedelay is provided to allow a pause to the generator ifthe normal supply is quickly restored.ManualmodeLocal controlIn manual mode, the engine can be started and stoppedlocally, through a pair of start and stop push buttonsNos 17 and 18, provided on the AMF panel (Figure16.1 1).Local remote control (switch SW3)This is provided to facilitate the manual start and stop ofthe engine from a remote point such as through a remotestation.Three-attempts startA three-attempts start facility is provided through asequential timer (ST) in auto mode. A starting relay(SR) gives three ignition impulses to the engine. If the

Captive (emergency) power generation 161507engine fails to start at three consecutive attempts, thestarting relay (SR) automatically locks out and emits nofurther ignition impulse. This feature is essential to protectthe engine against a possible hunting and a drain of thebattery.Switching offBus voltage relay (Relay code 27) provides an impulseto the generator trip circuit as soon as normal supply isrestored. The generator falls out of the circuit automaticallyafter a pause of 2- 10-30 seconds, and the engine stops.The relay now also has a delayed feature as it had duringthe start, to allow a pause to the main supply in case themain supply fails quickly again.Lubricating oilAn interlock is provided through a centrifugal type ofpressure switch (PS) to trip the engine in the case of lowlube (lubricating) oil pressure during a run. Since duringa start oil pressure has not built up, a timer, T, is introducedto bypass the trip interlock and avoid a false trip. A relay(contactor), d,, is used to provide lube oil pressureinterlock.Fuel oil solenoid valve (SV)This is to cut off the fuel supply to the engine on a tripor a normal stop of the engine.Speed and voltage controlPush button Nos 22 and 23 may be provided on the AMFpanel and also on the remote panel to raise and lower thespeed and voltage, when required to control the speedcf;) and voltage (E,) of the generator in order for it to besynchronized with another generator or an infinite bus.Figure 16.10 illustrates a typical scheme incorporatingall these features and interlocks.NotrThe above controls can also be achieved more precisely and almostinstantly by applying the same logic with the use of PLCs, asdiscussed in Section 16.13.16.8 Protection of a DG set16.8.1 Alarm and annunciation (Figure 16.10)Some or all of the following alarms and annunciationsmay be provided on the AMF panel, depending upon thesize of the DC set and the type of load it has to feed toforewarn the operator. and prevent a trip:Mechanical faults(a) Engine fails to start(b) Engine overspeeds(c) Engine high water temperature and high lube oiltemperature(d) Low level of fuel(e) Low lube (lubricating) oil pressureElectrical faults(a) Overcurrent alarm(b) Battery charger problems(c) Overcurrent and ground fault(d) Reverse power supply(e) A fault condition in the differential scheme(f) Breaker tripFigure 16.11 illustrates a general arrangement of such anAMF panel and the above alarm and annunciationprovisions.16.8.2 Electrical protection and meteringThe following protections and metering are considerednecessary for a DG set:For small ratings (say, up to 150 kVA)Short-circuit protection ~ through HRC fusesOvercurrent-cum-short-circuit protection, throughthermal overcurrent and built-in short-circuit releases.Figure 16.12 illustrates a general protection andmetering scheme.A kVAr meter to indicate the reactive power of thecircuit and to take corrective steps. when necessary.For medium ratings (say, up to 1000 kVA)(a) Voltage-controlled overcurrent and ground faultprotection (relay Code 51N/64) By a three-polevoltage controlled definite time delay relay with acurrent setting of 50-200% and a time delay rangeof 0.1-1 second. The relay is typical with acombination of an overcurrent-time relay and anundervoltage (u.v.) supervision unit. The purpose vfthe U.V. unit is to detect an U.V. due to a ground faultin the machine. During a fault the relay switchesover instantaneously to a lower response current valueand effects faster tripping. Typical characteristics canbe- An overload characteristic say, 50-200%, matchingthe thermal withstand characteristic of thegenerator. This operates during an overlvadcondition when the generator voltage is normal.- A ground fault characteristic determined by theU.V. unit monitoring the generator voltage. Duringa ground fault, causing an u.v., the sameovercurrent characteristic is reduced to anotherset value, resulting in a faster tripping, as desired.NoteThe relay can also be selected for a phase fault.(b) Reverse power supply protection (Relay Code 32)To detect a motoring action by a single-pole reversepower relay with a time setting of 2-10 seconds for

16/508 Industrial Power Engineering and Applications Handbookpressure interlockby - p a s svAMF scheme (continued)Figure 16.10 Typical control and annunciation scheme for an AMF panelboth active and reactive powers. The basic principleof the relay is to compare the current in any onephase with the voltage output across the other twophases. When the operating parameters exceed thepre-set values, the generator is switched OFF after apre-set delay in case the operating conditions havenot improved.(c) Differential protection (Relay Code 87) To detecta stator phase-to-phase fault by a three-pole differentialprotection relay, current setting 1040%. For schemediagrams, refer to Section 15.6.6( 1).(d) Overcurrent alarm (Relay Code 51A) This maybe provided to warn the operator of a likelyoverloading on the generator to prevent a trip andtake corrective action, by promptly shedding a partof the load. The current setting of the OK alarmrelay is kept lower than the setting of the OK triprelay, say, 105% for the alarm and 110% for the trip.A single-pole overcurrent relay with a setting of 50-200% and a time delay of 2.5-5 seconds may beconsidered ideal. Figure 16.13 illustrates a typicalprotection and metering scheme and Figure 16.11shows the provision of these relays on the panel.For a large power stationThis is a complex subject and requires detailed engineeringand application of the various protection schemes for

Captive (emergency) power generation 161509EOS pI I II I I€OH3 1 LLSI I IPI fI lI1444, I73a,aLa,0a,._mW2-w.- aLc LY al2m11-VAMF scheme (continued)Contact development(SW - 2)Contact development(SW - 3)KnobpositionManual Open CloseFigure 16.10 (Contd)different operating conditions. We briefly discuss belowa typical protection scheme that is generally adopted ata power station for the protection ofThe generatorThe power station's auxiliary supply system andThe generator supply system to the switchyard.See Figure 13.21 showing a general power generationand transmission system. For brevity we hae restrictedour discussion to the protection of the main equipmentat the generating station such as the generator (G), theunit auxiliary transformers (UATs) and the generatortransformer (GT), the protection of which is almost of astandard nature.The protection of the remaining system is a matter ofsystem design and appropriate application of the protectivedevices available depending upon system requirements.These three items have been taken out of the scheme ofFigure 13.21 and redrawn in Figure 16.14 for more clarity.Normal alarm schemes generally provided(i) Stator cooling water- Inlet temperature high- Outlet temperature high- Rectifier temperature high

16/51 0 Industrial Power Engineering and Applications HandbookI2nNLFor AMF panelLegend1. Hooter2. Annunciator3. Control supply 'on'4. Breaker 'on'5. Breaker 'off'6. Breaker 'trip'7. D.C. ammeter8. kW meter9. D.C. voltmeter10. A.C. ammeter11. Frequency meter12. A.C. voltmeter13. Ammeter selector switch14. Auto/Manual switch-Sw215. Local/Remote switch-Sw316. Voltmeter selector switch17. Start-D.G. set18. Stop D.G. set19. Test P.B.For generator control panel20. Accept P.B.21. Reset P.B.22. Engine speed raise23. Engine speed lower24. Breaker control switch25. Generator voltage raise26. Generator voltage lower27. R-phase28. Y-phase29. B-phase30. kWh meter31. kVAr meter32. Over current & ground faultrelay (51 N/64)33. Reverse power relay (32)34. Overcurrent alarm relay(51~)35. Differential relay (87)36. lncomer breaker feeder37. Lockout relay (86)38. Under voltage relay (27)Figure 16.11 Typical general arrangement of an AMF panel(ii) Hydrogen gas- Cooler hot gas temperature high- Cooler cold gas temperature high- Common cold gas temperature high- Machine gas temperature highR Y B NPower diagramFigure 16.12 Protection and metering scheme for a smallerrating DG set, say up to 150 kVA(iii) Machine temperature- Collector air idout temperature high- Generator field temperature high- Stator slot temperature high(iv) Bearing oil temperature high(v) Vibration level, etc.NoteSee also Section 12.8 for more details.Generator protectionThe generator protection arrangements usually employedare as follows:(a) Impedance protection (Relay Code 21G) Impedanceprotection or distance back-up protection (alsotermed external fault back-up protection). It isemployed to protect the generator from feeding thefault on a short-circuit on an adjacent system, whichmay be hanging due to the failure of its own systemprimary protection.(b) Anti-motoring protection through reverse powerrelay or low forward power interlock (Relay Code32-G) The generator acts as a motor when the

Captive (emergency) power generation 16/51 1W? PNLR Y B NFigure 16.13 Typical power diagram illustrating electricalprotection and metering for medium rating DG sets up to500 kVAwill operate as a synchronous motor and will drivethe prime mover (turbine). While the generator maynot be harmed by a motoring action, the prime movermay become overheated and damaged.If the field excitation is also lost, the generatorwill run as an induction motor again driving the primermover as above. As an induction motor, it will nowoperate at less than the synchronous speed and causeslip frequency current and slip losses in the rotorcircuit, which may overheat the rotor and damage it,see also Section. 1.3 and equation (1.9). A reversepower relay under such a condition will disconnectthe generator from the mains and protect the machine.(c) Loss of excitation or loss of field protection (RelayCode 40-G) Loss of excitation results in a loss ofsynchronism and causes operation of the generatoras an induction machine. It will result in the flow ofslip frequency current in the rotor windings, whichmay damage the rotor, as noted above.When an induction motor runs beyond thesynchronous speed, it behaves like an inductiongenerator and feeds power back to the supply system(Section 6.15). Below synchronous speed it behaveslike an induction motor and draws power from thesupply system. This protection trips the generator insuch an eventuality and protects the machine.(d) Negative phase sequence protection (Relay Code46-G) Unbalanced loads cause unbalanced currents,resulting in negative phase sequence currents. Thesecurrents are detrimental to the generator stator androtor windings (Section 12.2~). Their effects on agenerator are similar to those on an induction motorand the protection of a generator is the same as thatof a motor. Since unbalanced loads are naturalphenomena in a power network, they cannot beeliminated due to single-phase loads (mostly domesticand public utilities) and many industrial loads thatuse static drives and distort the sinusoidal wave form(Section 16.6.2), hence this protection.(e) ‘RTD’ protection through warning of windingoverheating (Relay Code 49-G) A generator is theheart of a power station and must not trip onmomentary overloads. It is therefore essential thatthe operating conditions of a generator are closelymonitored and normal operating conditions quicklyrestored, as far as possible, to eliminate a trip. The‘RTDs’ that are placed in the slots of the statorwindings can give a warning through an audio-visualalarm scheme and the operator can restore the normalconditions, if possible. The generator is allowed toexceed its rated current momentarily under thefollowing conditions:power input to the prime mover (steam supply to theturbine in a thermal station) falls below the no-loadlosses of the generator. This is a corollary to aninduction generator, discussed in Section 6.21. Thegenerator being still connected to the grid will drawpower from the grid through its stator windings. Sincethe field excitation does not change, the generatorTime (seconds) 10 30 60 120Generator current 226% 154% 130% 116%NoteOvercurrent protection is normally not provided in generator andgenerator transformers to save the machines from likely outages onmomentary overloads. In the event of overloading, the normal practiceis to shed some of the loads on the transmission network.

16/51 2 Industrial Power Engineering and Applications HandbookR Y BCTs(Differentialprotection)cProtection ogeneratortransformer(GT)-GCTs(DifferentialDrotection)IIImPIIIIIIII'l--t--t-mII.R I It y iMain BusjiTap off,5 auxiliarytransformersno.i&2CTs(Protection)CTs(Differentialprotection)CTs(Differentialprotection)J 3 3>f + 7I' i iIIGenerator neutformer (high impedancegrounding)G'Protection of generatorFigure 16.14 (Confd)

Captive (emergency) power generation 16/513Generator (G)214 : Impedance protection or distance back-up protection324 : Anti-motoring protection or low forward power protection404 : Loss of excitation or loss of field protection464 : Negative sequence protection49-43 : ‘RTD’ protection59-43 : Over voRage/Hz protection64-G & : Generator stator ground51-G fault protection64-GF : Field ground protection78-G : Pole slipping protection or out of step protection814 : Abnormal frequency operation87-G : Differential protectionGenerator Transformer (GT)49-GT : ‘RTD protection51N-GT: Ground fault protection63-GT : Gas protection64R-GT: Restricted ground fault protection87-GT : Differential protectionMain bus-IUnit auxiliary transformer (UT)49-UT : ‘RTD’ protection51-UT : Over-current protectionSIN-UT : Ground fault protection63-UT : Gas protection87-UT : Differential protection64R-UT : Restricted ground faultprotectionProtection of unit auxiliary transformers no. 1 and 2 (shown only one transformer)Figure 16.14 A typical protection scheme in a large generating station

16/514 Industrial Power Engineering and Applications Handbook(f) Overvolts per Hertz protection (V/Hz) (Relay Code59-G) Per unit voltage divided by per unit frequency,called V/Hz, is proportional to flux in generator andgenerator-transformer cores. (See also Section 1.2,equation (1.5) and Section 6.2.2). Excessive flux cancause serious overheating of metallic parts and meltingof generator core laminations. (See also Section 1.6.2.(Aiv) eauations (1.12) ~, and (1.13).) Excessive flux~ I ,(V/Hz) ;an be caused by voltage regulator failure,load rejection or excessive excitation. Even thoughV/Hz is more likely to occur when generator is offline,it can also happen when generator is on-line.Normally a two-stage protection is provided forV/Hz protection. If V/Hz exceeds 1.18 per unit, aprotection relay trips the generator within twoseconds (typical). For V/Hz 1.1 to 1.18 per unit, aprotection relay trips the generator within 45 seconds(typical).Generator stator ground fault protection (RelayCodes 64-G and 51-GN) The neutral of the generatorstator normally operates at a potential, close to ground,through a high-impedance single-phase distributiontypetransformer the secondary of which is shuntedthrough a resistor that has a voltage relay device 64-G connected across the resistor. In the event of aground fault, current flows through the resistor andthe relay 64-G operates due to a voltage drop acrossthe resistor. An overcurrent relay, 5 1-GN, connectedto the current transformer in the secondary windingof the grounding transformer is used as back-up for64-G relay.Exciter protection Protective devices are incorporated,as required for protection of the specificexcitation system components and are supplied bygenerator manufacturers as an integral part of theexcitation equipment. Some of the protective devicesareField ground protection (Relay Code 64-GF)The generator field winding is electrically isolatedfrom ground. Therefore one ground fault in therotor windings will usually not damage the rotor.However, two or more ground faults in the rotorwindings will cause magnetic and thermal imbalances,which may result in localized heating andmay damage the rotor. Protection provided in theexcitation system is used to trip the generator inthe event of a ground fault in the rotor windings.Field overheating protection This is a part ofexcitation system to prevent the field from prolongedovercurrent through an alarm or a signal. Itwill trip the generator only if it is absolutelyessential. The field winding can carry momentaryovercurrents for a short period, expressed in termsof field voltages, as noted below:Time (seconds) 10 30 60 120Field voltage 208% 146% 125% 112%Pole slipping or out of step protection (Relay code78-G) This is a protection against loss of synchronism.It can occur as a result of steady-state, dynamicor transient instability or due to loss of excitation orout-of-phase synchronism. Out-of-step (phase)operation can result in high peaks of current andfrequency, and cause winding stresses, pulsating torqueand mechanical resonance. All these may endangerthe generator windings. Loss of excitation relay mayprovide detection of such a situation, but it may notbe reliable under all conditions. An MHO-type relayis generally used to protect the generator from sucha situation.0) Abnormal frequency operation (Relay Code 81-G) For a generator connected to a power system,abnormal frequency operation may be the result of asevere system disturbance. An isolated unit, however,can operate at a low or high frequency, due to anincorrect speed control adjustment or a malfunctioningof the speed control device.The generator can tolerate underfrequency operationfor long periods, provided that the loads and thevoltage are proportionately reduced. It can also tolerateoverfrequency operation, when the voltage is withinthe permissible limits.If a system disturbance requires extra generatingcapacity, the generator may drop speed, reducing thefrequency proportionately. The under-frequency relay81-G will detect such a condition and operate asystematic load shedding in a programmed mannerin order to meet the load demand. If the abnormalcondition persists, the generator is taken off the mainsupply system.(k) Differential protection (Relay Code 87-G) Differentialprotection to detect stator phase-to-phase fault.(I) Surge protection This is similar to that for motors(Sections 17.10 and 18.8).Protection of unit auxiliary transformer (UT)49-UT RTD protection for warning of windingoverheating (see also Sections 12.7 and12.8 ).51 -UT Overcurrent protection5 1 N-UT Ground fault protection63-UT Gas protection through a Bucholz relay64R-UT Restricted ground fault protection87-UT Differential protection to detect a phaseto-phasefault.Prntertion of generator transformer (GT)49-GTOvercurrentprotection5 1 N-GT63-GT64R-GT87-GTRTD protection for warning of windingoverheating.For overcurrent protection, comments arethe same as for generator protection, 49-Gitem ‘e’Ground fault protectionGas protection through a Bucholz relayRestricted ground fault protectionDifferential protection to detect a phaseto-phasefault.NoteRelay Code numbers are according to ANSI designations. For detailsrefer to Appendix 2.

16.9 Parallel operation16.9.1 Theory of parallel operationThe performance of a generator varies with its operatingconditions. For instance, it is different on no-load andon-load when running in parallel with another generator.Similarly, it is different when running in parallel with aninfinite bus. For a better understanding and more claritywe analyse below, the performance of an incominggenerator when it is required to run in parallel with anothergenerator or with an infinite bus. Consider two generatorsGI and G2, operating in parallel, as illustrated in Figure16.15, having the following parameters:PM,, PM2 - Prime moversZ, (R, + JXd,) = Impedance of generator 12, iR2 + JXd2) = Impedance of generator 2NoteFor more clarity. particularly for a better illustration, we considerimpedances rather than only synchronous reactances. although theresistances. being small. are usually ignored.ZL- (RL + JX,) = impedance of the loadE,, E,E'VI2.fl. f2.fhI,. I?= e.m.f.s generated by the two generators= residual voltage across the internalcircuit of the same phases of the twogenerators, giving rise to circulatingcurrents.= bus voltage= frequencies of the two machines= frequency of the bus= Load sharing by the two generatorsCaptive (emergency) power generation 16/51 5I, = total load currentI, = circulating current within the two generators'circuit.Consider voltages E, and E2 being equal and in phase,a condition necessary for running the two generators inparallel, i.e.(16.1)( 16.2)( 16.3)In what follows we consider GI as the incoming machineand G2 as the machine already running. connected to thebus.When running in parallel with another machineA Performance on no-loudBy changing the driving torque or power inputSuppose the driving torque of GI is increased. GI willaccelerate faster than Gz and E, will advance E?vectorially by AO, the magnitude remaining same(Figure 16.16(a)) and- -EI - EL =E, ( 16.4)E, will cause a circulating current, I,. The circuitdiagram is drawn in Figure 16.16(b) where(16.5)Since Z, and Z2 are highly inductive. I, may beconsidered lagging E, by almost 90", as shown inFigure 16.16(a) and-Vh = i?~ - ic 21- - -= E? i I,. Zz(we have considered the voltageof the G2 bus as V,)If the two generators are identical. so that Z, = Z2,then.Figure 16.15bLoadRL\Common bus atfrequency QParallel operation of two generatorsE, + E2v, = 2 i 16.6)InferenceGI will generate an excess power compared to G,.Therefore while GI will operate as ;I generator, G,.receiving power from GI, will operate as a synchronousmotor. Since GI is overloaded compared to G?. it willtend to retard, and G2, receiving power from GI, willtend to accelerate. The net effect would be that bothgenerators will tend to synchronize on their own onceagain.By changing the excitation (field current) In theabove case if the field excitation of the incominggenerator, G1, was increased, causing the terminalvoltage E, to rise to E,', so that E; > E,. Then alsoa residual voltage, E,, would appear across the internal

16/516 Industrial Power Engineering and Applications HandbookG2 slowerIG1 faster\\\\. . . . . . .v, = €2 + lc ' 2 2NoteMotoring actionGenerating actionI.I. II. . I.. IMagnitude of iEl and E2 remains unchanged* EC is not appearing at 90" with lc because of lc . R, which is negligible but wehave considered it to be significant for better illustrationFigure 16.16(a) Effect of increasing the driving torque on no-load (magnified representation).:/E2IIIZlz2operate as a generator and G2 as a synchronous motorandb € C_IFigure 16.16(b) Residual voltage across the internalcircuit of generators GI and G2circuit of the two generators. The only difference wouldbe that now they will be in phase but different inmagnitudes, i.e.E; -E? = E,I, =E,ZI + 2 2I, will lag E, by almost 90' and lead E, by almost 90"(Figure 16.17). The net effect will be a demagnetizingarmature reaction for GI, tending to weaken its fieldand diminish E;, whereas for G2 it would be amagnetizing armature reaction, tending to strengthenits field and enhance E2. Both machines would thustend to synchronize once again. GI will now alsoIf z, = z,E{ +E,then V, = ~2(16.7)B Performance on-load1 By changing the driving torque or power inputThe performance of two or more generators, runningin parallel, on-load is not very different than analysedon no-load. A generator running in advance comparedto the others would share a higher load than thoserunning behind. The performance can be analysedmore easily by referring to Figure 16.16(a).For more clarily, consider any of the two generatorsto be identical and running at the same speed,generating the same voltage and sharing an equal loadof, I at a lagging p.f. cos $, i.e.G2 -Synchronous motorG1 - Generator

Captive (emergency) power generation 16/517When G, and G2 are sharing equallyE, = E?, z, = z2.I, = I? = I, andFigure 16.18I, + 1: = 21, as illustrated in Figure 16.18E, = €2Say, GI is made to run a little too fast compared toGZ, by increasing its input power. A residual voltage,Ec, will appear across any two identical phases, causinga circulating current Z,, lagging E, by almost 90". Thephasor diagram will change to Figure 16.19, which issimilar to the phasor diagram of Figure 16.16(a) exceptthe additional current phasors I; and I;. GI, whichis running ahead, will operate at a better p.f. than theother and share an extra load, equivalent to thecirculating current I,, such that it is 7 - 7 , + 7,. It1:is possible when GI is operating at a higher p.f., i.e.cos > cos 4 and Ti- = 1, - 7,. It is possible whenG2 is operating at a lower p.f., Le. cos q2 < cos 4 and-I; + I: = 7, + i2, i.e. the total load current remainingthe same.An unequal load sharing may be desirable whenthe generators are of unequal rating or for some reasonsone of them is to be loaded more or less than theother. The desired load sharing can be achieved byvarying the phase shift between the excitation voltagesE; and E; of the two generators as illustrated in Figure16.19. The higher the phase shift 6, the higher will bethe load shared by the generator that is running inadvance, compared to the other. In what follows weconsider G, to be faster than GZ. The phase shift canbe altered by changing the engine input power. Theload sharing can theoretically be determined with thehelp of speed m-load or drooping characteristics ofthe prime movers as illustrated in Figure 16.20.Consider the speed-load (drooping) characteristicsof the two machines as shown in Figure 16.20. Forease of illustration, the slopes have been exaggerated.Normally they are within 4% of the rated speed, asdiscussed earlier. When both machines are loadedequally, the total load may be defined by the load lineAA', at the bus frequency,fb, When the power input toPMI is increased, so that the drooping curve AO shiftsto curve BO', it shifts the load line AA' also to BB', sothat the total load shared by the two machines willstill remain the same. The load shared by GI is nowmore than before at P,: so that PI'> 6, and by G2less than before at Pi, so that Pi < P2 . The generatorsnow operate at a higher system frequency, ,fhl. If the,E1 = f2 trajectoryGenerator - GIPrimer mover - PM1...1,' = /I + IC + cos 4, > cos @- p.f. improves.. . .li = 12 - /C- cos o2 c cos 9- p.f. worsensWhen G, is running faster than G2, it shares more load-Load G,LoadG,--tLoad line AAP1 = P2 = Pand,'---\P, + P, = 2P'.\Load line BB'Load line CC'P; + Pi =2PP;' + P;' =2PFigure 16.19 Variation in load sharing by varying thedriving torque on loadFigure 16.20 Drooping curves of two machines,illustrating load sharing when running in parallel

16/518 Industrial Power Engineering and Applications HandbookI, . z, = /, ' z,IC Y-' ' 'ab = active component remains the same Le.,I. cos g= 1;. cos g;= 1;. cos g;bc, bc, and bc, = only reactive components vary\\R, = R2 = R (assumed)-.---*21 remains the sameFigure 16.21 Variation in the load currents with a change in the field excitation on loadsystem frequency is required to remain constant at fbit will be essential to reduce the power input to PM2,so that the drooping curve of PM2 shifts below to0°C' and the load line to CC' at the original frequencyfb, delivering the same total load yet again. The loadshared by G2 is now still lower than before.The load sharing by the two machines can thus bevaried by shifting the drooping curves of the primemovers by altering their power input.2 By changing the excitation (field current) Nowthat the engine input is not varied, there is no variationin the load sharing by the two machines. It is thebasic theory of change in the field excitation. Since achange in the excitation causes a variation in thegenerated e.m.f.s (E, and E2), the variation in voltagecauses a corresponding rise or fall in the reactivecomponent of the current. EI cos Q will remain thesame, except the variation in the copper losses (I,'R),which may vary the load sharing, marginally up ordown, depending upon whether it supplies I, orreceives. For instancc, at higher excitation, the e.m.f.will rise and so will the load current, but at a lowerp.f. the generator will have to feed extra losses andthus share a marginally lower load than previously.To illustrate the above, consider Figures 16.15 and16.21. Assuming that both machines were equallyloaded, an increase in excitation of G, will increaseE, to E;, which will tend to increase vh as noted inequation (1 6.7). A corresponding decrease in excitationof C2 from E2 to E; can, maintain the same level ofvh, as illustrated in Figure 16.21. The phasor differencebetween E,' and E; Le. E,, will give rise to acirculating current I,-, lagging E, by almost 90°, asnoted above and illustrated in Figure 16.21. The loadsharing can now be computed as follows.Asbuming that the above change in the excitationcauses the following changes in the basic parameters(i) I, to Z; at a p.f. cos 4,'(ii) l2 to 1; at a p.f. cos @;then 7; = 7, + 7,I;cos@;=I, COSQ, =IcosQI; = I, - r,I; cos 4; =I2 cos $2 = ICOS QThe active components will thus remain same, andonly the reactive components I( sin 4' andI; sin $2' will vary.NoteIf the machines were loaded unequally before making a change inthe excitation, the same ratio of loading would continue even afterthe change in the excitation of both machines provided that V,remains the same, i.eIf E, rises further, it will do so at still lower p.f.s and El will haveto be further reduced to maintain the same V, at yet higher p.f.s tomaintain the same level of active component and vice versa.ConclusionWith the change in excitation, only the reactive power,kVAr and the terminal voltages El and E2 altered can bewithout altering the active components of the load currentsor the power shared by the two machines. As discussedin Section 16.3.2, a generator is designed for a particularp.f. (0.8 lagging), having a defined value of kVAr. A

Captive (emergency) power generation 16/519reactive power higher than rated would cause reactiveoverloading of the machine and cause reactive circulatingcurrents. It would seriously affect the performance ofthe machine. There is therefore only limited scope invarying the excitation level of a generator.When synchronizing with an infinite busA bus maintaining constant vh and fh, irrespective of avariation in the loads connected on it, is called an infinitebus. An incoming generator would cause no change inthese parameters. unlike, when the two machines wererequired to run in parallel. The performance of an incomingmachine would therefore now be different than previously.Earlier a change in its input power or excitation wouldvary the output, frequency or the voltage of the othermachine. There is no such influence on an infinite bus.Performance on no-loudBy changing the driving torque or power inputThe situation as noted earlier would occur. If thegenerator GI accelerates faster and its voltage El getsahead of vh vectorially, the magnitude remaining thesame, Figure 16.16(a) would generally apply, withoutthe phasor E?, (Figure 16.22):_ - _El - Vh = E,EI, = 2 (for an infinite bus, Z2 may beZIconsidered to be zero) andv, = E, - r, '2,If E, is slower and falls behind V, vectorially, then GIwill operate as a synchronous motor and receivereactive power I,?Z,, from the infinite bus (Figure16.23) since v, = E, + i,Z,.By changing the excitation (field current) The samesituation would occur as noted above, if the excitationof GI is increased from El to E;. ThenE; - v, = EcF, v, = El + IC . z,GI slower and El falling behind V,(Operating as a synchronous motor)Figure 16.23 Effect of changing the driving torque ofGI on no-loadI, would lag E, by almost 90" and lead V,, by almost90" (Figure 16.24(a)). The net effect would be ademagnetizing armatlire reaction fnr GI, tending toweaken its field, diminish E( and synchronize it withthe infinite bus once again.If E, is reduced to E; then,Vb - E; = E,The direction of current would reverse, and I, wouldlead E;. GI now, instead of supplying the reactivepower to the bus, would receive reactive power andoperate as a synchronous motor. Theoretically, thearmature reaction, being magnetizing, would tend tostrengthen its field, enhance E; and help to synchronizeGI once again with the infinite bus (Figure 16.24(b)).But such a situation may not occur unless the generatoris designed for a leading p.f., as noted earlier.When, however, E, is equal to V,, and in phase, E,will be zero and there will be no current from GI to+ /C(a) When E; > Vb, G, operates as generatot(b) When €; < V,, G, operates as synchronous motorG1 faster and E, ahead of Vb- - - -v, = E, - I , .z,Figure 16.22 Effect of changing the driving torque of GIon no-loadvb-I L;I \I €1(c) When E; = V,, Ec = 0 and IC = 0, G1 floats on the busFigure 16.24 Effect of varying the excitation on no-load

16/520 Industrial Power Engineering and Applications Handbookthe bus or vice versa. The machine will only float onthe bus and the PM, will be supplying only themechanical losses, (Figure 16.24(c)).\D performance on-load(Considering GI as the incoming machine and referringto Figure 16.15)1 By changing the driving torque or power inputFixed parameters vb, fb, Z, = 0, and Z1Variable parameters Zl and cos &, while El will havea fixed magnitude but variablephasor disposition.When the power input to PMI is increased, the outputof GI increases. Since El is constant at a particularexcitation, it changes its phasor location only withrespect to V,. With a change in power input, therefore.El traverses through a fixed trajectory as shown inFigures 16.25 a and b, and with it changes its loadangle, el, load current, ZI and p.f. cos 4,. We haveconsidered four possible conditions, to define theperformance of the machine, under different levels ofpower input:When El is ahead of V, At a load angle 8, theload current, 11, will lag v b by an angle 41 (refer toFigure 16.25(a)). II is still considered to be laggingEf by almost 90", although it may be better onload.With the increase in the power input, E, will advancefurther and improve its p.f. At one stage, the p.f.will become unity and beyond this it will start leading(Figure 16.25 (b)). Incidentally, the maximum p.f.is achieved when the load angle el becomes 90",which is also the limiting stage, beyond which itwould become an unstable region, as the exciterwould cease to exercise any control over the voltage.At this point, refer to parameters E[, EA, I:, @{ ande; as 90".Any condition beyond unity p.f., i.e. GI > 0, wouldmean ZI leading and can compensate the reactivepower of the system and improve its p.f. The machineis now called a synchronous condenser (capacitor),which besides supplying power to the main bus,will also improve the system p.f. The above,however, is only a theoretical analysis. A generatordesigned for 0.8 p.f. lagging is not suitable to operateat a leading p.f., as the excitation system wouldcease to exercise any control over the voltage. Thevoltage rises rapidly beyond unity p.f. as a result ofpositive armature reaction (Section 16.4). When agenerator is required to operate at leading p.f.s, itsfield system must be designed for leading p.f.s.When, however, the power input to PMI is reduced,GI will gradually offload. Consider the situationwhen El, falls in phase with vb. (Refer to parametersE:, E,", 0: = 0 and I: at 4' in Figure 16.25(b).)GI will now feed no power to the bus, nor receiveany power from it. GI is now termed as floating on11"Figure 16.25 Performance of generator GI by varying itsdriving torque on load(b)the bus and the PM, supplies primarily only its noloadlosses.For the sake of argument, if the power input isreduced further, say by removing PMl totally fromthe generator, E;' will fall behind vb and I; willlead vb. (Refer to parameters E:, ET, e; and

Captive (emergency) power generation 1615212/,"at @: in Figure l6.25(a).) The machine will nowoperate as a synchronous motor rather than as agenerator and will absorb reactive power from thebus. Since the generator operates once again atleading p.f.s, the same condition will apply as notedabove.By changing the excitation (field current)Fixed parameters V,, jj,, Z, = 0 and Z,Variable parameters E,, I, and cos 4,The same theory would apply as discussed ahovein the case of two generators. Since there is no variationin the power input to PM,, the output of generator G1will remain the same, except for the marginal variationin the copper losses as noted earlier::. I, cos 4, = I ; cos @; = I('cos @('= constantIn other words, for the same bus voltage, V,, the activecomponent of the current for GI would remain thesame while the reactive component I, sin 4, I,' sin4' or /('sin 4" and therefore the reactive power (kVAr)would continue to vary. A change in excitation willchange E, and its load angle (Figure 16.26(a)) andconsequently will change I, and its p.f., cos @,. Thefollowing Dossibihties mav arise:When-dl operates at uhty p.f. is the most idealcondition. The generator will now deliver itsmaximum power at the least current value. Themachine is least stressed for its best performance.GI is ahead of the bus and is only sufficiently excitedsuch that cos @I = 1 andE, = v, + 7, gd, or E, cos 8, = V,When GI is overexcited, E, rises to E; and themachine starts to operate at lagging p.f.s, so thatE,' cos 8; > V,, and cos 8,' > cos 8,.When GI is underexcited, El will reduce to E('and the machine will start to operate at leadingp.f.s, so that E('cos e('< V, and cos 8: < cos 81.All these conditions are illustrated in Figure 16.26(a).In all three cases, the active component of current,OA, remains the same. A higher current than theactive component, either lagging or leading, is aloss component. It is desirable to operate GI asclose to unity p.f. as possible to keep this componentat its lowest. The variation in the generator current,I, versus field current, is shown in Figure 16.26(b).When the current II is leading, the machine absorbsreactive power and operates as a synchronouscondenser and in addition to supplying its activepower to the system also improves the system p.f.But, as noted above, for operating a generator as asynchronous condenser, its field system has to bedesigned accordingly.16.1 0 Procedure of parallel operation16.10.1 SynchronizationBefore switching an incoming generator on an existingIp f leading region 1 p f lagging regionFigure 16.26(a) Phasor diagram- I -p.f leading region I p.f lagging region(under excitation) (over excitation)Field current vs armature currentFigure 16.26(b) Variation in the load current of G, with thechange in the excitation on loadsource, which can be another generator or an infinitebus, it is essential to first fulfil the following basicconditions, to avoid a possible voltage or current transientcondition which may occur and cause electrodynamicforces in the generator and damage its armature or affectadversely other machines, connected on the system orthe bus system itself.1 The phase sequence of the incoming machine mustbe the same as that of the existing source (Figures16.27(a) and (b)).2 The terminal voltage, E,, of the incoming machinemust he almost the same as that of the other machine,E2 or the bus, V, (Figure 16.27(c)), i.e.E, = E2 or V,

16/522 Industrial Power Engineering and Applications HandbookB ..Y 'vb4(a)Phasor diagram ofthe existing sourcevbYI(b)Phasor diagram of theincoming generatorEl= VbandAV=O,AO=O(c)Both voltages inphase and equalHuntingAny error beyond permissible limits in AV, Af, or ABmay cause a shock and disturbance to the incomingmachine and the existing system. AB and Afmay causehunting which makes the rotor swing even beyond itssynchronous speed as a result of its own inertia. But thisdevelops an opposing torque too which retards theseoverswings. Thus, while hunting would attenuate on itsown, the machine would supply and absorb large amountsof power alternately during the course of hunting. As themechanical forces are proportional to the square of thecurrent drawn by the machine at a particular instant (F 0~12, equation (28.4)), they may be associated with largecurrent transients. The duration of such a situation wouldplay a very significant role in the stability of the systemand the safety of the incoming machine. This situationmust be dealt with as quickly as possible. Hence theimportance of keeping these variables as low as possible,and reaching a stable state in only two or three cyclesafter synchroniza-tion. Thus such reversals of mechanicalforces of the rotating masses are more important, ratherthan the magnitudes of the torques that the machine willhave to sustain. In large power stations, where such forcesmay assume very high proportions, because of large sizedmachines, they may even upset the normal supply systemby severe power fluctuations, outage of the system oroverstressing of the incoming machine through its statorand the rotor. For the significance of Afor AB refer toFigures 16.31(a) and (b). To achieve the requiredconditions of synchronization the following proceduremay be adopted.Phase displacement A8causing residual voltage ECEl = El,,,. sin wit. andVb = vb , sin @f.o1 and wb are the angular speeds (2nf)in rad./sec. of the two voltages(4Time (wJ --CFigure 16.27 Phase sequence and phase displacementand AV= El - E2orAV= E, - Vbwhere AV = difference in magnitudes of the twovoltages. Permissible variation: AV = within 1% ofV, or E2.3 The frequency of the incoming machine,f,, must bealmost the same as that of the other machine, f2, orthebus,fb. Permissible variation: Af= within 0.15 Hz.4 To check the phasor difference, if any, between E,and E2 or vb to check AB (Figure 16.27(d)). A6 givesrise to residual voltage EC, which is responsible forthe circulating current I,. (Section 16.9, equation(16.5)). Permissible variation: AB = within 7".To check the phase sequenceThis can be checked with the help of a phase sequenceindicator. This is a simple instrument that houses a small3@ motor, which rotates a pointer connected to the motorthrough a gearbox. The direction of rotation of the pointerwill determine the phase sequence of the system.To check the terminal voltage and frequencyFigure 16.28 suggests a simple method to measure theterminal voltage El and the frequencyf, of the incomingmachine:1 The voltage can be lowered or raised by varying thefield excitation through the AVR of the machine. Anydifference in the voltage of the incoming machinewith the voltage of the existing system will result inAV and AB.2 The frequency can he lowered or raised by changingthe speed of the engine by varying its power input,i.e. by controlling its fuel supply (diesel in a DG set,water in hydro and steam in thermal generation). Anyvariation in frequency will also cause a residual voltage,E,, and Figure 16.27(d) would apply,wherewhenE, = Elmax, sin colt - E2max, sin Yt etc.EL = E2 or EbE, = El max (sin w,t - sin Yt)and the frequency across the incoming generator breaker

Captive (emergency) power generation 16/523ExistingIncomingsystemMIGG,R Y Rl Y- . Fl’iExisting sourceVillOV laaaaalControl VTIncominggenerator-‘Common -w v,/110 vVb/fb : Existing voltage and frequency-meters€,if, : Incoming voltage and frequency-metersSYN . SynchroscopeControl VTFigure 16.28 Circuit to check Vand fwill be (.f\ +f2)/2 when it is operating in parallel withanother generator. In an infinite bus, the bus frequency,fh, will prevail (Section 16.9.1 C and D).To check the phase differenceSome methods to do this are noted below:1 Voltmeter method2 Dark lamp method3 Synchroscope method1 Check synchronizing relay5 Auto synchroniration1 Voltineter method (Figure 16.29)The incoming machine is brought up to its synchronousspeed by controlling the torque or power input to theengine and the voltage to the required level with the helpof the AVR (Figure 16.6). When the line voltage of theincoming machine and the other source are the same andfall almost in phase with each other, i.e. when thecumulative effect of AV, A0 and Affall within permissiblelimits, the three synchronizing voltmeter readings willread almost zero. This is the condition when thesynchronizing switch or the incoming machine breakercan be closed.2 Dork lamp method (Figure 16.30)This is a simpler method to check the phase displacementbetween the incoming and the existing voltage. Normallytwo lamps are connected in series to make them suitablefor 480 V as shown to withstand the maximum line voltage,in case the two voltages fall 180” apart. This voltage, AVcan rise toIncoming generator (E1, f,)Sl - Double throw switch0 - Generator breakerV - To compare voltages of the two sources-T-Synchronizingvoltmeters to@J&&measure AVand fc (A0and Af)Figure 16.29 Synchronizing by the voltmeter method-TIncoming generator (E,, fl)S, - Double throw switch0 - Generator breakerV - To compare voltages of the two sourcesR’iExisting sourceSynchronizinglamps to indicateAVandEc (A0 and Af)Figure 16.30 Synchronizing by the dark lamp methodor 480 V for a 4 IS V system1This is a better method, for it can compare both frequencyand phase displacement of the two voltages, as Af’wouidalso result in E, and is reflected by the flickering of thelamps. If the frequencies ,fl and fb are not equal, thelamps will flicker at the rate Afi i.e. (f, -,fb) times per

16/524 Industrial Power Engineering and Applications Handbooksecond. For example, for a difference of 2 Hz the lampswill flicker twice every second.In both the above cases the following will occur whenthe generator breaker is closed and the frequency of theincoming machinefl is not equal to the frequency of theexisting sourcefb. For ease of explanation, we considerthe dark lamp method. In the voltmeter method it is thevoltrnekr needle that will flicker rather than the lights.Ec will appear across each phase (Figure 16.31(a))and the lamps will flicker at (f,-fb) per second. Whenthe phasors El and Vb are closer, the brightness will bethe least (Figure 16.32) and when they are widest apart,it will be the maximum (Figures 16.31(b) and 16.33).To attain E, = 0, i.e. A0 = 0, it is essential that E, andV, are in phase.There will thus be a momentary bright and mostlydark period every second. During the dark period thetwo voltages are either in phase or are very close toeach other, such that the residual voltage E, is inadequateto make the lamps glow. When the lamps are dark,somewhere during the middle of the dark period, isthe ideal instant when the incoming machine can besynchronized with the existing source by closing itsbreaker. This is the condition illustrated in Figure 16.32.The rest of the performance of the incoming generatoris explained in Section 16.9.1, while dealing with thebehaviour of a generator during a parallel operation.There is, however, a disadvantage in the dark lampmethod as when E, is, say, less than-60% of E2 or V, (anincandescent lamp does not glow at less than 30% of itsrated voltage and there are two such lamps in series) thelamp will stay dark. A slight misjudgement may closethe generator breaker when E, may be large enough (upto 0.6 times or so of vb) across the generator windings tocause a dangerous situation, as discussed earlier.Before closing the breaker it is also essential to know,whether the incoming machine was running a little toofast or too slow. As discussed in method 5, the incomingmachine must run a little too fast compared to the machinealready running or the infinite bus while beingsynchronized. When so it will share a part of the existingload, no sooner it is synchronized and fulfils the purposefor which it is being synchronized with the existing source.To ascertain this before synchronizing, increase the speedof the incoming machine. An increase in the flickeringof the lamps will indicate a faster machine, while adecrease will indicate a slower machine. Paralleling of aslower machine is not desirable, as it may draw powerfrom the existing source and operate as a synchronousmotor rather than a generator and defeat the purpose ofparalleling.3 Synchroscope methodThis is the simplest method of all. A synchroscope is aninstrument that compares the speed (Le. the frequency)of the incoming machine,f,, with the frequency of theexisting source,fb (Le. Af) and is in the form of a rotatingpointer which rotates at a speed proportional to thedifference in the two frequencies. If the incoming machineis running a little too fast, it will have a clockwise rotationand if it is a little too slow, it will show a counter-clockwiserotation. The incoming machine will be synchronizedonly when it is a little too fast and the pointer rotatesclockwise at a very slow speed, i.e. when the frequencyof the incoming machine is too close to the other sourceand a little too high. The machine may be quicklysynchronized at the instant when the pointer moves throughits zero axis, as illustrated in Figure 16.34. For an accurateclosing, an indicating light is normally provided in theinstrument that glows at every zero, the instant at whichthe machine must be synchronized. Figure 16.34, suggestsa simple connection diagram of a synchroscope.IIiWhen the voltages are fallingphase apart (180” out of phase),fc = max at frequency AfVb = Existing source atfrequency, f,,E1 = Incoming source at\I II Iv” ‘4, - frequency AfFigure 16.31(a) Residualvoltage ,Ec across each phasewhen f, is not equal to fbFigure 16.31(b) Magnified representation of a frequency error Af

Captive (emergency) power generation 16/525- -Ec= El - v, = oFigure 16.32 Residual voltage is zero when El is in phasewith V,When El = V,Ec = E, + rb = 2VbFigure 16.33 Residual voltage is maximum when El is phaseapart with V,S, - Double throw switchG1 - Incoming machine (€l,f,)0 - Generator breakerV - To compare voltages of the two sourcesFigure 16.34Connection of a synchroscope4 Check synchronizing reluy (Relq Code 25)The purpose of this relay is to check the accuracy ofmanual synchronizing. It basically checks Af, AVand A0between GI and the existing source, as is done by asynchronizing monitoring relay. When the quantities fallwithin the permissible limits, the relay unlocks the G,breaker and only then may the machine be synchronizedmanually. It is a preferred practice to use such a relay asa safety precaution for manual synchronizing, to doublecheckthe pre-set quantities of Afi AV and A0, to preventinadvertent synchronization, particularly because of thelead time required to close the breaker after a closingimpulse (Table 19.1), which a manual mode may not beable to assess so accurately and the machine may besynchronized just before or just after the moment ofsynchronization.5 Auto-svnchr~nizatiolz~i~uti~i~In the preceding text we have discussed manual methodsof synchronizing two sources, more common for smallerinstallations, say, up to 500 kVA. For large installationsand power generating stations such procedures may notbe practical for manual methods may not be so accurateat the instant of synchronization. They are therefore likelyto cause a fault condition due to heavy circulating currents,as a result of higher A0 or Af than permissible. It mayalso lead to hunting. They are also time consuming.Moreover, the synchronization may be required at timeswhen the operator is not available. For such installations,an auto-synchronizing scheme must be used. This willcomparef, and E, of the incoming machine with that ofthe existing source and automatically control its speedVI) and excitation voltage (E,) to the pre-set values, sothat AJ AV and A0 are within the permissible limits atthe instant of synchronization. The recommended limits forsuch parameters, as noted above, may be considered asAF = within 0.15 HzAV = within 1 c/r of the rated voltageAQ* = within 7"(*Both of these give rise to higher residual voltage, E(..which may lead to hunting).All this can be achieved with the help of an autosynchronizingrelay, which is capable of monitoring thephase shift, A@, to perform perfect synchronization evenwithout an operator. A normally open (NO) contact ofthe relay is wired in the closing circuit of the interruptingdevice of the incoming machine. The relay sends out anadvance signal to account for the closing time of thebreaker circuit to close this contact at the instant whenAJ AVand A0 fall within their pre-set limits. Such relays,which may be solid-state (IC circuits) or microprocessorbased, are extremely accurate and fast-synchronizing.ImportantBesides the three basic parameters noted above, theincoming machine must also be running a little too fastcompared to the existing source, Le. ,fl > f2 or fh, at theinstant of synchronization. If it is not, it will furtherstress an already overstressed source. When the incoming

16/526 Industrial Power Engineering and Applications Handbookmachine runs a little too slow it will fall behind theexisting source and operate as a synchronous motor,drawing a reactive power from the existing source, ratherthan feeding to it, and thus stress it further. Such a situationis undesirable as the incoming machine is being switchedon the system precisely to relieve the existing source ofits overstress by sharing a part of its load. It is thereforemandatory that the incoming machine must be running alittle faster than the existing source at the instant ofsynchronizing. When it is, the incoming machine willimmediately share a part of the load equal to I, (Figure16.35) to the extent it was too fast. The synchronizingrelays are provided with an inbuilt feature to accomplishthis requirement. (Also refer to Section 16.9.1(A1) and(Cl), Figures 16.16(a) and 16.22.) IfI, = circulating current (load on the incomingmachine)1, or Ib =loading on the existing source then the newloading on the existing source,1; = 7, - ?, in case of another machine G2or I( = ib - in case of an infinite busThe incoming machine can then be loaded as desired.The total sequence of auto-synchronizing a standbygenerating unit with an existing system can thus besummarized as follows:On receiving a closing signal, the AMF panel startsthe prime mover of DC,. Through automatic speedand voltage controls, as discussed in Section 16.7, GIis brought up to its speed and voltage as desired.At this stage, an auto-synchronizing relay (Relay Code25) is brought into the circuit. This relay is suitablefor any size of a generating unit to be synchronizedautomatically with another unit or an infinite bus. Therelay executes three basic functions:1 As a frequency (AB comparator and frequencybalancing or equalizing unit (FNI) This unitcompares the difference in the two frequencies (Af)and controls it through an in-built frequency balancingrelay. The relay sends out a pulse to the motorizedgovernor of PM,, (Figure 16.36) to raise or lower itsspeed to attain the pre-set AL within 0.15 Hz, dependingupon the size of the machine and the flywheel usedwith the PM, The relay can be built into the autosynchronizingrelay or can be a separate unit.2 As a voltage (AV) comparator and voltage balancingor equalizing unit (UNI) This unit compares thedifference in the two voltages (AV) and controls itthrough an in-built voltage balancing relay. The relaysends out a pulse to the AVR of GI through a motorizedpotentiometer, which can be introduced in the QDCcircuit (Figure 16.6) to raise or lower its excitationautomatically to attain the pre-set AV, generally withinI% of the rated voltage.The relay can be built into the auto-synchronizingrelay or can be a separate unit.3 The auto-synchronizing relay monitors AV, Af andphase shift (A@, between the two voltage phasors. Inother words it monitors the residual voltage, E,. Italso ensures that GI is slightly ahead of the existingsource at the instant of synchronization.When these parameters are brought within the pre-setvalues, the relay transmits a closing impulse to theswitching circuit of GI, a little in advance to account forthe closing time of the breaker circuit. The breaker isthus switched at almost the same instant, when all theparameters fall within the pre-set limits. The total closingtime may be a few ms (say, 150-300 ms), dependingupon the closing time of the breaker and any other coilsor relays incorporated into the switching circuit whichmay add to the closing time. (Also refer to Table 19.1,for the closing time of breakers.)16.1 1 Recommended protection fora synchronizing schemeIn addition to the normal protection, as suggested inSection 16.8.2 the following is also recommended:E2 or V,Figure 16.35 Sharing of load by an incoming machine duringsynchronizing, when running faster than the existing source1 A reverse power relay (RPR) (Relay Code 32) Thisis meant for both active and reactive powers. If theincoming machine is slow, it will operate as asynchronous motor and draw power from the systemrather than feed it, not a desirable situation. The relaywill isolate the machine before it causes an overloadingof the existing source.2 Field failure relay (Relay Code 40) This monitorsthe exciter field current and detects the loss of fieldsupply or reduction in the field current.3 Mains decoupling relay (Relay Code 78) Thisprotection is applicable when the captive generator ishooked up with the main bus. The relay trips theincoming breaker instantaneously on failure of themain supply. Otherwise on rapid restoration of the

Captive (emergency) power generation 16/527I I I IiMl-M4 - Circuit breaker 32 - Reverse power relay 2 - Time delay relayGI-G3 - Generator LBR-A - Load balance relay active UNI - Voltage balance relayPM,-PM, - Engine LBR-R - Load balance relay FNI - Frequency balance relayGOV - Governor reactive SY1 - Synchronizing relay51-54 - Selector Sw. 78 - Main decoupling relay SFG-3 - Bus frequency relayAVR - Automatic voltage27 - Bus voltage relayregulatorFigure 16.36 Typical block diagram illustrating an auto-synchronizing and load-sharingmain supply, as a result a fast auto-reclosing, themachine may be subject to mechanical damage or afault condition. A small delay would throw the twosources out of synchronization and voltages up to180" apart, causing a residual voltage Ec, up to twicethe system voltage. In such a situation the relay mustoperate even faster than the reclosing time, as fast aswithin 1 cycle (20 ms for a 50 Hz system) and trip thebreaker in less than 10 cycles (200 ms) to account forthe tripping time of the interrupting circuit and theinterrupter.Note.\1 For schematic diagrams of relays, refer to the manufacturers'catalogues.1 The numbers given in parentheses denote the relay numbers asin the ANSI standard.16.12 Load sharing by two or moregeneratorsThe variable parameters in a generator that may affectits performance are as follows:Speed of the prime moverFor active load sharing,f-N( 1 ha)and P- N. T ( 1.2)The load sharing of the two generators is thereforedependent on the speed-load (drooping) characteristicsof the prime movers.Excitation voltage For reactive load sharing,P= \?.E.ICOS$

16/528 Industrial Power Engineering and Applications HandbookWe will notice during our subsequent discussions thatE and I cos @ are complementary. Although I andcos @cannot be altered directly, they are both functionsof excitation voltage E. A variation in E will varyboth I and cos@. The reactive or kVAr loading is thusdependent upon the voltage versus reactive loadcurrentcharacteristics of the generators and can bevaried through the QDC. Thus the power generatedor the load sharing by a generator can be altered inthe following ways;Active power (kW) sharing By changing themechanical torque of the prime mover by changingits driving force (fuel supply).Reactive power (kVAr) sharing By changing itsexcitation (field current) that will alter the generatede.m.f., E, in both magnitude and phase displacement,fremaining the same. A mismatch in excitation willresult in reactive unbalancing, causing a reactivecirculating current, which is not desirable. It iscontrolled through QDC by adjusting the droop evenfor identical machines.16.13 Total automation through PLCsWith the availability of PLCs (Programmable LogicControllers, Section 13.2.3) it has become possible toperform all the above controls and protectionsautomatically. The generating station can even be leftunmanned. Such schemes can be adopted for large captivegenerating stations or where a quicker and more accuratepower supply is desirable, such as large process plants(cement, paper, chemicals and refineries) which may havelarge captive generating units (two and more).When there are two or more machines it is also possibleto program the PLCs to select the machine to perform ina particular sequence so that each machine has almostthe same number of hours of operation. This makes iteasy to identify the next machine for routine maintenanceshutdown. A PLC is fully capable of performing thefollowing;To auto-start the machine in the same sequence asdiscussed in Section 16.7.To initiate operation of the next due machine when thepower demand on the existing machine or bus exceedsits rating.To perform all the duties of an auto-synchronizingrelay and monitor and correct AV, Afand A@.To run the incoming machine a little in advance, whileswitching it on the bus, to enable it share a part of theload immediately from the existing machine or thebus.The duty of load sharing between two or more machinesis performed similarly.When the load demand falls, the machine that hasoperated for more hours is stopped first.Any required sequence or programming is possible.The scheme can also be provided with the required- Metering- Protection- Indication- Alarm- Annunciation etc.Even the grounding of the generators can be monitoredthrough this scheme, so that only one machine isgrounded at a time, to avoid circulation of fault currents(Section 20.10.1).Example 16.2Consider two DG sets operating in parallel and having thefollowing parameters:G1Rating 750 kVA 750 kVAp.f. 0.8 lagging 0.8 laggingSpeed at full load 1500 r.p.m. (50 Hz) 1500 r.p.m. (50 Hz)Droop 3 Yo 4 %To determine the load sharing between the two, draw thedrooping curves as shown in Figure 16.37.G2For G1: FF, = no-load speed = 1.03 x 1500= 1545 r.p.m.and frequency = 1.03 x 50or 51.5 Hz.FF, = full-load speed= 1500 r.p.m. at a frequency of 50 Hz.FDA, = full active load= 750 x 0.8= 600 kWSimilarly for G2:FF, = no-load speed = 1.04 x 1500= 1560 r.p.m.and frequency = 1.04 x 50= 52 HZ600 400 200kWuG2e52 Hz-, 200 400 600kWGiFigure 16.37 Determining the load sharing between G1 andG2 with the help of prime-movers drooping characteristics

2-4Captive (emergency) power generation 161529at a frequency of 50 HzFF, = full-load speed = 1500 r.p.m.FoAl = full active load= 750 x 0.8= 600 kWTherefore, the generators would share a load of 1200 kWequally. If, however, the total load is reduced to, say, 1000kW, the loading will differ due to unequal droopingcharacteristics. The revised load sharing can be determinedas follows.Let 61B2 represent the full load of 1000 kW at a frequencyff3. If B,F3 and F3B2 are the load sharing by G2 and G,respectively, i.e. x + y = 1000 kW, then to determine therequired quantities, consider the two trianglesF2BiF3 and FzA1 Fowhen-- F2F3B1F3--FZFLl AIFOa X- 52-50-600or 600a = 2xWhile considering triangles F, B2F3 and flA,Fo,ora-0.5- Y1.5 600--~1000 - x600(a- 0.5) 600 = 1.5 (1000- x)From equations (1) and (2)0.5~600=2~-1.5(100-~)or 3.5~ = 300 + 1.5 x 1000:. x = 51 4 kW (for the generator having 4% droop) and y =486 kW (for the generator having 3% droop)2 x 514and a= ___600= 1.7 HZ... FF3 = 52 - 1.7= 50.3 Hz.While sharing a load of 1000 kW they will be unequallyloaded as noted above and operate at a bus frequency of50.3 Hz.Example 16.3Consider two DG sets operating in parallel and having unequalratings as noted below:G2Rating 1000 kVA 750 kVAp.f. 0.8 lagging 0.8 laggingSpeed at full load 1500 r.p.m. (50 Hz) 1470 r.p.m. (49 Hz)Speed at no load 1560 r.p.m. (52 Hz) 1530 r.p.m. (51 Hz):. Droop 4% 2%. It is governed bythe full-load speed ofthe larger machine(1500 r.p.m).and active load 800 kW 600 kWGI(A) To determine the load sharing between the two, draw thedrooping curves as shown in Figure 16.38.FF, = no-load speed of G1 at 51 HzFF2 = full-load speed of GI at 49 HzF2B1 = 600 kWFIB, = drooping characteristics for G,Similarly FF3 = no-load speed of G2 at 52 HzFf4 = full-load speed of Gz at 50 HzF4Aj = 800 kWF3A1 = drooping characteristics for G2Since the full-load speeds of the two engines are different,they will not be running at their respective full-load speedswhen running in parallel, but somewhere between thetwo. It will be seen that the maximum output will occurwhen G2 is operating at its full load. Consider trianglesFqF4A2 and F, F2B. Then-- FiF, F4A2F,Fz FzB1or ’-2 600.. x = 300 kW:. The maximum load the two machines can share whenrunning in parallel= 800 + 300= 1100 kW as against a capacity of 1400 kW(B) When considering a load of only 1000 kW the sharing ofthe two machines will be as follows. Now the machineswould run a little faster than before. Say, the operatingline is shifted to BiB;, at a frequency of FFo. Considertriangles F3 BiFo and 4A, F4 wheni.e.or- a- ~y 800800a = 2 ys52 Hzil Hz-X19 HztF600kW ~800 600 400 200 0 200 400 600- Load Load ~G2 - 800kW I GI - 600kWFigure 16.38Determining the load sharing

161530 Industrial Power Engineering and Applications HandbookSimilarly, consider triangles FIF& and Fi F26i whenand= 744 kWFOB2 = 256 kWor1000 - y2 600or (a- 1) 600 = 2 x 1000-2yFrom equations (1) and (2)1400a - 600 = 2 x 1000..a=- 26001400= 1.86 HZand frequency of operation = 52 - 1.86= 50.14 HZand Y=800 x 1.862(2)Conclusion1 A slight variation in the drooping characteristics causes avariation in the load sharing.2 A machine that has a higher droop (G2 in the above case)will share a larger load than the one that has a lowerdroop.3 The higher the droop, the higher will be the load variation.4 When there is a difference in the full-load speeds the loadsharing during a parallel operation will not be equal andthe generators will operate underutilized.Hence the significance of employing identical machines, whenrequired to run in parallel to achieve the optimum utilizationof their capacities.To determine the fault level of the system when two ormore machines are operating in parallel refer to Section13.4.1 (5).Relevant StandardsIEC Title IS BS60034-1/1996 Rotating electrical machines 4722/1992, BS EN 60034-1/1995Rating and performance 325/199660034-3/1988 Rotating electrical machines - BS EN 60034-3/1996-Specific requirements for turbine type synchronousmachinesSpecification for voltage regulation and parallel - BS 4999-140/1987operation of a.c. synchronous generatorsRelevant US Standards ANSI/NEMA and IEEEANSI/IEEE446/1995ANSUEEEC37.101/1993ANSIC50.10/1990ANSVIEEE115-1/1995NEMAMG-111993NEMAIMG-2/1989Recommended practice for emergency and standbypower systems for industrial and commercial applications.(IEEE Orange Book)Guide for generator ground protectionSynchronous machinesTest procedures for synchronous machines. acceptance andperformance testingMotor and generators ratings, construction, testing andperformanceSafety standards (enclosures) for construction and guide forselection, installation and use of rotating machinesNotes1 In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology andor its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, rcaders should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

Captive (emergency) power generation 16/531Appendix 1 : Selection of powercablesA16.1 IntroductionTo provide a reference for those working on power projectsor at sites, we provide some important data on differenttypes of LT and HT power cables in this appendix. Thecables described here are in use for all kinds of powerdistribution applications. Of these, XLPE cables are alsoused for power transmission applications. To help a userto select the most appropriate types of cables, we alsoprovide a brief comparative chart of the various types ofcables being manufactured. Tables giving the technicalparticulars of such cables in all voltage ratings have alsobeen provided.The selection process of power cables is almost thesame as that of a bus system discussed in Section 28.3.For simplicity we consider only the basic data for selectionwhich would suffice the majority of applications. Foraccurate calculations a similar approach will be essentialas for the bus systems (Chapter 28). For site conditionsand laying arrangements which may influence the basicrating of a cable, corresponding derating factors havealso been provided. The information covered here willbe useful to users to meet their cable requirements,although the data may vary marginally for differentmanufacturers. For more data on cables, not coveredhere, reference may be made to the cable manufacturers.The choice of any of the cables mentioned in TableA16.1 will depend upon the site conditions, fault leveland the voltage rating of the system. A brief comparisonof all these insulating systems is given in Table A16.2Table A16.1Insulating systems for cablesSI: no. Insulating system Constituents1234567Polyvinyl chloride (PVC)Paper insulated (PI)(Figure A16.1)Unfilled or filled crosslinkedpolyethylene (XLPE)insulated(Figure A16.2)Polyethylene (PE)Ethylene propylene (EP) rubberFlame retardant low smoke cables (FRLS)Fire survival cables (FS)A thermoplastic compoundImpregnated paperA thermoplastic compoundA thermoplastic compound. These are basically polyethylene compounds only, witha little crosslinking to save on cost. LT cables and those below 6.6 kV are costly toproduce and hence not in great useA synthetic rubber (butyl rubber)The outer sheathing of such cables is made with the base insulation of chlorinatedpolymers (e.g. PVC, polythene, CSP, PCP, XLPE or EP rubber) and hence they canbe manufactured for all system voltages. All polymers are self-extinguishing andfire retardant. But in the event of fire, they propagate fine and release large volumesof dense smoke, toxic gases and HC1. When combined with water, such as duringfirefighting, they produce corrosive acids which are highly dangerous for humaninhalation. A good FRLS cable must therefore possess the following properties:- Ability to restrict the propagation of flame- Emit low smoke,- The smoke emitted should not obscure visibility- Emit low acid (HCl) gas- Emit low toxic gasesTo make these polymers have the desired properties, certain additives (chemicals),as noted below, are added to the sheathing compound in specific ratios. The additivesact like flame retardants and diminish the ignitability of the insulation by loweringthe temperature of the cable, delaying ignition and resisting the spread of fire in theinsulation and the polymeric compounds- Alumina trihydrate - to achieve reduction of heat by cooling through an endothermicprocess that decomposes the flame- Molybdenum trihydrate - reduces smoke- Antimony trioxide - also provides a flame retardant effect- Zinc borate-forms a protective coating of a glass-like film, retards the burningprocess and protects the insulation- Calcium carbonate - emits non-flammable gases and helps to reduce the supplyof oxygen to the burning surfaces. The FRLS cables thus produced would possessthe required propertiesThese are silicon rubber, glass tape or glass mica tape sheathed, with an elastomer,having fire retardant and low smoke properties

~ ~~~~16/532 Industrial Power Engineering and Applications HandbookTable A162 Comparative study of different insulating systemsSI:no.123456ParticularsLoss factor, tan 6 at 50 Hz and 20°(Dielectric constant at 50 Hz and 20'A.C. dielectric strength (kVlmm)Volume resistivity at 20°C (R-cm)Specific gravity (g/cm3)Thermal resistivity,"C (cm/W)PVC0.075.8181012-10'5Very heavy,1.35-1.46650Paper'nsulutrdLow, 0.0033.5$010'5Heavy, 1.1550XLPE insulatedPVC JheutedVery low, 0,00052.34010'7Very light, 0.93350PE_~Very low,0.00052.3301017Very light,0.93350?P rubber

~ Thearc\~~ Power~ cable~ 3I''I~Captive (emergency) power generation 16/533~- _ _ ~ -20 Likely applicationscable for allLT applications and upto 11 kV inHTPupperIn commonuse up to11 kV but noinhibition upto 33 kV~-Power distribution andcontrol wiringUtilities, lifts, elevatorsUnderground railwaytransportMiningSubmersible pumpsPaper, miningChemicals and fertilizer?All locations that are notfire hazardousIXLPE insulutedPVC shented2 Fire retardant, has lowsmoke and low halogenemission under fireconditions3 Strong resistance todeformation at hightemperatures1 High thermal stability5 Until a few years ago paperinsulated cables had adominant position but notso with the advent ofXLPE cables (a developmentof the 1960s). inview of their higher thermalrating and availability in allvoltage ranges up to 400 kVand above. This situation isalmost similar to SF,,technology over vacuum(Chapter 19). While vacuumis preferred it has limitationsin HT above 33 kV as havcpaper insulated cables, whichare available up to 33 kVand have limitations beyondthis. Hence the use ofXLPE cables for HV andEHV installationsplants for hulktransfer of energy fortransmission anddistribution of power- Chemicals and fertilizers- Underground metro- Mining- High fire risk zones- All possible applicationsPEIEP richhei________2 Resistant to oil,fire and heatThey are copperconductor cables tosave from breakages4 Being highlyflexible. are idealfor materialhandlingapplications suchas cranes. hoists,lifts and allIn~Vd~latloIl\mounted onmoving platform\.Uqed as a trailingfor reelingand unreelingoperationsM z i a i d c- Reeling-unreelingoperations such as- Hoists- Conveyors- Lifts- Dumpers- Trailing cables- Portable drill\i - Earth-movingmachinery- Offshore platformfeeder cables- Ship wiring- Transportation- Railways- Ovens- Furnaces- Steel rolling mills~.NOIt,.\I Flame retardant low smoke (FRLS) cablesWhere greater safety for equipment and personnel against a likely occurrence of fire is a pre-requisite FRLS coating is used on practically all types ofcable?, described abovc. as in IEEE 383, IEC 60332 I and IEC 60754-1. This coating will restrict the spread of fire and produce a low smoke.Applications- Hazardous location.-High rise huildingr- Ho\pitals-Hotels-Schools-Power stations-Office\, banks-Public places, for safety to life and property-Offshore platforms etc2 Fire survival cables (FS)These are heat- and flameproof cables suitable for high f-ire risk zones. In severe fire conditions, when the outer protection and insulation have beendestroyed, these cables would still maintain continuity of essential services.-Emit vrry low smoke.-For installations prone to ignition and fire, these cables can sustain fire for three hours as in IEC 60331, to minimize the extent offire and consequentdamage.-This. however. I\ pos\ible only in special EP rubber compounds as noted above.Applications- Firefighting sy\terns-Pumping-Safety alarm\-Emergency lighting- Offshore platforms-All places of high fire risks

16/534 Industrial Power Engineering and Applications Handbook1. Conductor2. Conductor screening3. Impregnated-paperinsulation4. Screening (metallizedpaper)5. Tape6. Belt insulation7. Lead sheath8. Bedding9. ArmouringIO. Outer servingBelted construction in whichinsulation in the form ofcommon belt, in addition tocore insulation, is employedFigure A16.1Screen construction (H type) inwhich each core has its owninsulation and screenScreened SL construction (HSLtype) which, in addition topossessing its own insulationand screen also has a separatelead sheath for each core asagainst a common lead sheathin the other two typesDifferent constructions of paper-insulated cables7 Extruded semi-conducting conductor shieldingAluminium conductorTROPOTHEN-X insulationTROPODUR outer sheathIExtruded semi-conducting core shieldingCopper tape screeningFigure A16.2Core identification tapeThree-core XLPE cable (Courtesy: CCI)

~~~~~~~ ~Table A16.3(Contd)Type Na. of cares Conduc- Thickness Thickness Armoiiring Thickness Appros. Appros. Mar. A~~TOX Appros. Apprux. Current ratinyShort- Normal deliveryund cmss- tor fAlJ ofPVC of of PVC ad. net wt d.c. a.c. reactance capacir- circuilcertionul minimum insula- common Flat Round outerofcuhlr re.siatunce re,iAtunce at 50 H7 ance Der Direct iri In duct In air ratingfor Lengrharea nu. uf tion covering wire vvire sheuthat 20%phase groundI secondwires (nom.) minimum ~ize (niin.)extrudedmmmmmm Diu. mmmmmmatoperatingtemp70 "Chg/km Wkm R/kmpF/km Amps AmosAmps k4 m(rm.s.)Drum size Approx.grossweightAYFY 3'/,x25/16 sm/m 63'/2x35/16 smlrm 63'/,x50/2.5 sm 63'/,x70/35 sm 1231/2x95/50 sm 1531/2x120/70 sm 153'i2x150/70 sm 1531/2x185/95 sm 3031/2x240/120 sm 30311- ,x300/150 sm 303'/,x400/185 sm 533'/2x500/240 sin 531.2/1.0 0.31.2/1.0 0.314/1.2 0.31.4/1.2 0.41.6/1.4 0.4I fd.4 0.51.8/1.4 0.52.0/1.6 0.52.2/1.6 0.62 411.8 0.62.6/2.0 0.73.012.2 0.74.0x0.8 1.404.0x0.8 I 404.0x0.8 1.564.0x0.8 1.564.0x0.8 1.564.0xn.x 1.724.0xn.x 1884.0x0.8 2.044.OX0.8 2.204.0~0.8 2.364.0x0.8 2.684.0x0.8 2.8426 027.031.034.039.042.047.052.058.065.072.082.0930 1.20 1.441060 0868 1.041340 0.641 0.7701690 0.443 05322150 0.320 0.3842570 0.253 0.3053000 0.206 0.2493700 0.164 0.1994660 0.125 0.1525630 0.100 0.1236990 0.0778 0.09758710 0.0605 0.07670.0940.0930 0930.0900.0900.0880.0880.0880.0x50.0850.0850.0850.620 760.660 920.700 1100.730 1350.760 1650.780 1850.795 2100.810 23s0.820 2750.825 3050.830 3351.100 3.55637795I 1514015517520023526029031570 190 50086 2.66 500I05 3.80 500ID0 5.32 500155 7.22 so0180 9.12 500205 11.4 500240 14.1 500280 18.2 500315 22.8 500375 30.4 450405 38.0 350A1206 600A1004 610A1206 810A1407 1040A1506 1360A1608 1620A1608 1830B1809 229082010 2860B2212 352082414 401082414 4780AYFY 4x25 sm 6 1.2 0.3 4.0x0.8 1.404x35 sm 6 1.2 0.3 4.0x0.8 1.404x50 sm 6 1.4 0.4 4.0~0.~ 1.56260 970 120 I44 0097 0620 76 61 70 I 90 500 ~1206 62028 5 1170 0868 104 0094 0660 92 77 86 266 500 A1004 670n o 1510 0641 0770 0093 0700 110 95 105 3 80 500 A1407 950T\,pe de., ignationAYCEWFGbYwwFFCunductur type,rermrmivsmAluminium conductor- when it is the first letter of type designation. Whcn type designation doe\ not havc an 'A' prefix then cable is with copper conductorsWhen it is at first or second place in type designation it stands for PVC insulation.Individual core screening.Round stecl wire armouring.Flat steel wire (strip) armouring.Steel tape counter-helix.When last in type designation, it stands for PVC outer sheath.Steel double-round wire armour.Steel double-strip armour.Circular, wlid conductor.Circular, stranded conductor (non-compacted)Circular, stranded compacted conductor.Sector shaped, stranded conductor.Coiirtese.CCI

N m -I - - N Tcm - - N m4 s 4 2 2 4 ci0x 0x 3 x k k g z0 0 3 0 0dd-Id-Iddtk

~~~~~ ~~~ ~~~~~ ~~ ~~~~~Table A16.4Copper cables 650/1100 VConductortro\\-\ec tlOJ7mm'Mux. (1.c'.resistanceat 20°CUkmApprox. u , ~resistuiice (11opercitingtemp. (70°C)WkmCurrent rating /Or uric-wrr"In (IllA inpAmpAmpAny7AmpAnrpShort-circultrtitingfor 1 scadkA(rm..s.)1512.114.522212023202021 17170.173257.4 I8.8730292732272727 24240.28844.615.5239383541353536 30300.46063.083.6949484450444545 38190.690101.832.1965646070586060 50521.15161.151.38858382907s7877 64661.84250.7270.870I101 10I1011597I0599 81902.88150.5240.627I30I25I30I40I20125120 99I 104.03500.3870.463155I5016-iI65I45I55145 I251355.75700.2680.32 11901752 05175 1501658.0595I201 500.1930. I530.1240.23 I0.1840.149220250280200220245245280320210 I75240 I95270 22520023026510.913.817.3I 850.099 10.120305260370300 25530521.32400.07540.09 I2345285425345 29535527.6700400500610800I0000.060 I0.04700.03660.02830.02210.01760.07390.05920.04680.03790.03 I40.027 I- ~ ~"Three single-core cables laid in trefoil formationCofir/eyy: CCI375400425470510590110735355775405435475550590660125870385 335425 360440 39040045550034.546.057.572.592.0115.0

~~~~~~~~Table A16.5 HT Cables up to 11 kV: Armoured three-core power cables 1.9/3.3 kVaNo. of cores Conductor Thickness Thickness Armouring Thickness Approx. Approxand cross- (AI) of PVC of common,flat wire of PVC 0.d. net wt.sectional minimum insulation coverin1 size outer ofarea no. of wires (nom.) minimum sheath cablewrapped(min.)No. x mm2 mm mm mm mm3x25 sm3x35 am3 x50 sm3x70 sm3x95 sm3x120 sm3x150 sm3 x185 sm3x240 sm3x300 sm3 x400 sm3x500 sm666121515153030305353__2.32.32.32.32.32.32.32.32.32.52.73.00.30.30.40.40.40.50.50.50.60.60.70.74.0 x 0.84.0 xO.84.0 x0.84.0~ 0.84.0 xO.84.0 x0.n4.0 x0.X4.0 xO.84.0 xO.84.0 xO.84.0 xO.84.0 xO.8__-1.401.561.561.561.721.721.882.042.202.362.522.84mmkghm28.5 106031.0 126034.0 147037.0 176040.0 214043.0 246045.0 281049.0 330054.0 405060.0 489067.0 603076.0 7510Max. d.c. Approx. Approx. Approx. Current rating Short- Normal deliveryresktance a.c. resi- reactance capacitancecircuitat stance ut at per phase raring20°C operating 50 Hz Direct In In air for Length Drum Approx.temp. in duct I secondsize gross70OC ground weightR/km R/km Rkm pFAm Amps Amps Amps kA m kg(~m..s.)1.200.8680.6410.4430.3200.2530.2060. 1640.1250.1000.07780.06051.441.040.7700.5320.3850.3050.2490.1990.1520.1230.09750.07670.1040.1000.0960.09 I0.0880.0860.0840.0830.0810.0800.0800.0800.4450.5000.5550.6300.7 150.7850.9901.0801.2101.2301.2751.29014891 05I3016018020523026529532534561 7075 8692 105110 130135 155150 180170 205195 240230 280250 315280 375305 4051.90 5002.66 5003.80 5005.32 5007.22 5009.12 50011.4 50014.1 50018.2 50022.8 50030.4 50038.0 450A 1206A 1206A 1407A 1407A 1506A 1608A 1608B 1809B 2010B 2212B 2414 2212G670770930107013601560174020902550315037204240'These cable can be used on 3.3 kV grounded or ungrounded systemsCourtesy: CCITable A16.6 Armoured three-core power cable 3.N6.6 kV (grounded system)No. of cores Conductor Thickness Thicknesr Armouring Thickness Approxand cross- (4) of PVC of commonflat w,ire of PVC o.d.sectional minimum insulation covering size outerurea no. of wires (nom. J minimum sheathwrapped(min.)NO. x mn?mmmmrnmmmApprox.net wt.ofcablekR/kmMax. d.c. Approx. Approx. Approx. Current ratinxresistance U.C. resi- reactance capacitance __ -at stance at at per pha.rr20°C operating 50 HzDirect In In airtemp.in ducr70°CgroundR/km R/km R/km pF/km Amp5 Amps Amps~__Short- Normal delivetycircuit____ratingfor Length Drum Approx.1 second si7e grossweightkA m kg(cm..s.)3x253x353x503x703x953x1203x1503x1853x2403x3003x4003x500smsmsmsmsmsmsmsmsmsmsmsm666121515153030305353__Courtesy: CCI3.63.63.63.63.63.63.63.63.63.63.63.60.40.40.40.50.50.50.60.60.60.70.70.7-4.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0~0.84.0x0.84.0x0.84.0x0.84.OxO.X4.0x0.8I .56I .561.721.721.881.882.042.202.362.522.682.843538414447495256616672791530174020302370276030503450401048005630669080201.20 1.44 0.1190.868 1.04 0.1130.641 0.770 0.1080.443 0.532 0.1030.320 0.385 0.0980.253 0.305 0.0950.206 0.249 0.0930.164 0.199 00900.125 0.152 0.0880.100 0.123 00860.0778 0.0975 o on40.0605 0.0767 0.0820.3400.3750.4150.4700.5200.5650.7200.7750.8600.9401.0301.12574 61 7089 75 86105 92 105130 110 130160 135 155180 150 180205 170 205230 195 240265 230 280295 250 315325 280 375345 305 4051 .YO2.663.805.327.229.1211.414.118.222230.438.0500500500500500500500500500500450450A 1407A 1506A 1506A 1608A 1608B 1809n 18wB 2010B 2212B 22126n 2414B 2414960116013001 52017101960216025303100352038704470

~~ ~~~ ~~~~~~ -~~ ~~ -~~ ~~~~~ _~ ~~ ~~~ ~ ~~ ~~ ~ ~ ~~ ~~~ ~~~~ ~~~Thick-~~ ~~ ~~~~~ ~~~~ ~ circuitTable A16.7Armoured three-core power cables 6.35/11 kV (grounded system)aNo. nf i’uresund crow.\l~cti~Jliulurt’t’uConductiir Thickness Thicknrr Armuurit~,q Thickntw Apprur.(AI) of PVC of’r.ommon flu/ wire of PVC‘ (I (1.niinimurn rtirulotrnri covering .\i:e outermi. of wire.\ (nom.) minimum shvutlrn~ricpprd(inin,)No x mm23x253x353x503x703x953x1203x1503x1853x2403x3003x4003x500rdvrdvrdvrdvrmivmm/vrm/vrndvrdvrm/vrmlvnn/v66612ISIS153030305351111111 111111 (mm! mm ?111314.2 0.542 0.54.2 0.54.2 0.54.2 0.64.2 0.64.2 0.64.2 0.74.2 0.74.2 0.74.2 0.74.2 0.74 oxn.8 1.72 463.0x0.8 1.8X 494.Ox0.8 1.88 514.0x0.8 2.04 554.0~0.8 2.20 604.oxn.x 2.20 644.0x0.8 2 36 674.Ox0.8 2.52 724.0xn.x 2.68 784.OxO.8 2.68 834.0x0.8 3.00 924.0~0.8 3.00 99- ~~_ _ _ _ _ ~Tables of 6.3511 I kV giade (grounded system) arc also considered suitable for use on 6 6/66 kV (ungrounded system)Coirrteq: CCI24702800306035604160471052306090717081209670I1380Table A16.8 6.35111 kV (Grounded system) paper insulated belted cables (18:692)1.20 1.44 0.139O.XhX I 04 0.1320.641 0.770 0.1260.443 0.532 0.1190.320 0.385 0.1130.253 0.305 0.1080.206 0.249 0.1050.164 0.19~ n.in30.125 0.152 0.0990.100 0.123 0.0960.0778 o.0975 0.093n.n6os 0.0767 0.079I1 Paper insulated cables up to 33 kV_ ~0.436 73 61 730.4xn 86 72 870.525 IO0 86 I 050.589 125 Ill5 1300.661 150 125 1600.728 165 140 1850.7~0 191) 160 2100.859 215 1x5 2350.953 245 210 2701.040 27s 235 3101.170 315 275 1651.250 365 320 4351.90 5002.66 5003.80 5005.32 5007.22 son9 12 son1 1.4 50014.1 50018.2 45022.8 35030.4 20038.0 200- ~A160X 1570A160X 1730~1x09 1970~2nic 211082010 261082212 71l60B2212G 1320B2111 185082414 409082414 3700B2414G 2190B2616G 1240~~ .Conductor (aluminium)Thickness of Lead sheath Armouring Apprm. Approx. Normul Drum Approx. Ma.ri- Apprm Apprrix. Approu. Currenr mting Shortin.culution- ~~fiver net delivery Size gro.u mum 1i.c. react- capa- ~~Number CroJJ- Confi- Mini- Appror. Thid- Appro.r. all diu. weight length type weixht d.c. resist- uncc per citance Direct In In air ratingof srctiona1,~uration mum Cund/ Cond/ ees\ dim ness of diu. of cable A of resi.\t- uncc .f phaye uf per in ducr ,fur Icures ureuno. oj uind. yheath over each over delrvep unce ut condirc- SO Hz phase gruund ,econdwfri’\ (nom. I (nom j (nom.) armour- tape armour- Imgrh 20°C tor aring 1nom.j ingincluding operating(num)(nom.)drumtemp. 6S0C1171i12m m mm mm mm kgkm nimkg fNm n/km IZhni pF/ktii Amps Ampx Amps kA__~__ -~____~1 16 RM 7 5.3 3.8 I .5 27 0.5 32 38 3000 500 I506 1790 1.91 2.26 0.120 0.265 58 49 so 1.253 25 SM 6 53 3.8 I .5 26 0.5 31 37 2990 500 1407 1685 1.20 1.42 0.110 0.315 72 64 68 1.953 35 SM h 5.3 3.8 I6 29 0.5 34 39 3410 son I506 1990 0.X6X 1.03 0.107 0.345 x4 74 80 2.733 50 SM 6 5.3 3.X 17 31 0.5 36 42 3910 700 1x10 3240 0.641 0.757 0.102 0.390 I 05 92 100 3.903 70 SM I2 5.3 3.XI .x 34 0.X 40 37 4Y 70 850 2010 4820 0.443 0.523 0.096 0.435 I30 115 125 5.463 95 SM 15 5.3 3.8 I .x 37 0.x 43 50 460 700 20104490 0320 0.378 0.094 0.4x0 155 135 155 7.413 I20 SM 15 5.1 38 2.0 390.8 46 52 6320 600 20 10 4390 0.253 0.3oo 0.091 0 52s I70 155 175 9 36-3 I50 SM 15 5.3 38 2.0 42 0.X 4X 54 6850 son 2012 4030 0.206 0.244 0.08’9 0.565 I90 I75 200 I I .73 I85 SM 30 5.3 38 2.1 46 0.8 52 59 7880 400 2012 3750 0 I64 0.196 0.087 0.605 220 200 230 14.43 225 SM 30 5.3 38 2.2 49 0.8 55 62 8830 350 2012 3690 0.134 0.160 0.086 0.670 240 220 260 17.63 240 SM 30 5.1 3.8 2.3 49 0.8 56 63 9220 350 2012 3830 0.125 0.150 0.085 0.710 250 225 275 18.73 300 TM 30 5.3 38 2.4 56 0.8 62 6’9 1 OR90 500 2414 6411) 0.100 0.121 0.084 0.750 2x0 250 310 23.43 400 SM 53 53 3.X 2.6 00 0.8 66 73I 2570 400 2314 5990 0.077~ 0.0’956 ll.081 0.x70 125 2‘95 370 31 23 500 SM 53 5.3 3.8 2.X 60 OX 72 XO I4960 300 2616 5570 0.0605 0.07s~ o.080 0.935 365 320 435 39.0-~ -Note Thickness of lapped hcdding approx. 1.5 mni. Thickneys of lappcd serving approx. 2 0 mmCourtav. CCI

~~~~~~Table A16.9 11/11 kV (Ungrounded system) paper insulated belted cables (IS 692)Conductor (aluminium) Thickness os Lead sheathinsulation ____-Number Cross- Confi- Mini- ____ Thick- Approx.of sectionalgurution mum Cond/ ConaY ness diu.cores area no. of cond. sheath overwires (nom.) (nom.) (nom.) leadsheathmm2 mm mm mm mmArmouring Approx. Appros. Normal Drumover net delivery SizeThick- Appros all diu. weight length lypeness of diu, of cable Aeach overtape armour-(nom.) ing[nom.)mm mm mm kg/km mmApprox. Muxigrossmumweight ax.of resistdeliveryairce atlength 20°Cincludingdrumkg n/kmApprox. Approx. Approx. Current ratingShorta.c.react- capa- circuitresist- ance citance Direct In In air ratingance of per per in duct for 1conduc- phase at phase ground secondtor at 50 Hzoperatingtemp.6.5 "CWkm Nkm pF/km Amps Amps Amps !iA33333333333333Note16 RM 7 5.3 5.3 1.625 SM 6 5.3 5.3 1.635 SM 6 5.3 5.3 1.750 SM 6 5.3 5.3 1.770 SM 12 5.3 5.3 1.995 SM 15 5.3 5.3 1.9120 SM 15 5.3 5.3 2.0150 SM 15 5.3 5.3 2.1185 SM 30 5.3 5.3 2.2225 SM 30 5.3 5.3 2.3240 SM 30 5.3 5.3 2.4300 SM 30 5.3 5.3 2.4400 SM 53 5.3 5.3 2.7500 SM 53 5.3 5.3 2.930 0.5 3529 0.5 3432 0.5 3734 0.8 4037 0.8 4340 0.8 4642 0.8 4845 0.8 5149 0.8 5552 0.8 5852 0.8 5958 0.8 6563 0.8 6969 0.8 75Thickness of lapped bedding approx. 1.5 mm. Thickness of lapped serving approx. 2.0 mm4140434650535558626566727683351034903940475056006220686075708660964010050115701352016000500500500850700550500350350300550400350250-15061506160820102010201022122012201222162414241426162616204020302300463045204020421032503630373064905590581050801.911.200.8680.6410.4430.3200.2530.2060.1640.1340.1250.1000.07780.06052.261.421.030.7570.5230.3780.3000.2440.1960.1600.1500.1210.09560.07550.120 0.2450.110 0.2750.107 0.3000.102 0.3400.096 0.3850.094 0.4200.091 0.4600.089 0.5000.087 0.5350.086 0.6000.085 0.6350.084 0.6650.081 0.7750.080 0.825-~58728410513015517019022024025028032536549647492I 1513515517520022022525029532050 1.2568 1.9580 2.73100 3.90125 5.46155 7.41175 9.36200 11.7230 14.4260 17.6275 18.7310 23.4370 31.2435 39.0Table A16.10 12.7/22 kV (Grounded system) paper insulated screened (H) cables (IS 692)Conductor (Aluminium) Thickness Lead sheath A rmouring Approx. Approx. Normal DrumofoverNumber Cross- Confi- Minimum insulation Thick- Approx. Thick- Approx. all diu.of sectional guration no. of (nom.) ness dio. over ness of diu.cores area wires (nom.) lead each oversheath tape armour-(nom.) ingmm2mmmm mm mm mm mmApprox. Maxinetdeliver)' size gross mumweight length type weight d.c.of cable A of resistdeliveryance atlength 20°Cincludingdrumkg/km mm kg NkmApprox. Approx. Approx. Current ratingShorta.c.react- capa- circuitresist- ance per citance Direct In In air ratingance of phase at per in duct for Iconduc- S0Hz phase ground secondtor atoperatingtemp.65°CR/km Wkm pF/km Amps Amps Amps kA333333333333Note2535507095I201 50I85225240300400RMNRMNRMNRWVSMSMSMSMSMSMSMSM6 5.8 2.06 5.8 2.16 5.8 2.212 5.8 2.315 5.8 2.215 5.8 2.415 5.8 2.430 5.8 2.530 5.8 2.630 5.x 2.730 5.8 2.853 5.8 3.04345485152545660646470750.80.80.80.80.80.80.x0.80.80.80.80.8Thickness of lapped bedding approx. 1.5 mm. Thickness of lapped serving approx. 2.0 mm4951545858606366707176815558616565687 0747878849065207 190790089009030100301067011910131301361015430177505003503503003005004503503003002502002212201220122216221624142414261626162616261628 18s40403120337035 1035505980576052505020516049404910-~1.200.8680.64 10.4430.3200.2530.2060.1640.1340. I250.1000.07781.421.030.7570.5230.3780.3000.2440.1950.1590.1490.1200.09410.1470.1390.1320.1250.1 I60.1120.1080.1050. I020.1010.0980.0950.2160.2390.2650.2960.3550.3870.4190.4560.4960.5100.5570.6177288105I30155170I90220240250275310667997120140I55I7520522s23025528572881051301601802 I O2452752853203801.952.733.905.467.419.3611.714.417.618.723.431.2

~~~~~~ ~ ~~~.~ ~ __~~~~~~~ ~-~ ~ ~____~~~~~ ~~ ~~~~~~ ~~~~ -~~ _____~~ ~~~~Table A16.11 19/33 kV (Grounded system) paper insulated screened (H) cables (IS 692)Conductor (Aluminium) Thickness Lead .rhruth Arniriurrnx .&lpro.r. .. Approx. Normal Drum Approx. Muxi- Appro\. Approx. Approx. Cirrrerir ratingof -~__ oi’er-Number Cross- Conji- Miniinurn insulation Thick- Approx. Thd- Appr(i.1. nll dio wiKhf length type tvei#ht d.c.. r-esisr- unw p’i- (ituni.c Directof sectional gurotion no. of (nom. j ntw dia. ovet ne.s.s of diu.of i ableA of re.sitt- anceof phusrarper incorec urea wires leud 6wch overdeliver\. unw at conduc.- SO H: phuw ground(nom.) shrutli tupe urmoiii--length 20°C tor ut(nom j ingincliidiii,q operuting(nom.)drumtemp.65°Cnim’ mm mm- ~mm rim mm tnni mni Ig/km mmkg Wkni {Vhm Wkm- ~ ~__ __~~ ~___fiF/kni Ani/>.\___3 70 RMN 12 8.0 2.6 61 0.8 67 753 95 SM 15 7.6 2.5 60 0.8 66 733 120 SM 15 7.6 2.6 62 0.8 68 763 150 SM 15 7.1 2.6 62 0.8 68 763 185 SM 30 7. I 2.7 66 0.8 72 803 225 SM 30 7.1 2.8 69 0.8 76 843 240 SM 30 7.1 2.9 70 0.8 76 843 300 SM 30 7.1 2.9 75 0.8 82 903 400 SM 53 7. I 3.1 81 0.8 87 95~~~ ~~ ~~~ ~ ~~Note Thickness of lapped bedding approx. 1.5 mm. Thickness of lapped serving approx. 2.0 mmConrmy CCI11670 35011420 40012290 35012460 35013780 30015090 25015600 25017260 20019690 1502616 5160 0.443 0.523 0.IDX 0.239 1302414 5530 0.320 0.378 0.125 0.293 1552616 5380 0.253 0.299 0.120 0.319 1702616 5440 0.206 0.244 0.115 0.360 1902616 5210 0.164 0.195 0.111 0.390 2202616 4850 0.134 0.159 0.108 0.423 2402818s 5260 0.125 0.149 0.107 0.435 2452x18s 4810 0.100 0.1 19 0.103 0.474 2702x20s 4330 00778 0.0937 o.ioo 0.523 310I20I40155175200220230255295135I60I80210240275285320380Shortcircuitratingfor Isecond/ul5.467.4 19.3611.714.417.6I x.723.431 2-(111) XLPE Cables up to 33 kVTable A16/12Armoured cable 3.86.6 kV (Grounded system)No. of Conductor Nominal(‘ores and min. no. thickne.ucross- uj wires of.sectiunulXLPEarea ofinsulutionconductorNo. Xmm?mm3x25 d v 6 2.83x35 rm/v 6 2.83x50 rm/v 6 2.83x70 rm/v 12 2.83x95 rmlv 15 2.X3x120 rm/v 15 2.83x150 rm/v 15 2.83x185 rm/v 30 2.83x240 rm/v 30 2.83x300 rm/v 30 3.03x40 rm/v 53 3.33x500 rm/v 53 3.5CourtcwCCI- ~~Minimum Armouringthicknec.s nom.ddimens ions uf PVC diameter weight of length &r resistance reAisfancr atcommon offlat wire outerc ablenormal at at operat- 50 Hrcovering (strip) cheathdeliver). 20°C ing temp.(wr-upped)length)90°Cmm inmxmm mm mm kg/km mkg CYkm Wkm Wkrn- _ _ ______~__ ~. ~~ -0.4 4.0x0.8 156 40 I830 500 A 1506 I 200 I 20 1.54 0.1300.4 4.0x0.x 172 43 21 10 500 A 1608 1380 0.868 1.11 0.1230.5 4.0x0.8 I72 46 2180 500 A 1608 1520 0.64 I 0 822 0.11:0.5 4 .OXO. 8 I88 50 2820 500 B 1809 1850 0.443 0 568 0.1 110.5 4.0x0.8 188 54 1300 500 B 2010 2180 0.320 0.4 IO 0. 1060.6 4 0XO.8 2 04 59 3830 so0 B 2010 2440 n.253 0.325 0. IO20.6 4.0x0.8 220 62 4320 500 n 2212 2860 0.206 0.265 0.0990.6 4.0X0.8 2 20 66 4850 500 B 2212G 1120 I). 164 0.21 I 0.0960.7 4.0x0.8 216 I? 5870 500 B 231 1 3740 0.125 0.161 0.0930.7 4.0x0.8 252 78 6950 400 B 2616 1740 0.100 0.130 0.0920.7 4.0xo.x 2 84 89 8520 3ou B 2414 7420 0.077X 0.102 0.0900 7 4.0x0.8 700 97 I0270 200 B2414G 29100.0605 0.0782 0.089Minimum Approx. Approx. Normal Drum Approx. Mur. Appmx. Approx. Approx Current ratingthickness ocerall net deliverv six Pruss wt. d.c. U.C. rearranre cupucitance- _ _ ~pffimAmps0.235 930.260 I 100.285 I300.325 I600.370 1900410 2150.440 2400.480 2700.540 3150.555 3550.575 40s0.605 455Amp.\-IO0I201451802202552853303854405 10590-Shortcircuitratin#fur IsecondkA (rm..5.)2.353.294.706.588.9311.3I4 I17.422.628.237.647 0

1 .Table A16.13 'Armoured cable 6.35/11 kV (grounded system)No. of Conductor Nominal Minimum Armouring Minimum Approx. Approx. Normal Drum Approx. Max. Approx. Approx. Approx. Current ruting Shortcoresand min. no. thickness thickness nom. thickness overall net delivery size gross wt. d.c. a.c. reucrance capacitance circuitcross- of wires of of dimensions of PVC diameter weight of length (for resistance resistance ut per phase Direct In air ratingsectional XLPE common offlat wire outer cable normul at at operut- 50 Hz in ground for Iarea of insulation covering (strip) sheath delivery 20°C ing temp. secondconductorNo. xmm'3x25 nn/v3x35 d v3x50 rm/v3x70 rmh3x95 rm/v3x120 rm/v3x150 d v3x185 d v3x240 d v3x300 rdv3x400 rdv3x500 rm/vmm6 3.66 3.66 3.612 3.615 3.615 3.615 3.630 3.630 3.630 3.653 3.653 3.6(wrapped)mm mm xmm mm mm kg/km0.40.50.50.50.60.60.60.70.70.70.70.74.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.81.72 441.72 471.88 501.88 542.04 582.20 632.20 652.36 702.52 762.68 812.84 893.00 972140242027103150370042804660533063007260881010310m-500500500500500500500500500350300200A I608A I608A I8098201082010B2212B22 I28231 I82311B2414B2414824140length)ks14001540I790210023802840303034603950340035002920R/kmI .200.8680.6410.4430.3200.2530.2060.1640. I250.1000.07780.060590 "CRkmQ/km1.54 0.1371.11 0.1300.822 0.1240.568 0.1170.410 0.1110.325 0.1070.265 0.1040.21 I 0.1010.161 0.0970.130 0.0950.102 0.0910.0782 0.089PYkm__0.1950.2150.2350.2650.3000.3300.3550.3850.4350.4750,5300.590AmpsAmps93 100110 120130 145I60 I80190 220215 255240 285270 330315 385355 440405 510455 590kA (xm.s.)2.353.294.706.588.9311.314.111.422.628.237.647 .0Table A16.14 Armoured cable 12.7/22 kV [grounded system)No. of Conductor Nominal Minimum Armouring Minimum Approx. Approx. Normal Drum Approx. Max. Approx. Approx. Approx Current raringcores and min. no. thickness thickness nom. thickness overull net deliven, size gross wf. d.c. a.c. reacfance capacitunrecross- of wires of of dimensions of PVC diameter weight of length (for resistance resistance at per phase Direct In airsectionalarea ofconductorNO. xmm'3x35 rm/v 63x50 rdv 63x70 rdv 123x95 rm/v 153x120 rm/v 153x150 rdv 153x185 rm/v 303x240 rm/v 303x300 rdv 303x400 rdv 533x500 d v 53XLPE common ofgar wire outerinsulation covering (strip) sheathmm mm x mm mm mm-6.06.06.06.06.06.06.06.06.06.06.00.60.60.60.70.70.70.70.70.70.70.74.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.84.0x0.82.042.202.362.362.522.682.682.843.003.003.00596266707478828994101IO8cablekg/km3470383043304940558061806820804090801042012170m500500500500500400350300250200200820108221282212082311B23l I82616B24148241482616GB26160B26160normaldeliverylength)kg22602620286032703590343032503270323030403390at20°CR/km0.8680.6410.4430.3200.2530.2060.1640.1250.1000.07780.0605at ouerat- SO Hiin grounding temp90 "CQfkm Nkm pfkm Amps Amps1.110.8220.5680.4100.3250.2650.2110.1610.1300.1020.07820.147 0.1500.140 0.1650.132 0.1800.125 0.2050.120 0.2200.116 0.2350.113 0.2550.108 0.2850.105 0.3100.101 0.3450.097 0.380110130I60190215240265310350400450I20145180220255285330385440510590Shortcircuitratingfor I.secondkA (rm..s.)3.294.706.588.9311.314.117.422.628.237.647.0Courtesy: CCl

~ ~~ .~~~~~ ~~~~ ~~~~Table A16.15Armoured cable 19/33 kV (grounded system)No. (fcores undcros 7-\ertionulureu ofconductorNo. x wiwi'Conducror Nominul Minimum Arinourmginin. no. rhickne\\ fhicknr\,\ nom.ojwirc\ of of dimrnsionsXLPE COWllJlOn of,flor wireinsulation coveriiig(w,rup/Jed)(sfrip)mwi in m tmn x nrniMiiriwrum Al7pro.r.tkicknes5 merullof I'VC diuiirereirnwicflhlrkg/kmIWDl UIlI\i;e3x50 rm/v6 8.80.74.nxo.x76500B2.31 I4.703x70 rmh 12 8.8074.0~0.880350824146.SX3x95 rdv 15 x.x0.74 .OXO. 884350B26168.933x120 rm/v 15 8.80.74.0x0.8893no8241411.33x150 rmiv 15 8.83~1x5 rm/v 30 8.80.70.74.Ox0.84.nxo.x9296200?onB2414ciB2414G14 I17.43x240 rmiv 30 8.80.74 .OXD. 8102200B2616G22.63x300 rm/v 30 8.80.74 0xn.xI07200B2616G2x21x400 rmiv 53 R8__C'oiirte\y: K I0.74.OXO.8I14200A2X18G37.6-~

16/544 Industrial Power Engineering and Applications HandbookA16.2 Technical detailsWe have reproduced in a few Tables (A16.3-A16.15)for cables that are used more com-monly in all voltageratings and with aluminium conductors. For other cablesand copper conductor cables, refer to the manufacturersof their catalogues.A16.3 Service conditionsThe standard parameters of installation on which theratings of the cables are based as in the previous tables,are noted in Table A16.16.Voltage at thispoint is subjectto drops fromreceiving pointup to this point39 supply HT or LT atreceiving end. [Thisvoltage may be lowerthan rated because ofdistribution drops]Transformer (ifthe receiving endsupply is HT)Table A16.16Standard service conditionsParameterGround temperatureO CAmbient air temperature "CDepth of laying in groundUp to 1.1 kVcrn3.3 to 11 kV cm22 and 33 kV cmThermal resistivity of soil,"C c mThermal resistivity of cableinsulation, "C c mPVC30401590-150650PILC XLPE~ _ _ _30 3040 40- -90 90105 105150 150550 350A16.4 Recommended derating factorsThese are common for all types and sizes of cables,except where noted. Derating of cable ratings resultingfrom site conditions and laying parameters are providedin Tables A16.17-A16.26.A16.5 Voltage dropIt is essential to keep the voltage drop in a power cablewithin the permissible limits, particularly for long LTcables say, 25 m and above, similar to a bus system(Section 28.6.2). This may not be necessary in HT cables,where the voltage drop as a percentage of the systemvoltage may be low. The maximum permissible voltagevariation on a system as discussed in Chapters 1 and 12is +6% of the rated voltage. Therefore, during normaloperation the voltage drop in an individual feeder cableshould not be more than I-2% of the rated voltage forcorrect operation of the drive and the load. This is due tothe fact that there may already be many more drops inthe power network from the receiving point up to thefinal load point and all may add up to exceedthe permissible limits. (Refer to Figure A16.3 for moreclarity.) During a motor start it may be kept within 3-5%. The voltage drop in a cable during start can beexpressed byVoltage drop = I,, . Zwhere Z, = starting current of the motor in amperes,andZ = impedance of the cable for that lengthin ohms.Figure A16.3systemVoltage drop at various points in a distributionThis value can be determined from the data available ona per km basis in cable manufacturers' catalogues and asprovided in the previous tables. For the given a.c.resistance and inductive reactance it can be determined by:Z = ,/RZc + X: ohmwhereR,, = 8.c. rcsistance in ohmX, = inductive reactance in ohmIt is possible that at certain installations, even afterselecting the size of the cables on the basis of the siteconditions and the laying parameters as discussed above,a larger cable may become imperative as a consequenceof a higher voltage drop.ExampleConsider a 55 kW motor to be switched direct on-line andinstalled, say, at 75 m from its controlgear. To select the mostappropriate cable size refer to Table 12.4, where/r = 100A, andMaximum ISt = 7 x 1 OOA

Captive (emergency) power generation 16/545:. recommended cable size as in Table 12.4 = 3 x 70 mm2.Considering an armoured cable, as in Table A16.3, a.c.resistancelp haseR, = 0.532 - CY1000 m at 70°Cand inductive reactancelphase, X, = 0.083 CY1000 m at 50 Hz.:. impedance Z = 40.532' + 0.083'= 0.538 W1000 m75and voltage drop during start = 7 x 100 x 0.538 x -1000= 28.25 Vwhich is almost 6.8% for a 415V system and is notrecommended. This cable size is therefore inadequate forsuch a feeder length and the next size, Le. 3 x 95 mm2, maybe chosen.However, during normal running, 3 x 70 mm* cable willhave a voltage drop of just 4.03 V (100 x 0.538 x 75/1000 =4.03 V), which is less than 1% of 415 V. Therefore, dependingupon the duty the motor may have to perform, and otherloads connected on the same bus, the design engineer wouldbe a better judge to decide whether to select a higher size ofcable or be content with this marginal case.For clarity, voltage drop in the next size of cable, i.e. 3 x 95mm', is also calculated asR, = 0.385 W1000 m at 70°CX, = 0.083 CY1000 m at 50 Hz:. Z= d0.385' + 0.083'= 0.394 C2/1000 mTable A16.17Rating factors for variation in ambient air temperatureAir temp. "C 15 20 25 30 35 40 45 50 55Rating factor PVC 1.40 1.32 1.25 1.16 1.09 1 .oo 0.90 0.80 0.68PILC - - 1.30 1.21 1.10 1 .oo 0.88 - -XLPE 1.25 1.2 1.16 1.11 1.05 1 .oo 0.94 0.88 0.82Table A16.18Rating factors for variation in ground temperatureGround temp. "C 15 20 25 30 35 40 45 50 55Rating factor PVC 1.17 1.12 1.06 1 .oo 0.94 0.87 0.79 0.71 0.61PILC - - 1.30 1.21 1.10 1 .oo 0.88 - -XLPE 1.12 1.08 1.04 1 .oo 0.96 0.91 0.87 0.82 -Table A16.19Rating factors for multicore cables laid on open racks in airArrangement 1dNo. of racksNo. of cables per rack1 2 3 6 91 1 .oo 0.98 0.96 0.93 0.922 1 .oo 0.95 0.93 0.90 0.893 1 .oo 0.94 0.92 0.89 0.886 1 .oo 0.93 0.90 0.87 0.86Arrangement 2No. of racksNo. of cables per rack1 .oo2 3 6 90.84 0.80 0.75 0.732361 .oo 0.80 0.76 0.71 0.691.00 0.78 0.74 0.70 0.681 .oo 0.76 0.72 0.68 0.66

~~~~i &&pJ&Arrangement25 mm&,&2d& + i.No. of racks1231 6No. of circuits per rack2 3.~1 .oo 0.98 0.96I .oo 0.95 0.93.1 .oo 0.94 0.921 .oo 0.93 0.90Table A16.21 Rating factors for groups of twin and multicorecables laid directly in ground in horizontal formationTable A16.25 Rating factors for variation in thermal resistivityof soil (multicore cables laid directly in the ground)No. ofRating factor for axial spacingTouching 15cm 30 cm 45 cm 60cm~~2 0.79 0.82 0.87 0.90 0.913 0.69 0.75 0.79 0.83 0.864 0.62 0.69 0.74 0.79 0.826 0.54 0.61 0.69 0.75 0.788 0.50 0.57 0.66 0.72 0.76Table A16.22 Group rating factors for circuits of two singlecorecables, side by side and touching, in horizontal formation,laid directly in groundNo. ofSpacing (between centres of circuits)circuits Touching 15 cm 30 cm 45 cm 60 cm2 0.79 0.86 0.91 0.93 0.953 0.69 0.78 0.84 0.88 0.914 0.64 0.73 0.81 0.86 0.886 0.56 0.67 0.77 0.83 0.878 0.51 0.65 0.75 0.82 0.86Table A16.23 Group rating factors for circuits of three singlecore cables in trefoil and touching, horizontal formation laid directlyin groundNo. ofSpucing (Between centres of circuits)circuits Touching 15 cm 30 ern 45 ern 60 cm2 0.78 0.81 0.85 0.88 0.903 0.68 0.71 0.77 0.81 0.834 0.61 0.65 0.72 0.76 0.796 0.53 0.58 0.66 0.71 0.768 0 48 0.54 0.62 0.67 0.72Table A16.24 Rating factor for groups of twin and multicorecables laid directly in ground in tier formationNo. ofRating factor for axial spacingcircuits Touching 15 cm 30 cm 45 cm 60 cm4 0.60 0.67 0.73 0.76 0.786 0.51 0.57 0.63 0.67 0.698 0.45 0.51 0.57 0.57 0.61.~.~Nominalarea ofRating factor for value of thermal resistivity of soil inC cm/Wconductor(mm’) 100 120 150 200 250 30025355070951201501852403004005006301.14 1.08 1.001.15 1.08 1.001.15 1.08 1.001.15 1.08 1.001.15 1.08 1.001.17 1.09 1.001.17 1.09 1.001.18 1.09 1.001.18 1.09 1.001.18 1.09 1.001.19 1.10 1.001.21 1.10 1.001.22 1.10 1.000.910.9 10.9 10.900.900.900.900.890.890.890.890.890.890.84 0.780.84 0.770.84 0.770.83 0.760.83 0.760.82 0.760.82 0.760.81 0.750.81 0.750.81 0.750.81 0.750.81 0.750.81 0.74Table A16.26 Rating factors for variation in thermal resistivityof soil, three single-core cables laid directly in the ground(three cables in trefoil, touching)Nominalarea ofRating for value of thermal resistivity of soil inC cm/Wconductor(mm’) 100 120 150 200 250 3002535507095I20150185240300400500630 to1000Courtesy: CCI1.19 1.09 1.00 0.88 0.80 0.741.20 1.09 1.00 0.88 0.80 0.741.20 1.09 1.00 0.88 0.80 0.741.21 1.10 1.00 0.88 0.80 0.741.22 1.10 1.00 0.88 0.80 0.741.22 1.10 1.00 0.88 0.79 0.741.22 1.10 1.00 0.88 0.79 0.731.22 1.10 1.00 0.88 0.79 0.731.22 1.10 1.00 0.88 0.79 0.731.22 1.10 1.00 0.88 0.79 0.721.24 1.11 1.00 0.88 0.79 0.721.24 1.11 1.00 0.88 0.79 0.721.24 1.11 1.00 0.88 0.79 0.72

Captive (emergency) power generation 161547and the voltage drop during start= 7 x 100 x 0.394 x 3 volts1000= 20.7 Vwhich is still almost 5% of a 415 V system and is again, onlymarginally suitable for such an application. For still greaterlength of feeder cable line, a yet higher size of cable may beneeded but such lengths are seldom required..mm.A16.6 Skin and proximity effects in amulticore cableThe influence of skin effects in a multi-core cable isalmost the same as that of a multiphase busbar system,discussed in Sections 28.7 and 28.8. However, unlike abusbar system, the resistance and inductive reactancefor various sizes of cables can be easily measured andare providcd by leading manufacturers as standard practicein their technical data sheets. To this extent, making anassessment of skin effects in cables is easy compared toa busbar s stem. Since all the phases in a cable, of a 3-core or 3$?-core are in a rcgularly twisted formationthroughout the length of the cable, they represent thecase of an ideal phase transposition (Section 28.8.4(3))and almost nullify the effect of proximity.A16.7 Short-time rating of cablesTune seconds (t)(a) For PVC cables up to 11 kV. % fi = 0.076Calculating the minimum size of cable for a particularfault level is enough and requires no more elaboratecalculations as for a bus system for the suitability of itsstructure, mounting supports and hardwareetc., as discussedin Section 28.4.2. Thereason is the constructional flexibilityof cables and their direct laying into trenches or on cableracks. In the event of a fault the cables will not damagethe trenches or racks. The enclosure, mounting supportsand hardware etc. are absent unlike in the case of a bussystem. One can therefore determine only the suitabilityof aluminium size for a given fault level (thermal effect),and need not consider any mechanical factor. Since, allthe 3 or 3'/2 cores of the cable are in the form of almosta solid mass, during a fault, dissipation of heat would beslightly less while the direct thermal influence of one coreon the other would be slightly more than in the case ofa bus system. But for all practical purposes equation(28.1) and Figure 28.5 for determining the minimumconductor size will also be valid in the case of cables withvery little variation, i.e.(28.1)The values of 0, and 8 are now based on the short-timetemperature and the continuous operating temperatureassumed for the various types of cables as in Table A16.2.Based on these values, the above equation can be reducedto the following for different types of cables:(b) For XLPE cables 6.6 kV to 33 kV. % fi = 0.094Figure A16.4 Current - time characteristics for aluminiumcables for the selection of minimum cable size for a given faultlevel

~16/548 Industrial Power Engineering and Applications HandbookAx=0.076 for PVC cables*= 0.078 for paper insulated cables,* and= 0.094 for XLPE cables etc*.For the selection of cables, the same guidelines wouldapply as for a bus system in a switchgear assembly (Section13.4). The outgoing circuits that would have a diminishingvalue of fault current, due to circuit impedance, are alsonormally protected through a current limiting device. Anormal size of cable, therefore, commensurate with thethermal rating of the circuit, will be adequate in mostcases, subject to applicable deratings and voltage drops,discussed above. However, for incoming cables connectingthe incoming source of supply and the power-receivingend, an exercise to determine their minimum size for thesystem fault level would be essential, based on graphs ofFiguresA16.4(a) and (b) or similar graphs for other typesof cables. Refer to Figure A16.5 for more clarity. Therewill be no need to check on the mechanical strength ofcables and their supports.A16.8 Termination of cablesThis plays an important role and requires utmost care.While termination of LT PVC cables is easier with thehelp of a crimping tool, HT cables need a proper kit forjointing and end termination. The jointing material isalso manufactured by the cable manufacturers. We arenot providing details of these kits and their jointingprocedures. These can be obtained from manufacturers'catalogues.e-/n .~ICable to besuitable for systemCable to besuitable for systemfault level onthe LT side*These factors are provided by CCI. They may vary slightly fromone manufacturer to another, depending upon the grade of aluminiumbeing used and the value of q, considered by them. Based on theabove tactors, manufacturers provide the short-time rating of theircables for different types and sizes, in the shape of a graph, I,,versus t. Refer to graphs drawn in FigureA16.4(a) and (b) for PVCand XLPE cables respectively, based on the above factors. Similargraphs can be drawn for other types of cables also.Figure A16.5systemSelection ofRelevant StandardsIECTitleIS BS IS060055-l/1997Paper insulated lead sheathed cables for voltages up to 33 kVTests on cables and their accessories692/1994BS 6480/1997 -6005 5 -2/ 1 9 89General and construction requirements60 183/1990Guide to the selection of HV cables60227- 1/1998General requirementsPVC insulated electric cables for voltages 3.3 kV to 11 kV60227-2/1997Test methods60227-3/199760227-4/199760227-5/1998Non-sheathed cables for fixed wiringSheathed cables for fixed wiringFlexible cables (chords)s 1554-1/1988. BS 6004/19956500 and 674660227-6/1985Lift cables and cables for flexible connections

~~~ ~ ~~~~~~ .~ ~~.~ -~~Captive (emergency) power generation 16/54960228/ 199360228A/198260331/197060332- 1 / I 99360332-2/198960502- I / I 99860502-24 I998607.53- lil99460751-2/199160840/1988Conductors for insulated cablesGuide to manufacturers of cables and cable connectionsSpecification for 60011000V and 1900/3300 Varmoured electric cables having thermosettinginsulationFire resisting characteristics of electric cables:Performance requirements, sampleand test conditionsTests on electric cables under tire conditions.Test on a single vertical insulated wire or cableTest on a single small vertical insulated coppeiwire or cableCross linked poly-ethylene insulated PVC sheathed cablesPower cables with extruded cross-linked insulation(XLPE cables) for voltages from 1 kV-3 kV(V,,,= 3.6 kV)Power cables with extruded cross-linked insulation(XLPE cables) for voltages from 3.8/6.6 kVto 19/33 kVTest on gases evolved during combustion of materials fromcablesDetermination of the amount of halogen acid gasDetermination of the degree of acidity of gasesevaluated during the combustion of materials takenfrom cables by measuring pH and conductivityTests for power cables with extruded insulation forrated voltages above 30 kV up to 150 kV (XLPE cables)Flexible cables for use at mines and quarriesElastomer insulated cablesFor voltages up to 1100 VFor voltages from 3.3 kV to I I kV~~~.- . -8 130/1991 BS 6360 -- BS 546711997 ~7098- l/l988--~BS 4066-1/1995BS 4066-21 199s7098-2/1988 BS 7835/1996 -BS 6622119997098-3/ I99869 1/199 1.102611991Relevant US Standards ANSYNEMA and IEEE9968- 1/19889968-2/199 1BS 6425-]/I990 ~BS 6425-211993 -- -BS 670811998 -NEMAwc711992(ICEA S-19)NEMAwc7/1991(ICEA 5-66-524)NEMAWC8/1998ASTM D-2863IEEE 383/1992SS 324 24 17Rubber insulated wire and cable for the transmission anddistribution of electrical energyCross linked polyethylene insulated wire and cable for thetransmission and distribution of electrical energyEthylene propylene rubber insulated wire and cable for thetranqmission and dihbution of powerFor oxygen index testType test of class 1E electric cables for nuclear power generating stationsSwedish specifications for XLPE Cables 12-420 kVNote\I In the tables of relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references. readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication

16/550 Industrial Power Engineering and Applications HandbookAppendix 2: ANSI standard device function numbersDevice numberFunctionDevice numberFunctionMaster elementMaster sequence deviceTime-delay relayChecking or interlocking relayMaster contactorStopping deviceBlush operating or slip-ring short-circuitingdevicePolarity or polarizing voltage deviceUndercurrent under power relay (low forwardpower relay)Starting circuit breakerBearing protective deviceAnode circuit breakerMechanical condition monitorControl power disconnecting deviceField relay or loss of excitation relayKeversing deviceField circuit breakerUnit sequence switchCircuit breaker, contactor, or starterReserved for future applicationOverspeed deviceSynchronous speed deviceUnderspeed deviceSpeed frequency matching deviceReserved for future applicationShunt or discharge switchAccelerating or decelerating deviceStarting to running transition contactorElectrically operated valveDistance relayEqualizer circuit breakerTemperature control deviceReserved for future applicationSynchronizing check relayManual transfer selector device or local/remoteselector switchUnit sequence starting relayAtmospheric condition monitorReverse phase or negative sequence relayPhase sequence voltage relayIncomplete sequence relayThermal overload relayInstantaneous overcurrent relayDefinite time delay overcurrent relayInstantaneous ground fault relayInverse time overcurrent relayInverse time ground fault relayInverse time generator ground fault relayVoltage controlled overcurrent relayGuard relayIDMT with instantaneous overcurrent relayUndervoltage relayA.C. circuit breakerFlame detectorExciter or D.C. generator relayIsolating contactorReserved for future applicationAnnunciator relayPower factor relaySeparate excitation deviceField application relayDirectional reverse power relayShort-circuiting or grounding devicePosition relayRectification failure relay

~~ ~~ ~~ ~~~Devic e number Function Device number FunctionCaptive (emergency) power generation 1615511596061626364656667h6X69707172717375767778Overvoltage relayVoltage or current balance relayReserved for future applicationTime-delay stopping or operating relayBucholz relayRestricted ground fault relayGovernorNotching or jogging deviceA.C. directional overcurrent relayDirectional ground fault relayBlocking relayPermissive control deviceRheostatLevel switchD.C. circuit breakerLoad-resistor contactorAlarm relayPosition changing mechanismD.C. overcurrent relayPulse transinitterPhase-angle measuring or out of step relayUnderfrequency relayD.C. reclosing relayAutomatic selective control or transfer relayOperating mechani5mCarrier pilot-wire receiver relayLock-out relayDifferential protective relayAuxiliary motor or motor generatorDisconnecting switchRegulating deviceVoltage directional relayVoltage and power direction relayField-changing contactorAnti-pumping relaySupervision relayUsed for specific applications where other\ymbols are not suitableFuse failure relayFurther readingMotor protection relayAuxiliary relay tor Y79x0Auto reclosing relayFlow switch~~Bawl on ANSMEEE C 37 2Birnbhra, P.S., Generalised Theo? qfElectric~trl Machine% KhannaPublications, India.Siemens Elecrrictrl Engineering Handbook, Siemens AG. Munich.and Heyden and Sons Ltd, London, UK (for general information).

PART 111Voltage Surges,Overvoltages andGrounding Practices

171555Voltage Surges-Causes, Effectsand Remedies~. .- _I...---,- .-_.I17.2.2 Sudden change of load 171

Voltage surges-causes, effects and remedies 17155717.1 IntroductionVoltage surges are generally a phenomenon of high voltage(HV) power systems and can be considered as the mostsevere pollutant to the insulation of the power system andthe terminal equipment. In this chapter we analyse thelikely amplitude and steepness of surges that may ariseunder different system conditions and the most appropriateinsulation coor-dination between the equipment connectedon the same system. Insulation coordination provides acritcrion in selecting the right equipment with a moreeconomical insulation level for different applications andlocations. Generally, locations away from the source ofvoltage surges, i.e. equipment installed in the downstreamof a powcr system is subject to diminishing surge effects.For example, a rotating machine, which may be a motoror a generator, would rarely be subject to a direct lightningstrike as it would seldom be connected on a bus exposedto direct strikes. It is usually connected through a bus ora cable which is fed through a transformer. All theseinterconnecting devices would withstand most of the effectsof a lightning strike and it would be only somewhatattenuated and dampened surge to which the terminalequipment would be subject.This concept of diminishing value of voltage surges isa logical parameter to economize on the cost of insulationas far as pcrrnissible, without jeopardising the adequacyof protection to the system or the associated equipment.Different equipment installed at different locations onthe same power system may thus have varying degree ofbasic insulation level (BIL), as discussed in Section 18.3.One may notice the variation in BIL from Tables 11.6,13.2, or 14.1, and 32.1(A), for motors, switchgears andbus systems respectively when installed in the same powersystem. Similar variations would apply for other equipmentconnected on the same system. The aim here is tocover the subject as widely as necessary for a properunderstanding without going into extensive details.The type of a surge is identified by its shape, and itsseverity by its amplitude (V,) and time (t,) to reach thisamplitude. All overvoltages discussed are termed surges,since their severity would last only for a few microseconds(ps). In our discussions here and elsewhere in this book,we classify the overvoltages into two categories for easyidentification. One is the temporary or dynamicovervoltages discussed in Chapters 20 and 24 existing inthe system for a slightly longer duration say, one half ofa cycle to two to three cycles at the power frequency (50or 60 Hz), and the other as voltage surges, appearing forjust for a few ps, at transient frequencies of a few kHz.17.2 Temporary overvoltagesThese are of a relatively longer duration, and may haveseveral successive peaks, lasting from one-half of a cycleto a few cycles at the power frequency, depending uponthe time constant (= R/X,) of the circuit that gives rise tosuch overvoltages. The likely causes of such overvoltagesare discussed in Chapters 20 and 24. It has, however,been felt necessary to give a brief review of the same forgreater clarity of the present topic:Ground fault (Section 20.1)Sudden change of load (Section 24.6.2(ii))Resonance and ferro-resonance effects (Sections20.2.1(2) and 24.4.1 and 24.4.2)).17.2.1 Ground faultHigh overvoltages occur in the healthy phases on a groundfault:When the system is grounded through an arc suppressioncoil and is undercompensated whereas the arcinggrounds give rise to voltage surges.When the system has an isolated neutral.When the system is solidly grounded.For more details on grounding systems and the extent ofovervoltages refer to Chapter 20.17.2.2 Sudden change of loadThis is more pronounced on high-voltage and extra-highvoltagesystems (66 kV and above) when:Carrying large powers and where there may be widevariations in the load demandLoad rejection: The load side interrupter, feeding alarge load at the far end, tripsThe load demand falls sharply as a result of substantialload rejection, when the generator feeding the systemis suddenly underloaded and tends to overspeed. raisingits terminal voltage. While the field control systemand the turbine governor will act immediately to regulatethe system, the time to normalize the situation may bea few seconds, hence the necessity to protect the systemagainst such overvoltages.17.2.3 Resonance and ferro-resonance effectsSuch a phenomenon may occur when a circuit comprisinga capacitance C and inductance L is switched ON orOFF and when such circuit parameters undergo a changeduring normal operation, as a result of a sudden changein load. A power circuit will invariably possess suchparameters. For example, leakage capacitances betweenphase to phase or phase to ground are present in a cableor a conductor and these would rise when series capacitorbanks are connected on the same system, say, to improvethe system regulation. Similarly, leakage inductance isalso present in a cable or a conductor and that will alsorise when a transformer or a shunt reactor, having nonlinearmagnetizing characteristics (Figures 27.2b, and b,)is also connected on the same system. According to thefield data on this phenomenon it has been observed thatit is more pronounced on HV and EHV systems (66 kVand above) particularly under the following conditions:IWhen switching a lightly loaded circuit, having atransformer, and the natural frequency of the linearpart of the system corresponds to one of the harmonicsof the magnetizing current.

171.558 Industrial Power Engineering and Applications Handbook2 When the systems that are series compensated areconnected to a lightly loaded transformer or shuntreactor, under certain line conditions (Section 24.4).3 When harmonic filter circuits are connected to apower system with saturated reactors (Figure 27.2(c))resonance may occur between the reactor and thefilter capacitors to give rise to overvoltages.4 Resonance may also occur between the lineinductance, series reactors and shunt capacitors.5 Resonance may also occur between the linecapacitance or the ground capacitance and theinductance of a series-connected limiting reactor,or the inductance of a transformer, connected on thesystem.17.3 Voltage surge or a transientThe occiirrence of a surge or a transient is not intentional,unlike an impulse, as noted later and may appear on anHV and EHV system as a result of system disturbances,such as during:A lightning strike (Figure 17.1)A switching operation*Contact bouncing orA fast bus transfer"Because of a surge transference from higher voltageto the lower voltage side of a power transformer, andDuring faults such as during a ground fault in a resonantgrounded system or an isolated neutral grounded system.These surges are of very short duration and may be definedby the following two parameters:1 Prospective amplitude, V, (Figure 17.4), to define theinsulation endurance of the current-carrying system.Only the first highest peak is of significance for thispurpose, which will contain the maximum severity.The subsequent peaks are of moderate magnitudesand of little consequence for the system or the terminalequipment.2 Time of rise, t,, as illustrated in Figure 17.3. Dependingupon the time of rise, a surge may be classified intothree groups:Switching surgesLightning surges andVery steep or front of wave (FOW) surges.17.3.1 Switching surgesThese are slow rising surges and have a front time ofmore than 10 ps. They are considered as long-durationsurges due to their high total effective duration t2 (Figure17.2(b)). But they discharge very high energy, even duringtheir short duration, and may deteriorate or damage theinsulating properties of the system or the equipment that*A direct on-line switching is a single transient condition, whereasa quick bus transfer is OFF and ON, is. adouble transient condition,hence more severe.is subject to such a surge and has to absorb its severity.For the purpose of energy discharge by a surge, theamplitude and duration of these surges alone is considered(Section 18.6.3). The duration of such surges are themaximum of the other two types of surges.17.3.2 Lightning surgesThese are fast rising and may have a front time of almost1 ps or even less (Figure 17.2(a)). They are thereforeconsidered short-duration surges.17.3.3 Very steep or front of wave (FOW) surgesThese are very fast rising and of very short duration andmay have a front time as short as 0.1 ,us or less (IEC6007 I - I). As a result, while their energy discharge maybe too small to be of any significance, their rate of riseis very rapid. This makes them capable of damaging asmall part of the current-carrying conductor of the terminalequipment, rendering it highly vulnerable to an insulationfailure. Sometimes a restrike of the interrupting contacts,or a quick re-closing of a power system, may also causesuch surges. The situation may become worse:When interrupting small reactive currents, such asduring the opening of an unloaded power line, anunloaded transformer or a motor running at no load.In all such cases it may cause current chopping, leadingto extremely steep switching surges (Section 19.6) orWhen the system already had a trapped charge beforea reclosure.All electrical equipment are designed for a specificBIL,asindicatedinTables 11.6, 13.2, 14.1,and32.1(A)for motors, switchgears and bus systems respectively,and Tables 13.2 and 13.3 for the main power system(line clearances and insulators). If the actual severityof a prospective surge, Le. its amplitude and/or risetime or both, is expected to be higher than these levels(higher amplitude and lower rise time) the same mustbe damped to a safe level, with the use of surge arresters,surge capacitors or both as discussed later.17.4 Transient stability of overheadlines17.4.1 Auto-reclosing schemeAuto-reclosing of a power system, normally overheadlines, after a transitory fault is a type of protective closingand network automation to avoid a supply interruptionas such faults and to improve the system's transientstability limit. A transient stability limit refers to themaximum power that the system (all the generators feedingthe network) can deliver without loss of synchronism.Such transitory faults may be due to system disturbancesas discussed earlier, such as sudden switching of loadsON or OFF causing severe power fluctuations, lightningstrikes, or passing objects like birds, gales and stormshitting the overhead lines. In most cases, such disturbances

Voltage surges-causes, effects and remedies 171559are of a momentary nature. Field studies have revealedthat such causes contribute nearly 90% of the total tripping.During such faults, the interrupters at both ends of atransmission line may trip as a result of travelling surgesin both directions. An auto-reclosing scheme, which maybe applied to an overhead transmission or a long HTdistribution system, can reclose the interrupter on such atrip in about 10-15 cycles and maintain the supply systemintact, preventing loss of synchronism. Thus it helps toachieve a high degree of system stability.Since the fault is of a transitory nature, the schememay be applied even on a per-pole basis, allowing thehealthy phases to remain intact and the reclosingnecessitated only in the affected phase rather than all thephases, thus further enhancing the system’s transientstability. However, it would require an independentinterrupting mechanism and individual relaying andtripping schemes for each pole.It is possible that the breaker may trip on the firstreclosing after a fault, as the fault may not have cleared.In such cases a delayed reclosing, with a delay of onesecond or so, may also be incorporated into the sameswitching scheme to supplement the fast reclosing, tosave the system from a saving and a consequent trip. Bythen the fault would clear in all probability to allow thebreaker to reclose thus maintaining the continuity of supplyonce again and saving the system from falling out ofsynchronism and a tandem trip of all the feeding lines ofa power grid. If the fault persists and the breaker tripsagain, the breaker will lock out and will not close againuntil the fault is removed and the breaker is reset.Delayed reclosing may be adopted where the systemhas large interconnections (mesh system) as at a powergrid, and where loss of one phase may not cause a lossof synchronization in such a duration and hence restorethe transient stability of the whole system.17.5 Causes of voltage surgesIn actual operation, disturbances on a power system,causing sudden changes in the system parameters, arequite frequent and may generate temporary overvoltagesand voltage surges, as summarized above. The systemdisturbances may be of two types, external or internal,as explained below.negative, which creates a voltage gradient between theclouds or between the clouds and the each. The voltagegradient may exceed the breakdown value of the air andcause a flashover. This flashover is similar to anelectrostatic discharge of the atmosphere to the ground,as illustrated in Figure 17.1, and is termed a lightningstrike. There may be up to 20 pulses in each lightningstrike, each having a duration of about 50 ps. Thephenomenon is comparable to the discharge of a highlycharged condenser, the clouds forming one plate, groundthe other and the air as the dielectric.Electrostatically induced chargesDue to the presence of thunder clouds in the vicinity:A charged cloud above and near a large object, anoverhead line for instance, induces a charge ofopposite polarity than its own in the line. When thiscloud bursts suddenly and discharges, its inducedelectrostatic charge travels in both directions of theline, with a near velocity of light and equalizes thepotential at all points. The potential at any pointalong the line rises suddenly from its normal valueto the amplitude (V,) of the travelling wave.Due to the friction of dust or free snow blowingpast the conductors.Electromagnetically induced currentsDue to lightning in the vicinity of the overhead lines. Itis an indirect effect of a lightning strike. The lightningsurges may impose very severe stresses on the line andcause damage to the line insulators and the terminalequipment without causing a trip by the protective device.These surges can be contained at the receiving end withthe use of a surge arrester or a diverter (discussed later).17.5.2 Internal causesMaking or breaking a power circuit causes a change in17.5.1 External causesThese are mainly due to atmospheric disturbances asnoted below. The effect of such surges is totally differentfrom that of switching surges, and the amplitude isindependent of the system voltage.Lightning strikeThere is no clear explanation of lightning. However, themost popular theory is the charging of clouds at highvoltages, up to 20 million volts and a charging current5-100 kA or so, due to the movement of hot air upwardsand big droplets of water downwards. This process tendsto make the tops of the clouds positive and the bottomFigure 17.1 Discharge from clouds

17/560 Industrial Power Engineering and Applications Handbookthe circuit parameters and produces voltage surges. Thosearising out of switching operations are attributed to internalcauses. In this chapter we limit our discussions to thephenomenon of voltage surges, as related to internal causesand particularly as a result of switching. The requirementand the type of protection remain the same for externalor internal causes of system disturbances.17.6 DefinitionsIn the following text we will use a few new terms. Forthe sake of more clarity on the subject these are definedas follows.17.6.1 An impulseAn impulse is an intentionally applied voltage or currentin a laboratory. It is in the form of an aperiodic and unidirectionalwaveform (Figure 17.2). It rises rapidly withoutappreciable oscillations to a maximum value and thenfalls, usually less rapidly, to zero, with small, if any,loops of opposite polarity (Figure 17.4). The parametersFigure 17.3Time (ps)-f, = Virtual front time= 1.67tt' = 0.3t~ = 0.5tDefining a voltage or a current impulse waveformwhich define a voltage or a current impulse are its polarity,peak value, rate of rise (front time) and time to half itsvalue on the tail, as noted later.A transient of an external or internal nature, as discussedabove, is then related to one such type of impulse, forlaboratory testing a particular equipment or system, toestablish its suitability.Front time (ps) --tA 12/50 ps waveform(a) Lightning impulse waveform- 250 Front time (ps) 2500A 250/2500 ps waveform(b) Switching impulse waveformFigure 17.2 Standard impulse waveformsIdentifying an impulse waveformThe distinction between a lightning impulse and aswitching impulse may be made on the basis of theduration of the wave front (shape), rather than of itsorigin. A voltage impulse with a wave front duration ofless than 1 ys, up to some tens of ps, is generallyconsidered as a lightning impulse (Figure 17.2(a)) whereasan impulse with a front duration of some tens of thousandsof ps is a switching impulse (Figure 17.2(b)) accordingto IEC 60060- 1. The type of wave is generally designatedas tl/t2 (Figure 17.3), where1 tl = virtual front time of an impulse:For a voltage impulse having a front duration ofless than 30 ys (normally lightning surges)tl = 1.67 x time taken by the impulse to rise from30% of its peak value to 90%For a voltage impulse having a front duration ofmore than 30 ys (normally switching surges)tl = 1.05 x time taken by the impulse to rise from0% to 95% of its peak valueFor the sake of laboratory testing and analyticalstudies, these surges have been represented bystandard impulse waveforms. According to IEC60060-1 a 1.2/50 ys impulse is called a standardlightning impulse, and a 250/2500 ys impulse astandard switching impulse.For equivalent current impulses, such as 8/20 ys

Voltage surges-causes, effects and remedies 171561for a 12/50 ps voltage surge and 3W60 ps for a250/2500 ,us voltage surget, = 1.25 x time taken by the current to increasefrom 10% to 90% of its peak value.2 t2 = time interval between the origin and the instantat which the impulse has decreased to half of itspeak value.17.6.2 Transient recovery voltage (TRV) and itsrate of rise (r.r.r.v.) in an induction motorThis is an important parameter. A very fast-rising transientcan cause the surge voltage to be non-uniformly distributedover the entire length of the motor windings and affectsthe first or the entrance coil of the motor windings, asdiscussed in more detail in Section 17.8. This is thevoltage t V,) that will reappear immediately on a currentinterruption (current zero) (Figure 17.4) and cause a currentacross the parting contacts yet again, known as post-arccurrent. It oscillates at a very high surge frequencyf, andis composed of a number of surge frequencies, as theleakage inductance L and the lumped capacitance C ofthe interrupting circuit undcrgo rapid changes with thepropagation of the surge wave. The surge frequency is afunction of circuit constants L and C (equation (17. I)).Figure 17.4 drawn for one particular frequency, is onlyan illustration.The severity of the recovery voltage of a surge is definedby its r.r.r.v., which is a function of its amplitude, V,, andthe front time tl (Figure 17.4), tl in turn being a functionof the surge frequencyf;. The higher the surge frequency,the shorter will be the front time tl. The shorter the timef,, the higher will be the rate of rise and the steeper willbe the recovery voltage and the more severe will be itseffects on the terminal equipmcnt.Referring to Figure 17.4, if V, is the peak value of thevoltage surge of a particular transient voltage waveformin kV and tl, the virtual front time or the time of rise ofthe transient voltage from its zero to peak value in ps,then the rate of rise of recovery or restriking voltage,Transient ProspectiveProspective recovery recoveryor peak voltage (TRV) at voltage at Powervoltage surge frequency 6 frequency- ~ront -I t, Time (ptimeFigure 17.4A transient recovery voltage ('TRV')vtr.r.r.v. = - kV/pstlThe significance of this term can be realized by the factthat the voltage stress of a surge, having a maximumamplitude of 4.5 P.u., with a front time t, of 5 ,us, willroughly be the same or even less severe compared to asurge with an amplitude of only 2 p.u. and a front timeof 0.2 ps (see Insulation Sub-committee, RotatingMachinery Committee, 1981).When a fault occurs near the source of supply, the linelumped leakage capacitance C is small and the frequencyof oscillations high. of the order of a few kHz (equation(17. I)). But when the fault occurs at a distance from thesource, C tends to become higher and the frequency ofoscillations lower, of the order of a few hundred Hz. Anintroduction of some resistance in the interrupting circuitwill tend to dampen (attenuate) the oscillations but whereR > 2 (@of cos 4 > 45"). the system will remainoscillatory. R may be introduced in the circuit throughinterconnecting cables, interrupting devices and overheadlines etc., and thus reduce the severity of transients duringan arc interruption. If some R is introduced such thatR > 2- (4 of cos 4 < 45") then the high-frequencyoscillations can be totally dampened.This practice is usually adopted in oil circuit breakers(BOCBs and MOCBs). In air blast circuit breakers(ABCBs), and SF6 circuit breakers, a resistance isconnected in shunt across the contact gap, such that R isintroduced in the circuit during the making and interruptingprocesses only.17.6.3 Surge frequencyThis is the frequency at which the surges travel. Thisfrequency can be very high. of the order of 5-100 kHzor more, depending upon the circuit parameters. Thenatural frequency of oscillations of the transient recoveryvoltage of the circuit in terms of circuit parameters canbe expressed as:(17.1)wheref; = surge frequency in HzL = leakage inductance of the circuit in henry (H ) andC = lumped leakage capacitance of the circuit in farad(F) (L and C being the circuit parameters).Refer to a typical oscillogram of a switching surgeshown in Figure 17.5. Such a surge may exist on thesystem only until the interrupter is conducting, i.e. up toits contact making or contact opening, whatever theprocess of switching. It may be for only one half of acycle to two cycles respectively (10-40 ms for a SO Hzsystem) in terms of normal frequencyJof the system butmany times more than this, in terms of surge highoscillating frequency f?. The product of L and C willvary with a change in thc circuit parameters. For instance,when a transient wave travels through a power system,having a number of equipment and devices connected to

17/562 Industrial Power Engineering and Applications Handbookit such as a switching device to a cable, a cable to anoverhead line or a transformer and an induction motoretc, then the frequency of oscillations will continue toalter after every junction. At each junction the travellingwave will encounter a wave reflection and add moreimpedance in its circuit, as it will propagate ahead. Afterevery reflection, the circuit parameters will change aswill the frequency of uscillations. Thus, a number ofoscillatory frequencies may exist in the system at a time,leading to a more complex phenomenon and making itdifficult to accurately determine the effective surgefrequency of the system and the r.r.r.v. By the use ofoscillograms such as that shown in Figure 17.5, it ispossible to determine such complex quantities.The higher the frequency of the transient recoveryvoltage (TRV), the steeper will be the slope of the TRV(Figure 17.6), Le. the higher will be the r.r.r.v. The r.r.r.v.is a measure of severity of the TRV that the terminalequipment and the devices may have to endure.17.6.4 Surge impedanceThe shape and characteristics of a surge wave areinfluenced by the circuit parameters, Le. the leakageinductance L of the interconnecting cables and the currentcarryingcomponents of the equipment through which it1 2 3For same peak voltage 'V,'tl < ti < f; andfSl > f*z ' fs3VI vt vtand r.r.r.v = - - > -tl t; t;Time Ips)Figure 17.6 The variation of r.r.r.v. with the front time (tl 0~ 2)W Nr1510-200 0.5 1.0 1.5 2.0 2.5 3.0Time(a) Without surge arrester maximum peak to peakvoltage = 26 kV1510150 0.5 1.0 1.5 2.0 2.5 3.0Time(b) With surge arrester maximum peak to peakvoltage = 13.5 kVFigure 17.5 Oscillograms showing the effectiveness of a surgearrester on a 6.6 kV motortravels and the leakage capacitance C of such cables andthe motor dielectric lumped capacitance etc. (Section9.6.1). The relation between L and C that will determinethe shape of the travelling wave is known as the surge ornatural impedance 2, of the system and is expressed as:z, = Et2 (17.2)whereL = circuit leakage inductance in henry (H), andC = circuit lumped leakage capacitance in farad (F).Influence of surge impedance on terminalequipmentThese values (L and C) help to determine the likely surgevoltages that may develop during a switching operationwhile using different interrupting devices. For accurateanalysis, it is advisable to obtain these values from theirmanufacturers. Slamecka (1 980) and others have alsoestablished such curves and have provided them in theform of nomograms for easy identification of thisparameter, as reproduced in Figures 17.7(a) and (b). Thesurge impedance Z, varies with variations in the designparameters, hence from manufacturer to manufacturer.These curves have therefore been provided in the formof bands. To account for likely variations in the designparameters by the different manufacturers one can, atbest, obtain an average value of the surge impedancefrom these curves for a particular machine, as illustratedfor a 500 h.p. (= 500 kVA) motor. But for an accuratevalue, it must be obtained from the machine manufactureralone, as shown in Figure 17.7(c), for curves producedby Siemens for their motors.

Voltage surges-causes, effects and remedies 171563100005 3500 + $2= 3000 R1000N w--.k?a,C-2.- Fa,?v)100100.01 0.1 1 .o 10.0500 kVA (6.6 kV)MVA _tmFigure 17.7(a) Surge impedance of rotating machines as a function of input M V A / , , (from ~ Pretorious and Eriksson, 1982)33 kV 66 kV 11 kVMotors Motors Motors500 kVAFigure 17.7(b) Typical value rating of equipment of surgeimpedances for HT motors and transformers‘i-ELV t-0-io 20 50 102 zoo 103 2000 5000 io4500kWFigure 17.7 (c) Surge impedance of HT motors as a function ofoutput powerThe magnitude of the TRV, which is the product ofZ, x IJI, being the current chopped), can be determinedwith the help of curves shown in Figure 17.8. Thesecurves have been established with the help of the curvesin Figures 17.7(a) and (b) and the values of I,. assumedfor various types of interrupters. The curves in Figures

17/564 Industrial Power Engineering and Applications Handbookterminal equipment should be able to absorb. This canbe determined by a simple equation:V,’W = - x t x 10’ kW-s or kJ (1 W = 1 j/s)z,(17.3)whereW = energy released in kW-s or kJV, = prospective crest of the surge in kVZ, = surge impedance of the power system and theterminal equipment in SZt = duration for which the surge will exibt (in seconds).Accordingly the energy released by a switching surge isusually many times greater than a lightning strike or afront of wave (FOW) surge, since t2 (switching) >> t2(lightning) > t2 (FOW).Subsequent examples will illustrate the amount of suchenergies and the means to dissipate them safely.@ Oil-circuit breaker@VCB with a copperchromiumcontact@ SF6-circuit breaker-10 100 1000 4500 10000Circuit surge impedance (Zs) R@VCB with a copper-bismuthcontactZ, = Surge impedance of the circuitlC = Current choppedFigure 17.8 Likely surge voltages that may be developed bydifferent types of circuit breakers as a function of circuit surgeimpedance17.7(a) and (b) reveal, that for larger ratings of machines,the surge impedance will tend to diminish, while forsmaller ratings it will tend to rise. The curves in Figure17.8 will thus reveal, that while a lower value of Z, willtend to dampen the transients, the higher values willtend to enlarge, as corroborated in Section 18.5.1(reflection of waves). Smaller machines are, therefore,subject to higher TRVs compared to larger ones.Power cables connecting various equipment on a powersystem can be considered as lumped capacitances andcan help in the natural dampening of the amplitude (V,)of the prospective surges, besides reducing their steepness( Vt/tl) (Figure 17.6). Typical values of surge impedancesfor cables may be considered around 35-60 a, but thecable length has to be a safe one. Longer cable may raisethe V, beyond the BIL of the machine, as discussed inSection 18.6.2.17.6.5 Surge energyWhen these surges strike a power system or equipment,they release very high energy, which the system and the17.6.6 Velocity of propagationThe velocity of current or voltage waves in any mediumis called the velocity of propagation of electricity in thatmedium. The velocity of electromagnetic waves(electricity) through a conductor is a measure of line orconductor parameters through which it propagates andis represented by(17.4)whereU = velocity of propagation in km/s and is free fromthe frequency of the travelling waveLo = line or conductor mutual inductance in H/km,through which it travels. This will depend uponthe quantum of skin effect and the mutualinduction between two or more adjacent currentcarryingconductors. In overhead lines, it is verylow due to wide spacing, while it may be veryhigh for a cable, as noted later.Co = leakage capacitance in Fkm.The system parameters will vary with voltage. With arise in voltage, the leakage capacitance C will rise andmutual inductance L will fall due to larger clearancesbetween the conductors, and hence, less induction in theadjacent conductors. The product of LC will remain almostthe same for similar types of systems, and hence thevelocity of propagation in one type of system. For instance,the velocity of propagation in an overhead line, even fordifferent system voltages, will almost universally be inthe same range. This is evident from the typical lineparameters in Table 24.l(b), where the product of L andC is almost of the same order. It may, however, differwith the velocity in a cable. In a cable, the conductorsare placed in close proximity to each other and aretransposed (twisted) to nullify the effect of proximity.Also the cable may have a metallic sheathing and hence,provide an electrically symmetrical current-carryingsystem. Cables possess a very low mutual inductance(L) and a very high leakage capacitance (C).The product of L and C in cables is very high and so

Voltage surges-causes, effects and remedies 17665the velocity of propagation very low. For a 33 kV,3 x 300 mm- XLPE cable, for instance, taking parametersfrom Table A16.15 Appendix 1 of Chapter 16XI,= 0.1 15 R /km at 50 Hzorand.. . u=L=- 0'115 = 0.336 x IO-' Hkrn27TjC = 0.23 p F/km140.366 x 10 -3 x 0.23 x lo-"= 1.09 x 10' km/s or 109 m/pskm/sand for a 3.3 kV. 3c x 95 mm2 PVC cable from TableA163X, = 0.088 Wkm at SO Hz:. L = __ 0'088 = 0.28 x Hkrn2x.fand C = 0.7 I5 x I 0-6 pF/km:. u= I40.28 x IO-' x 0.715 x= 0.706 x IOs km/s or 70.6 dps.Similarly, consider an overhead line having thefollowing parameters:System voltage = 220 kVFrequency = SO HzXI- = 8.249 x IO" R/kmand R or - = 1.42 xx,Q/km..1.. u= -\:2.63 x IO-' x 0.45 x= 2.907 x 10' km/s(Also refer to Table 24.l(b). The variation in lineparameters considered here and produced in Table 24.1 (b)may be due to change in line configuration and spacings.)This is almost equal to the speed of light in free spaceand is true for straight conductors without any joints ordiscontinuities.The velocity of propagation in a cable is much lessthan in an overhead line, as noted above. The frequency(f') of the electromagnetic waves at which they propagatehas no bearing on the speed of propagation. The lightningsurges, which propagate at very high frequency (considera 12/50 ps wave, where 1.2 ps means the duration ofone quarter of a cycle, the frequency of this wave in freespace will be = 1/4 x 1/( 1.2 x = 208 kHz), will alsohave a velocity of propagation equal to light, i.e. 3 x1 Os kds (186000 miles/s) or 300 dps. In motor windingsit may be much less, due to discontinuities and a yethigher product of LC. For the purpose of derivinginferences of the influence of such waves on theperformance of certain equipment, such as an electricmotor, susceptible to such waves, the typical values fora 3.3 kV or 6.6 kV motor may be considered around 1 S-20 d ps in slots and 150-200 m/ps in windings outsidethe slots.17.6.7 Per unit voltage (P.u.)The general practice to express an overvoltage is in p,u,This abbreviation will be used frequently in ourdiscussions. One p.u. is the maximum voltage betweenthe phase and the neutral, i.e.-1 p.u. =v,\ 2 .-(17.5)2317.6.8 Current zeroIn an alternating current system the voltage and currentcomponents travel in the shape of a sinusoidal waveform(Figure 17.9) and oscillate through their natural zeros,100 times a second for a SO Hz system.The instant at which the power frequency current willpass through its natural zero, i.e. where i = 0, is termedits natural current zero (or current zero), wherei = I, sin o t ( 17.6)andi = Instantaneous current component at any instantI = r.m.s. value of the currentI,,, = peak value of the currentw = angular velocity = 2 q'"t = timeNoteThe term 'natural' is used to differentiate a power frequency currentwave from an oscillating or transient high-frequency current wave.17.6.9 Arc time constantThis is the time required by the quenching medium of aninterrupting device to regain its original dielectric strengthafter the final current zero.,/b Naturalcurrent zeros /'I-Time (mt)Naturalcurrent zeroFigure 17.9 Representation of a sinusoidal current waveform

17/566 Industrial Power Engineering and Applications Handbook17.7 Causes of steep-rising surges17.7.1 Arc re-ignition or restrikesThis may occur between the parting contacts of aninterrupting device while interrupting a low-power factor(p.f.) current. It is the re-ignition of the arc plasma betweenthe parting contacts of the interrupting device. It mayreappear after a natural current zero, due to a high TRV,which the dielectric medium between the contact gapmay not be able to withstand and break down. Figure19.3 illustrates the theory of ionization and de-ionizationof arc plasma. The same is true during a closing sequencewhen the contact gap, just before closing, falls short ofthe dielectric strength of the medium in which it is making,and breaks down. It is commonly termed as the restrikeof the arc. This does not mean that the contacts that areparting will close again but the interrupting circuit, thatwill close momentarily, through re-ignition of the arcplasma, until the next natural current zero, or until contactsare made during a closing sequence.This may better be understood with the help of Figure17.10, when the making contacts begin to separate. Thearc plasma becomes de-ionized by the immediate nextcurrent zero. At this current zero a recovery voltage acrossthe parting contacts will appear the magnitude of whichwill depend upon the p.f. of the circuit and the instant atwhich the interruption occurs on the voltage wave, asillustrated in Figure 17.11, through curves (a) to (d).The higher the p.f. of the interrupting circuit (Le. underhealthy conditions), the closer the system voltage andcurrent phasors will be. This will cause the recoveryvoltage to be of a moderate magnitude (zero voltage atunity p.f. curve (a)) causing no re-ignition of the arcplasma and the circuit will interrupt. At lower p.f.s,however, such as during starting conditions of a motoror a transformer, or on fault, the recovery voltage at acurrent zero will be very high (curves (c) and (d)) and itmay be enough to break down the dielectric mediumbetween the parting contacts if they are still close toeach other. This will cause a re-ignition of the arc plasmathat had de-ionized immediately before current zero. Thearc will thus remain ionized and will temporarily makethe electrical contacts until the next natural current zero.The TRV, now higher than before due to reflections, willtry to break down the dielectric strength between theparting contacts if these fell short of the required level ofthe impressed TRV. This will establish an arc yet again,while the contacts have travelled farther apart. The processof such arc re-strikes is termed multiple re-ignitions ormore commonly multiple restrikes. They continue untilthe arc plasma is finally extinguished and the breakerhas fully interrupted. It may be noted that the movingcontact of an interrupting device is spring loaded and ona trip command tries to part with some force from thefixed contact. It will continue to move away until itreaches its far end and interrupt the contacts fully.Irrespective of the dielectric strength of the arc chamberand amplitude of the TRV, the situation will attenuate ata stage when the contact gap has become large enoughand the TRV is unable to break the dielectric mediumacross the gap to cause a further restrike.Since TRV and the dielectric strength between theparting contacts both rise gradually and rapidly (TRVafter every restrike due to reflections and dielectric strengthdue to longer contact travel) there is almost a race betweenthe two. Depending upon which rises faster than theother, will there be a restrike of the arc or an attenuationof the TRV. Whichever may happen faster will prevailuntil the next current zero at least, and until a fullattenuation of the situation or circuit interruption isreached. The theory of arc re-ignition is, therefore, alsotermed the dielectric race theory. Figure 17.10 roughlyillustrates this phenomenon. The actual waveforms willbe much more complex being of a transient nature andcan be obtained through oscillograms.Arc extinguishesCurrent throughthe interrupterCommencement offault, beginning ofcontact separationi TRV after first i TRV after second irestrikeImmediate restrike1st current zero2nd current zero 3rd current zero(Arc restrikes) (Arc restrikes) (Arc restrikes)(Dielectric strength betweenthe contact gap is enoughto sustain the TRV and thecircuit is interrupted)Figure 17.10 Approximate representation of arc re-ignition during a fault-interrupting process

ZVoltage surges-causes, effects and remedies 17/567VoltagewaveCurrent wavelmnufiusin w,P.f. = 10$= 0sin 0" = 0, v, = 0-+c- d=OY -I = I , sin 4VoltagewaveCurrentwavepf = 8i$ = 36 87'sin 36 87"= 0 60.6 V,v, =Voltagewave-4 C d= 72.54"p.f. = 0.3d = 72.54"sin 72.54" = 0.95:. V, = 0.95 V,Voltagewavevmwv,= v,1 1pf =o$ = 90"+ I, $ = 90- sln 90" = 1CurrentwaveFigure 17.11(C)-m,(d) -Amplitude of recovery voltage at a current zero during a switching operation of an inductive circuit at different p.f.s17.7.2 Switching surgesSwitching surges in an inductive circuitWhen a coil of inductance L, consisting of Z number ofturns. is energized through a voltage e, the flux, 4, linkingthe inductive circuit will undergo a rapid change fromone peak (+ve) to the other (-ve), within one-half of acycle of the voltagc wave (Figure I .3), i.e.+e = ~ (d$ldt = - L (dildt) /The negative sign indicates the direction of the e.m.f.that opposes the change in the flux and thus the current. 3 %-Referring to a normal magnetizing curve of an inductivecore. it will be observed that the normal peak flux occurs 0near the saturation point (Figure 17.12). At the instant ofswitching, the flux may go up to 2&, (Figure 1.3) andcause an excessive saturation of the magnetic core, makingthe induced e.m.f. rise disproportionately. Hence thephenomenon of switching surges. If the core already hassome residual flux in its magnetic circuit, at the instantof switching, the surge voltage will attain yet highervalues and further intensify the switching surges.Excessive saturation of the magnetic core reduces thevalue of L, and raises the magnetizing current disproportionately,say up to ten times the normal Value Or Figure 17.12even more. as discussed in Section 1.2.inductor coil-'-I, I,(max.) = 10 I ,Magnetizing current (I,,,)or soA normal magnetizing characteristic of an

17/568 Industrial Power Engineering and Applications HandbookTo analyse switching surges we may consider thefollowing two possibilities:Surges generated during a switching ON or contactmakingoperation, andSurges generated during a switching OFF or contactinterruptingoperation.(ii) Surges generated during a switching ‘ON’operationField data collected from various sources have revealedfailure of a motor’s windings, even during an energizingprocess. It has been shown that a motor may be subjectto an overvoltage of 3-5 p.u. with a front time ‘t]’ as lowas 1 ps or even less (signifying the steepness of theTRV) during switching ON. The phenomenon of aswitching surge is thus a matter of a few microsecondsonly, resulting from the restrike of the transient recoveryvoltage (TRV) during its first few cycles only, at a veryhigh transient frequency, in the range of 5-100 kHz.To explain this, consider a switch being closed on amotor. The moving contacts will approach the fixedcontacts of the switching device and before they close, astage would arise when the dielectric strength of thegradually diminishing gap between the closing contactswill no longer be able to withstand the system voltageand breakdown, causing an arc between the closingcontacts. Under this condition, the voltage across thegap and thus also across the motor terminals, may rise inthe following way:By the first pre-strike of the arc gap, the TRV mayreach 1 P.u., when the motor is assumed to be at astandstill without any self-induced e.m.f. before closingof the contacts.When this TRV of 1 p.u. reaches the motor terminals,it will cxperience a surge reflection and almost double,subjecting the motor windings to a stress of almost 2p.u. (more appropriately 1.5-1.8 P.u.) as measuredduring actual tests (see Pretorius and Eriksson, 1982).It is less than 2, due to the circuit impedance, andprovides a dampening effect (Section 18.5.1) as wellas other effects, such as more than one junction of theinterconnecting cables from the switching device upto the motor tcrminals, which will also dampen thequantum of the reflected wave.To make all the three poles simultaneously of aswitching device during a switching ON sequence israther impractical, whatever precision of the closingmechanism of the switching device be achieved. Thisis due to possible variations, even negligible ones, inthe three contact travels or actual contact making. It isgenerally one pole of the device that will make firstand then the other two poles will make. There couldbe a gap of a few milliseconds between the first contactmaking and the second. This aspect is vital whenanalysing the surge conditions during a switching ON.The closing of one pole causes an oscillation in thecorresponding phase of the motor winding which leadsto oscillations in the other two phases that are still open.It has been seen that if the first pole prestrikes at themaximum voltage, i.e. at 1 p.u. across the making contacts,the peak voltage across the other two poles may approacha value of0.5 P.U. + (1.5-1.8 P.u.)or up to 2.0 to 2.3 p.u. (See Cormick and Thompson,1982)At this voltage when these two poles prestrike they causea surge voltage at the motor terminals of1.5-1.8 times the 2.0-2.3 p.u.i.e. 1.5 x 2 p.u. or 3 p.u. to 1.8 x 2.3 p.u. or 4.14 p.u.The above would be true if the motor is assumed to beat a standstill. If the motor had been running at almostfull spccd, this voltage would assume yet higherproportions, say, up to 5 p.u. due to the motor’s ownself-induced e.m.f., which may fall phase apart with thesystem voltage. Such a situation may arise during a fastbus transfer, when a running motor is switched overfrom one source to another or during a re-accelerationperiod after a momentary power failure. But since theamplitude and rate of rise of the recovery voltage (r.r,r.v.)at switching ON are influenced by the surge impedanceof the closing circuit, which is formed by the surgeimpedance of the motor and the interconnecting cables,the time of rise, t, (Figure 17.6) and the amplitude of thesurge voltage, Vt, would rise with the length of the cablebetween the switching device and the motor terminals(Section 18.6.2). Such transient voltages will exist onthe system just lip to the contact making and are thus ofextremely short duration (in ps).Such prestrikes before contact making are a naturalphenomenon and may occur in all types of switchingdevices, such as an OCB, MOCB, ABCB, SF6 or vacuuminterrupters. But the severity of the prestrikes and themagnitude of the transient voltages will depend upon themedium of quenching, which will determine the contactgap, before occurs a prestrike (Figure 19.1) and its speedof closing.Consider a vacuum interrupter having a dielectricstrength of 50 kV at approximately 1.2 mm contact gap(Figure 19.1). Assuming the dielectric properties to bealmost linear in this region, the contact gap, whileswitching a 6.0 kV motor for instance, will break downwhen the contacts are:6 x1.2- or 0.144 mm apart50Considering the speed of the moving contact as 0.6 m/s(typical), then the duration of prestrikes before this contactwill touch the fixed contact will be= 240 psFor the parameters of the motor circuit, if the frequencyof the TRV is considered as 17 kHz. arising out of

Voltage surges-causes, effects and remedies 171569prestrikes, then the number of prestrikes at transientfrequency before the contacts make will be2" x 17 x IO3 x 240 x 10 'or 8 ("each one cycle will cause two strikes).The prestrikes will give rise to voltage surges from 3 to5 p.u. as discussed above and for which all the terminaland the interconnecting cables must be suitable. Thefollowing are field data collected from actual operationsto illustrate this phenomenon:On switching ON a 6.0 kV motor, a front time as lowas 0.5 ps and the peak voltage transient, V,, up to ISkV. i.e. 3 p.u.(I p.u. = :t! 1' 2 x 6 kVhas been measured.Voltage transients up to 3.5 p.u. with front times asshort as 0.2 ,us are mentioned in Working Group 13.02of Study Committee 13 (1981) and Slamecka (1983).Thus a switching 'ON' phenomenon may give rise tosteep-fronted waves, with a front time as low as 0.2 ,usand high to very high TRVs (V,) with an amplitude up to3.0-5 p.u. Both are the causes of damage to the windings'insulation of a motor.NoteAbove we have analysed the case of an induction motor during aswitching sequence for systems of 2.4 kV and above. Thephenomenon of voltage surges in transformers, capacitors,interconnecting cables or overhead lines etc. is no different, as thecircuit conditions and sequence of Fwitching will remain the samefor all.It is another matter that all such equipment would have a betterinsulation lcvel (BIL) compared to an induction motor and may nothe as endangered by such surges as the motor. In the subsequenttext we have placed more emphasis on motors. being typical of all.(iii) Surges generated during an interrupting sequenceThis can be analysed in the following steps:When the interrupting circuit has a high p.f. (cos @),i.e. a healthy tripping and also circuits that are noncapacitive.When the interrupting circuit is underloaded and has alow p.f., i.e. when it is highly inductive and also circuitsthat are capacitive.Interrupting large inductive or capacitive currents.17.7.3 Healthy trippingIn the first case, it is easy to interrupt the circuit, whichposes no problem of current chopping (premature tripping)before a current zero or a prolonged arc after a currentzero for the following reasons:1 No current chopping: An induction motor or atransformer should be operating at near full load, topossess a high p.f. and carrying a near full load current2during such an interruption. Even the most modern,very fast operating interrupting devices, such as aVCB (Section 19.5.6) will not be able to cause apremature interruption. The interruption will thus bedevoid of any surge voltage.No restrike at current zero: When the voltage andthe current phasors are close to each other (Figure17.ll(b))orapproachingaunityp.f. (Figure l7.ll(a)),they will reach a current zero almost simultaneously.Even ap.f. such as 0.85 (@= 30") and above will havea residual voltage of not more than 50% of the systemvoltage at the current zero. Generally, an inductionmotor running almost at full load will have a p.f. ofmore than 0.85 (Table 1.9) and a recovery voltage ofnot more than 50% across the parting contacts at acurrent zero. This voltage will be insufficient to breakthe dielectric strength across the gap between theseparating contacts and establish an arc. Hence, theinterruption of a circuit at a high p.f. will be devoidmultiple restrikes that are responsible for the surgevoltages. The circuit will interrupt at the first currentzero. The situation will be similar in a transformer.17.7.4 Interrupting small inductive or capacitivecurrentsin the second case, when the circuit has a low p.f. orcarries a capacitive charging current a condition whichmay occur in the following cases (for ease of analysiswe have classified them into two categories):When interrupting an underloaded induction motor, ora transformer, such as when they are operating at ornear no load.Interrupting a cable or an overhead line when extremelyunderloaded, such as when operating at a near noload.Normally the interruption should take place at the firstnatural current zero. Modern high-speed interruptingdevices, however, may interrupt too small a current thanrated, such as noted above, prematurely, i.e., before anatural current zero. This may lead to steep-fronted highTRVs, similar to those discussed earlier, capable ofrestriking an arc between the parting contacts, until atleast the next current zero. In actual operation, this TRVis seen to risc to 2.5-3 p.u. The phenomenon is termedcurrent chopping and is dealt with separately in Section19.6. It is detrimental to a successful interruption of theswitching device and may cause damage to the terminal'sequipmcnt, such as the interconnecting cables, inductionmotors and transformers etc.17.7.5 Interrupting large inductive or capacitivecurrentsInterrupting when the motor is in a locked rotorcondition (Section 1.2) or is still accelerating andcarrying a highly inductive curretit of the order of sixto seven times the rated current at a p.f. of about 0.2-0.3.

orseconds171570 Industrial Power Engineering and Applications HandbookInterrupting an induction motor under a stalled conditionwhich is almost a locked rotor condition.Tripping an inductive or capacitive circuit immediatelyafter a switch ON as a result of a momentary fault orfor whatever reason and even during a fast bus transfer(temporary tripping and reswitching).Interrupting the equipment such as an induction motor,a transformer, a cable or an overhead line on a fault,such as on a short-circuit or a ground fault.Interrupting a charged capacitor.In all the above cases, whether the load is inductive orcapacitive, the voltage and the current phasors are morethan 70" apart. At a current zero, the system voltage willreappear almost in full (around 95% or so) across thepasting contacts of the interrupting device (Figure17.1 l(c)). This voltage is detrimental to the successfulinterruption of the circuit at the current zero, unlike inthe previous case.Circuit interruption is a transient condition, as itconstitutes abrupt changes in the circuit parameters Land C, which also alter the characteristics of the transientwave and its behaviour. The characteristics of a transientwave depends upon the circuit's surge impedance, Z,,which, in turn, depends upon the circuit parameters Land C:Therefore, if the arc does not extinguish at the first currentzero, it will give rise to voltage surges (TRVs) at veryhigh surge frequencies(fs=&)of the order of 5-1 00 kHz or more.The generation of voltage surges is almost the same asdiscussed when the circuit was being closed (Section17.7.2(ii)). The sequence of generation of the voltagesurges can be analysed for more clarity as follows.The magnitude of surge voltages will depend upon theinstant at which interruption of the switching device willtake place, on the voltage and the current waves. Referto Figure 17.1 1, illustrating the relative displacement ofthe voltage and the current waveforms at low p.f.s of,say, 0 and 0.3, which may occur during such interruptingconditions as noted above. This is when the load wasinductive. Had the load been capacitive, the voltage andcurrent phasors would have again been almost 90" apart,the current leading the voltage by nearly 90", and sameinterrupting conditions would arise as in the case of aninductive load.Under such switching conditions the voltage and currentwaves would be out of phase by almost 70' or more, andat every current zero the recovery voltage would be almostat its peak (95% or so), which would reappear across theinterrupting contacts, re-ignite the arc plasma and causea reflected wave of similar magnitude. The situation isfurther aggravated by the motor's or the transformer'sself-induced e.m.f. or the capacitor's own charge, whichmay also fall phase apart with the recovery voltage andadd up to the same. The cumulative effect of all suchvoltages may cause a significant rise in the TRV, whichmay become steep fronted and assume almost a similarmagnitude as we have analysed in the case of a 'switchingon'sequence, i.e. up to 3-5 P.u., at surge frequencies,with a front time of even less than 1 ,us (Section 17.7.2(ii)).The phenomenon is quite complex. Theoretically onlyan approximate analysis can be carried out, as we havedone in the case of a switching ON. Use of oscillograms,as illustrated in Figure 17.5, may be an accurate meansof measuring the exact magnitude and shape, i.e. steepnessand front time of the TRVs, to determine the r.r.r.v ofsuch high-frequency TRVs. The consequences of suchsteep-fronted TRVs on circuit interruption may pose thefollowing problems in successful interruption:The magnitude of surge voltage will depend on theinstant at which interruption of the switching deviceshall takes place, as illustrated in Figure 17.11. ATRV of 3-5 P.u., for a normal interrupting device,may be large enough to prevent the arc from beingextinguished at the next current zero. It may be capableof breaking the dielectric strength of the interruptingcontacts and re-establishing an arc after a current zero,until the next natural current zero at least. In thesubsequent half cycle, after the first current zero anduntil the next current zero, the arc will repeatedlyrestrike at a very high transient frequency of the orderof 5-100 kHz or more, depending upon the circuitconstants L and C. For a 12 kHz surge frequency, forinstance, the arc will restrike for at least 120 cyclesoff, or 240 times, as determined below, before thenext current zero:Time for half a cycle of a 50 Hz normal system= -Ix -1or1~2 50 100In this time the surge frequency will undergo12 x IO' x1~ 120 cycles100The repeated restrikes of the arc for about 240 timeswill result in steep-fronted, extremely severe TRVs,sometimes even more severe than a lightning surge.They are capable of inflicting serious damage to theinterconnecting cables and the end turns of theconnected equipment, be it a motor or a transformer,besides damage to the interrupting device itself andits interrupting contacts. The generation of surgevoltages beyond 5 p.u. is, however, seldom noticeddue to self-attenuation. The circuit parametersthemselves provide the required dampening effect.In the above one-half of a cycle of the natural frequency,the moving contact would separate by 1/100 s. Infast-operating interrupting devices, this time isnormally adequate to achieve a sufficient contact gapand to restore adequate dielectric strength to interruptthe circuit by the next current zero and allow no furtherrestrikes of the arc. Otherwise the arc may be reestablishedand the process may continue until theTRV itself attenuates or the contact moves farther

Voltage surges-causes, effects and remedies 17/571away, to extinguish the arc naturally as a result of alarger contact gap. The fast-operating interruptingdevices such as an SF, or a VCB may interrupt thecircuit at the first current zero and the worst, by thenext current zero whereas other interrupting devicesmay take a half to one and a half cycles of normalfrequency to interrupt the circuit and extinguish thearc. This is long enough to cause severe damage to allthe connected equipment and components unlessadequate measures are taken to protect them byproviding surge suppressors or surge capacitors orboth.3 The same theory will apply in capacitor switchingexcept that the interrupting current will now be leadingthe voltage by almost 90" instead of lagging. See alsoSection 23.5 to determine the amount of surge voltagesand inrush currents.NoteWhile interrupting low p.f currents it is observed that the generatedvoltage surges were higher when interrupting higher currentscompared lo interrupting lower currents. The statement is generalizedon the basis of experiments conducted on an induction motor byCHIP (5ee Central Board of Irrigation and Power. 1995).17.7.6 Switching surges on an LT systemThese do not occur on an LT system due to inadequateswitching voltage. An LT system is therefore not affectedduring a switching operation like an HT system. At mostthere can be an overvoltage, up to twice the rated voltageacross the terminal equipment, when a switch is closedon a circuit that is already charged, such as a capacitorbank or an induction motor, and the impressed voltagefalls phase apart with the induced e.m.f. The followingis a brief analysis to corroborate this statement:The BILof the LT system and the equipment connectedon it are suitable for withstanding such overvoltages(Tables 11.4, and 14.1 or 14.2).As ii result of a better insulation and insulating mediumof an LT interrupting device, compared to its voltagerating, there are no restrikes of the contact gap after acurrent zero. The TRV is insufficient to establish anarc.However, abnormal transient voltages are sometimesnoticed on the LV side of a transformer as a result oftransference from the HV side or due to contactbouncing, which are not necessarily switching surges.Although the line side impedance would greatly dampensuch transferred surges and in all probability may notbe a cause for concern it is advisable to protect largeLT motors which are installed next to such a source,such as a transformer and which experience a transferredsurge, which may be enough to cause damage. Formoredetails on transferred surges refer to Section 185.2.17.7.7 Voltage surges caused by contactbouncingThis occurs in a contactor during it$ closing operation,and may result in pitting and burning of contacts. It mayalso give rise to high-voltage transients (not necessarilyswitching transients), similar to the phenomenon of acurrent chopping (Section 19.6). The moving contacts,while moving, acquire a certain amount of velocity andmomentum, and release this, in the form of kinetic energy,to the fixed contacts as soon as they make contact withthem. The contact mechanism, being flexible and springloaded (Figure 17.13), has a tendency to bounce back.After being bounced, the contacts attempt to close again.The situation is a near replica of an open transient conditiondiscussed in Section 4.2.2( 1) and so are its consequences.During bouncing, however, the circuit is never broken,as arcing persists between the contacts. The contactsmay bounce again, and the process may be repeated,depending upon the design of the closing mechanism,until the process attenuates and the contacts finally close.Figure 17.13 illustrates such a phenomenon. This is nota desirable feature, yet it is not totally avoidable, due todesign limitations. But extremely low-bounce closingArcing Bridge ,y- ~ Spring loadinghornContact holder1 - Moving contactrntactFixed terminal\ Fixed terminal\Base ~ _ _ _Fixed contact1. Contacts in open position2. Contacts just making33. Contacts fully made and under spring pressure(before bouncing back)Figure 17.13(a) The phenomenon of contact bouncing

17/572 Industrial Power Engineering and Applications HandbookLeaf springsMovinqFixedcontactsCOllmounted onthe movingcontactsSpring-loaded moving contact assemblyFigure 17.13(b) A typical view of a contactor showing arrangement of fixed and moving contacts (Courtesy: L & T)mechanisms have been developed by the leadingmanufacturers to ensure an almost bounce-free closingof the contacts. These mechanisms have a closing timeas low as 0.5 ms or one fortieth of a cycle of a 50 Hzsystem by adopting a near-balanced closing action througha low-inertia magnetic coil, leaf spring cushioning effectof moving contacts and contact spring, etc.When switching an induction motor, there may thereforealso be a transient current for a few milliseconds in additionto the starting current. The magnitude of surges willdepend upon the instant at which the contacts will makeon the voltage wave. It may be up to fifteen to twentytimes the rated current, or two or three times the startingcurrent. Thus, a contactor, on bouncing, will enhance itsvoltage amplitude as discussed in Section 17.7.2(ii) andwill continue to do so until the process attenuates. Thecontacts may even weld together in such circumstances.The situation becomes even more difficult when the loadis inductive or capacitive, which it is in most cases. Thecontact interruption, although it poses similar problemsto those discussed in Section 17.7.2(iii), is less severe inthis case than contact making, as there is now no contactbouncing.To tackle the problem of arc quenching, large LTcontactors, say, 100 A and above, and all HT contactorsare provided with an arc chamber with splitter plates,(see Figures 19.11 and 19.12).It is for this reason that a contactor is classified by theduty it has to perform, according to IEC 60947-4-1, asnoted in Table 12.5.17.7.8 Voltage spikes developed by static drivesYet another source of voltage surges is the output of thesolid-state drives as discussed in Section 6.13.17.8 Effect of steep-fronted TRVs onthe terminal equipment (motoras the basis)A healthy contact making or interruption, not associatedwith any restrike voltage transients, can be assumed ascarrying only the nominal switching surges, as definedby a 250/2500 ps impulse, with a front time of 10 ps andabove (Section 17.3.1, Figure 17.2(b)). This will usuallycause no harm to the terminal equipment. But a switchingsequence, involving a restrike of the arc, between themoving and the fixed contacts of the interrupting device,gives rise to steep-fronted transient voltages of up to 3-5 P.u., as discussed earlier. The switching surges mayexist on the system for not more than a half to one anda half cycles of the power frequency, and can have afront time tl as brief as 1 ps or less. In extreme cases, itcan even reach a low of 0.2 ps (see Working Group13.02 of Study Committee 13 (1981) and Slamecka(1983)) and become capable of causing severe damageto the terminal equipment. All such waves are termedfront of waves (FOWs).When such a surge penetrates electrical equipment

Voltage surges-causes, effects and remedies 17/573such as an induction motor’s winding, most of its stressmay appear only across a small part of the windings (theline end coil or the first coil). The front of the surge willbecome less steep as it penetrates the windings, due tolumped capacitances C of the winding insulation andalso because of partial reflections and damped refractions(Section 18.5.1) from the discontinuities of the windingsas well as eddy currents (Section I .6.2.A-iv) which willadd to ‘L’. The interturn voltage stress will thus be higherfor the line end coil of the windings than the subsequentcoils as shown by the following example.Example 17.1A simple switching surge of, say, 250/2500 ps (Sections 17.3.1and 17.6.6 for velocity) would mean that in free space,considering the speed of propagation as 200 m/p, it willpropagate by250 x 200 x lo6 = 50 km106 1000by the time it reaches its first peak. Considering an averagevelocity of propagation in windings inside the slots andoverhangs as 100 m/ps (typical), the surge in windings willtravel 25 km, by the time it reaches its first peak. This distanceis too large for the normal length of a motor or a transformerwindings or the interconnecting cables and/or the overheadlines etc. For all theoretical considerations, therefore, we canassume this switching surge to be uniformly distributed overthe entire length of the current-carrying conductors and thusis only moderately stringent. It is possible that the surge maynot even reach its first peak by the time it has travelled theentire length of the windings. The maximum transient stressper unit length of the conductor, %/I, and also between theturns of the coils will be low (V, being the amplitude of theprospective surge and Y the length of the winding per phase).The circuit constants L and C through which the firstpeak of the switching surge will travel determine theshape and severity (steepness) of the surge. As the wavetravels forwards it will change these constants at everyjunction and the discontinuities of the windings.Consequently change the shape of the wave front, whichwill be diminishing gradually in steepness until it losesall its severity.If the switching is also associated with repeated restrikesof the contact gap of the switching device causing steepfrontedvoltage transients with a magnitude of 3.5 p.u.or so and a front time, r,. as low as 0.2 ps (assumed), theahove concept of transient voltage distribution over thelength of windings will change abnormally. It wouldnow travel only 0.2/106 x 100 x IO6, i.e. 20 m, by thetime it reaches its first peak. This distance is too shortand may involve just one coil or a few end turns of it, beit motor or a transformer windings, and very short lengthsof the interconnecting cables or the overhead lines etc.Such switching surges will be non-uniformly distributed,as the voltage stress of 3.5 p.u. will be distributed justover a length of 20 m. For a system voltage of 6.6 kV,the transient stress per unit length would be913 V/mIf we consider some length of the interconnecting cablesand deduct this from the total length of 20 m. the interturnstresses may assume much more dangerous proportions.The interturn insulation of the windings must be suitableto withstand even a higher impulse level than calculatedabove. As standard practice, the whole coil, if it is preformed,is tested for the prescribed impulse withstandlevel as in Table 11.6. It should be ensured that thc actualsteepness and amplitude of the FOW is well within theprescribed BIL.The first few turns of the line end coil of a motor ortransformer and short lengths of interconnecting cablesand overhead lines and their associated terminalequipment, will thus be subject to severe stresses andwill be rendered vulnerable to damage by such steepfrontedtranbient voltages.Various experiments conducted by different agencieshave revealed that the voltage stress across the first coilalone may be as high as 70-904 of the total transientvoltage acr05b a motor windings, having a front time. t,as low as 0.2 ,us. (see Working Group 13.02 of StudyCommittee 13 (1981) and Slamecka (1983)).Slamecka ( 1983) has produced curves. to illustrate theTRVs appearing across the entrance coil for differentfront times t,, for different lengths of conductor of thefirst coil (Figure 17.14). From these curves, one candraw an inference that the greater the length of the entrancecoil of the machine, the higher will be the intcrturn stressesto which it will be subject and vice versa. As illustratedin Example 17. I, a transient voltage wave having a fronttime of 0.2 ps will propagate about 20 m in the windingsand will affect only this length of the entrance coil. Thisis illustrated in Figure 17.14. As the TRV propagatesahead and penetrates deeper into the slots, it will haveits amplitude dampened and will subject the coils aheadto much less TRV. The curves in Figure 17.14 also suggestthat if the length of the entrance coil is reduced, it willbe subject to fewer stresses, as the remaining TRV willbecome distributed to the following coils. A slightly lesssteep surge, which is more probable, may travel throughthe length of the interconnecting cables and yet envelop01 02 04 07 1 2 4 7 10Front time (t,) ps -----c-8a,cmca, c> a lf-06 mFigure 17.14 A transient voltage stress across the first (entrance)coil of an induction motor as a function of front time (t,) andlength of entrance coil ’(’ (from Slamecka, 1983)_J a,

17/574 Industrial Power Engineering and Applications Handbookthe whole windings. It has been observed that TRVswith a front time of 0.5 ,us and more are fairly evenlydistributed over the entire length of windings.Different installations with different lengths ofinterconnecting cables, type of switching device, sizeand design parameters of a machine (particularly a rotatingmachine), will influence the steepness of surges and theirdistribution over the length of windings. Different studiesat different locations have revealed the followinginformation:If the surge is very steep, say, with a front time of 0.2ps or less, and the cable length is short, say, 10 m(presumed), it may inflict all its severity to only theentrance coil of the machine and sometimes even onlya few entrance turns of this coil. However, with thetechnological improve-ment of switching devices, thearc prestrikes are now less predominant or non-existentand hence, the switching surge may not be as steep asdescribed here.In all probability the steepness may be between 0.2and 0.4 ,us, in which case, depending upon the lengthof the interconnecting cables and of each motor coil,the maximum surge may appear across the last fewturns of the entrance coil or the first few turns of thesecond coil, as a result of multiple reflections at thediscontinuities and the joints. Multiple reflections mayraise the amplitude of the incidence wave up to twiceits initial value, as discussed later. Alternatively, if itbe fairly evenly distributed through-out the wholewinding. The last coil where it makes its star point ora few last turns of this last coil may be subject tosevere voltage surges due to multiple reflections at thestar point.Generally, it is the steepness of the surge that has agreater severity on interconnecting cables and machinewindings. The travelling waves and their partial reflectionsat the discontinuities of the windings influence the surge’samplitude and distribution over the length of the windings.The windings’ length, shape, inductance L and leakagecapacitance C and speed and size of the machine may betermed vital parameters that play a significant role indetermining the prospective amplitude (V,) of a voltagesurge and its distribution over the length of the windings.Below, we briefly discuss such parameters to betterunderstand the phenomenon of voltage surges and theirinfluence on the terminal equipment and interconnectingcables.17.8.1 Surge impedanceThe value of L and C of a machine will determine itssurge impedance Z, = and surge frequencyf, = 1/(27cTm). A low surge impedance will help todampen the prospective amplitude, V,, of the surge voltageand hence the stresses on the windings. A low surgefrequency will help to limit the number of restrikes ofthe interrupter and in turn the amplitude of the surge, VVA lowerf, will also reduce the steepness, V,/t,, of thesurge. Hence, a low Z, as well as a lowf, are alwaysdesirable. It is possible to modify the Z, of a machine atthe design stage to contain the prospective voltage, V,,within safe limits if the type of interrupter, length ofinterconnecting cables and the characteristics of the likelyprospective surges are known.NoteThe length of the interconnecting cables plays a vital role in containingor enhancing the severity of the incidence wave. After the interrupter,the surge enters the cable and propagates ahead. As it propagates,it rises in amplitude, at a rate of Vt/t, (Figure 17.15) until it reachesthe far end of the cable. The longer the cable, the higher willbecome the amplitude of the incidence wave which will be moresevere for the terminal equipment (refer to protective distances,Section 18.6.2). The length of the interconnecting cable is thereforerecommended to be as short as possible. The manufacturer of theinterrupting device can suggest a safe length for different sizes ofcables, depending upon the voltage of the system and the equipmentit is feeding.The approximate values of the surge impedances of motors andtransformers of various ratings and voltages may be obtained fromtheir manufacturers and drawn in the form of a graph, as in Figure17.7. With the help of these values, one can determine the likelysurge voltage from the graphs of Figure 17.8 that may developwhen using different interrupting devices.To this value is added thc peak phase voltage of 1 P.u., of thesystem to determine the total peak voltage likely to arise on a noloadinterruption. This total peak voltage should be less than theimpulse voltage withstand level of the equipment. For motors, itshould be well within the impulse test values given in Table 11.6.lhe effect of cable length is ignored, presuming that the cablelength is short and does not contribute in enhancing the severity ofthe incidence wave.Example 17.2Consider a 500 hp (e 500 kVA) 6.6 kV motor. The surgeimpedance from Figure 17.7(b) Z, = 4500 R. This surgeimpedance may generate the following surge voltages duringan interruption by different interrupting devices, dependingupon their chopping currents, according to Figure 17.8:t, > r;’ > r;v, > v,- z v,.Longer the cable,higher will be ‘t’ andmore severe willbecome the ‘TRV’-tlRise time (t)Figure 17.15 Influence of cable length on the TRV

~ ~ ~ ~ ~ ~ ~ ~~~ ~ ~ ~ ~~ -~ -~~~ ~ -~ -srtrgcs~~ ~~~~ ~~ ~~~ ~~~~ ~~ ~~ ~~ ~~~ ~ ~-Voltage surges-causes, effects and remedies 17/575OCB (MOCB or LOCB):5 kV + 5.4 kV = 10.4 kV-~[l P.U. = 12 x 6.6 = 5.4 kVEntrance coil(suitable for aBIL of (3p.u.16))SF, breaker: 7.5 kV + 5.4 kV = 12.9 kVVacuum breaker with copper-chromium contacts:6 kV + 5.4 kV = 11.4 kVVacuum breaker with copper-bismuth contacts:35 kV + 5.4 kV = 40.4 kV.From the above it can be noted that the surge voltagesthat may develop during a switching operation, with all typesof breakers except Cu-Bi VCBs, are well within the prescribedimpulse withstand voltage of 16.2 kV as in Table 11.6, column2. In Cu-Bi VCB switching it may be as high as 40.4 kV,which is far beyond the prescribed impulse withstand level of16.2 kV. Such breakers, as standard manufacturing practice,are fitted with metal oxide surge arresters. Otherwise a suitablesurge arrester must be provided near the motor.17.8.2 Number of turns per coilThis is another vital parameter of a machine, influencingthe distribution of voltage surges over the windings. Asnoted earlier, protection of interturn insulation is veryimportant, particularly when the machine has a multiturncoil arrangement and where the turn insulation islikely to be overstressed by the incident waves. Singleturncoils. having the same turn and coil insulation andcoils deeper in the slots, are less influenced. A fast-risingvoltage wave at the motor terminals will lift the potentialof the terminal (entrance) turns, while turns deeper inthe windings or the slots will be subjected to lesser severitydue to discontinuities and partial reflections. They are,therefore, sluggish in responding to the arriving surgesand are subject to attenuated severity of the surge andare less stressed than the entrance turns. To protect theline end coil or its first few turns. it is essential to keepthe length of the coil as short as possible, as illustrated inFigure 17.14, to subject the whole coil to an equal interturnstress. Otherwise the severity of the surge may affectonly the entrance coil or its few entrance turns.Example 17.3Consider a winding having six numbers of coilslphase asshown in Figure 17.16, and suitable for a surge voltage (BIL)of 3 p.u. (Table 11.6). If the surge travels only to the first coil,which is suitable for 3/6 or 0.5 P.u., then all the effect of thesurge (3 p.u) will impinge across this coil alone and greatlyoverstress its interturn insulation. Thus, a 3 p.u. surge wouldstress the interturns by 3 p.u./0.5 p.u. or six times the voltagefor which the insulation of the coil or its interturns are suitableor six times more than a uniformly distributed transient wave.Hence, the relevance of the length of the entrance coil in asteeply rising voltage wave.Consequently an FOW, whose amplitude, V,, is well belowthe insulation level (BIL) of the machine, may adversely affectthe interturn insulation of the first coil. Even if the winding iscapable of withstanding such transients, their repeated-Jyvl= 3p. .Winding showing 6 slots or 6 coils per phaseand one turn per coilIf, there be n number of turns per coil.Vtthen voltage across each turn = -6nFigure 17.16 One-phase winding, illustrating Influence of thearriving surge on the entrance coiloccurrence may gradually degrade its insulating strength to apoint of a possible failure over a period of time. A steepfrontedwave may damage the insulation in the form ofmicroscopic holes called 'pinhole' failures. While a singlepinhole may be of little relevance, a number of them maycause hot spots, leading to eventual failure of the insulation.17.8.3 Rating of a rotating machineIt has been seen that smaller rating HT machines aremore susceptible to steeply rising voltage surges, comparedto larger ratings, due to their relatively weaker interturninsulation and require more careful attention to theiradequate protection. We attempt to explain thisphenomenon as follows:1 In smaller ratings although the ratio of copper toinsulation of the windings in the slot is low (the contentof insulation high), the turn insulation is low, roughlyTable 17.1 Typical ratios of copper to insulation the turns percoil in smaller frame sizesFrame size Ratio of copper Turns per- coil Turn rnsulation(Table 1.2) to insularion to wlrclge~~- ~~6.6 kV 11 kV 6.6 kV II kV- ~- - ~~~400 1.5 0.8 12 20 Highlyvulnerable-~- - -~ ~~500 2.0 1.2 9 15 Moderately630 2.2 1.3 6 IO vulnerable~Note For our discussion, the frame size 400 alone is more critical.Frame sizes 500 and 630 are only indicative to illustrate thecomparison.

17676 Industrial Power Engineering and Applications Handbookin the same proportion. This is because of the highernumber of turns per coil which diminish the dielectricquantity, hence the endurance of the turn insulation,as witnessed during tests. The higher number of turnsalso pose a problem in forming the coils, particularlyat bends, as it is difficult to maintain the same qualityof insulation. Table 17.1 (see Pretorius and Eriksson,1982) suggests typical ratios for different voltagesystems in three smaller frame sizes used for an HTmotor.Lower ratings have low L and C and therefore have ahigher Z, (Figures 17.7 and 17.8) and do not help indampening or taming the surges.3 Similarly, high-speed machines too have more coilsper slot, subjecting the interturn insulation to higherstresses.4 The length of the entrance coil is another importantparameter, which may be subject to most of the voltagestress up to 70-90%, as noted earlier, for very fastrising(tl 50.2 pus) TRVs. So are the last few turns ofthe entrance coil, or the first few turns of the secondcoil, for fast-rising surges (tl = 0.2 - 0.4 ps), or thelast few turns of the last coil making the star point ofthe windings, when the TRV is almost uniformlydistributed (tl approaching 0.4 ps). The steepnessesare only indicative to illustrate the severities of surgesand their influence on the turn insulation. For aparticular size and design of motor, the steepness andits influence may vary somewhat.Other than the above, the interconnccting cable lengthbetween the switching device and the machine also playsan important role in the distribution of the fast-risingvoltage surges as discussed earlier. It is the cable thatwill bear the initial severity of the rising surges, beforethe surges reach the motor terminals. Hence, upon therise time, ti, will depend the safe length of the interconnectingcables to provide the maximum dampening effect.For very fast-rising waves, in the range of 0.2-0.4 ,us orso, cable lengths in the range of 30-300 m are seen toprovide the maximum dampening effect. Actual simulationtests or studies of similar installations are, however,advisable, for a more accurate assessment.For a less steep surge, the situation will be different.Now, the longer the cable, the higher will be the amplitudethat the surge will attain by the time it reaches the motorterminals. For example, for a surge with a risc time, tl of1 ps, the cable must be at least 100 m or more in length(considering the speed of propagation as 100 mlps), andsometimes it may not be practical to provide this. Thesteepness of surge is thus a very vital parameter in decidingan ideal cable length to achieve the desired dampeningeffect through cables. A slightly shorter length than thismay subject the terminal equipment to a near peakamplitude of the arriving surge. It is therefore advisableto keep the length of the interconnecting cables as shortas possible, to subject the terminal equipment to only amoderate amplitude of the arriving surge, much belowits prospective peak. (For more information on the subjectrefer to Section 18.6.2 on protective distances of surgearresters).17.8.4 The need to protect a rotating machinefrom switching surges, contact bouncingand surge transferencesSwitching surges have been seen to be the severest ofall. But it has been a greatly debated subject whether aprotection is overemphasized against switching surges,particularly when the occurrences of such high TRVsmay be rare yet are significant to cause concern, opinionsdiffer. Technological improvements and application ofthe latest techniques in the extinction of arc plasma,making use of high-speed interrupting devices, usingSF6 or vacuum as the insulating and the quenchingmedium, meticulous design of the arc chamber, designand material of the making contacts achieve an interruptionalmost devoid of restriking and adopting the latestinsulating practices for motor insulation (Section 9.3)such as vacuum impregnation and additional bondageand bracing of the windings’ end turns have diminishedif not totally eliminated the effect of these surges onmachines. Nevertheless, to take account of the possibilityof these surges causing damage to the terminal equipment,generally 2.4 kV and above, it is advisable that protectionbe provided as a preventive measure to protect the costlymachines and, all the more costlier, against the risk of ashutdown of a plant in the event of a possible failure ofthe machine. More so when a rotating machine has a lowlevel of insulation (low BIL) compared to an oil-filledtransformer.Dry type equipment, such as rotating machines, havelower impulse withstand levels compared to a liquidtypemachine, such as a transformer or a switchgearassembly. A comparison of impulse voltage withstandlevels for the same system voltage for motors (Table11.6) and for switchgear assemblies (Tables 13.2 or 14.1)will reinforce this point. A motor is always vulnerable toboth internal and external voltage surges. Circuit switchingis the most onerous of all and can overstress the windingsof a machine if it is not protected adequately, leading toan eventual breakdown, if not an immediate failure. Infast-acting devices such as a VCB with Cu-Bi alloycontacts (Section 19.5.6) the manufacturers provide asurge suppressor.During an interruption, an SF, interrupting device isfound to be normally devoid of a switching surge, as thereis no chopping of current. During a closing sequence,however, in both a VCB and an SF, breaker, the switchingsurges are almost within the same range, say, 1.5 to 2.5 P.u.,as recorded during a simulation test on a 400 kW, 6.6 kVmotor (see Central Board of Irrigation and Power, 1995).Misconception1 It is a misconception that only large high-voltagemotors need be provided with surge protection, inpreference to small machines, because they are morelikely to encounter dangerous surges. Analysis ofvarious motor circuits, as noted earlier, indicates thatsmaller and higher-speed motors are more subject tothe effects of voltagc surges rather than the larger orlower-speed motors due to one or more of the followingreasons:

Voltage surges-causes, effects and remedies 17/5770 The lower ratings of rotating machines, having fullload current of around 60 A and less, i.e. 600 kWor less for a 6.6 kV system and 300 kW or less fora 3.3 kV system, are generally prone to causedangerous steep-fronted TRVs when being interruptedon no-load by a VCB or a vacuum contactor,as a result of possible current chopping. For a fullload current of 60 A, the no-load current would beapproximately 50%, i.e. 30 A or even less (Section1.7). It is therefore possible that current choppingmay take place just before a natural current zero ataround 10% if it, i.e. 3 A or so. Refer to Figure17.17. The latest vacuum interrupters with Cu-Cralloy contacts (Section 19.5.6) may not allow thecurrent to reach its natural zero for a normalinterruption and may chop it somewhere near 3-5A and cause TRVs. Moreover, the lowest rating ofthe interrupting device itself may be large enoughfor this current to be interrupted at current zero andcause current chopping.2 Transient voltage damage usually appears in someother form of dielectric failure such as caused bythermal overloading, undervoltage and stalling etc.,totally masking the original cause, which could bedue to switching. Field research and statistical studyof failures have revealed that almost 35% of totaldielectric failures in power stations have beencaused by surges, rather than any other reason (CentralBoard of Irrigation and Power, 1995). Surge protectionfor a smaller motor, say, up to 300 kW in a 3.3 kVand 600 kW in a 6.6 kV system is thus advisable.v = V, sin wt0 95 vm appearing acrossthe contactscurrent choppingVtime (of) p.f = 0.3i=O1 I,i$ = 72.54'sin 72 54" = 0.95or K! I,:. V, = 0.95 V,,,\20.1:. sin ot = - and wf = 4.06' and1'2TRV at current chopping = V, sin 76.6" 0.97 V,Figure 17.17 Current chopping, say, at 10% of an inductivecurrent of 0.3 p.f. giving rise to almost a full system voltage.ClGRE envelope(based on the data obtained froma few manufacturers on a new winding)/-0 1 2 3 4 5 6Front time (t,) psFigure 17.18Dielectric envelopes for a 6.6 kV motorManufacturers of the rotating machines. being thebest judges, to suggest the most appropriate protectionrequired for their machines, depending upon the surgeimpedance of the machine and the likely voltage surgesthat may develop using different types of switchingdevices.CorollaryIn case of static drives also generating similar switchingsurges, their manufacturers provide the safe cable lengthsas a standard practice, Section 6.14.1.Dielectric envelopeThis is a curve that defines the limits of surge voltagesand the corresponding front times that a machine will beable to sustain without a failure during a switchingoperation or a lightning strike. Such a curve must beavailable for all machines and is provided by theirmanufacturers. Figure 17. I8 shows a dielectric envelopefor a 6.6 kV motor as recommended by IEEE, alsoconsidering motor ageing. The figure also shows a curveprovided by Electra (1981; see also Gibbs et (11.. 1989).This curve is based on the data obtained on new machinesfrom a few motor manufacturers.Surge protectionProtection for the machine should be such that the voltagesurges and their rise times, whenever they occur in thesystem, shall fall within this envelope of the machine.(Refer- to Section 17.10 for a total surge protection.)17.9 Determining the severity of atransient17.9.1 Simulated test circuitFrom the above it is essential to predetermine the nature.magnitude and steepness of such transients to decide themost appropriate protection or preventive measures foran HT motor, particularly. for critical installations. such

Motor17/578 Industrial Power Engineering and Applications Handbookas a power station or a process plant, to avoid the leastrisk. In our discussion so far, the circuit conditions, asassumed and the analysis of the voltage surges asgenerated, were purely statistical and varied from machineto machine and from one installation to another. Thepurpose so far has been to provide a general analysis ofsurge voltage phenomenon, its consequences on themachine and possible remedies and/or protection. Forabsolute motor protection, accurate transient conditionsmust be known. To determine the transient conditionsaccurately, a working committee of the IEEE-Cigre hassuggested a simulated test circuit (Electra, 198 I; Gibbset al., 1989) for all manufacturers of interrupting devices(1) to determine the behaviour of their devices, withpredetermined circuit parameters, almost representing anormal supply side of the machine; (2) to assess thebehaviour of the machine during a switching operationand subsequent restrikes, if any, as a guideline for theuser; and (3) to decide on the more appropriate switchingdevice for the machine and its surge protection, based onthe dielectric capability of the machine (Figure 17.18).The test circuit is illustrated in Figure 17.19 simulating:On the supply side of the interrupter, an equivalentbusbar, inductance and lumped capacitance with aprovision to connect p.f. correction capacitors, ifrequired, to represent a replica of the actual installation.On the load side, the interrupter is connected througha cable to the equivalent circuit, with the requiredquantities of lumped resistance and reactance, torepresent the motor to be tested, under a locked rotorcondition. The circuit would also represent aninterruption immediately after a start, to check for themost onerous operating condition for the interrupterto generate the highest surges as discussed earlier.Test resultsThe test circuit may be connected to the test voltage andthen switched to obtain the required oscillograms duringswitching operations and assess the following:Any restrikes and their numberAmplitude of voltage surge (V,) generated andTheir rise time (t,).With these data, one can determine the dielectric curvethe machine must have when switched with such aninterrupting device. This can be compared with the actualdielectric curve of the machine (Figure 17.18) obtainedfrom its manufacturer to decide the compatibility of theinterrupting device for the machine or vice versa and theextent of surge protection, if necessary. For more detailsand results of similar simulation tests, see Central Boardof Irrigation and Power (1 995).17.1 0 Protection of rotating machinesfrom switching surgesFor adequate protection of the machine it is essential toknow the amplitude, V,, and the rise time, tl of the severestvoltage surge (FOW) that may occur on the system. It isrecommended that the actual field tests be conducted forlarge installations according to the recommendedsimulation test circuits, noted above, to ascertain thesesurges.The tests may be conducted on a few ratings withdifferent sizes and lengths of cables to simulate the nearactual condition of the installation. One should be moreconservative than liberal on cable lengths to obtain safeI I I 1' Supply source ' Busbar section Test circuitI I I 18 ,1 1 breakerCable ~IsubstituteIIL RTypical parameters as recommended by ClGRE2, = Grounding impedance = 5 QL = Load inductanceL, = Source inductance = 4 mHR = Load resistanceC, = Compensation capacitance = 0.0 or 7.0 pF C, = Load parallel capacitanceC, = Busbar capacitance = 40 nFRp = Load parallel resistanceL, = Busbar inductance = 25 pHCable = 100 m screened, radial field 3 XUnipolar Z, = 40 a, 95 mmZ AI unrolled and grounded at both ends.Figure 17.19 A simulated test circuit

Voltage surges-causes, effects and remedies 171579results. The most appropriate impulse level must be chosen,such as 3.5 p.u. for a rise time of 0.2 ,us, for the machinesto suit the system conditions, according to Table 11.6 forrotating machines, or Tables 13.2 and 14.1 for switchgearassemblies, Tables 32.1 (A) and 32.2 for bus systems andTables 13.2 and 13.3 for all other equipment. It shouldbe noted that steeply rising surges that fall within therange of machine windings and interconnecting cablesalone are of relevance and not those that are faster orslower. The steepness of the surge will determine itspropagation through the windings and its effects on thecoils or the interturns, such as whether the first coil,second coil or the last coil or the first few or the last fewturns require protection as discussed earlier. Since theactual installation may differ from that considered inanalytical studies, it is possible that the results obtainedmay differ slightly from those analysed. It is thereforead\,isable to be more conservative in selecting theamplitude and rise time of the prospective surges. Whilethe machine may be selected with a standard impulselevel, and if the probable amplitude and/or the rise timeare expected to be more stringent, separate protectionmust be provided, as discussed later. The general practicehas been to insulate all the coils equally, according to thestandard impulse level of the machine. The followingare preventive measures that can mitigate the effects ofsuch surges:1 By improving the PF of the interrupting circuit: Thisis to achieve extinction of the arc, i.e. interrupting thecircuit at the first natural current zero, as far as possible,and averting any restrikes. It may be noted that atevery current zero the arc due to contact separationextinguishes and there is no conducting path betweenthe contacts until a restrike takes place. This can beachieved in the following ways:Modern interrupting devices have an amount ofresistive interruption as a result of appropriate designand choosing the right material for making contacts,(Section 19.2) which will help to moderate the highlyinductive interrupting current. Some manufacturerseven provide a resistance shunt across the partingcontacts. which forms part of the circuit duringinterruption only, similar to the theory of a surgearrester (Section 18.1).By installing p.f. improvement capacitors tocompensate the no-load magnetizing current of thecircuit being switched (Section 23.13). These maybe installed in the same switching circuit to switchtogether with the inductive load.2 In capacitor switching, introduction of an inductorcoil (Section 23.1 1) can contain not only the inrushcurrent but also tame the current phasor to shift closerto the voltage and thus limit theTRV on an interruption.3 By selecting the rating of the interrupting device asclose to the full-load current of the system or machineas possible. An excessive rating than necessary mayhave tendency towards current chopping.4 More care needs to be taken when there are a numberof such motors connected on the same bus and switchedin tandem, tending to multiply the switching effects(Figure 17.20).0 @ @Motor (Surge impedance = Z,)MPR - Motor protection relayFigure 17.20 A number of motors connected in tandem on acommon bus5 Protection against electrodynamic forces. These arecaused by transient currents, such ils on faults. andmainly affect the overhangs or the parts of the windingsthat fall outside the stator slots. These parts are speciallybraced and strengthened at the time of manufacture.Since the standard insulation level (BIL) of a machine,equipment or a system is already defined, according toTables 11.6. 14.1, 32.1(A). 13.2 and 13.3, the machinesare accordingly designed for this basic insulation (BIL)only. When the prospective surges are expected to be moresevere than this, separate protection becomes imperative.This is particularly important for arotating machine which,besides being a dry equipment, also has only a limitedspace within the stator slots and hence has the smallestBIL of all, as is evident from Table 11.6. compared toTables 14.1, 32.1(A) and 13.2. For its comprehensiveprotection it can be considered in two parts.17.10.1 Major insulation areaThis is the winding insulation to the body, which is morevulnerable to prospective voltage peaks, V,. as a result ofTRVs. When the TRV exceeds the BIL of the machine,it can be dampened to a safe limit with the use of a surgearrester, say, from peak a, to u7, as illustrated in Figure17.21. Details of a surge arrester and the procedure forits selection are discussed below. See also Example 17.6.The selection of the arrester will also depend uponthe method of star (neutral) formation of the stator’s

17/580 Industrial Power Engineering and Applications HandbookTRV3TRV,TRV,TRV,Required condition : etl < et,Where, etl = Actual turn to turn voltage that may developacross the machine windinget, = Design voltage of the winding turn to turn.= VJn, where V, is the permissible or designimpulse voltagen = Number of turns= Length of winding per turnFigure 17.22 Turn-to-turn insulation“1 1Curve-Oa, - Original steep fronted TRV,, r.r.r.v. = - tlCurve-Oa2 - Damped TRV, with the use of asurge arrester alone, r.r.r.v. maystill be higher than recommended forv,,a particular winding, r.r.r.v. = 7 tiCurve-Ob, - Tamed TRV3 with the use of a surgecapacitor alone, amplitude remaining nearlythe same as the original wave, r.r.r.v. - VtltiCurve-Ob, - Tamed and damped TRV, with the use ofa surge capacitor and a surge arresterin parallel, r.r.r.v. = - It should be lesst;”’than the impulse withstand level of the equipmentFigure 17.21 Taming and damping of a steep-fronted TRVwith the use of a surge capacitor in association with a surgearresterstar-connected windings. If it is solidly grounded, thereflections will be less severe, as the incident surge willbe discharged through the ground and cause lessreflections. But when the star point is left isolated, itmay cause severe voltage stresses to the end turns of thecoil due to repeated reflections. The procedure to selecta surge arrester takes account of this.vi,17.10.2 Minor insulation areaThis means the insulation turn to turn (etl = VJn, n beingno. of turns, Figure 17.22), which is more vulnerable tothe steepness of a surge ( Vt/tl). If e,2 is the design insulationlevel of the winding turn to turn then, for a safe condition,et2 > etl. If the design voltage falls short of the arrivingsurge, surge protection must be provided.A surge arrester is not able to reduce the steepness(r.r.r.v) of a surge. While motors with single-turn coilsmay be safe as turn to turn insulation becomes the sameas the major insulation with respect to the body, motorswith more turns per coil may not remain safe. In suchcases the steepness of the arriving surge must be reducedto a safe level. Very fast-rising waves can be tamed (r.r.r.vcontrolled) with the use of surge capacitors or surgesuppressors. Surge capacitors possess a good energyabsorbing capacity and can reduce the steepness ordiminish the r.r.r.v of a fast-rising wave (FOW) to a safelevel, less than the turn-to-turn impulse withstand levelof the coil, i.e. less than the steepness of a lightningsurge (t, > 1.2 ps). Normal practice is to tame to 10 psor so, so that the surge is uniformly distributed over theentire windings and a normal surge arrester, when theamplitude is also higher than prescribed, can protect itsafely such as from a, to bl or uz to b2 , as illustrated inFigure 17.21. Surge capacitors in the range of 0.1-0.25pF (generally 0.25 pF) are ideal for neutral groundedmachines and twice this level for the neutral isolatedmachines. The multiple reflections at the star point mayalmost double the voltage at the star point (neutral). Theyare connected between each motor terminal and groundor on the switching device within the interrupter housing.17.10.3 Surge capacitors (see also Section 18.8)A surge capacitor offers a near-open circuit during normaloperation (at or near the power frequencies) and a nearshort-circuitto the arriving surges at surge frequencies,A, while the inductance of the motor windings (-fJ willrise rapidly and offer a near-open circuit to the arrivingsurges. It thus attracts an arriving surge and reduces itssteepness due to high ‘C’ in the circuit parameters, whichalso reduces both f, and Z,, as discussed earlier.Example 17.4Consider a surge capacitor of 0.25 pF. ThenwhereZ, = capacitor impedance in R, andC, = surge capacitance in F

Voltage surges-causes, effects and remedies 17/581at a power frequency (50 Hz)z, =12rrx 50 x 0.25 x 10-6= 12.73 k0which IS large and will offer a near-open circuit to the powerfrequency voltage and remain unaffected under normalconditions. For a system voltage and 6.6. kV, it will draw acurrent of onlyor < 0.3A& x 12.73while in the event of a surge voltage, say at 13 kHz, it willbecomez, =12rrx 13 x 103x 0.25 x lo-‘= 48.95 Qwhich is quite low compared to a very high X, of the windingsand will share the bulk of the transient current and providethe required low effect of Z,, to reduce the steepness of theTRV. Since they have high-energy ($V’ ) storing capabilitycapacitors when charged with d.c. can store high energy. Asurge has a d.c. component, hence the effectiveness of thesurge capacitors. They absorb most of the energy of thearriving surge without changing its amplitude (V,) and reducethe surge’s front steepness and hence the rate of rise (r,r,r,v,),Fit,, similar to the way in which an arrester reduces amplitude(VJ.It is possible that small motors are protected through theuse of surge protective capacitors alone and only large motorsare protected through both the surge arrester and the surgecapacitor. They also help to dampen the surge transferencesdue to electrostatic coupling from the higher voltage side ofa power transformer to the lower voltage side of it, as a resultof a reduced Z, and a changed and reduced electrostaticratio (Cd(Cp + C, + C)) (Section 18.5.2).These capacitors differ from standard p.f. improvementcapacitors, as they are designed to withstand higher testvoltages and have a low internal inductance. They shouldpreferably be non-inflammable, synthetic liquid impregnatedand provided with a built-in discharge resistance. Forspecifications refer to VDE 0675 and VDE 0560 111.Example 17.5To illustrate the above, consider a surge capacitor of 0.25 pFbeing used in parallel with a 350 kW, 6.6 kV motor. If thelikely FOW is presumed to have an amplitude up to 5 p u,then the energy this capacitor can absorb= 1 x 0.25 x [6.6 x lo3]’ Joules2= 5.4 Jouleswhereas the energy of the FOW will beW , = . *2T Joules5where, V, = 5 x 5.39 = 26.95 kV(17.3)”Factor 2 is considered to account for reflections. For moredetails refer to equation (1 8.10).-4 2[ 1 P.U. = x 6.6 = 5.39 kV2, = 4000 Q from Figure 17.7(c).J = 20 ps for an FOW, Table 18.1:. w,,,,, = x 2 x 20= 7.26 Joulesthat means the surge capacitor is capable of absorbing mostof the energy released by the FOW and taming its steepness(r.r.r.v.) within safe limits. This (energy absorbed) is so muchdesired also as the arrester, during the FOW discharge, iscapable of absorbing only a small energy as calculated belowand most of it has to be absorbed by the capacitor only.Since the amplitude of the FOW is more than permissible(16.2 kV for a normal design and 26.9 kV for a special designmotor Table 11.6) it is advisable to dampen the amplitudewith the help of a surge arrester. Using a 6 kV station classsurge arrester of Example 17.6,Vres(FOW) = 21.5 kV from Table 18.8(It is advisable to select another arrester with a smaller V,than this to make use of a normal motor.)The energy that will be absorbed by the arrestervt -W(arrester) - v, . 2J Joulesfrom equation (18.10)~2,- 26.95 - l4 x 14 x 2 x 204000= 1.81 JoulesTotal energy absorbed by the arrester and the surge capacitor= 1.81 + 5.4= 7.21 JoulesLeaving only a small part (0.05 Joules) for the machine toabsorb, which it should be compable to do in view of its ownimpulse voltage withstand capability at V, = 14 kV.Note: The arrester has to be a duration class to absorb thelong duration (250/2500 p), high energy switching surgeswhich it has to absorb first before te long duration surges(after restrikes) become steep fronted FOWs. See Section18.8 for a total surge protection of a motor, which is stillhigher than 5.4 Joules and will require a surge capacitor of ahigher capacity. It is, however, advisable to change the arresterwith a V, < 16.2 kV.Illustration1 Refer to Figure 17.21. Consider a steep-fronted transientwave a, with a front time t,, which has been dampened toa moderate and less severe transient wave a2 with a fronttime r; by the use of a surge arrester. The resultant turnto-turnvoltage stress, et,, for a given length /of the machinewindings will be limited to the maximum allowable voltagestress, et2, of the windings’ insulation, et, being a designparameter which may be obtained from the manufacturer.The dampened intensity of the switching surge may evenmake up for an HT machine, having a low insulation level(less than 3 P.u., Table 11.6) suitable for operating on asystem that could attain a transient voltage as high as 3-5 p u during normal switching, with an extremely lowfront time t,. Figures 17.5(a) and (b), for instance, illustratethe oscillograms of a switching surge, with and without asurge arrester. The peak TRV, V,,, in normal switching,which was of the order of 26 kV (4.8 P.u.) for a 6.6 kV

~~17/582 Industrial Power Engineering and Applications HandbookT(' InterrupterP9Transformer( Interrupter- - -TCS1 (i_ T;'k0 0,GC, : Surge capacitor -0.1 - 0.25 pFSA : Surge arrester@ @ @ @ : Represent amplitude and steepness ofthe arriving surge at different locations correspondingto curves of Figure 17.21.Figure 17.23 Use of a surge capacitor in association with asurge arrester to provide taming as well as dampening effectsto a steep fronted TRVsystem has been dampened to a V, of only 13.5 kV (2.5P.u.) with the application of a surge arrester.2 If the turn-to-turn voltage, et,, and the front time, fy, soachieved falls within the design parameters, no additionalmeasures would be necessary to further reduce the TRV.However, if it is felt that the dampened voltage, et,, mayexceed the required value of r.r.r.v. during operation, asurge capacitor may also be introduced in the circuit asillustrated in Figure 17.23 to reduce the wave front a? tob,, so that the front time is enhanced to a permissiblet;'.Example 17.6For the selection of an arrester:Consider the same 6.6 kV motor having the following designparameters:BIL - Lightning impulse 31 kVFOW For a front time 0.2-0.4ps, the motor may be designedfor 16.2 kV (3 P.u.). If a higher front of wave voltagewithstand capability is required, the motor may bedesigned and insulated for 26.9 kV (5 P.u.) (Table11.6, column 3).Nominal system voltage = 6.6 kV (rms.)Maximum system voltage = 7:2 kV (rms.)7.2 x &1 D.U. = - = 5.9 kV43Phase to ground voltage V, = = 4.2 kV1/3M~-For a lightningimpulse wave of5 kA (8120 ps)For a switchingimpulse wave of1 .O kA (30/60 ps)For a FOW(1/2 ps) of 10 kAProtective BIL of motorlevelProtectivemarginavailable17.7 kV 31 kV 175% ... OK15.3 kV [governed by OKthe lightningimpulse]21.5 kV 16.2 kV for Not suitablenormal design26.9 kV forspecial design125% :. OKThe distribution system may be considered as solidly groundedhaving a GFF' of 1.4 and devoid of any other TOVs.+Therefore, maximum rated voltage = 1.4 x 4.2= 5.88 kVSelect the nearest voltage rating of a station class surgearrester (see also Section 18.8) from the manufacturer'scatalogue, (Table 18.8) as 6 kV which has the followingprotective levels:With this particular arrester, the motor has to be speciallydesigned for the higher level of FOW impulse voltage withstand.But the manufacturer can always modify the protectivecharacteristics of the arrester, depending upon the system'srequirements. The matter may therefore be referred to themanufacturer for recommendations. The arrester may be fittedwith a 0.25 pF surge capacitor in parallel, to reduce thesteepness of the FOW (for the arrester of example 17.5, it is14/0.2 kVlps) to a safer value.17.10.4 Surge suppressorsA surge capacitor can only reduce the steepness of asurge, it is incapable of dampening its amplitude. Theuse of a surge arrester to dampen the amplitude of thesurge, in association with a surge capacitor, may requiremore space, besides being a more expensive arrangement.This is particularly the case when an FOW switchingsurge in motor switching has only a moderate amount ofenergy to discharge. A compact and economical alternativeis found in a surge suppressor, which makes use of alow-value dampening resistance R in series with the surgecapacitor C. The resistance can now absorb a part of theenergy of a high-amplitude surge, dampen it to a desiredlevel and make the C-R unit suitable for taming as wellas dampening even an FOW surge. The combination ofC and R is appropriately termed a surge suppressor. Figure17.24(a) shows a general arrangement of such a suppressor,while the power circuit diagram illustrated in Figure17.24(b) is almost a replica of a non-linear resistor typesurge arrester, discussed in Section 18.1.1 and shown inFigure 18.1(a). The non-linear resistor NR of Figure18.1 (a), is now replaced by C. The dampening resistor Rof the surge suppressor is able to provide the arriving*GFF = Ground fault factor (Section 20.6).tTOV = Temporary overvoltage.

Voltage surges-causes, effects and remedies 17/583R Y B1 1 117.10.6 ConclusionSwitching of an induction motor was typical of all toillustrate the pheno-menon of switching surges in aninductive circuit. The situation would remain the samewhen switching a power system, transformer or powercables also, in all conditions of loading (fully loaded,under loaded or on no load). Switching surges in acapacitor circuit is discussed in Section 23.5.1 with thelikely levels of voltage surges. Below we discuss theinsulation coordination and protection of other machinesand systems.17.11 Theory of surge protection(i nsu lat ion coordination)Figure 17.24(a) C-Rtypesurge suppressorFigure 17.24(b) Typical powercircuit of a C-R-type surge suppressorsurge, a low resistance path, and a means of absorbingthe excess energy of the surge to dampen it to a requiredlevel while the surge capacitor arrests the steepness ofthe surge and reduces it to a desired r.r.r.v.The design of the C-R combination maintains anegligible level of leakage current through the suppressorin healthy conditions (to contain resistance loss). It iseasily achieved, as C provides a near-open circuit inthese conditions and permits only a very small leakagecurrent to flow through it.A C-R suppressor also helps to reduce the surgeimpedance, Zs, of the circuit and thus limits the amplitudeof Vt as discussed in Section 17.8.NoteSince a surge capacitor or a C-R combination surge suppressoroperates only at very high frequencies, such as those related to anFOW, they maintain a near-open circuit for the arriving long-durationswitching surges. They are thus required to handle only a moderateamount energy of an FOW and hence are suitable for such duties.17.10.5 Setting fast-responding relaysNo system is permitted to trip on the occurrence of amomentary disturbance, such as by travelling surges ofany kind. To overcome this:The motor protection relay, as discussed in Section12.5, is generally provided with a delay feature tobypass these transients and delay the tripping by twoor three cycles.The same is true for an overcurrent or a ground faultrelay when used in circuits that are prone to surges.The insulation of a current-carrying system, machine orcomponent is to provide it with the required insulationlevel (BIL) to withstand system voltages during normaloperation, as well as temporary overvoltages (TOV) andmomentary voltage surges, up to a certain level, duringsystem disturbances. A safety margin is built into theirequipment by the manufacturers according to Tables 11.6,14.1, 32.1(A), or 13.2 and 13.3 as standard practice tosustain such voltages without failure or rupture of theinsulating system. Repeated application of such voltages,even if they are below the BIL of the insulating system,or longer duration of overvoltages, may lead to failureor rupture of the insulating system as a result of insulationfatigue. It is possible that in operation, TOVs or voltagesurges may exceed the safe (prescribed) power frequencyor impulse withstand voltages (BIL) respectively, of thepower system or the terminal equipment. For therecommended safe insulation levels of different equipment,refer to tables mentioned above.For instance, when lightning of, say, a nominal dischargecurrent of 10 kA strikes a 400 kV (r.m.s.) overhead line,having a surge impedance of 350 Q, then two parallelwaves will be produced each of amplitude 10 x 350/2 or1750 kV which may be more than the impulse withstandlevel of the system and cause a flashover between theconductors and the ground, besides damaging the lineinsulators and the terminal equipment (Table 13.2). It istherefore imperative that the system is protected againstsuch eventualities.Surge protection, therefore, becomes essential when itis felt that the surges generated during an operation at aparticular installation may rise beyond the permissibleimpulse withstand capacity (BIL) of the equipment. Asystem that is prone to frequent occurrences of temporaryovervoltages. Induction motors for instance, which conformto the impulse withstand levels prescribed for this machineaccording to Table 11.6 may be supplemented by a surgearrester or suppressor, when it is felt that the amplitude orsteepness (or both) of the surges in operation may exceedthe prescribed levels. Coordination of insulation of theequipment to be protected with that of the level of theprotectivedevice, whichmay be a switchgear or alightningarrester, is called insulation coordination. This woulddepend upon the type of installation, such as the type ofswitching device and the length of the interconnecting

17/584 Industrial Power Engineering and Applications Handbookcables. For an assessment of the possible magnitudes anddurations of the TOVs and the prospective amplitudes andsteepnesses of the lightning or switching surges, particularlyat critical installations such as a generating station or alarge switchyard, it isessential tocarry out system transientanalysis (TNA) as noted later.Where this is not necessary, it can be assessed bythe equipment’s exposure to such TOVs and surges andthus determine the appropriate level of BIL. IEC 6007 1 -2 provides guidelines for the most appropriate surgeprotection scheme and is discussed in Section 18.6.Relevant StandardsIEC Title IS BS60060-1/1989 High voltage test techniques. General definitions andtest requirements60071- MY93 Insulation coordination - Definition, principles andrules6007 1 -2/1996 Insulation coordination - Application guide60470/1974 High voltage a.c. contactors60947-4- 1/1990 Contactors and motor starters. Electromechanicalcontactors and motor starters61024- 1/1990 Protection of stmctures against lightning - Generalprinciples- Capacitors for surge protection for systems 0.65 kV to33 kV207 1 - l/19932165- M99 I,2165-2/199137 16/ I 99 I ,9046/199213947-4- 1/19932309/ I989BS 923-1/1990BS EN 60071-1/1996BS EN 60071-2/1997BS 77~19n4BS EN 60947-4-111992BS 6651/1992,DD ENV 61024-111995Relevant US Standards ANSVNEMA and IEEEIEEE.4/1995ANWIEEE1313.1/1996Standard techniques for HV testingInsulation Coordination, Definitions, Principles and RulesNotes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedSurge energyRate of rise of recovery voltagevtr.r.r.v. = - kV/pst,V, = peak value of the voltage surge in kVW =Vt’- x t x 10’ kW-s or kJ (17.3)zcW = energy released in kW-s or kJV, = prospective crest of the surge in kVZ, = surge impedance of the power system and thet, = rise time in ps terminal equipment int = duration for which it exists (in seconds)(17.1)f, = surge frequency in HzVelocity of propagationL = leakage inductance of the circuit ‘in henry (H)C = lumped leakage capacitance of the circuit in U = ~(1 7.4)farad (F)Surge impedanceU = velocity of propagation in kdsLo = line or conductor mutual inductance in HkmCo = leakage capacitance of the same medium inzs = J$n(17.2) F/km

~ v,Voltage surges-causes, effects and remedies 17/585Per unit voltageI p.u. = \ 2- ( 17.5)13Current zeroThis is defined by i = 0, wherei = I,,,,, sin w f ( 17.6)i = instantaneous current component at any instantI,,,;,, = peak value of the currentw = angular velocity = 2 ttfr = timeFurther readingAdjayc, R.E. and Cornick, K.J.. ‘Distribution of switching\urges in the line - end coils of cable connected motors’, IEE7runsrrcfioti.s. E/ec.fric PoMw Applicurions, 2 No. I, 1 1-2 1( 1979).Andra. W. and Sperling. P.G., ‘Winding insulation 5tressingduring the switching of electrical machines’, Sienrens Review43, No. 8, 345-350 (1976).Beeman. D. (ed.). Indu.\tria[ Power System.\ Hundhook.McGraw-Hill. New York.Central Board of Irrigation and Power, ‘Re-switching transientsand their effects on the design and operation of large inductionmotor drive\ for power station auxiliaries‘, New Delhi, India.March (1995).Cigre Working Group 13.02. ‘Proposal ofa teat circuit to simulate\witching of small inductive currents in high voltage motorcircuits’. Lkctm. No. 75. March (1981).Cormick. K.S. and Thompson, T.R., ’Steep fronted switchingvoltage transients and their distribution in motor windings’.Part I - System measurements of steep fronted switching voltagetransient

Surge Arresters:Application andSelection

Surge arresters: application and selection 18/58918.1 Surge arrestersWhen surge protection is considered necessary, surgearresters" may be installed on or near the equipmentbeing protected. This is a device that limits the high TVs(transient voltages) generated during a system disturbanceby diverting the excessive part of it to the ground andreducing the amplitude of the transient voltage waveacross the equipment to a permissible safe value lessthan the impulse withstand level of the equipment (Tables1 1.6, I4.1,32.1 (A), 13.2. and 1 3.3. The rate of rise (r.r.r.v.)remains the same, i.e.- "" (curves (I, and u2 of Figure 17.21)I, t;'thus shielding the connected equipment from such dangerousvoltage surges. This is achieved by providing aconducting path of relatively low surge impedancebetween the line and the ground. The discharge currentto the ground through the surge impedance limits thevoltage to the ground. During normal service thisimpedance is high enough to provide a near-open circuit.It remains so until a surge voltage occurs and is restoredimmediately after discharge of the excess surge voltage.CorollaryAn arrester can be considered a replica of an HRC fuse. While afuse i\ ;1 cun-ciit limiting device and protects the connected equipmentby linuting the prospectlve peak fault currents, Is, (Figure 12.18)an arrester is a voltage limiting device and protect, the connectedcquipinent by limiting the prospective peak aurge voltage, V, (curve3. Figure 17.21 1.Arresters or diverters are generally of the followingtypes and the choice between them will depend upon thepower frequency system voltage, characteristics of thevoltage surges and the grounding system, i.e.(i) Gapped or conventional. and(ii) Gapless or metal oxide.18.1.1 Gapped surge arrestersThese are generally of the following types:I Expulsion These interrupt the follow current by anexpulsion action and limit the amplitude of the surgevoltages to the required level. They have low residualor discharge voltages (Vrcs). The arrester gap is housedin a gas-ejecting chamber that expels gases duringspark-over. The arc across the gap is quenched andblown-off by the force of the gases thus produced.The enclosure is so designed that after blowing offthe arc it forcefully expels the gases into the atmos-*Basically they arc surge diverters but conventionally are calledarresters upto 245 kV lightening surges and beyond 245 kV switchingsurges arc found to he more severe, Section 18.3. It ir cwtomary.therefore, to call an arrester up to 245 kV a lightning arrester andbeyond 245 kV a surge arrester. For ease of reference. we havedescribed them as surge arrestcrs or only arresters for all types.phere. The discharge of gases affects the surroundings,particularly nearby equipment. The gas-ejectingenclosure deteriorates with every operation and.therefore, has only a limited operating life. Moreover.these types of arrester are of specific ratings and anexcessive surge than the rated may result in its f ai 'I ure.They are now obsolete in view of their frequent failuresand erratic behaviour and the availability of a moreadvanced technology in a metal oxide arrester.2 Spark gap These have a pair of conducting rodswith an adjustable gap, depending upon the sparkovervoltage of the arrester. Precise protection is notpossible, as the spark-over voltage varies with polarity,steepness and the shape of the wave. These arrestersare also now obsolete for the same reasons.3 Valve or non-linear resistor In this version, a nonlinearSic resistance is provided across the gap andthe whole system works like a preset valve for thefollow current. The resistance has an extremely lowvalue on surge voltages and a very high one duringnormal operations to cause a near-open circuit. It isnow easier to interrupt the follow currents.A non-linear resistor-type gapped surge arrester maygenerally consist of three non-linear resistors (NR) inseries with the three spark gap assemblies (see Figure18.1 (a)). The resistance decreases rapidly from a highvalue at low currents to a low value at high currents,such that R/ -- constant (Figure 18.1 (b)). Hence, V-/is an almost flat curve, as illustrated in Figure 18.1 (b).Thyrite* and Metrosil* are such materials. The purposeof non-linear resistors is to permit power frequencyfollow currents, after the clearance of surge voltages,while maintaining a reasonably low protective level(Vres). Across the spark gaps. known as current limitinggaps, are provided high-value resistors (HR) backedup with HRC fuses. The non-linear resistors have avery flat V-/ curve, i.e. they maintain a near-constantvoltage at different discharge currents. The flatnessof the curve provides a small residual voltage and alow current. When the switching surge voltage exceedsthe breakdown voltage of the spark gap, a spark-overtakes place and permits the current to flow throughthe non-linear resistor NR. Due to the non-linearcharacteristics of the resistor, the voltage across themotor terminals is limited to approximately the dischargecommencing voltage (V,,,), which is significantlybelow the 3-5 p.u. level. It may be noted thatthe use of resistor across the spark gap stabilizes thebreakdown of the spark gap by distributing the surgevoltage between the gap and the non-linear resistor.Figures 17.S(a) and (b) are oscillograms illustratingthe effect of a surge arrester in arresting the surgevoltages caused during a switching operation or alightning strike.The current limiting gaps, as noted above, in serieswith the non-linear resistors make it possible to adjustthe protection level of the surge arrester for differentvalues of dircharge currents. They also help to maintain*Thyrite is a brand name from General Electric. USA.Metrosil is a brand name from a GEC company in the UK

18/590 Industrial Power Engineering and Applications Handbookswfl -3CR Y Bresistance used in the manufacture of an efficient surgeprotection device must offer the least impedance duringa discharge. This is to provide a free flow to the excessivedischarge current to the ground, on the one hand, and todraw a negligible current under normal system conditions,to make it a low-loss device, on the other. The alternativewas found in ZnO, ZnO is a semi-conductor device andis a ceramic resistor material constituting ZnO and oxidesof other metals, such as bismuth, cobalt, antimony andmanganese. Since the content of ZnO is substantial (around90%) it is popularly known as a ZnO or metal oxidesurge arrester. It has no conventional spark gap and hasexcellent energy absorption capability. It consists of astack of small ZnO disks (Figure 18.2(a)) in varyingsizes and cross-sections, enough to carry the dischargeOCRFigure 18.l(a) A typical power circuit of a non-linear resistortypesurge arresterV, profile becomes flat (constant) and lessthan impulse voltage withstand capability (BIL)of the equipment under protectionFigure 18.2(a) ZnO blocks and their small sizes-V,AtUt5-Figure 18.1 (b) Characteristic of a non-linear resistora near-constant voltage at around the switching surgeor lightning surge spark-over voltages during the flowof surge currents while clearing a surge.18.1.2 Gapless surge arrestersFrom the above it is evident that material for non-linearFigure 18.2(b) Distribution class surge arrester

Surge arresters: application and selection 18/591currents, mounted in series in a sealed porcelain housing(Figure 18.2(b)). The surface area (size) of disk can beraised to make it capable of absorbing higher energylevels. The design is optimized to minimize the powerloss. Figure 18.3(a)-(c) show the general arrange-mentsof a few types and sizes of gapless surge arresters.Under rated system conditions, its feature of high nonlinearityraises its impedance substantially and diminishesthe discharge current to a trickle. Under rated conditions,it conducts less than 1 mA (Figure 18.4(a)), while duringtransient conditions it offers a very low impedance to theimpending surges and thus raises the discharge currentand the discharge voltage. However, it conducts only thatdischarge current which is essential to limit the amplitudeof the prospective surge to the required protective levelof the arrester. The housing is sealed at both ends and isprovided with a pressure relief valve to vent high-pressuregases, such as those caused by heavy currents during avoltage surge or a fault within the arrester, and to preventan explosion in the event of a housing failure.18.2 Electrical characteristics ofa ZnO surge arresterIn view of the limitations in spark gap technology, asdiscussed earlier, the latest practice is to use gaplesssurge arresters. Accordingly, the following text relatesto gapless arresters only. For details on gapped surgearresters refer to ANSMEEE-C-62.1, ANSImEE-C-62.2and IEC 60099-5, as noted in the Relevant Standards.ZnO blocks have extremely non-linear, current-voltagecharacteristics, typically represented byI=K.V" (18.1)where the conductance (1/R) in the conventional formula(I = + . v)is replaced by K, which now represents its geometricalconfiguration, cross-sectional area and length, and is ameasure of its current-carrying capacity. a is a measureof non-linearity between V and I, and depends upon thecomposition of the oxides used. Typical values areIn Sic - 2 to 6In ZnO - it can be varied from 20 to 50.By altering a and K, the arrester can be designed forany conducting voltage (Vres) and nominal currentdischarge (I"). Vresand I, define the basic parameters of asurge arrester, as discussed later. Figures 18.4(a) and (b)CastingsupportplateSpringZnOdiscsContactplatesPorcelainhousing'.supportplateDoublesealingsysteml - 4Figure 18.3(a) Sectional view of a metal oxide Figure 18.3(b) A 400 kVsurge arrester (Courtesy: W.S. Industries) surge arrester (Courtesy: Elspro(International))Figure 18.3(c) 12-550 kV zinc oxide surgearresters (Courtesy: Crompton)

18/592 Industrial Power Engineering and Applications HandbookPredominately resistiveHigh current non-linear region(current predominately capacitive)(Rated voltage) V,(MCO") vc/D =I" = Conducting current at2, =Peak discharge currentor protective currentthrough the arrestervra,Impedance of theprotected circuitImpedance of thearrester at VresCharacteristic of ZnO(m - 20 to 50)Characteristic of Sic(= - 2 to 6)Characteristic of a linearresistance (= = 1)Very low leakage I, I"V,,,= -vi vresI, = -current in pA in pA ZP ZS(0.4-1 0 mA)(2.5-40 kA)Current through the arrester .--)(peak value of8/20 ps wave)Figure 18.4(a) Characteristics of a ZnO block400200r:E 6043' 4018.2.1 Electrical representation of aZnO elementA ZnO element basically represents a capacitive leakagecircuit. In its leakage current range it may be electricallyrepresented as shown in Figure 18.5, where Zzno is theleakage current, capacitive in nature, and I, and I, itscapacitive and loss components, respectively. The successof an element will depend upon its low loss-component,which would mean a lower loss during continuous20io10-7 io-6 10-~ io4 io3 IO-^ io-' ioo 10' io2 io3Amps./cn? -@ = 25°C0 = 5 0"~@ = 100°C@ = 125°CFigure 18.4(b)V-l characteristics of a ZnO blockillustrate typical electrical characteristics of a ZnO arrester,suggesting that in the event of a surge voltage, witha prospective amplitude V,, its resistance will fall rapidly,and it will absorb much of the current and energy up to(V, - V,,,) discharged by the voltage surge. The rest, i.e.the conducting voltage ( Vre,) depending upon the arrester'sprotection level, will appear across the arrester and theequipment it is protecting. It has an excellent energyabsorption capability. Some of its basic characteristicsaccording to Figures 18.4(a) and (b), are noted below.Note I, would consist of resistive as wellas 3rd harmonic component.Figure 18.5 A ZnO element

Surge arresters: application and selection 18/593operation, on the one hand, and a lower temperature riseof the element, on the other.18.2.2 Maximum continuous operating voltage(MCOV), V, (point 1 on the curve ofFigure 18.4(a))This IS the power frequency operating voltage that canbe applied continuously (2 2 hours) across the arresterterminals without a discharge. It continuously draws anextremely low leakage current, IZnO, capacitive in nature,due to ground capacitance. The current 15 in the range ofa few ,LA. Therefore maximum continuous operatingvoltageVI,,V, = - (phase to neutral)L 318.2.3 Rated voltage, V, (point 2 on the curve ofFigure 18.4(a))This is the voltage for which the arrester is designed.The arrester can withstand this voltage without a dischargefor a minimum 10 s under continuously rated conditions(when the arrester has reached its thermal stability),indirectly indicating an in-built TOV capability of 10 s.Now it also draws a current, capacitive in nature, in therange of a few mA. The lower this current, the lowerwill be the loss and the heat generated during anovervoltage and hence better energy absorption capabilitj.Below the knee-point, in the MCOV and TOV regionsparticularly, the V-I characteristic (Figure 18.4(b)) issharply drooping with the temperature rise, causing ahigher leakage current. It is a deterrent to otherwise goodperformance for it means higher losses and heat undernormal as well as TOV conditions. Although this verylow level has been achieved, research and improve-mentin the formation of a ZnO compound is a continuousprocess by leading manufacturers. It aims to optimizethe use of this material for a still better performance byattempting to flatten the droop as far as possible.18.2.4 Reference voltage (point 3 on the curve ofFigure 18.4(a))This is a voltage close to the knee-point of the characteristic.It is a point where it commences its conductionand draws a current, resistive in nature, in the range ofa few mA. Typical values are 0.4-10 mA.The design and configuration of ZnO disks is suchthat this resistive current is more than the capacitiveleakage current it draws in the MCOV region.18.2.5 Temporary overvoltage (TOV)(Figure 18.4(a))The operating point 1 temporarily shifts to near point 3.Now also the current is also resistive and in the range ofa few mA.18.2.6 Transient voltages (Figure 18.4(a))The operating point shifts to near point 4 and beyond onthe curve. It may conduct a current of 2.5-20 kA ormore, during a very fast-rising voltage surge.18.2.7 Protective level (Figures 18.4(a) and (h)For TOVs this is determined by its low current region(d) of less than 1A and for prospective transient voltagesby its high current region (e) of 2.5-20 kA (8/20 ,uscurrent surge). In this region the arrester must have ahigh capability for energy absorption.The ZnO protective characteristic curves of Figure18.4(b) also suggest that these protective characteristicsremain unaffected by variations in operating temperaturesexcept in very low current ranges is that and immaterial.18.3 Basic insulation level (BIL)BIL is the basic insulation level of equipment. When thesystem TOV or voltage surges exceed this level. theequipment may yield. In the latest international andnational standards it is defined as follows:1 For systems 1 kV< V,, 5 245 kV.(i) Rated lightning impulse withstand level (LIWL)(ii) Rated short time power frequency dielectricstrength.(iii) Prospective steep-rising TRVs (FOWs) that maybe caused during a switching operation, asdiscussed in Section 17.7*.For motors, switchgears and bus systems see Tables11.6, 13.2, 14.1 and 32.1 (A) and for other equipmentTable 13.2. For more clarity refer to Section 17.1.2 For systems having V, > 300 kV to 765 kV:(i) Rated lightning impulse withstand level (LIWL)(ii) Rated switchingimpulsewithstand level (SIWL)(iii) Prospective steep-rising TRVs (FOWs) that maybe caused during a switching operation as notedabove or during a fast bus reclosing (Section17.4) particularly with the line trapped charge.Refer to Table 13.3".*There is no rated withstand levels specified in these standards forsuch surges. This will depend upon the system parameters as notedlater and must be specified by the user to the equipment manufacturer.Equipment may he designed for more than one BIL values asnoted in the various tables referred to above for motors, switchgearsand other equipment. The choice of BIL for equipment for a particularapplication will depend upon the extent of exposure the equipmentmay be subject to in normal service and the security level requiredby the system and the surge protection. For more details refer toSection 13.4.1(3).It is advisable to select the lower value of the BIL whereverpossible, to save on the cost of equipment, particularly when surgeprotection is being provided. Equipment, however, exposed moreto such onslaughts may he selected with a higher BIL. Examplesare those mounted some distance from the surge arrester and havea higher protective distance leading to more reflections andtransference of surges.

18/594 Industrial Power Engineering and Applications HandbookThe types of surges referred to above are defined inTable 18. I .18.4 Protective marginsOn the BIL discussed above a suitable protective marginis considered to provide sufficient safety to the protectedequipment against unforeseen contingencies. IEC 6007 1 -2 has recommended certain values to account for theseand they are given in Table 18.2.BIL of the equipmentProtective margin = (18.2)Impulse protection levelof the arrester ( VreS)The protection level of an arrester, V,,,, is a functionof the magnitude of arrester discharge current (I,), andthe time to peak of the surge (tl), and is influenced bythe following.18.4.1 Steepness (tl) of the FOWThe protection level of the arrester diminishes with thesteepness of the wave. As tl falls, V,,, of the arresterrises, leaving a smaller protection margin across theprotected equipment. Refer to the characteristics of anarrester as shown in Figure 18.7, for a 10 kA, 8/20 psimpulse wave. For a front time of, say, 0.5 ps, it willhave a V,,, of approximately 118% of its rated V,,, at 8ps, and hence will reduce the protection margin as inequation (18.2).Table 18.1 Defining a surgePredominant Maximum system Power system Voltage shapesurgevoltagevmEquivalent currentshape at whichthe arrester istested'BIL of theequipmentLightning > I kV- Secondary 1.2/50 ps 8/20 ps See Section 18.3245 kV transmission orprimarydistributionLightning Mainly 1.2/50 p5 8/20 ps> 245kV' transmission See Section 18.3Switching 250/2500 ps 30/60 ps(total time t2 = 2750 ps,Figure 17.2(b)~- .- ._~Rise time, t, (Figure 17.3)may be less thanFOW > 245 kV Mainly 0.1 ps but total time up to 1/20 ps Note 3(switching) transmission 3000 ps and surgefrequency 30 kHz to < 100MHzNotesA lightning strike may commence at around IO2 to 10' Volts (1000 kV) between the clouds and the ground. By the time it reaches theground, it loses a part of its intensity. Although it may still be around 1000 kV at ground level, it is possible that sometimes switchingsurges at an EHV system above 245kV are more severe than a lightning surge, the more so because the amplitude of a switching surgerises with the rise in system voltage, while a lightning strike remains nearly constant irrespective of the system voltage. For thesevoltages, the national and international standards have prescribed separate impulse withstand levels as noted in Table 13.3 for switchingas well as lightning surges. They have also classified these severities in categories I , 2 and sometimes 3, depending upon the extent ofsystem exposure to lightning as noted in Section 13.4.1(3).In a surge arrester, it is easier to assess the severity of a voltage wave through an equivalent current wave, but it is found that thecharacteristic of an equivalent current wave is not exactly identical to the required voltage wave. It is noticed that the time of rise ofa voltage wave is generally shorter than its equivalent current wave, and hence more severe than the current wave. The reason is thenon-uniform distribution of the current through the cross-section of the conductor, because of skin effect and discontinuities asdiscussed earlier. Refer to Figure 18.6 explaining this. To ovcrcomc this deficiency, the actual time of rise of the test current surges,while simulating the characteristics in a laboratory, is slightly shortened (for an 8/20 ps wave, the test wave rise time will be slightlyless than 8 ps), to ensure the same severity of the test current wave as the actual voltage wave. A surge arrester is required to clearsuccessfully all the three types of voltage surges as prescribed. It is imperative to ensure that the selected arrester is capable of clearingall such voltage surges with the same ease and safety. Accordingly, protective curves are established by the arrester manufacturers overa wide range of likely surges, in terms of lightning, switching and FOWs. They provide those to the user for ease of arrester selection(Figure 18.7).For steep-rising waves (FOWs), no steepness or impulse withstand level is prescribed in these standards, as both the rise time andamplitude of such waves cannot be predefined. They will depend upon various system parameters, such as grounding method, cable orline length, other equipment installed on the system, their surge impedances, switching conditions (current chopping and restrike ofinterrupting contacts etc.) and the trapped charge, such as on a fast bus transfer etc. The choice of impulse level for a particular fastrisingwave for equipment to'be exposed to such transients is a matter of system study (such as TNA or EMTP, Section. 18.5). The usermust define these for the equipment manufacturer.~.

~~ ~~~~~~~ ~Surge arresters: application and selection 18/595Table 18.2 Recommended protection margins-Voltage impulscVoltagerangev,,,KVRecommended mat qui\For swtrthrng For lightning Forsurges \urge\ FOWyITable 13.2I1Table 13.32 3.6-245Decided by thelightning wrge. I 05 I ISwhich 15 moreEevereRise time for Is - tl(/s) 4Rise lime for V,- tl(~,)r, (1s) ’ ti ( VI)Rise bme (ps) --tFigure 18.6 Arrester voltage and current oscillograms for10 kA, 8/20 ps current impulse test18.4.2 Discharge or coordinating current (I,)V,,, rises with an increase in the discharge current throughthe arrester and vice versa (see Figure 18.7 having itsrated V,,, on the 10 kA characteristics at 8 ps). For a 15kA discharge current, for instance, V,, will rise furtherto approximately 1.18 for an FOW of 0.5 ,us, and reduceits protection margin further.The arrester manufacturer will provide the protectioncharacteristics for different discharge currents I, and fronttimes, t,, for each type of arrester to facilitate the usermake an easy selection of the arrester.18.4.3 Margin for contingenciesAn additional protection margin may be considered forthe contingencies noted below, depending upon thecriticality of a system or its susceptibility to overvoltages.1 Higher overvoltages than considered. during an actualfault, say, because of unfavourable grounding conditions.2 Non-simultaneous opening (Section 19.7) or closingof the interrupting poles (Section 17.7.2).3 More than two overvoltages occurring at the sameinstant such as a load rejection associated with groundand phase faults.It is, however, recommended to select the smallest arresteras this will provide the greatest margin of protection forthe insulation. A higher rating (kJ) of the arrester mayprolong its life but may reduce the margin of protection.It is therefore better to strike a balance between the lifeof the arrester and the protection of the equipment.5040t30---- - 1/(2-20) ps impulse-.-.- 30160~s impulse0Current (kA) A ps) Courtesy A B AFigure 18.78/20 ps)Protective characteristics for arresters type EXLIM Q (maximum residual voltage in per cent of residual voltage at 10 kA,

18/596 Industrial Power Engineering and Applications Handbook18.5 Protective level of a surgearresterThis is the maximum voltage at an arrester’s terminalsthat it can sustain while it discharges to the ground excessvoltages in excess of it without damaging the terminalequipment or disrupting the continuity of the supplysystem. In other words, it is the breakdown (for gapped)or discharge value (for gapless) surge arresters at whichthey would initiate operation and is the basic parameterthat forms the basis of selection for a particular installation.The purpose of a surge arrester is to safeguard a systemagainst probable transient conditions, particularly thosethat may exceed the safe impulse withstand level of theequipment. A brief criterion to determine the protectivelevel of an arrester is given in Table 18.3. The sparkovervoltage refers to conventional type gapped arresters,while the residual voltage refers to gapless type surgearresters. An arrester must protect the terminal equipmentagainst each kind of transient condition separately. Itsprotective level must therefore be checked separatelyfor all such transient conditions. While for a lightningand switching surge, it would be enough to define it byits amplitude, the FOW will be defined by its amplitudeand the front time, t,.The severity of the transient conditions can be establishedon the basis of past experience or data collectedfrom similar installations. However, for large and morecritical installations, such as a generating station or alarge switchyard, it is advisable to carry out transientnetwork analysis (TNA) or electromagnetic transientprogramme analysis (EMTP) with the aid of computers.For more details refer Gibbs et al. (1989) in Chapter 17.Where this is not necessary, the system may be analysedas follows to arrive at a more appropriate choice ofprotection level.1 Level of exposureWhen equipment is exposed to direct lightningstrikes Equipment connected directly to anoverhead line, or even through a transformer, willfall into this category. Select the highest value ofBIL and even then a surge protection will becomenecessary for critical installations.When equipment is shielded This is when it isinstalled indoors, like a generator or motor. Now itmay be subject to only attenuated surges. One maynow select a lower value of BIL. In most casessurge protection may not be essential for directlightning strikes.When equipment is exposed to severe internaldisturbances This is when equipment is exposedto switching surges, particularly when the surgesare steep-fronted, as in switching of HT motorsand all range-I1 equipment that are exposedto switching surges (Section 17.7). Now both ahigher level of BIL and surge protection may benecessary.2 Influence of surge reflections3 Influence of surge transferences4 Effect of resonanceThese are only basic guidelines. It is difficult to defineexposed or shielded equipment accurately. Equipmentinstalled indoors may never be subject to lightning strikesor their transferences, but may be exposed to severeswitching surges and require surge protection as for anexposed installation. There is no readymade formula byTable 18.3 Establishing the protection level of a surge arresterTransient condition as inTable 18.1Protection level of a gapped surge arresterProtection level of a gapless surge arrester*(I) Lightning surge forsystems I245 kV(2) Switching surge forsystems 2 300 kV(3) Steep-fronted waves(FOWs) tl 5 1 ps -originating from apremature interruptionor multiple restrikesduring a switchingoperation* Also refer to Example 18.5 and Table 18.11The highest lightning impulse spark-over orbreakdown voltage of the arrester should beless than the lightning impulse withstand levelof the equipment being protected less by theprotection margin (Table 18.2)The highest switching impulse spark-overvoltage of the arrester should be less than theswitching surge impulse withstand level ofthe equipment being protected less by theprotective margin (Table 18.2)The highest front of wave spark-over voltageof thc arrcstcr should be less than the FOWimpulse withstand level of the equipment lessby the protective margin (Table 18.2)The highest impulse residual or discharge voltage across thearrester at the nominal discharge current (item 10, Table 18.9)should be less than the lightning impulse withstand level of theequipment being protected less by the protective margin (Table18.2)The highest switching impulse residual or discharge voltageacross the arrester at a specified switching impulse current (item11, Table 18.9) should be less than the switching impulsewithstand level of the equipment being protected less by theprotective margin (Table 18.2)The highest FOW residual or discharge voltage across the arresterat a steep fronted impulse (1/20 ps) (item 12, Table 18.9) shouldbe less than the FOW impulse withstand level of the equipmentless by the protective margin (Table 18.2)Notes1 The protective levels of the surge arresters, at different system voltages are Curnished by the manufacturers in their product catalogues,Tables 18.9 and 18.11 furnish typical data for a few established manufacturers.2 In our subsequent text, we have limited our discussions only to the more prevalent gapless surge arresters.

which such levels can be quickly established, exceptexperience. The project engineer is the best judge of themost appropriate level of BIL, depending upon the surgeprotection scheme. Below we briefly discuss the effectof surge reflections and transferences on the safety ofequipment to arrive at the right choice of BIL and thesurge protection criteria.E-.Surge arresters: application and selection 181597I Junction (E + E’)18.5.1 Reflection of the travelling wavesThe behaviour of’ a transient wave at a junction of twoconductors, such as at junction J in Figure 18.8, is similarto that of water, when it passes through one large-diameterpipe to another of a smaller diameter. Some of the waterwill flow ahead and the remainder will backflow at thejunction. Similarly, a transient wave will also reflect inpart or in full at a junction between two conductors ofdifferent surge impedances, depending upon the surgeimpedance of the circuit ahead of the junction. This wouldgive rise to two types of waves, i.e.Refracted wave: a wave that is transmitted beyond thejunction.0 Reflected wave: a wave that is repelled at the joint.See Figure 18.9To analyse this phenomenon refer to Figure 18.8,If Z,, = surge impedance of the incoming circuitZ, = surge impedance of the outgoing circuit (Figure18.8(a))E = voltage of the incident wave (incoming wave)E’ = voltage of the reflected wave.E” = voltage of the refracted (transmitted) wave.then the voltage of the reflected waveE’ = E , zs2 - zsiZS? + zsi(I 8.3)-Surge impedance Surge impedance -VSl)(ZSiE = Incident waveE’ = Reflected waveE” = Refracted waveFigure 18.9 Illustration of the reflection of a TRV at a junctionand the voltage of the refracted waveE” = E + E‘=E+E-( z, - z,,zsz + zsi)(I 8.4)If the outgoing circuit is inductive (Figure 18.8(b)) as ina motor, transformer or an inductor coil, with an inductanceL then(1 8.5)and if it is capacitive (Figure 18.8(c)) with a capacitanceC then( 18.6)JunctionY-7(a)Figure 18.8E 3Different parameters of switching circuitsThis can also be derived for a combined R, L and Ccircuit to obtain more accurate data. Generally. the figureobtained through equation (18.4) is simpler, quicker andprovides almost correct information for the purpose ofsurge analysis, and is used more in practice. Where,however, more accurate data are necessary, such as foracademic interest, then the morc rclcvant formulac maybe used.Surge impedance thus plays a significant role indetermining the magnitude of the reflected wave thatmatters so much in adding to thc TRVc. (Also refer tographs of Figure 17.7 corroborating this analysis.)When the circuit is open at the junction thenzs2 =E‘ = E, Le. the travelling wave will reflect in full.The voltage at the junction=E+E‘=2E

18/598 Industrial Power Engineering and Applications HandbookThe incoming circuit is therefore subject to twice thesystem voltage and the voltage of the refracted waveE” = E + E’=2EThis means that the travelling wave will transmit infull, and the system will encounter a voltage of twicethe system voltage. Refer to Figure 18.10(a).When the circuit is shorted at the junction thenandzs2 = 0E’ = - Eand voltage at the junction = 0.This means that the travelling wave will reflect in fullbut with negative polarity, thus nullifying the systemvoltage. The voltage of the refracted wave will also bezero, and obviously so, as there will be no refractionat the shorted end. Refer to Figure 18.10(b).When the travelling wave at the junction enters a circuitwith equal surge impedance, such as in the cable beforeor after an interrupter, then Z,, = Zs, and E‘ = 0. Thismeans that there will be no reflection and the incidencewave will transmit in full, Le. E’ =E. (Refer to Figure18.10(c).) Hence such a junction will cause no damageto the terminal equipment or the interconnecting cables.Thus, the voltage wave at a junction will transmit and/or reflect in part or in full, depending upon the surgeimpedances as encountered by the incident and therefracted voltage waves. Each junction exposed to atravelling wave may thus be subject to severe voltagesurges up to twice the incidence voltage, dependingupon the surge impedances of the circuits before andafter the junction. When the circuit parameters causesuch high voltages, care must be taken in selecting theequipment, particularly for their connecting leads andend turns as the subsequent turns will be less stresseddue to an attenuated refracted wave.Example 18.1Consider a 33 kV overhead distribution network connectedto a terminal equipment through a cable (Figure 18.11). If thesurge impedance of the line is considered to be Z,, = 450 Rand the surge is travelling into the terminal equipment througha cable having a surge impedance of Z, = 60 R then,The voltage of the refracted wave, at junction ‘a’,E” = 2E. 60~450 + 60= 0.235 Ewhich is much less than even the incidence wave andhence, safe to be transmitted.The voltage of the reflected waveE‘= E. ~60 - 450450 + 60=-_ 390 E510= - 0.765 EI/C circuit is /--\subject to 2 E y ‘1E = Incident waveE‘ = Reflected waveE” = Refracted waveE‘ = E- Surge impedance(ZSl)I/C circuit isvoltage- Surge impedanceNo change insystem voltage(a) Junction open circuited.E“= E+ E=2ESurge impedance -(ZSZ = 4Junction(b) Junction short circuitedE”=E-E=OSurge impedance -(ZSZ = 0)JunctionE‘=O- Surge impedance Surge impedance -(ZSZ = Zsd(c) When Z, = ZslFigure 18.10 Magnitudes of refracted and reflected waves underdifferent junction conditions.Thus most of the incidence wave will reflect back withnegative polarity and reduce the effect of the incidencewave. But the situation reverses as the surge travels aheadto a transformer through junction ‘b’, as illustrated, andencounters a higher surge impedance. The cable has avery low Z, compared to a transformer. Now the refractedand reflected waves are both of high magnitude. Thereflected wave also has a positive polarity and enlargesthe incidence wave. The cable and the terminal equipmentare now both subject to dangerous surges as illustratedbelow:If the surge impedance of the transformer is considered as4000 0, then the voltage of the refracted wave

SwitchingdeviceCable junction, low/ refraction and reflection/ Arrester essential if V, >/’ BIL of the cableI-// Ikz33kV overhead line.Interconnecting cable( z sr- -loon)Ij Electrostatic capacitances,’help to tame and damp thearriving surger-‘ (VI = VreJ&GInterconnecting cable/ (ZS2= loon)Surge arresters: application and selection 18/599which will also raise the incidence wave to roughly 2E.Then, there will be multiple reflections between the junctionsuntil the reflected surges will attenuate naturally. It istherefore essential to protect the cable against surges atboth the ends as shown, particularly when the travellingwave is likely to be of a higher value than the BIL of thecable. It is, however, noticed that there is a naturaldampening of the travelling waves as they travel aheadthrough the power system due to the system’s lumpedcapacitances and inductances. Even the multiple reflectionstend to achieve a peak of just twice the incidence surge. Itis, however, advisable to take cognisance of all suchreflections and refractions while carrying out the engineeringfor a surge protection scheme and deciding the locationfor the surge arresters.Surges originating at some distance from the equipmentare of less consequence, for they become dampened asthey propagate due to circuit parameters L and C. For thepurpose of surge protection, each segment must beconsidered separately as the surges may generate at anysegment and hence separate protection is essential foreach segment.V,. (V,,, + 2.S.T -natural damping)< BIL of cable1‘*ynsformer(33111 kV)z,,= 2000 (1 z g m$8 E*pNote Cable junction ‘b’ has a high refractionand reflection. Arrester would be essential toprotect the cable rather than the transformer, if2V,. > BIL of the cable. If the cable is longenough say, > 50 metre or so, the natural,dampening of the incident wave up to junction b,may be enough and may not cause any harmfuleffect even without the arrester* All cables are sheathed,\*\)RelayFigure 18.11 Surge protection of cables, transformer and motorE”= 2E- 400060 + 40002Eand of the reflected waveE, = E , 4000 - 6060 + 400018.5.2 Surge transference through a transformer(from the higher voltage side to the lowervoltage side)This is another phenomenon which can be observed ona transformer’s secondary circuit. Voltage surges occumngon the primary side of the transformer, during a switchingoperation or because of a lightning strike, have a part ofthem transferred to the secondary (lower voltage) side.This is termed ‘surge transference’.A transformer has both dielectric capacitances andelectromagnetic inductances. Surge transference thusdepends on the electrostatic and electromagnetic transientbehaviour of these parameters as noted below.Electrostatic transferenceAt power frequency, the effect of electrostatic capacitancesis almost negligible as they offer a very high impedance(X, 0~ 1o;f being too low) to the system voltage. Thetransformer windings behave like a simple inductivecircuit, allowing a normal transformation of voltage tothe secondary. A system disturbance, such as a groundfault, lightning strike or switching sequence, however,will generate surges at very high frequencies,f,. Whensuch high-frequency surges impinge the windings, thelumped (electrostatic) capacitances offer a near-shortcircuitto them while the electromagnetic circuit offers anear-open circuit (X,-f,). The transformer now behaveslike a capacitive voltage divider and causes voltage surgesdue to capacitive coupling, in the lower voltage windings,tertiary (if provided), cables and the terminal connectedon the secondary side. The capacitive coupling may beconsidered as comprising the following.Capacitance between the turns of the windings0 Capacitance between higher and lower voltage mainwindingsCapacitance between windings and core.= E See Figure 18.12. The transformer as a voltage divider is

18/600 Industrial Power Engineering and Applications Handbookillustrated in Figure 18.13, and transfers a substantialamount of the first peak of the incidence surge on theprimary side to the secondary side. The surge voltagetransfer can be expressed bycIC - c, + c,v -A.VI .P(8.7a)whereV,, = voltage of surge transferenceCp = lumped capacitance between the primary and secondarywindingsC, = lumped capacitance of the lower voltage side.These values are provided by the transformer manufacturer.V, = Prospective voltage surge that may appear on theprimary side. If an arrester is provided on the primaryside, this voltage is limited to the residual voltageof the arrester (Vre5). In both cases, consider thehigher voltage such as during an FOW. In fact, thelumped capacitances will provide the arriving surgewith a short-circuit path to the ground and help todampen transference to the secondary to some extent.But these effects are not being considered to bemore conservative.p = a factor to account for the power frequency voltagealready existing when the surge occurs. IEC 60071 -2 has suggested a few typical figures as noted below:(a) For a lightning surge and FOW :For Y/A or A/Y transformers, p = 1.15For Y/Y or &A transformers, p = 1.07L.V.H.V.Figure 18.12 Distribution of winding inductances and leakagecapacitances in a transformer shown for one windingf, = Ips- 1 1V, - Surge on the primary sidefit, - Surge transference on the secondary sideI,C, - Lumped capacitance between the primary and the secondarywindingsC, - Lumped capacitance of the lower voltage sideC’ - Protective capacitanceC - Capacitance of cable and equipment connected on the lowervoltage sideFigure 18.1 3 A transformer as a capacitor voltage divider, drawnfor one phase(b) For a switching surge, p = 1.0 in both the abovecases.A lightning surge and an FOW have more influencecompared to a switching surge due to the former’s highersurge frequencies, f,.Margins can be added to account for the severity ofthe surges, depending upon the type of installation andits criticality.For high transformation ratios when V,lV2 is high, Cp>> C, and the incidence surges tend to transfer the wholeof their severity to the secondary side. Cp/(Cp + C,) isthe ratio of transference when the secondary is opencircuited. Transference is highest when it is open circuited.This ratio will generally lie between 0 and 0.4 (IEC60071-2), but the exact figure must be obtained from themanu-facturer, when designing the protection scheme.In service, there are a number of load points connectedto it, influencing the electrostatic value in the denominator.If ‘C‘ is the capacitance of the cables and the equipmentconnected on the lower voltage side of the transformer,the transferred surge will be reduced tov -CPIC - c, + c, + cvt .P(18.7b)The front of the transferred surge will, however, beless steep and dampened than on the primary side due tocapacitive dampening. But sometimes this may also exceedthe BIL, particularly of the tertiary (if provided) andalso the secondary windings of the transformer, as wellas the cable and the terminal equipment connected onthe lower voltage side. This is especially the case whenthe primary side voltage is very high compared to thesecondary. Protection of the secondary windings, in allprobability, will be sufficient for all the cables and terminalequipment connected on the secondary side.

Surge arresters: application and selection 18601Moreover, as the surge travels through the primary tothe secondary of the transformer. a part will becomedampened due to partial discharge of the surge to theground through the capacitive coupling and also partlythrough the inductive coupling of the transformer. Asthe surge travels forward it will encounter the system’s(interconnecting cables and the terminal equipment)capacitive and inductive couplings, and will continue toattenuate in steepness as a result of electrostatic discharges,and in amplitude due to inductance of the circuit. In fact,additional surge capacitors (C’) can be provided acrossthe secondary windings as illustrated in Figure 18.13, tofurther dampen the arriving transferred surges. In fact,this practice is sometimes adopted.For adequate insulation coordination it is mandatoryto first check such transferences with the BIL of thetransformer’s tertiary and secondary windings. The tertiaryis a crucial winding and any damage to this will mean amajor breakdown of the transformer. For the purpose ofprotection and to be more conservative, these calculationsmay be carried out with the LV side open-circuited.Similarly, on the primary side, the most severe surgesuch as an FOW may be considered. If the transferredsurge exceeds the BIL of the tertiary and secondarywindings, one or more of the following protective measuresmay be considered:When the primary is provided with an arrester, selectone with a lower V,,,,to shield the secondary side also.Consider tertiary and secondary windings with a higherBIL. if possible.But the tertiary must be specifically protected by theuse of an additional surge arrester between each of itsphases and the ground. It is possible that this arrestermay discharge rather too quickly compared to the mainarrester on the primary in view of larger transferences,compared to a very low voltage rating of the tertiary.If this occurs, the arrester at the tertiary may fail. Therating of the tertiary arrester, therefore, must bemeticulously coordinated with the V,,, of the primaryarrester. The V,,, of the tertiary arrester may have tobe chosen high and $0 the tertiary must be designedfor a higher BIL.Use surge capacitances across the secondary windings.Generally, an arrester on the primary should be adequateto protect the secondary windings. When it is not, aseparate arrester may be provided between each phaseand ground of the secondary windings.The terminal equipment connected on the secondaryside of the transformer is thus automatically protectedas it is subject to much less and attenuated severity ofthe transferred surges than the secondary windings ofthe transformer. Nevertheless, the BIL of theinterconnec-ting cables and the terminal equipmentmust be properly coordinated with the BIL of thetransformer secondary, particularly for largerinstallations. say, 50 MVA and above, to be absolutelyufe. Example 18.2 will explain the procedure.Electromagnetic transferenceThis is for systems having secondary voltages up to 245 kVand that are subject to the power frequency withstandtest.During a high-frequency (FOW) surge. the inductiveimpedance of the windings becomes very high and offersan open circuit to the arriving surge, and there is noinductive transference of voltage surges to the secondary.But at lower frequencies, such as during overvoltages,long-duration switching surges (250/2500 ps), and evenduring lightning surges, the windings acquire enoughinductive continuity to transfer a part of these voltagesto the secondary, depending upon the J; of the arrivingsurge, in the ratio of their transformation (V2/VI). It isgenerally noticed that such transferences hardly exceedthe power frequency withstand level of the windings andare thus less critical. Nevertheless they must be countercheckedwhile designing the surge protection schemefor the whole system. If it is higher, thenThe arrester on the primary side may be selected witha lower residual voltage (VreJ. orThe tertiary and secondary windings may be selectedfor a yet higher BIL if possible. ore An additional arrester on the tertiary and secondarysides must be provided.IEC 6007 1-2 suggests the following formula todetermine such voltages:V,V,, = p.q.r - (I 8.8)wherep = factor for power frequency voltage already existing,when an overvoltage or a long-duration switchingsurge occur as noted above.q = response factor of the lower voltage circuit to thearriving long-duration surges (for power frequencytransferences 9 = 1 and for FOWs 4 = 0).(i) For secondary open-circuited.lightning surges q 5 1.3, andswitching surges q 5 1.8.(ii) For loaded secondary q < I .O.It is seen that normally it may not exceed 1 .O due tomany factors, such as the secondary may not be opencircuited,and the circuit parameters. L and C. that thearriving surge may have to encounter with both having adampening effect:r = a factor that will depend upon the transformerconnections, as indicated in Figure 18.14V, = a prospective long-duration switching surge voltagethat may appear on the primary side. If an arresteris provided on the primary side. this may besubstituted with the switching surge residual voltageof the arrester, V,,,n = transformation ratio of the transformer ( VI/V2)Example 18.2Consider segment Xof Figure 18.25 for the purpose of surgeprotection. The detailed working is provided in a tabular form,for more clarity, as under:

18/602 Industrial Power Engineering and Applications HandbookTransformer connectionsSurges on one phase onlyV,=lp.u, v,= v,=oSurges of opposite polarity on two phasesv, = 1 p.u, vy = - 1 p.u, vb = 0HVwindingLVwindingTertiarywindingHVwindingValue of rfor LVwindingHVwindingValue of rfor LVwinding/L 0 - 000AlS- -1/A- -1AAv( QAAA10 - 0I0 - 0-3-,.0 00 0x 40 0-113 -113mOq-m-113 ,xL- -113Oa-1ld3-1143* +bo x0. -1143- 1 1 3 q 'I3-1134b.S-0 - -1L0 - -10 -1A 0 - -1a fi -21d3 lN30-1Lf lId30 -1,Id3-+ -21130.a'O -1 G-1g - Grounded stari - isolated starA Star connected windingsb yDelta connected windings"& Zig-zag windingsFigure 18.14 Values of factor 'r'

=~~~~~~~~Surge arresters: application and selection 18/603ConsiderationsTransformer voltage ratioConnectionsRatingApprox. surge impedance from graphs of Figure 17.7(b)by extrapolation (obtain accurate value from themanufacturer)Surge travels from higher voltage side of the transformerto the lower voltage side. Consider a surge protectionon the primary, with details as follows:BIL of transformer from Table 13.2.(Choosing a higher level, as the system being exposedto the atmosphere).Primary sidevm1 p.u.Secondary side V,1 p.uMax. continuous operating voltage, MCOV (Vm/% 3);HV sideLV sideCharacteristics of the arrester chosen from Table 18.9,class 111; Standard rating, V,(for detailed working and exact design parameters forselecting the arrester, refer to Example 18.3)Max residual voltage (lightning) at 10 kA, V,,, (8/20 ps)Max. residual voltage (switching) at 1 kA, V,,, (30/60 ps)Max. residual voltage (FOW) at 10 kA, Vr, (1/20 ps)TOV capability for 10 sReflection of surges:Z, of jumpers through which will travel the surge to thetransformerAs the transformer HV side is already protected by anarrester, it is not necessary to consider the influence ofrefraction of surges at point A, which is quite meagre inthis case,The reflected wave will dampen the incidence surge bySurge transferences through the HV side of thetransformer(i) Capacitive transference (initial voltage spike)CAssuming = 0.4c, + csV, = 2.5 P.U.Since an arrester is provided at location A, it is appropriateto substitute V, by V,,, (FOW) = 386 kV peakParameters132/11 kV%ID60 MVA50 0Power frequencywithstand voltageLlWLHV side 275 kVr , 650 kV peakLV side28 kvrms75 kV peak145 kV ,145lr2x = 118 kVlr312kVrrms\' 21' 312 x - = 9.8 kV145 = 83.7 kVv312 = 6.9 kV-\I 3120 kV355 kV peak294 kV peak386 kV peak203 kV peak (140% of V, which is OK)200 R2XV1X- 5050 + 200= 0.4 V,50 -V, .200~ - 0.6 V,50 + 200

18/604 Industrial Power Engineering and Applications HandbookConsiderationsp = 1 .I5 for a lightning surge in a KID transformer,which is too high compared to LV side LlWL of thetransformer of 75 kVPeak (protective level not to be morethan 75/1.2 = 62.5 kV peak, Table 18.2), and calls foreither an arrester on the LV side too or provision of afew surge capacitors across the secondary windingssuch that,The value of C' can be calculated if values of C, and C,are known, which can be obtained from the transformermanufacturer.Note Even then a surge protection is essential for thetertiary if the tertiary is provided.(ii) Inductive transferenceAssuming, p= 1 for a switching surge (in inductivetransference we have to consider longdurationsurges only)9= 1.8 for a switching surge743r= - from Figure 18.14 for a 'f&fitransformer with surges of oppositepolarity appearing on two phases.n = 145/12 = 12.1V, = V,,, (switching)= 294 kV peakand the power frequency withstand capacity of the LVwindings:. Protective marginwhich is too low. It is, however, possible to make it up byselecting the arrester on the primary side with a lowerswitching V,,,. Consult the arrester manufacturer for itor provide an arrester on the secondary side also.Moreover, the response factor, q is considered very high,which may not be true in actual service and an arresterat the secondary may not be necessary in all probability.The BIL of the interconnecting cables and the terminalequipment on the secondary must be at least equal tothe capacitive and inductive transferences of the primarysurges as determined above. If it is not so, the V,,, ofthe primary arrester must be re-chosen or an arresteralso provided on the secondary side.Parameters-:. V,, = 0.4 x 386 x 1.15= 177.6 kV peakCP 62.5c, + c, + c < 177.6V,, = p.9.r. -v,nv,, = 1 x 1.8 x -& x -294:_2 12.1= 37.9 kVpeak= 28 kV, s.or 28 fi kV,- 28fi - 1.0437.918.5.3 Effect of resonanceThis applies to systems up to 245 kV, where inductiveimpedance is significant.It is possible that at certain frequencies the capacitiveand inductive couplings of the transformer may resonate(X, = X,) (Section 24.4.1 and 2) during the course of along-duration surge and give rise to yet higher voltagetransferences. For critical installations it is advisable toidentify (e.g. by TNA) the likely surges and theirfrequencies that may cause such a phenomenon to helpto take corrective steps or modify the parameters C andL of the transformer at the design stage.An arrester basically is for equipment protection andmust be installed at all main equipment heads that areexposed to internal or external surges and whenever theamplitude of such surges, V,, is expected to exceed theBIL of the equipment.Figure 18.15 shows a power network with generation,transmission and distribution of power, illustrating the

~~~~~~ ~~Surge arresters. application and selection 18/605influences of the different kinds of surges that may appearin the system and which must be taken into account,while engineering a surge protection scheme for such asystem or a part of it.18.6 Selection of a gapless surgearresterTo provide the required level of surge protection forequipment or a power system against possible transientfrequency voltage surges, ZnO gapless surge arrestersare the latest in the field of insulation coordination. Wepro\;ide below a procedure to select the most appropriatetype and sizc of a ZnO surge arrester for the requiredinsulation coordination. Based on the discussions above,the f'ollowing will form the basic parameters to arrive atthe most appropriate choice:Service conditions: As for other equipment (e.g. motors,transformers or switchgears) a surge arrester isinfluenced too by unfavourable operating conditionssuch as noted in Table 18.4.Unfavourable operating conditions will require aderating in the rating of a surge arrester or specialsurface treatment and better clearances. Refer to themanufacturer for the required measures and/or deratings.Mechanical soundness: Such as strength to carry theweight of conductor and the stresses so caused, pressureof wind and, in extremely cold climates, the weightof ice.Maximum continuous operating voltage (MCOV) V,:This voltage is selected so that the highest systemvoltage, VI,,, as in column 2, Tables 13.2 or 13.3,when applied to the arrester is less th?! or equal tothe arrester MCOV, that is, V, 2 Vln/y'3.The BIL of the equipment being protected (Section18.3).'l'he arrester's nominal discharge current (In): Thisifies an arrester and is the peak value of a lightningcurrent impulse wave (8/20 ps) that may pass throughthe arrester for which it i5 designed. It may be one ofthe following:Table 18.4. .- ~~~~I1.5, 2.5, S, IO and 20 kAPirrcirnrrrr.\Standard operating conditionsAmbient temperature2 Sl stem frequency3 Altitude1 Seismic conditionc5 Pollution/contaminationStaridrird conditions-40°C to + 40°C48-62 HIIO00 mLocations prone to experience anearthquake of magnitude M = 5 or aground acceleration of 0.1 g and more(Section 14.6)Due to excessive rain humidity, smoke,dirt and corrosive surroundings etc.,wlricli may influence the arrester'sporcelain housing outer burface, hencethe insulating properties ofthe arrester.18.6.1 TOV capability and selection ofrated voltage, V,TOV is considered only to select the MCOV and therated voltage, V,, of the surge arrester. This is a referenceparameter to define the operating characteristics of anarrester. It plays no part in deciding the protective levelof the arrester, which is solely dependent on the transientconditions of the system, as discussed later. V,. is used tomake the right choice of an arrester and its energyabsorption capability to ensure that it does not fail underthe system's prospective transient conditions.To determine the level of TOVs and their duration, itis essential to analyse all the possible TOVs the systemmay generate during actual operation, and then decideon the mosl crucial of them. as shown in Table 18.5.Surge arrester manufacturers provide their TOV capabilitycurves in the shape of TOV strength versus duration ofthe TOV. Afew typical TOV capability curves are shownin Figure 18.16(a) for distribution class and Figure18.16(b) for station class surge arresters. They indicatethe ratio between V, and Vr which may vary frommanufacturer to manufacturer and is termed the TOVstrength factor K. For curve 18.16(b) i1 is typicallyv, = 0.8 v,Generalizing this for different condition4 of TOV andmake of surge arrester,V, = K ' V,For each kind of TOV and its duration, a correspondingfactor (K) is obtained and with this is determined therequired rating, V,. of the arrester. The most crucial TOVmay be selected as the rating of the arrester. If it is not astandard rating as in the manufacturer's catalogue asshown in Tables 18.9 and 18.11, one may select the nexthigher rating available. A simple procedure is outlinedin Table 18.5 for a quick reference.To determine TOVThe main cause^ of TOV, a\ discu\\ed above, may beone or more of the following.Load rejection (Section 24.6.2).Resonance and ferro-resonance effects (Sections 24.3.1and 2 and 20.2.1(2)).Ground fault: Section 20. I . It is essential to know thegrounding conditions to determine the OV factor asbelow:Amplitude: As discussed in Section 20.1, groundingconditions, besides influencing the ground fault current,also raise the system voltage in the healthy phases.It is established that for an isolated ur resonantgrounded system this can rise to 1.73 times and fora solidly grounded system up to 1.4 times the ratedvoltage. When system parameters such as Ro, X,, andX, are known a more accurate assessment of thisfactor can be made by using OV curves as shown inFigure 20.16.

~ (b)~ 200118/606 Industrial Power Engineering and Applications Handbook*IIPrimary transmission(132,220 or 400 kV)+Secondaryi transmissionIsolated phase bus ' Jumper or mlJumDer ! rnhlpLc0'0 Occurrence oflikely surges:@ Shape@ Type of surges@ Effect ofreflections andrefractions[Not to beconsideredwhere arresteris provided]@ Exposure toatmosphere-TraAsferred IDirect surges intsurges both directions -4I50IlowI1IIIII30 I ,II ,I I300,l a I(a) They may be transferred surges iConventional shape is shown, but amplitude as well as shape will varydepending upon the type of surge and Zs of line and equipmentLightning, switching or steep fronted (FOW), depending upon thesystem voltage and switching conditionsI Ii 300iI 40I,III1 v. high low low v. highI' Indoor 1Indoor and OutdoorpartlyI outdoor 1Station classfrom HV side orDirect surges on the LV side in Iboth directionsII 1IOutdoor Outdoor OutdoorStation classlowIOutdooi@ Type of arrester@ Check effectof surgetransferences:(1) Capacitive(2) Inductive@ Check forresonance effectsL For all voltages ck- For systems up to 245 kVNote:1. Transferences from LV to HV side are not significant and need not be considered.2. Transferences need not be considered when arrester is provided on the LV side.b- For critical installations up to 245 kV -a,._c* Values are only indicative to assess the likely reflections at different junctions.Obtain accurate values from the equipment manufacturers.JFigure 18.15 (contd.)

Surge arresters: application and selection 18/607cablek kstributionI (3.3, 6.6 or 11 kV)I f'S ,kstributionI(400'440v)Cable orbus System ILT loads@ Occurrence oflikely surges:1'L b L(a) They may be transferred surgesfrom HV side or(b) Direct surges on the LV side inboth directionsISurge protection on HV side is, adequate to absorb direct surges.The transferred surges from HV side! to the LT side are too feeble to causeany harm. In case of static! switchings, on LT side, however,I spike arrester will be essential,I Section 6.14.1I@ Shape@) Type of surges@ Effect ofreflections andrefractions[Not to beconsideredwhere arresteris provided]@ Exposure toatmosphere@ Type of arrester@ Check effectof surgetransferences:(1) Capacitive(2) Inductive@ Check forresonance effectsConventional shape is shown, but amplitude as well as shape will varydepending upon the type of surge and Zs of line and equipment.Lightning, switching or steep fronted (FOW), depending upon thesystem voltage and switching conditions.IIlowII50 IIhighI loo!I IIIv.highI 1I2000 IOutdoor Outdoor Outdoor or IndoorIindoorStation classIIIIDistribution classFor all voltagesciFor systems up to 245 kV -41 NoteI 1. Transferences from LV to HV side are not significant and need not be considered.I 2. Transferences need not be considered when arrester is provided on the LV side.II EFor critical installations up to 245 kV*Values are only indicative to assess the likely reflections at different junctions.Obtain accurate values from the equipment manufacturers.Figure 18.15 A power generation, transmission and distribution network and strategic locations for the surge arresters

~~~ ~forfor18/608 Industrial Power Engineering and Applications HandbookTable 18.5Determining TOVTOVs Cause of TOV OVjactor Tripping timeTOVfacror Kyrom Figure 18 16(b) for apurticular brand of surge arresterTOV,"Ground fault(i) for a solidly grounded system 5 1.4 1 or 3 seconds 1.16 - for 1 secondI. I3 - for 3 seconds(ii) For an isolated neutral system 1.73 IO seconds to a few 1.1 ~ IO secondshours and more0.93 ~ 2 hoursTOV, Load rejection 1.1 1 second 1.16TOV," Phase-to-phase faulth1 or 3 seconds1.16-forlsecond1.13 - for 3 seconds"Consider only one eventuality to occur at a time.bA phase-to-phase fault is a transient condition and may give rise to TRVs in the healthy phase. Such a situation may exist for one to threeseconds, depending upon the protection scheme adopted. It is a long-duration condition for a surge arrester. Such an occurrence may be rareand need not be considered if the GF condition is already taken into account. However, for the sake of accountability, the voltage rise inthe healthy phase under such a fault condition may be considered similar to the first pole-clearing condition of a circuit breaker. Althoughit may not be appropriate to equate the two conditions, the magnitude of overvoltage may fall in a similar range. IEC 60056 has suggestedthe following TRVs across the first pole during a brcaker-opening sequence, while the other two poles are still in a cluaed condition:up to 72.5 kV 1.3 V,100-170 kV 1.3-1.5 V,245 kV and above 1.3 V,(V, being the rated voltage of the system)The overvoltage factor, due to a phase fault, may thus not exceed a ground fault (GF) condition. It would, therefore, be appropriate toconsider the GF condition instead of a phase fault and, in all probability, it will be adequate to account for this contingency.Duration: This will depend upon the ground faultprotection scheme adopted and may be considered asfollows:(a) For a solidly grounded system: 1-3 seconds (normally,generation and transmission systems areprovided with a longer tripping time of up to 3seconds, and a distribution system still longer, say,up to 10 seconds).(b) For an isolated, impedance or resonant groundedsystem: from a few minutes to several hours (whenit is 2 hours or more, it may be considered continuousfor the purpose of selection of an arrester).Short-circuit condition (Section 13.4.1(6)).Based on these discussions and experience from differentinstallations, the likely power frequency overvoltage(TOV) over a long period of operation for different voltagesystems can be determined. For a more accurate value ata particular installation, it is advisable to carry out asystem study. Generally, not more than one contingencymay occur at a time. However, depending upon the typeof system, which may be critical and more sensitive toload variations, a fault condition or frequent switchings,more than one contingency may also be considered.Example 18.3For a 400 kV systemV, = 420 kV and420V, = - = 243 kV45and minimum rated voltage V,, = 243/0.8 = 304 kV, when thesystem is not subject to any TOV. Consider a solidly groundedsystem, with a protective scheme of 3 seconds and theeventuality of load rejection, which may occur simultaneously::. Total TOV = 1.4 x 1.1 = 1.54From Figure l8.16(b), the TOV factor, K= 1.13 for 3-secondtripping.:. Required V, = 1.542431.13= 331 kV (which is higher than 304 kV)t-0.01 0.1 1.0 10 100 1000 10000TimeCurve - 1, No prior energy.Curve - 2, With prior energy.Figure 18.16(a)TOV capability of a distribution class arrester

Surge arresters: application and selection 18/609”., ,0.1Figure 18.16(b) Typical TOV capability of a station class arresterFrom the reproduced protective characteristics of this arrester,as in Table 18.11, the nearest next higher rating available is336 kV for a 420 kV system voltage. This is the basic ratingof the arrester which must be further checked forWhether its protective level can protect the BIL of the systemand/or the equipment it is protecting, andits energy capability to clear safely the long-durationprospective voltage surges.NoteIncreasing the TOV level, i.e. choosing a higher protectivelevel (V,,,) of the arrester, may increase the life of the arresterbut will reduce the margin of protection for the protectedequipment. Selection therefore should be done judiciouslybearing all these aspects in mind.18.6.2 Selecting the protective level of thearresterFor prospective voltage surges, as described in Table18.1. which may arise in the system during normaloperation the protective characteristics of the surge arrestera must be well below the BIL of the equipment at allpoints. For a minimum protective margin refer to Table18.2. It is the basic parameter of a surge arrester thatdefines its protective level. This is the voltage that willappear across the arrester terminals and thus also acrossthe equipment being protected during a current dischargeon the occurrence of a voltage surge. It should be wellbelow the BIL of the equipment under protection. Referto the load diagram shown in Figure 18.17. for moreclarity.To determine this. consider the simple power circuitdiagram of Figure 18.18(a). whereZ, = surge impedance of the line on which is connectedthe equipment to be protected from the source ofsupply up to the arresterZz = surge impedance of the equipment to be protectedZ, = surge impedance of the arrester at C’[Then applying Thevenin’sY’ theorem, the equivalent circuitcan be represented as shown in Figure 18.18(b). Onapplication of a voltage surge. the arrester will start2’Thevenin’s theorem: This is one of the many theorem5 deducedmathematically to solve intricate circuit. This theorem hasestablished that it is possible to replace any network with linearparameters and comtant phasor voltage source\. a5 vieu cd fromterminals, a and h. Figure 18.19(a). by a \ingle phasor voltagesource E with a single series impedance Z. Figure 18.19rb). Thevoltage E is the same that would appear across the terminals (I andh of the original network when these terminal\ are open circuited.and the impedance Z. is that viewed from the same terminals whenall voltage sources within the network arc short-circuited. Thi\ canhe illustrated by the following example.Consider the simple switching circuit of Figure 18.191~). If wewish to find the current through the impedance Z3. on thc closureof the snjitch, 5, then:I=Voltaee across thij circuit (Zq) beforeIclosing the switch. ,EEauivalent imnedance. 2. of the remaiiiinz circuitas seen from this \witch (Figure 18.19d)- E- Ez, ‘Z2z, + ~%, +z,and the voltage across this circuitI

18/61 0 Industrial Power Engineering and Applications Handbookby drawing its elements' I 0~ V" characteristics as shownin Figure 18.17 and drawing a load line on it to representthe characteristic of the prospective surge. It is drawnwith V, as the ordinate and the current through theequipment I, (had there been no arrester) as the abscissa:VIThen I, = -z,and the protective current through the arrester on adischargeVoltage absorbed or the voltage that isdischarged to groundI, =Impedance of the line and the equipment4-/ VI - Vre,/" p - 7Discharge current1. Surge arrester protective characteristics2. BIL of the equipment.3. Actual voltage surge.Figure 18.17 Load diagramwhereV, = prospective surge voltageV,,, = residual surge voltage across the surgearrester (the impulse protective level of thearrester) which must be below the equipmentBIL, with a sufficient protective margin,and Z, = equivalent impedance of Z, and Z2.(the impedance of arrester, being too small,is ignored)The criteria to determine the safe protective level ofan arrester are the BIL of the equipment, as shown inTable 1 1.6 for motors, Tables 13.2 or 14.1 for switchgears,Table 32.1(A) for bus systems, and Tables 13.2 and 13.3for all other systems. Motors have a comparatively lowerBIL, but they are not connected directly on an outdoorqe7Thesr4Z,... r - ----L ~, ~.I ____.1 ,: ZI I;-- r ---L.., I . Ii(b)-G T GZ, = Source impedanceZ, = Impedance of equipment being protectedZ, = Impedance of surge arrester. At V,, it is negligibleFigure 18.18 Equivalent circuit applying Thevenin's theoremNetwork withlinear circuitparametersand constantvoltage sourcesConstantimpedanceax9mMay be connectedto any other (b) bnetwork.Thevenin'sequivalent circuitZ,conducting at a certain voltage and carry a certaindischarge current. The voltage at which the conductionwill start is the impulse protective level of the arresterand is termed the residual voltage (Vre5) of the arrester.By a manufacturer it is determined for each particulararrester to establish its protective level. It is established(c)(d)A simple power circuitEquivalent impedanceFigure 18.19 Thevenin's theorem

Surge arresters: application and selection 18/61 1overhead line. and are also reasonably shielded by theline transformer, switchgear and cables. Switchgear andthe bus systems, which may or may not be as shielded asthe motors, have comparatively a higher BIL than a motor.Their prescribed impulse withstand level for more exposedinstallations is given in Table13.2. list I1 or 111, while forshielded installations, it is lower and given in list I. Onemay notice that list I is still higher than a motor. On thisBIL is considered a suitable protective margin to providesufficient safety to the protected equipment as in Table18.7.Protective characteristics of an arresterThe protective characteristic of a surge arrester is definedby its V,.,,, as a function of its nominal current (I,) andthe time of rise, t,. in the impulse region as noted above.It is seen that the characteristic of an arrester varies withthe front time of the arriving surge. Steeper (faster rising)waves raise the protective level (Vres) of an arrester, asillustrated in Figure 18.20. and reduce the protectivemargin for the equipment it is protecting. Refer to Figure18.17 for more clarity. Figure 18.20 gives typical characteristiccurves of a leading arrester manufacturer, drawnfor different magnitudes of current waves (340 kA),V,,, versus t,. From these curves can be determined therevised V,.,, during very fast-rising surges to ensure thatthe arrester selected is suitable for providing adequatean protective margin during a fast-rising surge.Protective distanceThe protective level as determined above is true onlywhen the surge arrester is mounted directly on the protectedequipment (Figure 18.21). But this is seldom possible,as ther-e is usually a gap between the surge arrester andthe equipment, due to arrester height, connecting leadsand the working gap required between the mounting ofthe arrester and the protected equipment. On the occurrenceof a voltage surge, while the arrester will conductand absorb the part of surge voltage that is in excess ofits protective level (V,.,,), the residual voltage, V,.,,, willtravel ahead with the same steepness (r.r.r.v.) until itreaches the equipment under protection. It may regain asufficient surge voltage to endanger the BIL of theequipment. Since the voltage will continue to rise as ittravels ahead, as illustrated in Figure 18.22, the equipmentwill be subject to higher stresses than the protective levelconsidered for the surge arrester. The distance betweenarrester and the equipment and the r.r.r.v. will determinethe excess stress to which the equipment will be subject.This can be determined byV, = V,,, + S.2.T(See ABB 199 1 ) whereV, = actual surge voltage at the equipmentS = r.r.r.v. of the incoming wave in kVlp( 18.9)The factor 2 is considered to account for the reflectionof the incident surge at the equipment (equation ( 18.3)):T = travelling time of the surge to reach the equipmentfrom the arrester terminals.If 1 is the distance in metres from the arresterterminals to the equipment, then(considering the propagation of surge in the overheadlines at 0.3 kmlys, Section 17.6.6).The longer the distance, I, the greater will be the severityof the oncoming wave which would reduce the protectivemargin of the arrester dangerously. Safe protectivedistances are normally worked out by the arrester manu-The curves provide residualvoltages at different front timesin per cent of residual voltageat 10 kA 8/20 us10.1 0.2 0.3 0.4 0.5 0.6 0.8 1.0 2.0 3.0 4.0 5.0 6.0 8.0 10.0Figure 18.20Front time (ps) -Variation in protective level of an arrester with front time

200018/61 2 Industrial Power Engineering and Applications HandbookNote-------+Protection to benear the equipmentand not on the line- GSurge arresterFigure 18.21 Ideal location of a surge arresterA lightning surge protective level is found to be more stringentthan the switching surge or FOW protective levels. Hence, itis sufficient to check the protective distance requirement forthe lightning surge alone.In such cases, care must be taken to reduce the protectivedistance, I, if possible, otherwise equipment with a higher BILmay have to be selected or an arrester with a lower V,,,,considered. Alternatively one may have to provide two arrestersin parallel, which is also an acceptable practice.18.6.3 Required energy capability in kJkV,Figure 18.22 Effect of distance on the protective levelfacturers and may be obtained from them for more accurateselection of an arrester.Example 18.4For the arrester of the previous example,V,,, = 420 kV,V, = 336 kVV,,, = 844 kV for a lightning surge protective margin at 20 kAdischarge current, from the manufacturer (Table 18.11)It we consider the lightning surge with a steepness of 2000kV/ps, then for a total distance of, say, 8 m from the arresterto the equipment- 844 ~ x 2 x 8 xs -= 844 + 107= 951 kV0.3If we maintain a protective margin of 15% (to be more safe),then the minimum BIL that the equipment under protectionmust have= 1.15 x 951= 1094 kVwhich is well below the BIL of the equipment (1175 kV) as inTable 13.3, list I. One can therefore safely select the arresterwith the next higher rated voltage, V,, at 360 kV and a V,,, at904 kV.This is the energy the arrester has to absorb while clearinga switching surge. It also depends upon the distance ofthe arrester from the equipment it is protecting, asdiscussed above. The basic parameters that will determinethe required energy capability (I'Rt), are current amplitude,steepness, duration and number of likely consecutivedischarges. The energy capability of an arrester dependsupon the size of the ZnO blocks. By resizing these blocks,an arrester of a higher energy capability can be designed.Hence, an arrester with a higher kJ/kV, capacity will belarger than one with a lower KJ/kV,. ZnO blocks cangenerally absorb much higher energies at low currentswith long durations (i.e. power frequency stresses undernormal operation) than higher currents for short durations(i.e. surge voltage capacitive discharges). To select theright type of arrester the energy capability of the arresterin kJ, therefore, must be determined by what it has toabsorb on every discharge. System studies on and pastexperience of similar other installations can be a goodguide for the likely number of consecutive discharges anarrester may have to perform at a time. Generally, twoconsecutive discharges are considered to be adequate.Like other equipment, a surge arrester becomes too heatedtoo during normal service, even when it is not conducting,due to its continuous charging current, Zzno (Figure 18.4)however small the loss content. When the arrester isconducting, the level of discharge current (I = kV-) is afunction of the protective level of the arrester, Vres, andthe rest of the severity of the voltage surge is absorbedby the arrester.The rating of an arrester is therefore defined in termsof its energy capability to absorb at least the requiredenergy on each discharge and the number of dischargesthe arrester may have to perform in quick succession. IEC60099-4 has prescribed the minimum energy capabilitythat an arrester must possess, at each discharge. Thegraphs in Figure 18.23 indicate these levels for differentclasses. It may be noted that an arrester with a higherenergy capability level will mean less strain or risk offailure for the arrester, but at a higher cost. A brief procedureto determine the energy level that the arrester may haveto absorb on each discharge is given below.Of the three types of surges noted earlier (Table 18.1)which the arrester may have to sustain and absorb, theswitching surge energy would be the maximum that wouldexist for the longest duration, compared to lightning orvery steep-fronted FOWs (Section 17.3) which are ofvery short duration. Normally, therefore, a switching surgealone is considered for this purpose. During the operationof a surge arrester, a part of the surge, equivalent to the

~~ ~Class\\ISurge arresters: application and selection 18/61 3be considered as three consecutive discharges forall practical purposes. It is, however, seen that inservice, even two consecutive discharges are rareand three may never occur. It is, therefore, sufficientto consider two consecutive discharges for selectingan arrester. The normal practice by leadingmanufacturers is to specify only the total energycapability for which their arresters would besuitable during consecutive discharges.-1 3 3.51.95VreJ v,Figure 18.23 Classification of arresters in terms of specific energykJ/kV, versus V,,,/V, as in IEC 99-4/1991protective level of the surge arrester, will conduct andappear as the residual voltage across the arrester in termsof its protective level and the rest would be absorbed byit. What is absorbed would determine its absorptioncapability. For a switching circuit, as shown in Figure18.18(a). this can be theoretically determined by:( 1 8. 10)whereW = energy absorbed in kWs or kJ (1 W = I J/s)V, = prospective switching surge crest voltage (kV)V,.,, = Switching surge residual voltage of the arrester(kV)N0tpAs in IEC 60099-4. the lowest value of Vrcb must heconsidered which occurs at a lower switching surge dischargecurrent. such as a switching surge, V,,,. at 1 kA < 2 kA < 3hA. etc. Tahles 18.9 and 18.1 I illustrate this.Z, = surge impedance of the affected lineT = travelling time of the switching surge from thearrester to the equipment in ,us. The factor 2 isconsidered to account for the reflection of theincident switching transient wave at the equipment(equation (18.3)). The virtual duration of the surgepeak. considered in Table 18.6, also corroboratesthis.PI = number of consecutive discharges. The energycapability of an arrester is its capability to dischargethree such switching surges at an interval of 50-60 seconds each (IEC 60099-4). But since thethermal time constant of ZnO blocks is high (inthe range of 60- I00 minutes) the time interval of50-60 seconds does not really matter, and it mayApparently a higher level of V,,, would mean a lowerlevel of energy absorption by the surge arrester in termsof V,. For an excessive level of W required, it is better toselect a higher V,. If it jeopardizes the required protectionlevel, then select another type of surge arrester with ahigher energy capability.To determine W from the above it is essential to knowthe system parameters. Based on system studies ofdifferent voltage systems (transmission lines particularly)many data have been collected. Typical data as suggestedby IEC 60099-4 are provided in Table 18.6 for a generalreference. For secondary transmission or primary distributionnetworks up to 245 kV too. where a lightningsurge forms the basis of selection for the protective level(VTe5j of the arrester, only the switching surge must beconsidered to determine energy capability.Example 18.5 in Section 18.10 illustrates the procedureto determine the energy capability requirement of anarrester for a particular system. The ultimate selection ofan arrester is a compromise between its protective level,V,,, TOV capability and energy absorption capability.18.7 Classification of arrestersThese are classified by their nominal discharge currentsI, (8120 s) and surge energy absorption capability duringa discharge (k . J/k . V,j. Each discharge current is assigneda system voltage according to IEC 60099-4, as noted inTable 18.7. The energy capability of an arrester will varywith the overvoltage conditions of the system. It istherefore essential to ensure that the arrester chosen hassufficient capability to sustain the required system TOVand surge conditions during long years of operation. IECTable 18.6Line Line dischurgrdischarge cluss ofcurrent iirrester(kA 1I 0 1IO 210 320 420 STypical parameters of a transmission lineBased on IEC 60099-4'These are the line discharge test values to test an arrester. asrecommended by IEC and have been considered here lor the purposeof selection of an arrester. Refcr to Example 18.5.

~~18/61 4 Industrial Power Engineering and Applications Handbook60099-4 has also recommended the line discharge classesof arresters, based on the required energy level in k . J/k.Vr and the ratio of Vr,,/Vr (Figure 18.23). The energycapability of a normal arrester produced by leadingmanufacturers is generally higher than recommended bythe IEC for a particular class of arrester. The last columnof Table 18.7 and Figure 18.23 recommend the minimumenergy capability of different classes of arresters. Theclass of arrester for a system, particularly for heavy duty,can be easily chosen with the help of these figures for arequired level of k.J/k.V, and a ratio of Vre,/Vr. Arrestersclasses 1 to 5 are regarded as heavy duty and classifiedas station class by ANSI/IEEE-C62.11. For light duty, adistribution class may be selected to economize on cost.The nominal current, rated voltage, VI, level of protection.V,,,, and energy capability for such arresters are indicatedin Table 18.7 for a particular manufacturer.The subsequent example will provide a simple stepby-step-procedure to select an arrester for a particularapplication >245 kV.The application and rating of an arrester is a matter ofsystem study and systematic insulation coordination, likelydischarge current during a possible lightning dischargeat the location of the installation and experience gatheredfrom similar installations. Different countries may adoptdifferent practices, depending upon the stability of theirnetworks and the type of surges that may arise duringthe course of operation. Hence, only broad guidelinescan be given to choose an arrester. The rest is a decisionby the application engineer and his experience of powernetworks.18.7.1 Application of distribution class surgearrestersThe application of such arresters will be guided by theirfollowing basic characteristics:1 They possess a low-energy handling capability andhence cannot be applied for installations and equipmentthat may be subject to long-duration (250/2500 ps)switching surges requiring high-energy handling capability.They are, however, suitable to withstand an FOW,as this requires a very low energy handling capabilitywhich such arresters do possess.2 They also pose limitations for installations that connecta number of cables and capacitor banks, as these alsorequire a higher energy handling capability.3 They offer a relatively higher VI,, and may generallynot be suitable for protection of equipment that has alow BIL. The BIL of a rotating machine or a dry typetransformer, for instance, may fall lower than the V,of such arresters and the equipment may remainunprotected.4 Basically they are light-duty arresters and must beinstalled at locations that are not directly exposed tolightning strikes.These arresters are therefore employed for installationsand equipment that are not so critical as to cause ashutdown of the power system. Likely applications maythus be small distribution networks and equipment suchas a distribution transformer, feeding residential or smallindustrial loads, not directly exposed to the atmosphereand hence to direct lightning strikes. Such installationsare reasonably shielded, as when installed within a city.In Table 18.8 we provide typical data for an 11 kV surgearrester widely employed for the distribution of power.For higher ratings, consult the manufacturer or their datasheets. These arresters are normally used for protectionagainst lightning surges rather than switching surges.Lightning surges discharge only low energy which sucharresters can withstand.With the above in mind, 1.5 and 2.5 kA arresters arenot being used by many power networks, presumablyTable 18.7 Classification of arrestersArrester line dischargeclassNnminul discharge currentut lighming impul~e (8/20/pAs in ANSIAF in IEC 6 i ) O g d kA1 10"IO"10"20"20"up to 400 kV750 kV and aboveI3 I V, I 360} 360 < V, 5 750Generally more than recommendedas in graphs of Figure 18.23.Distributionclass52.5'1 1.5'up to 132up to 36hNot significant for the purpose ofarrester selectionNotes"Selection of any arrester will depend upon the duty it has to perform (light duty for indoor and heavy duty for outdoor installations) andthe required energy absorption capability.bYet to be defined by IEC 60099-4, hut generally for low-voltage systems. Not often in use for power system applications.'The application of such arresters is noted in Section 18.7.1.

Surge arresters: application and selection 18/61 5Table 18.8Technical details of distribution and stations class lightning arrestersParameters Distribution Station classaclass(for motor protection)Rated voltageNominal discharge current (8120 psec.)MCOVkV r.m.s.kA peakkV r.m.s.Max. residual voltage (kV peak) at:(a)(i)Lightning current impulse (8120 ps) peak2.5 kA(ii) 5 kA(iii) IO kA(iv) 20 kA(b) Switching surge residual voltage(1.O kA, 30/60 ps)kV peak(c) Steep current impulse (112 ps) 5 kA peak10 kA peakLong duration discharge class: (defines the energy capability)(a) Current A peak(b) Virtual duration of peakHigh current (4/10 ps)PkA peakTemporary power frequency overvoltage capability(kV r.m.s.),(i) 0.1 second(ii) 1 .O second(iii) 10.0 seconds(iv) 100 seconds957.65293033363033407510006510.5109.59.26105.5-17.719.02115.3-21.5--1003102.8-9.3110.011.58.1-11.5--100Notes"For 11 kV motors refer to Table 18.9.bTOV capability of the arrester is provided by the manufacturer as standard practice, either in the form of voltage versus time data, as givenin column 1, or a graph. A typical graph is shown in Figure 18.16(a).'To assess the loss parameter of the arrester, under normal operating conditions (power frequency), the following data, typical (for the abovearresters) may also be obtained from the manufacturer. Also refer to Section 18.2.Reference current: 1 .O mALeakage current at MCOV,(i) Resistive 3.0 mA(ii) Capacitive 0.8 mACourtesy: Elpro Internationalbecause of their frequent discharges and rapid failuresand an underprotection to the power network or theequipment. They are generally employed as resistors toabsorb switching voltage spikes* or surge transferencesin household appliances, uninterrupted power supply(UPS) systems, computers and similar applications. Forpower applications, such as small distribution networksand distribution transformers etc., 5 kA surge arrestersalone are in common practice. For distribution linesrunning through long open terrains, such as for ruralelectrification, however, the practice may be differentonce again, in view of higher degree of exposure of thelines and the terminal equipment to direct lightning strikes.18.8 Surge protection of motorsThe following are a few basic aspects that must be keptin mind while attempting to protect a rotating machinefrom system surges:*Voltage spikes may have high amplitudes but are of extremelyshort duration, like an FOW.1 The surge arrester must be suitable for absorbing energyof long-duration (250/2500 ps) switching surges.2 A rotating machine is a low BIL equipment. The arrestermust offer a low V,, to protect the machine adequately.3 A capability to sustain FOWs if they arise duringswitching operations.A distribution class arrester may not offer adequateenergy absorption capability, particularly for the energyof a long-duration switching surge. The residual voltage,V,,, of such arresters is normally high and may not offeradequate protection for the machine. They are, however,suitable for withstanding and absorbing energy of a shortdurationFOW. A distribution class arrester is thus notcapable of protecting a rotating machine.A surge capacitor too, when used in parallel tosupplement such an arrester, will not be able to make thearrester suitable for such a duty. This is effective onlyagainst very steep surges having a very high surgefrequency of 10 kHz and more. The frequency of a surgecan be determined by103f, = - kHz (t, is in ps)4t1(18.11)

~~~~~18/61 6 Industrial Power Engineering and Applications HandbookTable 18.9Protective characteristics of gapless station class surge arresters for a nominal discharge current of 10 kAa12345b678910111213System voltage (50 Hz)kVRating of surge arrester(at the reference current)kVDischarge classEnergy dissipationcapability cumulativeoperation (UkV,)High current impulsewithstand of 4/10 pswave shapekAReference current ofthe arrester at ambientternperaluremAComponents of thecontinuous leakagecurrent at COVResistive pA peakCapacitive pA peakWatt loss at MCOV perkV of rated voltageWkAMaximum continuousoperating voltagekV r.m.s.Max. residual voltageat lightning impulse of8/20 ,us wave kV peak5 kA10 kA20 kAMaximum residual voltageat switching currentimpulse of 30/60 ps,I kAMaximum residualvoltage at steepfrontedimpulse (1/20ps, IEC 60099-4) at10 kATemporary power frequencyovervoltage capability(a) 0.1 second kV peak(b) 1.0 second kV peak(c) 10.0 seconds kV peak(d) 100.0 seconds kV peak420 245 245 I45 123 72 36 36 24 12 12390 216 198 220 96 60 36 30 18 12 9111 111 111 III I11 I1 I1 11 I I IIO 10 IO 6.5 6.5 4.5 4.5 4.5 2.5 2.5 2.5100 100 100 100 100 100 100 100 100 100 1005 5 5 3.25 3.25 2.25 2.25 2.25 1.5 1.5 1.5400 400 400 400 400 400 400 400 400 400 4001500 1500 1500 1500 1500 1500 1500 1500 1500 1500 15000.25 0.25 0.25 0.25 0.25 0.25303 178 168 102 81 52860 567 518 333 249 162900 602 550 355 264 175957 668 610 390 304 202780 496 457 294 221 1381050 654 600 386 288 202705 397 364 220 156 97580 382 350 212 149 93565 367 336 203 142 89550 305 280 169 135 8430.5103110I28871240.25 0.255856535125.5859010471101484644420.25 0.2515 IO60 3865 4075 4652 3273 4529 1928 1826 1725 160.257.65333541284014141312a These are typical figures and can be varied by the manufacturer to suit an applicationTo identify the suitability of the arrester on direct lightning strikes.Courtesy Elpro InternationalThe first peak being more relevant, we have thereforeconsidered it’s the rise time to determine the virtualfrequency of the surge. This frequency, for a long-durationswitching surge (high t,), can be quite low for the surgecapacitor to acquire enough reactancesuch that the surge arrester alone may have to sustain thebulk of the surge. The surge capacitor provides onlymeagre support. The basic use of surge capacitors is toabsorb only the FOWs and reduce their steepness to makethem safe for the turn insulation of the machine. When aswitching operation commences, initially it is only alongdurationswitching surge that generates. It is only afterrestrikes of the interrupting contacts that it amplifies toFOWs. The capacitor will activate only after the longdurationswitching surge amplifies to an FOW. For allpractical purposes, therefore, the arrester alone has to besuitable for handling such surges, irrespective of whetherit is supplemented by a surge capacitor. See also Example17.5.Arresters for such applications are therefore designed

Surge arresters: application and selection 18/61 7especially with low V,,, at steep surges and a high energyabsorption capability. Only a station class surge arresteris normally preferred for such applications, and wheresteep-fronted surges are envisaged, these arresters toomay be supplemented with a surge capacitor. Table 18.8also provides data for motor protection station class surgearresters. For 11 kV voltage systems, refer to Table 18.9.Since a surge arrester is normally an engineered item tosuit a particular application, these data are only for generalreference. For exact application and type of installation,it is always advisable to consult the manufacturer.18.9 Pressure relief facilityA surge arrester is a sealed unit to save it from allatmospheric hazards. It is normally filled with air.Explosions of surge arresters have been noticed duringservice. Explosion of a porcelain housing is dangerous,as the shell splinters can cause great damage to nearbybushings, insulators and other equipment and alsomaintenance personnel if working in the vicinity.It is possible that ZnO elements may break down intime due to thermal cracking as a result of system TOVsoccurring frequently or existing for long durations. orwhile clearing a lightning or a switching surge and evensubsequent to that. Breakdown of ZnO elements intosplinters may collide with the main porcelain housing.But most damage is caused by a flashover between theZnO elements and the side walls of the housing, whichmay result in puncture or crackdown. It may lead to acascading effect and cause an eventual short-circuit withinthe housing and result in a very heavy ground fault current( V, / \;3. Zg) through it.The fault current may cause a flashover and rises inthe internal temperature and pressure of the housing. Itis extremely important to make provision to release thispressure, otherwise it may lead to an explosion of thehousing, scattering splinters like bullets in the vicinity.This is dangerous and must be avoided. The pressurerelief capacity must be such that there will be no violentexplosion of the housing during a failure. Although itmay shatter, the arrester’s fractured pieces should notfall beyond a circle of a radius equal to its height,somewhat similar to the properties of safety glasswindscreens used in a car. International specificationsrecommend that on a pressure build-up, the hot gaseswill escape through the pressure relief diaphragm or thehousing may simply collapse as a result of thermal shock.An arrester that is vented must be quickly replaced.This is achieved through a pressure relief system inthe form of a pressure relief diaphragm at the top of thehousing. The pressure relief system is designed for thesystem fault level. To test the pressure relief system,IEC 60099-4 and IEC 60099-1 have also specified thetest current of the same magnitude as the fault current ofthe system, but for a very short duration, of the order of0.2 second or so. This brief period is enough to burst orshatter the housing of the arrester, hence a longer durationis of no relevance.The practice for heavy-duty arresters is to eliminatethe causes of internal flashover between the ZnO elementsand the housing at the manufacturing stage. Differentmanufacturers have adopted different methods forachieving this. One such method is providing a barrierof an insulating material not affected by heat and arcing,such as an FRP (fibre reinforced plastic) tube betweenthe housing and the ZnO elements.18.10 Assessing the condition of anarresterTo ensure adequate safety for a system and its terminalequipment against overvoltages and voltage surges it isessential to ascertain the soundness of the arrester atregular intervals. It should be possible to do this when itis in service without taking it off from the lines. Ifdeterioration of the ZnO elements is detected it mayneed more frequent future services or replacement of thearrester. It can now be planned well in advance. Therequirement is similar to ascertaining the condition ofpower capacitors when in service (Section 26.2). Like acapacitor, an arrester deteriorates too with time due todegradation of the dielectric strength of its ZnO elements.ZnO is a highly non-linear resistor element. The successof an arrester will depend upon its low, continuous resistiveleakage current Figure 18.4(a) to maintain low loss andlow heating over years of continuous operation. Whenthe ZnO blocks start to deteriorate which is a slow processas discussed earlier, the leakage current starts rising fromits original level. The rise in current is rich in thirdharmonic component due to the non-linear characteristicof the ZnO blocks. Other reasons for degradation in thedielectric properties and a rise in current may be one ormore of the following:Ingress of moisture through the seals, although Silicagelis provided beneath the arrester sealing to absorb themoisture.Failure of ZnO elements during or after clearing afew surges.Premature ageing of the ZnO elements.Temperature variations. The rise in I, is rapid at higheroperating temperatures.Frequent system voltage variations. andBeing continuously energizedA rise in leakage current is not desirable and is indicativeof deterioration of the ZnO blocks, which may lead tofailure. It is therefore necessary to monitor the leakagecurrent and detect a possible failure beforehand and takecorrective steps in advance. The maximum safe leakagecurrent is specified by the arrester manufacturer as arelation between I, and its third harmonic component,13r. 13T is expressed in terms of IZnO as discussed later. Itvaries with deterioration of the arrester and is used as areference parameter to assess the arrester’s condition.As the actual leakage current measured through IZnOstarts rising and approaches the maximum leakage currentin healthy condition closer monitoring of the arresterbecomes essential, to avoid an abrupt failure or explosion.Refer to Table 18.10 to monitor the condition of the

18/618 Industrial Power Engineering and Applications HandbookTable 18.10 Log book to monitor the condition of an arrester when in serviceDate of measurementHealth) IZnO(h, mdx'provided b, the arre\termanujacturerPAActual nieasurement of IZn O,,,at sitehPAAF R of IZnO,h,~_____ ~~Likeh condition of the arrester1st year3rd year5th year7th year (December)'8th year (May)'8th year, (November)'25070 2890 361 I5 46I60 64220 88300 I20GoodGoodGoodSigns of deteriorationSign of rapid deteriorationRecommended for servicing aroundthis timeRequires servicing or replacement"IZnOl,, = maximum leakage current in normal conditions.bIZnOl,, = actual leakage current measured at site."Closer monitoring recommended.The above figures are purely hypothetical.NotesMonitoring by such an instrument can be carried by installing it permanently or by using it as a portable instrument. Even when connectedpermanently, the measurements are taken at long intervals for short durations for periodic checks of the arrester. The table suggests onlylikely check periods, which may vary with environmental conditions and the quality of the system voltage. Higher discharges or frequentswitchings of the line breakers, for instance, may deteriorate the ZnO blocks more quickly and require more frequent readings than forsystems having lower discharges or switchingsarrester. To measure I,, let us analyse the basic circuit ofan arrester as considered in Figure 18.5, whereI, = loss component and is about 5-20% of I, undernormal operating conditions. Any change inits value will contain a third harmonic componentbecause of its non-linearity. It will vary withsystem voltage and operating temperature.I, = leakage capacitive component leads I, by 90".It also depends upon the system voltage andthe operating temperature. It is, however,independent of I, and remains almost unaffectedby the deterioration of ZnO blocks, i.e. changein I,.IZnO = Total leakage current of the arrester. It also riseswith system voltage and operating temperature.Under healthy conditions this current is verylow (in the range of PA).Ideally, measuring the variation in IZnO should be enoughto determine the condition of an arrester. But it is not so,as it does not provide a true replica of I, for the followingreasons:System voltage harmonics With deterioration ofthe arrester, I, rises and so does its third harmoniccomponent, because of non-linearity of the ZnOblocks. But IZnO also measures the harmonics presentin the system voltage, particularly the third harmonic.The system harmonics are also magnified by theleakage capacitances of the arrester. Since an arresteris connected between a phase and the ground, thethird harmonic of the system finds its path throughthe grounded arrester and distorts the third harmonicof the ZnO blocks caused by their deterioration.Uneven distribution of C along the arrester(Lundquist et al., 1989)Pollution by the surroundings, such as by dirt andingress of moisture. In normal conditions this alsoraises I, and IZnO andCorona effect All such factors may influence theI, and IZnO in different proportions and be detrimentalin assessing the actual variation in I, throughIZnO. IZnO therefore cannot be regarded as a truereplica of I, monitoring IZnO may not accuratelyassess the actual condition of the arrester. To useIZnO to assess the condition of an arrester, it isessential to separate I, from it.The greatest effect of ageing is reflected by the variationin its resistive current, which is rich in the third harmonic.Variation in I, is used in assessing the condition of anarrester. By conducting laboratory tests to determine thecharacteristics of an arrester, we can establish a ratiobetween the total leakage current, IZnO and the contentof I,, to assess the condition of the arrester. If we canmonitor this current, we can monitor the condition of thearrester. Below we discuss briefly one such method bywhich this component can be separated out.Leakage current monitor(A diagnostic Indicator of metal oxide surge arresters inservice).NoteInstruments operating during discharges alone are basically surgecounters and can only indicate the number of discharges. Theyprovide no information on the condition of the arrester.To measure the resistive leakage current, I,, alone is a

Surge arresters: application and selection 18/61 9complex subject. However, there are a number of methodsto determine this. IEC 60095-5 mentions a few methodsby which the resistive leakage current, I,, can be measuredthrough IZnO. The main problem faced in all thesemethods is the presence of a system voltage third harmonicthat finds its way through the grounded arrester and showsup in IZnO. Therefore, unless the system voltage thirdharmonic is eliminated from IZnO, it will not provide atrue replica of the arrester's condition. Below we discussone more recognized method (See Lundquist et al., 1989)by which an attempt is made to separate out the systemvoltage third harmonic from IZnO. The method is basedon extraction of I, by third-order harmonic analysis ofIZnO. This is achieved by providing an electric fieldprobe located at the grounding end of the arrester. Theprobe compensates the third harmonic present in thesystem voltage so that the current measured at the groundend of the arrester contains only the third harmonic of I,.Harmonics, other than the third even if they are presentin the system or IZnO, are of little relevance, as theinstrument analyses only the third harmonic.The instrument separates out I, and I, and provides adirect reading of I, and hence the condition of the arrester.Refer to Table 18.10 providing a brief procedure to monitorthe condition of an arrester through such a monitor.A typical layout of such an instrument and its accessoriesis shown in Figure 18.24(a). It consists ofMonitorMonitorground wireI1Currentprobe withAdapter cableFigure 18.24(a) Leakage current monitor with accessories(Courtesy: TransiNor As)Power lineLeakage current monitor - this can be connectedperma-nently for continuous reading or periodicmonitoring. The normal practice is to measure onlyperiodically for a short period to take averagemeasurements on an hourly, daily, monthly or yearlybasis. When not connected permanently, theinstrument can also be used as a portable kit to monitorthe condition of other arresters installed in the vicinity.Field probe - to compensate the third harmonic ofthe system voltage to make the IZnO free from thethird harmonic of the system voltage. This methodof I, measurement therefore provides more accurateand closer monitoring of the arrester.Clip-on CT - to measure IZnO and is mounted atthe grounding end of the arrester.Current probe - to measure the third harmonic componentof I,. It is then converted to actual I, from the ZnOcharacteristic data provided by the arrester manufacturer,I, versus 13,, corrected to the site operatingtemperature and voltage. The value of I, is then usedto assess the condition of the arrester.Adapter (connector) - to connect the CT with theinstrument."I I I"ZnO blocks(Showingsectional view)Leakage groundcapacitancescausing leakagecurrentPorcelain discsinstrumentFigure 18.24(b) illustrates the use of a leakage currentmonitor. The instrument can be used to display or monitoron a computer remotely and store data at intervals asrequired to provide diagnostic information. Now it iseasier to take corrective measures in time. The instrumentcan also be programmed to give an alarm at a presetvalue of I, when the actual operating conditions exceedthis.lzno - Surge arrester leakage current free from system voltage 3rdharmoniclc - Free from system voltage 3rd harmonicI, - With 3rd harmonic current componentFigure 18.24(b) Measuring leakage current through an arresterduring normal operation

~ =~~18/620 Industrial Power Engineering and Applications HandbookExample 18.5Step-by-step selection of a surge arrester for the protectionof equipment mounted on a 400 kV transmission line.Considering conditions (i) and (ii) occurring at the same instant,:. total TOV = 1.4 x 1.1 = 1.54 for 3 secondsTOV factor as in Figure 18.16(b), K = 1.13For this voltage, we have selected a Class 4 arrester, whichhas the nearest next higher rating as in Table 18.11 asV, = 336 kVAs regards the primary selection of the arrester, it is completeat this stage. But this selection must be checked for its adequateprotective level and the energy absorption capabilityConsiderationsNominal system voltage (4)Maximum system voltage (V,)Data400 kV(r.m.s.)420 kV(r.m.s.)In per unit (P.u.) 1 p.U. = 420 x&-6= 343 kVMaximum continuous operatingvoltage (MCOV), 7 vrn = vc4206243 kV(max.)Temporary overvoltages (TOVs) OV trippingas in Table 18.5 factor time(i) Due to a ground fault for asolidly grounded system witha GFF of 1.4 TOVl 1.4 3 sec(ii) TOV due to load rejectionTOV2 1.1 1 sec(iii) TOV due to a short-circuitfault TOV3 1.1 3 sec(A) Checking for the protective levelFor this rating, the arrester protection levels are:Vres(max) for a switching impulse wave of 1 kAfor the purpose of energy capability = 654 kV.For switching impulse protective margin at 2 kA= 676 kV.Vres(max) for a 20 kA lightning impulse (8/20 ps)= 844 kV.Vre,(max) for a 0.6 ps FOW from Figure 18.20= 1.12 x 844 = 945 kV.BIL of the equipment to be protected, from Table 13.3list I (considering the lower side to save on the cost ofequipment)For switching surges = 950 kVFor lightning surges = 1175 kVFor very fast-rising surges (FOWs) = 1175 kV (assumed).This value must be obtained from the manufacturer ofthe equipment to be protected.To check the protective margins:Type ofsurge(i) For switchingsurges(ii) For lightningsurges(iii) ForFOWsMargins 1 Margin Suitabilityrecommended available of theas in Tablearrester18.21.151.051.15950 = 1.40 OK676-- 1175 - 1.39 OK844-- 1175 - 1.24 OK945Table 18.11 Protective characteristics of a gapless station class surge arrester for, (1) Line discharge class 4, and (2) Singleimpulse energy capability = 7 k.J/k.V,System Rated Max. MCOV TOV capabilityvoltage voltap cont. as in foroperating ANSIvoltage tests(CO V)420 330 264 267 383 363336 267 272 390 370360 267 291 418 396372 267 301 432 409378 267 306 438 416381 267 308 442 419390 267 315 452 429396 267 318 459 436420 267 336 487 462aAny impulse with a front time longer than 30 ps.Courtesy: ABBMaximum residual voltage with current waveSwitching surgesa1 kA 2 kA 3 kAkVpeak kVpeak kvpeak642 664 680654 676 693701 724 742724 748 767736 760 779741 766 785759 784 804771 796 816817 845 866Lightning surges 8/20 ps5 kA 10 kA 20 kA 40 kAkVDeak kVDeak kVDeak kvpeak71 873 1783810823829849862914759773828856870877897911966829844904934949957979994105591 1928994102710441052107710931160

=Surge arresters: application and selection 18/621NotesThe above selection of the arrester may be consideredas having taken account of reflections at the terminals.When such an arrester is installed on the system, ittakes care of switching, lightning or steep to very steeprising surges, irrespective of their amplitudes.It is presumed that the protective distance (distance ofthe protected equipment from the arrester) is short andwithin safe limits. It too must be checked before a finalselection of the arrester, as explained in Section 18.6.2.The above selection, however, seems to be appropriatefor a distance up to 8-10 m as considered in Example18.4.The arrester selected has protective margins muchgreater than the minimum required. In fact evenconsidering a protective distance up to 8-10 m, it wouldbe possible to select the arrester with the next higherV, to enhance the life of the arrester without jeopardizingthe safety of the equipment.(B) Checking for the energy capabilityIt is sufficient to check this for systems of range II alone, asthe energy requirements for systems of range I may bemoderate, since now the lightning surges are found to bemore severe than the switching surges. The lightning surgesdissipate a very low energy due to their extremely shortduration, while a normal arrester would be adequate to absorbmuch more energy than this.W (for each discharge)(1 8.10)To assess the energy capability of the arrester more realistically,let us consider the established system parameters as in Table18.6 to determine the maximum energy discharge throughthe arrester,whereVt = 2.6 V,= 2.6 x 336= 874 kVV,,, = 654 kV (for I, = 1 kA)Z, = 0.8Vr= 0.8 x 336= 269 R2.T = 2800~~NoteThe time of travel from the point the lightning strikes on theoverhead lines to the arrester, considering a line length ofapproximately 420 km and speed of propagation of the surgeas 0.3 km/,us,T=420~ 1400 PS0.3Accounting for reflections, the total time must be consideredas 2 x 1400 or 2800 ps. Accordingly, 2800 ps is specified byIEC 60099-4 for the testing of the arrester under the worstconditions. The arrester must also have the capability tosuccessfully discharge the prospective surge under suchconditions.:. w = (874 - 654) x 654 x 2800 x kJ269_- - 220 x 654 x 2800 x269= 1498 kJ or ~1498 = 4.46 kJ/kV,336The arrester chosen has a total energy capability of 7 kJ/kVr(from the manufacturer’s catalogue). In service the arresterwill be required to discharge much less energy than this asthe distance of the arrester from the point of surge originationmay be much smaller than considered.Line arrester class can also be selected quickly from Figure18.23, corresponding to Vres/Vr. In the above case, it is 654/336, i.e. 1.95, corresponding to which also the class of arresterworks out as ‘4’, as considered by us, and which will have anenergy absorption capability of more than 4.5 kJ/kVr perdischarge.(C) Checking for reflections and transferencesWe have considered protection of both 400 kVtransformers, one for primary transmission and the otherfor secondary transmission. We will now analyse theinfluence of surge reflections and transferences of a surgeoccurring on the 400 kV primary transmission bus asshown in Figure 18.25 and its effects on segments A andB.

18/622 Industrial Power Engineering and Applications Handbook15.75 kV, 200 MW15.751400 kV 250 MVA,400 kV Bus (primarytransmission)132 kV Bus (secondarytransmission)IX0)Jumper, 2s = 200 R132111 kV,60MVAZS~ = 50R* For surge protection of motors, refer to Section 18.8Cable 2s = 100 RfD 11/3.3or6.6 kV.Cable 2s = 100 C2IIHT distributionLT distributionLT loadsFigure 18.25 Surge protection analysis of potential locations for the network illustrated in Figure 18.15

Surge arresters: application and selection 181623Segment ASegment BBasic parameters Z,, = 300 R Zsl = 300Zsz = 300 RZ,, = 40 R(a) Reflection of surges: Low Low(b)As the arresters are being provided at theprimary of each transformer, this aspect neednot be consideredSurge transferences to the LV side(i)Capacitive transference(initial voltage spike)AssumingCPcp + cs0.4The actual value may be much less than this whenobtained from the transformer manufacturerp for a YlA transformer 1.15Vt(F0W)945 kV peak:. v,,= 0.4 x 945 x 1.15= 434.7 kV peakBIL (lightning) of the transformers’ LV sides fromTable 13.2Minimum protective level requiredfor 15.75 kV LV(V, = 17.5 kV)= 95 kV peak9511.05 = 90.5 kV peakwhich is too low compared to VI, and requires an arrester on the LV sideor provision of surge capacitors across the secondary windings, such thatcp

2309/198918/624 Industrial Power Engineering and Applications HandbookRelevant StandardsIEC Title IS BS60056/1987 High voltage alternating current circuit breakers 13118/1991 BS 5311/199660071 - 1/1993 Insulation coordination - Definitions, principles 2165-1/1991, BS EN 60071-and rules 2165-2/1991 1/199660071-2/1996 Insulation coordination - Application guide 37 16/1991, RS EN 60071 -2/199760099- 1/1991 Surge arrester - Non-linear resistor type gapped 3070-1/1990 BS EN 60099-surge arresters for a.c. systems. 1/199460099-4/1991 Surge arresters - Metal oxide without gaps for a.c. 3070-3/1993 BS EN 60099-power circuits 4/199360099-5/1996 Surge arresters. Selection and application 4004/1990 BS EN 60099-recommendations 5/199761024- 1/1993 Protection of structures against lightning ~ DD ENV 61024-1/1995General principles BS 6651/1992Relevant US Standards ANSINEMA and IEEEANSI/IEEE-C62.1/1994ANSI/IEEE-C62.2/1994ANSIIIEEE-C62.11/1993ANSI/IEEEC 62.22/1997ANSI/IEEE13 13.1/1996NEMNLS- 1/1992Standard for gapped silicon carbide surge arresters for ax. power circuitsGuide for application of gapped silicon carbide surge arrester for a.c. systemsMetal oxide surge arresters for a.c. power circuitsGuide for the application of metal oxide surge arresters for a.c. systemsInsulation coordination, Definitions, Principles and RulesLow voltage surge protection devicesNotes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publicationList of formulae usedElectrical characteristics of a ZnO surge arresterCharacteristic of a ZnO blockI= K.V" (18.1)K depends upon geometrical configuration, crosssectionalarea and length of ZnO Blocka = is a measure of non-linearity between V and IProtective marginsZ,, = surge impedance of the incoming circuitZ,, = surge impedance of the outgoing circuitVoltage of the refracted wave,E" = E + E'(18.4)E = voltage of the incident wave (incoming wave)E' = voltage of the reflected waveE"= voltage of the refracted (transmitted) waveWhen the outgoing circuit is inductive,BIL of the equipmentProtective margin = ImDulse Drotection level (18.2)df the arrester (V,, )Reflection of the travelling wavesVoltage of the reflected wave(1 8.3)L = inductance of the circuitWhen the outgoing circuit is capacitive,C = capacitance of the circuit(18.5)(18.6)

Surge arresters: application and selection 18/625Surge transference through a transformer(i) Electrostatic transference(1 8.7a)Vi,. = voltage of surge transferenceC,, = lumped capacitance between the primary and thesecondary windingsC, = lumped capacitance of the lower voltage sideV, = prospective voltage surge on the primary side= a factor to account for the power frequency voltagealready existing when the surge occursDampened \urge transference to account for the capacitanceC of cables and terminal equipment.V = CPc,, + c, i C(ii) Electromagnetic transferenceV, ' I' (1 8.7b)( 18.8)11 = factor for power frequency voltageq = response factor of the lower voltage circuit to thearriving long-duration surgesI' = a factor that will depend upon the transformerconnectionsII = transformation ratio of the transformer ( V,/V2).Selecting the protective level of an arresterProtectihe distancev, = \',

19/627Circuit Interruptersbreaker5 (SF6) IYi637-current chopping 19/6481647rruptinz devices 191650

Circuit interrupters 19162919.1 Circuit interruptersThese are only switching mechanisms and not completebreakers. The operation of an HT interrupter, whether toclose or open a circuit, causes certain types of switchingsurges. Details of such surges and causes of theirgeneration have been given in Chapter 17. Generally,surges are phenomena of HT systems of 2.4 kV andabove.Surges that may appear on an LT system as a result oftransference from the HV side of a transformer (Section18.5.2) are different and not related to switching. Thoseon an LT system due to switching of static devices (Section6.13) are not related to the switching of the circuit but tothe static devices themselves.In this chapter we discuss the types of insulating andquenching mediums, their switching characteristics andmerits and demerits of their use. We also consider thebasic interrupting devices developed over the years, usingsuch mediums, keeping switching surges as the basiccriteria in mind. The theory of arc interruption is thesame for all switching devices. The following types ofbreakers have been developed for the purpose of switchingand they mostly relate to the HT systems, except wherenoted:I Bulk Oil Circuit Breakers (BOCBs)2 Minimum Oil or Low Oil Content Circuit Breakers(MOCBs or LOCBs)3 Air Circuit Breakers (ACBs) - generally for an LTsystem only4 Air Blast Circuit Breakers (ABCBs)5 Sulphur Hexafluoride Circuit Breakers (SF6 breakers)6 Vacuum Circuit Breakers (VCBs)It is for the user to choose the most appropriate circuitbreaker to suit requirements. application and cost. Herewe discuss briefly the philosophy of circuit interruptionand the effect of insulating and quenching mediums onthe arc extinction of these breakers. We also deal brieflywith the constructional features and application of suchbreakers. For more details one may refer to the manufacturers'catalogues and literature available on the subject.To begin our discussions, let us first have a brief reviewof switching surges discussed above.19.1.1 During a switch closureSwitching surges may develop during a closing operationjust before the contacts are able to make. It mayoccur at a stage when the gap (i.e. the dielectric) betweenthe contacts becomes incapable of withstanding theimpressed voltage and breaks down. When this occurs,it causes an arc between the contacts leading to suchsurges.Switching surges will also develop when all the threepoles of a switching device do not make at the sameinstant. (Refer to Section 17.7.2(ii)). The surges mayhave a peak value up to 3-5 p.u. at a surge frequencyof 5-100 kHz or more, depending upon the closingcircuit constants Land C. They may exist in the systemfor a very short duration of much less than even onehalf of a cycle, i.e. up to the closure of the switch only.Typical speeds of interrupters may fall in the range of1-1 .S mm/ms. Extremely steep (fast-rising) transientsurges, up to 3.5 p.u. have been noticed in certainswitching circuits, with a front time as low as 0.2 pus.19.1.2 During a switch interruptionThese surges develop when interrupting a highly inductiveor capacitive circuit, such that the current phasor lags orleads the voltage phasor by so much that up to a nearfullsystem voltage may appear across the parting contactson a current zero and cause re-ignition of the arc plasma.(Refer to Section 17.7.3). These surges may also have apeak value up to 3-5 p.u. at a surge frequency of 5-1 00kHz or more, depending upon the interrupting circuitconstants L and C. They may exist in the system for aslightly longer duration of one half to one-and-a-halfcycles (10-30 ms for a 50 Hz system), i.e. up to circuitinterruption. Extremely steep transient surges, up to 5p.u, in certain interrupting circuits. have been noticedwith a front time as low as 0.2 ps and even le\a.The phenomenon of a switching surge is related to theperformance of the switching device, i.e. its speed ofoperation and ability to quickly rebuild its dielectricstrength (deionization of the arc plasma) between theparting contacts after a current zero.19.2Theory of circuit interruptionwith different switchingmediums (theory ofdeionization)When a live circuit is interrupted, an arc is invariablyformed between the parting cbntacts. the intensity andmagnitude of which would depend upon the quantumand the quality (p.f.) of the current being interrupted.The arc, due to its excessive heat, under high pressure orvacuum (the medium in the breaker is maintained thus),forms a plasma in the medium which causes decompositionof the insulating and the quenching medium to a fewgases and vapours. The gases so formed then ionize intoelectrons and protons, which are charged particlesconducting in nature, and make the arc conducting aswell. How to disperse the heat of the arc plasma quicklyfor a successful interruption of the circuit is the theoryof arc extinction. The types of' gases produced and theirbehaviour, as a consequence of ionization of the insulatingand quenching mediums are as follows:Oil in BOCB or MOCB This decornposcs intovapourized and dissociated hydrocarbon. which in turnionizes into H2 and other gases and vapours. H2constitutes around 70% of all the gases and vapoursproduced.Air in ACB or ABCB This ionizes into N2. O2 andvapours; N? constitutes most of it.

19/630 Industrial Power Englneenng and Applications HandbookSF6in SF6 circuit breakers This ionizes into sulphurand fluorine.In a VCB This is not the vacuum but the metal of theparting contacts that becomes vapourized.The main problem of circuit breaking arises out of theformation of the arc and its prolonged extinction, whichmay delay the circuit interruption and lead to a restrikeof the arc plasma after a current zero. The basic conceptof a circuit breaking thus leads to the quickest extinctionof the arc plasma. It has caused many engineers andscientists to undertake extensive research and developmenton the subject over the past 50 years or so to find moresuitable mediums and to evolve better techniques toextinguish the arc plasma. The present-day high technology,adopted by the various manufacturers in the fieldof arc quenching, is the result of these long years ofconsistent and continuous research and development work.To achieve a quicker extinction of the arc it is imperativeto create one or more of the following conditions:1 To quench the arc plasma caused during the interruption,quickly and continuously, to ensure that bythe next current zero, the arc path is devoid of anytraces of arcing. In other words, the contact gap mustrestore its dielectric strength before the next currentzero.2 To lengthen the arc as shown in Figures 19.11 and19.12. This is an effort to render the restriking voltage(TRV) insufficient to re-establish an arc across theparting contacts after a current zero. The processincreases the resistance of the arc plasma that helpsto absorb a part of the TRV by causing a voltage dropacross the resistance so created, besides improvingthe p.f. of the interrupting circuit and thus dampeningthe restriking voltage (TRV) to far below its peakvalue by the next current zero. Dampening of TRV atimproved p.f. may be observed from curves LI and hof Figure 17.1 1.3 Splitting the arc into a number of series arcs (Figure19.1 1 ) so that the input power to the arc becomes lessthan the heat dissipated during the process of deionization.The more efficient the process of cooling, thebetter will be the chances of avoiding a restrike andachieving a quicker extinction of the arc.4 A forced interruption before a current zero, as mayoccur in an ABCB or VCB, may cause current chopping(Section 19.6, Figure 19.27) giving rise to high TRVs,is not desirable. It is therefore important that the designof the interrupting device be such that a live circuitinterrupts only at a natural current zero, as far aspossible. to avoid generation of voltage surges. Thefollowing techniques have been developed to achievethis:Use of high pressure at the arc plasma to driveaway the same.Adopting forced cooling to quench the arc plasma.Use of such constructions that can elongate arclength and reduce the concentration of ions in thearc plasma and hence enhance the dielectric strengthbetween the parting contacts.Pre-inserting a resistor in the interrupter unit tocause a voltage (TRV) drop across it and to alsoimprove the p.f. of the interrupting circuit andmaking arc extinction easy. The mechanism is madesuch that a resistance commensurate with the systemparameters and switching conditions (which the userhas to stipulate for the manufacturer) is insertedinto the switching circuit. Insertion is made throughthe interrupting mechanism immediately, say, byhalf a cycle before the contacts make or open andis shortened or disconnected immediately on closingor opening of the contacts.These techniques have been successfully implementedin interrupting devices as noted in Section 19.1.2, beingcommercially produced by various manufacturers fordifferent voltage systems and applications.The dielectric properties of different mediums atdifferent contact gaps are illustrated in Figure 19.1. Itmay be observed, that except the medium of vacuum,which has a near constant or very little rise in dielectricstrength from about a gap of 10 mm and pressure aboutTorr or less, all other mediums, even air, have anear-linear rise in their dielectric strength with the contactgap.The dielectric strength can also be enhanced with therise in pressure of the medium, except oil, which cannotbe compressed, and can be considered as having a nearconstantdielectric properties. The characteristic of air atvery low pressures is illustrated in Figure 19.2. Thebehaviour of air at very low pressures (below 1 O4 Torr)is extensively utilized in vacuum interrupters.250t 2oo5'1505sF22 loac25c-0 5 10 15 20Confact gap (mm)Figure 19.1Dielectric strength of different mediums as afunction of contact gap

1000 ,1240AtCircuit interrupters 191631A molecule of gas, undei-high temperature andpressure or vacuum.1-1 ' 1Pressure (Torr) -+760 Torr =1 bar (1 atm.)Figure 19.2 Dielectric strength of air at different pressures ata uniform contact gap of 10 mm (1 Torr = 1 mm head of Hg)by very lightweightelectronS 4Moves towards anode and allelectrons get absorbed there.The leftout neutrons comeback to plasma unchargedIMoves towards cathode, collects anelectron and gets neutralized. Theneutrons so resulted, come back tothe plasma uncharged19.3 Theory of arc plasmaThe arc plasma is caused during an interruption of thelive contacts, and also just before closing the contacts,when the contact gap falls short of the required dielectricstrength to withstand the impressed voltage. When anarc is caused the gases present in the arcing chamber,under the influence of high temperature of the arc plasmaand the high pressure or high vacuum maintained withinthe arcing chamber, become ionized. They liberate protons(positive ions, positively charged, heavier particles) andneutrons (uncharged particles) surrounded by electrons(negatively charged lighter particles (Figure 19.3)). Thetheory of arc extinction relates to the physics and behaviourof these electrically charged particles that are responsiblefor a restrike of the TRV even after a current zero. Theeffectiveness of the medium and the design of the arcchamber to diffuse these electrically charged particles toneutrons as quickly as possible determines the ability ofone type of breaker over others. In fact, the theory of arcextinction is the theory of deionization (neutralization)of the electrically charged protons and electrons. Thetheory may be briefly explained as follows.The positive ions (protons) present in the arc plasmamove to the cathode (negative pole) and neutralize(deionize) the electrons there. Similarly, the negative ions(electrons) move to the anode (positive pole) and becomeabsorbed there. Thus the process continues until all theplasma is neutralized and contains only neutrons toextinguish the arc. In fact, the concentration of ionsbetween the parting contacts gradually becomes dilutedby the deionization of protons and the absorption ofelectrons at the anode, and a stage is reached whenadequate dielectric strength between the parting contactsis restored to extinguish the arc. It is not necessary forall the ions present in the plasma to be deionized toextinguish the arc, rather a stage when they are not ableto hold the arc. This process also increases the resistanceof the arc plasma as a result of reduced contact pressureand arc contact area.Note: The mass of a nucleon (proton or neutron) IS 1835 timesheavier than an electron and moves much slower than an electronsince, m,v: = rn,v: = constant. Where rn IS the mass and v, thevelocity of an electron or a proton. The bulk of the arc is thereforecaused by electrons rather than protons.Figure 19.3 Theory of ionization and deionization of gasatoms to extinguish the arc plasmaA proton, being 183.5 times heavier than an electron,moves sluggishly compared to an electron sincem,uf = rn2ui = constantwhere m is the mass of an ion and u its velocity. Then,for the mass of electron as rn, the velocity of the protonwill beIf Vp = velocity of the proton andV, = velocity of the electronthen Vp = ~ -- which is too slow compared to@mi- 43'an electron. The conductance of the arc plasma is thusthe result of the movement of electrons, rather than protons.which contribute only a small amount.19.4 Circuit breaking underu nfavoura ble operatingconditionsLong years of experience in the field of circuit breakingwith interrupting devices have revealed that under adverseconditions of circuit parameters, interruption may not besmooth. It may result in excessive voltage surges, as aconsequence of restriking of the parting contacts. A wrong

19/632 Industrial Power Engineering and Applications Handbookchoice of interrupting device may result in insulationfailure of the terminal equipment, such as a powertransformer, an induction motor or interconnecting cables.This situation may arise when:1 Interrupting small magnetizing currents, such as interruptingan induction motor or a transformer on noload,a situation, when the current may lag the impressedvoltage by nearly 90".2 Interrupting a charged capacitor bank, when the currentwill lead the impressed voltage by nearly 90".3 Interrupting an unloaded transmission or distributionline or a cable, Le. interrupting a line charging current,which is capacitive and may lead the system voltageby nearly 90".4 Interrupting an induction motor immediately after aswitch on, when the current is large and highly inductive.5 Interrupting fault currents that are mostly inductive(Section 13.4.1) and occur at very low power factors.They are excessive in magnitude, and cause highthermal effects and electromagnetic* forces on thearc chamber, the contacts and the contact mountingsupports.Under the above conditions, the arc, as usual, willextinguish at the first current zero but will have a tendencyto re-establish immediately again, after the current zero(Figure 17.1 l(c)) while the contacts are still parting. Thisis because the TRV across the parting contacts may exceedthe dielectric strength of the contact gap achieved so far.Restoration of the dielectric strength will depend uponthe speed of the moving contact and the insulating mediumof the arc chamber. There may be a number of restrikesbefore a final extinction is achieved. The frequency ofrestrikes may be extremely high (equation (17. l)),depending upon the L and C of the interrupting circuit,which would have the characteristics of a surge circuiton formation of an arc. In terms of actual rated frequency(f), restoration of the dielectric strength may not takemore than one half to two cycles, i.e. 10-40 ms (for a 50Hz system). The behaviour of circuit breaking thus dependsupon the design and the quenching medium of theinterrupting device.19.5 Circuit interruption in differentmediums19.5.1 Bulk Oil Circuit Breakers (BOCBs)Refer to the general arrangement of this breaker in Figure19.4. In this device the moving contacts make and breakin an oil bath. When the arc is formed during aninterruption, the oil becomes decomposed due to excessiveheat, and produces a few gases and vapours such as H2* The breaker will interrupt only during a transient state (Figure13.20) by which time the d.c. component responsible for the dynamicforces, has subsided.-Tulipcontacts. TerminalbushingsArc controlpotsMovingcontactsOil tankFigure 19.4 Rear view of a bulk oil circuit breaker assemblyshowing single-break contacts, self aligning cluster isolatingcontacts, terminal bushing and arc control pots (Courtesy:GEC Alsthom)(70%), C2H4 (20%), CH4 (10%) and free carbon, say, 3g per 10 litres of oil decomposed at a very high pressureof 100-150 bars, in the shape of a bubble around the arc(Figure 19.5). H2 is an extremely good medium forquenching and does most of the cooling of the arc plasma,exting-uishing it while passing through it. The gases thusproduced also cause turbulence in the oil in the neighbourhood,causing rapid replacement of the oil with cool oilfrom around the contacts, thus achieving a double coolingeffect. At each current zero, it almost recovers its dielectricstrength and also increases its post-arc resistance as aresult of cooling and arc extinction, making the interruptionall the more easier and complete.Simultaneously the bubble also pushes the oil awayfrom around it and reduces the cooling. Proper design,however, can ensure adequate cooling during interruption(arcing) by adjusting the speed of the parting contact,supplementing the cooling of oil through an additionaloil chamber, such as a side-vented explosion pot or crossjet pot, by adjusting the gap between the fixed and themoving contacts.While breaking smaller currents, the formation of gasmay not be adequate to provide the desired cooling effect.This is, however, immaterial because of less intensivearc formation requiring much less cooling. Extinctionmay be slightly prolonged but may be achieved by thenext current zero.

00 Cb8002 m3OOE a9OOP 100s009

19/634 Industrial Power Engineering and Applications HandbookOil level indicatorVent liberated gases in the breakerOil filling ventUpper pole headContact tubeConnecting terminalsSpring supported tulip contacts(fixed contacts)Arc quenching deviceArc chamberMoving contact rodContact roller guideConnecting terminalsContact rollersLower pole headFigure 19.7 Minimum oil content circuit breaker (MOCB)(Courtesy: NGEF Ltd.)breaker during the late 1960s and onwards. Their rupturingcapacity is also much higher than that of a BOCB andthey are extremely suitable for distribution systems withmoderate fault levels. The trend, however, has tilted infavour of more advanced technologies, now available inthe form of vacuum or SF, breakers. MOCBs are availablefrom 6 kV to 420 kV and have a rupturing capacity of250-25 000 MVA.19.5.3 Air Circuit Breakers (ACBs)Refer to the general arrangement of this breaker in Figure19.9(a), and (b).The moving contacts make and break in air as shownin Figure 19.10. During interruption, the arc is formed(Figure 19.11) producing N2 (80%) and O2 (20%) andmetallic vapours. The quenching and extinction of arcplasma is achieved through the elongation of arc, whichincreases the area of cooling, on the one hand and requiresa higher TRV to cause a restrike, on the other. To obtainthis, arc chutes are provided on the top of the interruptingcontacts, as illustrated in Figures 19.12(a) and (b). Thedesign of the arc chutes is such that it drives the arcplasma upwards and elongates it to provide the requiredcooling effect. This is achieved by constructing the arcchute housing of a non-magnetic material, such as glass,asbestos, ceramic or Bakelite, suitable to withstand thevery high temperature of the arc plasma. Metallic arcsplitter plates (fins) are fixed inside the arc chute housingOil drain plugLevel connecting operatingmechanismFigure 19.8 Cross-sectional view of a typical pole assembly ofan 11 kV MOCB (Courtesy: NGEF Ltd.)so that the arc plasma produces a magnetic field throughthese splitters and rises upwards, and splits into a numberof shorter arcs, to lose all its heat through convection.This renders the TRV insufficient to cause a restrike.The long arc length and subsequent cooling increasesthe resistance of the arc plasma and improves the p.f. ofthe interrupting circuit. It thus helps to bring the currentphasor closer to the voltage, and make interruption on acurrent zero less severe as a result of low TRV. (Refer tocurves a and b of Figure 17.1 1 .) The gradual rise of arcresistance after a current zero dampens the TRV andmakes such breakers almost immune to switching surges.For higher currents, the arc splitter plates may be alteredto have a variety of designs, such as with offset slots,serpentine splitter plates or similar features to effectivelyarrest the arc plasma within the arc chutes, rendering itincapable of causing a restrike after a current zero.Such breakers are normally produced for use on an LTsystem only. At higher voltages, while interrupting heavycurrents (such as on a fault) the arc energy may be sohigh that a disproportionate size of arc chutes may berequired to arrest and extinguish the arc, leading todisproportionate size of ACB.

Circuit interrupters 19/635(Courtesy Siemens)Figure 19.9(a) Views of air circuit breakers(Courtesy: GE Power Controls)(Courtesy: Alsthom)Figure 19.9(b) ACB in housingFigure 19.10 Typical contact arrangement of an LT air circuitbreakerACBs were the first to be produced commercially.They are simple to operate and cause no fire hazards.But at atmospheric pressure, they possess a low dielectricstrength and are therefore normally manufactured onlyin low voltages. Air has less contamination and thereforethese breakers require negligible maintenance, comparedto oil. They require no contact cleaning. Since there is alimit to producing these breakers for HT systems, theirnormal application is for LT systems alone, where theyare used extensively. In fact, they are the only breakersto meet the needs of an LT distribution system.19.5.4 Air Blast Circuit Breakers (ABCBs)Refer to the general arrangement of this breaker in Figure19.1 3( a).These are similar to ACBs, except that the process ofinterruption is accelerated by impinging a high pressure

19/636 Industrial Power Engineering and Applications HandbookSplitter orquenching platesArc plasma risingto the arc splitters- Housing of nonmagnetic material(Asbestos, bakelite,FRP or DMC etcLengthening ofarcArc movinginto arc chuteFixed & movingarcing contactsFigure 19.11 Process of arc formation and quenching in anACB using splitter platesaxial air blast through the arc plasma, when the contactshave just begun to separate. (See Figure 19.13(b).) Thecompressed air has greater dielectric strength and thermalproperties than ordinary air at atmospheric pressure.The conventional pressure of air blast is generally 218to 900 lb/in2 (1.5 to 6.2 MN/m2) up to 110 kV, up to3000 lb/in2 (20.7 MN/m2 (for voltages up to 400 kV, andstill higher pressures for higher voltages. The pressurerequirement will change from one manufacturer to another,depending upon contact design, interrupting mechanismand design and value of resistors. Since most quenchingis through a predetermined force of an air blast, the forceof arc plasma quenching and extinction (deionization)remains the same for a particular size of breaker,irrespective of the amount of current the interruptingdevice may have to break. This factor inherits a tendencyto break small currents, even before their natural currentzeros, causing current chopping (Section 19.6). Currentchopping may raise the TRV up to 2.5-3 p.u. However,it is possible to design these breakers to contain the valueof a TRV to a non-striking level.The dampening of a TRV is achieved through lowandhigh-resistance units provided across the contacts.The low unit will short at higher TRVs and the higher unitat lower TRVs. The arc interruption is fast and generallyat the first current zero due to dampening. When thebreaker is in the closed position, the resistors are opencircuited. As soon as the main contacts begin to interrupt,the contacts of resistor make first, before the main contactsseparate and provide the required dampening. Afterextinction of the arc, the resistor circuit opens againautomatically and restores to the original position.Until a few years ago these breakers had been quitecommon for medium voltages, up to 33 kV. Since theyrequire a powerful blast of air at high pressure and velocityinto the arcing region, they require a reliable source ofair supply. Air should be clean and dry and at the correcta Top terminalb Moulded-plastic baseb5 Fixed contactd Arc chutedZ Arc runnerd4 Arc splittersd5 Leaf springel Moving contacte2 Contact carriere, Tension springe, Flexible connectore9 Compression springf2 Operating shaftg1 Insulating barrierg2 Insulating barrierh4 Bimetal striph, Intermediate shafths Current transformern3 Intermediate shaftn4 Magnet corep Bottom terminalFigure 19.12(a) Operating mechanism of an LTACB showingthe arc chute with splitter plates (Courtesy: Siemens)pressure and volume at all times. This requires an elaboratearrangement for air compression, an air storage facility,a network of feed pipes, valves and safety devices besidestheir regular maintenance and upkeep. All this is costlyparticularly when only a few breakers are at a particularinstallation.Moreover, compressed air is released through an orificeat the exhaust point at a high velocity and causes a soundlike thunder. This may be frightening to people in thevicinity. Sub-stations involving a number of such breakersare a nuisance to residents nearby. The larger the breaker

Arc probeCircuit interrupters 19/637,/TerminalArc chute with arcsplitter platesArc in initial andfinal positionsLow PressureHigh pressureFigure 19.12(b) Arc chamber with splitter plates in a powercontactorPiston -Air flowTerminalInternal electricalconnectionControl valve andblast valveControlinsulatortBlast tubeFigure 19.13(b) Process of arc formation and quenching in anair blast circuit breakerCompressed air tankPneumaticDouble AinterruptingchamberInterrupting chamberdriving mechanismFigure 19.13(a) One pole of ABCB 72.5-420 kV with verticalcompressed air tank (Courtesy: ABB)and the higher the kV of the system, the greater thepressure of the air blast and its sound. Silencers, however,are provided to contain such sound hazards but they aremore appropriate for large installations which require alarge number to be installed in the same system toeconomize on the compressed air supply arrangement.19.5.5 Sulphur hexafluoride gas circuitbreakers (SF,)Refer to general arrangements of such breakers in differentratings as shown in Figures 19.14-19.16.This is the latest technology in the field of arc extinction.It was introduced in the 1960s and attempts to achieve ahigh dielectric strength between the contacts. At roomtemperature SF6 is a chemically inert, non-toxic and noninflammable,colourless, odourless gas, having a molecularweight of 146 and provides excellent arc quenching as aresult of electronegative behaviour.At atmospheric pressure, its dielectric strength is twoto three times that of air, as illustrated in Figure 19.1,and its arc-quenching ability many times more than air.This gas undergoes no chemical change at high temperatures,except small decomposition into SF2 and SF, gasesand some metallic fluorine in the form of an insulatingpowder while interrupting and quenching an arc. Thesegases and powder, however, are readily absorbed byactivated alumina placed in the filters in the closed-loopcircuit of the gas, as discussed later. The gas cycle issuch that after every interruption the consumed gas isreplenished through a reservoir filled with SF6 gas at ahigh pressure, say, sixteen times that of the atmosphereand connected to the main interrupting chamber through

19/638 Industrial Power Engineering and Applications HandbookFigure 19.14 11 kV SF6 circuit breaker (Courtesy Voltas)pressure valves and filters. As soon as the pressure in theinterrupting chamber falls below a pre-set value, the valvein the reservoir opens and builds up the lost pressure.Due to the very high pressure in the reservoir, comparedto only almost three times the atmosphere in the chamber,it is possible that the pressure inside the interruptingchamber may sometimes exceed the required value. Inthe interrupting chamber, therefore, are also providedhigh-pressure release valves to pump the excess gas backto the reservoir through a compressor and a filter. Thetotal gas circuit is a closed cycle without any venting tothe atmosphere.This gas is electronegative and its molecules quicklyabsorb the free electrons in the arc path between thecontacts to form negatively charged ions. This apparenttrapping of the electrons results in a rapid build-up ofdielectric strength after a current zero. The detailedsequence of arc extinction may be summarized as follows.The contacts begin to compress a quantity of SF6 gasas soon as they start opening. This opening also causesarc plasma between the contacts. The temperature of thearc plasma ionizes the gas into sulphur and fluorine atomsand quickly becomes quenched through the turbulence ofthe compressed gas through a very strange process ofnegative ion formation. At higher temperatures, the Satoms become ionized into S+ protons and SN neutrons.The S" electrons of the neutrons are immediately absorbedby the fluorine atoms to form fluorine ions (F) which areheavy and are sluggish. They contribute little to maintaining(a) Exploded view of a pole(b) Breaker in a draw-out positionFigure 19.15 General arrangement of a 3-12 kV, SF, circuit breaker in a housing (Courtesy: Voltas)

Circuit interrupters 19/6391. Interrupter unit 5. Operating mechanism2. Multi-shed insulator 6. Coupler3. Base tube 7. Operating rod4. Hydraulic storage cylinder 8. Control unit(a)Figure 19.16(a) An SF6 circuit breaker 123-145 kV, 31.5 kA. Itcan also be pneumatic or spring operated, depending upon thearc quenching technique adopted and energy required to extinguishthe arc (Courtesy: BHEL)the conductivity of the arc plasma. (See also Section19.3.) This quickly immunizes the free electrons, restoresits dielectric strength, quickly quenches the arc plasma,extinguishes the arc and builds up the dielectric strengthafter a current zero. After a current zero, the processquickly, quenches the arc in the beginning itself bysweeping away the arc plasma, thus improving the dielectricstrength between the parting contacts and achievingsuccessful extinction of the arc. The arc extinction processmay be slightly delayed when the contacts open veryclose to the next current zero, and the quenching mediumblows it out with force, before the current zero, leadingto a case of current chopping. But with continuousimprovement in the techniques of arc extinction, it hasbeen possible to achieve an interruption devoid of acurrentchopping or a restrike of the arc plasma.As noted above, to quench the arc plasma successfullyit is essential to create a turbulence in the SF6 gas aroundthe arc plasma to destabilize and blow it out. This can beachieved in two ways:1 Puffer technique (a puffer identifies the exhaling ofgas forcibly by compression)2 Rotating arc technique3 Arc blast and arc assistance technique.1 - Breaking chamber2 - Central housing3 - insulating column4 - Insulating rod5 - Opening spring6 - Lower housing7 - SF6 monitoring block8 - Operating mechanism9 - Closing springFigure 19.16(b)SF6 circuit breakers 300-500 kV (Courtesy: Alstom)

19/640 Industrial Power Engineering and Applications Handbook100%50%33%25%Puffer The r m a I Arc Doubleblast assisted volumeFigure 19.16(c) Comparison of energy requirement for arcextinction in an SF, breaker, using different techniques(Courtesy: Alstorn)Puffer techniqueDestabilization of the arc plasma is achieved by forcedconvection of gas created by the movement of the mainand arcing contacts through a puffer piston. This is anintegral part of the moving main and arcing contacts(both being concentric). In the light of more advancedtechniques of arc extinction now available, the manufactureof such breaker is now limited to about 145 kV.Sequence of arc quenchingRefer to Figure 19.17. On a trip signal the main movingcontacts start separating alittle ahead of the moving arcingcontact and compress gas through the puffer piston insidethe tubular chamber. On separation the main fixed andmoving contacts are transferred to the fixed and the movingarcing contacts as shown. As soon as the moving arcingcontact starts separating, an arc is formed between thefixed and the moving arcing contacts and the alreadymoderately compressed gas is compressed further. Thiscompressed gas is impinged with full force through theblast nozzle (Figure 19.18) at right angles to the arc plasmafrom all sides to achieve instant destabilization of the arc.Through radiation also, the arc plasma dissipates apart of its heat which supplements the quenching. Butthis is too meagre a contribution, as heat dissipation occursonly through the outer surface of the arc plasma.Nevertheless, it is the major cause of gas impedimentgiving rise to the phenomenon of clogging, discussedlater, and which helps in ai extinction.Pole in closed position. Maiand arcing contacts closecPole at the moment ofseparation of arching contacPole after arcquenching@- Fixed arcing contact0- Tubular gas (blast) chamber0 - Puffer piston@ - Main fixed contact@ - Main moving contact@ - Fixed contact assembly@ -Arc chute@ - Moving contact assemblyFigure 19.17 Sequence of arc extinction through the puffer technique in an SF6 breaker through the cross-section of a pole(Courtesy: NGEF Ltd.)

Circuit interrupters 1916411. Fibre glass arc chamber tube2. Lower pole head3. Cap4. Gas filling valve5. Main fixed contact6. Fixed arcing contact7. Moving contact arrangement8. Moving arcing contact9. Blast nozzleIO. Expansion chamber11. Lower contact12. Lower terminal13. Top terminal14. Spline shaft lever15. Switching lever16. Two level pressure switch-1 st level contact for alarm-2nd level contact for trip and lockout17. Enclosure for alumina18. Explosion proof safety valveFigure 19.18 Cross-sectional view of a pole of a 12-36 kV SF,breaker (Courtesy: NGEF Ltd.)The events are so fine-tuned and the size of chamber,pressure of gas, travel, distance of the moving contactand the size of blast nozzle so designed and adjusted thata near-strike-free interruption can be achieved for lowreactive currents (inductive or capacitive) as well as fullloadand very heavy fault currents. The advance compressionof gas through the movement of main contactplays an important role by storing a part of the gas evenbefore opening of the arcing contacts.At high instantaneous currents the arc may occupymost of the contact area between the arcing contacts andmay impede the flow of gas through the arc plasma. Thisphenomenon is termed the clogging effect, but it assistsarc extinction in the following manner.The gas around the arc plasma takes away a part of itsheat by radiation. At high temperatures, the gas loses itsspecific gravity, becomes light weight and diminishes inmomentum (- mu2). As a result, the gas is renderedincapable of penetrating through the arc plasma to quenchit. The flow of gas through the thick of the arc plasma isthus impeded.As the moving contact moves away, so the arc plasmaelongates, losing its initial intensity, and as it approachesthe current zero, it loses the most of it. The gas, on theother hand, cools and regains its lost mass, while itspressure in the chamber continues to build to its optimumlevel, making it more capable of extinguishing a lesssevere arc plasma. The interrupter can thus be adjustedto blow out the arc at the first current zero, while clearingheavy to very heavy fault currents.Similarly, at lower currents, the volume of arc plasmais too small (- Z2) and so is the clogging effect. Thepressure and volume of the quenching gas can be adjustedto interrupt the current now also at current zero. All theseadjustments are pre-set and sealed by the manu-facturer.Since the arc extinction technique is highly effectiveand quick and occurs when the arcing contact is stillmoving, arc length and hence contact travel, can bereduced as can the arc energy and the excessive heatingas well as erosion of the arcing contacts. An extendedcontact life can thus be achieved by this technique.In lower voltage ratings, as noted above, the puffertechnique is quite prevalent and is adopted by all leadingmanufacturers. For constructional details and moreinformation on this mechanism, refer to the manufacturers’catalogues. Figure 19.18 illustrates a typical pole assemblyof a 12 kV, SF6 circuit breaker.Rotating arc technologyThe process of arc quenching can be enhanced byincreasing the turbulence by ionizing more gas atoms, toincrease the gas pressure and therefore turbulence. Thisis possible by bringing more gas into contact with thearc plasma, and this can be achieved by rotating the arcplasma between the contacts and displacing the arc by amagnetic blow-out. A general method of achieving thisis by providing an electromagnetic field around the fixedcontact, which gives the required rotating motion to thearc plasma, when the moving contact moves away fromthe fixed contact, as illustrated in Figure 19.19, similarto a motoring action. Figure 19.20 illustrates one pole of

19/642 Industrial Power Engineering and Applications Handbooksuch a breaker, provided with a magnetic coil. The breakersbased on this principle are known as rotating arc circuitbreakers. Figure 19.21 illustrates the rotating arc formationand the direction of a magnetic field during an interruption.As the arc is made rotating over the arcing contacts, theheating and thus the erosion of the contacts is low inthese breakers, and they have an extended contact life.This technique, although good, is cumbersome and istherefore generally not practised now by the manufacturers.Instead, some have improvized the puffertechnique itself to assist and smooth the arc-quenchingprocess. This they have achieved by optimizing the useof arcing heat through the thermal blast and arc assistancetechnique.Thermal blast and arc assistance techniqueThe design of the arc chamber is improvized to augmentthe arc-quenching capability of the arcing chamber, byfurther compressing the gas that had already expandedduring arcing and impinging this on the arc with a greaterSuccessive arcpositionsContact barField coil assembly(Replica of left-hand rule, Figure 1.1)Figure 19.19 Making the arc rotate in a magnetic fieldCylindrical’ electrode0-0-The arrows indicate the direction of magnetic forcesMagnetic forces on the spiral arcPole duringarc quenching@ Fixed contact assembly @ Arc chute@ Magnetic field coil @ Moving arcing contact@ Fixed arcing contact @ ArcIFigure 19.20 Typical design of one pole of a rotating arc SF6circuit breakerFigure 19.21 The spiral arc in a rotating arc SF6 circuit breaker(Courtesy: South Wales Switchgear Ltd.)

Circuit interruDters 191643blayt. The blast also helps the main moving contacts toinow farther away and thus reduce the energy requirementbl the moving mechanism to interrupt the breaker. Thismakes the whole process of arc extinction easy and smooth.This technique. instead of equalizing the arc heat, reducesthe arc energy itself, facilitating a quicker and smootherextinction of the arc. The moving tnechanim that in anormal puffer is usually hydraulically or pneumaticallyoperated (Figure 19.16(a)) due to the higher energyrequirement by the moving mechanism can now beachieved through a simple spring mechanistn (Figurel9.l6(b)).Figure 19,16(c) illustrates a comparison of energyrequirements between the various techniques in practice.Future technology on which some manuf. acturera arealready working may be in the form of a double volumetechnique. In which an attempt is being made to makethe I'ixed main and arcing contacts also moving. Thisodd enhance gas compression as well as the separation01' the arcing (interrupting) contacts, and further reducethe arc energy requirement, making arc extinction easier.smoother and quicker. The approximate improvised energyrequire-inent is shown in Figure 19.16(c).As SF6 breakers can be designed to provide a verysmooth interruption of an arc, devoid of current choppingor ;I restrike of the arc plasma by accurately controllingthe supply of gas to the required level of cooling. theyma! also be termed soft break interrupters. SF, breakersare the most extensively used and are suitable forpractically all applications and voltage systems up to765 kV and more. The other advantage with SF,switchgear is a space saving of up to 70-90'3 over theconventional type of switchgears. Since these breakersare totally enclosed and sealed from the atmosphere,the! are also the most recommended choice for a11installation\ that are ha,mdous and prone to explosions.Pwiiiwrtioii resistoi.These breakers, when used for switching long transmissionline\ at 120 kV and above, are pro\,ided with a preinsertionresistor across each interrupting contact to limito\ ervoltages that may occur during a closing or openingsequence. as a result of heavy charging currents as notedin Table 24.2. The value of the resistor may be around400 R for the line parameters, considered in Table 24.2and may \ary with line parameters. The resistors areconnected so that during a closing sequence they shortcircuitthe making contacts before closing the maincontact\ for. say. 8-1 0 milliseconds. and open immediatelyafter the contacts are made. This also happens during anopening wquence.I,ow gas velocity and pressure minimizes the tendencytowards current chopping.A closed recycling of gas causes no noise or contatninationof the atmosphere.There is no carbonization and therefore no tracking.(conduction of the insulating medium).Because of the extremely good dielectric propertiesof SF, yas. the arc gap and the contact travel and56hence the arcing time are lob. requiring lex\ energyto interrupt (Figure 19.6).As the arcing time is very low. it causes no or only :Ismall amount of contact erosion.It is highly suitable for hazardous locations.19.5.6 Vacuum circuit breakers (VCRs)Refer to the general arrangement ofa loow breaker xhou tiin Figure 19.22 and its housing (Figure 19.23).The electrical breaking capacity in vacuum has beenlong known. But it was not until 1970 that it wii\ used inthe making and breaking of currents at high voltages. Ithas been a widely recognized and accepted breaker. lea\ ingbehind all other techniques of arc breaking and extinctionin its voltage range. In vacuum. a IO mm gap at aboutl/IOhmm vacuum of mercury is capable ofwithstandinga peak voltage up to around 240 kV (Figure 19. I ). Thesebreakers require no maintenance and are very compact.Unlike other mediums, the dielectric strength of avacuum increases with a gap. but only marginally. whichis the limiting factor in producing such breakers beyond36 kV. These breakers are therefore used only for niediunvoltagesystems (2.4-36 kV). Some inanufacturcn haveattempted to produce them up to 66 kV but they have notshown the desired results so far. The application of thesebreakers therefore continues to be up to 36 kV onl!.A comparison of dielectric strength of high wciiutnwith the other available mediums is shown in Figure19.1. The very high dielectric strength 01' vactiiini makesit possible to quench an arc with a ver) small contactgap and breakers with very coinpact di mensions can bedesigned.Because of the low contact gap. lo\\ arc resistance andfast clearance. the arc energy dissipated in vacuum for :Iparticular current is 1/10 that of oil and l/1 that of SF,.and is illustrated in Figure 19.6.Vacuum is finally judged to be the best medium toquench the arc plasma and interrupt ;I circuit under themost adverse conditions. Figure 19.23 gives cross-\ectionalviews of one pole of a vacuum circuit breakcr and atypical construction of the arcing contact\ and Figure19.25 shows its assemblq.AdvantagesSome advantages of vacuum circuit breakers arc sutiimarizedbelow:I They have a lungcr life span. during which they donot deteriorate or lose their dielectric properties.2 They require extremely low niainteiiance.3 At lower currents. say, up to I kA, the inaxitnuinduration of arc even at low p.f.s. is of the order of justonc-half to one cycle of the natural frequency of thesystem. as against nearly twoThe low exciting currents, at low p.f.s. are moredifficult to interrupt rather than large current\ at highp.f.s. due to an extremely adverse wltage-currentphasor disposition. For more clarity rcfer to Section17.62. The current now is nearly 90" lagging theapplied voltage and the TRV approaches ;I full applied

19/644 Industrial Power Engineering and Applications Handbook11 kV vacuum contactor with HRC fuses (Courtesy: Joyti Ltd.)7.2-36 kV vacuum circuit breaker (Courtesy: Siemens)Figure 19.22voltage and hence a tendency to cause a restrike. AVCB is devoid of a restrike after a current zero, asexplained later.Thus they have an extremely low energy requirementto actuate the operating mechanism and an equallyshort breaking time.They cause no fire hazard.They make no noise.They do not emit any gases.They are the only devices that are independent of theoperating system, as the breaking capacity is dependentmainly on the material and contour of the contactstructure and the quality of the vacuum.DisadvantagesVacuum breakers have a few disadvantages as noted below:1 They may inherit current chopping tendencies at verylow currents of 3-5 A, varying from one manufacturerto another and depending upon the contact materialused. This is due to their extremely fast operation asa result of a high vacuum pressure of the order ofTorr (1.333 x lo4 N/m2) or more (one Torr beingthe pressure equivalent to hold a column of mercury1 mm high). Thus they cause a high TRV, particularlywhen interrupting a highly inductive or capacitivecircuit (Figure 17.1 l(d)).Very high vacuum may have a tendency to cold weldingof the making contacts. Two pure metals, when joinedhave a tendency to stick together under high vacuum.This phenomenon is termed cold welding. The contactson closing may require a lot of force to separate themwhich may prove to be detrimental in clearing a faultpromptly.At no load, opening of contacts may lead to rougheningof the surface, due to the breaking of the cold weldingthat takes place.They also have a tendency to melting and welding ofthe contacts while making or carrying large currents.However, this is overcome by suitably designing andcontouring the contacts so that the arc impinges overa large area of contacts rather than at one point onlyto prevent melting of contacts. The material of thecontact is chosen so that it will produce less gas content,have good anti-weld properties and low currentchopping tendencies (high contact resistance). Thenormal metal alloys in use are:Low resistance-high kA alloy (high melting point):copper-bismuth has a good resistance to cold

~~ stemCircuit interruDters 19/645Fixed contact stemExplosion covers1 Met!t;ng ~relay chamberFixed terminal padGlass ceramicGlass ceramicMetallic bellowsMoving contactMoving contactFigure 19.24 Sectional view of a 12 kV up to 2500 A. 40 kAvacuum interrupter (Courtesy: BHEL Ltd.)Figure 19.23 General arrangement of a 7.2-36 kV vacuumcircuit breaker in a housing (Courtesy: Siemens)welding but has a higher probability of currentchopping. Refer to curve 4 of Figure 17.8.High resistance-low kA alloy (low melting point):copper-chromium (CLR) which also has a goodresistance to cold welding and a lower probabilityof current chopping, similar to in OCBs. Refer tocurve 2 of Figure 17.8.5 Since a very small gap in vacuum can withstand axry high voltage, a larger gap than required will notincrease its dielectric strength. This is the limitingfactor for a VCB to exceed a certain voltage system,presently 36 kV.6 They may cause contamination of the vacuum due tocas produced by arcing.7 They may lead to deterioration of the insulation ofthe insulating container due to condensing ofthe metalvapour on the inner surface of the container (more intrans\erse magnetic field type breakers).The theory of arc extinction, as related to vacuum, istypical. No arc can take place in the absence of a gas.The molecules of the gas alone under heat and pressurewill cause to ionization. responsible for the arc plasmaand subsequent deionization, which extinguishes it. Invacuum. the content of gas is missing. In fact it shouldhave been an ideal condition to interrupt a circuit withoutthe formation of an arc and thus make the interruptiondevoid of high TRVs and the phenomenon of arc restrikes.But it is not so, as the heat generated at the partingcontacts causes boiling of the contact material (generallyalloy of copper as mentioned above). This boiling producesmetal vapour, usually of copper atoms (copper, of all theother alloy metals, has the lowest melting point). Mostof the metalized vapour is thus formed of copper atomsonly. An electric field within the contacts quickly generatesfree electrons of this metal vapour and a constrictedlocalized plasma is established. Beyond a certain currentvalue, the behaviour of the arc is suddenly modified andthe constricted form of the arc plasma transforms to adiffused form. The cathode spot becomes di\ided intoseveral very small spots, which then move very rapidly.repelling each other continually. This phenomenon isused in current breaking in vacuum. In other mediumsand conventional interrupters. the current maintains onlya single arc column. These spots have an extremely highcurrent density which can reach millions of amperes persquare centimetre. The result is that very high densitystreams of electrons are emitted without a commensuratequantity of metal vapour. As the current falls to zero, atthe next current zero the metal vapour solidifies. leavingbehind no medium to hold the arc and the electrons ceaseto cross the contact gap. The dielectric strength reachesits maximum. The anode. being cool is no longer able toemit more electrons, hence it is not able to restrike aftera current zero. Arc extinction in such a medium is thereforeextremely quick. The arc plasma depends largely uponthe alloy being used as the contact matcrial. It i$ of vital

191646 Industrial Power Engineering and Applications HandbookI1. Upper interrupter support2. Top terminal3,7. Glass ceramic housing4. Arc chamber5. Fixed contact6. Moving contact8. Metallic bellows9. Moving contact stem10. Bottom terminal11. Lower interrupter support12 Lever13. Insulated coupler14. Contact pressure spring15. Release pawlFigure 19.25 A pole assembly of a vacuum circuit breakerimportance to limit the excessive boiling of the contactsdue to the arc heat. It is possible to achieve this bysuitably designing the contour of the contacts to increasetheir area. Depending upon the design of the contacts’contours, the breaker may beAxial magnetic field type orTransverse magnetic field typeIn axial magnetic field type the shape of the contactsmay be as shown in Figure 19.26(a). With this design ofcontacts the arc plasma will spread out axially and increasethe contact area whereas in transverse magnetic fieldtype breakers the contacts are like spiral slits in the formof petals. The design of contacts causes the current toflow radially outward along the contact, and gives it arotational movement under the influence ofelectrodynamic forces (similar to a rotating arc SF,interrupting device, Section 19.5.5 and Figure 19.21).The rotational movement adds to the contact area andprotects the contacts from damage and a reduced life.But in this case the arc will fall perpendicularly to themagnetic field. It is possible that it may impinge on theinside insulating lining of the contact chamber and rupturethe interrupter. Axial magnetic field type contacts aretherefore generally adopted by manufacturers.19.6 Current choppingWith advances in technology in the field of circuitinterruption, fast to extremely fast interrupting deviceshave been developed, aided by high-performing arcquenching and extinguishing mediums, as discussed above.While such techniques have helped in the interruption ofsystem currents, particularly on faults (at very low p.f.s),they have also posed some problems in certain types ofcircuit breaking. For instance, an air blast circuit breakerand a vacuum circuit breaker are both extremely fastoperating. When interrupting on a fault, their operationis as desired but at much lower currents than rated, suchas at no-load, they may operate rather faster than desiredand interrupt the circuit before a natural current zero.Premature interruption of a circuit such as this is termedcurrent chopping and may occur just before a naturalcurrent zero when the current is small. In a VCB it is ofthe order of 3-5 A.When the p.f. of the interrupting circuit is low, aswhen interrupting an induction motor or a transformer,running on no-load and drawing a small but highlyinductive current, and when interrupting a highlycapacitive circuit, such as a live but unloaded cable oroverhead line carrying a high capacitive charging current,in all such cases during a circuit interruption the current

Circuit interrupters 19/647level in the arc chamber of the interrupting device, thearc may restrike again and cause higher TRVs. This shallfurther complicates the process of interruption andextinction of the arc. See Blower et al. (1979), Telanderet al. (1986) and IEEE transactions (1977).Arcing with contrate or slotted cup contact(a) Slotted cup type contacts of a 7.2 kV, 25 kA vacuum interrupter(Axial magnetic field type) (Courtesy: Siemens)NoteThe surge frequency at which the TRV will restrike will be extremelyhigh. It may be of the order of 1ClOO kHz, depending upon thecircuit constants L and C. To interrupt such high-frequency currentsis difficult for an ordinary breaker. But with the use of hightechnologies, such as adopted in SF6 and VCB interrupting devices,which make them fast operating (for arcing time see, Table 19.1),it is possible to interrupt, such high-frequency TRVs promptly.19.6.1 Influence of frequency on the systemAn a.c. current waveform passes through a natural zeroevery one half of a cycle. This is a highly redeemingfactor in an a.c. system. It helps to extinguish an arcpromptly, which is not so in a d.c. system. A higherpower frequency than rated would in fact support theextinction of an arc, irrespective of its other magnetizingeffects, while a lower power frequency than rated willdelay and add to the complications of extinguishing anarc. At surge frequencies the situation becomes different,as the zerospoccur so frequently that the contacts areill1Ill1TRV = 2.5 pu. 1Contacts break andarc extinguishes here(b) Arcing with spiral petal contact(Transverse magnetic field type)Figure 19.26 Magnetic arc control to increase the contactarea of the arcing contacts in a VCBmay interrupt before a natural current zero and cause anear peak system voltage across the parting contacts.Figure 17.11(d) has been redrawn in Figure 19.27 formore clarity. Under the cumulative influence of thereflected wave and the equipment’s back e.m.f., it mayattain a value of high TRV, capable of breaking thedielectric strength across the parting contacts of theinterrupting device. It may cause yet higher TRVs, untilat least the immediate first natural current zero of theinterrupting current. See TRV at current zero (Figure19.27). Thus it would endanger the terminal equipmentand the interconnecting cable. Such surge voltages (TRVs)have increased to 2.5-3 p.u. in normal operation. It ispossible that the arc does not extinguish at the first currentzero, point ‘a’ on the current wave. If a sufficiently highdielectric strength between the contacts is not attained,even by the next natural current zero (point ‘b’) by virtueof an extremely low contact gap or an inadequate insulation*at the immediatecurrent zero ‘a’I II I Voltage wave --eI II . I . 0.02 part of one halfI I cycle of the current wave(enlarged for clarity)ICurrent wave-* Parting contacts are subject to this voltage which they are notable to withstand and cause an arc raising the TRV up to 2.5 puFigure 19.27 Approximate representation of assumed voltageand current waveforms illustrating a current-chopping effect andits attenuation while interrupting a circuit having 0.3 p.f.

I,19/648 Industrial Power Engineering and Applications Handbooksubject to frequent restrikes, and are vulnerable to damage,while the actual extinction of an arc will take place onlyat a natural current zero, To cope with such situationsthe interrupters must be fast operating. Figure 19.27illustrates the restriking phenomenon of the partingcontacts during current chopping that is assumed to occurat point ‘a,’, on the current wave. The actual current andvoltage waveforms may differ from assumptions,depending upon the speed of the breaker, rate ofdeionization, current being interrupted, its p.f. and theinstant at which the interruption initiates. In addition,the surge impedance of the circuit being interrupted. Weassume that the TRV may rise to 2.5 p.u. and is interruptedby the immediate first current zero (i.e. at point ‘a’ inFigure 19.27) in about 0.02 part of one half of a cycle ofa 50 Hz wave. During this period, if we consider thesurge frequency of the interrupting circuit to be of theorder of 13 kHz* the arc may restrike for nearly 2.6cycles or 5 times before a final interruption as determinedbelow.Time for 0.02 part of one-half of a cycle of a normalfrequency wave of SO Hz, during which current choppingoccurs:1 1= - x - x 0.02 second2 50:. Number of completed cycles of the TRV at the surgefrequency of 13 kHz1 1= 13 x lo3 x - x - x 0.02 cycles2 so= 2.6 cycles (approximately S restrikes)By the immediate first current zero it is assumed thatthe contacts have travelled sufficiently apart to achievethc rcquircd deionization and have built up adequatedielectric strength to withstand at least 0.95 V,. If thecircuit does not interrupt at the immediate current zeroat ‘a’, which is so near to the point of chopping ‘a,’, theintcrruption will take place only by the next current zeroat point ‘b’ and result in another 260 strikes by then. Tostudy more accurate behaviour of an interrupter, withthe number of restrikes and the formation of the actualtransient voltage waveforms on current chopping,oscillograms similar to those during a short-circuit testmay be obtained (Section 14.3.6).19.7 Virtual current choppingThis may occur during the interrupting process of aswitching device when not all the three poles will interruptsimultaneously. It is corollary to a closing phenomenon(Section 17.7.2(ii)) when not all the three poles make atthe same instant. This will also endanger the insulationof the other two phases. The interphase dielectric leakagecapacitances, as illustrated in Figure 19.28, are the cause.When one of the poles, say of phase R, starts opening*It is seen that the surge frequency during current chopping mayrarely exceed 20 kHz and which, in the context of switching surges,may be considered as low-frequency oscillations, easy to handleand interrupt.qsb; T&T -Motor ~windings I---- .......*.-------!= Load currentir, iv, ib = Charging or restriking currentsC,, C, C, = Interphase dielectric leakage lumped capacitancesFigure 19.28 One pole opening.Restriking phenomenon in phase ‘R’ causing charging currentsin phases ‘Y and ‘13’ which are still closedfirst and faces a restrike of the arc, leading to surgefrequency currents, similar (balancing) currents will beinduced in the other two phases that are still closed, inaddition to the normal current, I,, that these poles willstill be carrying. The result will be that when these twopoles also open the charging currents at a surge frequencymay virtually force a faster or premature current zero inphase B, as illustrated in Figure 19.29. This is termedvirtual current chopping and may cause an additional TRV.The amplitude of this TRV, however, may not be largedue to a generally low surge impedance of the interruptingcircuit during an interruption. It may achieve a level ofonly 0.6-0.7 p.u. (see Telander et al. 1986), which maysometimes prove fatal for the insulation of the terminalequipment due to its steepness. Charging currents woulddevelop in phase Y also (at point ‘y’, but not shown toavoid overlapping of curves). But current chopping is notpossible in this phase because pointy’ will fall further awayfrom a current zero, on the one hand, and the Y-phase wouldcarry a near-maximum current at this instant, on the other.Current chopping is a phenomenon of small currents.19.8 Containing the severity ofswitching surges19.8.1 Theory of energy balancingFrom the above we can deduce that at the instant ofcurrent chopping the arc extinguishes for a moment and

Circuit interrupters 19/649Virtual current chopping in phase 'B'- otLocation, r - Interruption commencesr' - Restrike of arc occurs in phase 'R' causing a chargingcurrent 'i:y- Charging current 'i; develops in phase 'Y (Not shownto avoid overlapping)b - Charging current '6' develops in phase '6'b' - Charging current ib causes a virtual forced current zeroor virtual current chopping at this pointFigure 19.29 An approximate illustration of a virtual currentchopping in Phase 'B as a consequence of restrike of arc inPhase 'Rre-establishes as soon as the TRV reappears. At the instantof arc extinction, it may be considered that the arc energyis transferred to the dielectric medium of the partingcontacts. This energy reappears in the form of a TRVacross them, and tends to cause a restrike of the arcplasma once again. We can generally consider the arcenergy as the magnetic energy caused by the magnetizingcurrent of the interrupting circuit and the energy receivedby the dielectric medium as the capacitive energy.If Ine is the inductive current in amperes (no-load currentof the motor or the transformer, Figure 1.15) and L theinductance of the circuit being interrupted in henry, thenthe electromagnetic energy, J, initially contained by thearc plasmaJ = 4 . L . Zzl JoulesNoteIn fact, this energy should be less by the hysteresis loss (Section1.6.2A.(iv)) which has been ignored in the present analysis. If V, isthe prospective peak surge voltage (TRV) in volts and C the dielectriccapacitance of the contact gap in farad at the instant of restrike,then the capacitive energy J, received across the contact gap isJ = 3. C V: JoulesFor successful interruption of the arc plasma it isessential that the energy emitted by the arc plasma is atleast equal to the capacitive energy received by thedielectric medium. At the instant of arc extinction, thisphenomenon is termed energy balancing, whenTInterruptorArc duringinterruption -----,or a restrike{' +$y "i\ I Contact gap having aInterrupting +dielectric capacitancedevice'C which is subject toa TRV of V, during arestrikeNo-load magnetizing circuit of anTransformer say,11 kVB.6 kVInterruptorCcable@Motor6.6 kWFigure 19.30(a)Figure 19.30(b) An energy-balancing phenomenon whileinterrupting a highly inductive circuit as shown in (a)

19/650 Industrial Power Engineering and Applications Handbookor V, = I,, J$ 2-I,, Z,We represent this through a circuit diagram (Figure 19.30)whereis the surge impedance, Z,, L and C the circuitconstants of the interrupting circuit, as discussed in(Section 17.6.4). C represents the dielectric capacitancebetween the parting contacts of the interrupter. V, mustbe prevented, as far as practicable, from reachingdangerous levels with the use of surge arresters.In( of equipment cannot be altered once it is manufactured.The surge voltage V,, that will appear acrossthe contacts therefore becomes a function of the inductance,L, of the circuit being interrupted. Lis the inductanceof the motor or the transformer windings up to theterminals of the interrupting device, including theinductance of the interconnecting cables or the bus system.The dielectric strength is the capacitance, C, betweenthe contacts. The C of the contacts will change with thetravel of the contacts and the deionization of the arcplasma. The influence of the inductance of the circuit L,can be reduced by introducing some capacitance orresistance into the circuit. Capacitance can be introducedby installing a few p.f. correction capacitor banks acrossthe motor or the transformer terminals and the resistancecan be introduced temporarily across the moving contactof the interrupting device through its closing mechanism.This is a practice adopted by leading manufacturers toprovide a resistance where the interrupter is required tooperate under adverse switching conditions.19.9 Comparison of interruptingdevicesTo assist in making an easy selection of an interruptingdevice for a particular application, we have provided acomparison between the various devices as in Table 19.1.

Table 19.1Comparison of interrupting devicesI 2-I Specification\Mcditini of Oilinsulation andquenching of arcplasmaNominal voltageratingsHT 3.3-200 kVVCB"7Oil Air Air blast Sulphur hexallutri-idc Vxuum3.3-470 kV and above Up to 22 kV(not in practice in lightof bettei- technologies inVCB's and SF6)I 1-765 kV 6.6-76.5 kV and above 3.3 hV to 30 k\iLT 66OVNot economical 660 VNot applicable Not applicable Not applicable(iii) Nominal current Manufactured in all Only one break per pole, Up to 6000 AUp to 4000 A and higher Up to 4000 A and highcr A\ requiredratingspossible current ratings therefore current ratingsin higher ranges arerestricted(iv) Closing time 5 cycles4-6 cycles 1-2 cycles 2-3 cycle? 4-5 cycles 2-3 cycles(between the instantof application ofvoltage to theclosing coil and theinstant when thecontacts touch)c(v) Opening time(between the instant 4-5 cyclesofapplication ofvoltage to the tripcoil and the instantof separation of the2-3 cycles i-l cycle1-2 cycles 1-2 + cycles1-2 cycle.;arcing contacts)'(vi) Maximum duration 0.75-1.25 cyclesof arc plasma (arcingtime)'2 cycle\+-1 cycle1-1 + cyclei-I cycle(vii) Total breaking time 4.75-6.25 cycles 4-5 cycle\(between the imtantof application ofvoltage to the tripcoil and instant offinal arc extinction)'1-2 cycles2-4 cycle.;1;-3 cyclcs7 Theory of arc plasma Emits ions of H2 aroLintl 70% and remining a\ Emits ion\ of N, (80% 1 antl 02 (20'2 1 and ~nctallic Emits IOII\ ol S' (I~)~(III\). Emits metallic vapourfromtion~~ation of gares at acetylene antl rnetallic vapoiir\\,ipour\S" (neutron\). iiccom- the contact material. Sincehigh temperature

1 2 3 4 56 73 Theory of arc Cooling of arc plasma is based on the bubble theory, Deionization of N, (0, Deionization of N, andquenching and H, cooling it by the turbulence caused by H2 bubble having a small content, has 0, by a strong blast ofextinction, through no significant influence) air. The deionizationdeionization of arcby lengthening of the arc force remains the sameplasma plasma through the arc for one size of breaker,chutes, as a result of mag- irrespective of thenetic field induced in the currentmetallic splitters of the arcchute by the inductive arcplasmaDeionization of freesulphur electrons S",takes place through theirabsorption by thefluorine ions F(electronegative theory)boiling point, the vapouris largely composed ofcopper ionsDeionization of copperions is natural and extremelyfast4 Number of no-load Min. 1000operations (dependsuponthearcenergy: thelower the arc energy,the higher will be thenumber of operationsand vice versa).Min. 1000 Min. 1000 Min. 1000Min. 5000 Min. 10 000-20 0005 Maintenance High, to check the conditionof oil, contacts andarc quenching tins anddevicesChecking the condition of Very low maintenance except for checking the Requires very little Requires very littleoil is more frequent, in condition maintenance, except maintenance, exceptview of smaller quantity of the arcing contacts for any wear, tear or pitting periodic checks of periodic checks ofof oil pressure and condition vacuum and condition ofof the contactsthe contacts. Maximumcontact erosion 2-3 mm6 Dielectric strength > 2compared to air>2 1 Compressed air has better 2-3than 1Nearly 10 times (5 timesof oil)7 Arc energy compared 1to oil8 Whether suitable tointerrupt smallinductive(magnetizing) orcapacitive currents,without causingcurrent choppingYes (but for a limitednumber of switchingoperations), because ofdouble break perphase that enables quickrestoration of thedielectric strength1No, because of generallyone break per pole, forbreakers 12 kV andabove, unless the breakeris specially designedwith extra quenchingsystem through a jet ofoil- - Roughly - I (Figure 19.6) Roughly - I (Figure 19.6)3 10(Contact erosion low and suitable for frequentoperations)On an LT system, such a No, bccausc of force ofphenomenon has no air blast even at smallrelevance, because of currents which mayhigh dielectric strength develop the tendency tobetween the parting current chopping. On thecontacts compared to the other hand, it is possiblesystem voltagethat the breaker may failto interrupt the steepfronted TRVsYes, in all types of arcquenchingtechniques:(i) Puffer type: becausethe flow of gasthrough the arc is afunction of magnitudeof current(ii) Rotating arc type:because themagnetic field thatGenerally not, becauseof extremely fastdeionization. It has thetendency to currentchopping. However, inview of continuousdevelopment in thisfield, it has now beenpossible to achieve achopping current as low(Confd)

I 2 3 1 5 6 I9 Application switchingofInduction motorsGeneratorsTransformersReactorsInductive circuitsCapacitor banksDistribution linesHigher-speed autoreclosing(a) Suitable for allapplications. not (a) Suitable for allrequiring frequent applications notswitching operations. requiring frequent(b) Until now they were switching operationsbeing extensively used fora11 LT applicationsExtensively used for allLT applications andsuitable for frequentoperationsGenerally witable forHT distribution only,and at installationsemploying a number ofsuch devices to makedry compressed airsystem handy andeconomicaldri\e\ thc arc i\produced by theinteri-upting cui-rentitself, and will varywith the current.and so will varythe cooling gas indirect proportion.The smaller thecurrent to beinterrupted, thesmaller will be themagnetic field andthe cooling force,enabling the arc toinrerrupt at anatural current zeroonly. But thistechnique is nowseldom practised(iiij Thermal blast and arcassistance type:As (i) ahoveSuitable for allapplications. Switching ofcapacitor banks, however,may pose problembecause of high TKVs.These may cause a restrikeof the arc plasma and giverise to switching surges;I\ 2A and even la\.lherelore, exceptinterrupting current,lower than this. theseinterrupters are suitablcto interrupt huch currentswithout current choppingSuitable for all types ofindustrial needs andlrequent opcration. Theirability for quickinterruption and fastbuilding up of dielectricstrength made themsuitable for onerousduties and high TRVsIO Whether a rurgearrester is essentialNoNoRV Better to provide Ciencrally not(depending upon thetype of equipment it is\witching. the minimumcurrent it has to interruptand the likely amplitudeof TRV)Ye\. particularly for dryinsulated equipment (suchas an induction motor, adry type reactor or a drytype transformer), whichhave low insulation le~~els .An oil-immersed equipment(such as a powertranrformerj. having arelativcly much bettcidielectricstrength, may be\witched without anai-rester(C,Jll/d)

~ ~~I 2 3 4 5 6 7I1 Effect of leakage onperformance:(i) Operating safety Since the oil is filled at atmospheric pressure, there is Not applicable Not applicable (i) Generally, the use ofgenerally no leakageSF, is safe, inhaling ispoisonous. If there is aleakage it would bedangerous. SF, itself isnon-toxic, but decomposedSF, is moderatelytoxic and must behandled under controlledconditions. Training andguide-lines are essentialfor personnel workingon such equipment. Sitedata collected fromvarious sources,however, suggest aleakage rate of less than0.1 % per annumGenerally a safeequipment. But in theevent of vacuum leakagewhich, although remote,may cause a fire hazardand X-rays. X-raywarning and propershielding may beessential(ii) Dielectric strengthNot applicableNot applicableNot applicableNot applicable Still high Zero(iii) Intempting capacity Not applicableNot applicableNot applicableNot applicable. But if the In the event of aleakage is in the compres- leakage, the breaker willsed air supply system, the still be in a position tobreaker may not interrupt interrupt the normalon fault, depending upon currentsthe air pressure the systemis able to maintain on aleakageIn the event of aleakage, the breaker willnot be in a position tointerrupt even normalcurrents12 Fire hazardsGenerally, oil is fire hazardous, but not prone to itunless the dielectric strength reaches a very lowlevel and the medium itself becomes conducting.Nevertheless, they are not suitable at locationswhich are fire hazardous. Normal make or break,even on fault, may only deteriorate the quality of'oil, but yet not make it fire hazardous, unless asnoted aboveThey may be rated as more fire hazardous than a Arc formation and extinction takes place inside aBOCB or MOCB in view of arc formation taking sealed chamber, thus emitting no gases or vapouraplace in the open, although under controlled conditions. to the atmosphere, which may be a cause of fireThey are not suitable for installations prone to fire hazard. They are the most appropriate choice forhazards, unless the sub-station or the control room such areaswhere they are installed is isolated from the area ofhazards (Section 7. I I). They are generally suitablefor all other areas13 Noise levelQuiet operation Quiet operation Except during an Thunderous noise during Quiet operationinterruption, it has quietoperationan interruption, because ofair blast. The intensity canbe controlled by providingsilencersQuiet operation

~ ~~~~~~ ~~I4 Trend\Thcw breaker\ wcre cxten\i\elq used until a ternyear5 ago (say, 1970s) hut are non fast losing thcitholdin favour of SF, and VCB\ for HT and ACBsfor 1.T \y\tetn\3nly hreaket for all LT~pplication\In lower range\. say. upto 66 kV. this breaker ISnow fast lo\ing it\ holdin favour of SF,, duc tocost considerations aridthe cumbcrsome dry airpreswre sy\tem. Inhigher ranges. say. 120-765 kV, however. theyare \till being used hutonly rarelyThis has become the This has the latotiiiobt preferred breaker technology in the at-cwith thc wide\t voltage interruption and suitablerange and suitability for for all applications a\all application\noted above, particularlyin view of havingachieved a very lowchopping current, of theorder of just 2A and evenle\s. In view of currentchopping and limitedvoltage range. however,it is facing limitations inits extensive application.The development workon higher ranges is stillunder way and not manybreakers above 36 kVhave been produced sofar. Tendency to a coldweld, during closing andchange in the arcchamber pressure, duringan interruption whichmay damage the metallicbellows (Figure IY.24),have also been noted asdisadvantages in theirextensive use"These comparisons relate only to HT interrupters."For LT systems only.'Referred to at power frequency (SO or 60 H7) and ratcd current. These figures are approximate and for a general reference only. They may vary with rating, loading and manufacturer. Forlower voltage ratings. they may he shightly higher.

19/656 Industrial Power Engineering and Applications HandbookRelevant StandardsIEC Title IS BS60056/1987 High voltage alternating current circuit breakera 13118/1991 BS 5311/199660298/1990 A.C. metal enclosed switchgear and controlgear for 3427/1991 BS EN 60298/1996rated voltages above 1 kV and up to and including 52 kV605 17/1990 Gas insulated metal-enclosed switchgear for rated voltages of72.5 kV and above60947-2/1998 Low voltage switchgear and controlgear,Circuit breakersRelated US Standards ANSVNEMA and IEEE- BS EN 60517/199713947-2/1993 BS EN 60947-2/1996NEMA/SG-3/1995NEMA/SG-4/1990NEMA/SG-6/1995ANSI C 37.06/1994ANSI C 37.16/1998Lnw voltage power circuit hreakersA.C. HV circuit breakersPower switching equipmentA.C. HV circuit breakers rated on a symmetrical current basis - preferred ratings and related required capabilitiesLow voltage power circuit breakers - preferred ratings and application recommendationsNotes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it ia pusaible that reviaed ediliunshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Somc of thc BS or IS standards mcntioncd against IEC may not bc idcntical3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.Further readingI Bettge, H., ‘Design and construction of vacuum interrupters’,Siemens Circuit, XVIII. No. 4 (1983).2 Blower, R.W., Distribution Switchgear, Collins Professionaland Technical Books, 1986.3 Blower, R.W., Cornick, K.J. andReece, M.P. ‘Use ofvacuumswitchgear for the control of motors and transformers inindustrial syatenia’, E/ec/ric Power Applictrrions, 2, Nu. 3.June (1979)4 Bouilliez, O., ‘Mastering switching voltage transients withSF, switchgear’, Cahiers techniques Merlin Gerin, No. 125.5 ‘Current chopping phenomena of MV circuit breakers’, IEEETransactions, PAS -96, No. 1 (1977).6 Pelenc, Y. ‘Review of the main current interruption techniques’Voltas Limited, Switchgear Division, India.7 Ramaswamy, R., ‘Vacuum circuit breakers’, Siemens Circuit,XXIII, April (1 988)8 Telander, S.H., Wilhelm M.R. and Stump, K.B., ‘Surge limitersfor vacuum circuit breaker switchgear’, CH 2279-8/86/0000-00375 1 .00, IEEE (1986)

201657Te m po ra ryOvervoltages andSystem Grounding

Temporary overvoltages and system grounding 20165920.1 Theory of overvoltagesA three-phase balanced system has all the three phasorsof voltage and current 120" apart. as illustrated in Figure20. I (a) for a conventional anti-clockwise rotation. Thesephasors are known as positive sequence components.During a fault. this balance is disturbed and the systembecomes unbalanced being composed of two balancedcomponents, one positive and the other negative sequence(Figures 20.l(a) and (b)). For a description of the efiectsof these components, refer to (Section 12.2(v)). Duringa ground fault, zero phase sequence components alsoappear, which are single phasor components and combinethree equal phasors in phase, hown in Figure 20. I (c).This is the residual voltage, Vg, that appears across theground circuit, i.e. between the neutral and the groundas illustrated in Figure 20.12. This voltage is responsiblefor a fault current, I, . Is will flow through the groundedneutral when it is a three-phase four-wire neutral groundedsystem, as shown in Figure 20.12. It will also flow througha thrce-phase three-wire artificially grounded system whenit is $rounded through a neutral grounding transformer(Section 20.9.1 ) as illustrated in Figures 20.17 and 20.18.In a three-phase three-wire system, which has neither itsown grounded neutral nor an artificially created groundedneutral. there will be no direct ground fault current. Butcharging currents through the ground leakage capacitances,particularly on an HT system, may still exist, as illustratedin Figures 20.2-20.3.These currents may develop dangerous overvoltagesacross the healthy phases, under certain ground circuitimpedance conditions, as discussed in Section 20.2. I( 1).It is thus possible to encounter a ground fault, even whenthe system is not grounded, the fault current finding itsreturn path through the ground leakage capacitances. Whilean LT system. in view of a far too low ground voltage, V,(equal to line voltage, Section 20.2.1( I)), as comparedto high ground capacitive leakage reactance. Xcg. wouldcause a near open circuit (V,/X,,being too meagre) andstay immune. leaving the grounded conductor floating atv, = v,Figure 20.2 An ungrounded or isolated neutral system(circuit completing through the ground leakage capacitances)Figure 20.3 Case of ground fault within the loadBvlt) L V R v>/ -2% 1-__ vIv v454361NY iVY V(a) (b) (C)Positive sequence Negative sequence Zero sequence orresidual quantities =Gv, = v, = v, = v - t',Figure 20.1 Phasor representation of an unbalanced power Figure 20.4 Case of a ground fault on a power system onsystem on a ground faultthe load side

20/660 Industrial Power Engineering and Applications Handbook20.2 Analysis of ungrounded andgrounded systemsThis is related more to an HT system.20.2.1 Analysis of an ungrounded systemThis is common toA three-phase three-wire, star connected isolated neutralsystem,A delta connected system, orA three-phase four-wire, ungrounded system.An ungrounded a.c. system is more often subject to overvoltages.The reason is that even when the system is notconnected to ground, it makes a ground connection throughthe coupling of the leakage ground capacitances, asillustrated in Figure 20.2. For all purposes these capacitancesmay be regarded as equal and uniformly distributed,providing a balanced system. In the event of a groundfault, this circuit is closed as shown and a capacitivecurrent, I',, flows back to the healthy phases via thesecapacitances Cg, Analysing the circuit of Figure 20.2,the following may be derived:IfC, = electrostatic or leakage capacitance to theground per phase, providing the return path tothe fault current Ig, in the event of a groundfault on the system. This current will becapacitive in nature.Xcg = ground capacitive reactance per phaseUnder transient conditions, f becomes surge frequency,f s andXc, becomes,where fs =12xfs . C,127c J lL being the inductance of the circuit.If I, = ground fault currentV, = ground potential, which is the same as zerosequence voltage or residual voltageVt = line voltage andVp = phase voltagethen on a ground fault, say, in phase R, the voltage acrossthis phase will reduce to zero, while the ground potentialacross the healthy phases will rise to V, from a zero levelearlier. It will cause a fault current Zi, in the healthyphases also, through the ground leakage capacitancessuch that I: = VI/X,, (assuming the ground circuit hasno other impedance) and the ground fault current throughthe grounded circuit,_ -I, = I, + I,= r,' + I,' (both phasors being 60" apart)= JIi2 + Ip" + 2 . I; .I; . COS. 60" (as in Section15.4.3)= 1; (but this current may not be enough to tripa protective relay)The voltage across the healthy phases is now fi timesor 73% more than the phase to neutral voltage underhealthy conditions.Under certain line to ground impedance conditions,when the ground circuit may also contain some inductivereactance, the voltage of the healthy phases may risefurther, much above the system voltage, due to resonanceand ferro-resonance effects (see below). All this in turnmay tend to swing the system to an unstable state. It mayalso cause arcing grounds* at the supporting insulators,leading to voltage surges, travelling in both directionsalong the line, and may be enough to cause damage tothe line insulators and the terminal equipment.The magnitude of overvoltage will depend upon theactual inductive reactance introduced through the groundcircuit. The unintentional introduction of an impedanceinto the ground leakage capacitive circuit is generally aresult of a ground fault in the main equipment such as agcncrator, a transformer, power capacitors or an inductionmotor and when one or more of their phase windingsforms a part of the ground circuit, as illustrated in Figure20.3. Under such conditions, they will introduce theirown inductive, capacitive or resistive impedances or acombination of them into the ground circuit and alter theparameters of the ground capacitive circuit. Figure 20.5illustrates the system of Figure 20.2, drawn simply on asinglc phase basis. The current through the ground circuitwill be zero when the system is healthy, the ground leakagecapacitances finding no return path (Figure 20.6). It isnot necessary that the equipment itself should develop aground fault and then only its impedance will be inductedinto the ground circuit. It may be introduced into thecircuit even when there is a line or a system groundfault, as illustrated in Figure 20.4. The ground circuitmay give rise to dangerous voltages in the healthy phases,as analysed below.(1) When the external impedance is a resistanceor a capacitive reactanceReferring to Figure 20.5, if Z; is the external impedanceintroduced into the natural leakage ground circuit, as aconsequence of a ground fault, as shown in Figures 20.3* Arcing grounds: This occurs during a temporary ground fault onan ungrounded HT system. It causes an arc between the line conductorand the ground, which may be direct or through the flashover of theline insulators, due to overvoltage. The ground leakage capacitorsthat may be considered as charged with the line voltage are dischargedto the ground on developing a ground fault. The supply voltagewould charge them again and so the process will repeat until thefault exists. The repeated discharges and charges of ground capacitorsduring the fault produce arcs and are termed arcing grounds, givingrise to voltage surges.

I emporary overvoltages and system grounding 201661May attenuate at about 6-10 . &v,XcQ = Ground leakage capacitive reactanceZ; = Additional impedance (R, X, or X,)Figure 20.5 Equivalent ground circuit representing the powersystem of Figure 20.2I- ?-1 24 5Ratio of additional impedance i'i to Xcg --c@ Resistive reactance@ Capacitive reactance@ Inductive reactanceOn a healthy system the ground leakage capacitances find no returnpath and therefore do not activate. They do so when there is a groundfault as illustrated in Figures 20.3 and 20 4Figure 20.6 A healthy systemand 70.4, then the voltage across the impedance 2; canbe determined as follows:When the impedance is a resisthie reactanceThe peak voltage %:,,,dxj across R will be(20.1)Computing values of Virnah) by approximating, in termsof circuit resistance Rand ground capacitive reactance XCg,whenK = 0, G,,lJyl 0Figure 20.7 Influence of external impedance Zb in theground circuit on the system voltage in the event of groundfault in an ungrounded systemreactance, R, in the ground circuit. is shown in Figure20.7, curve 1.InferenceIrrcspcctive of the value of the external resistance (R) inthe ground circuit, the maximum voltage the healthyphases may have to sustain will not exceed &V,, inthe event of a ground fault on an ungrounded system.When the impedance is n cnpacirive reacfanctReferring to Figure 20.5, the peak voltage Vl' across X ,will bev;max' 43= x, +xcgv: (20.2)Computing VinlaXalong similar lines to those above. i.e.whenx, = 0, = 0R2= 2X,,, V,',,,, = --.fiV, 2- 89% of fiV,JsAs K approaches infinity:AV, .will tend to approachThe variation in r/,',,,,,, with the variation in resistivex, = x,, , v,', = + . av,x, = 1.5 x,,,= 50% of dv,1.5Vi",,, = 2.5 aV, = 60% of &V,X, = 2Xc, . Vimax2= - fi V, = 66.7% of $V,3As X, approaches infinity. Vimaxwill tend to approachaV,. The variation in V,',,, with the variation incapacitive reactance, X,, in the ground circuit is shownin Figure 20.7, curve 2.

201662 Industrial Power Engineering and Applications Handbooklnfe ren ceThis is same as for resistive impedance. In these twocases, when the external impedance is resistive orcapacitive, there is no excessive voltd e rise across thehealthy phases ofthe system beyond &V(, The voltagedeveloped across the ground capacitance, Xce, and theexternal impedance R or X, is shared in the ratio of theirown values, the sum total of which will remain constantat Av(.(2) When the external impedance is an inductivereactanceNow the situation is different, as resonance and ferroresonanceeffects in the series inductive-capacitive circuitsmay cause dangerous overvoltages.(i) Resoizaizce effectReferring to Figure 20.5, the peak voltage Vi across XLwill bexLx,, -xL'V,',,'!, = fiV((20.3)Computing Vi along similar lines to those for resistiveimpedance,i.e when XL = 0 V,',,,,, = 0(ii) Ferro-resonance effectThe above analysis of overvoltages in the healthy phasesof an ungrounded system in the event of a ground faulton one of the phases was based on the assumption thatthe inductive reactance of the electromagnetic circuit,Le. the magnetic core of the connected equipment (whichmay be a transformer or an induction motor) was linearover its entire range of operation. But this may not alwaysbe true. It is also possible that some components such asa CT or a CVT, may have their magnetic core graduallysaturated* during normal operation, under certain circuitconditions and resonate with the ground capacitivereactance Xcg. This may lead to a high voltage in a healthysystem, even when there is no ground fault, the groundcircuit becoming completed through the grounded neutralof such a device (Figure 20.8).'Clt, krRYBXI, = Xcp, V,',,,,,will tend to approach infinityThis is known as a resonating condition. It is, however,seen that in view of some in-built impedance in the groundcircuit it will tend to attenuate the alarmingly risin voltageVil,,,;,, to oscillate at around 8 to IO times dV,. Thisvoltage will tend to raise the ground potential substantially,depending upon the value of the external impedance X,.It will also raise the ground potential of the healthy phases,which may cause arcing grounds and become dangerousto the line insulators and the terminal equipment. This isknown as the resonance effect of the inductive reactance.WhenAs X,> approaches infinity, Vilndxwill tend to approach&V,, This variation in V$, with the variation ininductive reactance, X,, is also shown in Figure 20.7.The inductive reactance, XL, will tend to offset theground capacitive reactance X,, and diminish thedenominator to a certain value of X,, say, until itcompletely offsets the content of Xcr (X,= Xcg). At higherratios, when XL > 3 Xcg, the denominator will rise morerapidly than the numerator and will tend to allmuate theV,'lndxa~ with R and X:, but at a slightly higher value ofVinlax (Figure 20.7, curve 3).During saturation X, 1 Xcg leading to the phenomenon of ferroresonanceFigure 20.8 Case of a ferro-resonance in a residual VTleading to overvoltages even in a healthy systemThe phenomenon of saturation of the magnetic coreof such a device during normal operation and its resonancewith the ground capacitive reactance, Xcg, is known asferro-resonance. It would have the same effect on thehealthy phases/systein as in (i) above.The inductive reactance of the magnetic circuit onsaturation may fall to a much lower value than the linearinductive reactance of the magnetic core, as a consequenceof design requirements, and lead to a condition of lowXL to Xcs ratio (say XL = 1.5 to 2 Xcg) in curve 3 ofFigure 20.7. Under such a condition, the voltage acrossX, will tend to oscillate automatically between certainovervoltage limits. The effective X, is seen to match theground capacitive reactance Xcg such that it helps todampen the overvoltages across the healthy phases, tooscillate at around two to three times aV,.*Saturation of transformel-s also produces high currents, rich inharmonics.

Temporary overvoltages and system grounding 20/663Such a situation may also arise in electromagneticequipment, which is subject to a varying system voltageduring normal operation. One example is a residual voltagetransformer (RVT) (Section 15.4.3) which is required todetect an unbalance (zero sequence) or a ground fault inthe primary circuit and may reach early saturation. Asimilar situation is also possible in a measuring CT andeven a protection CT, as both may saturate at a certainlevel of fault current in the primary. The same situationwould arise in a CVT (Chapter 15). Such devices (CTsand CVTs) are generally grounded as a safety requirement,and may give rise to such a situation during normaloperation, even in a healthy system.20.3 The necessity for grounding anelectrical systemAn ungrounded system, in the event of a ground fault, issubject to an overvoltage, as noted in Section 20.2.1. Itis prone to cause yet higher voltages in the healthy phaseswhen its ground circuit becomes inductive, as discussedabove. Thc ovcrvoltage may damage the supportinginsulators and the terminal equipment. Should a systembe left ungrounded? This aspect must be viewed with theabohe phenomena in mind. It may also cause arcinggrounds and prove fatal to a human body coming intocontact with the faulty equipment or the conductor.Assuming that a total power system, from its generatingstation to the far end LT distribution network (Figure13.21) is left ungrounded and a ground fault occurssomewhere on the LT side, the fault current through thehuman body will find a return path through the groundingcapacitances, no matter how feeble it may be. This currentmay be dangerous to a human for which even 10 mA isfatal. as discussed in Section 21.2.1.Generally, neutral grounding should be adopted inprinciple to avoid generation of overvoltages and toeliminate the phenomenon of arcing grounds. Even asolidly grounded system, with a very low resistivereactance, can avoid such overvoltages as the fault currenton a ground fault will find its return path through theshortest solid ground conductor, having a low resistance,rather than through the ground capacitances, which havea relatively much higher capacitive reactance. (See Figure20.12.)Below we briefly discuss the critcria and theory ofselecting a grounding system to achieve a desired levelof fault current to suit a predetermined ground faultprotection scheme, i.e. type of grounding and groundingimpedance to suit the system voltage, type of installation,and location of installation.z,This is measured between phase to phase or phaseto neutral, depending upon the availability of theneutral. The test current is kept at the rated valuefor the equipment or the system under test. For thesystem shown in Figure 20.9(a).VI=-v'3 ' I,22 = negative sequence impedance.This is the same as above, but now a negativesequence voltage is applied (by interchanging oneof the phases of the source of supply)Zo = zero phase sequence or residual impedance. Thisis measured between the three-phase terminals ofa star winding shorted together and the neutral(Figure 20.9(b)) and is calculated bywhere - V = test voltage and43I = test current, to be around the currentcarryingcapacity of the neutral.Also Zo = 3% + 3%where Ro = zero phase sequence or residual resistance,andX, = zero phase sequence or residual reactanceZg = total impedance through the ground circuitIg = ground fault current, zero sequence currentor residual current through the ground circuitVg = ground potential, which is the same as thezero sequence voltage or residual voltage. Ina ground fault, it would remain at v;/&unlike at Vt in an ungrounded system (Section20.2).The residual voltage may also be measured by a residualvoltage transformer (RVT) (Figure 20. IO). Refer to Section15.4.1 for more details.Supply source420.4 Analysis of a grounded systemConsider Figure 20.12 again, when the neutral N issolidly grounded. Also refer to Figure 20.1. IfV, = line voltage2, = positive sequence impedanceI(a) To measure positive(b) To measure zerosequence impedancesequence impedanceFigure 20.9 Measuring system impedances

20/664 Industrial Power Engineering and Applications HandbookNBA three-phase four-wire system may be grounded inthe following ways:1 Solid neutral grounding system or2 Impedance neutral grounding system20.4.1 Solid neutral grounding system (alsoknown as effectively grounded system)Residualvoltagetransformer(RWResidual voltage Vg = 36Figure 20.10 Detecting a ground fault in an isolated neutralsystemThen the total impedance through the ground circuit2, = 2, + z, + 2"and the fault current through ground circuitI =- 3 ",z,The residual current may also be measured by a three-CT method as illustrated in Figure 20.11.When some extra impedance R, X,-, XL or a combinationof these is introduced into the ground circuit it willbecome possible to alter the magnitude and thecharacteristic of the ground circuit current, Ig, to suit analready designed ground fault protection scheme asdiscussed below.We have already discussed a solid neutral groundingsystem in Section 20.3. The residual volta e or the groundpotential rises to the phase voltage Vt/& and does notalter the voltage of the healthy phases. To analyse thissystem, we have redrawn the circuit of Figure 20.2 inFigure 20.12, grounding the neutral solidly. The impedanceto ground, Zg, through the neutral circuit will be extremelysmall and resistive in nature, compared to the groundcapacitive reactance Xcg, i.e. Zg

Temporary overvoltages and system grounding 20/665Figure 20.13Z, = Tapped reactor or arc suppression coilAn impedance grounding neutral systemIf I," = fault current through the healthy phases due toground capacitive reactance Xce, then the currentthrough the ground capacitive reactances= c+= \3.1:And the total ground fault currentThe value of I, can thus be varied in magnitude andphase displacement to suit a particular location ofinstallation or protective scheme by introducing suitableR and X , into ~ the neutral circuit. When the impedance isinductive. the fmlt current 1; will also be inductivc andwill offset the ground capacitive current I:. In such agrounding, the main purpose is to offset the fault currentas much as possible to immunize the system from thehaurds of an arcing ground. This is achieved by providingan inductor coil. also known as an arc suppression coil,of a suitable value in the neutral circuit.With such a system the possibility of an arcing groundis almost eliminated and it is now possible even to allowthe ground fault condition to prevail until it can beconveniently repaired. Now it will cause no harm to ahuman operator, supporting insulators or the terminalequipment. On long-distance transmission or distributionnetworks such a situation may rather be desirable toprevent the system from tripping instantly until at leastthe off-peak periods or until the supply is restored throughan alternative source. The process of finding the faultand its repair may be allowed to take a little longer.A neutral grounding reactor (NGR) is also desirableto achieve auto-reclosing of a faulty phase during a onepoleopening as a result of a fault of a transient nature onthis phase (Section 17.4.1). The reactance grounding willlimit the high unbalanced capacitive currents throughthe healthy phases to ground and prevent an unbalanceand so also an unwanted trip as illustrated in Figure20.14. Referring to Figure 20.14, a ground fault of atransient nature on phase B would cause the two linebreakers b and h' to trip and result in heavy ground faultcurrent through the capacitive coupling of the healthyphases, eventually also tripping the breakers ofthe healthyphases. The NGR would neutralize such currents andprevent an unwanted trip, achieve the desired auto reclosingand improve system stability. For such a situationto arise it is tuned to a zero p.f. to achieve a near-resonantcondition so that the fault current, Ig, is almost zero, andthe capacitive current is substantially offset by the powerfrequency inductor current, i.e.where the total ground capacitive reactanceIfL = inductor coil inductance in henryC,,= ground capacitance per phase in farad andf = system frequency in Hz20.5 Arc suppression coil or groundfault neutralizerThis is also known as a Petersen coil, named after itsinventor. With an inductive reactance, X,, in the groundcircuit the ground fault current can be substantiallyneutralized by tuning the inductor correctly. A smallresidual ground current however, will still flow throughthe ground circuit as a result of its own resistance, insulatorleakage and corona effect. In all likelihood, it would besufficient to operate the protective scheme. Since thefault current is now nearly in phase with the voltage ofthe healthy phases, it will prevent the interrupting devicefrom causing a restrike while interrupting the fault (Section17.7.2(iii)). Such an arrangement is more appropriatefor systems that are above 15 kV and are subject tofrequent ground faults, for example an overheadtransmission line or a long distribution system.Coupling (leakage)capacitancesgrounding NeutralFreactor (NGR)Figure 20.14Transmission line,\- i' P I 'I I 4 41grounding Neutra:{reactor (NGR)t t t\A ground fault neutralizer.R

20/666 Industrial Power Engineering and Applications Handbookthenor L 1henry3(2~f)~ . Cg(20.4)It is likely that for reasons of system disturbances,frequency fluctuations and switching of a few sectionsof the system, both X, and Xcg may vary in actual operationand upsct the resonance condition, leading to transientovervoltages. To overcome this, the inductor coil maybe made variable (the setting of which may be alteredautomatically) through a motor-driven tap changer toachieve the tuning again. If the ground fault persists theinductor coil may be rated continuously rather than for ashort time and for the full fault current that the coil mayhave to carry.To overcome the generation of overvoltages acrossthe inductor coil (Section 20.2.1(2)) its inductance, L, isgenerally selected high, so that the resonance conditionwith the ground capacitors, C,, occurs near the naturalfrequency of the system (SO or 60 HL) and the voltagedeveloped across the inductor coil, Vi, may oscillateonly at around Vel&. Generally, the Petersen coilneutralizer is a high-reactance grounding and is also termedresonant grounding, free from overvoltages and restrikes.Equating the voltages developed across the Petersen coilappropriate ground fault protective scheme as well asthe insulation level for that system. It is defined as theratio of the highest voltage to ground, Vg (r.m.s.), of thehealthy phase or phases during a ground fault to thecorresponding power frequency phase voltage V, /&when the system was healthy. Refer to Figure 20.15."E:. Ground fault factor (GFF) = - (20.5)v, 143which is usually more than 1.It is also established that in an effectively groundedsystem the voltage to ground, Vg, of the healthy phasesdoes not exceed 80% of the line-to-line voltage V, andconsequently the GFF does not exceed 0.8 x 4, i.e.1.4. The system may be considered as effectively groundedwheno= between 0 and 3, andx+R2 = between 0 and 1X+whereXo = zero sequence reactanceRo = zero sequence resistanceX, = positive sequence reactanceRThe voltage developed would thus oscillate around thenormal voltage and fall in phase with the fault current toachieve a near-strike-free interruption of the interruptingdevice on a ground fault.Note1 Aa in IEC 60071-1,Za higher insulation level (BIL) w~ll benecessary for all insulators and terminal equipment when theground fault persists for more than 8 hours per 24 hours or atotal of more than 125 hours during a year.2 Because of likely de-tuning and generation of overvoltages,this system is seldom in practice.Sometimes such a situation may arise on its own,even on a normally grounded system, not intendcd forground current neutralizing. It can happen when anoverhead line snaps due to a storm, winds or any otherfactor and falls on trees, hedges or dry metalled roadsand remains energized in the absence of a proper returnpath and cause a low leakage current, insufficient to tripthe protective circuit. This is a situation not really desirableon a normally grounded system, as it may lead to anungrounded system and may develop overvoltages.20.6 Ground fault factor (GFF)This is an important indicator that shows the groundingcondition of a system and helps to determine the most6--1v, Id 3r GG(a) Healthy system=G(b) Faulty systemR & Y - Healthy phases6 - Faulty phaseV, - (rms) value of ground voltage measured on faultFigure 20.15 Determining the ground fault factorY

Temporary overvoltages and system grounding 20/667To achieve the desired conditions of grounding, thefollowing are the generally adopted grounding practices.20.6.1 An ungrounded or isolated neutral systemWhen the system is totally isolated from the ground circuit,except through indicating, measuring or protective devices,which are normally grounded and possess a highimpedance to ground, the ground fault factorThis may become higher, depending upon the circuit'sconditions (for example, the ferro-resonance effect,Section 20.2.1.2).20.6.2 A grounded neutral systemWhen the system is grounded through its neutral, eithersolidly through a resistance or through an arc suppressioncoil (inductor), it becomes a grounded neutral system.This type of system grounding may be classified asfollows:Effectively groundedNon-effectively grounded andResonant grounded systems.Effectively grounded systemThis is to achieve a higher level of fault current to obtaina quicker tripping on fault. It is obtained when the systemhas a ground fault factor not exceeding 1.4 ( Vg I O.SV,),as noted above. A solidly grounded system will provideeffective grounding. This system will reduce the transientoscillations and allow a current sufficient to select aground fault protection. It is normally applicable to anLT system.Non-effectively grounded systemAn impedance grounded system will fall into this category.The GFF may now exceed 1.4 (Vg > 0.8Ve).Resonant grounded systemWhen the system is grounded through an arc suppressioncoil (reactor) so that during a single phase to groundfault, the power frequency inductive current passingthrough the inductor coil would almost offset the powerfrequency leakage capacitive component of the groundfault current. However, such a situation is not allowed topersist for more than 8 hours in any 24 hours or for atotal of 125 hours during a year. If it is more, a higherlevel of insulation (BIL) will become necessary. Theground fault factor in this case may also exceed 1.4.When the system parameters Ro, Xo and XI are known,the value of GFF may be determined more accurately bythe use of Figure 20.16, which has been established fordifferent grounding conditions, Le. for different valuesof Ro. Xo and XI.07651:uo 32101 2 3 4 5 6 7 %X,,/Xl -R,, = Zero sequence resistanceX,, = Zero sequence reactanceXl = Positive sequence reactanceFigure 20.16 Relationship between RdX1 and XdX, to determinethe values of ground fault factor kg based on IEC20.7 Magnitude of temporaryove rvo I t ag esAt a particular location this can be obtained by multiplyingthe peak system phase to ground voltage by the groundfault factor at that location. For example, for a 6.6 kVnominal voltage system, having an isolated neutral(ungrounded system), the maximum temporary overvoltageduring a ground fault may rise toVph(peak) x GFFwhere Vph (peak) = fi x6.6- kV = 1 p.u.and GFF =afor an ungrounded system:. Temporary overvoltage = 1 pa. x 8= 4p.u.20.8 Insulation coordinationFor non-effectively grounded systems, having a GFF ofmore than 1.4, a higher level of insulation (BIL) will beessential for all equipment being used on the system towithstand a higher level of a one-minute power frequencyvoltage test as well as an impulse voltage withstand testif such levels (Lists I, I1 and 111) are prescribed in therelevant standards. If not, then it may be assumed, thatthe prescribed test values take account of such an

20/668 Industrial Power Engineering and Applications Handbookeventuality. Refer to Tables 11.6, 14.1, 32.1(A), 13.2and 13.3 for more details.20.9 Application of different typesof grounding methods (for HT,HV and EHV* systems)In the preceding paragraphs we have analysed thebehaviour and characteristics of a system when groundedor left isolated. The ground fault factor (GFF) plays avery significant role in the selection of insulation level(BIL) and its coordination with the different equipmentconnected on the system. The application of a particularmethod of grounding would thus depend uponThe ground protection scheme envisaged to decide onthe magnitude of the ground fault currentCriticality of the supply system, e.g. whether an immediatetrip on fault is permissibleInsulation level of the main equipment connected inthe system.The grounding of a generator, for instance, whichmay be designed for 6.6, 11, 15 or 21 kV, and all otherequipment connected on this system may be solidlygrounded to have the least GFF and hence, add no extracost to the machine for a higher level of insulation or alarger size. On the other hand, at such voltages the groundfault current may also not be excessive to be of concernto cause an extra burden to the windings of the machineor in the selection of protective devices. It is, however,noted that a few application engineers may prefer anisolated neutral system at certain installations wherecontinuity of supply is mandatory, even on a groundfault, until an alternative arrangement is made. Examplesare auxiliary drives in a power generating unit or essentialdrives in a process plant. At such installations, it isimperative to ensure that the generator and all theequipment connected in the system are designed for thehigher GFF and a larger size or greater cost of the machinesare immaterial. But the occurrence of another groundfault before clearing the first will lead to fatality andcause total damage of the faulty equipment. The groundcurrent can now find its way through the earlier groundfault and cannot be prevented, as there is no protectionavailable. Isolated systems are therefore generally notrecommended. Instead, for such require-ments, aresonance grounding system may be adopted, limitingthe ground fault current to a desired low value to protect*I We use HT for all voltage systems above 1.1 kV unless acomparative reference is necessary.2 To identify the windings of a transformer, we use HV for thehigher voltage and LV for the lower voltage side.3 When we refer to a transmission system, we classify the differentmaximum voltage systems as follows:HT - up to 66 kVHV - above 66 to 245 kV andEHV - above 245 kV.the machine from heavy fault currents and prevent thesystem from tripping on a ground fault. Such a system ismore prevalent in overhead transmission or longdistributionnetworks to save the whole system from anoutage on a ground fault.The more recommended practice is to ground the neutralsolidly or through an impedance, commensurate with therequirements of the protective scheme and the fault currentlimited to a desired level. The terminal equipment and thewindings of all the machines may now be designed for avoltage corresponding to the relevant GFF.20.9.1 Artificial neutral grounding of athree-phase three-wire systemIn the previous section we have discussed the theory ofproviding a ground fault protection when a neutral wasalready available on the system. This could be utilizedfor a solid or an impedance grounding to achieve therequired level of fault current on a ground fault, andmeet the requirement of the protection scheme or thestability of the system. Here we discuss circuits whichdo not have a neutral as a matter of system design, suchas for the purpose of transmission and long HT distribution,where the power is transmitted on a delta circuit toeconomize on the initial cost, such as from the generatorto the generator transformer (Figure 13.15) where thegenerator is star and the transformer can be YlA, say,15.75/400 kV. In such a case, when grounding is requiredon the delta side of the transformer, this is possible bycreating an artificial neutral point. The basic need forsuch a provision, where a neutral does not exist, may benecessary, primarily to achieve the following:1 To reduce the high-voltage transient oscillations in anisolated neutral system and to prevent the voltage ofthe healthy phases from rising beyond their line toneutral voltage, as far as possible. Prolonged existenceof overvoltages may have a tendency to cause a shortcircuitfrom line to ground, even in the healthy phases.2 A ground fault protection scheme that is easy to handle,clear the fault quickly and prevent it from spreading.3 To eliminate prolonged arcing grounds as a matter ofsafety to human lives. A live conductor falling onground will remain live if not grounded and cause anarcing through ground leakage capacitances. It maygenerate excessive heat and become a hazard to lifeand property.4 On higher voltage systems, due to ground leakagecapacitances, the voltage of the two healthy phasesmay increase to twice the voltage, similar to doublecharging, when switching a capacitor unit (Section23.5.1).5 Electrostatic induction may take place on overheadpower-carrying systems, through charged clouds, dust,rain, fog and sleet and due to changes in the altitudesof lines. If these induced charges are not freed throughgrounding, they will continue to rise gradually andaccumulate on the system. This is called floatingpotential and may result in a breakdown of the systeminsulation or the terminal equipment.

Temporary overvoltages and system grounding 20/669If the above are not taken into account the devicesand components used on the system may have to beselected or braced for a higher system voltage, say, up totwice the rated voltage or even more. The overvoltagecondition may almost be the same as for ungroundedcapacitor switching (Section 23.5.1 (ii)).Neutral grounding transformersThe artificial neutral point on the delta side can be createdby providing an interconnected star neutral, also knownas a iig-zag transformer. It can also be provided througha star-delta transformer. Both arrangements are illustratedin Figures 20.17 and 20.18 respectively. Such transformersare of a standard core type, with only a single windingon each transformer limb, split into two halves as shown.The total winding arrangement is like a 1:1 autotranstormer,with the provision of altering the groundimpedance, similar to that in a normal ground circuit, asdiscussed above. The additional resistance or impedance,as required in the ground circuit, may be provided eitherby inserting it between the neutral point and the groundor in the three individual phases as shown in Figures20.17 and 20.18.The rating of the auxiliary transformer can be shorttime.sufficient to feed the fault current and its own noloadlosses fot, say, 30 seconds or so, according to themaximum tripping time of the protective scheme. Thenormal rating of such transformers is generally between5 and 100 kVA, sufficient to carry the full ground faultcurrent. The winding is designed for line-to-line systemvoltage.NoteThe magnetic core is designed so that it will not jaturate duringnormal operation to avoid a ferro-resonance condition. The kneepoint voltage is kept at about I .3 times the system line voltage.Residual voltage transformer (star/open deltatrans former)This is not a method of providing an artificial neutral. asin the previous case, but to detect an unbalance or residualvoltage (zero sequence voltage) in a three-phase threewireor a three-phase four-wire ungrounded system. Theresidual or zero sequence voltage that may appear acrossthe open delta will be the reflection of an unbalance or aground fault in the system (Figure 20.10). Also refer toSection 15.4. I for more details.20.1 0 Important parameters forselecting a ground faultprotection scheme1 Single-phase loads or unevenly distributed loads on abalanced three-phase system do not lead to a faultycondition as a result of unbalanced currents. The singlephaseload currents will always have the return paththrough the neutral and not the ground. The currentthrough the ground circuit will flow only when thereis a ground fault and the circuit completes throughthe ground loop. Refer to Figures 20.19(a) and (b).Delta connectedtransformerwindingsDelta connectedtransformerwindinas(a) Current-limiting resistor in neutral (bj Current-limiting resistors in line (c) Phasor representation ofthe grounding transformerFigure 20.1736 interconnected star-neutral grounding transformer

"(201670 Industrial Power Engineering and Applications HandbookDelta connectedDelta connectedtransformertransformerwindings IQ windings 1,Neutral may beresistor groundedor solidly groundedG(a) Current-limiting resistor in neutral(a) Healthy system(b) Faulty systemFigure 20.19 In a healthy system the unbalanced current(other than a ground fault or phase to phase and ground fault)will flow through the neutral and not the ground(b) Current-limiting resistors in lineFigure 20.18 34 Staridelta neutral grounding transformer(c) Phasor representation ofthe grounding transformer2 The magnitude of the ground fault current is a matterof system design and will largely depend upon thesystem voltage and the ground loop impedance, asdiscussed above. The theory of grounding protection,however. is different for an LT and an HT system.While an LT system will require a reasonably highground fault current and thus a low ground circuitimpedance to detect the fault promptly, an HT systemmust have as low a ground leakage current as possibleto avoid a dangerous potential gradient at any pointon the ground loop and the ground (step and touchvoltages, Section 22.9) and to eliminate the danger toa human coming into contact with it and also to preventarcing grounds. HT systems as standard practice aretherefore designed to have a high ground circuitimpedance.3 The required ground impedance may be determinedon the following lines, ifkVA = rated capacity of the supply source, whichcan be a generator or a transformerVi = system rated line voltage in voltsI,. = system rated full-load line current in amperesZg = impedance of the grounded neutral circuit inohmsI, = required level of the ground fault current inThenz, = ~amperes.\15 . I,R1000 'and I, =kVaAmp47. v,IF is generally defined in terms of Ir, such as 1040%or 20-80% of Ir, depending upon the protection scheme.If I, is, say, n in units of I, then

Temporary overvoltages and system grounding 20/671z,,V,?os =11,n(20.6)1000. kVAFrom this equation one can determine the requiredcalue of neutral circuit impedance for a particularlevel of ground fault current. The external impedancewill be Z,, less the ground impedance. In HT systemsone can also determine the likely value of a groundinductor coil to achieve a near-resonance condition,to eliminate the arcing grounds, on the one hand, andfacilitate a strike-free extinction of an arc by theinterrupting device, on the other.Example 20.1For a 1600 kVA, 11/0.415 kV transformer, considering the LTsidez, =(0 415)' x 1000'nx1000x1600For an industrial power distribution network, if the setting ofthe protection relay is considered to be 20% of I, then(0 415)'~ 10002 ~I' - 0 2 x 1000 x 1600z -= 0 54 CLThe natural zero phase sequence inductive reactance of thegrounded neutral may be considered to be too small comparedto this and ignored for ease of calculations Thus the resistanceRo of the grounded neutral circuit may be considered as itsimpedance, i e0I ,415 x 1000=L 3 x 0 54== 444 AmpsThe impedance calculated thus will form the basis ofdetermining the adequacy of the grounding stations provided.Probably, for such a low value of grounded neutral impedancethe grounding stations may have to be more elaborate andgreater in numbers (arranged in parallel). This is to ensurethat at no stage will the impedance of the system increasebeyond 0.54 R (inclusive of the impedance of the ground).Otherwise it will render the protective scheme ineffective andallow the ground fault to persist.The above is the case for an LT system. For an HT system,similar calculations may be carried out to determine the groundneutral impedance. Now it will be much higher than the abovedue to a high Vg to limit the fault current and thus also theground potential difference to below a dangerous level at anypoint on the ground circuit. The impedance of the groundcircuit in such cases may be increased through a resistanceor an inductor coil.20.10.1 Grounding of generatorsSolid groundingGenerators that are solidly grounded have a differentgrounding practice from others due to their zero phasesequence reactance. which is much less than its positiveor negative phase sequence reactances (Section 13.4.1(5))and Table 13.6). As a result, the ground fault current ina generator circuit is greater than its three-phase symmetricalfault current. This current rises further when all ofthem are individually grounded and more than one unitare running in parallel at a time. It is worth mentioningthat when two or more generators are running in paralleland all of them are grounded, they form a closed circuitto cause circulating currents. This may occur even in ahealthy system due to unbalance, not because of singlephaseloads but unequal generator phase currents due toeddy currents. These are important aspects and must beconsidered while deciding on the grounding method of asolidly grounded generator. Consider Figure 20.20,illustrating four identical generators operating in parallel.Each has the following reactances:positive phase sequence instant p.u. (sub-transient)reactance = XCnegative phase sequence p.u. reactance = x2zero phase sequence p.u. reactance = xo:. Total p.u. reactance of the ground circuits = x;' + X? + .XgFor the sake of illustration, consider the values of thesereactances as in Example 13.4:xy = 11.7%x2 = 8.89= 4.0%The instant (sub-transient) fault current, lssc, through agenerator in a symmetrical three-phase system, irrespectiveof the condition of neutral as defined in Table 13.9 willbelist (3-$ fault peak value) =where v = p.u. voltageV21

0.088201672 Industrial Power Engineering and Applications Handbook8 . uand ZSsc (phase-to-phase fault) = ~x: + xz- 43. ti- 0.117 + 0.088= 8.45uIf there is more than one generator operating in parallel,when the fault occurs, the reactances of all such machineswill fall in parallel and diminish the effective reactanceof the faulty circuit and enhance the fault current. Forinstance, when all four machines are operating in parallel,the effective reactance will becomeand the enhanced fault currenti.e. 4 x 8.5511 or 34.2~Generalizing the above, the symmetrical fault current,when n number of machines are operating in parallel,2)Issc = 7. ’2xdThis is irrespective of whether the generator neutral isisolated or solidly grounded, and if grounded, whetheroperating singly or in parallel, whereas the ground fault,when the generators are grounded but only one machineis operating at a time,3u 3.vI =-=E x x:+x2+xoand for the values of reactances considered;I, =3ti0.1 17 + 0.088 + 0.04= 12.24~(Table 13.9)(which is much higher than 8.55~011 a symmetrical threephasefault). When more than one machine are operatingin parallel at a time and all of them are individuallygrounded, the effective reactances for n number ofmachines will bex 0x, = - nReactances with the suffix ‘e’ denote the equivalentreactances when all the machines are grounded andrunning in parallel1and x, = - (x; + x2 + x,) )nand the new fault current will be(20.7)and for the impedances assumed, when all the fourmachines arc running in parallel= 12.24 x 4~or 48.96~ (as against 34.2~ on a symmetrical three-phasefault).The ground fault circulating current is now as manytimes more, as many grounded machines will be operatingin parallel. Such high ground fault currents may causedisturbance in the supply system and severely damagethe machines. This can be reduced by grounding just onemachine at a time, as illustrated in Figure 20.20, andadopting the normal protective scheme. Equation (20.7)will now be modified tox:where x:~, = - nx?x2,, = -andxoc, =xgReactances with the suffix ‘el’ denote the equivalentreactances when all the machines are running in parallel,but neutral if only one machine is grounded at a time.(20.8)n n ”For the reactances assumed, when all the four machinesare running in parallelI” = 3v0.117 ~4 4= 32.88~+,04which is much less than before and even less than thethree-phase symmetrical fault current. To restrict theground fault current, when more than one generator areoperating in parallel it is advisable to ground only onemachine at a time. Although provision must be made forall the machines for grounding, so that when one ormore machines are out of service, one of those in servicemay be grounded. To achieve this, all the neutrals maybe connected to a common neutral bus through individualneutral grounding switches as shown in Figure 20.20,and the common neutral grounded solidly.Impedance groundingThe above problem can also be overcome by impedancegrounding rather than solid grounding. It can be aresistance R or inductance L or both, as discussed above.In the present case, if we consider a p.u. resistance r ofjust 9% in every neutral, the improved ground fault current

Temporary overvoltages and system grounding 201673will become as follows when all the four generators aregrounded and are operating in parallel and one of themdevelops a ground fault:I"=-g- 3v - -x4xz+ x2 + xo + 3r-4 x 3vJ(O.117 + 0.088 + 0.04) + 3 x 0.09- 4 x 3v0.27 + 50.245---4 x 3v0.365or 32.88~This is less than the corresponding symmetrical faultcurrent of 34.2~ and incidentally equal to the groundfault current when only one machine is grounded at atime. By this method, the ground fault current can becontrolled to any desired level. We have considered aresistance with a view to improving the p.f. of the faultcurrent and thus, making it easier to interrupt.Level of ground fault current for large generatorsManufacturers of large generators, 200 MW and above,recommend the ground fault current, Ig to be limited inthe range of 5-15 A and a fault clearing time of the orderof 5-30 seconds to protect the machine and avoidoverheating of the grounded steel frame. It is alsoGeneratortransformerTo gridII___________,________---l3.1,I, - Ground and other leakage currents per phase.Is, - Ground fault current through the generatorR - Secondary grounding resistanceFigure 20.21 Grounding method and flow of ground faultcurrent in a generator on a ground fault (Example 20.2)recommended to limit the TRV by inserting a smallresistance into the grounding circuit to make the groundfault current (Ig) somewhat resistive than capacitive dueto capacitive coupling between the generator and theassociated equipment and the ground as illustrated inFigure 20.21. To achieve this, the ground fault lossrepresented by R (Igr being the active current and Rthe ground resistance) should be higher than the electrostaticloss to ground, as explained in Example 20.2.Since a reactor can only offset or over-compensatethe capacitive kVA, it will not yield the same result as aresistance. Resistance grounding is therefore preferredto reactance grounding. The GFF, however, will now behigher and may rise to fi times. The phase to neutralvoltage in the healthy phases may rise to the line voltageduring a ground fault, as in an isolated neutral system.Machine insulation and all equipment and devicesassociated with the machine must take care of this. Thelow-resistance grounding may be achieved through adistribution transformer, with a low resistance on thesecondary side, as shown in Figure 20.21.Example 20.2To determine the grounding parameters, consider a generatorrated for 200 MW, 15 kV and the ground fault current limitedto 15 A. Considering GFF as 6, the voltage ratio of thegrounding transformer with a 220 V secondary will be15, & kV:220 V6or 15 kV:220 VConsider safe /gr as 10 A.1. Electrostatic kVAConsider the following ground and other leakage capacitancesfor the sake of reference:Generator to ground = 0.5 pFGenerator bus duct to ground = 0.15 pFLow-voltage winding of GT to ground = 0.007 pFSurge capacitance = 0.2 pF:. Total coupling capacitance = 0.857 pFand coupling reactancex cg - 2n x 50 x 0.857 x lo4= 3.72 x 103 n1(for a 50 Hz system)15 000and coupling current IC, = &i x 3.72 x io3and electrostatic kVA= 15 x 6.99d3= 60.53 kVAr2. Ground fault lossSecondary current on faultv, * I, = v, . /2= 2.33 Nper phase3 . ICc = 6.99 A for three phases

~~~~ ~~201674 Industrial Power Engineering and Applications Handbook.. 12 = 682 Aand required secondary resistanceR = g682= 0.32 R... /ir R = (682)' x 0.32= 148.8 kVA:. Iir R >> kVArand actual current at the Doint of faultI, = 2/10' + (2.33 x 3)'= 12.2AThis is illustrated by a phasor diagram in Figure 20.22./gr = 10 AI, = i,, + 3. iccFigure 20.22 Phasor diagram for the ground fault current ofExample 20.23. Rating of transformer= 15 x lo3 x 10 VA= 150 kVA. 15 kVl220 V,~Short-time rating 5-10 minutes, although the relay will tripmuch earlier.Relevant StandardsIEC Title IS BS6007 1 - 1/1993 Insulation coordination - Definitions,principles and rules2 165- 1/199121 65-2/199 1BS EN 60071-1/199660071-2/1996 Insulation coordination - Application guide 37 16/199 1 BS EN 60071-2/1997ANSVIEEE13 13.1/1996Relevant US standards ANSVNEMA and IEEEInsulation coordination, definitions, Principles and Rules__ .Notes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.23Some of the BS or IS standards mentioned against IEC may not be identical.The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedWhen the external impedance is capacitiveAnalysis of an ungrounded systemWhen the external impedance is resistive(20.1)Vi = voltage across the external resistance RR = external resistanceXcg = ground capacitive reactanceVe = line voltageVi = voltage across X,X, = external capacitive reactanceWhen the external impedance is inductiveXL = external reactance(20.2)(20.3)

Temporary overvoltages and system grounding 201675Arc suppression coil or ground fault neutralizerTo substantially offset the capacitive current by the powerfrequency inductor current1- = inductor coil inductance in henry= ground capacitance per phase in farad andI' = system fi-eclucncy in HzGround fault factorI, = required level of the ground fault current inAmperesGrounding of generators(i) When all the machines are inclividiially groundedu = p.u. voltage of generators;' =positive phase sequence ~ .LI.(sub transient) rcaclancex2 = negative phase sequence p.u. rexIance(ii) When only one machine is grounded a1 ;I timeI/, = highest voltage to ground (r.m.s)Selecting a ground fault protection schemez=\',QII 1000 kVA(20.6)kVA = rated capacity of the supply sourceC I = system rated line voltage in voltsI, = system rated full load line current in Amperes/ I = unit of /,Z1 = impedance ofthe grounded neutral circuit in Ohms.q, = zero phase sequence p.u. reactatice/ I = no. of machines operating in parallclFurther reading(20.8)

Grounding Theoryand Ground FaultProtectionSchemes

Grounding theory and ground fault protection schemes 21/67921.1 Protection of a domestic or anindustrial single-phase systembody can endure, without damage has a relationship withthe leakage current through the body and its duration,i.e.21.1.1 Effects of current passing through ahuman body and the body’s tolerablelimitsAn electric current, rather than voltage, through a humanbody may cause shock and can damage vital organs ofthe body as follows:It may cause muscular contraction, unconsciousness,fibrillation of the heart, respiratory nerve blockageand burning. These are all functions of body weight.The muscular contraction makes it difficult to releasean energized object if held by the hand and can alsomake the breathing difficult. The heart, being the mostvulnerable organ of a human body, is damaged most,mainly by ventricular fibrillation, which may resultin an immediate arrest of the blood circulation (electrocution).In Table 21.1 we provide likely intensities ofbody currents when lasting for more than a heart beat(nearly 60-300 ms or 3-15 cycles for a 50 Hz system).It is generally seen that a human body can sustain amuch higher current at a lightning or switchingfrequency (5 kHz or above) due to the extremely shortduration of such surges (30 ps or less).The current can pass through the heart, when the currentpasses through hand to hand, or through one hand and afoot. Current flowing between one foot to the other maynot be considered dangerous, but may cause muscularcontraction and pain. The subsequent body fall, if it occurs,may, however, be fatal as now the current can also flowthrough the hand involving the heart. Ground faultprotection emphasizes keeping the fault current belowthe fibrillation threshold and for a period of less than aheart beat, in the range of 60-300 ms. It has beenestablished that the electric shock energy which a humanTable 21.1 Likely intensities of body currents whenlasting for more than a heart beat (= 60-300 ms)Sr. Body At frequency Intensityno. current(d) (HZ)1 100 50-60 Lethal2 100 3- 10 kHz Tolerable (because of3 60-100 50-60extremely short duration)Fatal4 16-60 50-60 Breathing may becomedifficult due to muscularcontraction5 10-16 50-60 It is the let-go current andmakes it hard to release anenergized object if held byhand6 5-9 50-60 No dangerous effects7 1 4 50-60 Threshold of sensation(21.1)wheresb = shock energy in watt-seconds (Ws)Ib = r.m.s value of the leakage current through the bodyin Amperes (A)tb = duration of leakage current in seconds (s)The energy, S, for a 50 kg body is regarded safe at0.0135 Ws and for a 70 kg body at 0.0246 Ws0.0135 :. I, (50 kg) = /?0.116--and I, (70 kg) = /yKOE- = -(21.2)(21.3)Figure 21.1 as in IEC 60479-1 illustrates the variouszones of ground leakage current versus duration of faultand its effect on a human body.To enable the body to sustain a high fault current it isessential that the fault interrupting device (relay or release)is quick responding. For domestic applications, it isrecommended to be less than a heart beat.NotePrevention from ground leakage as such is important, as it causescorrosion through electrolysis and may damage the insulation ofwires due to ageing.21.1.2 Use of ground leakage circuit breakers(GL CBs) *It is likely that smaller installations, such as a domesticlight or power distribution network, and single-phaseindustrial or light loads may not always meet therequirements, as discussed in Section 21.2, Table 21.2,due to the circuit’s own high impedance (other than theimpedance of the ground conductors). It is thereforepossible that the fuses, if provided for the short-circuitprotection of the system, may be too large to detect aground leakage. The ground leakage circuit breakers(GLCBs) provide an effective solution to this problemand are easily available and extensively used. They areextremely sensitive to very feeble ground leakages, ofthe order of 10 mA or more, and trip the faulty circuitwithin a safe period on the smallest leakage current andprovide effective protection to the human body. Figure21.2 illustrates the tripping scheme for such a breakerand the principle of operation is discussed below. Normallytwo types of GLCBs are in use, i.e. 30 mA and 300 mA.*More commonly known as earth leakage circuit breakers (ELCBs)or residual current circuit breakers (RCCBs). They operate on theprinciple of residual current.

21/680 Industrial Power Engineering and Applications Handbook100005000Effects of A C currents (50 Hz) on a personat least 50 kg in weight2000Zone-1 - Usually no reactlonZone-2 - Usually no dangerous physiological1000 effect500Zone-3 - Usually no danger of heart failure(ventricular fibrillation)Zone-4 - Probability of heart failure up to50%1 :::Zone-5 - Probability of heart failure more3 50than 50%s20E 10K5210 1 02 05 1 2 3 5 10 20 30 50 100 200 500 1000 2000 5000Current passing through the human body in mA (rms) +Figure 21.1EffeCcts of current leakage through a human body-- ~ --- - --------R- 7and electrocution, while the latter protects the circuitIYfrom fire risk. Normally the 300 mA GLCB is used asIthe incomer and 30 mA as the outgoing for the individualIB feeding circuits.I-a-I4-III1IIZ, -Ground fault loop impedanceIA - Leakage current through the appliance But for simplicity thewhole current is considered through the body onlyif -Leakage current through the human body* -Test button to check periodically the operation of the ELCBFigure 21.2(a)Schematic drawing of an ELCBRefer to Figure 2 1.2 showing a single-phase circuit breakerhaving a thermal overcurrent and a magnetic short-circuitelement, OCR. A core-balanced transformer is arrangedin the circuit as shown. When the system is healthy, thecurrent through the phase, I,, and the neutral, I,, will beequal and the magnetic effect of the load current on theprimary windings, P, and P2, will be almost zero. In theevent of a ground leakage, I(, this balance is disturbedand some of the leakage current will also flow throughone of the primary windings of this core-balanced CT, asshown in the circuit. This will cause a magnetic flux inthe core, inducing a voltage in the secondary winding S.In the circuit shown, it feeds a rectifier circuit providedon the secondary side of the circuit. This unit facilitatesa direct current flow through the leakage trip coil, whichtrips the breaker. These breakers are extremely fast operatingon fault (operating time may be le2s than 5 ms), andhave a low let-through energy. Refer to I- - t characteristics(Figure 21.1). The ground fault current will also becomprised of a current through the appliance, I, (Figure21.2). But since the basic purpose of this device is toprotect the human body from an electric shock orelectrocution, we have considered the leakage current,I(, through the body alone to be safer. To ensure totalsafety, the device should be able to detect this leakage,assuming there was nu ground current through theappliance. When the device will be able to do so, thecircuit in any case will be protected.

Grounding theory and ground fault protection schemes 21/6811. Core(Courtesy: MDS switchgear Ltd.)2. Primaly winding3. Relay4. Ground fault indicator5. Test button tripping6. Outgoing terminalthey can be interchanged7. Incoming terminal(Courtesy: English Electric)Figure 21.2(b) Two-pole ground leakage circuit breakers (ELCBs or GLCBs)21.2 Ground fault on an LT system21.2.1 System protected through over currentreleases and HRC fusesThe value of ground loop impedance is always predetermined,depending upon the system requirement and thetype of protection available to the system, its accuracyto detect the fault and time to operate. For systemsprotected through overcurrent releases or HRC fuses only,the ground loop must have a comparatively lowimpedance. It will allow a high ground fault currentthrough the faulty circuit, sufficient to trip the over currentcum-short-circuitreleases of the breaker if the circuit isprotected through such releases of the breaker or blowout the HRC fuses, if the circuit is protected throughHRC fuses. Such a requirement is more desirable forhigher rating systems, where discrimination between ahealthy and a faulty condition by such devices may bedifficult. Medium-rating systems may cause a relativelymuch higher fault current and be automatically protected,as the normal ground fault current would be sufficient totrip the short-circuit releases or blow out the HRC fuses.The rule of thumb to determine the ground loopimpedance is to consider the ground fault current as oneand a half times that of the overcurrent setting of the circuitbreaker for breaker-controlled systems (a fault conditionfor a breaker) or three times the rating of the fuses, forfuse-protected systems (an overcurrent condition for thefuses). Based on this rule, Table 21.2 suggests the optimumvalues of ground loop impedances for circuits of differentcurrent ratings for an LT system. At these values of currents,the overcurrent releases will trip in about 130-370 seconds.Refer to ‘Z2 - t’ characteristic curves of such releases asshown in Figure 21.3. The HRC fuses will blow out inabout 40/60 seconds. Refer to ‘Z2 - t’ characteristic curvesof fuses, as shown in Figure 21.4.Table 21.2 Maximum impedances of ground loop, when protectedby overcurrent releases of circuit breakers or fusesCurrent rating of Overcurrent of Maximum desirablecircuit fuse the circuit breaker impedance of groundMCCB or ACB fault loop on a 240 V‘a’ ‘b’ circuit (415/&)z,240= ~3a or 1.5bR510152030406080100125150175200300400 etc.102030406080120160200250300350400600800 etc.1685.342.721.3310.80.640.530.460.40.270.2

21/682 Industrial Power Engineering and Applications HandbookNoteGround fault protection through overcurrent releases or HRC fusesis not a reliable practice. It requires a very low ground circuitresistance to maintain a high ground fault current, which may notbe possible in the long run. As a consequence, in certain cases thesystem may remain intact on a ground fault and damage the healthycircuits. It is therefore good practice to provide a ground leakageMCB, ground leakage MCCB or a conventional MCCB with aCBCT and a ground fault leakage relay for low-rating incomingfeeders.For domestic light and power distribution this practice is common.Modern industrial installations can now adopt a fuseless systemand use MCCBs and provide a more sensitive ground leakageprotection scheme for more safety and reliability.21.2.2 System protected through relaysA system protected through a separate ground fault relaywill require a different concept from that discussed above.A relay can beI -0 Electro-magnetic (being quickly outdated)0 Static, based on discrete ICs or0 Solid-state microprocessor based.2221 1.2 ' 2 3 4 5 6 7 8910150% lrFigure 21.3 /'-t characteristics of an OCR (Figure 12.13(a),reproduced for illustration) (Courtesy: L & T)The relay is very sensitive and quick responding and canreliably actuate at low leakage currents (0.1 A or less). Itis therefore operated through the secondary of a CT (1 or5 A), provided in the ground circuit. Figures 21 S(a) and

Grounding theory and ground fault protection schemes 211683System or equipmentunder protectionRYB1 ,.q &+ 44 I + 4-G - G = G -G(a) (b) (4(a) Only for a ground fault (G/F) protection(b) For 3-0/C and l-G/F protections(c) For 2-0/C and l-G/F protectionsFigure 21.5(a)Unrestricted G/F protection schemes for three-phase three-wire systemsSystem or equipmentunder protectionRYBNFigure 21.5(b)=G -G(a) (b) (C)(a) Only for a ground fault (G/F) protection(b) For 3-0/C and 1-G/F protections(c) For Z-O/C and l-G/F protectionsUnrestricted G/F protection schemes for three-phase four-wire systems(b) show different methods of CT connections for a groundfault protection. The fault current in this case need notbe hery high. It should rather be limited to a low value,to limit damage to the equipment. As a rule of thumb itmay be up to the rated current of the feeding transformer,or the rating of the incoming feeder, or twice the ratingof the largest outgoing feeder, depending upon the circuitto be protected.The most common practice, however, is to limit theground fault current to only half the rated current of thesystem, or the circuit that is being protected. This is alsoin line with the universal practice of having the neutralof half the size that of the phases. The neutral is normallygrounded to form a complete circuit through the groundconductor in the event of a ground fault. Refer to Figure2 1.6. showing a typical distribution network illustratingthe grounding circuits.The preferred normal settings in a ground fault relayare 10-40% or 20-80% of the rated system current,depending upon the application. For a balanced load, asetting of, say, IO-20% is sufficient. For a system havingmore single-phase light or power loads, a higher settingwould be necessary, depending upon the likely singlephaseloads. This is to avoid false tripping on a healthysystem, as all the out-of-balance current will flow throughthe neutral only.The theory of operation of such a protection scheme isbased on the principle that in a balanced circuit the phasorsum of currents in the three healthy phases is zero, asillustrated in Figure 21.7, and the current through thegrounded neutral is zero. In the event of a ground fault,i.e. when one of the phases becomes grounded, this balanceis upset and the out-of-balance current flows through thegrounded neutral. A healthy three-phase circuit. however,

~21/684 Industrial Power Engineering and Applications Handbookbalance of the system. Similarly, on a single phasing,although the balance is upset, there is no current throughthe ground circuit and this scheme would not detect asingle phasing.rG'21.3 Ground fault protection inhazardous areasThese are highly sensitive areas and a little higher levelof a ground fault current can be catastrophic. It is thereforemandatory at such locations to keep their ground leakagecurrent (rather than the ground fault currents) low bymaintaining a certain level of ground loop impedanceand then be able to detect and isolate these currentspromptly.The leakage current at hazardous locations such asrefineries, petrochemical plants and mines should notexceed 15% of the rated current of the circuit or 5 A,whichever is greater. Table 2 1.3 indicates the maximumpermissible ground leakage currents for such areas at15% of the rated current and the recommended maximumground loop impedances.The use of core-balanced CTs is quite common forsuch applications. They are specially designed to detectthe ground leakage current of a circuit. This ground leakageis then used to trip the faulty circuit through a grounddoes not necessarily mean that each current phasor R, Yor B individually is equal in magnitude and phase. Evenif it has unbalanced currents in the three phases (whichis a likely situation in a three-phase system, see alsoSection 12.2(v), it will not cause a current to flow throughthe ground circuit, as illustrated in Figure 20.19(a). Thecurrent through the ground circuit will flow only whenany one or more of the phases has a ground fault andforms a complete circuit through the ground loop conductor,as illustrated in Figure 20.19(b), and disturb theTable 21.3 Maximum impedances of ground loop for protectionby ground leakage relays in hazardous areasFigure 21.6 An HT to LT distribution system showingCurrent rating Maximum permissible Recommended maximumgrounding circuitsof circuit ground leakage ground loop impedanceA>kcurrenP 'C'on a 240 V phase toAmp Amp ground circuitbz, =-240I S xc5 5 3210 5 3215 5 32/E IY 220 5 3230 5 3240 6 26.760 9 17.8IR = Iy = /B80 12 13.3100 15 10.7_ . .:. IR + / y + Is = 0I50 22.5 7.1125 18.7 8.6Figure 21.7 Phasor sum of a balanced three-phase system is 175 26.2 6.1zero200 30 5.3300 45 3.5400 60 2.7NotesaHighly sensitive ground leakage relays can sense a current as lowas 0.1 A and less and at a much higher ground loop impedance.bCalculated to allow at least 150% of the maximum permissiblecurrent to be on the safe side. For example, for 15A maximumground leakage current, maximum impedance at 240 V--= - 240 10.7 !21.5 x 15

Grounding theory and ground fault protection schemes 21/685Figure 21.8 Schematic of a core-balanced CT (CBCT)fault relay. wired at the secondary of such CTs, as shownin Figure 21.8.The theory of operation of a ground leakage circuitbreaker (GLCB) is also the same as the combination of;I core-balanced CT and a ground leakage relay. Forindustrial application, use of a core-balanced CT with aground leakage relay and for domestic application use ofa GLCB is more common.21.4 Ground leakage in anHT systemIn HT systems, the situation is different from that in LTdue to high phase to ground voltage Vg. It is now capableof causing high ground fault currents. To contain thesecurrents in such a system the normal practice is to increasethe ground impedance by inserting some impedance intothe ground circuit through the grounded neutral. Thecriterion of choosing a resistance or a reactor is discussedin Section 20.4.2. An HT system at the main distributionpoint is normally protected through relays for differentprotective schemes and not through fuses or overcurrentreleases. Since a relay is much more sensitive and isquicker to respond, even a small ground leakage currentwill be enough to actuate it and isolate the faulty circuit.The ground loop impedance in an HT system is thereforekept much higher than on an LT system. Moreover, sincemodern relays are able to sense a current as low as 0.1 Aor less, the universal practice is to have as large a groundloop impedance as possible to limit the ground leakagecurrent, such that under no conditions will the touch andstep voltages exceed the permissible tolerable levels asdefined in Section 22.9.or impedance, to limit the ground leakage currents, mayalso require this type of fault detection.A core-balanced CT is in a toroidal (circular) orrectangular form, like a conventional protection CT, exceptthat it is designed, with a large core opening toaccommodate all the 3-, 3i-or 4-core feeder cablespassing through it (Figure 21.8). The ha+ differencebetween this and conventional protection CTs is the lowunbalance or leakage current at which a CBCT operates.A normal protection CT would operate between its ratedand accuracy limit currents, as discussed in Section 1S.6.S.The important design parameter in a CBCT is its magnetizingcurrent at the relay operating voltage. rather thanthe class of accuracy and accuracy limit factor.21.5.1 Design parameters for a CBCTAs these CTs have to detect small to extremely feebleground leakage currents their operating region is requiredto be very low, near the 'ankle point' (almost IO% of theknee point voltage) on the magnetizing curve. The designcriterion is thus the minimum exciting current requiredat the relay operating voltage (details available from therelay manufacturer) to actuate the relay.Consider the equivalent tripping circuit diagram ofthe CBCT in Figure 2 1.9. IfRcT = resistance of the CBCTRL = resistance of the connecting leadsR, = resistance of the protective relayI,, = relay operating currentV, = relay operating voltage at the relay currentsettingI, = primary unbalance or ground fault current_ - -= I, + I, -I I,n = turns ratio of the CBCTI,( = magnetizing (leakage) or no-load current ofthe CBCT at the relay operating voltage V,Then Vr= Ir,(RcT + R, + R,) and the primary groundfault current--I, = n(1, + J,,t)0Relay21.5 Core-balanced currenttransformers (C BCTs)These are generally employed to detect small amountsof ground leakage currents, such as in mines and othersensitive installations. They are also used to protectsensitive equipment against small ground leakages. Installationshaving isolated neutral or using ground resistanceFigure 21.9. . .lp = /R + / y + IB;, = Relay operating currentin, = CT no load or excitation currentEquivalent circuit of a CBCT protection circuit

21/686 Industrial Power Engineering and Applications Handbook- areA number of cables connectedto a common loadA number of cables connectedto a common load* The relay contactswired in seriesfor indication, alarmor trip circuit.(a) Separate CBCT for each cable.Figure 21.10(b) Separate protective circuit for each cableMethods to wire a protective circuit through CBCTsSince on a fault the power factor is normally low, Ire andIn( may be considered almost in phase, whenIp 2- n(zre + I n 0Since V, and In( are interdependent parameters theoptimum design is achieved when In( at Vr is of the samemagnitude as the relay operating current, i.e. In( 2- Ire.For high current systems, using more than one cablein parallel, the number of CBCTs will also be the sameas the number of cables, as illustrated in Figure 2 1.10, inwhich caseIP = n[/, + N . In(]for N number of CTsAll such CBCTs have to be identical in turns ratio andmagnetizing characteristics to avoid circulating currentsamong themselves. To order a CBCT the followinginformation will be essential:Minimum primary ground leakage currentNominal CT ratio. This may be such that on the smallestground fault the current on the secondary is sufficientto operate the relay. Normal I, = SO, 100 and 200 A. Itis recommended to be such that V, 2- 0.1 x knee pointvoltage of the CBCTRelay settingCT secondary current, 1A or SAMinimum excitation current required at the relayoperating voltageKnee point voltageNumber of cables in parallelLimiting dimensions and internal diameter (ID) of theCT. ID will depend upon the size of the cable.21.5.2 Insulation levelIrrespective of the system voltage, a CBCT may bedesigned for an insulation level of only 660 V. The cableinsulation of the HT conductor is sufficient to providethe required insulation between the conductor and theCBCT.21.5.3 Mounting of CBCTsThe following is the correct procedure for the propermounting of these CTs:1 It is necessary to pass all the 3, 3 i or 4 cores of thecable through the core of the CBCT to detect theunbalance or the ground leakage in 3-core cables andonly ground leakage in 33- and 4-core cables. Toexplain this see Figures 21.11(a) and (b). A 3-corecable will detect an unbalance in the three phases,whether this is the result of unequal loading in thethree phases or a ground fault. However, 33- or 4-core cables will detect only a ground leakage as theamount of unbalance, when it occurs, will be offsetby the flow of this unbalanced current through thereturn path of the neutral circuit. When using only 3-core cables, the load must be almost balanced otherwiseit will send wrong signals or a higher setting of therelay will become essential to account for the out-ofbalancecurrents due to the feeders’ unequal loading.2 In armoured cables, armouring must be removed beforepassing the cable through the CBCT to avoid an inducede.m.f. through the armour and the correspondingmagnetizing current which may affect the performanceof the CT.3 As such CTs are required to detect small out-of-balancecurrents, the connecting leads should be properlyterminated and must be short to contain the leadresistance as far as possible.4 For high-rating feeders using more than one cable,there must be one CBCT for each cable. Not morethan one cable must pass through such CTs. Thesecondary of all such parallel CTs may, however, beconnected in series, across the common relay (Figure2 1.1 O(a)).All CBCTs being used in parallel and intended forthe same feeder must have identical magnetizingcharacteristics and calibration in order to relay identicalsignals. Even then small variations in the output may

y the neutral circuit3-g load3-g four-wire load I,, - Unbalanced current due to load-AThis CBCT will detect the unbalance aswell as the ground leakage currentsThis CBCT will detect only theground leakage currentI,, - Unbalanced current due to load-AIub - Unbalanced current due to load-8Iga - Ground fault current of load-AIgh - Ground fault current of load-B(b) On a ground faultFigure 21.11Detecting the unbalance and the ground leakage currents through a CBCToccur which could affect the sensitivity of a circuitusing more than one CBCT. If such a reduction insensitivity is considered detrimental to the protectionof the system or the equipment, it is advisable to usea separate relay with each CBCT (Figure 21.10(b)).For a common indication, alarm or trip, the relays'trip conlacls may be wired in series.5 When using only single-core cables, the same methodshould be adopted, as discussed above. Groups maybe made of 3, 3; or 4 cores (R, Y, B or R, Y, B andN) of single-core cables, and one CBCT used in circularor rectangular form, whichever is more convenient,for each group.NoteIn an HT system it is recommended that each circuit he protectedseparately for a ground leakage as far as possible. This is tolocalize the effect of the fault and to trip only the feeder of thefaulty circuit. A common protection in the incoming is notadvisable, except for cost considerations, which also will benominal for such a protection scheme compared to the cost ofthe main equipment. Otherwise a ground leakage in thedownstream may trip the whole system. which may not bedesirable. It is also possible that at the downstream. the groundleakage is very feeble due to high ground loop impedance andmay not be sufficient to be detected by the common groundleakage protection probided at the incoming. It is also possiblethat the setting of the relay is not so low as to detect this. Thusto provide a ground leakage protection for each individual circuit

21/688 Industrial Power Engineering and Applications Handbookwill be worth while without any serious cost implications. Sucha philosophy, however, will not hold good for a system whichis protected only through its incoming feeder and all the outgoingfeeders are merely isolators. In this case the ground leakageprotection will have to be centralized for the entire system andprovided in the incoming feeder only.21.6 Ground fault (G/F) protectionschemesA scheme for a ground fault protection will depend uponthe type of system and its grounding conditions, i.e.whether the system is three-phase three-wire or threephasefour-wire. A three-wire system will require anartificial grounding while for a four-wire system the typeof grounding must be known, Le. whether it is effectively(solidly) grounded or non-effectively (impedance) grounded.Grounding protection will depend upon the measurementof the residual quantities (Vo or 1,) that will appearacross the ground circuit in the event of a ground fault.As discussed above, in a balanced three-phase systemthe voltage and current phasors are 120" apart and addup to zero in the neutral circuit. In the event of an unequallydistributed system or a ground fault, this balance isdisturbed and the out-of-balance quantities appear acrossthe neutral or the ground circuit respectively. The currentthrough the ground circuit will flow only when there isa ground fault, the fault current completing its circuitthrough the ground path. The normal unbalanced current,due to unevenly distributed single-phase loads or unequalloading on the three phases, will flow only through theneutral circuit.In a phase-to-phase fault, however, the system will becomposed of two balanced systems, one with positivesequence and the other with negative sequence components.The phasors of these two systems individuallywill add up to zero, and once again, as in the above case,there will be no residual quantities through the neutral orthe ground circuit, except for the transient and spilloverquantities.The ground fault current may be detected through threeor four CTs, one in each phase and the fourth in theneutral circuit (Figures 21 S(a) and (b)). Through theneutral to discriminate the fault, as discussed later.NoteIn a G/F the three CTs will also measure the unbalanced loadcurrent, if any, in addition to G/F current. For an appropriate settingof the relay, therefore, it will be essential that the likely systemunbalanced current be measured and the relay set in excess of thisto detect a G/F. For systems feeding single-phase or unbalancedloads, prone to carrying high and widely fluctuating unbalancedneutral currents, it may be difficult to determine the likely amountof unbalance and provide a suitable setting for the G/F relay. Insuch cases use of four CTs or core-balanced CTs (if it is a four-wiresystem) would be more appropriate.Below we discuss the more widely adopted practicesto detect a ground fault, i.e.:Protection through a single CTRestricted G/F protectionUnrestricted G/F protectionDirectional G/F protection andDifferential G/F protection (high impedance differentialprotection is discussed in Section 15.6.6(1)).21.6.1 Protection through a single CTThis is the simplest method to protect an equipment againsta G/F (Figure 21.12). It can, however, be applied only atthe source, which is a generator or a transformer, providedthat the source has no other parallel grounding paths inthe vicinity. This is to avoid sharing of the fault currentand false or inadequate detection of the fault current bythe relay. This scheme is therefore more functional at themain generating source, such as at the generator or thegenerator transformer, having a low impedance solidlygrounded neutral.For any other equipment or system, such as shown inFigure 21.13, the fault current may be shared by thevarious grounding stations in the vicinity and the relaymay not sense the real extent of the fault, even when thesystem is effectively grounded. Apart of the fault current,may now flow through the other nearby grounding stations.Moreover, for a fault on another feeder spill currentsmay also pass through such relays and trip them (unwanted)when the relays are highly sensitive or have a low setting.Such a scheme will also not discriminate when required,and hence will have limitations in its application. Nevertheless,it is common practice to apply single CT protectionthrough neutral circuits of the grounded transformersanywhere in the system, generation, transmission ordistribution. Multiple groundings may cause problems,but this is taken into account at the desigdplanning stage.Neutralsolidlygrounded \TiFigure 21.12'Ground fault protection through a single GTI I IFigure 21.13 Limitation in using a single CTfor a G/F protectionwhen the equipment has more than one parallel ground path

Grounding theory and ground fault protection schemes 21/68921.6.2 Restricted ground fault protectionWhen it becomes essential to discriminate between afault within the circuit to be protected from one outsidethe circuit. this scheme may be adopted. While doing so,it must be ensured that adequate ground fault protectionis available to the remaining feeders, if connected on thesame system.For a three-phase three-wire system(generally HT systetns)The scheme for a three-phase three-wire artificiallygrounded system will require four CTs, identical in designparameters. turn ratio, error and magnetizing characteristics.Otherwise spill currents may occur sometimessufficient to operate inadvertently a low-setting or highlysensitive ground fault relay. The fourth CT through theground circuit is used with the same polarity as the threeCTs of the phases. Then only would the residual currentof the phase CTs fall 180" apart from that of the groundcircuit CT. Such an arrangement will provide the desireddiscrimination, to detect the fault occurring within theprotected zone. Figures 21,14(a) and (b) illustrate thisdiscrimination through the use of the fourth CT. Theresidual current of the three-line CT, in a healthy condition,in the event of an unbalance in the system, is taken careof by raising the setting of the relay to account for theunbalance. as in a core-balanced CT. The fault currentthrough the relay will flow only on a G/F occurringwithin the restricted zone as illustrated in Figure 21.14(a).For faults occurring outside the restricted zone, as shownin Figure 3 1.14(b), the fault current through the groundcircuit CT will be offset by the residual current of thethree-phase CTs and thus the relay will remain immuneto such a fault.of residualinducedImpedancegroundedsystem7 o h To trwNote All the four CTs must be wired with the same polarity(a) Relay operates when the fault occurs within the protected zone/'Equipment tobe protectedDirection ofphase inducedcurrents in theFor a three-phase four-wire system(generally LT systems)A three-phase three-wire system is generally a balancedsystem and has negligible unbalanced residual currentthrough the three-phase CTs. The relay for a groundfault protection can thus be set low. The scheme discussedabove is thus satisfactory for a G/F. But in three-phasefour-wire LT systems. which are generally unbalanced,the above scheme poses a limitation as there may now bea substantial residual unbalanced current through the relay.For G/F protection. therefore, such a scheme will requirea higher setting of the relay to avoid a trip in a healthycondition. It is possible that this setting may becomesufficiently high. for highly unbalanced systems to detectan actual G/F and defeat the purpose of G/F protection.Such a situation can be averted by providing a fifth CTin the neutral circuit as shown in Figure 21.15, whichobviously will have its excitation current direction oppositeto the residual current of the three-phase CTs. Hence, itwill offset the bame, and the relay can be set low. It istherefore mandatory to use five CTs in LT systems foradequate restricted G/F protection.Note All the four CTs must be wired with the same polarity* Induced residual fault current through the phase CTs.~ _ _(ft = ir + iy + I b +is equal and falls opposite to theresidual current is through the ground circuit CT.(7, + 7, + l b) is considered zero under healthy condition.For small unbalances the relay is set a little higher.(b) The relay stays immune to a fault occurring outside theprotected zone.Figure 21 .I4 Scheme for a three-phase three-wire restrictedground fault protection for an HT systemApplication1 A restricted ground fault is recommended for equipmentthat is grounded, irrespective of its method ofgrounding. Unless the protection is restricted, theequipment may remain unprotected. Generally, it isan equipment protection scheme and is ideal for theprotection of a generator, transformer and all similar

21/690 Industrial Power Engineering and Applications Handbookequipment or circuits, requiring individual protection.Figure 21 .I6 illustrates the operation of the relay whenthe fault occurs within the protected zone. The schemewill prevent isolation of the equipment for faultsoccurring outside the restricted zone (Figure 21.17).2 A delta-connected or an ungrounded star-connectedwinding should also be protected through a restrictedground fault scheme, otherwise it will remainunprotected. There is no zero sequence or residualcurrent in such a winding to detect a ground fault.' IBN* The residual currents are equal and opposite, hence nullify andthe relay stays inoperative.Figure 21.17 Scheme for restricted G/F protection for athree-phase four-wire system. Fault occurring outside theprotected zoneUnder healthy condition the unbalance residual current of the 3 phaseCTs is nullified by the equal and opposite current in the neutral CT.There is thus no current through the relay.~ _ ~ _/r + /v + / b = / uFigure 21.15 Scheme for restricted G/F protection for athree-phase four-wire system. Healthy conditionThe arrangement will be similar to that for directionalprotection of a delta side of a transformer, as discussedlater, with the use of a grounding transformer (Figures21.19 and 21.20).21.6.3 Unrestricted ground fault protectionThe CT provided in the ground circuit is now removedand the same scheme becomes suitable for an unrestrictedG/F or a combined G/F and phase fault protections. It istrue for a three-phase three-wire or a three-phase fourwiresystem. This scheme may also be arranged for acombined O/C and G/F protections as illustrated in Figures21.5(a) and 21.5(b).The residual currents are additive and the relay can be set low.Figure 21 .I6 Scheme for restricted G/F protection for athree-phase four-wire system. Fault occurring within theprotected zoneApplicationThis is the most common scheme in normal use for anypower system with more than onc fceder, connected to acommon bus, such as for distribution and sub-distributionpower networks, having a number of load points,controlled through a main incoming feeder. In a switchgearassembly, for instance, common protection may beprovided at the incoming for a ground fault or combinedO/C and G/F protections as discussed above. In suchcases, a restricted G/F protection may not be appropriateor required, as the protection now needed is systemprotection, rather than individual equipment protection.The incomer must operate whenever a fault occurs atany point on the system. Moreover, for an LT system,where it may not be desirable or possible to provideindividual protection to each feeder, such a scheme isadopted extensively.21.6.4 Directional ground fault protectionIn the previous section we discussed non-directionalprotection of an equipment or a system. But for systems

gw’cGrounding theory and ground fault protection schemes 211691I @---Residual ~voltage Apolarizing coil ~-L-[,/EiPrimarywindingsstar connectedsecondaryFigure 21.19 A typical scheme for a directional GIFprotection relayNotea, b - are non-directional GFR’sa’, b’- are directional GFR’s* Reverse direction current, Iga through relay b would trip breakerB’ and reduce Iga to zero.Figure 21.18 System using directional ground fault relayswith more than one source of supply in parallel, such asa power grid, receiving power from more than one source(Figure 13.21) or an industrial load, having two or moresources of supply, one of them being a captive powersource, it is possible that a fault on one may be fed bythe other sources and may isolate even a healthy system,thus rendering the system unstable. Such a situationrequires discrimination of a fault and can be preventedby the application of directional G/F protection. Theprimary function of a directional G/F protection is thusdiscrimination.In the above situation, even an overspeeding motor ona fault elsewhere would feed back the supply source andrequire such protection. The protective scheme isolatesthe faulty source from being fed by the healthy sources.Figure 21.18 illustrates a simple power circuit providedwith a directional G/F relay. In the event of a fault insyqtem B, source B alone would isolate. Source A wouldnot feed the fault as relay b’ would trip the breaker B’and eliminate Iga. The relays are necessarily set at loweraettings and at lower tripping times than the non-directionalGFR to isolate the faulty feeder quickly than to wait forthe trip by the non-directional relay. This is to avoid atrip of the non-directional relay ‘a’ on system A.A directional G/F relay basically is a power-measuringdevice, and is operated by the residual voltage of thesystem in conjunction with the residual current detectedby the three CTs used for non-directional protection, asshown in Figure 21.19. To provide directional protection,therefore, a residual VT is also essential, in addition tothe three residual CTs. The voltage phasor is used as areference to establish the relative displacement of thefault current. In healthy conditions, i.e. when the currentflows in the right direction, V, = 0, (refer to Section15.4.3 for details), and the relay remains inoperative.The relay operates only when the current flows in thereverse direction.NoteThis relay may be used only under unrestricted fault conditions,with three CTs as shown. If the scheme is used under a restrictedfault condition, with the fourth CT in the neutral. the directionalrelay will remain immune to any fault occurring outside the zone ofthe three CTs, as the fault current through the fourth CT will offsetthe residual current, detected by the three CTY (Section 21.6.3).rendering the whole scheme non-functional.Current polarizationVoltage polarization depends upon the location of therelay and the location of the fault. It is possible that theresidual voltage, at a particular location in the system, isnot sufficient to actuate the voltage coil of the directionalG/F relay. In such an event, current polarization is usedto supplement voltage polarization. Current polarizationis possible, provided that a star point is created on thesystem, even through a AID power transformer. if sucha transformer is available in the same circuit, Figure2 1.20. Else a grounding transformer may be provided as

'L21/692 Industrial Power Engineering and Applications Handbookillustrated in Figure 21.21, and grounded neutral utilised,to provide the required residual current polarization, toactuate the ground fault relay.CTsm'1 s2DM PowertransformerNoteThe two currents, residual and polarizing, are capable of operatingthe relay.21.6.5 Current setting of a ground fault relay1 Selection of CTs for ground fault protection particularlyneeds more careful consideration of the fault conditionsand impedance of the ground circuit, in addition tothe location of the CTs. This is due to a rather lowsetting of the G/F relay, 1040% or 20-80% of thefull load current or even lower, as in mines and othersensitive locations. Too low a setting may even trip ahealthy system due to ground capacitance leakagecurrents (more so in HT circuits) or unbalanced currentsthrough the neutral (more on LT circuits). In corebalancedCTs detecting small leakage currents, it ispossible that during a phase-to-phase fault there maybe transient spill currents through its residual circuit,which may operate the low-set and more sensitiveG/F relay, which may not be desirable. To overcomethis, the time setting of the relay may be suitablyCTsResiduacurrentcoilCurrentpolarizingcoilground faultrelayCurrent polarization through the groundedneutral of aA/D power transformerA /D PowertransformerFigure 21.20 Typical circuit illustrating current polarizationscheme to operate a direction GFRResidual1current ~coil ,Current ~polarizing Icoil ~-_----Directionalground faultrelayCurrent polarization through the grounded neutral of a groundingtransformerFigure 21.21 Typical circuit illustrating current polarizationscheme to operate a directional GFRadjusted so that the overload relay will operate fasterthan the G/F, or slightly higher setting for the relaymay be provided or time delay in the trip circuitintroduced.2 In another situation, when the ground circuit has ahigher impedance than designed it may be due topoor soil conditions, dry soil beds, rocky areas, poorgrounding stations or inadequate maintenance. In suchconditions, the ground circuit may provide a lowerfault current than cnvisaged. Sometimes an overheadconductor may snap due to strong winds and fall ondry metallic roads, hedges and shrubs causingextremely low leakage currents, creating a hazard tolife and property. This may cause an arcing ground,leading to fire hazards. For all such locations andsituations, very low current settings (of the order of5% of I, or even lower) or leakage current detectionthrough core-balanccd CTs may be adopted.3 For circuits protected by HRC fuses for short-circuitconditions, the G/F relay must be a back-up to thefuses, and trip first on a ground fault. In other words,Pt (relay) < 12t fuses.

Relevant StandardsGrounding theory and ground fault protection schemes 21/693IEC Title IS BS1008-1119901009- 11199160034-1/199660255-6/198860364-1 to 760479-11199460479-2/198760755/198360898/199561024- 1/1996Residual current operated circuit breakers, without integralovercurrent protection, General rulesResidual current operated circuit breakers, with integral overcurrent protection, General rulesRotating electrical machines-Rating and performanceElectrical relays - Measuring and protection equipmentElectrical installation of buildingsGuide on effects of current passing through the humanbody - General rulesGuide on effects of current passing through the humanbody. Special aspectsGeneral requirements for residual current operatedprotective devicesElectrical accessories - Circuit breakers for overcurrentprotection for household and similar installationsProtection of structures against lightning. GeneralprinciplesCode of practice for earthingSpecification for grounding transformersCode of practice for undesirable static electricityPart 1 General considerationsPart 2 Recommendations for particular industrial situations4722/1992325/19963842-1 to 12732/19898437- 1119938437-2/199312640/19888828/19962309-19893043/19913151/1991768911989BS EN 60034-1/1995BS EN 60255-6/1995BS 7671/1992PD 6519-1/1995PD 6519-2/1988BS 4293/1993BS EN 60898/1991BS 6651/1992DD ENV 61024/1995BS 7430/1998-ANSIC 37.16/1988NEMA280/1990Related US Standards ANSI/NEMA and IEEELow voltage power circuit breakers - preferred ratings andapplication recommendationsApplication guide for ground fault circuit interruptersBS 5958-111991BS 5958-2/1991Notes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedCurrent passing through a human body(21.1)0.116I, (50 kg) = -&(ii) for a 70 kg body0.157Ib (70 kg) = -(21.2)(21.3)Sb = shock energy in watt-seconds (ws)Ib = r.m.s value of the leakage current through the bodyin Amperes (A)tb = duration of leakage current in seconds (s)Safe currents through a human body(i) for a 50 kg bodyFurther readingGeneral Electric Co. Ltd, Protective Relays and Application Guide,GEC Measurements, St, Leonards Works, Stafford, UKTaylor, H. and Lackey, C.H., ‘Earth fault protection in mines’, TheMining, Electrical and Mechanical Engineec June (1961)

Grounding-Practices221695

Grounding practices 22/697SECTION I22.1 Grounding electrodes forindustrial installations andsubstationsThe following are a few types of grounding electrodescommonly used for the grounding of industrial installations,equipment grounding or small and medium-sizedsub-stations.22.1.1 Plate groundingRefer to Figure 22.1. The approximate resistance to groundin a uniform soil can be expressed byR (derived from equation (22.12) (22.1)where p = resistivity of soil, considered uniform in Q m.A = area of each side of the plate in m2.NoteThe minimum thickness of plate is recommended asFor cast iron - 12 mmFor GI or steel - 6.3 mmFor copper - 3.15mrnand size not less than 600 mm x 600 mm15 mm @GIwater pipeIWire meshIction chamberx 500 x 600 mmExample 22.1The resistance to ground for a 600 mrn x 600 mrn plategrounding, considering a sandy soil, treated artificially andhaving attained an average soil resistivity of 10 Rrn.=q3144 2 x 0.6 x 0.6= 5.22RIf the plate is 1200 rnm x 1200 mm thenR= J 3.144 2 x 1.2x 1.2= 2.61 QFrom the above the resistance to ground of a plate groundingis inversely proportional to the square root of the lineardimension (a) of the plate. The variation in resistance withthe size of the plate is shown in Figure 22.2. Considering theresistivity of soil as 10 Qm, since the ground resistance isproportional to the resistivity of soil, there would be differentparallel curves for the ground resistance for different valuesof resistivity of soil.22.1.2 Pipe or rod groundingRefer to Figure 22.3. In this case, the approximateresistance to ground in a uniform soil can be expressedby:100 .pR = 2n.I [log, y -11 RwhereI = length of pipe in cmd = internal diameter of pipe in cm(22.2)A homogenous layerof coke/charcoal/salt00 x 1200 x 12 mmNoteThe depth (2000 mm) would vary with themoisture content and quality of soilFigure 22.1 A typical layout of a plate electrodeLinear dimensions f&i)(m) --+Figure 22.2 Variation in resistance to ground with the linear dimensionsfor a plate gounding, for the same resistivity of soil

22/698 Industrial Power Engineering and Applications Handbook15 mm$ GIwater pipeIWire meshICast iron/ chamber cover0ml1250Y CementA homogenous layerof coke/charcoal/salt& sand2500Figure 22.3station100 mm ID, CI13 mm thickA typical arrangement of a pipe electrode groundingNoteThe diameter, thickness and length of the pipe is recommended asfollows:Cast iron pipes - 100 mm internal diameter, 2.5 to 3 m long and13 mm thick. (This is a cumbersome and costlier arrangement,is not often used)MS pipes - 38 to 50 mm diameter, 2.5 to 3 m long (also notoften used)Copper or GI rods - 13, 16 or 19 mm diameter, 1.22 to 2.44 mlong.This type of electrode grounding is more suited for a soil possessinghigh resistivity, and the electrode is required to be longer and drivendeeper into the soil to obtain a lower resistance to ground. Theapproximate variation in resistance with the length of electrode fora particular value of resistivity of soil is shown in Figure 22.4, forgeneral reference.Example 22.2The resistance to ground of a 19 rnrn internal diameter pipe,2.44 rn long, with p as 10 Rrn-244Length of electrode (cms)Figure 22.4 Approximate variation of resistance to ground with thelength of pipe or a rod electrode for a particular value of resistivity ofsoilR=- 100 x 10 8 x 244 - 1 Ohms2n x 244 loge 1.9= 0.65 [log, 1027.37 - 11= 0.65 x [6.93 - 11= 3.86 RFigure 22.4 is drawn for p = 10 R rn.22.1.3 Strip or round conductor groundingIn this case the approximate resistance to ground in auniform soil can be expressed byloopR=- 2n. I x log, + Q Ohms (22.3)whereI = length of strip or rod in cmh = depth of thc strip or rod in cmw = width of the strip or diameter of the conductor rodin cmQ = -1 for strip and -1.3 for round conductor grounding.1 The minimum cross-sectional area of the strip or therod should be chosen according to the ground faultcurrent and its duration (Section 22.4.1 and equation(22.4)). The minimum area of cross-section is recornmendedasFor copper strip - 25 x 1.6 mm2For MS or GI strip - 25 x 4 mm22 We have considered a single length of electrode. Ifthere is more than one length the values can be obtainedfrom BS 7430 for different electrode arrangements.

60%Grounding practices 221699The performance of this type of electrode grounding isalmost the same as for the pipe grounding (Section 22.1.2)as is the variation in resistance to the ground with thelength of thc clcctrodc as in Figure 22.4.Example 22.3The resistance to ground for a 100 mm x 5 mm, 5 m copperstrip, buried at a depth of 1.5 m, having a soil resistivity of100 ilm100 x 100 X log, ___ 2x500' -2 x ilx 500 150 x 10= 3 185 log, 33333- 1 = 3 185 x 581 - 1R== 17 50 ilIf the length of the strip is 25 m. other parameters remainingthe same, then100 x 100 2 x (2500)'R=-12 i l ~ 2 5 0 0 ~ 150x10 ' ~ ~ ~= 0.637 log, 8333.33 - 1= 0.637 x 9.03 - 1= 4.75 022.1.4 Choice of grounding method1 The normal ground impedance in LT systems, isgenerally high. To achieve a ground fault current ofthe order of 1 to 3 times the rated current, necessaryto protect a low current system against a ground faulta\ discussed in Section 2 1.2. I , would be dificult unlessadequate measures are taken with the groundingstations to have as low a ground impedance as possible.To achieve this. the grounding stations are madeelaborate. at adequate depth, with proper chemicaltreatment and watering arrangements to ensuresufficient moisture throughout the year. This is attainedby a perforated pipe driven froin ground level up tothe electrode. See Figures 22.1 and 22.3, showingtypical arrangements of a grounding station, with aplate grounding and a pipe grounding respectively.Despite this, the ground resistance may still be toohigh to meet the design parameters.To overcome this, a number of such groundingstations (two being a minimum to provide a doublegrounding system) may be essential and connected inparallel to achieve the required low value of groundresistance. A cumulative resistance of up to 2.5 R isconsidered satisfactory. However, for more effectivenessand to make the ground protective circuit moresensitive. a resistance of up to 1 R would be better.The grounding stations may be separated, centre tocentre. by 2 in or more, to be out of each other'sresistance Lone. A better gap would be around 4.5 to6 m, when full ground resistance may be achieved byeach electrode without infringing on the resistancerime of the other electrodes as well as economizingon the number of electrodes. This use of more thanone grounding station in parallel so that each stationis out of the resistance zone of the other stations,would alter their cumulative resistance, generally asfollows according to Hund Book ofElectrica1 histullulionPructic.e.s by E. A. Reeves:-0 1 2 3 4Number of electrodes (stations)Figure 22.5 Likely variation in ground resistance with more thanone electrode connected in parallelPlate groundingFor two electrodes - 50%For four electrodes - 256Pipe, rod or strip groundingFor two electrodes ~For three electrodes - 45%For four electrodes - 35%For more accurate calculations refer to BS 7430. Thelikely variation is illustrated in the form of the graphin Figure 22.5.The resistance to ground would vary with depth ofthe electrode. A minimum depth of 1.5 in from groundto the top of the electrode is considered mandatory.and even deeper to reach damp soil.A pipe or strip grounding is more effective than aplate grounding.The size of the plate, the length of the pipe or thestrip may be altered to obtain a lowei- balue of groundresistance.The choice of the metal (Section 22.4) for the groundingelectrode will depend upon the corrosion factor ofthe soil. But all metals are equally good and possessa life span of 12 years and more. For a longer workinglife, the thickness of the electrode may be increasedas discussed in Section 22.4.1. GI* being a more-*Apparently GI seems to be the best metal as B groundiiig electrode.But if part of the zinc coating of the metal is chipped due to poorcoating or to any other reason. the metal is rendered prone io rapidcorrosion and erosion and may fail with passage of time. Someusers therefore prefer lo use hare MS conductor rather than GI.

22/700 Industrial Power Engineering and Applications HandbookTable 22.1Likely resistivity of soilType of soilAlluvium and lighter clays, such as sandyand/or muddy soilClays (excluding alluvium)Marls like keuper marl such as marblePorous limestone like chalkPorous sandstone such as keuper sandstoneand clay shalesQuartzite's compact and crystalline limestonesuch as carboniferous marbleClay slates and slaly shalesGraniteFossil slates, schists, gneiss, igneous rocksBased on BS 7430"Depends up on the water level of the localityClimatic conditionsNormal and highrainfall (more than500 mm a year)Resistiviry of soilRm(Likely value)5102050100300100010002000Low rainfall or desert condition(less than 250 mm a year)(Likely range of values)a-5 to 2010 to 3030 to 10030 to 300100 to 1000300 to 30001000 upwardsa10 to 1005 to 300lUndergroundwater salinity--1 to 51000 upwards 30 to 100preferred metal for grounding purposes. Choice ofthe metal for the ground electrode would, however,depend upon the underground municipal services, suchas water, sewerage and telephones and also thestructures and foundations of nearby buildings to savethem from corrosion and erosion. Copper, beinggalvanic, under damp conditions forms a completeelectrolytic circuit between it and other metals andcauses corrosion, which erodes the other metals (aneffect similar to that discussed in Section 29.2.5, whilemaking a bimetallic joint).6 Any of the grounding methods may be adopted forindustrial installations, equipment grounding or smalland medium-sized sub-stations, depending upon thetype and condition of the soil. A sandy soil will beeasy to dig and plate grounding will be easier, whilea rocky soil will present problems in digging and apipe, rod or strip grounding would be easier. Similarly,dry soil will require deeper digging, where a pipe,rod or strip grounding would be a better choice.22.2 Resistivity of soil (p)This will depend upon the type and quality of soil and itschemical composition, i.e. the composition of salts andminerals, content of moisture and normal rainfall duringthe year. Table 22.1 obtained from BS 7430 and Table22.2 from IEEE 80 show the likely resistivity of differenttypes of soils in ohm-metres. These are only likely valuesfor a general reference. It is recommended that where agrounding station is to be installed, the soil is tested atTable 22.2Type of soilRange of soil resistivityWet organic soil 10Moist soil 102Dry soil1 o3BedrockBased on IEEE-80Average resistivityRmI o4nearby locations and an average value of the soil resistivityis determined, as discussed in Section 22.11. The conditionof soil, such as its moisture, content temperature andcontent of salts and other minerals have a large bearingon its resistivity. Figure 22.6 illustrates the effects ofsuch factors on the resistivity of soil. While the temperatureof the soil is a fixed parameter, at a particular location ofthe grounding station the soil can be artificially treatedto improve the content of moisture and chemical composition,to achieve a lower value of soil resistivity. Ithas been found that the resistivity of soil can be reducedby 15-90% by a chemical treatment with the followingsaltsNormal salt (NaCI) and amixture of salt and soft cokeMagnesium sulphate (MgS04)Comer subhate (CuSOA)Caichm chloride'(CaC1;)Sodium carbonate (Na2C0,)~~~~~~~~l and mostcommonly used saltsMore common salts

, I0 10 20 30%Content of moisture +(a)f-103EA2.:8 10’c3:I.z 10d-Sod temperature(b)1-20 O* 20 40 60°CGrounding practices 221701-0 5 10 15% 20%Effect of salt(C)Figure 22.6Effect of moisture content, temperature and salt on the resistivity of soilRefer to Figures 22.1 and 22.3, illustrating a normalarrangement of grounding stations with provision forchemical or salt treatment. The salts used need not be indirect contact with the electrode.22.3 Measuring the groundresistanceI),AGroundThc above tables can give only a general idea of thetheoretical value of resistivity of the soil at a particularsite for the purpose of design work. The exact resistanceof the grounding station must be determined at the siteof installation to support theoretical assumptions and thegrounding conditions adjusted, if necessary, to obtainthe required ground resistance. The resistance of a groundingstation can be measured with the help of a groundtester. which generates a constant voltage for accuratemeasurement. The tester has two potential and one currentprobe. The procedure of measurement is illustrated inFigure 72.7.One of the potential probes A is drilled into the groundat ahout 15 m from the grounding station G, whoseresistance is to be measured. The second probe B is placedbetween the two. The current lead of the meter is connectedto the grounding station. The meter will indicate someresistance. which may be noted. Two more readings arealso taken by shifting the centre probe B by almost 3 mon either side of the original location. For an accuratedue of the ground resistance, the values obtained mustbe same. If they are not, the probe B is still within theresistance area of the grounding station G. Shift awayprobe A by another 6 m or so and place probe B betweenG and A, and repeat the test. If the three readings arenow the same. consider this as the actual ground resistanceof station G, otherwise shift probe A farther away until aconstant reading is obtained.The same test can also he conducted with the help ofa battery, voltmeter and an ammeter, as illustrated inFigure 22.8. The voltmeter must now indicate the samereading at all three locations. When Vbecomes constant,read the current I. Then the ground resistancestationFigure 22.7 Measuring the ground resistance with the help of aground testerA Ground’= levelPotentialprobesFigure 22.8 Measuring the ground resistance with the help of anammeter and a voltmeter

~ ~~ ~~ ~~ ~ ~~~~~22/702 Industrial Power Engineering and Applications Handbookr7VR,=-!2I22.4 Metal for the groundingconductorCopper, aluminium, steel and galvanized iron are themost widely used metals for the purpose of grounding.Choice of any of them will depend upon availability andeconomics in addition to the climatic conditions (corrosioneffect) at the site of installation. In Table 22.3 we providea brief comparison of these metals for the most appropriatechoice of the metal for the required application.22.4.1 Size of the grounding conductorThis is a matter of system design and is different for LTand HT systems, as discussed above. The main criterionwhen determining the size of the ground conductor is tosustain the rated short-time ground fault current of thesystem for the required duration, without damage to orpermanent deformation of the ground conductor and tolimit its temperature rise within permissible limits. Itwill also limit the voltage drop within 55 volts betweenany two grounded points with which a human body maycome into contact. However, for all practical purposes,the minimum size of conductor as determined below fora required fault level will generally be adequate to limitthe voltage drop within the safe limits. For more dctailsrefer to IEC 60298.The ground conductor can be of aluminium, GI orcopper, as discussed earlier. A humid or a chemicallycontaminated location is corroding in nature. Aluminiumhas a rapid reaction and is fast corroding. At such locations,use of GI or copper conductor would be more appropriate.Table 22.4 suggests the ground conductor sizes foraluminium conductor power cables for small and mcdiumratingfeeders when aluminium is used for the groundTable 22.3 Comparison of grounding metalsNo. Churucteristics Copper Alum in iurn Steel Galvnnized iroriI2 3 41 Conductivity (96) 100 61 30-40 8.5(for annealed copper) (for EC grade aluminium) (for copper-clad steel core) (for Zn-coated steel)2 Resistance to High. Being cathodic Highly corrosive and is, Corrosive. Copper-clad High, and is extensivelycorrosion with respect to other therefore, less preferred steel may be used to used for groundmetals, which may be compared to other overcome this deficiency connections and gridsburied in the vicinity metals, for undergroundconnections or groundelectrodes. For surfaceconnection\, however,where it is less corrosiveand highly conductive,compared to steel or steelalloys it is preferred3 Galvanic effect" Copper is a galvanicmetal and causescorrosion, in the presenceof moisture, in nearbymetals, such as cablesheathes, steel structureand water, gas or drainpipes, buried in itsvicinity. With all suchmetals, it forms acomplete electrolyticcircuit and corrodesthem. Tinning may giveprotection against itsgalvanic effects but thisis an expensivepropositionThe are not galvanic but become anodic in the vicinity of copper and erode4 Approximate cost 100 50 IO 15considerations (4)Therefore most appropriateand economical~ ~~ ~ ~ ~ ~ ~ ~~ ~"Thia occurs when two dissimilar metals in an electrolyte have a metallic tie between them. There is a flow of electricity between the anodicand cathodic metal surfaces, generated by the local cells set between dissimilar metals. One metal becomes an anode and the other a cathodeand causes an anodic reaction which represents acquisition of charges by the corroding metal. The anode corrodes and protects the cathode,as current flows through the electrolyte between them.

Grounding practices 221703Table 22.4 Size of aluminium ground conductor for differentsizes of power cables for a grounding system1 up to 25 mm’ Same as the cable size7 Above 25-50 mm’ 25 mm23 Above SO mm2 Approximately half the size of themain cable size. Say. for a 400mm2 main cable, a groundconductor of 185 mm’, will beadequateconductor. For a GI ground conductor, this size may beroughly doubled.For large feeders and HT systems the ground faultcurrent would be controlled naturally through the groundcircuit impedance and a smaller ground conductor maysuffice.NoteIn LT systems, where the neutral is grounded, the neutral as well asthe ground conductor may have to carry unbalanced currents up 10half the rating of the line currents due to single-phase loads. Aground conductor should also he rated for the same size as theneutral. irrespective of the setting of the relay.Now equation (22.4) as suggested by BS 7430 willapply, which is based on our discussions in Section 2 1.3.1,where the ground system is normally predetermined forthree times the rated current of the circuit for an HRCfuse-protected system or one and a half times for an overcurrent release-protected system:(22.4)whereS = crosyectional area of a bare ground conductor inmm- .!g = r.m.s. value of the ground fault current in amperesK = r.m.s. current density in A/mm2. This will dependupon the material of the conductor and its maximumpermissible temperature. For more common metalsit may have the following values, assuming the initialtemperature of the conductor to be 4OOC.Copper = 205 A/mm2, assuming the finaltemperature to be 395°CAluminium = 126 A/mm’, assuming the finaltemperature to be 325°CSteel or GI = 80 A/mm’, assuming the finaltemperature to be 500°Ct = duration of fault in seconds (operating time of theprotective device).NoteFor other grounding materials. or hazardous locations requiring amuch lower end temperature. refer to BS 7430.Example 22.4Consider a power distribution system having the main incomingfeeder rated for 400 A and the outgoing feeders rated up to200 A. To calculate the main ground conductor size, assumethat the system is protected through HRC fuses. Then, basedon the previous assumptions,Ground fault current, I, = 3 x 400= 1200 Aand interrupting time of 400 A HRC fuses, referring tocharacteristic curves of Figure 21.4 = 60 seconds:. Ground conductor size for an aluminium conductori.e. 25 mm x 3 mm [ 1” x $1or any other cross-section of an equivalent areaIf the conductor is of GI then,s=- l2O0 ,/6080or = 116 mm2or 25 mm x 5 mrn (1” x 4”)or any other cross-section of an equivalent area.It could similarly be calculated for the individual outgoingcircuits, or considered equivalent to half the cable size beingused to feed the circuit.Example 22.5If a distribution system is fed from a 1600 kVA, 11 kV/415 V,transformer, thenI,1600 x 1000 A=x 415= 2225 AIf the system is protected through overcurrent releases, thenapplying the same assumptions as before:Ground fault current, I, = 1.5 x 2225 Aand the maximum tripping time at this current, referring tocharacteristics curves of Figure 21.3.= 370 seconds:. Ground conductor size for an aluminium conductorS= 1.5 x 2225 %/m126= 509.5 mm2or 100 mm x 5 mmor 100 mm x 6 mmor any other cross-section of an equivalent area.If the conductor is of GI thens = 1.5 x 802225 ,,m= 802.5 mm2or 80 mm x 10 mm orany other cross-section of an equivalent area.

0'1s722/704 Industrial Power Engineering and Applications Handbook22.5 Jointing of groundingconductorsAs discussed in Section 29.2.5, jointing of two differentmetals (copper being one) causes electrolysis at the joints,leading to corrosion and failure of the joint. To avoidthis, it is recommended that the same procedure be adoptedas discussed in Section 29.2, and where the electrodeand the connecting ground strip are of the same metal,that the joints are riveted or welded with the same metalafter making the surface. Soldering is not recommended.22.6 Maintenance of groundingstationsTo ensure that a grounding station has not deterioratedand its ground resistance has not increased due to soildepletion it is mandatory to carry out a few checksperiodically to ascertain the resistance of the groundingstation. If the ground resistance is found more than itwas designed for, it is possible that by proper moisteningof the soil or by adding more salts or chemicals to thegrounding pit, the desired level of ground resistance isachieved once more. If not, then additional groundingstations may have to be installed to obtain the originallevel of the ground resistance.22.7SECTION IIGrounding practices in a powergenerating stationThis is a vast subject, on which extensive research hasbeen done by many authors over the years. The groundingstations in such areas are normally spread over the entirestation, and sometimes may even extend beyond itsboundary to achieve the desired results. Here we discussbriefly, the basic criteria behind the requirement of agrounding system in a power station and its designconsiderations.The magnitude of ground voltage in such areas in theevent of a ground fault is very high, due to high systemvoltage. On a ground fault, the ground path resistancemay become a source of a high potential gradient acrossthe grounding conductors at a particular location. It maybecome high enough to prove fatal to a human operatorcoming into contact with it. To limit this potentialdifference at all locations within a tolerable value andachieve an equipotential distribution of a ground conductorover the station is the basic criterion on which is basedthe design of a grounding system for a power generatingstation. Our discussion is also applicable to outdoorswitchyards and large sub-stations. For detailed working,it is advisable to refer to IEEE-80.The following basic data are important for the designof such a grounding system.22.8 Tolerable potential difference ata locationWe discussed in Section 2 1.1.1 the maximum tolerablecurrents through a human body and their duration. Thepotential difference in a ground conductor at any pointwhere a human body may come into contact with it duringthe course of a ground fault should be such that theresultant current through the human body will remainwithin these tolerable limits.22.9 Voltage gradientsThe likely positions, in which a human body may comeinto contact with a phase or a ground conductor, and thecorresponding potential differences, that he may beexposed to, are illustrated in Figure 22.9.22.9.1 Tolerable step voltage, E,This is the difference in the surface potential to which ahuman body may be subject when bridging a distance of1 metre on the conducting ground through the feet withoutbeing in contact with any other conducting groundedsurface (position 1, Figure 22.9).The safe step voltage, E,, should not be more than thetotal resistance to ground through the body, R2f,b (Section22.10.1) x safe body current, Ib, as a function of time,whereR,fsb = 6 . c, . ps + 1000 R (from equation (22.10),discussed later)and I, = - l6 for a SO kg body, as in equation (21.2)fi--0.157 for a70 kg body, as in equation (21.3)Jt.*. Es(5") I (6 ' C, ' ps + 1000) x ~and(6 . C, . pE + 1000) x ~O. for a 50 kg bodylii(22.5)for a 70 kg bodylir22.9.2 Tolerable touch voltage (E,)(22.6)This is the potential difference between the groundpotential rise (GPR) and the surface potential at a pointwhere the person is standing on the conducting groundwith one hand in contact with a conducting groundedsurface (position 2, Figure 22.9). The safe touch voltage,Et, should not be more than the total resistance to groundthrough the body, R21ps (Section 22.10.1) x safe bodycurrent, Ib, as a function of time, whereRZfps = 1.5 C, . p\ + 1000 R (from equation (22.11)discussed later)

Grounding practices 22/705YGenerator transformerGeneratorIndoor/@ - E, - Step voltage@ - E, - Touch voltage@ - E,,, - Mesh voltage(max touch voltage)@ - - Transferred voltageUtility auxiliary transformerSwitchgearFigure 22.9 Likely positions in which a human body may come in contact with phase and the ground conductors and the corresponding voltagegradients (illustrated through the layout of Figure 13 21)E,,,,,, 5 (1.5 c, ' p, + 1000) x ~and E,,7o10.1 16\f0.1575 (1.5 C, p,. + 1000) x -\tfor ;I SO kg body (22.7)for a 70 kg body (22 8)Of the d e step and touch voltagcs, the requirement ofthe safe touch voltage E, is more stringent. These arebasic design parameters, which will decide the requiredsize of ground mat and the design of mesh. In all ourfuture assumptions. we consider the average weight of ahuman body to be 70 kg.22.9.3 Mesh voltage or safe design voltage (E,)This is the maximum touch voltage ofa grounding stationthat may occur under thc worst situation (position 3,Figure 22.9). The design of the grounding station mustensure that in actual service this voltage does not exceedthe permissible tolerable limits noted in Section 12.9.6.22.9.4 Ground potential rise (GPR)This is the maximum voltage, that a station groundinggrid (ground mat) may attain relative to a remote grounden route to the grounding network considered to be atthe potential of the remote ground. In normal conditions

22/706 Industrial Power Engineering and Applications Handbookthe grounded grid potential may be regarded as zero,except transferred voltages and surge pilferages that maybe caused during a transient state (Section 18.5.2). Duringa ground fault the current will flow through the groundinggrid and cause its potential to rise with respect to a remoteground. This voltage rise is seen to go up to 25 kV, butgenerally not beyond 10 kV (IEEE 367) and can beexpressed byGPR 0~ I, . R,where Ig = fault current through the grounding gridand R, = grid resistance at the station grounding grid,with respect to the remote ground.The larger the grounded grid area, the lower will be thegrid resistance, and the lower the GPR and the meshvoltage.22.9.5 Transferred voltage (Etr)This may also be considered to be a type of touch voltagewhere the voltage may be transferred into a switchyardor a generating area as a result of a ground fault somewherein the power network in one of the supply sources and aperson standing in the local area of one grid station comesinto contact with a grounded conducting part, groundedat a remote grid station or vice versa (position 4, Figure22.9). In such a situation, if a ground fault occurs, thepotential to ground may exceed the full GPR of the localgrounding grid where the person is standing. Thetransferred voltage may exceed the sum of the two GPRsof the two grounding grids, due to the induced voltagesin the steel structures, neutral wires and metallic pipes inthe vicinity. It is not practical to make provisions forsuch an eventuality in the design of a station groundinggrid. To safeguard a human body from such voltages,IEEE 80 has recom-mended providing isolating devices,surge arresters or display danger boards at suitablelocations. For more details, refer to this Standard.22.9.6 Design parametersThere are a few important parameters that must bedetermined before beginning the detailed engineering ofa grounding station.Maximum ground grid current and its durationThis is the maximum grid ground fault current, IG, thatmay occur during the lifetime of the power plant. It mayincrease to the sum of two GPRs as noted above, ;.e. upto 80 kA or even higher (IEEE 367). For system faultlevels refer to Table 13.10. It is advisable to carry outfault current studies every few years to assess the actualfault level compared to those considered at the time ofdesigning the grounding system. It is possible that thegenerating capacity of the power station and so also itsfault level has increased with time.t, - duration of fault. Typical values may range between0.25 and 1.0 stsl - shock duration. The value may be considered bykeeping a safety margin in the allowable body currentand its duration. In an automatic reclosure powersystem, reclosure after a ground fault is commonpractice in modern power systems. This may resultin repeat shocks in quick succession to a humanbody coming into contact with the ground conductor.Although this situation may last for less than 0.5 s,it may prove fatal. A reasonable allowance for suchan eventuality should be made when deciding onthe clearing time.Since a switchyard is normally connected to more thanone supply system, the ground fault current in a powerstation is contributed by the power plant as well as bythe switchyard and the transmission networks. Thefollowing possibilities may arise:1 When there is a fault in the local generating area(Figure 22.10(a)) the return path will be through thegrounded neutral of the generator. The switchyard'sother remote power sources may also contribute toi.RemoteLocalsourcepower stationThe step and touchvoltages are not affectedin- the generator area+ 191LGTGTT G L---+III+Il - - - - - - - - - - - - - - - - - - - - - - - - - - - ~ L f -O/G11 I11 I11 I192 I1 41 ,'Ig2 will exist only if the source is grounded star[If it is A or isolated neutral /g2 = 01Remotesource(a) Fault at the local areaOIG19 1Y~L------------)-----------------'192 (b) Fault at a remote locationLocalpower stationFigure 22.10 Contribution to ground fault currrent by other supplysources when more than one system IS operating in parallel

Grounding practices 221707this fault as illustrated, provided that they are groundedstar. An isolated star or delta-connected source willremain unaffected by remote faults. See also Table13.5 for more clarity. The step and touch voltages inthe generator area will not be affected. The GT(generator transformer) area and the nearby steelstructures will develop high step and touch voltages.2 Similarly, when a fault occurs some distance fromthe generating area, this area will feed the remotefault as did the remote sources in the generator areain the previous case, thus, developing step and touchvoltages in the GT area (Figure 22.10(b)).The flow of circulating currents in the groundingconductors or ground of region two caused betweentwo or more interconnected grounding stations, for afault occurring in region one is termed the telluriceffect.3 When the generator and the switchyard groundingmats are interconnected the ground fault current willdivide between the two, depending upon their groundresistances, in inverse proportions (Figure 22.1 1) suchthatRemotesource(a)Localpower stationI Power plantground circuit1 Switchyardground circuitand I,,I, 'R,= -RlI, 'R,and I,, = -where R, = ~R2R, 'R2R, +R2For the grounding grid to remain effective over longyears of operation, in view of existing ground parallelpaths provided by other grounding stations in the vicinityand expansion of the power system in future, moremeticulous design would also consider the followingfactors:Resistance of the grounding gridDivision of the ground fault Current, 1,. between theother parallel ground pathsThe decrement factor to account for future exDansion,Power plant Switchyardarea area(b)I, = Total symmetrical fault currentIgl = Fault current shared by the power plant area/g2 = Fault current shared by the switchyard areaR, = Resistance of the power plant ground circuit4 = Resistance of the switchyard ground circuitFigure 22.11 Sharing of fault current by the power plant and theswitchyard areaswhereIG = maximum ground fault currentIg = symmetrical ground fault currentDf = decrement factorSafe design voltageOf all the grounding grid voltages derived above, themesh voltage or the maximum touch voltage, E,,,, mustfall within the safe limits and it forms the basic designparameter. The design of the grounding system mustensure that on a ground fault the actual touch voltage E,,,will not exceed the maximum tolerable touch voltage E,mentioned above. An ideal design would mean a potentialassuming that there is no crushed rock and ps = 0. Thegraph reveals that a human body weighing 50 kgcan endure a shock voltage of 65 V for almost 3.2 s and130 V for almost 0.8 s. Similarly, a body weighing 70 kgcan endure a shock voltage of 65 V for almost 5.8 s and130 V for almost 1.46 s. A higher touch voltage than130 V would require a yet faster isolation of the fault.22.1 0 Determining the leakagecurrent through a body22.10.1 Body resistance1 The proportion of the leakage current through a humanbody will depend upon the resistance of the body

22/708 Industrial Power Engineering and Applications Handbook2 To determine the total resistance of the ground circuitthrough the human body, the following may be adopted.R2rs = resistance between the two feet in seriesRZfp = resistance between the two feet in parallelThere are many formulae to determine the above, allleading to almost the same results. The most adopted,assuming a layer of crushed rock (gravel) over theground surface, is expressed byR2fs = 6 . c\ . P.;and total touch resistance RZfsb through the bodyAllowable fime (sec.) --tFigure 22.12 Limits of touch voltages as a function of timecompared to the resistance through the ground. Todetermine the likely body current it is therefore essentialto determine the average body resistance. On thissubject many studies have been made and the followingdata established (Figure 22.13):a = resistance hand to hand = 2300 Rb = resistance hand to feet = 1130 R(A leather shoe is considered as a part of the body)c = resistance between the two feet =lo00 QIt is observed that the body's resistance diminishesat higher voltages, above 1 kV and currents morethan 1 A, passing through the body, due to a punctureof the skin tissues. For all safety measures and grounddesign consideration, the average human bodyresistance is considered universally, as 1000 R whichhas yielded satisfactory results.RZfsb =' cs ' Ps -k Rb= 6 . C, . ps + 1000 (22.10)and RZfp = 1.5 x C, . p,and total step resistance, R2fp,, through the bodyR2fp, = 1.5 x C, . p, + Rb= 1.5 x c, ' p\ + 1000 (22.1 1 )where Rb = body resistance= 100OQC, = reduction factor for derating the nominalvalue of surface layer resistivity, correspondingto a crushed rock layer of thickness h,and a reflection factor kwhereP k= - PIP + Psandp = ground resistivity in Riiip\ = crushed rock (gravel) resistivity in RmNoteTo achieve a liigli contact rcsiatancr ah a measure to provide highersafety to personnel working in the power plant and switchyardareas. common practice is to spread a layer of concrete or crushedrocks (gravel) over the finished ground surface. In the power plantarea. a layer of concrete ( 150-300 mm, depending upon the stationvoltage) is spread to provide a resistivity of nearly 500 Rm ormore. In the switchyard area. a layer of crushed rocks is spread(75-150 mm) to provide a resistivity of nearly 2500-3000 Rm ormore. The value of C, can be read from the h, versus K curvesprovided by IEEE 80, as in Figurc 22.14.TCrushedrock/-SOllFigure 22.13 Resistances of different body partsExample 22.6Consider a large sub-station grounding system, having alayer of crushed rock, 150 mm thick at the surface, having aresistivity of 3000 Rm and the soil resistivity of 150 Rm:150 -:. k =3000 - 2850150 + 3000 3150= - 0.90:. C, from Figure 22.14, corresponding to a rock surface of150 mm= 0.7:. Ground resistance between the two feet in series

"-0 0.04 0.08 0.12 70.16 0.20 0.2470.15 0.25hs (metres)Figure 22.14 Reduction factor C, as a function of reflection factor Kand thickness of crushed rock (gravel) h,Ras = 6 x 0.7 x 3000= 12 600 Rand in parallelR2fp = 1.5 x 0.7 x 3000= 3150 RHaving determined the actual ground loop resistance throughthe body, one can find the ground leakage current that mayflow through a human body during an actual ground faultunder different body touch conditions with the grounding grid.For example, referring to Example 22.9 and Table 22.6,the safe touch voltage, Et, in the power plant area is estimatedat 267 V. For this voltage, the leakage current, le, through thetwo feet when in parallel which is a more severe case,dwhere4 AandGrounding practices 22/709= ground resistance at the surface of the soil- P = ground resistance of the total buried length(L) of the conductorsR, = station ground resistance in Qp = average resistivity of soil in QmThis will depend upon the condition of thesoil and its moisture content. This is why it isusually high where the moisture content isless than 15% of the weight of soil. Thevariation in soil resistivity is, however, lowwhen the moisture content exceeds 22%.A = area of the grounding grid(i) in a rectangular gridA = u . b m2wherea = length of the grid in m andb = width of the grid in m(ii) In a circular gridA = + a 3 m2(Figure 22.15)wherer = radius of the grid in mL = total length of the buried conductors in m= L, + L, (see also Section 22.14.4)L, = total length of conductors, used in the grid in mand L, = total length of the grounding rods in mNumber ofgrounding rods = ngGL \Gounding1rods--267 x 1000 mA31 5022.10.2 Ground resistanceThe ground resistance is a function of the area occupiedby the grounding station and the stratification of the soil.The stratification of the soil is usually of a non-uniformnature and may vary the resistivity of soil vertically aswell as horizontally, thus varying the resistance of soil.The minimum value of ground resistance (resistance ofthe grounding station) at a certain depth h from theground surface may be expressed byh - Average depth of grid below the ground surfaceFigure 22.15Area of a grounding grid and length of buried conductors

22/71 0 Industrial Power Engineering and Applications HandbookIf there are ti, number of conductors lengthwise and/ih widthwise and nlf = number of grounding rods used ina grounding grid at a depth of h, then the total length ofthe buried conductorsL = 12,b + nh . n + rig . h metres(22.13)With an increase in the length of the buried groundconductors, the value of their ground resistance diminishes.It has been found that equation (22.13) is more accuratefor a grid depth up to 250 mm. At greater depths ofstation grids, a more accurate representation is found inthe following equation:1 )] (22.14)1 + h,i201A-22.1 1 Measuring the averageresistivity of soilIt is important to determine the average resistivity of soilat every site where a grounding station is to be located.To do this, a soil test is essential. For this, samples maybe collected from a number of nearby locations at thesite to arrive at an average value. As a result of soilstratification, samples must be collected at different depthsto ascertain variation in the resistivity to decide on asuitable depth for the grounding grid. For simplicity,Tables 22.1 and 22.2 suggest the likely average range ofresistivity for different kinds of soils and their moistureconditions.A simple way to measure the resistivity of soil is afour-pin method in which four probes are drilled into theground along a straight line at equal distances a anddepth b. Then a voltage V is applied to the two innerprobes and a current, If, is measured in the two outerprobes (Figure 22.16). This test can also be conductedwith the help of a ground tester as discussed in Section22.3, which normally also has a provision for this test.SwitchBatteryThe soil resistanceV R =-I! I,and p=Since generally:. p = 2na R,4 n. a , R,2 LE1 +=----L4"' + 4b2

10.72kAGrounding practices 22/71 1where1, = symmetrical r.m.7. value of the zero sequence faultcurrentVt = line voltageZ, = positive sequence equivalent system impedance,Q/phase at the location of the faultZ7 = negative sequence equivalent system impedance,Q/phase at the location of the fault.Z,, = zero sequence equivalent system impedance, a/phase at the location of the fault.It is also possible to estimate this, if the system unitimpedance is known, when the maximum fault currentFull load current of the system= Unit impedance of the system= - 1 x 100%Z,l(13.5)Example 22.7Consider a switchyard receiving power from three of 200MVA, 15.75 kV generating sources, having an impedance of16 5 10% each and feeding the transmission lines through15.75/132 kV, 250 MVA generator transformers (Figure 22.1 7)with an impedance of 14 f 10% each (Table 13.8).Considering the system fault level at 32.8 kA, calculatethe ground fault contribution by the generating units;In this case the sources feeding the fault are generatorsand not generator transformers (GTs). The GTs are only currentlimiters and introduce their reactances into the circuit. Sincethe ratings of the generators and the GTs are different, it ismandatory to first convert them to a common base, say, 200MVA in this case. Therefore the combined unit impedance ofeach generator and the GT at point 6, considering the lowervalue of the impedances to be on the safe side,200= (16 - 1.6) + (14 - 1.4)- 250= 14.4 + 10.08= 24.48%Since all three generators are operating in parallel, theimpedance of each circuit, as calculated above, will fall inparallel and the equivalent impedance will become_- +-+-1 -- 1 1Z,, 24.48 24.48 24.48or z,,24.48= - 3or 8.16% (0.0816 P.u.)Full load current at base MVA= 0.875 kA0.875:. I - -' - 0.0816= 10.72 kASince the system fault level is 32.8 kA, the ground faultcontribution by the other sources connected on the gridor= 32.8 - 10.7222.08 kA15.75kV200 MVA 200 MVA200 MVA250 MVA 250 MVA-I250 MVA1L 4IlO/GPower plantI-- T I- i-lground bus 1 11Fault fed by other ~sources connectedon the gridII____ ~__-_ - _ __--IIi-5IlI ______- - 1,Figure 22.17 Illustration of Example 22.7

~~ ~~ ~22/71 2 Industrial Power Engineering and Applications Handbook22.14 Designing a grounding grid22.14.1 Minimum size of grid conductorsThis can be calculated from the formula derived by Sverakand recommended by IEEE-80:A= , I, (22.1 6)whereA = cross-sectional area of ground conductorin mm2I, = ground fault current in kA (r.ni.s.)This may be substituted with the estimatedmaximum ground grid current, IG (Section22.9.6), that may occur during the life ofthe grounding stationTcap = thermal capacity factor from Table 22.5,in J/cm3/"C. This is derived from formula(4.184) ph . ps in Ws/cm'/ "C (for detailsrefer to IEEE-80)ph = specific heat of ground conductor in Cal/gram/ "Cps = specific weight in g/cm't = duration of fault in secondsyo = thermal coefficient of resistivity at areference temperature of 20°Cp2,, = resistivity of ground conductor at a referencetemperature of 20°C in ,U R/cmKO (at 0°C) = reciprocal of w0- I 20Dc20tmax = maximum allowable temperature in "Ctamb = ambient temperature in "CTypical values of the above constants for the most widelyused metals are given in Table 22.5, based on IEEE-80.22.14.2 Corrosion factorCorrosion takes place in all metal conductors located in 3humid environment, ground electrodes, being one example.It is therefore mandatory that certain corrosion marginsare considered when choosing the size of ground electrodesto account for this in the long run, particularly during theconsidered life span of the generating station, switchyardor sub-stations. Handbooks on corrosion suggest likelycorrosion depths. For steel this is considered to be around2.2% per year. For GI it will be much less. Consideringeconomics of steel over GI it is a common practice to useonly steel for such extensive and elaborate groundingstations. For steel grids, the depth of corrosion in 12 yearsis estimated to be around 3.48 mm in a soil having a pHvalue of 7.4. This figure can be used to determine thedepth of corrosion for any number of years. Consider-ingthe lifespan of a power generating station as 40 years, thedepth of corrosion during this period would bei-+-+-+...-1 1 11 2 3- 3.48 x 4.27853.1032= 4.8 mm12This amount of corrosion will occur on each side of theelectrode,:. total corrosion during a span of 40 years of operation= 2 x 4.8= 9.6 mmThis is an average value of metal erosion during thelength of service and may vary with soil conditions. Incoastal areas, for instance, where the subsoil water issaline, erosion of metal would be much more rapid anda further safety factor must be considered. Field experiencewill be a better guide to assess this.22.14.3 Maximum touch and step voltages of agrounding stationIn Section 22.8 we discussed the tolerablc step and touchTable 22.5Material constantsDescriptionConducrivit>8 a20K,, = -- 20 t,,,. Fusing PZO Tcup[-:, 1 temperatiire "C pWcm J/tm-'/"CStandard annealed soft copper wire 100 0 000393 234 1083 17241 3422Commercial hard drawn copper wire 97 0 000381 242 1084 17774 3422Copper-clad steel core wire 40 0 000778 245 1084/1300 4397 3 846Copper-clad steel core wire 30.0 0.00378 245I084/1300 5.862 3.846Commercial EC aluminium wire 61.0 0.00403 228657 2.862 2.556- ~~~Aluminium alloy wire 5005 53.5 0.00353Aluminium alloy wire 6201 52.5 0.00347Aluminium-clad steel core wire 20.3 0.00360Zinc-coated steel core wire 8.5 0.00320Stainless steel No. 304 2.4 0.00130-263268258293149660 3.2226 2.598660 3.2840 2.598660/1300 8.4805 2.670419/1300 20.1 3.9311400 72.0 4.032Refer to Figure 22.18 giving the nomograms for more widely used metals to determine the size of conductor A in terms of duration of fault.

~ ~~~~~~ ~Grounding practices 22/713Example 22.8To determine the minimum size of ground conductor, consider a station grid made of Z, coated steel, having the followingparameters:Parameters AS in /€€€-80 As in IS 3043For an /g of 30 kA,size of groundconductor1 or 0.001 kA(to calculate a generalized factor in rnm2/A)10.003220.13.93141 940__- 200.0032= 293 (Table 22.5). A= ~or ~1' i/3931 00032x200 001 -,O 0061 log, 2 14- 0 001-I 0 6061 x 0 76= 0.0147 mm'IA=68A/mm20.0147= 0.0147 x 30 000= 441* mrn'0 001293 + 419xlo 1 ) loge ( 233 < 40 ]1 or 0.001 kA(to calculate a generalized factor in rnm'1A)10.004513.83.845040~- 200.0045= 2020 0013.8 x lo4-/[ 0 0045 x 1381 loge [ ?:2',=?: 1- 0.001I 0.006 log, 2.69- 0.001\ 0.006 x 0.99= 0.0123 rnm'1Aor ~ =81 Nmrn'0.0123Say 80 Almm'= 0.0123 x 30 000= 369* mrn'*If a future expansion in the generating capacity of the station is envisaged, the grounding grid conductor size so estimatedmay be enhanced by a suitable decrement factor, Df (Section 22.9.6)voltages that a human body can endure. In actual servicethese voltages of the grounding station should not exceedthe prescribed tolerable limits. Thc grounding stationdesign as carried out above must therefore be countercheckedfor these limits. If it exceeds these limits thestation must be redesigned, to contain the actual stepand touch voltages within the prcscribcd Icvcls.IEEE-80 has suggested the following formulae in termsof ground current and total length of buried conductorsto determine the actual step and touch voltages:p' K, . K, . ICMax. step voltage E,(actual) =(22.17)Land mesh or maximum touch voltage E,,(actual)(22.18)wherep = resistivity of- the soilK,. K,,, = geometrical factors, depending upon the moreimportant parameters such as area of thegrounding grid, its depth and conductor spacingand less important factors, such as diameter ofthe conductors and the thickness of the finishingsurface by concrete or gravel.The touch voltage diminishes up to 1 m depthof the grid and then rises rapidly. The idealdepth for economic considerations may be takenas 0.5 m when the touch and step voltages arereasonably low.Ki = corrective factor, accounting for the increasein current densities at the far ends of the gridsystem, the resistivity of the soil and the averagecurrent density per unit length. Zti/L. of buriedconductors.I, = maximum fault current contributed by the powergenerating units.L = total length of the buried conductors of thegrounding station (equation (22.13)).Estimating the step voltage E, (actual)In this caseand the maximum step voltage is assumed to occur ata distance equal to grid depth h. where /z is more than0.25 m and less than 2.5 m. Depths less than 0.25 m maybe rare:

20 -- 5025 -- 4030 I- 3040 -t 50-- 2o428loo-- 10i..,o3 200-- 5L400-- 2.5.3600-800 -1000 - 1 .operiphery of the grid. The ground current dischargingthrough a uniformly spaced ground grid is scanty at thecentre, dense at the edges and a maximum at the corners.Accordingly, the worst step and touch voltages wouldoccur at the outer meshes of the grid, especially at thecorners. To make the current density more uniform, amore non-uniform conductor spacing would thereforebe necessary with more number of meshes at the centreand fewer towards the periphery. Increasing the numberof meshes, i.e. reducing the conductor spacing, wouldtend to reduce the step and touch voltages until a saturationstage is reached, i.e. when, L, approaches Lc. The factor1.15 may now be increased to 1.2 based on field experienceandK, = 0.656 + 0.172 nwheren = the number of conductors on each side of a squaregrid. If the ground grid is not a square and the numberof conductors lengthwise is IZ, and widthwise nb thenr -n = \'ind . n,,Estimating the maximum touch voltage, E,,,(actual)This will largely depend upon the ratio of the currentdensities in the far-end conductors, i.e. conductors at theperiphery, and the innermost conductors and can beexpressed as follows, based on extensive research:Figure 22.18temperatureNomogram for conductor sizing at 40°C ambientwhereK+ 2 log,Kh n(2n- 1)wherew=;;-+-+-+1 1 1 1L 3 4 n-1For n 2 6WE- ' + log, (n - 1) - 0.4232(n - 1)The step voltage falls sharply at higher depths.L = L, + L,when there are only a fewground rods at the peripheryof the gridor = L, + 1.15L,. when there are proportionatelymore ground rods at the farend of the gridwhere Lc = total length of the grid conductors andL, = total length of the ground rodsThe factor 1.15 represents the higher current densityin the ground rods that are placed at the far ends orFor grids without or with only a few ground rods nearthe inner grid (not at the far ends):K, = 41+$where ho is the reference depth of grid = 1 mThe values of Ki and L can be determined along similarlines to those for the step voltage.NoteIn the above equations for maximum step and touch voltages thebest results will be obtained when the following parameters areachieved:n 5 250.25 2 h 5 2.5 md < 0.25 h andD > 2.5 m

Grounding practices 22171522.14.4 Estimating the value of groundconductor length (15)Having determined the safe touch voltage E, (equation(22.8)) the maximum mesh voltage E,, (actual) (equation(22.18)) should be equal to or less than this voltage.Thus by equating these two we can estimate the likelylength L of the ground conductors. Considering this foran average human body of 70 kgE, (actual) 5 E,oo,P K m K, I,,or0.1575 (1.5 C, pI + 1000) x TL \t-P Km K, I,, 1, tor L > (22.19)(ISC, p, + 1000) x 0.157hyardThis is the parameter that will help to decide the size andtype of the grounding grid and design of the mesh.Since there are too many variables and parametersrelated to a grounding station, the following practicalexample illustrates a step-by-step procedure to design agrounding station.Example 22.9To design a grounding grid, consider a power generatingstation as shown in Figure 22.19, transmitting power at400 kV through its own switchyard to another switchyardremotely located.Size of grounding conductorConsider GI for grounding, and the same parameters of Section22.14.1, leading to a minimum size of grounding conductoras 80 Nmm', based on IS 3043. Fault level for a 400 kVpower station as in Table 13.10 is 40 kA (envisaging nofurther rise in the fault level, IC = Ig)= 500 rnrn'If we consider circular conductors thenor4= 500d = 25.24 rnrnFor a power station, assuming a corrosion factor for a lifespanof 40 years as 0.12 mm/year thenCorrosion depth = 2 x 4.8 = 9.6 rnm:. minimum d = 25.24 + 9.6= 34.84say, 35 mmSharing of ground fault currentRated current of the power plant300I,= 3x-13 x 400= 3 x 0.433 kAgridFigure 22.19 General layout of a power plant, station groundinggrid, and switchyard grounding grid etc3 x 0.433g' - (16 + 14.5) x 0.9:. / -(considering the lower side of Zp, to be on the safe side)= 4.73 kAand /g2 = 40 - 4.73= 35.27 kA as illustrated in Figure 22.20where/g, = fault current sharing by the power plant groundinggrid/g2 = fault current sharing by the switchyard groundinggrid (transmission system)/g = total fault current, considered to be 40 kA for a400 kV system as in Table 13.10.For ease of understanding, the rest of the working is shownin the form of Table 22.6.

22/71@e-6 Industrial Power Engineering and Applications HandbookAfter the final designs are complete it is recommendedthat the actual touch E, (actual) and step voltage E, (actual)are rechecked for both power plant and switchyard areasseparately, to ensure that they are within the tolerable limitsas determined above. After the ground stations have beenfinally installed the actual step and touch voltages must bemeasured to verify the designs.NoteThe above example illustrates a simple procedure to designa ground mat in a large power generating station, interconnectedto external supply sources through a power grid.The procedure would be the same with a large switchyard,receiving and transmitting large powers.For small power houses, which may be captive and smallswitchyards or sub-stations, receiving and distributing currentsto industrial or domestic loads, such an elaborate design isnot required and simple grounding stations as discussed inSection 22.1 will be sufficient.Figure 22.20 A simplified layout of Figure 22.19, illustrating thesharing of fault current by the station grounding grid and the switchyardgrounding grid on fault

SoilSurface~'70Table 22.6Safe (tolerab1e)potential difference1 (i) ~ resistivity, p L2m ~(ii) ~ resistivity for cnncrctc of thicl\ne\\ /I, ~ab 250 mm, p\, + 1000) x T-, Ground re\istance (62)R,> = /L(equation (22.12)) L being Iargc= 4 \ A1 thereforei\ ignored cor eax of calculation1 knuinption\ hy experience: Area 90 000\' f70 - 550k = -70 + 550= -- 4*0 = -0.77620C, = 0.850.157= (1.5 x 0.85 x 550 + 1000) x ~\/ 1= (701.25 + 1000) x 0.157= 267 V0.157= 16. x 0.85 x 550 + 1000) x ~= (2805 + 1000) x 0.157= 591 V\'72500I .07070 - 2500k= ~70 + 2500=-~2430 = - 0.952570C, = 0.68= (2550 + 1000) x n. I 57= 557 v= (6 x O.6X x 2500 + 1000) x o.157;I=(l0200+ 1000) ~0.157= 1758V120 000Ground rcsi\ranccTotal ground resistance of power plant andswitchyard areas interconnecled inparallel (recommended practice) where 120.103 x 0.089K,, =0.103 + 0.089= 0.0177

log,56R,, = ground resistance of power plant areaR,, = ground resistance of switchyard areaR,, = total ground resistance of power plantand twitchyard areas in parallelFault current sharing by the two ground matsI(i) Due to power plant,(ii) Due to transmission system,kAkATo estimate 'L', to achieve safe potential differences:p. k, . k, 'I, fiequation (22.19) ' Safe touch voltage (E,)Assuming the following:Area of power plantConsider a rectangular grounding matSpacing between cross conductors (mesh),both lengthwise and widthwise D, m:. No. of conductors lengthwise and lengthand no. of conductors widthwiseand length,and nFor h = 1 md = 0.035 mkAI = 40 x0.0477g' 0.103= 18.54I,2 =40 x 0.04770.089= 21.464.73 x 0.0477 = 2,190.1034.73 x 0.0477 = 2, 540.08935.27 x 0.0477 = 16.3535.27 x 0.0477 =0.103 0.089To determine this, it is necessary that certain assumptions, based on field experience, are made for a possible grounding systemand if necessary, further modifications made to arrive at the desired results and design:90 000360 x 25012.5= 30 x 250 = 7500= 21 x 360 = 7560= 2/3ox21= 25= 0.656 + 0.172n= 0.656 + 0.172 x 25= 4.956=-- 1 - 1(2n)"" (2 x 25)""=-= I 0.731.367120 000400 x 30015--400 + I = 2815= 28 x 300 = 8400=-+1=2130015= 21 x 400 = 8200= \128x21= 24= 0.656 + 0.172 X 24= 4.784-1(2 x 24)2/24- I - 0.521.91= ul:+ 1 = 1.4142n+ 12.5* (12.5 + 2 x 1 15' (15 + 2 x 1)'-~loge[ 16 x I x 0.035 8 x 12.5 x 0.035 4 x 0.035 ] = &[loge( 16 x I x + 0.035 8 x 15 x 0.035 4 X 0.0358n(2 x 25 - I )+ ~0.52 8t 0.73 log, 1.414 rr(2 x 24 - 1)1.414

08'0 = x9.0 =

~ _____~~ ~22/720 Industrial Power Engineering and Applications HandbookRelevant StandardsIEC Title IS BS60298/1990 A.C metal enclosed switchgear and controlgear for rated 3427/1991 BS EN 60298/199660050- I95/1998voltages above 1kV and up to and including 52 kVCode of practice for earthing and protection against 3043/1991 BS 7430/1998electric shocks BS IEC 60050-195AN S UIEEE-37.101/1993ANSUIEEE8 1/1983ANSI/IEEE-80/1991ANSI/IEEE-l42/1991ANSUIEEE141/1993ANSMEEE24 1/199 1ANSUIEEE242/1991ANSUIEEE-367/1987C2- 1997Relevant US Standards ANSI/NEMA and IEEEGuide for generator ground protectionGuide for measuring earth resistivity, ground impedance and earth surface potentials of a ground system. Part I:Normal measurementsGuide for safety in a.c. substation groundingGrounding of industrial and commercial power systems (IEEE Green book)Recommended practice for electric power distribution for industrial plants (IEEE Red book)Recommended practice for electric power systems in commercial buildings (IEEE Grey book)Recommended practice for protection and coordination of industrial and commercial power systems(IEEE. Buff book)Recommended practice for determining the electric power station ground potential rise and inductivevoltage from a power faultNational Electrical Safety code (NESC)Notes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.23Some of the BS or IS standards mentioned against IEC may not be identical.The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedResistance to ground:Plate groundingp = resistivity of soil in RmA = area of each side of the plate in m2Pipe or rod groundingR=- 100. p x [loge 78x1 - 1} R2n. I1 = length of pipe in cmd = internal diameter of pipe in cm.Strip or conductor groundingloopR=- x log, - 2x12 +QOhms2n. 1 h.w1 = length of strip or rod in cmh = depth of strip or rod in cm(22.1)(22.2)(22.3)w = width of strip or twice the diameter of the conductorrod in cmQ = -1 for strip and -1.3 for round conductor groundingSize of the grounding conductorIs=L& (22.4)kS = cross-sectional area of a bare ground conductor inmm2I = r.m.s. value of the ground fault current in amperesK = a factor that would depend upon the material of theconductort = duration of fault in secondsTolerable step voltageEa(50) 5 (6 . c s ' P\ +for a 50 kg body470, 5 (6 ' c, ' P? +for a 70 kg body0.116000) x -.Jt(22.5)0.157000) x - Ji(22.6)

Tolerable touch voltageE,,50, 5 (1.5 c, p, + 1000) x ~for a 50 kg body0.116JiEt(,,), 5 ( I .5 c, p, + 1000) x __zitfor a 70 kg body0.157(22.7)(22.8)n, = no. of conductors lengthwiseb = width of the grid in mnb = no. of conductors widthwisea = length of the grid in m andn = no of grounding rodsa = depth of grounding rods in m(ii) For grid depths > 250 mmGrounding practices 22/721Design parametersI, = lg . DtI, = maximum ground fault currentZg = symmetrical ground fault currentD, = decrement factorBody resistanceand.f??t$h = 6 . c, . p, + Rh(22.9)= 6 . C, . p, + 1000 (22.10)f??fsh = total touch resistance through the bodyRh = body resistanceRzf,,\ = 1.5 x c\ ' p& + Rh= 1.5 x c, ' p\ + 1000 (22.11)R2fp5 = total step resistance through the bodyC, = reduction factor for derating the nominal valueof surface layer resistivity, corresponding toa crushed rock layer of thickness, h, and areflection factor kwherek=-- '-" andP + P\p = ground resistivity in Qmp, = crushed rock (gravel) resistivity in RmGround resistance(i) For grid depths up to 250 mmwhere A = a . bMeasuring average resistivity of soilp=2mR, (22.15)Minimum size of grid conductorsA=(22 161A = cross-sectional area of ground conductorin mm21; = ground fault current in kA (r.m.s.)= it may be substituted with I, (equation(22.9))Tcap = thermal capacity factor in Jlcm'/"Ct = duration of fault in seconds= thermal coefficient of resistivity at areference temperature of 20°Cpz0 = resistivity of ground conductor at areference temperature of 20°C in p Q/cm&)(at 0°C) = reciprocal ofI,,,, = maximum allowable temperature in "Ctamb. = ambient temperature in "CActual maximum touch and step voltages of agrounding station1;R, = station ground resistance in Rp = average resistivity of soil in Rm- 15 ground resistance at the4 ,;A=surface of the ground__ = ground resistance of the total buriedlength ( L) of conductorsTotal length of the buried conductorsL = IZ., . b + ilka + II.. . h metresh(22.12)(22.13)Max. step voltage, E,(actual) =and mesh or maximum touch voltage,E, (actual) =p. K,, . K, .I,LFor details refer to the textP' K, ' K, ' I CL(22.17)(22.1 X)Estimating the value of ground conductor lengthto design the grounding gridp' K, . K, .I, 4L> (22.19)( I .5 C, . p5 + 1000) X 0.157

22/722 Industrial Power Engineering and Applications HandbookFurther readingCentral Board of Irrigation and Power, India, ‘Corrosion of earthingequipment’, Review No. 1, Jan (1973).Central Board of Irrigation and Power, India, ‘Earthing systemparameters for HV, EHV and UHV sub-stations’, TechnicnlReport No. 4Y, Sept. (1985).Central Board of Irrigation and Power, India, ‘Design of earthingmat for high voltage sub-station’, Publication No. 223 Jan (1992).Central Board of Irrigation and Power, India, Workshop on Designund Earthing Sysrenu, April (1 994).Reeves, E.A., Handbook of Elecrricul fn.stul/arion Prucrices.

PART IVPower Capacitors

Power Capacitors:Behaviour,SwitchingPhenomena andImprovement ofPower Factor

Power capacitors: behaviour, switching and improvement of power factor 23172723.1 IntroductionIn view of the considerable increase in power distributionnetworks and their overutilization to meet increasingconsumer and industrial demands it has become imperativeto optimize the use of available power through efficienttransmission and distribution.Voltage and power factor (p.t‘.) are the two mostimportant parameters in a power system that influenceits utilization. The element of voltage is optimized byraising the transmission and distribution voltages as muchas feasible. The more prevalent of these are 400 kV a.c.for long-distance transmissions and 33-1 32 kV or evenhigher for secondary transmissions. Figure 23. I illustratesa typical transmission and distribution network. Continuedefforts are being made to raise the transmission voltageto 750 kV ax. or 500 kV d.c., or higher. Some countriessuch as the USA, Russia and Canada have already adoptedsuch systems. A d.c. system, we recall, has no skin effect(Section 28.7) and can transmit power at unity p.f. Athigher p.f.s, the line losses (I’R) are low for the samepower transmitted. It is estimated that a d.c. transmissionsystem can transmit about threefold more power for thesame cost as an ax. transmission system.The element of p.f. mainly affects the secondarydistribution system which serves industries, agriculture,public utilities and domestic loads. Most of them arehighly inductive and result in lowering the system p.f.These loads are largely responsible for most of thedistribution losses and voltage fluctuations at the consumerend. In developing countries it is estimated that usefulpower is lost mainly due to transmission and distributionlosses. In India, for instance, it is estimated to result ina loss of about 18-20% of the total useful power, mostof which occurs at the secondary distribution attributableto low p.f.s.The application of power capacitors, can tackleproblems of both low p.f. and voltage fluctuations andthese aspects are discussed here.23.2 Application of power capacitorsA capacitor draws leading current. When connected toan inductive circuit. it offsets its inductive (reactive)component and improves the p.f. of the circuit. It can beapplied in two ways and is accordingly classified asfollows:1 Shunt capacitor - connected across the inductive circuitto improve its p.f.2 Series capacitor - connected in series at the far end ofa long transmission or HT* distribution line to offsetthe reactive component of the line impedance, containthe voltage drop and enhance the receiving-end voltage.It can support a transmission or distribution systemin the following ways:Improving the regulation of the system at thereceiving endLimiting the system voltage swing during a loadrejection or off-peak periods, and protect it fromovervoltagesEnhancing the stability of the system by minimizingthe voltage fluctuations caused by load variationsandEnhancing the power-carrying capacity of the systemby reducing the I’R losses.This subject is dealt with in more detail in Chapter 24.23.3 Effect of low PFA low p.f. means a higher load current than necessaryand accompanying higher line losses. Inductive loadsare the main cause of a low p.f., with induction motorsthe major contributors. Under operating conditions a motormay often be operating underloaded due to one or moreof the following reasons:While making selection of even for a standard motor,it is generally not possible to exactly match the ratingof the machine with the load. The motor may havesome reserve capacity.Users may select a slightly larger machine to ensuresafety.When selecting a machine for more critical installations,*Here HT meanr all voltages of 2.4 kV and aboveExtra high voltage High1medium voltage H T distribution for bulk suppliestransmission transmission and large installations132/220/11115 7512 1/24 kVLT distributionGenerating station Receiving station 1 st distribution Main receiving and distribution stationFigure 23.1A typical transmission and distribution network

23/728 Industrial Power Engineering and Applications Handbookas discussed in Section 7.22 the number of deratingsfor all unfavourable operating conditions are assumedto be occurring at the same time. This results in theselection of an oversized machine as all the unfavourableconditions may not be occurring simultaneously.Wide voltage fluctuations may be prevalent in a ruraldistribution system, particularly in developing countries.In such cases it is common practice for users to selectan oversized machine for their needs. Accordingly,the motors employed for loads such as pumps, thrashersand winnowers are normally over-rated and underutilized.Also the same motor may have to cater fordifferent types of loads, at different times, and theseloads may be much less than the motor rating.The power factor of a motor decreases sharply at loadslower than rated as discussed in Section 1.8. All theabove factors, contribute to reducing the overall systempower factor, which is sometimes seen to reach a low of0.6 or even less on an LT distribution network.Higher line losses result in higher voltage fluctuationsdue to the greater line drop (IZ). In an HT system, thevoltage at the receiving end is not as much affected as inan LT system due to the lower voltage drop on an HTsystem as a percentage of the system voltage. An LTsystem experiences a much higher fluctuation, as apercentage of the supply voltage, due to excessive voltagedrops. A higher current for a smaller load, due to a lowpower factor, reduces the capacity of the distributionnetwork. In other words, it decreases the generating capacityof the power plant, feeding such a system. For instance,for the same power generated or distributed ‘P’P= &.V.I-cos$Therefore, for a given load current I,P = cos $Thus, at 0.70 p.f., the generating or the distributing capacityof the system, compared to a system having a p.f. of0.90, will reduce to approximately 77.8%, (0.7/0.9 x100%). In other words, if we consider a distribution p.f.0.70 to be improved to 0.9, then the useful power can beenhanced by0.90 - 0.70 100%0.70i.e. 28.6% of the existing system (see Figure 23.2).For the same generation or distribution current I, at a p.f.cos $,, utilization of the system will bePI = I x cos @Iwhereas at an improved p.f. cos 4, it will beP, = I x cos &Thus while the Z2R loss will remain the same in bothcases, at a lower power factor the utilization capacity ofthe system will reduce in the same proportion as thepower factors, i.e.cos @Icos $20.21,by0.21, i.e. m o r. f28.6% of the existingsystem for the sameI‘R tossesFigure 23.2factor0.71, 0.91‘Better utilization of power with improved powerAs a result of higher voltage drops, the receiving-endvoltage may sometimes fall below the required limit whichin, electric motors, is -6% (Section 1.6.2). Motors forsuch locations may be required to be suitable for voltageslower than the standard system voltages such as 400 V,380 V or even less, as against the standard 415 V. Amotor designed especially for a lower voltage than standardmay sometimes require a larger frame size at a highercost.23.4 Other benefits of an improvedpower factorReduced kVA demand will result in lower tariffs sincethe electricity companies usually charge users on thebasis of their maximum kVA demand.In certain countries the consumer may even be entitledto a rebate for maintaining a high power factor, insteadof paying a higher tariff.Lower kVA demand will reduce the load current (dueto reduced Z2R losses) and result in an economicalselection of switchgear components and cables.Lower voltage drops in the lines and cables and thuslesser voltage fluctuations.The electricity companies would also benefit due tobetter utilization of their distribution system and makemore power available to consumers.23.5 Behaviour of a power capacitorin operationBefore we discuss the application of this device, it isimportant that we study its behaviour during operation.A capacitor unit behaves like a short-circuit on beingenergized and retains its charge for a brief period evenwhen the source of supply is removed. This behaviourgives rise to the following:I,

Switching overvoltagesHarmonic effects and inductive interferences andExcessive charging currents.These aspects are discussed briefly below.Power capacitors: behaviour, switching and improvement of power factor 23/72923.5.1 Switching overvoltages (surges)This phenomenon is generally associated with an HTsystem. We have discussed in Section 17.7 the subject ofswitching surges as related to an inductive circuit.Switching phenomenon in a capacitive circuit is equallyimportant, and a little more complex. It therefore requiresmore careful handling of this device. Although theswitching behaviour of a capacitor circuit is almost areplica of an inductive switching, its current leading theapplied voltage by 90" results in more complex behaviourthan an inductive circuit during a switching operation.The phasor displacement between the voltage and thecurrent in an inductive circuit, particularly an inductionmotor or a transformer, is much less than 90" duringnormal operation. At a p.f. of 0.65, for instance, thecurrcnt will lag the voltage by 49" and at a p.f. of 0.9 by25". During the most severe conditions, such as on afault, at near a p.f. of 0. I, the current will lag the voltageby 84" and at a p.f. of 0.3 by 72". In capacitor switching,in addition to 90" phasor displacement between the appliedvoltage and the current, under all conditions of switchingthere will also appear an overvoltage across the partingcontacts, depending upon the grounding method of thecapacitor units, their configuration and the applied voltage.All such overvoltages must be limited within the impulsewithstand level of the capacitors, as noted in Section26.32. We have analysed thete aspects for the followingcapacitor configurations:Grounded star (a widely adopted practice in LT systems)Ungrounded starDelta connections andParallel switching of capacitor units.(i) Grounded star capacitor unitsSee the arrangement shown in Figure 23.3(a) and itsequivalent single-phase circuit in Figure 23.4. We arediscussing the phenomenon of switching overvoltagesassociated with an HT system. An LT system, althoughnot immune to over-voltages, causes no restriking betweenthe parting contacts of the interrupting device as a resultof inadequate transient recovery voltage (TRV; Section17.6.2) and generates no switching surges.Consider the opening of a switch on one pole (Section19.7) at the instant of contact interruption, say at point11' on the current wave (Figure 23.5). The current willlead the voltagc by 90" and the parting contacts will besubjected to a full system voltage of I p.u. Refer to point'a' on the voltage wave. In addition, the contacts willalso be subjected to the trapped charge of 1 p.u. withinthe capacitor unit, which it will retain for a while. If thecontacts fail to interrupt at point a', the arc will strikeagain (re-ignition of the arc plasma, Section 19.4). Bythe next current zero (point h' on the current wave),C1(a) Grounded star (b) Ungrounded star (c) Delta connectionFigure 23.3 Configurations and grounding methods of capacitorbanksRand X, are voltage damping impedancesFigure 23.4 A single capacitor switching17 c1when the arc ceases again and the contacts try to interrupt,they will be subjected to a TRV of 3 P.u., the capacitorunit still retaining the charge. Refer to point bb" on thevoltage wave. It is possible that the contacts are not ableto travel sufficiently apart to build up the required dielectricstrength, and the arc may strike again. At the next currentzero (point c' on the current wave), when the contactshave travelled by another one half of a cycle, they willbe subjected to four times the system voltage CC" on thevoltage wave, considering that the capacitor unit maystill be retaining its charge. The arc may now beextinguished as a result of self-attenuation, as the contactswill have travelled sufficiently far apart to withstand theTRV and interrupt the circuit. The process may be repeatedif the dielectric strength or the travel of the contacts isnot sufficient to withstand a TRV of four times the ratedvoltage. It will continue to do so until the process attenuateson its own due to circuit parameters Land C, the decreasingcapacitor charge and adequate travel of the moving contactof the interrupting device and extinguish the arc to finallyinterrupt the circuit. Field experiments have revealedthat overvoltages due to such repeated restrikes during

23/730 Industrial Power Engineering and Applications HandbookVoltaoe wave ’P‘ I I ISwitch recoveryic“3rd interruptionI cycle I cycle(90” leading)\ I I2nd current3rd currentzero zero zero1st 1 st 2ndinterruption restrike restrikeTime --mFigure 23.5Build-up of capacitive voltage (TRV) on restrike during contact interruption in a grounded star HT capacitor unitthe interruption of a grounded capacitor unit seldom exceed2.6 p.u. as a result of self-attenuation and high-speedmodern interrupters.(ii) Ungrounded capacitor units (both star- anddelta-connected units)Consider the arrangements shown in Figures 23.3(b) and(c). Now the opening of the switch in one phase will notisolate the capacitor in that phase from the system, as inthe grounded system as a result of a feedback throughthe lumped (leakage) capacitances of the other two phases(Section 20.1 and Figure 20.2). Assume the initiation ofinterruption in one of the phases at a current zero. Referto point a’ on the current wave (Figure 23.6(a)). Theparting contacts will now be subjected to an additionalvoltage equal to the system voltage of 1 P.u., because ofthe other two phases, which would also be conducting,besides 2 p.u. at point ‘a’ as discussed in the previouscase. The other two phases which are now conductingwill each retain a charge of 0.5 p.u. at every voltagepeak in the first phase, as illustrated in Figure 23.6(b).This charge will be redistributed in the first phase assoon as its contacts open. The unswitched phases, however,will retain a charge of0.5 p.u. + 0.866 p.u. or 1.366 p.u. (max.)where 0.5 p.u. = trapped charge and2170.866 p.u. = 7 p.u. is the excitation due to lumpedIcapacitances of the unswitched phases.The parting contacts will therefore now be subjectedto a TRV of 3 p.u. at the first interruption. If the contactsfail to interrupt at point a’, the arc will strike again, andby the next current zero (point b’ on the current wave),when they will try to interrupt again, will be subjected toa TRV of 4 p.u. Refer to point bb”on the voltage wave,assuming that the capacitor unit is still retaining its charge.The cycle of an arc restrike will be repeated at the nextcurrent zero c’, at 5 P.u., point cc” on the voltage wave.The process will continue until the circuit will finallyinterrupt by virtue of the interrupter’s ability to interruptor attenuation of the TRV, whichever occurs first. Fieldexperiments have revealed that overvoltages due torepeated restrikes during the interruption of an ungroundedcapacitor unit, connected in star or delta, will seldomexceed 3 p.u. due to self-attenuation and high-speedmodern interrupters. The system grounding has no effecton such ungrounded capacitor units. The interruption of

Power capacitors: behaviour, switching and improvement of power factor 231731Charge because oflumped capacitancesd't '-113uChargeretained -f--- - 1Voltagewave-IPU3puI--Arc extinguishes andcontacts interrwt;.:5PU! Iungrounded capacitor units i\ thus more \evere than ofgrounded units.(iii) Parallel switching of capacitor unitsThe amplitudes of overvoltages (TRVs) caused by a singlecapacitor switching, as discussed above. will diminishinto when the capacitor interrupts a circuit that alreadyhas energized capacitor units of comparable sizes in thesame system. At the instant of contact separation, the twounits, which may be charged at different potentials. willget discharged into each other and their cumulative voltagewill tend to settle at an intermediate value that will dependupon the size ofthe capacitors being switched. The largerthe line banks (already charged capacitors). the lowerwill be the amplitude of the TRV of the combined units.4PU5pu& 112 112 -Current wavecycle cycle(90" leading) ! ' I, Arc extinqutshes andinterruption restrike restrikeTmeFigure 23.6(a) Build up of capacitive voltage (TRV) onrestrike during a contact interruption in an ungrounded star HTcapacitor unitISummary of overvoltages (on HT capacitorswitchings)1 Switching overvoltages is generally a phenomenonof HT circuits.2 The choice of the type of capacitor connections willlargely depend upon the system. its fault level andthe presence of a communication system in thevicinity (as a result of harmonic effects and inductiveinterferences). Different electricity companies mayadopt to different practices of capacitor connectionsand their switchings, depending upon these Factors.A more detailed study on this subject (harmonicsand inductive interferences) is beyond the scope ofthis handbook, but these effects are discussed brieflyin Section 23.5.2. Various papers written on thesubject and a study of existing systems and the normalpractices to deal with these effects and interferencesare available and may be referred to for more details.See the further reading at the end of this chapter.Overvoltages, as developed on different groundingsystems, are discussed in more detail in Section 20.2.3 When interrupting a capacitive circuit while the1 YRAt every current zeroin phase 'R' totalcapacitor charge 'Cc',- --C,=V,+G+G2 2=2Vmor2pu[ V, = phase to neutralpeak voltage]Figure 23.6(b) Capacitor charge in an ungrounded system

231732 Industrial Power Engineering and Applications Handbook45678910capacitor is fully charged, the system will experiencean overvoltage and the arc between the moving contactsof an interrupting device (a breaker or a contactor)may restrike even more severely than an inductivecircuit, due to higher restriking voltages (TRVs).Some resistance andor inductance is recommendedto be introduced into the switching circuit to dampenthe restriking voltage by altering the p.f. of theswitching circuit during a make or a break, by helpingthe voltage and the current phasors to move closerso that on a current zero, the TRV is much less thanV,. See also Figures 17.1 l(a) and (b) for an inductivecircuit. The situation will be almost the same for acapacitive switching except that the current will nowbe leading the voltage by 90" instead of lagging atmuch less than 90".The impedance of the capacitor switching circuitwill determine the attenuation of the striking voltageand the time of arc extinction.Frequent switching operations must be avoided asfar as possible. In an HT system, they are as suchlow. However, where automatic switching deviceshave been installed, frequent switching operationsmay be likely. As standard practice, therefore,precautions should be taken to allow a pause beforethe next switching, or introducing a rapid dischargedevice across each capacitor unit to allow the lastdisconnected unit discharge to a safer value beforereswitching (Section 25.7). Nevertheless, switchingsurges must be dampened as much as possible byintroducing some resistance or inductance into theswitching circuit, as noted above.The following are the recommended values of theswitching transient voltages that may be consideredto select the switching device:(a) Grounded capacitors units - peak recoveryvoltage (TRV) on a healthy switching up to 2.6p,u.(b) Ungrounded star- and delta-connected capacitorunits -peak recovery voltage (TRV) up to 3 p.u.Where necessary, surge arresters may be used todampen the high transient switching voltages. Referto Section 17.1 1.A capacitor will retain its charge, even after disconnection,as it takes time to decay. If it is connectedacross a motor it is possible that the capacitormagnetizingkVAr may excite the motor during anidle period, and make it act like an induction generatorand result in an overvoltage (Section 23.13). Thisovervoltage may pose yet another switching problemin the inductive circuit, as discussed in Section 17.7.The capacitor manufacturers, as standard practice,provide a discharge resistance across each capacitorunit to discharge the trapped charge as soon as thecapacitor is disconnected. For rate of discharge seeSection 26.3.1(15).As an energized capacitor retains its charge, evenafter a disconnection and causes a voltage transienton a reswitching, it is recommended that its circuitbe closed only through a contactor or a breaker,with an undervoltage release. So that in the event ofa power failure, the circuit will interrupt automaticallyand can be reclosed either manually or through apower factor correction relay to allow a pause beforereclosing.23.5.2 Harmonic effects and inductiveinterferencesDistortions in voltage and current waveforms, from theirsinusoidal waveforms, are termed harmonic disorders.They are caused by equipment whose impedance varieswith the voltage or current change (non-linear loads).This includes induction motors, fluorescent lamps, batterychargers, silicon-controlled rectifiers (SCRs) and weldingequipment as well as any magnetic core that saturates atvarying degrees of applied voltage. A sinusoidal (a.c.)voltage that rises and falls with each one half of a cyclecauses such an effect. Such loads are saturable reactorsand power transformers (Section 1.2.1). The amplitudesof such distortions (harmonics) magnify with resonanceeffects between the system and the load impedances.The amplitude of the thus distorted sinusoidal waveformof the voltage or the current quantities is the phasor sumof all the individual sinusoidal waveforms of eachharmonic present in the system, as if all the harmonicsare superimposed on each other (Figure 23.7). A harmonicis the integral multiple of the fundamental frequency,such as the third, fifth, seventh and eleventh having afrequency of 3f, 5J 75 and I Ifrespectively ('fbeing thefundamental frequency of the system).Generally, a few harmonic quantities are always presentin an a.c. system. as normally some non-linear loadsalways constitute a power system. These harmonics maynot affect the operation of the inductive loads connectedon the system as much as a capacitive load or acommunication network that may be running in parallelor in the vicinity of such a power system.Capacitors themselves do not generate harmonics butthey do magnify their amplitudes, when connected in3rd harmonic in phase opposition with the fundamental wave5th harmonic in phase with the fundamental waveFigure 23.7 Effective amplitude of a particular waveform withthird and fifth harmonics

.Power capacitors: behaviour, switching and improvement of power factor 23/733the system. If the capacitors are grounded star, they willprovide a return path for the third harmonic quantitythrough their grounded neutral and help the system togenerate third harmonic disorders. As the harmoniccomponents affect a capacitive circuit more than anyother equipment connected in the system, our mainemphasis will be a study of the harmonic effect on acapacitive circuit.But as the harmonics do exist in the system, they doaffect an inductive load. They may also disturb acommunications network as a result of capacitive coupling,whose effects are magnified in the presence of capacitorunits in the power system. It is therefore consideredrelevant to discuss this subject in more detail to makethe harmonic study more informative.Here we briefly discuss the sources of generation ofharmonics of different orders, their likely magnitudesand possible influence on the capacitor and inductiveloads, connected in the system. We also study theirinfluence on a communications network. Such a networkis affected when it is running in parallel, and in closevicinity of long-distance HT distribution power lines.Sometimes the communication lines may be runningthrough the same structures on which the power lines arerunning.A Effects of harmonics on the performance of acapacitor unitB\ the harriionic voltage\The effective harmonic voltage can be expresed bywhere V,, = effective harmonic vultageV, = system voltage andV,,?, Vh5, Vh7 and Vh, etc. = magnitudes of theharmonic voltage components in terms offundamental voltage at different harmonic orders.Referring to the data available from experiments, asshown in Table 23. I , it has been estimated that a Vh of1. I VI should be sufficient to account for the harmoniceffects. For this dielectric strength is designed a capacitorunit and selected a switching or protective device.B: the knrrmtiic currentsA harmonic component affects the performance of acapacitor unit significantly due to diminishing reactanceat higher frequencies, which adds to its loadingsub\tantially and can be analyscd as follows:If n IS the harmonic order, such as 3, 5, 7 and 9 etc.. thenthe harmonic frequencyand harmonic reactanceThis means that the capacitor will offer a low reactanceto the higher harmonics and will tend to magnify theharmonic effect due to higher harmonic currents onaccount of this. In fact, harmonic currents have a greaterheating effect too compared to the fundamental componentdue to the skin effect (Section 28.7).where Ich and Xch have been considered as the capacitivecurrent and the reactance of the capacitor respectively.at a particular harmonic frequency& = n ..f: The effectivecurrent caused by all the harmonics present in the systemcan be expressed byI,,, = d(1: + 9 I& + 25 + 49 If,], + . . . t i2 lL'hn)(23.2)where I, = rated current of the capacitor. andIch3, lchs, lch7 and lchn etc = magnitude ol' theharmonic current components at different harmonicorders.Based on the system studies carried out and Table23.1, it has been assessed that in actual operation, effectivecurrent through a capacitor circuit may increase up to1.3 times its rated current, I,, Le. Ich = 1.3 I, to accountfor all the harmonic effects (V; ,f,,: equation (23.4)). Acapacitor unit is thus designed for at least 30% continuousoverload capacity (Section 25.6). Its switching andprotective devices are selected along similar lines.Summarizing the above, the harmonic quantities whenpresent in a system on which are connected a few capacitorbanks affect the capacitors as follows:Overcurrent will mean higher losses ([:h . R).Overcurrent will also mean an overvoltage across thecapacitor units, which would inflict greater dielectricstresses on the capacitor elements.Since the harmonic disorders occur at higher frequenciesthan the fundamental Uh >,f). they cause higher dielectriclosses due to a higher skin effect.Harmonic output of a capacitor unitThe rating of a shunt capacitor unitkVAr =fi 1Vand I, = -:. kVAr = ~or kVAr =(v in volts and 1, in amperes)1000xca. I"1000 . x,217 ' v2 ' 2nf.c.1000Generalizing, kVAr,, oi V;01kVAr, a p,2.fl,(23.3)(23.4)+ 3 . Vi3 + 5 . Vh5 + 7 + . . . II . V,; )(23.5)

23/734 Industrial Power Engineering and Applications HandbookThe rating of a capacitor unit will thus vary in a squareproportion of the effective harmonic voltage and in adirect proportion to the harmonic frequency. This rise inkVAr, however, will not contribute to improvement ofthe system p.f. but only of the overloading of the capacitorsthemselves.When determining the actual load current of a capacitor unit inoperation, a factor of 1.15 is additionally considered to accountfor the allowable tolerance in the capacitance value of thecapacitor unit (Section 26.3.1( I))::. Effective kVArh = 1.3 x 1. IS- 1.5 times the rated kVAiand for which all switching and protective devices must beselected. It may, however, sometimes be desirable to furtherenhance the overloading capacity of the capacitor and so alsothe rating of the current-carrying components if the circuitconditions and type of loads connected on the system are proneto generate excessive harmonics. Examples are when they areconnected on a system on which we operating static drives andarc furnaces.It is desirable to contain the harmonic effects as far as practicableto protect the capacitors as well as inductive loads connectedon the system and the communication network, if running inthe vicinity.Influence of harmonics on the performance ofan inductive loadIn an inductive circuit the presence of harmonics is notfelt so much as the higher harmonics tend to enhance theharmonic reactance, which causes a dampening effect:XL = 277 .f. LFor a harmonic order n and harmonic frequency nharmonic reactance will becomeXLh = 277. n .f. L i.e.n .ftheAt higher harmonic frequencies, X,, will rise proportionatelyand reduce the harmonic currents andprovide a dampening effect. The harmonic quantitiespresent in the system thus do not have any significanteffect on the performance of an inductive load, whichcan be a generator, transformer, motor, cables andoverhead lines etc. In an LT system, such effects arehighly dampened due to high system impedance. Nospecial efforts are therefore generally made to suppressthe harmonic effects in an LT system.The magnetic core of an electromagnetic equipment(generator, transformer, reactor and motor etc.) will,however, generate additional no-load iron losses athigher harmonic frequencies (equations (1.12) and(1.13)). The content of iron losses, being low, will riseonly marginalIy and can be ignored for all generalapplications.Electronic appliances that are highly susceptible tosuch effects are, however, provided as standard practicewith harmonic filters in their incoming circuits.C Effects of harmonics on telephone linesIn earlier years, to reach a remote area, where separatetelephone lines had not been laid it was normal practiceto run them through the same poles as the HT powerdistribution lines (generally 11-33 kV). This wasparticularly true of internal communications of theelectricity companies for ease of operation and to savecosts and time. This communication was known as themagneto-telephone system. But the proximity of telephoneIines to power lines adversely affected the performanceof the telephone lines due to generation of overvoltages(Chapter 20) and electrical interferences (conductive andinductive interferences, discussed later) on the telephonelines by the power lines. Some of these interferences,particularly system harmonics, had the same frequencyas the audio frequency of the telephone lines and affectedtheir audio quality.The running of telephone lines through power lines islong discontinued. They are now run on separate structureswithin a city and nearby areas at audio frequency (=0.3-3.4 kHz), and maintain enough distance from HTpower distribution lines. They are therefore almostunaffected from such disturbances. Nevertheless,interferences must be kept in mind when installing theselines so that they are out of the inductive interferencezone of the power lines. The latest method in the field ofcommunications to avoid disturbances is to useunderground optical fibre cables, where possible, asdiscussed later. Optical fibre cables are totally immuneto such disturbances.Remedies for electrical interferencesThe following measures are recommended to mitigatethese problems:Use of screened cables in the communication networkand grounding the screen effectively. Metallic-sheathor armoured cables are not recommended.Transposing the overhead communication lines, i.e.reversing the respective positions of thc two sides ofthe lines every I km or so, to avoid continuousparallelism (due to electrostatic and electromagneticinductions), as illustrated in Figure 23.8. See alsoSection 28.8.4(3) on phase transposition.It is common practice to leave the star-connectedcapacitor banks ungrounded when used in the systemor use delta-connected banks to prevent the flow ofthird harmonic currents into the power system throughthe grounded neutral.A 0Figure 23.8 Transposition of overhead communication lines

Power capacitors: behaviour, switching and improvement of power factor 2317354 Use of filter circuits in the power lines at suitablelocations, to drain the excessive harmonic quantitiesof the system into the filter circuits.Filter circuitsA filter circuit is a combination of capacitor and seriesreactance, tuned to a particulrv harmonic frequency (seriesresonance), to offer it the least impedance at that frequencyand hence, filter it out. Say, for the fifth harmonic, Xc5 =XLS.Since,fh, = 5 .f.. =27r.Sf-L2n-5f.Corx,= 5X,-S01' X, = 5' ' X,Generalizing, X, = N; . X, for the N,,th ordinal numberof the harmonic disorder.A capacitor is susceptible to variations in the systemvoltage, frequency and operating temperature. Failure ofcapacitor elements may also cause a variation in capacitivevalues in the three phases of the filter circuit. Underthese conditions, it is possible that a filter circuit mayfall out of tune during normal service and cause excessivecurrents in one or more phases due to sub-synchronousor ferro-resonance effects with the impedance of the mainsystem. If we consider the reactance of the detuned circuitas X, (considering it to be capacitive after detuning, asX, will rise when C of the capacitor units decreases),which may fall in phase opposition to the system reactanceX, (considering it to be inductive), as shown in Figure23.15(b) then it will lead to such a situation and causeexcessive harmonic currents through the filter circuit,due to its low effective impedance, 2, sinceZ = R,, + XL -Xc(R,, is the resistance of the main system)= very lowIt is therefore recommended that a small resistance ofa low I?. R loss be introduced into the filter circuits asshown in Figure 23.15(a) to limit such an excessive flowof currents through them. Knowledge of the systemparameters (resistance and reactance) is also essential todesign an appropriate filter circuit to avoid a possibleresonance in the first instance. If this occurs the resistancethus introduced will limit the excessive flow of current.5 Another remedial measure is the use of blockingcircuits by providing a high-impedance path in the groundcircuit when a phase-to-ground circuit is used for thecom-munication network, to block the entry of the thirdharmonic quantities into the system from other systems.Blocking circuitsThese are parallel resonant L-C circuits, and are tuned tooffer a high impedance to a particular harmonic frequencyto block its entry into the system through its neutralfrom another system. For other frequencies, particularlyat power frequency, it will offer a very low impedanceand permit them to flow into the system. The filter circuitmay now be designed for a notch frequency of1fh = 2n. JLC (23.6)At this frequency, the filter circuit will offer a very highimpedance and block the entry of that harmonic to flowinto the circuit through its neutral. For other frequencies,it will provide a very low impedance and permit them toflow into the circuit. Thus for any value offh the productof LC can be determined, and hence the blockingcircuit can be designed. To achieve a narrow bandwidthof attenuation, at near ,fh. the blocking circuit may bedesigned separately for each major harmonic frequencyCf,). To ensure a high Xch, it is advisable to keep thevalue of L low and hence limit the inflow of chargingcurrents.Application of such circuits is also provided in theform of a line trap, which is a blocking circuit and isused at each end of the line to which the carrier relayingis connected as discussed later. A bypass or filter circuitis also provided to drain the power frequency componentof the current when present in the communication network.One end is connected to the main line through a couplingcapacitor or CVT and the other to the ground throughthe PLCC (Power Line Carrier Communication), asillustrated in Figure 23.9(b). Leading manufacturerscombine the coupling equipment and PLCC couplingdevice into one unit for ease of application and installation.Use of tertiav (auxiliary) windingWhen a power transformer is y/)' or y/y connectedand is feeding a system on which harmonics are notdesirable, it is generally provided with a tertiary winding.A tertiary winding, which is also called an auxiliarywinding, is an additional winding connected in A. andsandwiched between the main primary and the secondarywindings of the transformer. It thus has the same magneticcircuit. This winding provides a closed path for the thirdharmonic quantities to circulate and prevents them fromappearing on the load side. It may also be used for thepurpose of metering, indication and protection of groundfault currents.Other than the system harmonics, electrical interferencesare also caused by line disturbances, which may be causedby lightning, switching, sparking or a fault. As discussedin Chapter 17, line disturbances occur at very highfrequencies but some may coincide with the audiofrequency of telephone lines. and cause disturbance inthe audio quality of the telephone system. AI1 thesedisturbances are referred to as inductive interferences.We provide more details on inductive interferences inthe box below, to make the subject of electrical interferencemore informative. We also provide a passing referenceto communication systems being adopted worldwide.

23/736 Industrial Power Engineering and Applications HandbookInfluence of HV and EHV system disturbances on communication servicesPower and communication lines are not run through the same structures due to very high transmission and distribution vol~ages.Referring to long HV” or EHV* power lines, electrical interferences are caused whenever these lines and the communication linesrun in parallel in close vicinity to each other and the distance between them falls short of the inductive zone of the power lines. Toprotect the communication lines from inductive interferences their distance from the power lines is maintained such as to containthe maximum radio influence voltage (RIV) of the power lines on the communication lines within 100 pV at 1000 kHz, as inNEMA-107. The main causes of electrical interferences in an overhead communication network are distortions in the voltage andcurrent waveforms in the overhead power lines, as discussed below.1 Electrostatic induction (a voltage effect)This is the prime cause of noise and distortion in an audio system. The capacitive coupling (conduction) betwccn the power and thecommunication lines gives rise to such an effect. It is associated more with the voltage of the system and particularly when it iscapacitor compensated. Even without the power capacitors, the leakage (coupling) capacitances between the HV or EHV powerlines, particularly 132 kV and above, and the overhead communication lines play an important role and give rise to this phenomenon.Systems lower than 132 kV do not cause such a situation as a result of the insignificant capacitive effect.Normally amall charging currents will flow from the main power lines to the communication lines through their couplingcapacitances C,, (Figure 23.9(a)). The charging currents find their return path through the grounded communication lines (when thecommunication lines have their return path through the ground), having a ground capacitance C,. These capacitances form a sortof voltage divider and induce an electrostatic voltage into the communication lines which can be expressed bywhere V, = induced electrostatic voltage in the communication linesC, = coupling capacitance between the power and the communication lines andC, = ground circuit capacitance of the grounded communication lines.Both C, and C,, are related to the system voltage, as can be inferred from Table 24.1(b). The field induced by them is termedan electric field (non-magnetic). This field, i.e. content of C, and C,, rises with an increase in the system voltage.V, = nominal voltage of the power system.The voltage, V,, is the main source of distortion in the audio quality of speech and the carrier waves. In normal conditions (at powerfrequency) the electrostatic effect may be feeble but during a line disturbance, the charging currents (V,,/Xch) magnify due to highersystem frequencies (highfh) and the diminishing value of coupling impedance,1x,, = -2Zfh c cand cause severe distortions in the audio and carrier waves. The higher the system frequency on a disturbnnce, the higher will be4Power line (only/-one phase shown)(23.7)./lr,-Communicationx,, linei ---- - 111- --------1c -‘The return path of thecommunication circuitis through the ground* Longitudinal voltage (Source of danger to the telephoneequipment and operating personnel)Figure 23.9(a) Electrostatic influence on a communication line* We have classified the different voltage systems as follows:HT - up to 66 kV,HV - above 66 to 245 kV, andEHV - above 245 kV.UHV - 1150 kV (ultra high voltage used in the USA)

Power capacitors: behaviour, switching and improvement of power factor 23/737the distortion level. even for their (harmonics) small contents (Table 23. I ). During a phase-to-ground Fault. for iii\taiice. the groundpotential may rise (Section 20.1 ) and cause much higher inductive interferences (electrostatic induction). that may jeopardix theconimunication sy\tems particularly those serving essential services such as railways. defence installationh, power generation andtransnii\sion. The ground conductor now acts like a large-diameter solenoid producing high induced currents in the groundedcommunication network and raises its ground potential. It gives rise to noise disturbances and affect\ its audio quality. It may alsodistort the carrier n,aves (-30-500 kHz) such as those u\ed in a power line carrier communication (PLCC) network. and mutilatethe tran5fer of vital data or \end$ out wrong signals. To achieve better reliability, \uch as on a critical powei- net\vork. couplingbetween phnsc to phaw is uwd. We provide the layout of ruch ;I \y\tein in Figut-e 23.9(b).2 Electroniagnetic induction (a current effect)This may occur as a result of electromagnetic induction between the power and the communication line\ due to their- proxiinit!. Thcmagnetic tield produced by the main power conductors may infringe upon nearby existing communication lines. As such. it ma) notc;iusc \trious disturbance due to adequate \pacing (>305 mm. Section 28.8) between the power and the comniunication line\ Butin ii grounded communication network the situation may deteriorate. during a single phase-to-ground fault in the power linea. Thegrounded conductor which will carry the ground fault current will no\\ act a\ a vel-y lai-ge-diameter solcnoid and produce highmagnetic fields. inducing heavy overvoltages in the coiniiiunication lines significantly affecting audio quality. Whci-e ii more reli.iblci~iiiiiiiiitiiciiti(iii network i': considered imperative. this is achieved by connecting the network hetween ph;i\r IO phaw than pha\eti) ground.All effects caud by electrostatic or electromagnetic inductions are termed Inductive Interferences. With the u\e of gla\\ opticalfibre cables in new installation\. this effect is werconie automatically. Optical fibre cable\. as discussed later, ha\e no metal contentand carrq no electrical \ignals. Therefore the above discussion is more appropriate for existing in\tallations and rilso to pro\ idc atheoretical a\pcct atid more clarity on the phenomena of inductive interference\. These can also be applied to other field\ rather thancommunications alonc.In the e:irIicr installations sensitiYc to such interferences the normal practice wa\ coordination between the gcnei-atrng and powciti-an\niis\iotiagencies and the authorities of e\senrial rervices (such as public telephones, defence services and railway\). whopro\ idc their own communicotion system

~ couplingvIjICarrier ,' j equipment I'II arrangement ,i2 D -nding I switch forImaintenanceICarrier frequency ,IICommunicationnetwork (PLCC)1. LineTrap: 4. Surge Arrester:To block carrier frequencies to travel beyond the communicationnetwork and weaken the vital signals.To drain out the excessive voltages transferred to the communicationnetwork through the CVT, caused by lightning, switching or a2. Drain Coil: fault.It has low impedance at Power frequency and high impedance atcarrier frequency(a) To drain out power frequency component of the current whenpresent in the communication network(b) It also protects the CVT from overvoltages.3. Coupling capacitor or CVT is used when it is required for measurementand protection. For details on CVT refer to Section 15.4.1.3,5. The carrier equipment generally comprises the following:(a) Carrier wave instruments and devices,(b) Protection signalling equipment and devices, and(c) Telephone, Telex, Fax and Data transmission equipment.Figure 23.9(b) Typical layout of a line coupling device and carrier equipment (PLCC) showing phase-to-phase coupling

Power capacitors. behaviour, switching and improvement of power factor 231739line from the systcm or bq preventing tripping ot a healthy line through hlocking \ignal\ during a \\.\tern di\turhancr. henceretaining continuity of upp ply of the healthy lines on ir \y\tem disturbance as fw. a\ po\\ihle.Carrier relaying permih high-speed clearing 01. faults.For protection \\here the direction of the current 01- thc phase angle of the two end\ of the line are coiiipni-ctl to tlccide ~hctlieitherei\ ii fa~ilt on a particular line.The tran\mi\.;ion of data to the load dispatch centre\ is conducted at different frequencies. The ciii-rier ti-equenq I\ dccidcd hythe user and may fall in the range 30 500 kHi. A typical communication and relaying ichcmc i\ illu\trnted in Fipire 23.9tb). Formore detail\ on the PLCC. refer to the further reading at the end of the chapter.,A PLCC \hould ensure a good quality of speech and tranjmission of vital data at ii carrier frequency for cIc;ir and rin;rnibtfiiiou\communication between tu0 \tation\. It should also avoid falw signalling so thatHigli-frcqueiicy dt~tnrbonccs ai-e ii~ttraiisfei-1-cd to the carrier equipmcnt thrvugh the coupling uqutpnicnl.Di\turhances do not mutilate the tfirnsfer oi' vital data or send el-roneow \igiials that may lead to ni~rlfLtiicti~~titii~ or \(\ITlr I itiilrelay\ or prevent them from taking timely corrccti\e action.The coupling of ii PLCC with the main lines i\ carried out through line traps. and coupling capacitor\ or CL'T. 4 C\ 7' I\ u\cdwhen it IS also required lor measurement and protection. For detail\ on a CVT refer to Section 1.5.1.1.The PLCC ciin he connccted between a phn\e and the ground ofthe tran\niission line ihelf. the ground forming the return pnth. ll\eot only one conductor saves on cost and makes the a hole syvtem economical. But this arrangement i\ not reliable duu to dependenceon a single conductor which during a problem or fiult on this phase may render the whole communication sy\tcm inoperative. Goodpractice i\ therefore to connect the PLCC between any two phases.A typical nelaork is shown in Figure 23.O(b). U\e ofdiyitnl PLCCquiprnent can provide :I more reliable communication sy\tem and more speech and data communication ch;inncl\.Optical fibre cables (OFCs)Thc optical fibre conimunication system employ\ the latest technology in the held of coniniunication dnd data trnn\mi\sioii mlireplace$ the use of conventional copper cables and ovei-head lines to underground lihre optical COIIIIII~IIIIC~~~~OI~ \!,htcmi. Theprinciple of thi\ technique i\ that the transmission of light signals by internal reflection\ through transparent fitires 01 :lass. cm becarried over long distances uithout losing clarity or strength.The optical fibre i\ an extremely pure core of glass fibre. surrounded by ir cladding to contain the light beam\ within the core.Metallic shedding ofnon-magnetic material is aI\o provided over the gin\\ fibre to protect the cable from damage dut-ing trm\portiition;ind handling.The electrical vgnals (pulses) of the iiudio wave\ are converted into light pulses optical energy through trmsducm This nptic;ilenergy i\ then ainied At cine end ofthe thin arid transparent optical fihre. The thickness of the fihre 19 even less than ;I human hair. \:I>~up to 25 pm. The corc of tihre glass transforms the optical (light) energy into a jingle wabe ( I330 x IO" or 15.50 x I Of' kin wavelength)of intense light. Thi\ wa\e posscsscs much iiiioie capacity fur cat-rying information than coppcr cable\. At the recciviny cnd :I decoderis used to convet-t the optical energy back to the original audio wa~es. A hunch of optical fibre cables can transmit nitIlion\ of ;iudio\iynals at the sime time. each fibre being capable of transmitting as much information as a thick copper cable with ii number of core\.Thi\ technique has completely transformed traditional overhead communication systems. As there is no u\r of coppet- orclcctricity in the optical fihre the optical waves are not subject to any electrical interference a\ di\cussed aha\ c. Corning Glas\Work\. USA. were the fir\^ to produce \uch cable\ in 1970.D Influence of harmonics on other electrical andelectronic circuitsBy the hririiioiiic idttigrsI ,411 elcctrical equipirierit and devices connected onthe system are subject to higher dielectric stressesdiie to a higher effective voltage (equation (23.1 )).2 High harmonic voltages may give rise to pulsatingand transient lorqucs in a motor oI a generator, insquare proportion to the voltage (7;, oc v; ). At higheramplitudes of such harmonics, it is possible that thedriving or the driven shafts may even shear off as aconsequence of transient torques. Transient torques.up to 20 times the rated torque, have been experiencedwith oversiLed capacitor units connected on the system,causing wide voltage fluctuations.3 Additional noise from inductive equipment such asmoton. generators and transformers and even overheadlines.3 Flickering of GLS lumps in homes and offices.By the hcrrrizonic currentsI Overheating of the windings of an inductive load,such ;IS a transfornmer and rotating machines. cablesand overhead lines etc.. connected on the systcm dueto the higher effective current (equation (23.2)).2 The electromagnetic relays, which have a magneticcore and are matched to the fundamental frequency,may be sensitive to such disturbances and maymalfunction. sending erroneous signals.3 They may cause magnetic disturbance\ and noise incommunication networks in the bicinity as discuswdabo\ie.3 They may also affect the pertormancc of electronicequipment operating on the same system. such ax acomputer or a static power factor correction relay.5 Increased errors in all types of measuring instruments,which are calibrated at a fundamental frequcncy.6 Possible resonance and ferro-resonance effects (Section20.2.1(2)) between the reactances of the generatorand the transformer windings and the line capacitances.In addition to the line capacitors. it is possible thatthe circuit is completed through the ground couplingcapacitances (Section 20. I and Figure 20.2) and giverise to high to very high system voltages. The effectwill be the same. whe~her the syslem is grounded ornot. The ground coupling capacitances providing thereturn path, as through a grounded neutral. Thisphenomenon would normally appl) to an EHV ry\tt'in.

23/740 Industrial Power Engineering and Applications HandbookBy the harmonic fzux and frequenciesIncreased iron losses in the inductive machines, operatingon such a system, due to saturation effects (equations(1.12) and (1.13).23.6a Generation of triple harmonicsin an inductive circuitA hysteresis loss in an electromagnetic circuit will occurdue to molecular magnetic friction (magnetostriction),as discussed in Section 1.6.2(A-iv). This causes a distortionin voltage and current by distorting their natural sinusoidalwaveforms. This distortion in the natural waveforms, interms of magnetizing current I,, induced e.m.f. e andmagnetic flux $J of the magnetic circuit is the source ofgenerating triple harmonic quantities, the magnitude ofwhich will depend upon the shape of the hysteresis loop,which is a function of the core material. A fluctuation inthe system voltage which causes a change in the fluxdensity B (equation (1.5)) also adds to the triple harmonicquantities. Permeability ,u of the magnetic core is differentat different flux densities B. This results in giving rise totriple harmonics due to magnetic friction (magnetostriction).Wherever the switching of a capacitor bank ismore frequent, the generation of triple harmonic voltagesis more severe.Uneven distribution of loads causes an imbalance in thesupply system which also generates triple harmonicvoltages. This is more pronounced in LT systems than inHT systems due generally to fewer switching operationsand an almost balanced load distribution in HT systems(also refer to Table 23. I). But an HT system too is not freefrom third harmonics due to unequal overhead lineimpedances which may be caused by unequal spacingsbetween the horizontal and vertical formation of conductors,asymmetrical conductor spacings and hence an unequalinduced magnetic field. Overexcitation of transformercores, which may be a result of voltage fluctuations or aload rejection, may also cause triple harmonic voltages.When capacitors are used to improve the system'sregulation, they may cause excessive voltages duringoffpeak periods, resulting in an oversaturation of thetransformer cores and the generation of triplc harmonics.23.6b Generation of harmonics by apower electronic circuitAnother reason for harmonic disorders is the presence ofpower electronic circuits in the system. The loads suppliedthrough solid-state devices such as thyristors generatenon-sinusoidal waveforms. These waveforms give riseto harmonic currents and distort the supply voltage. Linecommutatedconverters and frequency convertersemployed for a.c./d.c. drives, electrolytic processes,inverters, a.c. controllers, d.c. choppers, cyclo-converters,and SCR rectifiers (for details refer to Chapter 6) areexamples of electronic applications that may generateharmonic quantities and influence the normal currentdrawn by a power capacitor. With advances in powerelectronics, modern power systems have a widerapplication of solid-state technology in the control ofinduction motors and d.c. motor drives as discussed inSection 6.9. There is therefore more likelihood of thepresence of harmonics in the system that may result inhigh to very high harmonic currents and consequentharmonic voltages. This effect is felt more on LT andalso HT systems up to 11 kV where use of static controlsis more prominent. For the harmonics generated by powerelectronic circuits, the harmonics ordinal numbers thatmay be present in the system are expressed byharmonic ordinal number Nh = nk f 1where n = pulse number; 3 for a three phase and 6 for ahexaphase power electronics circuit, etc.This represents the type of converter connection andidentifies the number of successive commutations (Le.the number of discrete segments of the d.c. waveform,Figure 23.10) the rectifier unit will conduct during eachcycle of a.c. power input. The higher the pulse number,Likely waveforms generated by a rectifier unit in its outputfor each cvcle of a.c. inout. for different numbers of DulsesParametersEffective value'Mean valueIForm factoror harmonicdistortion factor(Rectifier peak factor)1.57 I 1.11 11.017 I 1.0009 1 1.000053.14 11.5711.2111.0511.01* - rrns value of all harmonicsFigure 23.10The content of ripples for different number of pulses (n)

,05Power capacitors. behaviour, switching and improvement of power factor 23/741All such harmonics are undesirable for the system andthe equipment connected to it. Refer to Table 23.1 forthe likely magnitudes of harmonic quantities which maybe present in a typical power system.8-1Form factor = % = 1 .ll Rectifier factor = - = 1.5721a2ixNote When referring to the d c side, it is the mean value and whenreferring to the a c side. it is the effective value that IS more relevantFigure 23.11 Effective and mean values in a sinusoidal waveformthe nearer will the mean and effective values of therectified voltages approaching the peak value. as illustratedin Figure 23.10. Six pulse rectifiers are thereforeconsidered ideal 10 achieve a near-peak voltage in bothits mean and effective values and are more commonlyused. In which case: form factor or harmonic distortionfactorEftective value (r.m.\. value of all harmones= 1.0009Mean valuePeak value ~and rectifier peak factor. i.e.Mean valueas against the values indicated in Figure 23.1 I . Figure23. IO illustrates some common types of pulse numbers.the likely shape of their waveforms and the approximatevalucs of their ‘form factors’ and ‘rectifier peak factors’.k = 1st. 2nd. 3rd. and 4th etc. waveformsFor X. = 1 and 2, the harmonic sequence will be(i) for a three-phase thyristor circuit3 x k I . i.e. 2 and 4, and3 x 2 t I. i.e. 5 and 7. etc.(ii) for a hexaphase thyristor circuit6 x I f I = 5 and 7. and6 x 2 t 1 = 1 I and 13. etc.The plus sign indicates a positive sequence harmonicand the minus sign a negative sequence harmonic. Theireffect is same as for the positive and the negative sequencecomponents disciissed in Section 12.2(v) and causespulsation in the magnetic field and hence, in thc torqueof a rotating machine.23.6.1 Theory of circulation of triple harmonicquantities in an a.c. systemTriple harmonic quantities are unbalanced quantities andcan exist only on a system, which has a grounded neutral.The configuration of the windings of the source generatingthe harmonics. the system to which it is connected, andits grounding conditions, thus play a significant role intransmitting the harmonics to the whole system asdiscussed below:When the system is star connectedIsolated neutral In a symmetrical three-phase threewirestar-coiinected system the summation of all phasequantities, voltage or current will be zero, as illustratedin Figure 2 I .7. When a few harmonics exist in such asystem, this balance is disturbed and the sum of thesequantities is riot zero. The third harmonic quantitiescan find their outlet only through a neutral of thesystem. Since there is no neutral available in thissystem, no third harmonic quantities will flow into orout uf the supply system and affect the equipmentassociated with it or a communication network ifexisting in the vicinity (Figure 23.12(a)).Ungrounded neutral (floating neutral) However,when the neutral is provided. each phase can completeits own circuit through the neutral. provided theconnections are made so that it can complete its circuitthrough the neutral. The supply system would containno third harmonic, as it cannot drainout this througha floating neutral. The supply system. being balanced.would contain no harmonic as noted above. In eachphase to neutral, however, the third harmonic mayexist but of a lesser magnitude, deperidirig upon theimpedance of the system. (Figure 23. I2(b)).Grounded neutral When the neutral is grounded.the third harmonic quantities will be able to find theirway through it. This system can thus be considered tocontain the third harmonic quantities and will affectoperation oftheequipmentconnected onit and communicationnetworks if existing in the vicinity (Figure23.12(c)).When the system is delta connectedThis is a similar system to that discussed in 1 above andwill contain no third harmonic quantities (Figure 23.12(d)).23.6~ ResonanceYet another reason for harmonic disorder in a powercircuit is the occurrence of series or parallel resonanceor a combination of both. Such a situation may occur atcertain frequencies generated by the harmonic generatingsources in the system. The capacitors installed in thesy\tem magnify the same.23.7 Effective magnitude of harmonicvoltages and currentsA waveform containing harmonics may be considered tobe a standard sinusoidal waveform superimposed withthe other harmonic waveforms. Figure 23.7 illustrates astandard sinusoidal waveform, superposed with third and

23/742 Industrial Power Engineering and Applications HandbookWY/, = 0lhlA - RvhHarmonic circuit is not completed,:_ lt, = 0 v, = 0(a) Isolated neutral systemIhlHarmonics can circulate freely to other systems and from othersystems to this systemlh and Vh, = High(c) Grounded neutral system1h = Ihl + /h2 + /h3Harmonic circuit is made only internally. There is no flow of harmonics/h = 0Lto outside or from outside to the system, /h = 0:. I, = low v, = 0 v, = low (d) Delta connected system(lh and Vh are harmonic quantities)(b) Ungrounded neutral systemFigure 23.12 Circulation of triple harmonic quantitiesfifth harmonic waveforms. Higher harmonics are notshown being of lower magnitudes. Harmonics tend todiminish in amplitude at higher orders, Le. amplitude ath3 > h5 > h7 etc., where h3, h5 and h7 etc. are theharmonic quantities at the third, fifth and seventh harmonicorders respectively. The effective value under the influenceof such harmonics will be the summation of all suchharmonics present in the system. Each harmonic wavemay have a different phase relationship with thefundamental waveform, as illustrated in the figure, andadd or subtract from it to give the waveform a shapesomewhat of a pulsating nature as shown in Figure 23.7.Accurate measurement of harmonic quantities and theircumulative effect is a complex subject, and is not possibleto determine theoretically with the help of algebraicequations. But it can be easily measured with the help ofa harmonic analyser. Cigre (1989) has made a comprehensivestudy of the likely harmonics and theiramplitudes that may be present in a power system. Theseare briefly described in Table 23.1.InferericrThe study based on which the table is produced hasrevealed the following important aspects:1 Hexaphase electronic circuits generate maximumharmonic disorders.2 Inductive equipment generates moderate amounts ofharmonic disorders, except the third harmonics on anLT system.3 The odd harmonics (non-multiples of 3), generatedby hexaphase rectifiers, are more severe than any otherharmonic disorders present in the system, except LTsystems, where third harmonics too are pronounced.An LT system and the devices connected on it musttake cognisance of these effects and suitable measuresbe taken, either to diminish such effects or to selecthigher ratings of the devices to operate on such systemswithout overloading. HT systems must take particularcognizance of odd harmonics.

~~~~~~~Table 23.1 Typical harmonic orders and their magnitudes as studied in a power systemOdd harmonic-~Likeh harmonit voltage (%JS\ \tern voltageSyttern voltage~Harmonicordern~~- ~up toI kVAbove 1-33 kVnetworksand primarydistribution.33-100kVAbove 33-100 kVnetworks andsecondarytransmissionabove 100-220 kV-Hdrmonicordernup toI kVAbove 1-33 kVnetworks,and primarydistribution,33-100kVAbove 33-100 kVnetwork\ dnd\econdarytran\mi\$ionabove 100-220 kVHarmonicordernUp toI kVAbove I-33 kVnetworksandprimarydistnbution,33-100kVAbove 33-100 kVnetworksandsecondarytransmissionabove100-220 kV57II13174 to 64 to 52.5 to 3.52 to 3I to 25 to 64 to 52.5 to 3.52 to 31 to 21 to 2I to 20.8 to0.8 to0.5 to.5.53915214 to 50.8 to5 0.35 0.21.5 to 2.5.5 0.8 to 1.55 0.35 0.20.8 to 1.50.5 to I5 0.35 0.224681 to 20.5 to 15 0.55 0.51 to 1.50.5 to 10.2 to 0.550.21 to 1.50.5 to I0.2 to 0.55 0.2Source lEEE 51 9, ,for general and dedicuted power systemsNotes1 Still higher harmonics have not been conqidered, as they are of even lower magnitudes.2 The values are expressed as a percentage of the fundamental voltage.3 The above data is for a general reference. Actual harmonic distortion on the sybtem during operation will depend upon the types of connected loads and their operating conditions, suchas variation in its loading and fluctuations in the system voltage.

~~~ ~-23/744 Industrial Power Engineering and Applications Handbook23.7.1 Recommended magnitudes of harmonicdisordersFor safe operation of power and control equipment anddevices operating in such systems it is essential to limitthe amplitude of the voltage distortions to a safe valueby installing filter circuits based on the system's actualoperating conditions. These limits are recommended byleading standards organizations are:1 UK engineering recommendation - G5/3 (BritishElectricity Council Standard). For harmonic voltagedistortions in the system as in Table 23.2.2 IEEE-519: Guide for harmonic control and reactivecompensation of static power converters, for harmonicvoltage distortions of general and dedicated powersystem, as in Table 23.3.Table 23.2 Premissible individual and total harmonic voltagedistortions - as in G5/3, UK engineering recommendationsNominal system Individual harmonic voltage 7imd harmonicvoltage distortions distortions~~kV Odd( 70) Even (lo) %0.415 4 2 56.6-1 1 3 1.75 433-66 2 I 3132 1 0.5 I .sFigure 23.13 Rise in apparent current due to harmonicquantitiesI, = apparent or current measured by an ammeterIh = actual current due to harmonic distortionsAll the above are r.m.s. quantities.@ = displacement angle between the system voltageand apparent current, defining the p.f. of theload6 = actual phase displacement due to harmonicdistortions. It is the actual p.f. which is less thanmeasured for a system containing harmonicdisorders.Quantity Ih is composed of two components:One at fundamental frequency. Its displacement withthe fundamental voltage is termed the displacementfactor, and for a linear voltage and linear load willdefine the p.f. of the load, i.e.Table 23.3 Premissible voltage distortions as in IEEE-519 forgeneral power systems and dedicated systemsaNominal system General power Dediccrredvoltage systems rystemskV lo 962.449 5 81 15 and above 1 .s 1 .sA dedicated system is one that is servicing only the converters orloads that are not influenced by voltage distortions.23.8 When harmonics will appear ina systemThe second component is caused by the differentharmonic quantities present in the system when thesupply voltage is non-linear or the load is nonlinearor both. This adds to the fundamental current,I, and raises it to Ih. Since the active powercomponent I, remains the same, it reduces thep.f. of the system and raises the line losses. Thefactor IJIh is termed the distortion factor. In otherwords, it defines the purity of the sinusoidal waveshape.:. cos 6 = displacement factor x distortion factorThere are three possible ways in which harmonics mayappear in a power system:1 When the system voltage is linear (an ideal conditionthat would seldom exist) but the load is non-linear:The current will be distorted and become nonsinusoidal.The actual current I,, (r.m.s.) (equation(23.2)) will become higher than could be measuredby an ammeter or any other measuring instrument, atthe fundamental frequency. Figure 23.13 illustratesthe difference between the apparent current, measuredby an instrument, and the actual current, where1, = active component of the currentFor example, if the apparent p.f. is 0.9, a distortionfactor of 0.85 will reduce it to 0.9 x 0.85 or 0.765.2 When the supply system itself contains harmonicsand the voltage is already distorted: now even thelinear loads will respond to such voltage harmonicsand draw harmonic currents against each harmonicpresent in the system, and generate the same order ofcurrent harmonics.3 When the system voltage and the load are both nonlinear:a condition which is common. The voltageharmonics will magnify and additional harmonics maygenerate, corresponding to the non-linearity of the

~ ReducingPower capacitors: behaviour, switching and improvement of power factor 23/745load, and hence further distort an already distortedvoltage waveform.Harmonics will mean- Higher voltage and current than apparent- Adding to line loading and losses, andthe actual load p.f.Harmonic effects are reduced by the use of shuntcapacitors installed to improve the p.f. of the systemor an individual load, as they provide a very lowimpedance path to harmonic quantities and absorbthem. But using shunt capacitors alone is not sufficientas at certain harmonic frequencies, the capacitivereactance of these capacitors, along with leakagecapacitance C,, of the line, if it is an HV or EHV line(Table 24.l(b)) may tune up with the line inductivereactance L and cause a series or parallel resonanceand develop high voltages. This phenomenon will bemore apparent on an HV and EHV system than anLT system. On LT systems, the line resistance itselfis enough to provide an in-built dampening effect toa possible resonance (z = R,) at least). To avoidresonance, filter-circuits are used. Filter-circuits maybc tuned for the lowest harmonic content or for eachharmonic separately as discussed later, when thesystem is required to be as near to sinusoidal waveformas possible. The system may contain second, thirdand fifth and higher harmonics, but it may be sufficientto tune it to just below the fifth harmonic, which willbe able to suppress most of the other harmonics alsoto a great extent. as it will make the circuit inductivefor all harmonics.In an HT system, either the star is not grounded or itis a delta-connected system and hence the third harmonicis mostly absent, while the content of the second harmonicmay be too small to be of any significance. For thispurpose, where harmonic analysis is not possible, or fora new installation where the content of harmonics is notknown, it is common practice to use a series reactor of6% of the reactive value of the capacitors installed. Thiswill suppress most of the harmonics by making the circuitinductive, up to almost the fourth harmonic, as derivedsubsequently. Where, however, second harmonics aresignificant, the circuit may be tuned for just below thesecond harmonic. To arrive at a more accurate choice offilters, it is better to conduct a harmonic analysis of thesystem through a harmonic analyser and ascertain theactual harmonic quantities and their magnitudes presentin the system, and provide a correct series or parallelfilter-circuits for each harmonic.As noted from general experience, except for specificlarge inductive loads such as of furnace or rectifiers, thefundamental content of the load current is high comparedto the individual harmonic contents. In all such cases, itis not necessary to provide a filter-circuit for each harmonicunless the current is required to be as close to a sinusoidalwaveform as possible, to cater to certain critical loads orinstruments and devices or protective schemes operatingin the system, where a small amount of harmonics maylead to malfunctioning of such loads and devices.Otherwise only the p.f. needs be improved to the desiredlevel. Also to eliminate a parallel resonance with thesource reactance, an inductor equivalent to 6% value ofthe capacitive reactance can be provided in series withthe shunt capacitors.If there are many large or small consumers that adistribution line is feeding, it is possible that the voltageof the network may be distorted beyond acceptable limits.In this case it is advisable to suppress these harmonicsfrom the system before they damage the loads connectedin the system. Preferable locations where the seriesinductor or the filter-circuits can be installed are:1 At the supply end and/or2 At the receiving end of the consumer. using power atHT and generating high harmonics (e.g. rolling mills,induction and arc furnaces, electrolysis plants,chemicals and paper mills). Electricity companies willalways prefer the consumers to suppress the harmonicsat their end. Even if this practice is adopted by some.small harmonics generated by many small consumers,who may not take any preventive steps to contain theharmonics generated by their plant, may still exist inthe system. As a result, the distribution network maycontain some harmonics which affect the performanceof other loads in the system. If the level of harmonicsis unacceptably high (Tables 23.2 and 23.3) they mustbe suppressed at the supply end.23.9 Filter circuits: suppressingharmonics in a power networkThe use of a reactor in series with the capacitors willreduce the harmonic effects in a power network, as wellas their effect on other circuits in the vicinity, such as atelecommunication network (see also Section 23.1 I andExample 23.4). The choice of reactance should be suchthat it will provide the required detuning by resonatingbelow the required harmonic, to provide a least impedancepath for that harmonic and filter it out from the circuit.The basic idea of a filter circuit is to make it respond tothe current of one frequency and reject all other frequencycomponents. At power frequency, the circuit should actas a capacitive load and improve the p.f. of the system.For the fifth harmonic, for instance, it should resonatebelow 5 x 50 Hz for a 50 Hz system, say at around 200-220 Hz, to avoid excessive charging voltages which maylead to0 Overvoltage during light loads0 Overvoltage may saturate transformer cores andgenerate harmonicsFailure of capacitor units and inductive loads connectedin the system.It should be ensured that under no condition of systemdisturbance would the filter circuit become capacitivewhen it approaches near resonance. To achieve this, thefilter circuits may be tuned to a little less than the definedharmonic frequency. Doing so will make the Land henceX, always higher than Xc, since

23/746 Industrial Power Engineering and Applications HandbookHT, 132 kV (typical)This provision will also account for any diminishingvariation in C, as may be caused by ambient temperature,production tolerances or failure of a few capacitor elementsor even of a few units during operation.The p.f. correction system would thus become inductivefor most of the current harmonics produced by powerelectronic circuits and would not magnify the harmoniceffects or cause disturbance to a communication systemif existing in the vicinity (see Figure 23.14). A filtercircuit can be tuned to the lowest (say the fifth) harmonicproduced by an electronic circuit. This is because LTcapacitors are normally connected in delta and hence donot allow the third harmonic to enter the circuit while theHT capacitors are connected in star, but their neutral isleft floating and hence it does not allow the third harmonicto enter the circuit. In non-linear or unbalanced loads,however, the third harmonic may still exist. For a closercompensation, uni-frequency filters can be used tocompensate individual harmonic contents by tuning thecircuit to different harmonics, as illustrated in Figures23.15(a) and@). For moreexact compensation, thecontentsand amplitudes of the harmonic quantities present in thesystem can be measured with the help of an oscilloscopeor a harmonic analyser before deciding on the mostappropriate filter circuit/circuits. Theoretically, a filter isrequired for each harmonic, but in practice, filters adjustedfor one or two lower frequencies are adequate to suppressall higher harmonics to a large extent and save on cost.Refer to Table 23.1, which shows the averagecumulative effect of all the harmonics that may be presentin a power system. If we can provide a series reactor of6% of the total kVAr of the capacitor banks connectedon the system, most of the harmonics present in the systemcan be suppressed. With this reactance, the system wouldbe tuned to below the fifth harmonic (at 204 Hz) for a 50Hz system as derived below.132/11 kVFilter circuits to compensate the harmonicsR = Resistance to limit the fault levelFigure 23.1 5(a) Compensation of harmonics bytuning each capacitor circuit at different harmonicstX, (Filter reactance)-R,!:,HT,132r (typical)Q132/11 kV T~,(typical)t X, (System reactance)If X, = X, then Z= RFigure 23.15(b)If X, is the 6% series reactance of the value of theshunt capacitors installed, thenXL = 0.06 X c& .Ic1 fc2G-L, and L, are reactorsFigure 23.14 Compensation of harmonic currents by the useof reactors in the capacitor circuits0.06or 2nfL = -2 xfcwhereL = inductance of the series reactor in henryandC = capacitance of the shunt capacitors in farad0.06or LC= -(2nf)’

Power capacitors: behaviour. switching and improvement of power factor 23/747When such a reactance is used to suppress the systemharmonics it will resonate at a frequency fh such that- '0for a SO HZ system0.06= 204 Hz. i.e. at the fourth harmonic.But when the third and/or second harmonics are alsopresent in the system, at a certain fault level it is possiblethat there may occur a parallel resonance between thecapacitor circuit and the inductance of the system (source),resulting in very heavy third or second harmonic resonantcurrents. which may cause failure of the series reactor aswell as the capacitors. In such cases, a 6% reactor willnot be relevant and a harmonic analysis will be mandatoryto provide more exacting filter circuits.It is pertinent to note that since a filter circuit willprovide a low impedance path to a few harmonic currentsin the circuit (in the vicinity of the harmonic, to which ithas been tuned) it may also attract harmonic currentsfrom neighbouring circuits which would otherwisecirculate in those circuits. This may necessitate a slightlyoversized filter circuit. This aspect must be borne in mindwhen designing a filter circuit for a larger distributionnetwork having more than one load centre.It is, however, advisable to conduct a harmonic studyof the system to select a more appropriate size of reactor,particularly where the installation is expected to experiencehigh harmonic disorders.IJse of a reactor will enhance the voltage across thecapacitor banks and must be considered in the design ofthe capacitor units. Refer to Figures 23.16(a) and (b)illustrating this. IfRI1.ri'WXL i(a) Power diagramVC > voh(b) Phasor diagramFigure 23.16 Voltage across the capacitor units rises with theuse of series reactorswherethenVph = system voltageVc = voltage across the capacitor banks = I C. XcVL = voltage across the series reactor = I,.X, = capacitive reactanceXL = inductive reactance. andI, = current through the capacitor circuitXI-Vph = Vc - Vl~ (phasor diagram, Figure 23.16(b))= ICXC - ICX,and the rise in voltage across the capacitor banks. as aratio of the system voltagek-- IC&- XCVp, I,X, -1, XL x, -x,For a series reactor of 6%. for instance.2- xcVph X, - .06X,= 1.0638(73.8)That is, the voltage across the capacitor banks will increaseby 6.38%. This voltage must be considered in the designof capacitor units.Similarly, the third harmonic may also be suppressedby grounding the generator or the transformer neutralthrough a suitable impedance (LC circuit), as discussedin Section 23.5.2(c) and equation (23.6).23.9.1 Compensating for the series reactorWhen a capacitor circuit is compensated through a seriesreactor, either to suppress the system harmonics or tolimit the switching inrush currents (Section 73.1 I ) orboth, it will require suitable adjustment in its voltageand capacitive ratings. The series reactor will dampenthe switching currents but consume an inductively reactivepower and offset an equivalent amount of capacitive kVAr.and require compensation. The following example willelucidate this.Example 23.1To determine the basic parameters of a 6% series reactorand its capacitive compensation, consider Example 23.6 with3000 kVAr banks (1000 kVAr per phase) rated for 33.4 kV:Then the voltage rating of the capacitor units should bechosen forsay, = ~= 1.0638 x Y 33.4>/ 335.5 kV instead of>b-33or =34kV73 \3from equation (23.8)(see Figure 23.1 7)Since the rating of the series reactor is = 0.06 x 3000 =180 kVAr.The capacitors' rating must be enhanced by the rating ofthe reactor to obtain the same level of effective kVAr.

23/748 Industrial Power Engineering and Applications Handbook33.4 kVI60 kVAv~=-xxooov 35.5d3Figure 23.17 Phasor diagram for Example 23.1:. The rating of the capacitor banks should be chosen for= 3000 + 180= 31 80 kVAr.For the basic parameters of the reactor on a per phasebasisI,= ~100035.5/&= 48.8 Aand rating of reactor, VI . IC = 60 kVAr. vl = 60 x 1000 v..48.8= 1229.5 V1229.5and X,= -48.8= 25.2 RIf this is designed for, say, 200 Hz, then its inductanceL=25.22 x ZX 200= 20 mHRating of the secondary distribution transformer = 30 MVA,3310.4 kV,zp = 10% (after applying the negative tolerance)Load currents at different harmonics are recorded as follows:Harmonic order1 (fundamental)35791113Load current (A)400a5027.5a105aThe third harmonic was almost absent. hence It was not considered.The second and other even harmonics were also insignificant.Similarly, higher harmonics (above the thirteenth) were alsoinsignificant. hence were not considered for ease of illustration:. Actual loading of the system./h = [400’ + 25 x 50‘ + 49 x 27.5’ + 121 x 1 o2 + 169 x 5’1”’(23.2)= 525.2, say, 525 A.The ampere meters, however, indicated only 400 A. While itappeared that the line was not fully loaded, in fact, it wasloaded to its rated capacity of 525 A and was utilized by only352 .400 x 0.88, Le. 352 A or - 1.e. by 67%, considering the525system p.f. at 0.88.Let us assume the following line parameters:Phase displacement at fundamental frequency, Le. apparentp.f. as measured by a normal p.f. meter.cos $, = 0.88 lagging (Figure 23.18(a))... $1 = 28.36”.23.9.2 Studying and nullifying the effects ofharmonicsExample 23.2A distribution network 33 kV, three-phase 50 Hz feeding anindustrial belt with a number of medium-sized factories somewith non-linear loads and some with static drives and somewith both. It was observed that while the lines were apparentlyrunning reasonably loaded, the active power supplied wasmuch below the capacity of the network. Accordingly, aharmonic study of the network was conducted and it wasfound that despite localized p.f. control by most factories, thep.f. of the network itself was well below the optimum level andthe voltage was also distorted by more than was permissible.To improve this network to an acceptable level, we haveconsidered the following load conditions, as were revealedthrough the analysis.Rating of primary distribution transformer = 30 MVA,132133 kVzp = 10% (after applying the negative tolerance)I, = 525 A (Figure 23.18(a))(2) Actual loading(3) Improved loading(4) Extra loading capacitygenerated by improvingthe p.f.(5) Improved actual loadingthe networkFigure 23.18(a) Enhanced capacity utilization of the networkwith the improved p.f.

xPower capacitors: behaviour, switching and improvement of power factor 2317494132133 kVR = 1.95Q- suppress 5th harmonicquantitiesI L I I I 117352Actual, p.f., cos $2 = -525= 0.67.. $2 = 47.9"Let us improve the apparent load p.f. to 0.98i.e. cos q3 = 0.98and$3 = 11.48'From Example 24.3 and Figure 24.25(b) inductive reactanceof line = 6 Q per phase.Reactance of each transformer = 3.63 (2.Total inductive reactance of the network = 13.26 (1 per phase:. line inductance L13.26= ~27c f= 42.23 mH/phaseSince the p.f. is to be improved from 0.88 to 0.98:. Shunt reactive compensation required1111111 h,TTTTTTTlIxc = 400 [sin 9, - sin $3]= 400 [sin 28.36 - sin 11.481= 400 [0.475- 0.1991= 110.4 Aand kVAr required = & x 33 x 11 0.4Arrangement of capacitors in double star, 11 20 kVAr in each arm.Figure 23.18(b) Network of Figure 24.25 shown with shuntcompensation= 6310 kVArSay, 6300 kVArThe shunt capacitors can be provided on the LT or the HTside, whichever is more convenient. In the above case, sinceit is large, the HT side of the load-end transformer will bemore convenient. The receiving-end transformer, however,will now be operating under more stringent conditions thatmust be taken in account or the capacitors may be providedon the LT side to relieve this transformer also from excessivecurrents. In Figure 23.18(b) we have considered them on theLT side.Improved actual p.f., cos =0.98~ 0.67 = 0.750.884; = 41.4"and improved actual current--- 0.67 x 5250.75= 469 A (Figure 23.18(a))Figure 23.18(c) Reducing the actual loading of line by tuningfor the fifth harmonicand improved apparent line current = 400 x 0.88 = 359 A0.98and X, =kV' x 1000kVAr- 33' x 10006300= 172.86 R106:. c=KF27c x 50 x 172.86= 18.42 pF

23/750 Industrial Power Engineering and Applications HandbookWe have ignored any leakage capacitance between line toline or line to ground at 33 kV.Tackling the harmonicsTo eliminate all the harmonic contents from the supply it isessential to provide shunt-filter circuits separately for eachharmonic disorder. As this is a costly arrangement it is notfollowed in practice. It is enough to drain the highest (inmagnitude) harmonic disorder. This will substantially improvethe current and voltage waveforms and make the circuitinductive for all higher harmonics, preventing a condition ofresonance at any harmonic disorder during a line disturbance.The line reactance and capacitance of the capacitors mayresonate at a frequency1rh = -2 n m= 1. 12n 442.23 x x 18.42 x lo6= 180.5 HZ(23.6)say, at about the third or fourth harmonics. But both theseharmonics are considered negligible in the system hence,the filter circuit will be essential for the fifth harmonic. Theseries reactor required in the capacitor circuit is tuned, say,at about 225 Hz to be on the safe side, to account for variationin the line parameters and the inductor itself to keep thecircuit inductive under all conditions of line disturbances::. 27~ x 225 x L' =oror12rrx 225 x CL'= 1(2n)' x (225)' x 18.42 x lo4= 27.2 mHXL' = 2n x fx 27.2 x 10-3= 8.54 Q at fundamental frequency.This is about 4.94% of Xc, and must be compensated byproviding additional capacitor units to maintain the requiredlevel of p.f. as discussed in Section 23.9.1.:. Modified size of capacitor banks = 6300 x 1.0494= 661 1.2 kVArLet us choose a bank size of 6720 kVAr for easy selection ofeach unit. To adopt a better protective scheme, let us selectthe configuration of a double star (Figure 25.5(b)).:. Each bank for each arm of a double star- 6720 -~ 6= 1120 kVArLet us arrange them two in series and seven in parallel,based on recommendations made in Section 25.5.1 (iii) andas shown in Figure 23.18(b).:. Size of each unit = 1120 = 80 kVAr2x736*and voltage rating = -45x2= 10.39 kV*NoteThe capacitor units are generally designed for the highestsystem voltage, as shown in Tables 26.4 or 13.1 unlessspecified for still higher voltages by some users. Accordinglyequation (23.8) does not apply in this case. Some localelectricity distribution authorities, depending upon the likelymaximum voltage variation on their systems, may sometimesask for still higher voltages. But such distribution networksmay not remain stable in the long run and must be improvedas far as possible with the use of reactive controls, as discussedlater.The above configuration or size of each unit is not mandatoryand can be altered, depending upon the economics of capacitorvoltage and size of units available. Protection for all types ofconfigurations is easily available through various schemesdiscussed in Section 26.1.From the above it can be inferred that for an accurateanalysis of a system, particularly where the loads are ofvarying nature or have non-linear characteristics it isnecessary to conduct a harmonic analysis. The abovecorrective measures will provide a reasonably stablenetwork, operat-ing at high p.f. with the harmonics greatlysuppressed. The improved actual line loading, eliminatingthe fifth harmonic component, which is compensated,II; = [4002 + 49 x 27.5' + 121 x 10' + 169 x 52]5= 461.9A as against 525 A352and improved actual p.f. = -461.9= 0.76and 44 = 40.5"A phasor diagram (Figure 23.18(c)) illustrates thereduction in the actual loading and enhanced load transfercapacity of the network which can be achieved withthe help of harmonic suppressions. For even betterutilization, the system may be tuned for higher harmonicdisorders also.23.1 0 Excessive charging currents(switching inrush or makingc u r re n t s)The circuit of an uncharged capacitor unit, when switched,acts like a short-circuit due to the very high natural(transient) frequency& of the switching circuit, causingthe capacitor reactance X,, to approach zero. The valueof the making current will depend upon the magnitude ofthe applied voltage and the impedance of the circuit (motorand cables) or of the system that will form the switchingcircuit. In other words, it will depend upon the fault levelof the circuit and the rating of the capacitor units connectedin that circuit. This switching transient current, as illustratedin Figure 23.19, as such being high, will be even higherif the circuit already has some switched capacitor unitsdue to partial discharge of the already charged units intothe uncharged units, as discussed in Section 23.10.2. Thebehaviour of the capacitor switching circuit under variousswitching conditions is described below.

4 InFigure 23.19, Charging currentTime -Current waveform on switching of a capacitor23.10.1 Single-capacitor switchingInrush currentThe capacitor inrush current is a function of steady stateIiid the transient components of the current, i.e.I,,, = 1, + I,,where I,,, = inaxinium inrush or making current whenswitching an uncharged capacitor unitI, = maximum steady-state current (capacitorpeak full-load current ( :2Ic j)I,, = transient component of the current, which isa function of the short-circuit power of theswitching circuit, i.e. the kVA of the transformer,if a transformer is feeding the circuit,and the kVAr of the capacitor being switched.; Short-circuit kVAi.e. I,, = I,1' Capacitor kVArPower capacitors: behaviour, switching and improvement of power factor 23/751andIV:I1-Figure 23.20 A single capacitor switchingTZ,, = short-circuit impedance 01 the .;witchingcircuit andZ = steady-state impedance of the switchingcircuit.If R. Ll (of cable, transformer and load inductances)and C, are the switching circuit parameters then-z = R + x,, - x,,During a steady-state condition, i.e. at power frequency,f Xcl >> XLl and during a transient condition, i.e. duringa switching operation whcnf

23/752 Industrial Power Engineering and Applications Handbookwheref, = switching or transient frequencyf = power frequencyX, = power frequency capacitive reactance of the circuit,andX, = power frequency inductive reactance of the circuit.Switching overvoltageThis is seen to rise up to 1.8-2 p.u. of the peak appliedvoltage, Vp, when switching an uncharged capacitor andaround 2.6-3 p.u. when switching an already chargedcapacitor (Section 23.5.1).Rating of the switching deviceThis is as recommended in Section 25.6 and is equallyapplicable to LT and HT capacitor units.Example 23.3Consider the same LT system of 41 5 V, as of Example 23.8,three-phase 50 Hz, and having a fault level of 36 MVA andemploying capacitor banks of 360 kVAr.Inrush currentCapacitor full-load currentI,360=& x 0.415= 500 Aand static current I, = & x 500 = 707 AandI,,, = 707[ 1 + IF)= 707 (1 + 10)= 7777 Aor eleven times the capacitor peak rated current. This iswhen totally uncharged capacitors are switched. If thecapacitors are already charged when they are switched, theinrush current may rise up to twice this, Le. almost twentytwotimes the rated peak current (but not exceeding the faultlevel of the system).NoteHere we have considered the fault level of the switchingcircuit to be same as that of the entire system, as they arethe total capacitor banks connected in the system and switchedtogether as a single unit.close proximity with another set of banks of almost similarcapacity such as a capacitor control panel, having a setof capacitor banks arranged to switch ON and OFF,depending upon load variations of that circuit. Capacitorunits installed on the same system but not in closeproximity may not influence the switching behaviour ofthese capacitor units as severely as when they are installcdin close proximity due to the extra line impedance thusintroduced.Inrush currentIn this switching, the cumulative effect of the appliedvoltage and the trapped charge of the already chargedcapacitors is of little relevance but the resultant currentis extremely high as derived below.Consider a typical circuit as shown in Figure 23.21:In=- VP (when the applied voltage was at its peak andZ, the switching capacitor had no trapped charge)where I, = peak inrush current in parallel switchingZ, = surge impedancewhere- I L2dC"q(equation (17.2))L2 = circuit inductance between the banks inhenry (ignoring the negligible selfinductanceof the capacitor banks)Ceq = equivalent series capacitance of the twocapacitors C, and C, in farad(23.12)(23.13)This is a generalized formula to simplify the equation.In fact, the actual switching current will be much lessthan this, due to circuit impedance. For more accurateand I,, = Vp JLzzSwitching frequency= 500 HZ23.10.2 Parallel switching of capacitor unitsThis shows yet more complex circuit behaviour andrequires more detailed analysis as discussed below.We have considered the switching of capacitor banksthat are installed on one power circuit and mounted inParallel operation of capacitor bank C,with an already switched capacitor C,Figure 23.21 A simplified capacitor switching circuitI

Power capacitors: behaviour, switching and improvement of power factor 23/753result\. the inductance of the capacitor banks may beintroduced into the total inductance, L2, Le., it should bethe summation of all the reactances of the capacitorsalready switched. plus the series reactance of the capacitorbeing switched. See Figure 23.22.:. Equivalent reactance of the capacitors already switched1L,,,, =-1 +1 1 1-~ + - + _..-L, L, L; L,,Iand L7uq = 1 +k+ I + +... -~L, L, L; LnL, = series inductance of the capacitor being switched(this is L, of equation (23.13))and ('I,,, = C1 + C2 + C? + ... C,(this is C, of equation (23.13))The figures of Lzcq and C,,, may be substituted inequation (23.13) for L2 and C, respectively to derive amore accurate switching inrush current.Nore, YI In the abo\e discussions it is assumed that the capacitor C'? hasno trapped charge when it i\ switched ON. If this is not so. thecurrent may rise up to 2 . I,,,. as discussed in Section 23.10.1.2 Field experiment\ have reLealed that such currents may be ashifh as up to 15-250 times the 5teady st;lte current I,. but willlast onl? up to the first current zero.Switching frequencyAs discussed above, a capacitor circuit is an L-C circuit,and the switching frequency ,f7 can be expressed aswhere(equation (17.1))1CC(, = (as in Figure 23.22)-1 -+ I-CI,, C,C, = capacitance of the capacitor being switched*L, being L2 and C, (now being switched) is C, of Figure 23.21Figure 23.22 Equivalent circuit for R number of capacitorbank already switched on the circuit and another (C,) beingswitched(23.13)The parameters CI,, and bcq are the same as those derivedabove.The switching frequency in capacitor switching is veryhigh. We have already witnessed this in a single capacitorunit. The situation becomes highly complicated whenswitching is effected in a circuit that already has a fewswitched capacitor units. In Example 23.4 we will seethat a circuit with only 6 out of 60 kVAr capacitor unitscan have a switching frequency as large a\ over 13 kHz(in operation, it may not exceed 5-7 kHr because of thecircuit's actual impedance), when five of these capacitorsare already switched and the sixth is switched. This highto extremely high switching transient frequency isdetrimental in giving rise to the switching inrush currents.of the order of 15 to 250 times and more, of the steadystatecurrent I,. Since the switching capacitive reactanceis inversely proportional to the switching frequency, itoffers an almost short-circuit condition during a switchingoperation.SummaryAt the instant of switching the surge impedance, Z,(equation (23.12)), and the natural frequency of theswitched circuit, i.e. the transient frequency,j; (equation(23.14)), determines the amount of inrush current.The natural (surge) frequency,,f,. of parallel capacitorswitching is extremely high, of the order of 5-7 kHror more. It may not exceed thi.: hecause of the circuit'sown parameters R and L that have been ignored inour analysis for easy illustration. The actual frequency,f,, will depend upon the size of capacitors beingswitched compared to the capacitors that were alreadyswitched, and their corresponding inductance in theswitching circuit. A highf, will diminish the capacitivereactance of the capacitor to an almost negligible value.This leads to a near-short-circuit condition during theswitching operation, causing extremely high transientinrush currents of the order of 15-250 times and evenmore of the capacitor's steady-state current I,.This extremely high inrush current at a frequency ofalmost 5-7 kHz will release an enormous amount oflet-through energy during contact making,i.e. = I,; . tIsay (250. I, )' . ~2 x 50(for an I, of 250. I, occurring at a transient frequencyof 6 kHz and existing up to the Pirst current Lero in a50 Hz system).The interrupting device, which may be a breaker ora contactor, must be suitable to sustain this energywithout deterioration of or damage to its contacts,while the fuses must stay intact when provided forbackup protection.

823/754 Industrial Power Engineering and Applications HandbookNotes1 For air break LT switches and contactors IEC 60947-3 and IEC60947-4-1 have specified the making capacities for heavy-dutyAC-3 components as follows (also refer Lo Section 12.10):(i) For rated current up to 100 A - 10 times the r.m.s. value(ii) For rated currents beyond 100 A ~ times the r.m.s valueThe manufacturers of such devices, however, may declarethe making capacity of such devices to be even higherthan this when the higher value may be considered todesign the inductance, to limit the inrush current (Section23.1 1).(iii) Light-duty switches and contiactors up to AC-2 duty arenot suitable for such applications.2 NEMA publication ICS.2-210 for general-purpose contactorsprovides data for ratings prevalent in the USA. These data aresummarized as follows:(i)Table 23.4 gives continuous current ratings and maximummaking currents, I,, for various sizes of contactors.(ii) Choice of contactors for different values of capacitorswitching currents is provided in Table 23.5.4 The transient condition will cease to exist after thefirst current zero, hence the extremely short durationof such transient currents.5 Such heavy currents are more prevalent in an HTSize of Continuous current ratingcontactor Amps (rm s )8" 1 18Eri~lusedOpen9 10201 21 302 45 503 90 1004 135 1505 270 3006 540 6007 810 9008 1215 13509 1 2250 2500Maximum makingcurrent, I,Amps (peaky)87140288483941158171636326947014,20525,380than in an LT system. An HT system will have amuch smaller impedance, 2, compared to the appliedvoltage, V,, than an LT system, which will have amuch higher value of Zcompared to the applied voltage"P.Thus, the analysis conducted above to derive I, on aparallel switching is more pertinent for an HT system,and offers only a theoretical analysis for LT systems.Nevertheless, for large LT installations such as, 2000-2500 kVA for a single circuit, as discussed in Section13.4.1(5), employing large capacitor banks close to thefeeding transformer, it may be desirable to limit the inrushcurrent to a desirable level on the LT side also. In allprobability, switching devices selected at 150% of thccapacitors' normal current as in Section 25.6 will besufficient to meet any switching contingency, even on aparallel switching.23.1 1 Limiting the inrush currentsLet ub refer to equation (23.13) for the inrush current,I, on parallel switching. If we can increase the impedanceof the switching circuit by introducing a resistance R orinductance L, or both, we can easily control this currentto a desired level.NoreIf we add L or R in the circuit it will increase the surge impedance,reduce the surge frequency of the switching circuit, and dampenthe 90' lead surge current to a moderately leading surge current,small in magnitude and steepness. At every current zero when thecontacts of the switching device tend to separate, the recoveryvoltage will be much less than its peak value due to smaller phasedisplacement between the current I,,, and the TRV, and thus help tointerrupt the switching device with a lesser severity.For more details see Section 17.7.2(iii). The principleof restrike in capacitive switching is almost the same asin the case of inductive switching. Consider Figure 23.23,where for ease of analysis, an additional inductance, L,is introduced to increase the circuit impedance, whichTable 23.5 Contactor rating for capacitor switching at different transient (switching) currents I,,,Size ofcontactorContinuous ratingof contactor.yAmps (rm.s.)459013527054081012152250Maximum size of three-phase capacitors in kVAr at different switching currents I,Capacitor switching curreiit I,3000 5000 IO 000 14 000 18 000 22 000A A A A A A25 16 8 6 4 453 53 31 23 18 1580 80 80 61 49 41160 160 160 160 160 149320 320 320 320 320 320480 480 480 480 480 480720 720 720 720 720 7201325 I325 1325 1325 1325 1325~.

,#3Power capacitors: behaviour, switching and improvement of power factor 231755T-IILl1V c1IIITFigure 23.23 Inductance L introduced into the switchingcircuitc2 Twill also dampen the resonant frequency of the switchingcircuit. The improved I,',, of equation (23.13) can nowbe expressed as:(23.15)where L7 and Ci can further be substituted by LIeq andC,,,, respectively.1and Xc = ~2n. f- c;_ X' -'x-- 1 - 1- 3 2n.f.C 2~-F.C':. equivalent C' = 3 Ci.e. inductance of each phase in star L' = 1.2 pH3= 0.4 pHand capacitance of each phase in star C' = 3 x 120 pF. For60 kVAr bank, the equivalent circuit can be drawn as shownin Figure 23.25.L'= 1-I.++ 1 +- 10.4 0.4 0.4_- - o: pHand C' = 3 x 120 + 3 x 120 + 3 x 120= 3 x 3 x 120 pFThe equivalent values in delta would beL= 1= - 1.2 or 0 4 pH~-+--+ 1 1 31.2 1.2 1.2and C = 120 +120+120=36OpFExample 23.4Consider the Ll system of Example 23.8, where we haveconsidered six banks of 60 kVAr each.Assume that three units of 20 kVAr each are used to makeeach bank of 60 kVAr, and let there be six such banks.Data available:System - 415 V, 3-4, 50 HzEach unit of 20 kVAr hasC = 120 pf Typical, may be obtained from theL = 1.2 pH I capacitor manufacturerConsider each capacitor unit of 20 kVAr to be connected indelta, as shown in Figure 23.24(a).Step 1There will be three such parallel circuits to make it 60 kVAr.lo calculate equivalent capacitance and reactance in delta,we may convert it into an equivalent star as shown in Figure23.24(b) by maintaining the same line parameters as in Figure23.24(a). If the impedance of each phase in delta is Z, thento maintain the same steady-state line current, I/;, in star alsolet the equivalent impedance of each phase in star be 2. Then,and415 # -/s (delta) = ~Z/,(star) = __41523 Z'or 415 43 = 415Z ,3 Z'orZ'= 1.3Since Xc=27r.f.L _.. x; =Ix2a f L3equivalent L' will be1= -L3415 V415 VL B 17Figure 23.24(a) Equivalent A circuit for a 20 kVAr capacitoiunit415 V415 V_tB&-Figure 23.24(b) A clrcuit, equivalent to A circuit of Figure 23.24(a)

231756 Industrial Power Engineering and Applications Handbook- 5 x 0.4 + 0.43x5pH01and C,eq = 9 x 120 + 9 x 120 + 9 x 120 + 9 x 120 + 9 x 120= 5 x 9 x 120 pFor 5 x 9 x 120 x FFigure 23.25 Equivalent circuit for a 60 kVAr bank in starmade of 3 x 20 kVAr unitsCapacitor normal current for each 60 kVAr bankSwitch and HRC fuse and contactor rating, as in Section25.6,or= 1.5 x 83.5125.25 AThus, for a single capacitor switching of a 60 kVAr bank, arating of 125 A will be required for the switch, fuse and thecontactor etc.Step 2: Effect of parallel switchingConsider the case when all five units are already energizedand the sixth is switched. The equivalent circuit can now berepresented as shown in Figure 23.26.:. Le, for five switched capacitor banks-& +0.4= -pH3x5and L2 eq for the total circuit11 1 10.4/3 0.413 0.413 0.4131 +- +--+ -0.413 pH9x 120PF5 x 9 x 120 x 10-6 x 9 x 120 x 1040.4x6 x10-6(5x9x120x10-6 +9x120x10-6)\I'= 4152 5x9x120x9x120-X~ ~ 6 x 9 ~ 1 2 03x5= 415 x 43750or 25 414.6 AmpsIf switching is effected when the incoming capacitor is alreadycharged, then the inrush current, I,,,, will be nearly twice this,i.e.= 25 414.6 x 2= 50.80 kAHowever, in no case will the switching current exceed theshort-circuit kA of the switching circuit.InferenceThis current is excessive for a normal switching device tomake frequently, even for one half of a cycle each time.However, on an LT circuit, it would be much below this dueto the circuit's own resistance and inductance, which havebeen ignored in the above analysis. Nevertheless, it wouldbe advisable to limit such a switching surge to a reasonablevalue to protect the system from damage during repeatedswitchings as well as the switching devices (selected at 150%of the capacitor's normal current).Step 3: Limiting the inrush currentConsider an inductance L pH introduced into each capacitorcircuit as shown in Figure 23.27(a) or (b) to limit the switchingcurrent within the making capacity of the switching device.Then(?+LjL& = 5 x 104H[y+L)x6L;eq = 5 x 104HandC;eq = 5 x 9 x 120 x 104F-1 2 3 4 5 6StepsUnit being< , switched nowAlready switched units ( G)(Lq and Cleq)Figure 23.26 Equivalent switching circuit for six capacitorunits of 60 kVAr eachThe equivalent switching circuit is shown in Figure 23.27(c)::. ImprovedJ2 5 x 9 x 120 x 10-6 x 9 x 120 x 10-6I' --~415m - 6 /(e+L)x6IC" ' x10-6x6x9x120x10-65

415 VIt10.413 pH9 x 120pHPower capacitors: behaviour, switching and improvement of power factor 23/757devices such as the switch or the contactor will generallyhave a making capacity of nearly eight times (peak valuefi x 8) its normal rating. (See the note below.) These dataare provided by the component manufacturers. Let us thereforeassume that /; is to be restricted up to & x 125 x 8, Le.& x 1000 A (the actual value may be more than this, whichmay be obtained from the manufacturers' catalogues).Therefore, in the above case,415' x2x5x9x12Ox5-~L=3 x 6 x 6 x 2 x 1000' 3 pHor= 43.06 -- 0.4pH342.93 pH415 V$-(a) Equivalent AIf we are able to provide an inductance of this value witheach capacitor bank of 60 kVAr the problem of excessiveinrush transient current can be overcome and the componentratings as chosen above will be sufficient to switch a parallelcircuit.415 VIStep 4: Reactive power of the series inductanceHowever small, the reactive kVAr of series inductance maybe, it would offset as much of the capacitive kVAr. It would beworth while therefore to keep this aspect in mind to ensurethat the capacitive kVAr chosen to improve the p.f. of thesystem to a certain level is not over-adjusted. Otherwise ahigher capacitive kVAr would become necessary to achievethe same level of p.f. as was envisaged initially. In the abovecase,1: .x,Inductive kVAr = 3 x -[X, = reactance in Wphase]1000where /, = the capacitive current I,(b) Normal connection A= 83.5AL0.413pHand x, = 2n x 50 x 42.93 x Wphase= 0.0135 ni-1 3 4 5StepsFigure 23.27or [y + L) = 415' x(c) Equivalent switching circuitEquivalent switching circuits2 x 5 x 9 x 120 x 53 x 6 ~ 6 ~ 1 ; ~-Ep~or L=415* ~ 2 ~ 5 ~ 9 ~ 1 2 O X 53x6x6xI&' 3J9 x 1206In this case we have considered the switching device tobe rated for 125Afor each bank. In an LTsystem the switching(83.5)'~ 0.0135:. Inductive kVAr = 3 x1000= 0.28which is even less than 0.5% of the capacitor kVAr per bankand may be ignored.Note1 For large LT or HT banks compensating large installations,transmission or distribution networks, the value of seriesreactance may become large. In which case, thesignificance of the above phenomenon will appear to bemore meaningful. It is, however, seen that in view of theline impedances (which have been ignored in the aboveestimate) the actual reactance may not normally be requiredby more than 0.5% to 1 Yo of the total installed kVAr. Usinga series reactance equal to 0.2% of the kVAr rating tocontrol the inrush currents is quite prevalent.2 In HT systems, where a series reactor is already beingused, to suppress the system harmonics this would alsoserve to limit the switching inrush currents and no separatereactor would be necessary.3 Since a series reactor is of a relatively small value, it maynot be able to withstand the system fault conditions. Inthat case, it is advisable to connect it on the neutral sideof the star-connected bank rather than on the line side.

23/758 Industrial Power Engineering and Applications HandbookConversion from delta to starWhenever an electrical network has different configurations,such as star and delta, it must first be convertedto an equivalent star or delta before conducting anyanalysis. As derived above, the following rules of thumbmay be applied:1Reactances in star = - x reactances in delta3i.e.L’=-L 13and capacitance in star = 3 x capacitance in delta, i.e.C’ = 3cwhere L = inductance in A per phaseL‘ = inductance in Y per phaseC = capacitance in A per phaseC’ = capacitance in Y per phase23.11.1 Transient-free switchingWith the use of static switchings through IGBTs orthyristors (SCRs), as discussed in Sections 6.9 and 24.10both switching overvoltages and inrush currents can becompletely eliminated. Switching are now possible atthe instant the applied voltage wave passes through itsnatural zero. Since such a switching scheme is free fromany overvoltage or inrush currents, the number ofswitching operations is no problem. Also refer to Section6.16.1 on soft switchings.Although costlier, when a smoother and faster p.f.correction is desirable, without causing an overvoltageor inflow current, static switchings should be chosen.They have most application in large automatic reactivepower controls, as discussed in Section 24.10.23.11.2 Designing an inductor to limit the inrushcurrentsA simple coil (even a straight length of cable will servethe same purpose, unless it is too long) as shown inFigure 23.28, will provide a self-inductance when asinusoidal current is passed through it. We will make useof such a coil to control the excessive currents. The inducedinductance in such a coil can be expressed by4 a’Figure 23.28 A circular coiled coilSection-aa’L=-. 4’N2 10’ r(1og. o,7&8y- r, - 2whereL = self-induced inductance of the coil in henryN = number of turns of the coilr = mean radius of the conductor in cm andr, = radius of the cross-section of the cable in cmExample 23.5 (see Example 23.4)For obtaining a self-inductance of 42.93 pH consider a coil of15 cm mean diameter (r= 7.5 crn) made of the same cablethat is connecting each capacitor bank through the switchingdevice.(a) Select a cable of 50 mrn2 of copper, rated for almost 125A at 45°C ambient. This cable can also withstand a shorttimecurrent of @ x 1000A for a few seconds. See alsoSection 28.4.1, the graph of Figure 28.5, and SectionA16.7,where -. 4‘x 1 Ji = 0.1250or t = 18 secondsusing a PVC insulated flexible copper cable, having anominal outer diameter of 12 mm and an insulationthickness of 1.4 mm (typical).Diameter of conductor = 12 - 2 x 1.4= 9.2 mmandr, = -9.2= 4.6 mrn = 0.46 cm2:. 42.93~10~ = -xN’x7.5 41Tlog,109 ( 0.7788 x 0.46-_4n- x N‘ x 7.5(10ge 167.48 - 2)1094n--x N’ x 7.5 (5.12-2)109or48= - x N2 x 7.5 x 3.1210942.93 x lo4 x lo9”=y 41~~7.5~3.12= 12.08 say, 12 turnsBy providing 12 turns of 150 rnm mean diameter of the 50mm2 flexible copper cable connecting each 60 kVAr capacitorbank a self-inductance of roughly 42.93 x 104H can beintroduced into each switching circuit, which will limit theswitching inrush current to almost the permissible value ofthe making current (I,,,) of the switching device.

xPower capacitors: behaviour, switching and improvement of power factor 231759Insulation ' 2 m m J L9 mm 4'(b) If we use two 25 mm2 cables as shown, thenr1 = 7.8 mm or 0.78 cmand the number of turns that will be necessary:42.93 x = % x N' x 7.5 log, 7.510 ( 0.7788 x 0.784n= - x N' x 7.5(10ge 98.77 - 2)= ~1 og4n109or N = 13.26Say, 13 turnsSwitching frequencyAs calculated above:C1 eq = 5 x 9 x 120 xN' x 7.5 x 2.593and C, = C2 = 9 x 120 x 1 O-6FF1 (5 x 9 x 120 x 10-6) + (9 x 120 x 10-6)- - ,-2n\ i""3x5x 10+)(5 x 9 x 120 x 10-6)- lo3 I 6 x 3 ~ 5 ~ 1 0 ~2rr d 0.4 x 6 x 5 x 9 x 120= 13.26 kHzThis will be the surge frequency of the switching circuit whenthe sixth capacitor is switched on a circuit that has five capacitorunits already switched. The improved value of this switchingfrequency, f;, after the introduction of inductance 42.93 pHin each capacitor circuit will become.L2e.q =/y+L)x65x 10-6HCleq, = C1 eq = 5 x 9 x 120 xC, = C2 = 9 x 120 x FTherefore the only change in the previous calculation is forLZeq, which is now substituted by (0.4 + 3L), i.e. (0.4 + 3 x42.93) or 129.19 for 0.4 in the numerator,F-2)7- -:. fl = 13.26 1- 0.4 kHz\I 129.19= 13.26 x 0.0556 x lo3 Hz= 738 Hzwhich is substantially dampened with the use of additionalinductance, and is only slightly more than the switchingfrequency on single capacitor switching as worked out inExample 23.3.23.1 2 Capacitor panel designparametersSince this device is also associated with the same powersystem as a switchgear assembly, it should generallymeet the same specifications (Section 13.4) except forsmall variations in operating conditions and testrequirements. Capacitors generate excessive heat whenin service. Installations employing large capacitor banksmust therefore have a capacitor mounting structure suitableto dissipate heat freely and permit circulation of fresh airduring normal operation. To achieve this, open-typeenclosures are usually preferred when it is possible tohouse the panel in a separate room or mount the units ona structure in an open switchyard. It may also be providedwith an expanded metal enclosure. Forced cooling withinthe enclosure or the room where such banks are installedis common practice to dissipate the excessive heat. Formore details refer to the following publications:IEC 6083 1 - 1 and 6093 1 - 1 for LT, and IEC 6087 1 - 1 andIEC 60143- 1 for HT systems23.12.1 Selecting the voltage rating of thecapacitor unitsWhen selecting the voltage of the capacitor units caremust be taken that during operation the voltage acrossthe capacitor units does not fluctuate beyond *lo%. Ifthis happens the voltage rating of the capacitor unitsmust be chosen so that the variation across the unitsunder unfavourable operating conditions does not exceed*lo%.Example 23.6If 3000 kVAr capacitor banks are required for a system of 33 kV+ 7.5'4, -5% and the capacitors are rated for 34/?'3 kV thenthe output of the capacitors at nominal voltage will reduce to(g)' x 3000 = 2826 kVArwhich will result in an undercompensation by 174 kVAr or5.8%. If the system voltage falls to, say, 31.5 kV during peakload periods the rating of the capacitor banks will fall furtherand may render the system unstable,Effective kVAr at 31.5 kV= [ x 3000= 2575 kVAr

23/760 Industrial Power Engineering and Applications Handbookwhich will result in an undercompensation by 425 kVAr, or14.2%. It is therefore advisable to select the voltage rating ofthe capacitor units at almost the average voltage of the system,which in the above case, will be- 33 x 1.075 + 33 x 0.95 = 33.4 kV2The capacitor banks may be designed for 33.4/ & kV.See also Example 23.1 to account for the voltage rise dueto series reactor in case series reactors are used.23.12.2 Determining the kVAr ratingConsider Figure 23.29(a), where, the p.f. of a powercircuit is to be improved from cos qbl, to cos q&. If kVArlis the reactive component of power at p.f. cos &, whichis to be improved to kVAr2, at p.f. cos &, through thereactive power compensation, then the reactive componentof power compensated or kVAr rating of the requiredcapacitor bankskVAr = kVArl - kVAr,and from Figure 23.29(a)-- kVAr,- tan 4kW:. kVAr, - kVAr2 = kW(tan $, - tan $?) = kW . K(23.17)where K is a multiplying factor. For quick application ofthis equation and to simplify calculations, the factor Khas been worked out for different values of cos #I andcos & and reproduced in the form of Table 23.6.Example 23.7For a load of 75 kW, having a p.f. of 0.75, the capacitor ratingto improve it to 0.95 can be calculated as follows:COS $1 = 0.75$1 = 41.41"and tan $1 = 0.882COS $2 = 0.95.. & = 18.19"and tan & = 0.329:. tan $1 -tan & = 0.882 - 0.329or K = 0.553(the same value can easily be determined from Table 23.6)and the required rating of capacitors,kVAr = 75 x 0.553= 41.5 kVArSay, 40 kbArSee also Figure 23.29 (b).Figure 23.29(a)capacitorDetermining the kVAr rating of a shunt23.13 Capacitor rating for aninduction motor75 kW VcLine current aftercornwensationFigure 23.29(b) Reduction in line current after power factorcompensationThe selection of capacitor rating, for an induction motor,running at different loads at different times, due either tochange in load or to fluctuation in supply voltage, isdifficult and should be done with care because the reactiveloading of the motor also fluctuates accordingly. Acapacitor with a higher value of kVAr than the motorkVAr, under certain load conditions, may developdangerous voltages due to self-excitation. At unity powerfactor, the residual voltage of a capacitor is equal to thesystem voltage. It rises at leading power factors (Figure23.30). These voltages will appear across the capacitorbanks when they are switched off and become a potentialsource of danger to the motor and the operator. Such asituation may arise when the capacitor unit is connectedacross the motor terminals and is switched with it. Thismay happen during an open transient condition whilechanging over from star to delta, or from one step toanother, as in an A/T switching, or during a tripping ofthe motor or even while switching off a running motor.In all such cases the capacitor will be fully charged and

~Power capacitors: behaviour, switching and improvement of power factor 23/761Table 23.6 Chart for selection of capacitor rating, to improve the existing p.f. to a higher levelPresent Factor K for required power factorpowerfactor 0.80 0.85 0.90 0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99 Unity0.40 1.537 1.668 1.805 1.832 1.861 1.895 1.924 1.959 1.998 2.037 2.085 2.146 2.2880.41 1.474 1.605 1.742 1.769 1.798 1.831 1.860 1.896 1.935 1.973 2.021 2.082 2.2250.42 1.413 1.544 1.681 1.709 1.738 1.771 1.800 1.836 1.874 1.913 1.961 2.022 2.1640.43 1.356 1.487 1.624 1.651 1.680 1.713 1.742 1.778 1.816 1.855 1.903 1.964 2.1070.44 1.290 1.421 1.558 1.585 1.614 1.647 1.677 1.712 1.751 1.790 1.837 1.899 2.0410.45 1.230 1.360 1.501 1.532 1.561 1.592 1.626 1.659 1.695 1.737 1.784 1.846 1.9880.46 1.179 1.309 1.446 1.473 1.502 1.533 1.567 1.600 1.636 1.677 1.725 1.786 1.9290.47 1.130 1.260 1.397 1.425 1.454 1.485 1.519 1.552 1.588 1.629 1.677 1.758 1.8810.48 1.076 1.206 1.343 1.370 1.400 1.430 1.464 1.497 1.534 1.575 1.623 1.684 1.8260.49 1.030 1.160 1.297 1.326 1.355 1.386 1.420 1.453 1.489 1.530 1.578 1.639 1.7820.50 0.982 1.112 1.248 1.276 1.303 1.337 1.369 1.403 1.441 1.481 1.529 1.590 1.7320.5 1 0.936 1.066 1.202 1.230 1.257 1.291 1.323 1.357 1.395 1.435 1.483 1.544 1.6860.52 0.894 1.024 1.160 1.188 1.215 1.249 1.281 1.315 1.353 1.393 1.441 1.502 1.6440.53 0.850 0.980 1.116 1.144 1.171 1.205 1.237 1.271 1.309 1.349 1.397 1.458 1.6000.54 0.809 0.939 1.075 1.103 1.130 1.164 1.196 1.230 1.268 1.308 1.356 1.417 1.5590.55 0.769 0.899 1.035 1.063 1.090 1.124 1.156 1.190 1.228 1.268 1.316 1.377 1.5190.56 0.730 0.860 0.996 1.024 1.051 1.085 1.117 1.151 1.189 1.229 1.277 1.338 1.4800.57 0.692 0.822 0.958 0.986 1.013 1.047 1.079 1.113 1.151 1.191 1.239 1.300 1.4420.58 0.655 0.785 0.921 0.949 0.976 1.010 1.042 1.076 1.114 1.154 1.202 1.263 1.4050.59 0.618 0.748 0.884 0.912 0.939 0.973 1.005 1.039 1.077 1.117 1.165 1.226 1.3680.60 0.584 0.714 0.849 0.878 0.905 0.939 0.971 1.005 1.043 1.083 1.131 1.192 1.3340.61 0.549 0.679 0.815 0.843 0.870 0.904 0.936 0.970 1.008 1.048 1.096 1.157 1.2990.62 0.515 0.645 0.781 0.809 0.836 0.870 0.902 0.936 0.974 1.014 1.062 1.123 1.2650.63 0.483 0.613 0.749 0.777 0.804 0.838 0.870 0.904 0.942 0.982 1.030 1.091 1.2330.64 0.450 0.580 0.716 0.744 0.771 0.805 0.837 0.871 0.909 0.949 0.997 1.058 1.2000.65 0.419 0.549 0.685 0.713 0.740 0.774 0.806 0.840 0.878 0.918 0.966 1.027 1.1690.66 0.388 0.518 0.654 0.682 0.709 0.743 0.775 0.809 0.847 0.887 0.935 0.996 1.1380.67 0.358 0.488 0.624 0.652 0.679 0.713 0.745 0.779 0.817 0.857 0.905 0.966 1.1080.68 0.329 0.459 0.595 0.623 0.650 0.684 0.716 0.750 0.788 0.828 0.876 0.937 1.0790.69 0.299 0.429 0.565 0.593 0.620 0.654 0.686 0.720 0.758 0.798 0.840 0.907 1.0490.70 0.270 0.400 0.536 0.564 0.591 0.625 0.657 0.691 0.729 0.769 0.811 0.878 1.0200.7 1 0.242 0.372 0.508 0.536 0.563 0.597 0.629 0.663 0.701 0.741 0.783 0.850 0.9920.72 0.213 0.343 0.479 0.507 0.534 0.568 0.600 0.634 0.672 0.712 0.754 0.821 0.9630.73 0.186 0.316 0.452 0.480 0.507 0.541 0.573 0.607 0.645 0.685 0.727 0.794 0.9360.74 0.159 0.289 0.425 0.453 0.480 0.514 0.546 0.580 0.618 0.658 0.700 0.767 0.9090.750.760.770.780.790.800.810.820.830.840.1320.1050.0790.0530.0260.2620.2350.2090.1830.1560.1300.1040.0780.0520.3980.3710.3450.3190.2920.2660.2400.2140.1880.1620.4260.3990.3730.3470.3200.2940.2680.2420.2160.1900.4530.4260.4000.3740.3470.3210.2950.2690.2430.2170.4870.4600.4340.4080.3810.3550.3290.3030.2770.2510.5190.4920.4660.4400.4130.3870.3610.3550.3090.2830.5530.5260.5000.4740.4470.4210.3950.3690.3430.3170.5910.5640.5380.5120.4850.4590.4330.4070.3810.3550.6310.6040.5780.5520.5250.4990.4730.4470.4210.3950.6730.6520.6200.5940.5670.5410.5150.4890.4630.4370.7400.7130.6870.6610.6340.6080.5820.5560.5300.5040.8820.8550.8290.8030.7760.7500.7240.6980.6720.0260.6450.85 - - 0.136 0.164 0.191 0.225 0.257 0.291 0.329 0.369 0.417 0.478 0.6200.86 - - 0.109 0.140 0.167 0.198 0.230 0.264 0.301 0.343 0.390 0.450 0.5930.87 - - 0.083 0.114 0.141 0.172 0.204 0.238 0.275 0.317 0.364 0.424 0.5670.88 - - 0.054 0.085 0.112 0.143 0.175 0.209 0.246 0.288 0.335 0.395 0.5380.89 - - 0.028 0.059 0.086 0.117 0.149 0.183 0.230 0.262 0.309 0.369 0.5120.90 - - - 0.031 0.058 0.089 0.121 0.155 0.192 0.234 0.281 0.341 0.4840.91 - - - - 0.027 0.058 0.090 0.124 0.161 0.203 0.250 0.310 0.4530.92 - - - - - 0.031 0.063 0.097 0.134 0.176 0.223 0.283 0.4260.93 - - - - - - 0.032 0.066 0.103 0.145 0.192 0.252 0.3950.94 - - - - - - - 0.034 0.071 0.113 0.160 0.220 0.3630.95 - - - - - - - - 0.037 0.079 0.126 0.186 0.329

23/762 Industrial Power Engineering and Applications HandbookV, > ES (1.R ignored)(a) Receiving end voltage rises with leading p.f.'sIV, < E, (/.R ignored)(b) Receiving end voltage diminishes with lagging p.f.'sFigure 23.30its excitation voltage, the magnitude of which dependsupon the p.f. of the system, will appear across the motorterminals or any other appliances connected on the samecircuit. The motor, after disconnection from supply, willreceive the self-excitation voltage from the capacitor andwhile running may act as a generator, giving rise tovoltages at the motor terminals considerably higher thanthe system voltage itself. The solution to this problem isto select a capacitor with its capacitive current slightlyless than the magnetizing current, I,,,, of the motor, say,90% of it. See also Figure 1.15.At voltages lower than rated, the no-load current, In,,and the magnetizing current, Z, of the motor is low andrises with the voltage. At loads lower than rated, althoughthe p.f. will diminish sharply as discussed in Section 1.8,the reactive component I, and the no-load current Ink,i.e. component OA of Figure 1.16, will remain the same.If these facts are not borne in mind when selecting thecapacitor rating, particularly when the p.f. of the motoris assumed to be lower than the rated p.f. at full load,then at certain loads and voltages it is possible that theTable 23.7 Recommended capacitor ratings for direct switching with induction motors, to improve power factor to 0.95 or better atall loads-MotorH. p:2.557.51012.5Capacitor rating in kVAr ut a motor rp.m3000 1500 1000 750 600 500rp.m. %p.m. rp.m. rp.m. rp.m. rp.m1 I 1.5 2 2.5 2.52 2 2.5 3.5 4 42.5 3 3.5 4.5 5 5.53 4 4.5 5.5 6 6.53.5 4.5 5 6.5 7.5 8Motor Capacitor ruting in KVAr at a motor r.p.m.H. p:3000 1500 1000 750 600 500rp.m.__-~rp.m. rp.m. rp.m. rp.m. rp.m10511011512012522 24 27 29 36 4123 25 28 30 38 4324 26 29 31 39 4425 27 30 32 40 4626 28 31 33 41 471517.52022.52527.53032.53537.54 54.5 5.55 65.5 6.56 I6.5 7.57 87.5 8.58 98.5 9.566.5789__9.510I111.5127.5 8.5 98 10 10.59 I1 1210 12 1310.5 13 14.511.5 14 1612 15 1713 16 1813.5 17 1914 18 201301351401451 5015516016517017518018519019520027 29 32 34 43 4928 30 33 35 44 5029 31 34 36 46 5230 32 35 37 47 5431 33 36 38 48 5532 34 37 39 49 5633 35 38 40 50 5734 36 39 41 51 5935 37 40 42 53 6036 38 41 43 54 614042.54547.5509 10 13 15 19 219.5 11 14 16 20 2210 11.5 14.5 16.5 21 2310.5 12 15 17 22 2411 12.5 16 18 23 2512 13.5 17 19 24 2613 14.5 18 20 26 2814 15.5 19 21 27 2915 16.5 20 22 28 3116 17 21 23 29 3237 39 42 44 55 6238 40 43 45 56 6338 40 43 45 58 6539 41 44 46 59 6640 42 45 47 60 61556065707520521021522022541 43 46 48 61 6842 44 47 49 61 6942 44 47 49 62 7043 45 48 50 63 7144 46 49 51 64 728085909510017 19 22 24 30 3418 20 23 25 31 3519 21 24 26 33 3720 22 25 27 34 3821 23 26 28 35 4023023524024525045 47 50 52 65 7346 48 51 53 65 7446 48 51 53 66 7541 49 52 54 67 7548 50 53 55 68 76

capacitor kVAr may exceed the motor reactive component,and cause a leading power factor. A leading p.f. canproduce dangerous overvoltages. This phenomenon isalco true in an alternator.If such a situation arises with a motor or an alternator,it is possible that it may cause excessive torques. Keepingthese parameters in mind, motor manufacturers haverecommended compensation of only 90% of the no-loadkVAr of the motor. irrespective of the motor loading.This. for all practical purposes and at all loads, willimprove the p.f. of the motor to around 0.9-0.95. whichis satisfactory.Table 23.7, based on the recommendations of motormanufacturers. suggests the likely capacitor ratings fordifferent motor ratings and speeds. For higher ratings.interpolate or calculate the rating as illustrated in Section33.12.2.23.1 4 Location of capacitorsRefer to a typical distribution network shown in Figure13.3 1. The capacitor is of maximum use when located asnear to the load-point as possible, especially in inductionmotors. because:Power capacitors: behaviour, switching and improvement of power factor 231763I- 1 feedersOutgoing-- --- --IMotor ControlCentre (MCC)lICTransformer - 1Power ControlCentre (PCC)IIICtransformer - 2-DistributionBoard (DB)for utilities IThe reactive load is confined to the smallest part ofthe system.The motor starter can be used to switch thc capacitorax well as the motor, eliminating the cost of an extra\witch and fuse for the capacitor.By employing the same switch for motor and capacitor,the capacitor is automatically controlled. Thc capacitoris required to be in the circuit only when the motor isin operation. The capacitor panel can be a part of themain power control centre (PCC) or the motor controlcentre (MCC) or il can also be a separate pancl, asshown in Figure 23.32. In group loads such as anindustrial or a power-house application, it may hemore economical and practical to install capacitors ingroups, whose kVAr value can be varied as desired.depending upon the system loading at a time. Thiscan also he done automatically through a preset powerfactor correction relay (Figure 23.38). The advantagesof a group installation can be:Diversity: When a number of motor loads areconnected on a common bus. normally not all themotors will be operating at a time. A capacitor banknear the MCC would permit the use of lcsscr totalkVAr than if the capacitors were located separatelyat each individual load.When many small motors are operatingsimultaneously it is cconomical to install largercapacitors in several sections than to have manysmall capacitors installed at each motor.Figure 23.33 and 23.34 illustrate the locations discussedso far. Figure 23.33 suggests the locations of the capacitorin respect of the motor whose p.f. is to be improved, e.g.Figure 23.31industrial unitMotor feedersMCB Distribution IBoard (DB) forReceiving and distribution of power in anin position I” the capacitor is connected on the motorside after the starter. The same starter will switch boththe motor and the capacitor. Since the capacitor willreduce the kVAr demand, the current through the starter*‘When adopting this location. care must he taken that the capacitor\are not wbjected to a quick reclosing (Section 2.5.6.2(1)). In ay/A or A/T switching the capacitors would be subjected to a qulckreclosing and may endanger the motor inwlation as well as it\own. In this ca\e it would be e\sentinl to make a wtable modificationin its switching circuit to keep the capacitor out of the circuitduring thc changeover, as suggested in Figure 25.7.

23/764 Industrial Power Engineering and Applications Handbookr xft_ _ChamberON OFF(60 kVAr)PFCR - Power factorcorrection relayAfM SSW - Auto-manualselector switchF - FusesC - Contactor@ - ON P.0.@ - OFF P.B.@ - Indicating lights- FrontON OFF(60 kVAr)rFront viewL xILouvers orexpandedmetalSection X-X'* Minimum clearance betweeneach unit to be at least thewidth of the capacitor unit.Figure 23.32 General arrangement of capacitor panelrelay will be low and it should be set at a lower valueaccordingly.In position 2 the capacitor is connected on the lineside, although switched by the same switching device.The relay setting is not affected since only the full motorcurrent will flow through the starter. In position 3 thecapacitor is connected to the circuit, through an additionalswitch fuse unit.Figure 23.34 suggests possible locations where thecapacitor banks can be installed for individual or groupcontrols, depending upon cost and simplicity. Location1 will be suited for individual loads and is effective whenthere are not many load points. For group controls, location-I/C switch orcircuit breaker( Location - 4Pe,06&6 l!G\\'G- 1 Location - 2 Location - 3LocationFigure 23.34loadsCapacitor locations for individual for groupI;-GLocation - tFigure 23.33 Capacitor locations for individual compensation3 is ideal and most economical. Compared to location 3,location 2 is not appropriate unless there are feeders onthe main bus that would require a power factorimprovement in addition to capacitors at location 3. Inthis case the capacitors at location 2 will be for the feederson the main bus and not for the downstream distributionfeeders. Location 4 controls the power factor from onepoint. This location is suitable for improving the system

Power capacitors: behaviour, switching and improvement of power factor 23/765power factor, reducing the kVA demand and the strainson the main supply and distribution network. Technicallyspeaking, it has no advantages for in-plant powerdistribution, nor does it help to reduce the kVA strain onthe feeding cables or the loading on the power distributionfeeders. But for simplicity and ease of control for theentire plant it is used most often.From main PCCl/C breakerTo PFCR23.15 Automatic PF correction of asystemAs discussed above, for an industrial or power plantapplication or an installation with a number of inductiveload points a grvup capacitor control is always moreeffective. simple and economical. Such an installationgenerally has a frequent variation in its load demand dueto some feeders coming on the bus and some falling outat different times. There may be variation in the individualfeeder’s load demand, such as a tool room, where not allthe machines will be working at a time, or a pulp milland paper mill, where the paper mill has a continuousload, the pulp mill an intermillent one. A water treatmentplant or a pump house are similar installations where allor some of the loads would be in operation at a time. Forsuch installations. the total capacitive load demand at arequired power factor level is worked out and the totalcapacitor banks are installed at a convenient point andsuitably grouped (banked) for the type of loading andsystem demand. Each bank is controlled through a powercontactor and a common power Lactor correction relay toautomatically monitor and control the power factor ofthe system to a predetermined level, preset in the relay,by witching a few capacitor banks ON or OFF, dependingupon the load demand and the power factor irieasured bythe relay. The relay actuates the required number ofcapacitor feeders through their contactors.Automatic correction is always recommended toeliminate manual dependence and to achieve betteraccuracy. It also eliminates the risk of a leading powerfactor by a human error that may cause an excessivevoltage at the motor and the control gear terminals.The following example illustrates the method ofselecting the capacitors’ value, their grouping and theircontrol for a system having a number of load points.PFCR - Power factor correction relayFigure 23.35 Single-line diagram for an industrial loadFrom main PCCiExample 23.8Consider a system shown in the single-line diagram of Figure23.35, where total load on MCC-1 as in the single-line diagramof Figure 23.36 is3 x 10 h.p = 30 h.p.4 x 5 h.p. = 20 h.p.2 x 20 h.p. = 40 h.p.2 x 40 h.p. = 80 h.p.2 x 30 h.p. = 60 h.p.1 ~ 100 h.p. = 100 h.p.1 x 50 h.p. = 50 h.p.__Total380 h.p. + spares and lighting loadNote1 Lighting loads are not considered for the followingreasons:Figure 23.36 Single-line diagram for MCC-1

~~ ~~~23/766 Industrial Power Engineering and Applications Handbook(i) All incandescent lamps are resistive and do notinfluence the power factor of the system.(ii) Mercury vapour lamps and sodium vapour lampsare invariably provided with built-in p.f. improvementcapacitors by their manufacturers.(iii) Fluorescent tubes used for industrial installations anddomestic use are also provided with p.f. improvementcapacitors.2 Similarly, 100 A and 25 A spares are also not considered,assuming that the utility of such feeders would be rareand for short durations (such as for maintenance).3 Spare motor feeders are also not considered for similarreasons.Total loads for other MCCs, as in the single-line diagramof Figure 23.35:MCC-2 - 75 h.p. + Spare and lighting loadMCC-3 - 150 h.p. + Spare and lighting loadMCC-4 - 200 h.p. + Spare and lighting loadThen the total inductive load connected on the main PCCcum-MCC:MCC-1 -MCC-2 -MCC-3 -MCC-4 -5 h.p. x 1 -20 h.p. x 1 -Total -380 h.p.75 h.p.150 h.p.200 h.p.5 h.p.20 h.p830 h.p.Considering the diversity factor (Table 13.4) for such aplant to be 70%, or as the system may require, which canbe determined by the type of industry and the processdemand::. Total inductive load on the system at any time= 830 x 0.7= 581 h.p.or 581 x 0.746 i.e. 433.43 kWAssume the plant p.f. before the power factor correction tobe 0.65. (We have provided in Table 23.8 the likely operatingpower factors for different types of industries, as a roughguide.) If this p.f. is to be improved to say, 0.95, then fromTable 23.6,Factor K = 0.84and total capacitor banks required = 433.43 x 0.84 = 364kVArThus capacitor banks of 360 kVAr should be adequate forthis system and can be arranged in six steps of 60 kVAreach.For the rating of the incoming feeder,IC = 360 - 500A& x 0.415:. Rating of the incoming switch = 1.5 x 500 = 750 AThe nearest standard rating available = 800 AFor the rating of outgoing fuse and contactorsIC = 6o - 83.5A45 x 0.415:. Rating of the outgoing feeders = 1.5 x 83.5 = 125 AAutomatic controlSelect a six- or eight-step power factor correction relay. Inthe case of an eight-step relay, the sventh and eighth stepscan be left as spares, which may be used later when morecapacitor banks are considered necessary. A typical controlscheme and the power circuit for such a system is shown inFigure 23.37. A general arrangement is shown in Figure 23.32for such a capacitor control panel.23.15.1 Other methods for PF correctionLT installations may be subject to frequent load variationsand inductive switchings. It will be more desirable insuch cases to provide them with an automatic p.f.correction scheme than manual correction. It may beTable 23.8 Likely operating power factors for different types of industriesT~pe of iridustryLikely poh'er factorTvpe .f industryLikely powerfiictorCold storage and fisheriesCinemasMetal pressingConfectioneryDyeing and printing (textile)Plastic mouldingFilm studiosNewspaper printingHeavy engineering worksRubber extrusion & mouldingPharmaceuticalsOil and paint manufacturingSilk millsBiscuit factoryPrinting pressFood productsLaundries0.76 to 0.800.78 to 0.800 57 to 0.720.770.60 to 0.870.57 to 0.730.65 to 0 740.580.48 to 0.750.480.75 to 0.860.51 to 0.690.58 to 0 680.600.65 to 0 750.630.92Flour millsGas worksTextilc millsOil millsWoollen millsPottcriesCigarette manufacturingCotton pressFoundriesTiles and mosaicStructural engineeringChemicalsMunicipal pumping stationsRefineries and petrochemicalsTelephone exchangeRolling millsIrrigation pump0.610.870.860.51 to 0.590.700.610.800.63 to 0.680.590.610.53 to 0.680.72 to 0.870.65 to 0.750.64 to 0.830.66 to 0.800.72 to 0.800.50 to 0.70Notu: These figures are only indicative

cmL0

23/768 Industrial Power Engineering and Applications Handbookimpracticable to keep a close monitoring of such systemvariations and alter the capacitive reactance manually tomaintain the desired p.f. level.The situation in an HT installation will, however, bedifferent. An HT system is mostly a fixed-load one, wherea variation in the load may be only occasional and atcertain periodic intervals. To provide an automaticswitching in such cases will be a costly arrangement dueto expensive switchgear equipment. In addition, it willalso cause cumulative switching surges as a result ofrapid reclosing and interrupting sequences unless a staticswitching control is adopted as discussed in Section 24.10,which will further add to the cost. To mitigate suchconstraints, the following may be considered as the morerecommended methods for p.f. control of such installations:1 Identify the likely variations in load and p.f. during a24-hour period. Maintain a minimum fixed kVArpermanently connected in the system, at the distributionpoint for the likely constant loads. For the rest, thebanks may be selected so that only one or two aresufficient to control the whole system for the desiredp.f. level or system regulation. This will also limitoperations of the banks to only three or four a dayand less than a thousand during a year, as recommendedin Section 26.1(2). Control may now be carried out:Manually, as variations in load would almost bespecific and periodic and easy to monitor orBy making it automatic, which would now beeconomical with one or two variable banks.2 In a secondary transmission or a primary distributionnetwork it is common practice to provide the maximumpossible size of fixed capacitor banks at the transmissionpoint or on the pole at the distribution pointwith the feeder transformer. This is done to maintainthe p.f. and the regulation of the system up to a certainminimum level. Capacitors mounted out of sight oraway from easy reach may sometimes be ignored formaintenance or periodic checks. This must be takeninto account while installing the capacitor units. Tomonitor the health of capacitor units one may installan ammeter or a kVAr meter in each phase and maintaina logbook to monitor the health of capacitor units/banks through the unbalance in the three phases ofthe capacitor circuit. When warranted, preventivemeasures such as cleaning the insulators, tighteningthe terminals, and replacing or adding of a few capacitorunits to make up for the lost capacitances and to balancethe system can be undertaken.23.1 6 Switching sequencesA switching sequence that can be employed for a particularload cycle may be one of the following:23.16.1 First in last outThis is the simplest type of switching. Capacitors areswitched ON in the sequence of 1-2-3 . . . n and switchedOFF in the reverse sequence i.e. n . . . 3-2-1. In thisswitching the last switched capacitor is made to switchagain. This switching is therefore more stressful for thecapacitor as well as for the system, due to surge voltages.During a switch ON therefore some time delay must beintroduced into the switching circuit for the capacitorcharge to decay to a safe level. In this sequence, thecapacitors of each step are normally the same. It is aprimitive and unscientific switching sequence andtherefore not in much use.23.16.2 True ‘first in first out’ (FIFO)Capacitors are switched ON in the sequence of 1-2-3. . . . n but an additional logic is used to switch OFF inthe same sequence 1-2-3 . . . n, i.e. the oldest OFF capacitoris switched ON first and the oldest ON capacitor isswitched OFF first. This apparently is the best switchingsequence, giving enough time to an OFF capacitor todischarge before it is switched again. Each stage capacitorrating must be equal to avoid a wide fluctuation in thep.f. correction and hence undesirable subsequentswitchings, which may be necessitated when the capacitorsare not of equal ratings.23.16.3 Pseudo (false) ‘first in first out’In a way this is more appropriate than the FIFO, for itcan have unequal stage capacitors. Capacitors, however,are arranged in ascending order such as 10, 20, 50 . . .kVAr etc., or large or small capacitors. But, as mentionedabove, this may not provide accurate correction in thefirst instance. Now the relay may have to operate on thetheory of probability and resort to a lot of sequencing toarrive at the right correction. Such a sequencing mayalso lead to hunting and cause voltage fluctuations, becauseof varying p.f. s and also cause switching currents andvoltage surges.23.16.4 1+2+2. . .The first capacitor is half the rating of the remainingones, which are all equal. The p.f. correction is delayed,due to allowing discharge time. The smallest unit is morestressed.23.16.5 Direct combinatorial or auctioneeringswitchingHere the relay assesses the kVAr requirement of the systemand switches ON all the required capacitorssimultaneously. Theoretically any size of capacitor unitscan be used. But when a capacitor is taken out formaintenance, this can create confusion in the p.f.correction. The relay will not be aware of this and actaccordingly. The bank size may also be large and causeswitching surges.23.16.6 Special sequencingThese are sequencers and can sequence the switching ofcapacitors in any fixed pattern. Capacitors can beautomatically taken out of the circuit and others introducedin their place by a device known as ‘the load rotator’.

Power capacitors: behaviour, switching and improvement of power factor 23l769A good relay can be modified to perform a particularswitching sequence during ON and OFF and bothsequences need not be same. The following is a type ofrelay that can be modified to perform any desired switchingpattern.23.16.7 Binary switchingThis is a highly recommended method of capacitorswitching for installations that are large and require veryfine monitoring and correction of p.f. with the smallestnumber of banks. The entire reactive requirement isarranged in only a few steps yet a small correction up tothe smallest capacitor unit is possible. The relay issequenced so that through its binary counter the requiredswitching is achieved in small steps, with just four or sixsets of capacitor units or banks. The operation of theentire sequence can be illustrated as follows:1 A four-stage binary The capacitor units or banksare arranged in the ratio 1 : 2 : 4 : 8. The four-stagecapacitor bank is switched in one to fifteen steps(summation of 1 + 2 + 4 + 8), giving fifteen differentvalues of reactive power control in small steps of oneunit each. To achieve this, the relay goes up and downonly one step at a time. For example, if we have toswitch 75 kVAr in small steps, this can be arranged inthe ratio of 5 : 10 : 20 : 40 kVAr units. These capacitorscan be switched in steps of 5 kVAr each (total 15steps). Table 23.9 illustrates the switching sequencefor switching ON or OFF all the capacitor units througha four-stage relay. Since the relay will switch ON insteps of 5 kVAr only, fast switching may be constrained,particularly when a charged unit, which had beenrecently out and may be fully charged, is required toTable 23.9 A four-stage binary schemeSwitching Capacitor units Totalstepsswitched1st 2nd 3rd 4th kVAr in5 kVAr 10 kVAr 20 kVAr 4OkVAr steps of5 kVAr each6 - X X - 307 X X X - 358 - - X 409 X - - X 4510 - X - X 5011 X X - X 5512 - - X X 6013 X - X X 6514 - X X X 7015 X X X X 75x indicates the units that are ON.be switched again. This delay can be eliminated withthe use of special discharge devices, which wouldmake it possible to achieve fast switching. Timersmay be introduced into the switching circuit to allowfor the thus reduced discharge time between twoconsecutive switchings.2 A six-stage binary Now the capacitors are arrangedin the ratio of 1 : 2 : 4 : 8 : 16 : 32 units, i.e. a six-stagecapacitor bank can be switched in 63 steps, withdifferent switching combinations, at a difference ofjust one unit. If the smallest unit is of 5 kVAr, a totalcorrection of 3 15 kVAr is possible in steps of 5 kVAreach with just six sets of capacitor units and banks inthe ratio of 5 : 10 : 20 : 40 : 80 : 160 kVAr. The rest ofthe sequencing is the same, as illustrated for fourstagebinary. The six-stage sequencing scheme canbe developed along similar lines to those in Table 23.9.A six-stage binary scheme should normally be adequateto switch even very large banks. Yet this sequencing canbe further modified to achieve a more particular sequencingpattern when required, as noted below.23.16.8 Modified binary switchingA fifteen- or sixty-three-step switching, as noted above,may be cumbersome and require too large a time gap toswitch. Switchings in such small steps may indeed notbe necessary. It is possible to modify the sequencing ofthe relay to switch all the units in just a few steps byaccepting a few specific jumps, skipping a few steps inbetween still retaining the feature of small corrections. Atypical modified six-stage scheme is shown in Table 23.10,achieving full switching in just 18 steps and yetmaintaining quite small steps, in this case 20 kVAr andcorrections, still in steps of 5 kVAr each. A four-stagescheme can be modified to only six steps, as indicated inthe first six steps of this table.The binary scheme can be modified to suit any otherrequirement also. For example, for an induction furnace,requiring very fast correction, the scheme may bemodified, to have six stages in the ratio of10 : 20 : 40 : 80 : 160 : 160 kVAr to make a total of 470kVAr and corrections in steps of 10 kVAr each.For loads with fast variations, this type of a modifiedversion too may be sluggish, due to step-by-step tuning.In such cases it is possible to employ two relays, one forcoarse correction, such as through a serial relay (FIFO)which can quickly switch the large banks, and the otherfor fine tuning, which may be a binary relay. These tworelays may be combined into one hybrid unit. The binaryscheme can thus be modified to suit any required dutywith prompt a d finer corrections and least strain on aparticular unit or on the system.23.17 PF correction relaysThe basic principle of this relay is the sensing of thephase displacement between the fundamental waveformsof the voltage and current waves of a power circuit.Harmonic quantities are filtered out when present in the

~~~Switchingsteps434751555963Capacitor bank1st 5 kVAr 2nd 10 kVAr 3rd 20kVAr 4th 40kVAr 5th 80 kVAr 6th 160 kVArX -X -- X-XXXXXTotal switchedkVAr51015355575951 15135155175195215235255215295315*Modified binary for a four-stage schemeNote: It is possible to achieve any correction within a step of one unit only (5 kVAr in the above case).system. This is a universal practice to measure the p.f. ofa system to economize on the cost of relay. The actualp.f. of the circuit may therefore be less than measured bythe relay. But one can set the relay slightly higher (lessthan unity), to account for the harmonics, when harmonicsare present in the system.From this phase displacement, a d.c. voltage output isproduced by a transducer circuit. The value of the d.c.voltage depends upon the phase displacement, Le. thep.f. of the circuit. This d.c. voltage is compared with abuilt-in reference d.c. voltage, adjustable by the p.f. settingknob or by selecting the operating band provided on thefront panel of the relay, as shown in Figure 23.38.Corrective signals are produced by the relay to switchON or OFF the stage capacitors through a built-insequencing circuit to reach the desired level of p.f.A little lower p.f. than set would attempt to switchanother unit or bank of capacitors, which may overcorrectthe set p.f. Now the relay would switch off a few capacitorunits or banks to readjust the p.f. and so will commencea process of hunting, which is undesirable. To avoidsuch a situation the sensitivity of the comparator is madeadjustable through the knob on the front panel of therelay. The sensitivity control can be built in terms ofphase angle (normally adjustable from 4 to 14 degreeselectrical) or percentage kVAr. Thc sensitivity, in termsof an operating band, helps the relay to avoid a marginalovercorrection or undercorrection and hence the hunting.As soon as the system’s actual p.f. deviates from thepre-set limits, the relay becomes activated and switchesin or switches out capacitor units one by one, until thecorrected p.f. falls within the sensitivity limit of the relay.The p.f. correction relays are normally available inthree versions, i.e.Electromagnetic (being quickly outdated). They arevery slow, and may take up to 2 minutes or more toinitiate a correction.Solid state-based on discrete ICs.Solid state-based on micro-controllers (microprocessors).A delayed correction may not be desirable in applicationsthat have a fast variation of loads such as rolling mills,induction furnaces, arc furnaces, all types of weldingloads, power presses, hammers and elevators etc. Suchtype of loads draw reactive power from the supply source.Since the loads change rapidly, unless they arecompensated promptly they may cause voltagefluctuations, low p.f. and high harmonics. A rapid voltagedrop may even trip the load, which may cause seriousdamage to machines in operation. For instance, in a rollingmill, the mill may be damaged as the process ingots maybe half-way through between the rolls when the millstopped.With application of solid-state technology, thisshortcoming of an electromagnetic relay is automaticallyovercome. The solid-state relays are available withswitching stages as many as from 2 to 16. For specialapplications, they can be designed for even higher numbersof steps.A time delay is built in to allow discharge of a chargedcapacitor up to 90% before it is reswitched. This isachieved by introducing a timer into the relay’s switching

Power capacitors: behaviour, switching and improvement of power factor 23/771PF ..-9r set i OnIJSupply2OperateFigure 23.38Typical front views of automatic power factor correction relayscircuit. The timer comes on whenever an OFF signaloccurs, and blocks the next operation of a chargedcapacitor, even on an ON command, until it is dischargedto at least 90% of the applied voltage. This feature ensuressafety against an overvoltage. Normally this time is 1-3minutes for LT and 5-10 minutes for HT shunt capacitors(Section 26.3.1(5)) unless fast-discharge devices areprovided across the capacitor terminals to reduce thistime. Fast-discharge devices are sometimes introducedto discharge them faster than these stipulations to matchwith quickly varying loads. The ON action begins onlywhen the timer is released.The time of switching between each relay step is,however, quite short, of the order of 3-5 seconds. Itincludes the timings of the control circuit auxiliary relays(contactors). It may be noted that of this, the operatingtime of the static relay is scarcely of the order of three tofive cycles.In rapidly changing loads it must be ensured that enoughdischarged capacitors are available in the circuit on everyclose command. To achieve this, sometimes it may benecessary to provide special discharge devices (Section25.7) across the capacitor terminals or a few extra capacitorunits to keep them ready for the next switching. It mayrequire a system study on the pattern of load variationsand the corresponding p.fs. Fast switching, however, isfound more often in LT systems than in HT. HT systemsare more stable, as the variable loads are mostly LT.The above discussion is generally related to IC-basedsolid-state relays and in most parts to microprocessorbasedrelays of the more rudimentary types.23.17.1 Microcontroller (microprocessor)-basedrelaysMicroprocessors are the latest technology in the field ofp.f. correction. Switching can be programmed throughany of the switching sequences described above. Correctionnow is much faster, as noted already.The relay can also be programmed to identify adischarged capacitor of the required capacity for the nextswitching, in which case it would require no sequentialoperation. By calculating the p.f. level of the system therelay can switch ON/OFF the necessary capacitor unitsfrom among the eligible units available for the nextswitching. The switching is so programmed that all theunits of the same rating have almost the same operatinghours. The normal operation of a microcontroller-basedrelay can be programmed along the following lines:

23/772 Industrial Power Engineering and Applications Handbook(1 unbalance if any, by measuring the kVAr in each phaseCT to monitorsystem p.f.CT to monitorcapacitor healthFigure 23.39 Location of monitoring CTs for static relaysAs soon as it is switched on, it shifts to a 'learn mode'.In 'learn mode', it measures the kVAr value of eachcapacitor unit in terms of C in pF by switching themone by one through the CTs provided in the incomingof the capacitor circuit. The CTs are provided so as tomeasure the current of the capacitor circuit alone andnot of other loads (Figure 23.39).It stores all these data in its memory.After every correction the relay again carries out themeasurements on the capacitor units (periodically, say,in 24 hours in supervisory mode), to monitor theirhealth and to detect for any deterioration. When thecapacitance of a unit falls below the acceptable levelit can emit an alarm.Similarly it can be programmed to measure the loadand provide an^ appropriate signal."0 At the start of the correction cycle the relay calculatesthe required kVAr, examines the available capacitorvalues stored in its memory, takes a decision to switchON or OFF the appropriate capacitors and switchesthem in rapid succession to reach the target p-f. in theleast possible time. Now it uses the CTs provided inthe main incoming circuit (Figure 23.39).It can identify the next ideal unit for switching.To achieve the desired level of p.f. it switches on thecapacitor unit which is nearest to the required ratingand available for the next switching or switches off afew units in a similar way.Each capacitor circuit has a software timer whichreswitches the unit only when it has timed out.It is not necessary to arrange capacitors in any particularorder, nor of particular ratings for the different stages.But, as mentioned earlier, when desired it can also beprogrammed for any switching scheme noted above.The units need not be very small or very large.The relay can be programmed to store the operatinghistory of the cumulative service hours for each unitto obtain uniformity in ageing of each unit.The relay can also be programmed to provide a counterfor the number of switching operations each capacitorhas already performed on per day, per month or yearlybases."Since the relay is connected in the same way as a meteris, it can read, calculate, and display all the desiredoperating parameters such as, V, l,L p.f., kVAr, kW, andkVA as well as kWh, kVArh and maximum demand.*The relay can be connected to a computer and transmitall such data to a remote station for further monitoringand control.NoteTo ensure proper sensing of the incoming current and its phasedisplacement by the relay it is essential that the CTs ratio and theirVA burden chosen for the required duty are close to the actualrequirement as noted in Table 15.8. Sometimes this fact is overlookedand CTs with a much higher VA burden or ratio or both are chosenwhile the secondary circuit may not be adequately loaded. In thiscase the CTs may not accurately transform the primary parametersto the secondary and, in turn, the relay may not send accuratesignals. Moreover, the relay itself may operate only at minimum1% or more of its rated current (1 or 5 A), depending upon itsdesign and type (IEC 60051-1).*These are special features that can be built into the relay. But eachmay involve an element of cost and the relay manufacturers maytherefore provide only the essential features as standard and theremaining ones on request.

Power capacitors: behaviour, switching and improvement of power factor 231773Relevant StandardsIEC Title IS BS60051- 1/199760947-3/ 199860947-4-1/199060143-11199260831- 1/199660871 - 1/199760931-1/1996Direct acting indicating analogue electricalmeasuring instruments and their accessoriesDefinitions and general requirements1248-1 to 9Specification for LV switchgear and controlgear. 13947-311993Switches, disconnectors, switch-disconnectors and fusecombinationunitsLow voltage switchgear and controlgear 13947-4-Electromechanical contactors and motor starters111993Series capacitors for power systemsShunt power capacitors of the self healingtype for a.c. systems having rated voltage up toand including 1000 VShunt capacitors for ax. power system havingrated voltages above 1000 V9835/199113340/199313925- 111998Shunt power capacitors of non self-healing types 13585-111994up to and including 1000 VGeneral performance, testing and rating. Safetyrequirements. Guide for installation and operationRelevant US Standards ANSUNEMA and IEEEBS 89-1/1990BS EN 60947-3/1992BS EN 60947-4-1/1992BS EN 60143-1/1993BS EN 60831-111998BS EN: 60871-111998BS EN 60931-1/1996ANSVIEEEC.37.012/1989ANSVIEEEC.37.26/1991ANSUIEEE-5 19/1993ANSVIEEE 643/1992ANSUIEEE 776/1998ANSI/IEEE 1036/1993ANSVIEEE 1137/1998ANSI C93.1/1981NEMA- 10711993NEMA/ICS-2/1993Application guide for capacitive current switching of a.c. HV circuit breakers rated on asymmetrical current basisGuide for methods of power factor measurement for LV inductive test circuitsGuide for harmonic control and reactive compensation of static power convertersGuide for power line carrier applicationRecommended practice for inductive coordination of electrical supply and communicationlinesGuide for application of shunt power capacitorsGuide for the implementation of inductive coordination mitigation techniques and applicationPower line carrier coupling capacitorsMethod of measurement of radio influence voltage (RIV) of HV apparatusIndustrial control and systems, controllers, contactors and overload relays, rated not morethan 2000 V a.c. or 750 V d.c.Notes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

23/774 Industrial Power Engineering and Applications HandbookList of formulae usedInfluence of harmonics on the performance of acapacitor unit(i) By the harmonic voltagev, =1/(v,2+v;+v;+v;+...v;") (13.1)V, = effective harmonic voltageVI = system voltageV,,, V,,, V,, and V,, etc, = magnitudes of the harmonicvoltage components in terms of fundamentalvoltage. at different harmonic orders(ii) By the harmonic currentsI,, =,/(I:+ 91:h3 + 251:h,+ 491:,,,+ ... n2 I&) (23.2)I,, = effective harmonic currentI, = rated current of the capacitorIch5, I,,, and ILhn ctc. = magnitude of the harmoniccurrent components at different harmonic orders(iii) Harmonic output of a capacitor unitRating of a shunt capacitor43. v'kVAr =1000 ' X'or kVAr =&.V'.2?rf.cIO00Effective harmonic outputInfluence of harmonics on telephone lines(I) Electrostatic inductionBlockirig circuit1= 2n.,fh = notch frequency(23.3)(23.4)(23.5)(23.6)(23.7)V, = induced electrostatic voltage in the CommunicationlinesC, = coupling capacitance between the power and thecommunication linesCg = ground circuit capacitance of the communicationlinesFilter circuits: use of reactor enhances voltageacross the capacitor hanksRise in voltage,Vph = system voltageVc = voltage across the capacitor banksXc = capacitive reactanceXL = inductive reactanceSingle-capacitor switching(i) Inrush current(23.8)I,, = I ,( 1 + F) (23.9)I, = maximum inrush or making current, while switchingan unchanged capacitor unit.I, = maximum stead -state current (capacitor peak fullload current (&Ic ))Isc = fault level of the system(23.10)L, and C, are the switching circuit parameters(ii) Switching frequencyi;'-=f.F(Short circuit kVA) (23.11)' =f (Capacitor kVAr) XL.f, = switching or transient frequencyf = powcr frequencyX, = power frequency capacitive reactance of the circuitX, = power frequency inductive reactance of the circuitParallel switching of capacitor banks(i) Inrush current(23.12)Z, = surge impedanceL2 = circuit inductance between the capacitors inhenryC,, C2 = capacitances of two capacitorsc, ' c2= JG(23.13)

Power caoacitors: behaviour, switching and improvement of power factor 23/775I,,, = peak inrush current in parallel switchingV, = line to neutral peak voltager = mean radius of the coil in cmrl = radius of the cross-section of the coil in cm(ii) Switching frequency(23.14)Determining the kVAr rating for improving p.f.from cos 4, to cos 42kVAr, - kVAr2 = kW(tan q5, - tan &) = kW.K (23.17)L,, = equivalent inductance of the capacitors alreadyswitchedLZeq = L,, + L,C,,, = equivalent capacitance of the capacitors already5 witchedC, = capacitance of the capacitor being switchedLimiting inrush currentsc, ’ c,I,:, = improved inrush currentL = series inductance being introducedV,, = line to neutral peak voltageDesigning an inductor to limit the inrushcurrentsL = self-induced inductance of the coil in henryN = number of turns of the coil(23.15)Further readingI AIEE Committee. Report 017 the Operation of SwitchedCupucirors, Vol. 74, Part Ill, 1255-1261. December (1955).2 Central Board of Irrigation and Power, 1ndia;Interference ofpower lines with telephone lines’ Technml Report No. 29.Research Scheme on Power, March (1982).3 ‘Cigre, Harmonics voltage unbalanced voltage dips and voltagefluctuations’. Study Committee 36, ’Interference,’ WorkingGroup 36.05, Published by Electra, March (1989).4 Johnson, LB., Schultz, A.J., Schultr. N.R. and Shores. R.B..’Some fundamentals on capacitance switching‘. AIEETrunsuctions. 727-736, August (1955).5 Narasimha, G., ‘Capacitor switching and its effect on LTswitchgear’, Electrical India, 5-8, 30 November (1982).6 Savir-Silbermann, S., ‘Surges caused by switching of capacitorbanks’, Electrical Energ?; 88-95. March (1958).7 Sharma. S.P., ‘Switching capacitor banks in L.T. switchgear’,IEEMA Journal, September (1989).8 Shenoy, G., ‘Reducing harmonic currents and improving powelfactorin installations having predominant thyristor loads’,Siemens Circuit, XVII, No. 1 (1982).9 Vansickle, R.C. and Zaborszky. J., ’Capacitor switchingphenomena’. AI€€ Trun.saction.s, 70, Part I, 151-159 (195 I).IO Vansickle, R.C. and Zaborszky, J., ‘Capacitor switchingphenomena with resistors’. AIEE Trmscrcrion.~. 97 1-977. August(1954).

System VoltageRegulation241777

~~ transmissionSystem voltage regulation 24177924.1 Capacitors for improvement ofsystem voltage regulationAnother important application of capacitors is to improvethe voltage regulation of a power supply system. Theregulation of a power system at the receiving end isdefined byRegulationVoltage at no load - Voltage at full load-~ x 100Voltage at no load(24. I)Higher regulation will mean a higher voltage fluctuationat the receiving cnd, resulting in poor stability of thesystem. Regulation up to 3-5% may be consideredsatisfactory. To improve the regulation of a system, powercapacitors can be used in series at the receiving end ofthe \ystem.24.2 Series capacitorsThe basic purpose of series capacitance is to offset thecontent of excessive line inductance, reduce the linevoltage drop, improve its voltage regulation and enhancethe power transfer capability and hence the stability levelof the system. It can accordingly find application at allhigh-current and high-impedance loads such asAn electric arc furnace. where heating is caused byarc plasma between the two electrodes. The arcingmakes the circuit highly inductive, besides generatingunbalanced currents (third harmonics), due to differenttouchdown arc distances in the three electrodes whichmake it a non-linear impedance load.An induction furnace, where the heating is due to eddycurrent losses induced by the magnetic field.Electric arc and resistance welding transformers as forspot. seam and butt welding.A long transmission line, say, 400 km and more. for aradial line and 800 km and more for a symmetricalline, as discussed later.It can also be applied to an HT distribution networkthat has ii high series inductive reactance to improveits receiving-end voltage.In all these applications a shunt capacitance is of littlerelevance. as it will not be able to offset the line inductivereactance, X,, with X,, and hence will be unable to containthe switching voltage dips at the load end in furnacesand also voltage drops during a change of load in atransmission or HT distribution network. A shunt capacitoroffsets the reactive component of the current (Figure23.2) while the line voltage drop, for the same line current,remains unaltered. Series capacitors are therefore moreappropriate where voltage regulation is the main criterion,rather than line loss reduction. Summarizing the above,the main functions of a series capacitance can be statedas follows:Transmitting-sidevoltage E,1 PrimaryI- transmission1 (Generator side)I!-ISeries capacitorsReceiving-endvoltage E,Secondary(Load side)Figure 24.1 A simple transmission network with seriescompensation1 To neutralize and reduce substantially the content ofinductive reactance of the line. Refer to a simpletransmission network with series compensation. shownin Figure 24.1.2 To alter the circuit parameters L and C, reduce theline impedance and hence the voltage drop, and alsoto enhance utilization. i.e. the power transfer capabilityofthe line.3 To improve the far end or the load-side voltage. inother words, the voltage regulation and the stabilitylevel of the I;ystem.NotesUnlike the above, a \hunt capacitor alterh the load currcnt byoffsetting the reactive coinponcnt of thc current (Figure 23.2)hy improving the load p.fand altering the characteristics of theload.A series capacitor has little application it1 an LT system due tothe high content of line resi\tance and bery little of inductance.Any amount of reactive compensation will scarcely influencethe performance of the line, a\ a rerult of the high content illIR. compared to IX,.Series and shunt cauacitors both urovide the same degreecof compensation. But it is the correct reactive supportthat provides a more stable system less prone to load andvoltage fluctuations. Thus a judicious choice betweenthe shunt and the series capacitors is required. In thefollowing our main thrust is to arrive at the mostappropriate type and extent of reactive support to achievea higher level of utilization of a power transmission ordistribution system, on the one hand. and more stability,on the other.24.3 Rating of series capacitorsReferring to Figure 24.2. thi4 can be expres4ed bykVAr=3 I,’ X, (24 3)

24/780 Industrial Power Engineering and Applications HandbookGTtE,LineparameterskVArLoad side 31x, = 1 0 0 0 ' ~31'Series E, c=1000 . kVAr. 2dcapacitorskVar = 3$ rating of the series capacitors1 = Rated current of series capacitorsFigure 24.2 The single-line diagram for Figure 24.1whereI, = line current. The value of line current to beconsidered for calculating the size of capacitor banksmust take account of the likely maximum loadvariation during normal operation or the overloadprotection scheme provided for the capacitors,whichever is higher.X, = capacitive reactance of the series capacitors perphase.and voltage rating = I, . X,.This rating will be much less than the nominal voltageof the system. But since the series capacitors operate atthe line voltage, they are insulated from the ground andfrom each phase according to the system voltage. Forthis purpose, they are generally mounted on individualplatforms for each phase, which are adequately insulatedfrom the ground. Figure 24.3 shows such an installation.24.4 Advantages Of seriescorn pensat ion(i) Automatic voltage regulation: Since the VAr of aseries capacitor 0~ I;, the voltage regulation isFigure 24.3 The installation of HT capacitor banks (Courtesy: Khatau Junker Ltd)

System voltage regulation 241781automatic. as the VAr of the series capacitors will\ary with a change in the load current. When themltage drops, the line current will rise. to copewith the same load demand and so will rise the VArof the capacitors also providing an automatic higherVAr compensation. When the voltage rises, thecurrent will fall and so will fall the VAr compensation.No switching sequence, as necessary in shuntcapacitors. is therefore required for series capacitors.(ii) They may be connected permanently on the system.;IS they compensate the line reactance. which is fixed,though the load reactance may be variable. unlikeshuni capacitors. which are to be monitored for theiraddition or deletion during peak and offpeak loadperiods respectively. The costs of switchingcquipment and operational difficulties are thereforelow in series capacitor\.(iii) They also provide the same degree of p.f. improvementas the shunt capacitors and do so by the leadingvoltage phasor rather than the current phasor.Liniitritiorzs(a) It i5 not advisable to use them on circuits that havefluctuating loads or frequent inrush currents. such asswitching of motor loads. During a start the latterwill cause an excessive current. I,,, and proportionatelyraise the potential difference across the capacitor units(I,, . X,) and overload them in addition to causinghigher dielectric stresses. Series capacitors for suchinstallations must he designed for very high voltages.say, up to I,, . Xc/l 5, which will not be economical.Also, protection against over\ oltages will still beessential as a safeguard against ;L similarcontingency,a\ discussed later.(b) It is possible that the natural (suh-harmonic) frequency( I/?rdLC, Section 17.6.3) of a system with seriescapacitors will fall below the fundamental frequencyand render the sptcm more prone to resonance. Nowresonance may occur below the fundamentalfrequency. This may prove fatal under certain loadingconditions and influence the source of supply in the1-ollowing way:1 Ferro-resonance effectsAn L-C circuit is more prone to ferro-rcsonanceeffects during voltage fluctuations as a result ofsaturation of the iron core. which may be of atransformer or an inductor coil. On saturation,the inductance reduces drastically and becomesmore prone to resonance with the capacitance ofthe circuit. Voltage fluctuations rnay occur due toswitching operations, particularly of an unloadedline or load fluctuations (Section 20.2.1(2)).Although the line impedance will provide as ii ffi e i en t d a m pe n i n g effect to automatic a I 1 yattenuate such a state. precautions are mandatoryto avert the same. An inductive compensation, ofthe order of40-50%. may be adequate Lo improvethe system parameters and also avert a lerroresonanceeffect. A more realistic approach to theproblem is possible if a mathematical model ofthe ferro-resonance is developed and supplementedby experimentals to pr~dti~e data to dosign ;inappropriate series coinpen\ation schcine.2 Sub-synchronous resonance (SSR)A series compenuted network \vi11 h;iie its natui-ulfrequency expressed b!The abo\e ciin also be eripic\\ed h\.,where,fi, = natural frequcnc) of thc writ\ circuit,f'= nominal frequency 01' the \y+mL = natural reactance of thc linc pcr I)h;isc.including that of gcncrator and loadX,. = 27t ..f . 1-Cs = serie\ c2ipacit;ince pet' phawIx,, = __-?E I c\The frequency.,fl,. will occur I'or only ;I f e cqcles ~during an abrupt change in the line pai-anietcr\.such as during a switching operation 01- occiii-reiiccof a fault etc. (Section 17.6.3). To ciisurc that thecircuit remains iiiductive Lindcr all condition.; ofload variation5 and Fault. to ;I\ oitl a capacitiwmode of operation and an crices\i\c chargingvoltage the content of X,, mu\t renuin highcr IhanXcs(XL > Xes). The natural frequcnc);.,f;,. thcrcl'orehas to be necessaril) lo\\ei. than the po\\crfrequency ofthe system (/i, X, and mayhave far- I-eac h i ii g i in p I i c at ions . [ 11 t tic \ u b-synchronous range of ;I steam ttirhine generatorthis frequency may causc a i.e\oniince with therotating masses of the turbogenerator rotor andgenerate electromechanical oscillations in the rotor.This may assume serious \ignificancc in largcgenerators which have ;I numbcr. of naturalmechanical frequencies of the I-otating ina\\u\. inthe range of 10-25 Hr. 'This trequcncy inaycoincide with the natiiral frcquency ol'the \y\temduring a line disturbance and niagnifj thcoscillations of the rotating maws beyond de\irablelimits. If unchecked. these oscilla[ions maycontinue to magnify and result in the shcaring offof the weakest part of the \haft. Although rare.serious dainage has occulred ill edici- year\. whenthe turbine shaft had actually rhearcd off becauseof this phenomenon. Hydroturbine generators art'less prone to such oscillations. 3% theii- naturalmechanical frequency lies below IO HL. The naturalfrequency of a serie.; compensated system ma)not rcach this level.These oscillations can he dwnpcned M ith theuse of filter circuits or by bypassing all or pari 01'the series compensation during ii line disttirbancc.Similar techniques are adopted whilc protecting

24/782 Industrial Power Engineering and Applications HandbookI, (fault)iving E, end@ Series reactor to add toline impedancelC (fault) = I, (fault)An R-L dampening circuit0) Isolators0@ Dampening resistor@ Auxiliary saturating reactor@ Main saturating discharge reactorFigure 24.4Dampening circuit across the series capacitors to limit the fault levelthe series capacitors against fault conditions, asnoted in Section 26.1.2(ii) and illustrated inFigure 24.4 (Figure 26.10.2(ii) redrawn). Forcritical installa-tions it is essential to first evaluatethe likely frequencies of the rotating masses andthen more exacting measures must be taken toavoid a resonance.If a large induction motor is switched on sucha system it is possible that its rotor may lock upat the sub-synchronous speed and keep runningat higher slips. This situation is also undesirable,as it would cause higher slip losses in addition tohigher stator current and overvoltage across theseries capacitors.3 System fault levelSince the line impedance, R + J (X, - Xc), willreduce with a series compensation, the fault levelof the system will rise. It should not matter if thefault level of the system is determined by theimpedance of the source of supply, ignoring anyother impedance of the circuit (Section 13.4.1(5)).Moreover, such a situation is automatically avertedthrough the protection of the series capacitors, asdiscussed below, by which the capacitors are bypassedduring a line fault, the line restoring itsoriginal impedance, hence the original fault level.Nevertheless, when it is required to limit the systemfault level, inductive coupling circuits may beprovided to reduce the fault to the desired level.This is also discussed below:The fault current can be limited by providing adampening circuit, such as a short-circuit-limitinginductive coupling, across the series capacitors,as illusti-ated in Figure 24.4. This can be acombination of an R-L circuit. During normaloperation this circuit will provide a high impedanceand remain immune. On a fault, the high voltageacross thc capacitors will cause a heavy inductivecurrent flow through the coupling circuit, whichwill neutralize the capacitive current through thecapacitors and help keep the capacitors almostout of circuit (Zc fault = I, fault), similar tu ashorting switch, as discussed later. The normalcondition is restored as soon as the fault conditionis cleared. It may also be regarded as a filter circuit,as it would also help to dampen the systemharmonics.24.5 Analysis of a system for seriescompensationConsider a simple system as shown in Figure 24.S(a).1 When the line resistance, R, is significant comparedto the line inductive reactance, X,, there will be a(a) Circuit diagram of an uncompensated line.(b) When 'R' is significant(c) When 'R' is insignificant! CV, LoadI, . RFigure 24.5 Receiving-end voltage phasor diagram on load,in an uncompensated lineV,E,

2limited use of series capacitors (Xcc) in view of alarge content of I.R. See the phasor diagram in Figure24.5(b). Also refer to Figure 24.5(c) when R isinsignificant. Now the receiving-end voltage, V,, canbe improved by offsetting the reactive componentwith the use of series capacitors.When R

241784 Industrial Power Engineering and Applications HandbookGeneratortransformer(Figure 13.21)\ -\Lumped leakage-G CaDacitances(a) Distributed line leakage capacitances(b)Distributed line parameters-Sending Line lengthendFigure 24.8 Charging current profile on no load in atransmission lineReceivingendcularly in an HT power system. The design of a powertransmission or distribution system is a different subjectand is beyond the scope of this book. Reference may bemade to work by many authors, some of which areprovided in the further reading at the end of the chapter.Some developing countries, where reactive powermanagement practices have not been predominant, forwhatever reason, consideration of cost being one, maysuffer froin fluctuating voltages, flickering lights, highline losses, reduced capacity of the power Lines and theirconsequent overloadings etc, leading to frequent trippingsand breakdowns. The active load current (I, cos 4) becomestoo low at low p.f.s. Frequent and wide voltage fluctuationslead to fusing of bulbs and failure of fluorescent lights,besides requiring a voltage stabilizer with each householdappliance. such as TVs, air conditioners, refrigerators,ovens and computers.All this increases the direct cost of the appliances, onthe one hand, and is an additional burden on the alreadyoverloaded lines, on the other, by permanent losses ofsuch voltage boosters, which further erode the p.f. of thes y s tein .(A)UG1-(c) Equivalent circuit diagram on no loadFigure 24.7 Representing an open circuited transmission linewithout compensationof the line is open circuited. Figure 24.8 describes anormal current profile of such charging currents.Capacitors play a vital role in the management ofreactive control in power transmission and distributionsystems. With industrial growth and growing demandfor power for public services, utilities and consumer needs,efficient reactive power management is highly desirable.In the following text, we broadly consider the purposeand applicatioii of reactive power management, parti-24.6.1 ObjectivesThe basic objectives of a reactive power managementsystem can be identified by the following for LT and HTdistributions:1 Load balancing and reduction in negative phasesequence currents. In Section 16.12 we discussed loadbalancing of two generators where more emphasis isplaced on the active control of power through speedcontrol of the prime movers rather than its field(reactive) control. Active load balancing is morcappropriate for the optimum utilization of a machinerather than reactive control. But in the case of a powertransmission or distribution network it is the optimumutilization of the available active power throughefficient reactive power management that is morerelevant.A load unbalance is a common feature in a powersystem, and can be the result of one or more of thefollowing:A higher neutral current due to unequally distributedsingle-phase loads.

~ Induction~ Large~ AndSystem voltage regulation 241785Saturation of power transformers as a result ofperiodic overloading and load rejections.Increased ripples in the rectifier circuits. causingharmonics.Malfunctioning of some equipment, possibly becauseof a fault.Oscillating torque in the rotating machines as aresult of load variations and harmonics present inthe system.Feeding non-linear loads such as:furnaces- Arc furnaces and arc welders- Steel rolling millsmotors with periodic loading- Thyristor drives- Railway traction which is mostly through d.c.drivesmany loads which may have to be frequentlyswitchedAll such loads generate harmonics and cause variationsin the fundamental power frequency of the supplyqystem which leads to distortion in the sinusoidalwaveform of the voltage. This distortion may affectthe quality of the supply system (voltage) beyonddesirable limits. A non-sinusoidal and distorted supplysystem may adversely affect the different loadsconnected on the system, besides leading to outage ofthe system itself.Maintaining a near-unity p.f.Maintaining the frequency to near constant by suppresingthe system harmonics.Maintaining the receiving-end voltage at almost therated voltage.(B) Transmission of power1 Enhancing the steady-state power transfer capabilityof the lines over long distances. or making short linescapable of transferring larger powers.2 Improving the stability of the system by supportingthe voltage at key points. Without compensation, thestability of the system becomes a limiting factor evenfor shorter line lengths and the system is renderedprone to frequent outages on small disturbances.3 Reducing system oscillations and flickering causedby voltage fluctuations and system harmonics as aresult of frequent and rapid changes in reactive powerdemand, loss of load, loss of generation or a systemfault. Excessive voltage swings may cause tripping ofindustrial drives and even system outages. High-speedSVCs (Static VAr compensators, Section 24.10(2))can overcome such situations by providing appropriatereactive support during system disturbances andmaintaining a near-flat voltage profile through thelength of the transmission line.3 Providing voltagc support when switching large loads.5 Improving voltage regulation: both overvoltages (OVs)and undervoltages (UVs) are undesirable. An OV maycause ageing ol‘ the equipment’s insulation and canlead to a flashover or eventual breakdown of theterminal equipment and line insulators. It may alsolead to saturation of power transformers operating inthe system. The transformers produce high currents,rich in harmonics, and cause ferro-resonance or subsynchronousresonance. An UV will result in highersystem loading than necessary and cause underutilizationof the system capacity.In the following we consider the case of a transmissionline, 132 kV and above, being more typical and complexfor the purpose of reactive control. Based on this, it wouldbe easier to apply appropriate reactive control to adistribution network and large inductive loads such as anarc or induction furnace.24.6.2 Analysis of an uncompensatedtransmission lineCurrent profileA transmission line can be represented, as shown in Figure24.7(b). In Tables 24.1 (a) and (b), we show typical lineparameters for different system voltages and lineconfigurations. Because of line charging capacitances,C,’s, between conductors and conductors and ground, asshown in Figure 24.7(a) (higher significance in HV andEHV lines of 132 kV and above), and series inductanceLo, there is a charging current, Io, even at no load andeven when the far end of the line is open-circuited (Figure24.7(c)). Figure 24.8 describes a normal profile for suchcharging currents. This current rises with the rise in linelength and is highest at the generator end. As thisphenomenon a function of system voltage it is negligibleor nil in HT lines up to 66 kV. This current is totallycapacitive, ignoring the effect of line resistance. Ro,. Atransmission line, being a high power transfer system.has a very low content of Ro (Table 24.1 (a)).The magnitude of the charging current, Io, will dependupon the content of C, which is a measure of line voltage,size of the conductor, spacing between the conductorsand between the conductors and the ground etc. Table24.2 provides the approximate values of charging current,Io, and charging power for a few system voltages withdifferent line configurations. The generated chargingreactive power, by the line charging capacitances (C’s).flows back to the generating source and has to be absorbedby it, even on no load, or a part of it during light loads.It is a strain on the field windings of the generator. as themachine under no load is underexcited. andunderexcitation is not a healthy situation for a thermalturbogenerator because,A capacitive circuit magnifies the harmonic effectswhen present in the system, as discussed in Section23.5.2, and gives rise to spurious voltages and currents,raising the normal VI and I, to V, and I,,, respectively.The stator windings are subject to overcapacitivevoltages as a result of this, and the end turns particularlyare endangered.Reduced field current reduces the voltage generated,which may affect the system’s stability.The generator manufacturer can define the lowest

~~~~~24/786 Industrial Power Engineering and Applications HandbookTable 24.l(a) Typical line parameters per circuit for HV and EHV transmission linesNominalvoltage, V,kV(zm.s.)765400400400400400220Conductortype1Quad Bersimib(QB)Twin Moose(TM)Twin AAAC(TA)Quad Zebra(QZ)Quad AAAC(QA)Triple Zebra(TZ)ZebraPositive sequence componentsRoWkm1.142 x2.979 x IO-*3.094 x1.68 x1.566 x2.242 x7.487 xXLOn/km2.619 x IO-'3.32 x IO-'3.304 x IO-'2.544 x IO-'2.682 x IO-'2.992 x IO-'3.992 x IO-'XCOWkm2.44 x 10'2.88 x IO52.82 x IO52.40 x IO52.29 x 1052.74 x IO'3.408 x io5Zo = J zzGR252.8309.22305.24247.09247.826286.32368.846Zero wquence components__RnXLn XC,n/km Wkm GWcm__2.633 x IO-' 1.053 4.161 x lo51.619 x IO-' 1.241.682 x lo-' 1.2379.133 0.9508.512 1.00212.186 1.1122.199 x IO-' 1.3394.46 x io54.37 x 10'3.73 x 1053.55 x io54.23 x 1055.421 x IO51321.622 x IO-'3.861 x IO-'3.416 x IO'363. I694.056 x IO-' 1.62226___-.____Table 24.l(b)Nominal Conductorvoltage, V, TvpeLine inductanceLine capacitanceWavelengthkV(zm.s.)765 QB400 TM400 TA400 QZ400400QATZ220 Z132 PXL"GWcm2.619 x IO-'3.32 x IO-'3.304 x IO-'2.544 x IO-'2.682 x IO-'2.992 x lo-'3.992 x IO-'3.861 x IO-'x LOL,, = -2n. fhenry (H)____8.33 x IO410.56 x IO410.51 x IO18.09 x IO48.53 x IO49.52 x IO412.70 x 10"12.28 x 10"u=- 1xc,Wkm__--2.44 x 1052.88 x Io"2.82 x io52.40 x IO52.29 x 10'2.74 x IO'1Ci, =27r. f . x,,n farad (nF)'13.0411.0511.2813.2613.8911.61kds3.034 X IO52.927 x 10'2.904 x io53.053 x IO52.905 x Io"3.008 x Io"krn6.07 x IO?5.85 x 10'5.81 x 10'6.11 xi035.81 x IO'6.02 x 10'3.408 x io5 9.34 2.904 X 10' 5.81 X IO33.416 x 105 9.3 1 , 2.958 x IO5 5.92 x10'* I nF = IO-' FNote: The line parameters will vary with system voltage, configuration of line conductors and their spacing between themselves and theground, tower configuration, etc.Based on Manual on Transmission Planning Criteria, CEA (Central Electricity Authority)excitation level below which the machine may beunstable.Figure 24.9 shows a typical output characteristic orreactive capability curve of a generator, illustrating thestability levels of the machine under different conditionsof operation. The machine must operate within theselevels and the voltage profile within the specified voltagelimits, as noted in Table 24.3.Example 24.1Consider a 400 kV, triple-Zebra line, having a distributedleakage capacitive reactance Xc0 of 2.74 x lo5 Wkm fromTable 24.l(b). Then the charging power per phase per km,":fco = -x co400Pco =2.74 x 105= 0.584 MVArIkm(iii) Voltage profileSince the charging current is capacitive in nature, theline voltage drop at the far end would raise the terminalvoltage on a no-load as shown in Figure 24.10. Thecharging current and rise in terminal voltage both at noloador during an underloading condition are undesirable.

Wkrn~ -~~~ ~~~Ferrtrnti~ in~ 10~ 1.036~~~~Table 24.2 Typical basic line characteristics (approximate)LZkm ~3 ! 4I25, ~____~-MW6I, (It Pi,--17 ZnAmp7Clia r,q ing ‘Churgitzg~e8ec.t per km252.82315 0 1,747 I812.40x IO-’ 59 33 x 10-3309.22SI7 4747 0 800.56 1.074 x lo-’ 61 51 x IO ’3.304 x IO-’ 2.82 x 10’ 305.24524 2757 0 820.567 1.082 x 1 of 61 97 x IO-’2.544 x IO-’ 1 2.40 x 10’ 247.09647 5935 0 960.67I .03 x I 0-1 58.99 x 10645 6932IO10.70I .082 x 10-3 61 97 x IO-’220 2 3 992 x 10 I 1 3 408 x 10’ 368 846558 9807 0 840.584 1.045 x 10-3 59 85 x 10131 2344 0 310.142 1082 xIO-’ 61.97 x 10 ’48 0210 0 220 051 1 063 x 10 60.88 x 10 ’__~- ~~ _____Tharging reactive power is proportional to V,’ and diminishes sharply at lower voltages. Charging reactive power mut be \uitably offset at the point of power generation to protect thegenerator and other connected eauipment from excesive overvoltagcs due to capacitive charging.This reveals that on a 50 Hz system the pha\e displacement between the sending and receiving end voltagca, due to electromaynctic propagation through the lines, is of the order of 6” per100 km. Accordingly, an uncompensated line should be restricted to about 400 kin in length in radial lines and 800 km in symmetrical lines to avoid an overvoltage at the receiving end (Table24.5)Atrip8-5x< 09rcrdintI.vIn degree511

_241788 Industrial Power Engineering and Applications Handbook90% generatorstability limit 7Prime movert I0 I I 0.901 .o 0.8 0.6 0.4 0.2 0 0.2 0.4 0.6 0.8 1 .oPer unitPer unitReactive power -+- Reactive power(condenser mode)(generator mode)Leading mode -+-- Lagging mode(field underexcited)(field overexcited)Figure 24.9Normal characteristics of a generator, illustrating the stability levels (safe operating limits)Table 24.3 Permissible voltage variations during temporarysystem disturbancesNominal voltage Rated maximuni Rated minimumstability level stabiliv levelV&V(r.m.s.).~ ~~kV(r.m.s.)" kV(r.m.s. Jb_ _ _ ~765 800 728400 420 380220 245 198132 145 122"As in IEC 60694'As in Manual on Trunsnzission Planning Criteria, CEA (CentralElectricity Authority).4 'coWhile the former would stress the generator windings,the latter may cause a voltage swing during a load rejectionor load fluctuation and result in a line outage. Thesefeatures, if not controlled, may render the system unstable.Overvoltages must therefore be controlled within anacceptable limit. Table 24.3 prcscribes one such limit.To achieve the above, the charging power must becompensated at the generating end itself, and this can beachieved through a reactive power control as illustratedin Figure 24.11. Figure 24.12 illustratcs general voltageand charging current profiles before a compensationand Figure 24.13 a desirable flat voltage profile that canbe achieved through a series compensation, discussedlater.W X f bl lxcoi 11IC0 tv,IIFigure 24.1 0 Receiving-end voltage rises on no-loadFigure 24.11 An open circuited series compensatedtransmission line

System voltage regulation 241789phase voltage at the receiving end in radial linesand mid-point voltage in symmetrical lines.(Symmetrical lines are those which are fed fromboth ends such as when the far end is connectedto a power grid).surge impedance of the lineSending Length of line __c Receivingendend@- V, profile without compensation@- V, profile with the use of series capacitors.0- I,, charging current profile.Figure 24.12 Voltage and current profiles when the line atthe far end is open-circuitedtIC0xccFigure 24.13 Receiving-end voltage is flattened with the useof series caDacitors24.6.3 Power transferIt has been established that the active power transferthrough a power system can be expressed by(24.3)The expression is free from p.f. A change in p.f. (cos #),however, will adjust the torque angle, 6. The higher thep.f. (low @), the, greater will be the torque angle 6 andvice versa. For more details to arrive at the above derivationrefer to the further reading at the end of the chapter.Assumption - that the line is lossless, i.e. Ro isnegligible.whereP = power transfer from one end of the line to thereceiving end per phaseE, = phase voltage at the sending endconstant for a particular line, irrespective of itslength, although L and C will rise with the linelength (L = e . L, and C = 4: . Co. if t is the lengthof the line).where Lo, C,, and R, are the line parameters, per phaseper unit length. In our subsequent analysis, we haveignored Ro, being negligible. The standard line parametersare normally worked out for different system networksoperative in a country by the power transmission anddistribution authorities on the basis of conductorconfiguration, spacing between them and the ground.Typical data for a few voltage systems have been providedin Tables 24.l(a) and (b).sin 8 = line length effect or Ferranti effect, discussedlater8 = In radial lines this is determined for the entirelength of line while for symmetrical lines. it iscalculated up to the mid-point. i.e. it refers to812.6 = load angle or transmission angle. This is the torqueangle between the receiving-end and transmittingendvoltages, and is responsible for the requiredpower transfer from the transmitting end to thereceiving-end.NoteThis should not be confused with AB as used in Section 16. I O inconnection with the paralleling of two generators. There it representedthe electrical shift between the rotors of the two machines or supplybuses. If it is not eliminated, it will cause a circulating currentbetween the two machines or the buces when running in paralleland will add to their heating.For short lines say, 200-300 km, for a 50 IIz systemsin 8 -- 8 (in radians)e.g. for a 250 km, 400 kV line, as considered earlier.from Table 24.2, for line type TZ,8 = 59.85 x IO-' degree/km = 14.96' for 250 kmand sin 8 = 0.258In radians 8 = L?L . 8 = 1.045 x per km or 0.261180for 250 km (both are almost the same)and 8 = E.aas in equation (24.6)Then from equation (24.8) noted later.

~24/790 Industrial Power Engineering and Applications Handbook= L0.2Zf. 1= X,, i.e. the inductive reactance of the entireline length.and equation (24.3) will becomeE, E,p=-. sin 6XLFor short lines, this is a very useful formula.24.7 Influence of line length(Ferranti effect)(24.4)Figure 24.15 The line length effect even when the sendingendand receiving-end voltages and currents are maintainedat unity p.f.E,The velocity of propagation of electromagnetic wavesand the line length have a great influence over the capacityof power transfer through a line under stable conditionsand also define the quality of the receiving-end voltage.The electromagnetic waves (electricity) travel with greatspeed, close to the speed of light (Section 17.6.6) andhence have a very long wavelength. SinceUwhere= wavelength in kmu=- 1JL,c,(24.5)= Velocity of propagation of electromagnetic waves-- 3 x IOs km/s (more accurate values are determinedin Table 24.1 (b) for the line parameters considered).The normal line lengths may vary from 200 km to 500km. As a result, the electromagnetic wave is able to travelscarcely a small fraction of its one full wavelength, up tothe far end of the line (Figure 24.14). The instantaneousvoltage at the receiving-end therefore is never in phasewith the voltage at the sending-end (Figure 24.15). Thisphase displacement, which is caused neither by the p.f.Receiving end voltageE, rises with 6'= (E,),,, sin wt= sin 8Sending e ni/L e I \,voltage E,(e= 0)400 kmE, cos 8UI:Q-A = 4.872 x lo3 krn2n radians or 360"Illustrating one wavelengthFigure 24.14 Phasor position of sending-end and receivingendvoltages in an overhead linenor by the mechanical positioning of the rotor of thegenerator, or the bus to which the receiving end may beconnected, is termed the Ferranti effect. It constrains theline lcngth within certain limits to transmit power understable conditions, as discussed later.This phase shift (6) for a particular line length can becalculated as follows:(24.6)where8 = phase shift between the transmitting-end and thereceiving-end voltages, in radians or degrees,depending upon the value of z considered, i.e.z 22= - or 180" respectively.7e = line length in km. For the various HV and EHVnetworks and their line parameters considered, 6 iscalculated in Table 24.2and the voltage at the receiving-end, when it is opencircuited,E, cos 6 = E, (Figure 24.15)Or E, =E,-(24.7)cos 6For the 400 kV, TZ line considered above, Er, for a400 km line length,6 = 400 x 59.85 x lo-' = 23.94".:. E, =E,cos 23.94"= 1.094 E,The Ferranti effect therefore, raises the receiving-endvoltage and becomes a potential cause of increased voltagefluctuations when existing in the system, similar to acapacitor magnifying the harmonic quantities. The longerthe line, the higher will be the voltage rise at the receivingend.This will cause wider voltage fluctuations duringload variations, particularly during light loads and loadrejections. Beyond a certain line length, this effect mayeven render the line unsuitable for the safe transmissionof power. For very short lines, however, the effect maybe negligible and may be ignored. The line length is

~ sin.System voltage regulation 24/791therefore chosen so that the receiving-end voltage ismaintained within the permissible limits under allconditions of its far-end loading. Thus the receiving-endvoltage is influenced by three factors:The line distributed parameters Lo and CoThe Ferranti effect due to Co andThe p.f. of the far-end load.The effect of p.f. can be controlled by shunt capacitors,near the load point and the Ferranti effect by altering theline parameters. Since2n@=-.ea+--I Tv, = E, = E$e12 +e12 ___jFigure 24.16(a) Series compensation by sectioning at themidpoint of the line200 -I6Stable= 2 ~ f m e .(24.8)whereL = Lo . e andc=c,.e$100NoteGenerally, an HT distribution network has a very short length (e),less than 10-15 km. Moreover, the leakage capacitance (C,,) forsystem voltages up to 66 kV is almost negligible. The Ferrantieffect is therefore not applicable to a distribution network.For the same system frequency, the Ferranti effect canbe reduced bySectioning or sin 6 effect: Line length compensationcan be achieved by sectioning, i.e. by dividing the lineinto two or more sections. This method indirectlyreduces the physical length of line (e). Each sectionnow operates as an independent line and is compensatedthrough shunt capacitors controlling the voltage withinthe required limits, at all such sections sincep = E .E u . sin62. sin 8Maximum power transfer is possible when 6= 90". Ifthe line is compensated, say, at the midpoint, as shownin Figure 24.16(a), then the maximum power transferwill improve toP, =E, . E,. sin68z,.-LLwhere P, = compensated power transfer at the midpoint.E, = E, = V,, which is the mid point voltageand is held constantZ,,0, and 6, are midpoint parameters andz, = 212e , = et20 45" 90" 180"Load angle (6)-Figure 24.16(b) Rise in power transfer with mid-pointcompensation6, = 612:. P, = ~2. v,' 6sinZ. sin,L= 2 ' P , ' sin 6Maximum power is doubled by a midpointcompensation and occurs at 6 =1 80", as shown in Figure24.16(b). Thus by changing the location of the shuntcompensation the utilization capacity of the line canbe altered. For a midpoint compensation, the line canoperate stably up to 612 or so, i.e. at about 90".This is a costly and cumbersome solution, and maybe resorted to where series compensation is not possible.But such a situation will seldom arise. Power is rarelytransported over very long distances and through radiallines. A transmission line is normally symmetrical, asits far end will generally be connected to a power gridat less than 1000 km or so. However, if such a situationarises, sectioning would be one viable solution.Reducing the electrical line length by reducing theproduct 2/Lc (equation (24.8)).

~24/792 Industrial Power Engineering and Applications HandbookTable 24.4 Far-end voltage, due to the Ferranti effect, in a400 kV TZ type line, at different line lengths1C= 2n. f. x,This product can be reduced by reducing XL, usingseries capacitance with a reactance X;, which willreduce X, to X, - X; . This is where reactive controlplays a major role. By meticulous reactive powermanage-ment, the Ferranti effect can be controlledand the electrical line length increaced to the deciredlevel. It is a different matter that the electrical lengthof the line cannot be raised infinitely, for reasons ofstability, as discussed later.To apply the corrective measures to limit the Ferrantieffect it is essential to first study its over voltage (OV)status at the far end of the line. Consider the earliersystem TZ of 400 kV 50 Hz and draw a voltage profileas illustrated in Figure 24.17, for the voltages workedout as in equation (24.7), at different lengths of theline. The voltages, for the sake of simplicity, are alsoshown in Table 24.4.From the voltage profile it is evident that up to aline length of almost 250 km the overvoltage at the farend is quite acceptable. For greater lengths than this,the far-end open-circuit voltage will rise beyondacceptable limits and may damage the line insulatorsand the terminal equipment. Moreover, during a linedisturbance or load variation this voltage fluctuationmay assume more dangerous swings. Generally, atransmission line is connected through a power gridwhere more than one supply source may be feedingthe system. When this is so, lines are called symmetricalas they are fed equally from both ends. The far endpoint shifts automatically to the middle of the line,diminishing the Ferranti effect, doubling the electricalline length. (Figure 24.18). In other words, such linesLine 0 from cos 0 v, = - E'km equutron (24 6)and Table 24 2Lo) e100 5 985" 0 995 100 5200 11 97" 0 978 102 2250 14 96" 0 966 103 5300 17 955" 0 951 105 1400 23 94" 0 914 109 4in 9i ofE,can automatically transmit power, within permissibleparameters, up to twice the length of a radial line,which is fed from only one end. In such cases, it isonly the midpoint voltage that is more relevant andmust be considered for the purpose of Ferranti effect.In the first case, if we had considered a safe linelength of 250 km, this would become 500 km for asymmetrical line. Figure 24.18 illustrates such acondition. Depending upon the length and type of line,a line length compensation may be required. Mosttransmission lines are seen to be within permissiblelengths and only a few may require such a compensation.Nevertheless, it may be worth reducing the phasedisplacement between E, and E, to less than 15"electrical, to further improve the quality and stabilitylevel of power transmission.24.8 Optimizing power transferthrough reactive controlTo transmit power over long distances is the basic requisiteof economical transmission. Let us study equation (24.3).If we are able to maintain a unity p.f. between thetransmitting and receiving ends, then for a lossless lineE, = E, = Vo110108t 106$? 104-1000 50 100 150 200 250 300 350 4000Line length in krnFigure 24.17 Voltage profile of a 400 kV/400 km radial lineon a no-load illustrating the Ferranti effect50 100 150 200 250 300 350 400Line length in krn ---+Figure 24.18 Voltage profile of a 400 kV/400 km symmetricalline on a no-load illustrating the Ferranti effect

System voltage regulation 24i793Under such a condition, the line will maintain a unityp.f. at all points of the line and the reactive powergenerated, due to the distributed line charging capacitances(Co), is offset by the reactive power absorbed by thedistributed line inductances (Lo). The generator is nownot unduly stressed by the reactive power feedback, i.e.-- "Z -I; .XL0xcowhere reactive power generated = V~/Xco per phase perunit length and reactive power compensated (absorbed)= Zi . XLo per phase per unit length. Io is the capacitivecharging currentor %=,/-IO11.93; I4/ -------15 30 45 60 75 90Load ang/e (6)--+a radiansPIPo, considered for a 250 km radial line length as per Table 24.5.IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII@ Stable regionStability level30 1.935F[ 2.740 1) @ Stable when series compensatedl3.7403.870Not so stable on severe line@ disturbances, even after aseries compensationFigure 24.19 Variation in load transfer with change intransmission angle 6= 45 = Zo (Zo is termed the natural orsurge impedance of the line)The voltage will now maintain a flat profile from thetransmitting end through the receiving end and all theinsulators or terminal equipment would be equally stressed.If Vo is considered as the nominal phase voltage of thesystem then equation (24.3) can be rewritten asvc?P = -. per phase(24.9)Zo sin 8The concept behind the above equation is that the voltagesand the currents, at the transmitting and receiving endsare maintained at the same p.f. The voltage at the receivingend, however, will shift in phase with respect to thevoltage at the transmitting end by an angle 8, due to theFerranti effect and that effect is considered in the abovederivation. Refer to Figure 24.15 for more clarity. In theabove equation the element V,/Zo is an importantindicator of the power transfer capability of a line, and istermed the natural loading or surge impedance loading(Po) of the line, i.e.P - vc? per phase (24.10)O - zoSuch a line is said to be naturally loaded and thisassumption is true only when the power is beingtransmitted at unity p.f. and there is a total balancing ofreactive powers. Since Zo is constant for anuncompensated line, so is Po, irrespective of its length.The magnitude of this will depend upon the line voltage,size of conductors and the spacings between them andfrom the ground (these parameters decide Co and LO andhence Zo). It is also an indicator of a normal loadingcapacity of a line. The recommended practice is to loadan uncompensated line to near this value or a little abovewhen the line is a little shorter, or a little less when theline is longer to retain the level of stability. Also refer tothe load curves in Figure 24.20 for more clarity.To optimize this power transfer through reactive controllet us study equation (24.10) for the parameters that canbe varied to achieve this objective. The active powertransfer will depend upon the following factors:Nominal voltage of transmission (Vo) .is a policydecision of a country, depending upon the likely powerloading of such lines and future power plans. Generally,the levels of voltage, V,, for primary and secondarytransmissions are gradually increasing to cope withgrowing power demands. A typical system of transmissionand distribution is illustrated in Figure 23.1.Load to be transferred, keeping suitable margins for afuture increase in demand.Likely expected load variations and p.f. of the load(which may be based on experience).

24/794 Industrial Power Engineering and Applications Handbook@ Uncompensated@ Partially compensated@ Fully compensatedFigure 24.20-INatural load (PIP,)Capacity utilization load curves with and without compensationLine length effect or Ferranti effect, sin 8, that willdetermine the optimum line length which will alsodepend upon whether it is a radial or a symmetricalline.Surge impedance of the line, Z,.Angle of transmission, 6.Equation (24.3) defines the active power as independentof p.f. However, depending upon the p.f. of the load, thiswill adjust the load angle 6. The larger the angle oftransmission, the higher will be the power transfer. Figure24.19 illustrates the power transfer characteristics of a250 km line selected from Table 24.5.Rewriting equation (24.9),P = -.sinPo6(24.1 1)sin 8For the system to remain stable under all conditions ofloading, switching, or any other line disturbances it isessential that an uncompensated line is loaded at muchbelow this level. Otherwise disturbances of a minor naturemay result in undampened oscillations, and may evenswing the receiving-end voltage beyond acceptable limits.It may even cause an outage of the system. It is thereforenot practicable to operate an uncompensated line to itsoptimum level. For this we will analyse this equation forsin 0 and sin 6 as follows.24.8.1 Line length effect (sin e)The element Polsin 0 can be considered as the steadystatestability limit of the line, say P,,,. A line lengthcompensation can improve the voltage profile and hencethe power transfer capability of the line as follows.Figure 24.20 illustrates three power transfer or loadcurves:Curve 1 : without any compensation, the voltage profilesags on small load variations and is not capable oftransferring even a natural load.Curve 2: with partial compensation, the voltage profileimproves and the line is able to transfer more load thanabove, but less than its natural loading. Voltage still sagsbut the swing is more tolerable.Curve 3: The line is fully compensated. The voltage profiletends to be flat and the line is capable of transferringeven morc than the natural load without an appreciablesag in the voltage profile.24.8.2 Influence of load angle (sin 8)A study of various systems has revealed that the loadangle for an uncompensated line should be maintainedat about 30" only. This means that an uncompensatedline may be loaded to just nearly half its steady-statelevel to retain a high level of stability during loadfluctuations, particularly during light loads or loadrejections, switching of large inductive loads or any typeof minor or major line fault.When the line is compensated, and a near-flat voltageprofile can be ensured so that during all such disturbancesthe receiving-end voltage will stay within permissiblelimits, the load angle can be raised to 45-60" 10 achievea high power transfer.Of all the above parameters, system voltage is alreadypredefined and considering that it cannot be changed,the only parameters that can be altered to optimize P areZ, and 0. Both parameters can be altered to any desiredlimit with the application of reactive power controls,subject toThe thermal capacity of the line conductors andRetaining the stability limit of the system thus modified.

System voltage regulation 241795After we have assessed the optimum power level itbecomes easy to decide the type and amount of reactivepowcr control required to achieve this level, assumingthat the lines can be loaded up to their thermal capacityand the optimum power derived above can be attained.Our main objective will now be to arrive at the stabilitylevel of the system and the parameters that define this.As noted above, the stability level defines the maximumpower that can be transferred through a line withoutcausing a voltage tluctuation and angular differencebeyond acceptable limits, or a consequent outage of theline. during a load variation, or a temporary linedisturbance. It should, in fact, maintain its continuityeven during a fast clearing of a major fault. To determinethe effects of Zo and 6 on the receiving-end voltage andconsequently the transfer of power, P, within stable limitswe will study the voltage equation of a losslesstransmission line (considering K, = 0, for an easyillustration), feeding a load P at a p-f. cos 4.Radial linesThe transmitting-end voltage in terms of line parameterscan be represented byE, = V,. cos 0,. + JZ,, . I, sin 0, (24.12)whereE, = phase voltage at the transmitting-endV, = phase voltage at the receiving-end0, = line length effect or Ferranti effect at the end of theline. in degreesI, = load current- P- JQ-L’,P = active loadQ = reactive loadK,, = line resistance per phase. It has been ignored andthe line is considered losslessZ,, = surge impedance of the lineThe voltage stability of a system is the measure ofvoltage fluctuations which must remain within permissiblelimits during load fluctuation or rejection or other linedisturbances and even temporary faults. We may thereforesolve the above equation for V, and P, to study thebehaviour of the system under varying load conditions,P. As there are two more variables, load p.f. and the linelength, which will influence P and V,, different sets ofload curves can be drawn as illustrated in Figure 24.2 1,for different line lengths at different p.f.s (at near unity,to obtain the best performance). From a study of thesecurbes one can identify the most appropriate linelengths which can extend the highest level of stabilityto the system. For example, set ‘a’ of curves are moreideal compared to set ‘b’, which correspond to verylong line lengths, compared to the ideal line lengths ofset ‘a’.After identifying the likely line lengths we can thenstudy the most appropriate p.f. at which the load must betransmitted to maintain the highest level of stability. Forour purpose, parts of the curves that lie near the ratedvoltage, say, within V, f 5%. alone are relevant for study.The line will perform best at p.f.s very near unity andcause the least possible voltage fluctuations by maintaininga near-flat voltage profile over reasonable variations ofload. Leading p.f.s are not considered for reasons ofcapacitive overvoltages. The p.f. of the load can beimproved by applying shunt capacitors near the load points,as discussed above.Irxje re n ceThe voltage stability level diminishes with an increasein the line length. For very long line lengths, the far-endvoltage may swing from high to very high values duringload variations, rendering it unsuitable for operation nearthe maximum load transfer level. During light loads toothe steeply rising voltage profile may cause a high-voltageswing on a small load variation. A load variation thereforewill cause wide to very wide voltage fluctuations andrender the system unsuitable rather than unstable for apower transfer near the required level. For transfer of aload under stable conditions the line lengths of theuncompensated lines will be too short and hence will notbe economically viable. We will seek a solution to theseproblems with the help of these curves which will providean introduction to the utility of reactive power controlsto improve the power transmission capacity of a line andits quality through the following discussion.Iiifluence of PFThe power transfer capability of the line rises ai. thep.f. swings towards the leading region and diminishesas it swings towards the lagging region. Since a powersupply system is not run at leading p.f.s for reasons ofdangerous overvoltages that may develop (as a resultof overexcitation of the capacitors, Section 23.13) acrossthe terminal equipment it is advisable to run the systemas close to unity p.f. as possible. Moreover, the fieldsystem of the generating machines is also designedfor maximum operation at lagging p.f.s only, asdiscussed in Section 16.4. At leading p.f.s (after acertain limit) there is a possibility of its field systemlosing control and becoming ineffective.The receiving-end voltage rises with leading p.f.s anddroops with lagging. This is illustrated with the helpof phasor diagrams (Figures 24.22(a) and (b).At unity p.f. the voltage variation and hence theregulation is the least and maintains a near-flat voltageprofile. This is the best condition to provide the highestlevel of system stability from a voltage point of view.The power factor can be improved with the use ofshunt capacitors at the load points or at the receivingend, as discussed above. It is not practical to have anear-fixed loading for all hours of the day. Moreover.there may also be seasonal loads which may upset theparameters considered while installing the capacitor banks.In such conditions the system may therefore have to beunderutilized or run under a high risk of instability during

24/796 Industrial Power Engineering and Applications HandbookOpen end voltagehigh to very high,out of stable limitsOpen end voltagewithin permissibleLine lengthmuch longerthan ideal(Set 'b' curves)Operatingregiont\v- Diminishes (referring to onlyoperating region) and linesoperate underutilized.,t1 .oE-POFigure 24.21A comparative study of load transfers for different line lengths at different p.fs. for an uncompensated lineadverse load conditions, particularly when the p.f. fallstoo low. Such a situation can be overcome by readjustingthe reactive needs of thc line by providing switchedcapacitor banks a few of which can be switched-in orswitched-out, depending upon the load demand. Theswitching may be manual or automatic with the help ofa p.f. corrcction relay (Section 23.15). The latter is alwaysrecommended.Influence of line length (Ferrunti effect)For each p. f. and line length the curve V, versus P describesa certain trajectory. Maximum power can be transferredonly within these trajectories. Each line length has atheoretical optimum level of power transfer, P , whichis defined by Polsin 8. In Table 24.5 we have worked outthese levels for different line lengths, for the systemconsidered in Example 24.1.A line can be theoretically loaded up to these levels.But at these levels, during a load variation, the far-endvoltage may swing far beyond the desirable limits of V,k 5% and the system may not remain stable. With theuse of reactive control it is possible to transfer power atthe optimum level (PmaX) and yet maintain the far-end(or midpoint in symmetrical lines) voltage near to V, andalso to have a near-flat voltage profile.Reactive control can alter the line length (= m) tothe level at which the system will have the least possibleswings. It is evident from these curves that an uncompensatedline of a much shorter length may not be able,to transfer even its natural load (Po) successfully. This isdue to the steeply drooping characteristics of the voltageprofile at about this load point, which may subject theV, > ES (I. R ingnored)(a) Receiving-end voltage rises with leading p.f's

~~~~ ~ ~ ~8- m1Say, for an actual line length of 800 km (symmetrical),04,,,, dzz 0~ x 400 ( 8,o0 = midpoint Ferranti effect)which must be improved for, say, a 250 km radial linei.e. 8250 - d m x 250for 8400 to be almost equal to 8 ,d m x 400 = Jm x 250Since a shunt compensation will reduce Xc (- Uc), itwill not provide the desired compensation. This can beachieved with the use of series compensation, C, remainingthe same. Then:. Compensation required = 0.61 XL0 per unit length.Series capacitors making up 0.61 X,,. e may be introducedinto the system to achieve the desired electrical line length.Influence of surge impedance (Z,)Since Po -1-Z OZ, p1ay.s a very significant role in the power transfercapability of a line. By reducing the value of Z,, thepower transfer capability of a system can be increased.Sinceand in absolute terms = d-orSystem voltage regulation 24/797= JmThis value can be reduced by decreasing the value of X,which is possible by providing series capacitors in theline. If Xcc is the series compensation, then the modifiedimpedanceand hence any value of power transfer can be achievedup to the theoretical P , (Table 24.5). But for reasonsof other parameters that may also influence the stabilityof the system, it is not practical to achieve the optimumcapacity utilization of the line without sacrificing thelevel of stability, even when the required degree ofcompensation is provided. Parameters that may influencethe stability can be one or more of the following:1 A small value of (X, - Xcc), i.e. Xcc approaching XL,will have more chance of a sub-synchronous resonance(SSR) with the rotating machines and a ferro-resonancewith the transformers during a switching sequence orline disturbance.2 Higher harmonic contents may magnify the harmoniccurrents and affect the loading capacity of the line.3 A very close compensation, Le., a low XL - Xcc, mayalso raise the fault level of the system beyondacceptable limits.To overcome such situations within acceptable parametersduring normal operation, it has been found thatan ideal series compensation is achieved at around 40-70% of XL, preferably in the range of 45-60% only. Thelevel of compensation will depend upon the expectedload fluctuations and the presence of harmonic disordersin the system.Example 24.2Consider the 400 kV, 50 Hz system and apply the abovetheory. If the system has relatively fewer load fluctuationsand the loads are reasonably linear, then we can consider ahigher compensation to the extent of, say 75% of X,. Thenv,2Po = -andz oTable 24.5Line lengthRadial linekmLevel of P , for a 400 kV, 50 Hz, TZ systemSymmetrical linekmIpmax8 from equation (24.6) sin 0__=- 1V$ES = 2%Po sin f3cos e100 200200 400250 500300 600400 8005.985" 0.104 9.61 100.511.97" 0.207 4.83 102.214.96" 0.258 3.87 103.517.955" 0.308 3.25 105.123.94" 0.406 2.46 109.4Note: Normal practice is to design a system to carry at least its natural load, Po, under stable conditions.

24/798 Industrial Power Engineering and Applications Handbookor P, = Po xSince there is no change in the shunt capacitance (X; = Xc ),71 10.25:. P, = Po x -\I- = 2 PoWhile it is possible that PmaX may be further raised by a stillcloser compensation, this is not advisable to retain the stabilitylevel of the system. The above compensation is even higherthan the line length compensation considered earlier and willfurther improve the electrical line length.Adding shunt capacitors would also reduce Z, but wouldraise the electrical line length; hence it is not considered.Moreover, on EHVs, the charging shunt capacitances, C, assuch require compensation during light loads or load rejectionsto limitthevoltage rise (regulation) atthe farendorthemidpoint.Hence no additional shunt compensation is recommended.NoteSeries compensation would mean a low value of Z, andhence a higher system fault level. This needs be kept in mindwhile designing the system and selecting the switching devicesor deciding on the protective scheme or its fault setting.Symmetrical linesEquation (24.12) is now modified toE, = V, . COS 0, + JZ, . I, . sin 0, (24.13)whereV, =voltage at the midpoint of the line (Figure 24.18)0, = line length or Ferranti effect up to the midpoint ofthe line.The rest of the procedure, even the inferences drawnabove, would remain the same as for a radial line. Theonly difference now is that the system would becomesuitable for twice the lengths of the radial lines as aresult of the midpoint effect which doubles the line length.ConclusionA compensated line can transmit much more power thanits natural loading within stable limits and hence fulfilthe requirement of economical power transfer. The abovewas a theoretical analysis which can provide quite accurateresults, depending upon the accuracy of the data assumed.The more scientific procedure to conduct this type ofstudy, however, would be through a load flow analysis ofthe steady-state component to study temporaryovervoltages and transient analysis through a TNA(transient network analyser) or an EMTP (electromagnetictransient programme). TNA is an analogue method whileEMTP is a digital method of system analysis. For detailsof system models and procedure to study a system, referto Miller (1982).A transmission line may have to operate under differentconditions of loading (I, and p.f.) at different hours ofthe day, and then there may also be seasonal loads. Thetype of reactive compensation therefore must be decidedfor the varying load conditions, so that they are able toprovide a continuous change in the VAr as demanded. Itis normal practice to have a combination of series andshunt reactive compensations to suit all conditions ofloading, some fixed (unswitched) compensators for normalload conditions and the remainder variable, to switchON or OFF depending upon the load conditions or loadfluctuations. The choice of different types of reactivecompensators may be considered on the following basis:1 Shunt reactors These are provided as shown inFigure 24.23 to compensate for the distributed lumpedcapacitances, C,, on EHV networks and also to limittemporary overvoltages caused during a load rejection,followed by a ground fault or a phase fault within theprescribed steady-state voltage limits, as noted in Table24.3. They absorb reactive power to offset the chargingpower demand of EHV lines (Table 24.2, column 9).The selection of a reactor can be made on the basis ofthe duty it has to perform and the Compensationrequired. Some of the different types of reactors andtheir characteristics are described in Chapter 27.Reactors add to Z, (Z, = JX,, Xco ) and hencereduce surge impedance loading (SIL), Po. But mostare the fixed type, depending upon the maximum loadGSCR - Shunt compensating reactortIntermediatoryswitching stationGFigure 24.23A shunt compensated transmission line

conditions, and the remainder are switched. Moreover,switchable reactors are switched only during atemporary disturbance, therefore they have no adverseeffect on Po. Ideally they can be made fixed tocompensate 50-60% of C, and the remainder switchedin a few steps. The size and number of steps willdepend upon the likely underloading, open-circuitingduring offpeak periods or other temporary disturbances.The overvoltages that may occur under such conditions,and which must be controlled through these reactors,is a matter of system design practice adopted by acountry or its central power authorities and may bebroadly based on our discussions in Section 24.6 andTable 24.3. To determine more accurate overvoltageconditions, however, a TNA or EMTP study wouldbe better for an existing system and earlier data andexperience for a new system.NoteOn 132 kV networks the MVAr loading is light. as most of thep.F. is controlled at the distribution level and the capacitivechargiiig MVAr demand is low (Table 24.2, column 9). Thecharging MVAr is normally not compensated because, on load,more than this is o ffw by the load p.f.Shunt capacitors These are used to supplement thenatural line capacitance, C,, under heavy inductiveloading depending upon the system voltage, but areused generally for p.f. improvement of the system,where the natural C, is not sufficient to maintain theline p.f. They reduce Z, and enhance SIL, Po, andboost the line voltage. They are normally switchedand not permanently connected to avoid resonanceon load rejection or an open circuit. Generally theyare used for systems up to 33 kV, i.e. at the distributionend. But when the p.f. IS not fully compensated at thedistribution end it can be compensated at the secondarytransmission level of 66 kV or 132 kV. Capacitors atsuch voltages may be connected through dedicatedtransformers, as illustrated in Figure 24.24.SwitchingDedicatedtransformer66 or 132 kV lineIbanksSystem voltage regulation 24/799NoteOn 66 kV networks the MVAr loading is normally high andtherefore one practice is to install MVAr meters and adopt amanual switching during variation of MVAr beyond thepermissible level, purely as a cost consideration.3 Series capacitors These are used for line lengthcompensation to help transmit power over longdistances and also improve the stability level of thenetwork. They are usually installed at the line ends orat the selected locations. They reduce Zo and enhanceSIL, PO, and boost the receiving-end voltage.Example 24.3 Application of series compensation onan HT distribution networkLet us consider the primary distribution network of Example23.2 as shown in Figure 24.25(a) feeding an LT load of 29.4MW at 0.98 p.f. through a 33/0.4 kV transformer. The followingline parameters have been considered:Resistance of primary distribution overhead lines, sectionB-B at the operating temperature,Ro = 0.13 akm per phaseInductive reactance of this section at 50 HzXLo = 0.4 akrn per phaseThere is no leakage capacitance, C, and hence no Ferrantieffect on such low voltages. We will use series compensationto reduce the line voltage drop and improve the regulationand hence the stability of the network as well as its loadtransfer capability,Load p.f. = 0.98 (L - 11.48')In Example 23.2 the system was not capable of transmittingits full capacity. Let us consider that with the use of seriescompensation it can be fully loaded up to30 MVA x 0.98 = 29.4 MW.The impedance of the transformerZz=Px v,"100 kVAx lo3_-- 10 33' x 106100 ~ O lo6 X= 3.63 R(from equation (13.3))For ease of calculation, let us consider the impedance ofthe transformer as its leakage reactance, ignoring resistanceand draw an equivalent circuit diagram as in Figures 24.25(b) and (c). Assuming the length of the primary distributionline to be 15 km, the total line parameters will becomeXL = 3.63 + 15 x 0.4 + 3.63= 13.26and R = 0.13 x 15= 1.95 RFigure 24.24 Use of dedicated transformer to connectcapacitors on networks 66 kV and aboveReceiving-end voltage before series compensationTo study the voltage fluctuation at the receiving-end withfluctuations of loads, let us do so in terms of variation inthe transmitting-end voltage, assuming the receiving-endvoltage remains constant at 33 kV. We are doing this forease of calculation and for drawing the phasor diagram.

24/800 Industrial Power Engineering and Applications Handbook94:11/400 kVIPrimarylil transmission400/132 kV Jk---j-ev.B 132/33 - kVISectionB-8 underconsiderationIIIA HV - -1I IJ jIISecondarytransmissionIdistribution PrirJlary133/0.4 kV-Secondarydistribution29.4 MW(a) Typical primary distribution network30 MVA,-,- 6PR= 1.95 Q1TShunt 4capacitors--_ B+LT loads = 29.4 MW at 0.88 p.f.(b) Details of section B-BB O @ O @ O@ Leakage reactance of transmitter-end transformer.@ Resistance of overhead lines of section B-B.@ Inductive reactance of overhead lines of section B-B@) Capacitive reactance of series capacitors.XcLeakage reactance of receiving-end transformer.(c) Equivalent circuit of Section B-BFigure 24.25Determining the value of series capacitors for a primary distribution networkTo study the impact of series compensation we consider16.62the full-rated current of the transformer and the line for = 22.43 tan21.43optimum utilization of the entire system.= 22.43 L 17.16"E, =33 - -- + I, . Z , in kV:. Voltage drop = 22.43 - 19.05J3= 19.05 + __(1000= 19.05 + 0.525 x 1.95 L - 11.48" + 0.525X 13.26 L (90 - 11.48')= 19.0s + 1.024[cos (-11.48') + J sin (-1 1.48')]+ 6.96 (cos 78.52" + J sin 78.52')= 19.05 + 1.024(0.98 - J 0.199) + 6.96(0.199 + J 0.98)= 19.05 + 1.00 - J 0.20 + 1.38 + J 6.82= 3.38 kVor 15.07 % of E,(Figure 24.26)3 6 . 9 6 k V17.16"525 AV, = 19.05 kV1.00 kv= 2 I .43 + J 6.62 Figure 24.26 Receiving-end voltage after shunt compensation

19.05System voltage regulation 241801It is difficult to operate such a system on full load. Itis bound to have wide voltage and load fluctuations,more so on a line disturbance. Such a system may haveto be operated well below its rated capacity to retain itsstability. even when the voltages on the primary and thesecondary transformers are adjusted so that the requiredrated voltage is available at the receiving-end. A voltagesuing of 158 between a full load to a load rejectioncondition is too wide and may lead to outage of thesystem on a line disturbance. Since there is no furtherscope to improve the above situation with the help ofshunt capacitors (the p.f. is already at 0.98). let us do sowith the help of series compensation. Since the far-endp.f. is being maintained at a high level, the system canachieve some stability. A series compensation to the extentof, say 60% of the total line impedance should not beexcessive, as the line already has some series resistance.Moreover, the transformers will have some resistancetoo which has been ignored in the above analysis andhence the chances of a sub-synchronous or ferro-resonanceoccurring will be remote.:. Series compensation = 0.6 ( I .c)S + J 13.26)Say. x,. = 8 Rand 5ize of series capacitors.-52s' x 8IO00= 3305 kVAr per phase= 8.04 R (in absolute terms)For 10% load variation, to be on the safe side. thecapacitors must be rated for:= ( 1 1 )I x 220s= 3668 kVA1and voltage across the capacitors,v, = I, . x,= 52s x 1.1 x 8= 4.62 kVIf we consider three units in series and nine in parallel(Figure 24.27(a)) then the size of each unit--- 2668 = 98.8 say. 100 kVAr3x9and the voltage rating of each unit- 4.62 = 1.54 kV3Thc improved line impedance= I.Y5+J 1326-JSO= 1.95 + J 5 26and the improved transmitting-end voltage, the loadremaining same.E\=1905+(0525L- I148")(1.9S+JS.26)= 19.05 + 1.024 (0.98 - J 0,199) + 2.761 (0.199+ J 0.98)= 19.05 + 1 .00 - J 0.02 + 0.55 + J 2.71= 20.6 + J 2.51I= 20.75 tan-2.51~20.6= 20.75 L 6.95":. voltage drop = 20.75 ~or 8.2% of E.= 1.7 kVThis is also the regulation of the system. See the phasorrepresentation shown in Figure 24.27(b).1njeftiuric.e.r1 Raising the compensation from 60% to, say, 70%may further improve the above situation but this maynot he advisable to maintain a high level of stabilityduring line disturbances. Moreover. the p.f. of thesystem has already reached a high of cos 6.95". i.e.0.99, which also is not advisable. Tu be more safe. thelevel of shunt compensation should be slightly reduced.NoteSince the load variation on an HT dictriburion network will beonly nominal. and the network will also have enough reAance.it should be possible to compensate the systeni up to 70% 01- so.to further improve the regulation of the network. hay up to 5%oI'E,. withom jeopardizing the level ofstnbility. Thc applicationengineer can take a more judiciou\ decijion. knowing thecondition of the network to he compensated.2 The voltage variation with the series compensation,although high, at about 8.2%, is still manageable byadjusting the tappings on the transmitting-endtransformer, for which a tramformer with highertappings may be selected or a transformer with a highersecondary voltage may be chosen. say, at 36 kV orso. For minor adjustments, the tappings on thereceiving-end transformer may be used. With this,the above system can be utilized to its optimumcapacity.3 The phasor displacement between the transmittingand the receiving ends, with the use of seriescompensation, is reduced and the receiving-end voltagehas moved closer to the transmitting-end voltage. whichwill provide more stability to the system during a linedisturbance.4 Even a higher cross-section of line conductors wouldbe able to improve the above situation by reducingthe line resistance and hence the voltage drop.S It will be pertinent to note that series compensationon HT lines will be more effective when the lineinductive reactance itself is high, as when the line isindividually feeding highly inductive loads, such asan induction or an arc furnace or other similar loads.Nevertheless, it can also be effectively applied onoverloaded distribution networks similar to the onewe have considered above. to raise the line capacityand reduce the voltage dip at the receiving end.

24/802 Industrial Power Engineering and Applications Handbook132/33 kVLoad 29.4 MVAat 0.88 p.f.* 27 Nos. 100 kVAr eachN, (Series group) = 3 HHN2 (Parallel group) = 9HHHHHHShunt capacitors2240 kVAr perphase(for arrangementrefer to Figure23.18b)(1) The capacitors are shown in the centre ofthe line which being the best location. Butthey can be provided near the receivingend transformer also.(2) They are to be mounted on platforms,insulated for 19.05 kV from the ground.(3) The normal practice is to mount eachphase units on separate platforms,insulated for 33 kV from each other.2700 kVArh)per phaseFigure 24.27(a) Application of series compensation on an HT distribution network6.96 kv 7 To provide reactive support for any power system ornetwork, suffering from voltage fluctuations or highline transfer losses the or required when it load is felt it is that important the system to carry cannot out= 20.75 kva field study first, to identify areas and suitable locations2.761 kV17.16"where reactive support would be more appropriate. A6.95"procedure along the lines of Example 24.3 to determine11.48" V, = 19.05 kV ,oo kVthe amount and type of reactive support should then525 A be adopted.* Series compensationNoteIn fact E, is the fixed phasor and V, the variable But for ease ofdrawing, we have considered V, as the base phasor.Above we have dealt primarily with the technical aspectsof reactive controls. For commercial implications, seeLakervi and Holmes.Figure 24.27(b)comDensationsReceiving-end voltage after shunt and series6 For large concentrations of loads, such as for anindustrial or a residential area and where addition ofmore loads in future is likely, forecasts of a realisticloading of lines may fail. Therefore, for growing citiesparticularly, it is advisable to install initially a slightlylarger primary distribution network to cater for theincreasing power needs. It is felt that for such loadcentres, an 11 kV or even 33 kV. distribution isinadequate. The commensurate primary distributionfor such locations may be considered at 66 kV, and inresidential and industrial areas or public placesunderground cabling should be adopted to rninimizcthe risk of running such high tension lines in the openand also save the scarce load area. Underground cablingis more expensive than an overhead system, but ismore safe in congested areas. The use of an overheador underground system will depend upon the location,safety and convenience, besides consideration of cost.See Lakervi and Holmes in the further reading formore details.NoteCountries like the USA and Japan have adopted undergroundcabling for transmission of power up to 1000 MW at 550 kV.24.9 Transient stability levelThis defines whether a system can restore normal operationfollowing a major disturbance, such as on aTransient phase to ground faultPhase to phase faultThe outage or failure of a generating unitFailure of the overhead line or a transformer andSudden opening of the line.The highest level of power it can transmit during suchdisturbances is its transient stability limit. Dampeningthe power oscillations on such disturbances and restoringthe power system to stable limits are the main objectivesof reactive control. From the load curves one can observethat an uncompensated line may have dangerous powerswings on small fluctuations of load, which may lead toloss of synchronism between two generators and evencause an outage of the line. To keep this system stableduring such disturbances, it must be operated at muchbelow its steady-state stability level.Consider a typical case during a line disturbance whenthe line is not adequately compensated. A transmission

network is being fed by more than one generator. Whenany transient fault occurs, the current flow through thefault will be shared more by the machine nearest to thefault and less by the one installed a little away. This willupset the earlier tandem operation of the machines andthey will become unequally loaded and may fall out ofstep. The one near the fault will slow down more thanthe other. The machine that shares the smaller amount ofthe load will slow down less and feed more, becomingovermessed. Now it will slow down and the other willpick up. The situation will reverse thus and so the situationwill continue creating a hunting effect. The followingmay result depending upon the transient stability limitof the system.The fault is cleared promptly and the normal conditionis restored. The setting and the speed of the protectiverelays should be commensurate with such a situationand isolate the faulty circuit as quickly as possible.If not, the machine being loaded most may fall out,which may not necessarily be the one nearer the faultorThe situation may have a cascading effect until all themachines fall out, resulting in a total blackout. 2To achieve a better level of stability it is desirable thatthe line be loaded a liltle less than the optimum power itis capable of transmitting to sustain the system disturbancessuch as load tluctuations, faults and switchingof large machines without an outage. The load curves(Figure 24.20) provide a guide to determine the level atwhich the line should be operated and from this can beassessed the magnitude of di$turbances that the line cansafely sustain and recover promptly without an outage.Series reactive support will become essential. wheneverthe line loading is expected to be more than the STL, P,,(generally on 132 and 220 kV networks). But seriesreactive support has been found extremely useful onexisting lines even up to I I kV, which are required tocater for higher power demands than were originallyen\,i\aged.24.10 Switching of large reactivebanksThe \cries capacitors are connected in series with thepower lines to provide reactive control to an individualload or to a power distribution or transmission system.They are therefore switched with the power lines and arethus permanently connected devices.But the shunt capacitors can provide reactive controlthrough unswitched, i.e. permanently connected, banks(fixed VAr) or through switched banks (variable VAr).The unswitched VAr may be used to aid stability againstpossible overvoltages of the network, during a loadrejection or an open circuit while the switched VAr isused to maintain the level of p.f. during load variations.VAr switching can be done in two ways.IManual control This is through switching devicesSystem voltage regulation 24/803by switching in or switching out a few units. In manualswitching it will be possible only in steps. and maynot provide a smooth compensation and may alsocause switching transients (Section 23.5. I). Moreover.conventional switching methods (mechanical switchingthrough contactors and breakers), are sluggish due tothe time of closing and interruption, which may be ;ismuch as three or four cycles, depending upon thetype of interrupter (Section 19.5 and Table 19. I ) aswell as the minimum time required for the dischargeof the capacitors. Human sluggishness may alsointroduce some delay. They are therefore ill-suited tomeet the system’s rapidly fluctuating needs.However, power systems that cater to almost fixedloads at a time and whose variations occur only atspecific times of the day may not require ii fastresponse. In such cases. it is possible to provide manualswitching methods which will give enough timebetween two switchings. Manual switching, however,has certain shortcomings, due to the human factorsuch as its accuracy and diligence, as noted above.The recommended practice is therefore to select fastreactive controls as noted below.Static VAr compensators (SVCs) Whenever a largereactive control is required, the SVC is always apreferred method. The static VAr controllers are moreexpensive, but respond very quickly. They cause noswitching transients and limit the magnitude of adisturbance. through extremely fast controls. Theycan handle large currents and peak inverse voltages,except voltage transients, such as switching surges orlightning strikes, which may have a front time as lowas 1-2 ps only (Section 17.3.3) while the switchingtime of a static device (a thyristor) may be as much asone cycle, as discussed later. But surges can be takencare of by a surge arrester. The use of an SVC or amanual switching will largely depend upon thecharacteristics of the line. the type of load it is feedingand its importance. For a system having almost thesame type of load demand during the day. inanualswitching may serve the purpose. But for a systemwith wide fluctuations, an SVC alone will be suitable.The decision will vary from one system to anotherand the system engineer can make a better choice./voteIf auto-control is selected through p.f. or voltage control. catemust be taken against frequent switchings of the capaciroi-swhen the load is of a varying nature which may c;iuse thecapacitors also to switch frequently. Fast switching?, can bemade possible by providing special di\charyc devices. and bycontrolling the number of switchings to within permisyihle limit\(Section 26.1.1(2) by carefully arranging the unity 21.; diicus\edin Section 23. IS. 1In SVCs the number of switchings is of no relevance,as they are free from inrush currents. Switching isperformed at the instant when the current wave ispassing through its natural zero. Static devices invarious combinations and feedback control systems,which may be computer-aided, can almostinstantaneously (5 1 cycle) generate or absorb reactivepower, as may be demanded by the system. Correction

24/804 Industrial Power Engineering and Applications Handbookis quick and matches the fast-changing load parametersof the power network at the receiving end. They arecapable of maintaining a near-constant voltage profileat all times at the receiving end. The correction achievedis accurate and smooth, besides being extremely fastand free from surges. They may be installed at strategiclocations along the line or at the receiving end. Theselection of location is an important aspect to optimizethe size of compensator and a more efficient voltageregulation.A fast VAr control is achieved through thyristorswitching, which by itself is capable of a steplessvariation. But switching of capacitors, which areswitched in banks, is not stepless. The SVCs may beof the following types.24.10.1 Thyristor-switched capacitor banks(TSC)Thyristor-switched capacitor banks are normally connectedin parallel with several banks of shunt capacitors to controlthe system voltage. Feedback sensors and controls monitorthe voltage level. When the voltage swings to either sideof the preset value, a few banks are switched in or switchedout. This is illustrated in Figures 24.28(a) and (b). Pointa indicates the operating point under normal conditions.During a load variation or disturbance the voltage dipsand the operating point shifts to 6. With the use of TSC,the load point is shifted back to c. Since the controlis in steps, it may be coarse. The steps may be limited tosave on the cost of thyristors. This step change in voltagecan be smoothed and a stepless reactive control achievedwith the use of a TCR (thyristor-controlled reactor) inparallel and operating it with the TSC banks in tandem.Such a scheme can be tailored to suit even the smallestreactive need of a system. The combination can be termedhybrid compensators. One such scheme is illustrated inFigure 24.3 1 and discussed later, in more detail.In TSCs the thyristors are used in anti-parallel to switcha capacitor bank ON or OFF but without any phase anglecontrol. A TSC therefore does not by itself generate anyharmonics, unlike a TCR.24.10.2 Thyristor-controlled reactors (TCRs)These consist of two oppositely poled thyristors, as shownin Figure 24.29 and conduct on alternate half cycles atthe fundamental frequency. Reactors may be switched orphase angle controlled. Three-phase SVCs can independentlycontrol each phase and the TCR can be used forphase balancing. When a phase angle is controlled, astepless reactive power control can be achieved, exceptfor generation of harmonics during the control process.The gate control at peak voltage (a = 90") can allow fullconduction of the reactor. The conduction can be controlledby varying the gate angle, a. For example, partialconduction is possible with a between 90" and 1 SO", buta from 0-90" is not used, as then the circuit wouldproduce asymmetrical currents with d.c. components.The effect of increasing the gate angle is to reduce theharmonic components of the current, and hence the powerlosses in the thyristor controller and the reactor. If theFigure 24.28(b)compensatorFigure 24.28(a) Switching instants for a TSC1, -Improvement in loading by use of a TSCreactors and the thyristors are connected in delta, tripleharmonics can be eliminated and filter circuits would benecessary only for the remaining harmonic quantities.Various combinations of thyristor circuits are possible toobtain a desired phase displacement between the voltageand the current (cos $) and hence suppress the variousharmonic contents present in the system. (Section 6.13provides more details on this.) See also the further readingat the end of this chapter.The number of thyristors in series, each selected foran impulse voltage of a little less than the impulse voltagewithstand level of the terminal equipment (Table 11.6)can effectively limit the switching overvoltages withindesired safe limits. Then connecting them in anti-parallelwill mean that the voltage will be forward for either of

System voltage regulation 24/805Figure 24.29Opposite poledthyristorsFor practical approach, thewhole reactor is divided intotwo parts One part ISconnected after the thyristorto limit the fault currentL:I:Interrupter ( 9ReactorScheme for a thyristor-controlled reactor (TCR)the opposing thyristors, hence protecting the systemagainst overvoltages in either direction.24.10.3 Transient-free switchingTo switch ON a charged capacitorIn a thyristor circuit ifa charged capacitor is left ungatedat ;I current zero there will be no conduction of currentwhile the capacitor will still hold the full d.c. charge, asillustrated in Figure 24.30, equal to positive or negativepeak ofthe system voltage. For a transient-free switchingthe capacitor i5 switched when the system voltage andthe capacitors’ induced e.m.f. have the same polaritiesand coincide almost in magnitude. This situation maytake up to one full cycle (Figure 24.30) and can delaythe switching by one cycle when switched immediatelyafter a switch OFF. The system’s protective devices maybe introduced with an additional time delay of at leastone quarter to one half cycle to bypass disturbances of atransitory nature.DC charge ona switch otf‘Id at everyone-half of a cyclcOne cycle_ _The next similarcondition occursafter one cycleFigure 24.30 A delay up to one cycle in transient-freeswitching ON, of a charged capacitorTo switch OFF a charged capacitorThis can be achieved at any current zero which occursevery half cycle (Figure 24.30).Reactor switchingUnlike a capacitor. an energized reactor on a switch-offretains no charge at a current zero and can be switchedON or OFF on a current zero at any point on the voltagewave without causing a transient. Hence there is a delayof, at most one half of a cycle between two consecutiveswitchings of a reactor. The balance ofthe two oppositelypoled thyristors, however, is monitored through the gatecontrol to avoid even harmonic quantities, although oddharmonics will still be generated when the gating anglesare balanced, i.e. are equal for both the thyristors.24.10.4 Response of SVC on a fault or linedisturbance of a transient natureAn SVC offers an extremely low response time, of theorder of just one cycle as noted above. But this time issufficiently high to respond against disturbances of atransient nature. For instance. during a fault condition.as expressed by the current-time oscillogram of Figure14.5, the SVC will respond during the transient periodonly and not during the sub-transient period. The subtransientperiod may be less than a cycle and not fallwithin the response range of an SVC. But a reactivecorrection is also not needed for conditions of such atransient nature, which is taken care of by the surgearresters (Chapter 18). A system is normally suitable toremain stable, without an outage, during disturbances ofsuch a transitory nature. Similarly, the SVC will stayimmune to lightning and switching surges.24.10.5 Combined TSC, TCR and fixedcapacitor banksWith the combination of switched capacitors (TSCs andTCRs), also known as a hybrid combination, each phasevoltage can be closely monitored to maintain a nearbalancedand flattened profile at all times at the receivingend. A typical scheme is illustrated in Figure 24.3 1 whichcomprises:A few fixed capacitor banks which are normallyenergized. When they are required to be switched,they cause a switching delay due to the closing oropening of the interrupting device. besides generatingthe switching surges. To avoid delays and switchingsurges, they may also be made as TSCs, if cost is oflittle consideration and faster and more accuratecorrections are more important. in view of highlyfluctuating and non-linear load demands.A few TSCs for finer reactive controls.A few TCRs to balance the reactive power supply.They may generate harmonics. which must besuppressed to avoid any resonance. TSCs and TCRsare monitored through a feedback control system.A filter circuit to absorb the harmonic currents generatedby TCRs and in certain conditions. when TCR is ‘OFF’.

24/806 Industrial Power Engineering and Applications HandbookI&TV.TdInterrupter (\A/ D Transformer forreactive power supply.Secondary at say, 11 kV400 kV transmission or132 kV distribution lineReactive powersupply busLine side reactorto limit inrush currentand suppress harmonicsSome fixed reactorto filter out subharmoniccomponentsand feedback0Note: * Capacitors and reactors may be A connected toeliminate triple harmonics@ Thyristor switched capacitors@) Filter circuit to absorb harmonic currents caused by TCR@ Thyristor controlled variable reactor (TCR) @ (i)-A few fixed capacitor banks that may be normally ON.(it may also be a saturated reactor)(it) Normally D connected. They are not grounded, when A connectedFigure 24.31Scheme for a reactive power control showing combined TSC,, TCR, and fixed capacitorsalso generate capacitive reactive power. Refer to Figure24.33 for more clarity.Consider a noymal load line (1) (Figure 24.32) havingthe initial operating point at (a). On a disturbance, theload line shifts to (2) and the operating point to (b). TheTCR would respond and some inductive reactance (X,)will be shed to raise the content of Xc. The load line willbecome less inductive and more capacitive to help thevoltage rise to point (c) within one cycle. If the voltageis still below the preset value, some capacitors can beswitched ON either electromechanically or through TSCs,depending upon the system adopted. The delay at point(c) will depend upon the method of switching of thecapacitor banks. The voltage will now jump to point (d)and final correction is achieved up to point (e). Thesequence a-e would complete in less than two cycles ifall the components are thyristor switched. The sequencesfrom a-b-c-d-e can be reduced to a-b-e allowing a littleover- or undershoots from c to d.NoresI Reactive control is also possible through synchronous condensers.As they rotate, the rotor stores kinetic energy which tends toabsorb sudden fluctuations in the supply system, such as suddenloadings. They are. however, sluggish in operation and veryexpensive compared to thyristor controls. Theii rotating massesadd inertia, contribute to the transient oscillations and add tothe fault level of the system. All these factors render them lesssuitable for such applications. Their application is thereforegradually disappearing.2 Reliability of shunt or series power capacitors is of utmostimportance for the security of the system on which they areinstalled. Their failure may disturb the system or result in asystem outage.3 Generally, SVCs are designed for I1 or 33 kV and connected toa higher voltage system through a dedicated transformer throughthe tertiary of the main transformer. Figure 24.33 illustrates atypical SVC system using a dedicated transformer.4 Since reactive controls are normally meant for large to verylarge installations, the practice so far has been to use thyristorsonly for such applications. With the advent of IGBTs, smallerinstallations can now be switched through IGBTs.

System voltage regulation 24/807400 kV lineDedicated400/33 kV& transformerY\\\ \\ \\\\\\\ \\ \7thharmonicfilter10 MVArCapacitive -+- lnductive- Load current -@- Normal load line@- Load line on a disturbance (overloading)@I - Corrected load lineNote: Similarly, load lines can be drawn on a load rejection, or thegenerator or the line outages.Figure 24.32 Voltage regulation during overloading throughreactive management- - - -Filter circuits-A "Range of compensation availableCapacitive : 0 to +140 MVArInductive : 0 to -140 MVArIllustration:(a) When the TCR is ON, MVAr support = -190Now filter circuits would also be on to suppressharmonics and absorb capacitive MVAr = +50:. Net maximum reactive support available = -140 MVAr(b) When the TCR is OFF, TSC would supplycapacitive MVAr = +90The filter circuits would also supplycapacitive MVAr = +50:. Net maximum reactive support available = +I40 MVArFigure 24.33 Typical SVC at 400 kV through a dedicatedtransformerRelevant StandardsIEC Title IS BS60358/1990 Coupling capacitors and capacitor dividers9348/1991 BS 75781199260519- 111984 Safety in electroheat installations 9080-1 to 4 BS EN 60519-11199360694/1996 Common specifications for high voltage switchgear and 342711991 BS EN 60694/1997controlgear standardsRelevant US standards ANSI/NEMA and IEEEIEEE 824/1994NEMA-CPU1992Standard for series capacitors in power systemsShunt capacitors, both LT and HTNotes1 In the tables of Relevant Standards in this book while the latest editions of the standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

24/808 Industrial Power Engineering and Applications HandbookList of formulae usedCapacitors for improvement of system regulationRegulation =Voltage at no load - Voltage at full loadVoltage at no loadRating of series capacitorskVAr = 3. If . X ,and voltage rating = I, . X,.(24.1)(24.2)I, = line currentXc = capacitive reactance of the series capacitors perphaseReactive power management(24.3)P = power transfer from one end of the line to thereceiving end per phase= phase voltage at the transmitting end= phase voltage at the receiving end, in radial linesand midpoint voltage, in symmetrical linesZO = surge impedance of the linesin 8 = line length effect or Ferranti effect6 = load angle or transmission angle(24.4)X, = inductive reactance of the whole line lengthInfluence of line lengthVelocity of propagation of electromagnetic wavesLo andlengthe=e=n=I =A=(24.5)C,, are the line parameters per phase per unit(24.6)phase shift between the transmitting and receivingendvoltages in radians or degrees22- or 180"respectively7line length in kmwavelength in kmVoltage at the receiving end, when it is open-circuited,EE, = I (24.7)cos eLine length effect or Ferranti effecte= ?nf.JL,c,.i (24.8)Optimizing the power transfer throughreactive controlv,zP = -.per phaseZ, sin eV,, = nominal phase voltageNatural loading or surge loading of a line,V;Po = - per phasezoAnalysis of radial linesEs = V, cos 0, + J Z,, . I, . sin 8,(24.9)(24.10)(24.11)(24.12)E, = phase voltage at the transmitting end.V, = phase voltage at the receiving end0, = line length effect or Ferranti effect at the end ofthe line, in degreesI, = load currentZo = surge impedance of the lineAnalysis of symmetrical linesE, = V,,, . cos e,,, + J Zo . I, . sin 0, (24.1 3)V, = voltage at the midpoint of the line0, = line length or Ferranti effect up to the midpoint ofthe lineFurther readingI Central Board of Power and Irrigation, India (Indian NationalCommittee for Cigre), Electric PoM,er Transmission at voltagesof 1000 kVatid Above, Oct. (1984).2 Central Board of Irrigation and Power, India, Static VArCompensators, Technical Report No. 41, March (1985).3 Central Board of Irrigation and Power. India, Workshop onSeries Compensation in Power Systems, Nov. (1986).4 Central Board of Irrigation and Power, India. Workshop onReactive Powercompensation Planning and Design. Dec. ( 1993).5 Westinghouse Electric Corporation, Elecfricril Trrmsnii.ssioiimid Di.stri6utiori. East Pittsburgh. Pennsylvania. USA.6 Engbarg. K. and Ivner. S.. Static VArSystemsfor Voltage ControlDuring Steady State arid Transieiit Conditions, May 198 I.ReactivePowerCompensationDepartment. Sweden. May (1981),7 Erinmez. LA. (ed.). Static VAr Compensators. Cigre WorkingGroup 38.01. Task Force No. 2 on SVC (1986).8 IEEE - Delhi Section, Advance Level Course on Reactii'e PowerCoiitrol in Electrical Power Systems, December (1984).9 Kundor. P., Power SI'stem Sfability and Corirrol, McGraw-Hill. NewYork.IO Lakervi, E. and Holmes, E.J.. Electricit?. Distributiori Nehi.orX-Design, Peter Peregrinus, London.11 Miller. T.J.E., Reactive Power Control in Electric Sy.srems,John Wiley, NewYork (1982).12 Praaad, J. and Ambarani. V.. Static VAr Compensator ,forIndustries, BHEL, India, 59th R & D Session, Feb (1994).13 Research Station. M.P. Electricity Board. Central Board ofIrrigation and Power, India, Series Capacitor Application toSitb-fraii.srinssior2 system.^ Case Studies. 'I'echnical Keport No.66. Nov. (1988).I4 Weeks, W.L.. Eaii.smi,s.sion arid Distribirtioii ojElectrical Energy(DrAigri rrsprcts). Harper & Row, New York.

Making CapacitorUnits and Ratingsof Switch i ngDevices25/809

Making capacitor units and ratings of switching devices 25/81 125.1 Making a capacitor elementA capacitor unit is made up of a number of capacitorelements stacked together in combination of series andparallel configurations to add up to the required voltagerating and VAr size. A capacitor element can be formedwith any of the available dielectric materials, as notedbelow, depending upon the voltage rating of the capacitorunit, the likely voltage variations, harmonic contents inthe supply system and switching operations.25.1.1 Metallized polypropylene (MPP):self-healing typeEach capacitor element is made up of two layers of thinPP dielectric films, 5-10 pm thick. Each film is metallizedwith an Al, Zn-A1 or Ag-Zn-A1 alloy on one side (Figure25.l(a)) by a vacuum deposition process. To ensureaccuracy of process in metallizing the film, the processis carried out under highly controlled environmentalconditions of temperature, moisture and dust. Thethickness of the metal film is extremely thin, in the rangeof 0.2-5 pm. These two layers are then interleaved andwound on a mandrel (Figure 25.1 (c)) to form an element.The size of the element will depend upon the design,voltage rating and size of the capacitor unit to be produced.These elements are the self-healing type and require nointernal fuses (Figure 25.2(a)). During an internal fault,only the metallized film is fused, removing a negligiblearea of the metal film which is not replenished. The PPfilms remain almost intact, with small localized damagewith a few pinholes. Their life may be considered to bearound 60 000 operating hours suitable for approximately5000 switching operations per year. But they are suitableonly for linear loads. The elements are stacked togetherto form the required size of capacitor unit and impregnatedwith a suitable dielectric, which may be a synthetic oilor epoxy resin.25.1.2 Double dielectric processBy altering the dielectric mix and its thickness it is possibleto change the characteristics of an element to make itsuitable for higher voltage variations and non-linear loads.Making use of this, some manufacturers have introducedan improvized version by increasing the thickness of thedielectric coating of individual elements and impregnatingthe entire capacitor unit so formed with a non-oil dielectric,such as epoxy resin. Epoxy resin is devoid of leakage,while leakage is possible in an oil dielectric. The unit,called a double dielectric unit, is now suitable to sustainhigher voltage fluctuations and the presence of systemharmonics, as required by fluctuating and non-linear loads.But provision of an inductor coil is now essential in eachcapacitor unit to suppress the inrush current. The repairof such units is simple and can be carried out at site ifspare elements are available.25.1.3 Mixed Dielectric (MD)This is the non-healing type, and is normally providedwith internal fuses. In this type, two layers of A1 foil areused and coated with a mixed dielectric, formed of tissueor kraft paper and polypropylene (Figure 25.1(b)).Elements so formed are not self-healing and are normallyprovided with internal fuses for their individual protection(Figure 25.2(b)). Their life, of approximately 90 000operating hours is better than an MPP element, and canwithstand around 20 000 switching operations per year.The elements are suitable for sustaining a higher voltagevariation and the presence of harmonics, as required byfluctuating and non-linear loads.The elements thus formed are then inserted into a sheetsteel container, vacuum dried and impregnated, withsuitable non-PCB dielectric, which may be an oil dielectricor epoxy resin. The capacitor shell is then hermeticallysealed, in oil dielectrics, to avoid any leakage of dielectricduring operation.Until the 1970s the chemical used as the impregnatingand dielectric medium for capacitor units was PCB(polychlorinated biphenyl) liquid. It was found to be toxicand unsafe for humans as well as contamination of theenvironment. For this reason, it is no longer used. Thelatest trend is to use a non-PCB, non-toxic, phenyl xylylethane (PXE-oil), which is a synthetic dielectric liquid ofextremely low loss for insulation and impregnation of thecapacitor elements or to use mixed polypropylene or allpolypropylene(PP) liquids as the dielectric. A non-oildielectric, such as epoxy resin, is also used.DielectricPP filmPP filmLayer ofAlternateFilm coating oftissue paperelectrodesAI, Zn-AI orprotrudingAg-Zn-AI alloy AI foil at the endsconductorTinned copper tabs(a) An MPP element (b) An MD or PP element. For PP element (c) MPP, MD or PP plates wound on a mandrelthere is no layer of tissue paperFigure 25.1Making an element

25/812 Industrial Power Engineering and Applications Handbookiapacitorc;_l$elements(a) Use of expulsion type external fuses(For self-healing type (MPP) capacitor unitsFigure 25.2(b) Use of internal fuses(For non-healing type (MO or PP)capacitor units)(c) Finished capacitor unitsArrangement of capacitor elements for making a capacitor unit (Courtesy: Savin Capacitors)Such capacitor units require careful meticulous checksby measuring the capacitance of the unit every six monthsor even less, depending upon the duty and the switchingoperations it is performing. Repair is not possible at siteas it requires a re-impregnation under vacuum and hermeticsealing once again which is possible only at the manufacturer’sworks. Since it may not be possible to exerciseclose monitoring of the health* of the capacitor unitwhen it is in operation it is possible that there have occurredprogressive isolation of elements after blowing off theinternal fuses. This may cause a reduction in the capacityof the unit and hence an undercompensation. The internalfuses cannot be replaced at site, being installed inside asealed unit.We provide in Table 25.1 a brief comparison of thedifferent types of dielectrics in use for impregnation andinsulation while forming capacitor elements. The tablewill also help make a proper choice of the dielectric andthe type of capacitors for a particular application.protected capacitor units, while the capacity of the unitmay have deteriorated below acceptable limits over aperiod of time. It is also possible that failure of a fewelements on a minor fault may lead to a cascade failureof the healthy elements connected in parallel, due to anovervoltage across them (Figure 25.3). This is a featurethat is almost absent in series-protected units (see alsoSection 25.4.2). But self-healing types of capacitors haveBellowAIDuring a severe faultbellows expand and itJ snap opens-Be4low25.2 A critical review of internallyprotected capacitor unitsWhile it is true that in the absence of external fuses it isnot possible to assess the health of a unit, it is equally truethat an intact external fuse is also no guarantee of a healthyunit. In fact, internal fuses isolate only small elementsand the capacitance of the unit remains almost intact overlong periods. Periodic checks, say, once every six months,are important to assess thd. health of the unit and this isapplicable to all types of capacitor units. Generally, externalfuses show a fault between the line and the shell only.They stay immune to mild immense faults occurring insidethe shell. Minor faults therefore do not show up in externally*with the availability of microprocessor-based p.f. correction relays(Section 23.17) which measure the health of a unit while performingswitching operations, and indicate this through a display, blinker oralarm, it is now possible to monitor the health of a unit while inoperation and to take prompt preventive or corrective measures.i_ Elements _Iin parallel* Failure of one element will cause an ovetvoltage across the others inparallel and may lead to a cascade failure.Figure 25.3 Cascade failure of healthy elements in theabsence of an internal protection

Table 25.1 Application of different types of dielectrics and approximate capacitor lossesMaking capacitor units and ratings of switching devices 25/813Type of dielectricLosses ( W/kVAr) (Typical)ApplicationCapacitorunitDischargeresistanceInternal fusesLT capacitors(i) (a) Metallized propylene(MPP)0.80.7Not provided inview of selfhealingfeatureSuitable for systems where frequent switching is notlikely and the system is free from harmonic generatingsources. They are the self-healing type and their outputmay become reduced with time (because of failures ofcapacitor elements as a result of switching inrush currentsand system harmonics). If the system conditions are notconducive, an inductor coil may be provided to limit theharmonic effects (Section 23.9). Some manufacturers,as standard practice, provide an inductor coil inside theshell to contain the inrush current and also dampen theharmonics.(b) Improvized MPP0.80.7Not provided inview ofself-healingfeature1 The thickness of the dielectric coating on individualelements is nearly double the above2345For total impregnation of the capacitor unit, a nonoildielectric such as flexible epoxy is used to eliminateleakageThe elements are still the self-healing typeThey are suitable for moderately fluctuating voltagesand moderately non-linear loadsUse of an inductor coil in each capacitor unit is stillessential, to suppress the inrush currents(ii) Mixed dielectric (MD)(paper and polypropylene)1.50.70.2512These are suitable for systems with varying loadsand requiring frequent switchings, such as whenthe capacitor units are connected on an automaticswitching modeWhere the system is having harmonic generatingsources (non-linear loads) and is prone to a largeamount of harmonics(iii) All polypropylene(PP) (film foil)HT capacitors(i) Paper dielectric(ii) Mixed dielectric (MD)[quickly outdated](iii) All polypropylene(PP) (film foil)0.50.70.252.2-1.5 0.250.2 0.25*3 Systems prone to frequent voltage variations4 Capacitor elements are non-self-healing and aregenerally provided with internal fuses5 An inductor coil may not be necessary with suchcapacitorsNormally not in practiceThese give a reduced probability of the shell rupturedue to internal fuses*When internal fuses are provided.When, however, expulsion type external fuses areprovided, this loss would fall outside the capacitor unit.Advantage: All polypropylene units, besides having alow loss, provide a reduced probability of shell rupturethe ability to regain their lost capacitance, hence they canperform better over long periods.The greatest advantage of internal fuses, which act ascurrent limiting devices, may be regarded in the protectionof the capacitor shell. Modern practice is to use oils andresins that may be inflammable liquids and are a sourceof a fire hazard in the event of a shell rupture on a severefault. Internal faults may cause arcing and ionize theimpregnating dielectric liquid and generate gases. Excessivegas pressure on a severe fault may cause the shell to explode.Shell rupture protection is a vital consideration in externallyprotected capacitor units. Since there is no control oversmall internal faults until they become major fault, protectioncan be provided only for the whole unit and the entireunit has to be dismantled after such a fault. In fact, thecapacitor bank may have to be shut down completely toreplace the lost unit with a new one to avoid an imbalance,besides making up for the lost capacitance.In units with internal protection the fault is not severe,as the fuses with each element will isolate the element

251814 Industrial Power Engineering and Applications Handbookquickly on the smallest fault and prevent the occurrenceof a major one. The operation is extremely fast as all theparallel elements, which are fully charged when the unitis in operation, discharge into the faulty element. Theoperation of the capacitor unit is restored within milliseconds.There is no necessity for shell protection ininternally protected capacitor units whereas for externallyprotected capacitor units, fuses are required to be closelyco-ordinated with the shell rupture characteristics, asdiscussed in Section 26.1.1(6). However, internallyprotected capacitors have some disadvantages too. Forexample, when they are used on a power distributionnetwork and mounted on poles they are exposed to externaldirect lightning strikes. Similarly, line-switched capacitorsare subject to internal switching surges and inrush currents.It is possible that all these surges will fuse many, if notall. the internal fuses in tandem of such capacitors becausethe fuses are of very small ratings and have a low transientwithstand capability (1*r). Such a fusing may render thewhole unit unusable while the dielectric properties maybe unaffected. It is also not easy to replace or closelymonitor the health of the capacitors mounted on poles.With external fuses this disadvantage is overcomeautomatically as a blown fuse in both events can be easilydetected and replaced.Nevertheless, unbalance is a major occurrence in eithercase and an unbalanced protection, in addition to shortcircuitprotection is imperative in all types of capacitorunits. Hence it is also possible to design series protectedcapacitor units in larger sizes and for higher voltages,which are more economical, compared to smaller units.Whether to use internal or external protection hasbecome a matter of debate due to the different practicesadopted by manufacturers. Users must decide on theprotection that best meets their requirements.ImportantFor loads, such as on motors where the capacitor unit isbeing switched with the machine, or where closemonitoring of the capacitor health is a prerequisite toavoid an eventual outage of the capacitor unit by gradualdepletion of its capacitance, internally protected capacitorunits may be preferred.25.3 Self-healing capacitorsThe failure of a capacitor element due to a change in thesystem parameters, such as fluctuations in voltage, switchingoperations or the presence of harmonic generatingsources on the system is quite common. The elementmay also fail as a result of internal defects in the elementitself or its dielectric quality. The dielectric film coating,being extremely thin (5-10 ps), may break down quicklyunder such unfavourable operating conditions and causea failure of one or more elements. At excessive faultcurrents, however, it may cause a blow up of the externalfuses, rendering the whole unit unserviceable and causingan extra voltage across the healthy units connected in thesame parallel group of capacitor banks. Such an occurrenceis rare but it may generate excessive pressure due toheating of the dielectric within the shell and endangersafety. To avoid explosion of the shell it is commonpractice to provide pressure relief bellows in the unit.These will expand under such conditions to relieveexcessive inside pressure, on the one hand, and isolatethe supply to the unit, on the other, by snap-breaking thecapacitor’s internal connecting wire between the capacitorterminals and its elements, as shown in Figure 25.3.The metallized film capacitors have the characteristicof self-healing. On a small dielectric failure the capacitorelement is not rendered completely unserviceable. Afterclearing the fault, the affected capacitor element returnsto the circuit and the capacitor unit functions normally.Only the punctured area is eliminated from the elementand causes a negligibly small reduction in its capacitancevalue. Such a characteristic is termed ‘self-healing’ andsuch capacitors are known as the self-healing type.The capacitor element in this case is made up ofextremely thin metallic films. When a dielectric failureoccurs in any of the elements, the current passes throughthe film. The film, being too thin to sustain the current,fuses only at the point of dielectric puncture clearing thefault quickly. The external fuses remain intact, and soremains the affected element in service. Somemanufacturers claim that 10 000 such failures and healingsmay reduce the rating of the element or the unit made upof such elements by barely 1%.25.4 Making a capacitor unit fromelementsFor low-voltage applications these elements may be madeup to 1 kVAr or so each and designed up to 440 V,depending upon the system requirement, while for highvoltageapplications, they may be designed up to 2-3kV, each rated for 20-100 kVAr. These values are onlyindicative and may vary from one manufacturer to anotherdepending upon the dielectric used, the process adoptedand the site requirements.When larger units are required, a number of suchelements with appropriate voltage and kVAr ratings, maybe connected in parallel, the size of which will dependupon the requirement and the economic size of thecontainer. For an LT system, each unit may be made upto 50 kVAr, while a more realistic size is found up to 25kVAr in steps of 1,2,3,4,5, IO, 15,20 and 25 kVAr. ForHT capacitors, they may be made in larger units, such as50,56, 84, 100, 150 and 166.7 kVAr. For higher voltages,the elements may be connected in series groups say,three or four series groups, to make them suitable for a6.6 kV system and four to six elements up to a 15 kVsystem. For yet higher voltages, the capacitor units maybe connected in series to achieve almost any operatingvoltage. These series elements may be connected in parallelto achieve larger ratings as may be required. Figures25.4(a) and (b) illustrate some arrangements.25.4.1 Restriction in series-parallel combinationsof capacitor elementsWhile making a capacitor unit, the same precautionswould be mandatory as discussed in Section 25.5.1 for

Making capacitor units and ratings of switching devices 25/815CapacitorelementsDischargeresistanceI-Elements 4in parallelEx p u I s i o nfuse 4(a) Use of internal fusesCapacitorelementsDischargeresistance25.4.2 Protection of capacitor elementsThe elements must be protected against dielectric failures.This may be done in two ways - through internal orexternal fuses - as discussed above. When the fuses areinternal they are provided with each element as illustratedin Figure 25.4(a). The purpose is to isolate any elementthat has developed a fault, rather than rendering the wholeunit unserviceable. If external fuses are provided, asillustrated in Figure 2S.4(b), they will isolate the wholeunit on a fault. Regular monitoring of health ofa capacitorunit is essential to ensure that its kVAr rating is notreduced more than is permissible (Section 26.3.1 ( 1 )). Areduced rating, besides undercompensating the reactiverequirement, will also cause a voltage unbalance in thesystem, leading to voltage fluctuations and magnifyingthe harmonic contents.It may, however, be noted that the dielectric failure ofan element during operation is not an unwarranted feature,so long as it does not cause the rating of the defectiveunit to fall below the permissible limits over long periodsof operation. The size of each element is too small tocause a significant variation in the rating of the unit.However carefully the process and quality controls aremonitored, during the impregnation and forming of coilssmall voids are still possible that may yield duringoperation. Close monitoring of the rating and unbalanceis, however, essential to ensure a permissible rating andvoltage unbalance on a large installation. as noted inSection 26.1.25.5 Making capacitor banks fromcapacitor unitsFor higher system voltages, it is common practice to usemore than one lower voltage rating, identical capacitorunits in series, to obtain the required voltage. With theuse of a number of such series connected units in parallel,one can make any size and voltage capacitor banks. Thefollowing are a few more common practices:(a) Single star:As illustrated in Figure 2SS(a)(b) Double star:As illustrated in Figure 2S.S(b)(b) Use of expulsion type external fusesFigure 25.4 Arrangement of capacitor elements for making acapacitor unitmaking a bank from a number of capacitor units. This isto ensure that failure of a few elements during operationdoes not cause an overvoltage that is more than permissible(Table 26. I ) across the healthy elements and an unbalancein the three phases.Example 25.1To make 76 kV, 14400 kVAr capacitor banks, use 75 kVArcapacitor units suitable for 11 kV each.:. Total no. of units required = '4400 = 192~75System voltage = 76 kV:. Voltage phase to neutral = - = 43.88 kV43:. Number of 11 kV capacitor units, required in series per43.88 phase = --11With adjustment in the voltage rating of the 11 kV capacitorunits, the bank may be made suitable for the required system76

25/81 6 Industrial Power Engineering and Applications HandbookParallelgroup = N2 R,11111111SeriesreactorJRl+n11111111Seriesreactor(a) Single star seriesiparallel capacitorbanks with neutral side series reactors(b) Double star series/parallel capacitor banks with line side series reactorsFigure 25.5voltage. Small voltage adjustments by the manufacturer mayalso be essential when series reactors are being used.No. of single phase units required per phase192= - = 643We may therefore consider a double star system, each havinga configuration for each phase of 4 x 8 units, that is, fourunits in series group to make up the required phase voltageand eight in parallel group to make up the required kVAr(Figure 25.6).25.5.1 Precautions against overvoltages on thefailure of some unitsFor large capacitor banks using a number of small identicalunits per phase the configuration of the units must besuch that a fault or a failure in one unit of a phase shouldnot result in an overvoltage of more than 10% across theremaining healthy units in that phase. This requirementbecomes more specific when the capacitor units are usedwith external fuses and a fault would mean the fallout ofall units. A unit with internal fuses, however, may notpose such an eventuality, as the fault would result in thefailure of just one or two elements of that unit and maynot cause an overvoltage. Similarly, metallized, selfhealingtype capacitor units may have fewer problems ofthis nature, as discussed in Section 25.3. We give belowthe formulae provided by leading capacitor manufacturersto calculate the magnitude of overvoltage in the event ofa failure of one or more units in a group of series/parallelcombination of capacitor units. It will help in rearrangingand balancing the capacitor units to contain the overvoltagewithin 10%. To achieve this, there must be a minimumnumber of capacitor units in parallel:(i) Single-phase units connected in parallel Thevoltage rise across the healthy units on the failure ofa unit of the many connected in parallel can beexpressed by;-7IIIIIIIIFigure 25.6Configuration of a double-star capacitor bank making 14 400 kVAr

~~Making capacitor units and ratings of switching devices 25/817v, = / q v (25.1)N2 -whereVI = enhanced voltage across the units on the failureof a unitN2 = number of units in parallel.It may be observed that when the number of units inparallel are fewer than six, the voltage across thehealthy units will rise by more than lo%, even on afailure of only one unit. The units in parallel thereforemust be a certain minimum number to ensure that inno case will the voltage across the healthy units inthe event of a failure of a few units exceed 10%.(ii) Single-star capacitor banks (Figure 25.5(a))(a) % overvoltage- 3.N.N, -2Nx 100 (25.2)(b) Fault current: The fault current on failure ofone unit may be expressed by3.N,.N,I* = (25.3)(3 N, - 2) ’Icwhere I, is the rated current of one unit.(iii) Double-star capacitor banks (Figure 25.5(b))6.N.NI -5N(a) % over-voltage = x 1006N2.NI -6N.NI +5N(25.4)(b) Fault current: The phase current in a double star isthe summation of phase currents of each star (Figure25.5(b)), Le.Iy = Iyl +- 472The fault current on a double star can therefore becalculated on the same basis as for a single-starcapacitor bank.- 3Nz . N1 - 3 N. N, + 2N4whereN = number of units failedN, = number of units in series per group perphaseN2 = number of units in parallel per groupper phaseExample 25.2Consider Example 25.1 for 7200 kVAr units arranged in asingle star and 14 400 kVAr units arranged in a double star.The computation of overvoltages and fault currents in theevent of failure of a single unit in the two configurations is ina tabulated form as follows:ParametersSingle starDouble starTotalkVAr120014 400Rating of each capacitor unit kVArLine voltagekV15 I l516I“PhVoltage rating of each unitNo. of series units per phase pergroupNo. of parallel units per phaseper group(a) Overvoltage (OV) %(on failure of one unit N = 1)kVkVN1N244i 44I l183,N.Nl -2N3N2 .N, -3.N.Nt +2N x 100 6.N.N, -5N6N2 . N, - 6N. N, t 5N x 100- 3XlX4-2Xl xloo6x1~4-5x1 xloo3x8x4-3xlx4+2xl6x8x4-6x1x4t5x1= 11.6= 10.98which exceeds the desirable limit of 10% which exceeds the desirable limit of 10%48In such cases, the overvoltage protection may still be set for 10% and even less, to contain thefault. It is a recommended practice to provide OV protection on the failure of only one half ofa unit, to limit the damage to a very low extent.OV in the event of damage toonly one half of a unit(b) Load current of each unit3xtx4-2xf3x8x4-3x;X4+2x+= 5.5%x 100I-- 15 = 6.82 A- 116xix4-5xi6x8x4-6xfx4+5x+ x 1001 = 5.2%

~ a~25/81 8 Industrial Power Engineering and Applications HandbookPhase current under healthy conditions:. Fault current, If:. Overloading7200x 7 61 = 54.7 A3,N, .Nz(3N, - 2) 'Ic- 3x4x8 x6.82(3 x 4 - 2)= 65.47 A-- 65.4754.71 = 19.7%x 76= 2 x 54.7= 109.4 A1 (for the phase with the faulted unit)- 65.47 + 54.7109.4= 9.84 %25.5.2 Special-purpose capacitor unitsLike a special-purpose electric motor (Chapter 7) acapacitor unit can also be required to perform underspecial operating conditions, such as:Frequent switchingsVarying loadsVarying supply frequencyHigh harmonic quantities (as for electric arc andresistance welders and electric arc and inductionfurnaces)High humidityHigh ambient temperature (such as near a furnace)Saline, corrosive or dust-laden surroundingsChemically aggressive orHazardous installations, contaminated with inflammablegases, vapour or volatile liquids (such as for marine(aboard a ship) or mining applications)Such types of loads may require special design ofcapacitor elements and their dielectric impregnation,cooling arrangement, size of shell or surface treatment.For all these applications therefore it is important toknow the actual operating conditions, behaviour andcharacteristic of the load and its duty cycle before selectingthe capacitors.25.6 Rating and selection ofcomponents for capacitor duty25.6.1 Current ratingThe factors discussed in Section 23.5.2 give rise directlyto the current drawn by the capacitor unit and indirectlyadd to its rating. The relevant Standards on this devicerecommend a continuous overload capacity of 30% toaccount for all such factors. A capacitor can have atolerance of up to +15% in its capacitance value (Section26.3.1( 1)). All current-carrying components such asbreakers, contactors, switches, fuses, cables and busbarsystems associated with a capacitor unit or its banks,must therefore be rated for at least 1.3 x 1. 15fc, i.e. 1 SI,.For circuits where higher amplitudes of harmonics areenvisaged, for reasons of frequent load variations or moreelectronics circuits, it is advisable to consider a moreliberal factor, say, up to 200% I,.. See the footnote inSection 26.1.1(3).Example 25.3For a capacitor bank of 500 kVAr in 10 units of 50 kVAr each,for an automatic power factor correction the rating ofcomponents would beI, = 500 (for a 415 V supply system)fi x 0.415= 695.6 A:. Incoming equipment should be suitable for695.6 x 1.5 Aor 1043.4 A, say, 1000 Aand for individual units 104.3 A, say, 100 AWe have summarized in Table 25.2 the basic parameters toselect the correct type of components for capacitor duty.NoteAs a rule of thumb consider the rating of switching andprotective devices in amperes, the minimum being twice thekVAr rating of the capacitor units in LT. For a more economicalselection, however, the manufacturers of switching devicesand control equipment may be referred to for theirrecommended size of equipment, as some sizes may have abuilt-in reserve capacity not requiring a higher componentrating.25.6.2 Precautionary measures for circuitshaving large capacitor banksA capacitor draws excessive charging currents andgenerates switching surges during a switching operation.It also magnifies harmonics when harmonics are presentin the system. All these are undesirable and may becomehazardous when the banks are large, the more so whenthe system is HT. The following are home remedialmeasures that can be taken to safeguard the connectedequipment from the adverse effects of such parameters:1 Electrodynamic and thermal stresses These aredue to excessive charging currents. All the currentcarryingcomponents used in the circuit such as the

Making capacitor units and ratings of switching devices 251819interrupter and the bus system must be suitable towithstand such stresses.Switching surges The switching device must becapable of making and breaking the circuit successfullyand generate the least possible voltage surges. Allother equipment and interconnecting cables used inthat circuit must be suitable to withstand the severitiessurges thus generated.Voltage harmonics When used in a transformercircuit capacitors may generate harmonics. At lightloads, the capacitors must be switched off or theirvalue reduced to avoid oversaturation and consequentdamage to the transformer. Otherwise they may causeresonance, between their capacitances and the inductanceof the transformer, giving rise to overvoltages,on the one hand and voltage harmonics, on the other(Section 20.2.1(2)). In rectifier and arc furnace applications,however, inductive reactances of such equipmentwill rise due to high harmonic frequencies and offsetto some extent the effect of the higher voltagesproduced by the third harmonic.Residual voltages (for all ratings) Having energizedonce, the capacitor takes time to discharge. There isthus a residual voltage across the capacitor terminals,even after the power supply has been switched off.This residual voltage is detrimental to the connectedequipment as well as to the operator. It is thereforemandatory that this residual voltage is dampened fromits peak value at fi V, to a maximum of 50 V within1 minute in LT and 10 minutes in HT shunt capacitorsand 5 minutes in HT series capacitors after a switchoft.(See also the footnote in Section 26.3.1.)When a capacitor switch is reclosed after an interruptionon a capacitor which is not fully discharged the trappedcharge of the capacitor will be superimposed on the appliedvoltage and lead to overvoltages. Suitable resistance unitsare fitted across the capacitor terminals as standard featuresby the capacitor manufacturers to dampen the residualvoltage from its crest value to the desired level or less,within the stipulated time, as noted above. When fasterreclosing becomes essential, as in an automatic p.f.correction scheme or for such load applications that changequickly with little discharge time between two successivereclosings as noted above then either the discharge device,as discussed in Section 25.7 or the switching schemeshould he designed $0 that the residual voltage acrossthe capacitor terminals decays to at least 10% of thesystem voltage by the next reclosing.The following are examples where the capacitors shouldeither he isolated from the circuit with the switchingdevice so that its induced e.m.f. does not impinge on theconnected equipment or it is reduced to a safe level bythe next switching.Protection from excitation voltagesStar/delta or auto-transformer srvitching of (I motorFor motors compensated individually and having a star/delta or auto-transformer switching, the changeover fromstar to delta windings or from one step of the auto-transformer to the other is a case of rapid reclosing. Thismay result in an overvoltage across the motor windingsdue to two residual voltages, one of the motor's owninduced e.m.f. at the reduced voltage of y or of tappingof the A/T and the other of the capacitor charge. if thecapacitor is connected across the motor windings(terminals ai, bl, ci), irrespective of the time of discharge.In such cases, the capacitor should be open-circuitedwhile the motor is in star or auto-transformer positionthrough a special starter. A typical power and controlscheme is illustrated in Figure 25.7.Hoist and crane applicationsDescending loads may overspeed the motor and overexcitethe capacitor when connected across the motor due 10motor generator action above the synchronous speed(Section 6.21). Such a situation may damage the motoras well as the capacitor and must be avoided.Electrodynamic brakingWhen a motor is statically controlled and employs dynamicbraking (Figure 6.32) and has a capacitor also connectedin the circuit it is possible that on an instant braking thecapacitor may discharge itself into the static circuit anddamage the electronic components used in the rectifieror the inverter units. This situation must also be avoidedby suitably designing the capacitor switching circuit.Stability of a power networkIn this case the next switching may almost be instant toretain the stability of the system and save it from anoutage. A static switching (Section 24. IO) alone may bethe correct solution for such applications.25.7 Fast discharge devicesTo accelerate discharge, one of the following method\may be adopted. This is a special requirement and themanufacturer must be informed.Use of extra resistance across the capacitor terminals(Figure 25.8).Use of high series reactor, with or without resistanceacross the capacitor terminals. Open-delta VT (Section15.4.3) is one such device, which is a combination ofan inductance and a resistance and is to discharge anHT capacitor unit quickly.Use of a switch which, in the open position, can chortcircuitthe capacitor with small resistance.NotesI Capacitor5 conncckd directly (without a switching device)across a motor or a transformer are automatically dischargedthrough their windings, being inductive, and need no additionalresistance or inductance for a faster discharpe.2 A capacitor being switched through a Y/r> \witchins of amotor, howevci-, poses a different kind of a problem a5 discussedabove and requires a different type of sicitching rchrme, asillustrated in Figure 2.5.7. This is because a charged capacitor

IECIECIECTable 25.2Rating and selection of components for capacitor dutyType ofcomponentReferenceStandardvoltage class Insulation level Current ratingFault level(M and time)Any other requirement1 2 3 4 5 6IIsolator LT ~ 60947-3HT - IEC 60129HRC fuses LT ~ 60269HT - IEC 60282-1,60549 and 60644CircuitinterrupterLT - IEC 60947-4- I(for contactors)HT ~ 600.56(for circuit breakers)As per thevoltage ratingof the capacitorbanks (Section23.12.1)(average voltage)As in Tables13.2 and 14.3for Series I orTable 14.1 forSeries I1 voltagesMinimum 150% ofthe rating of thecapacitor banks.For systems proneto generatingexcessiveharmonics, andwhen the size of thereactor, if used, isnot sufficient tosuppress these, ratingsup to 200% arerecommended.As for themain system.For LT isolators minimum duty AC23.(Table 12.5)1 For LT systems, it may be a switch anda contactor, ACB or MCCB. When acontactor, they would be minimum AC3or AC6b duty. (Table 12.5).2 For HT systems, refer to Table 19.1,for the selection of interrupters forcapacitor duty. The breaker should bechosen restrike free, as far as possible.Cables andconductorsSeries reactorRefer to TableA16.1 in Chapter 16._IEC 60289 Refer to column 5 I For smaller units, theyare used to limit onlythe inrush currentsduring a paralleloperation and may berated for 0.2% of thecapacitors’ kVArrating. They must beconnected on theneutral side of a starconnectedcapacitorbank. Also refer tothe notes below step 4of Example 23.4.The series reactormay not be suitable towithstand the systemfault level, nor addsignificantly, to thecircuit impedance todiminish its fault level.__ ~Refer tocolumn 51 For limiting inrush currents: Generally,it is not required for the LT systems.For larger banks, however, say, 500 kVArand above, when they are installedclose to the supply source, so that theline reactance up to the bank is toosmall to limit the switching currents,the use of reactor would be advisable.2 For suppressing harmonics: since on anLT system, the capacitor banks arenormally small, there is generally noneed to provide a reactance to suppressthe harmonics. It is, however, advisableto provide this for larger banks above500 kVAr to contain the overvoltageson account of this, to save the otherdevices also operating on the samesystem. For the likely magnitudes ofharmonics on an LT system, refer toTable 23.1.

.' 1'' iI9-PI7009 381~-9-PI7009 331" II-CPOOY 3819t t Z I

25/822 Industrial Power Engineering and Applications HandbookPower supplyR Y B NCFT T T 1 1-T-YtD IM1'1h3Ydelaytrip-NffPower circuitControl circuitM - Main contactorD - Delta contactorS - Star contactorC - Capacitor contactorh,, b, h3 - Indicating lightsTD.0. - Time delay open contactOCR - Over current relayT - TimerIMotor-statorwindingIFigure 25.7duringTypical power and control scheme for ,k/P starting of a motor with capacitor bank, illustrating capacitor shorting(by I/& V,) is now switched onto a motor winding which,besides the applied voltage, is also carrying an induced e.m.f.equal to the reduced applied voltage (I/& V, ).25.7.1 The value of discharge resistanceThe discharge resistance as in IEC 60252 can bedetermined byR, = 5 (25.5)where R, = discharge resistance in R.If it is very high, so that its power loss V,' lRd is lowthen it may be left permanently in the circuit. Otherwiseit should be introduced into the circuit only duringinterruption through the switching device) (wired acrossthe NO contact of the switching contactor). This willprovide the resistance across the capacitor banks whenbeing interrupted and short-circuit it when closed.C = capacitance in faradz = time constant in seconds. It is the time by when theexponentially varying voltage VI, will diminish to0.6321 V,. In the most stringent switching operation,it can be expressed as(25.6)wheret = changeover time or required discharge time inseconds. For normal applications not requiringfrequent changeovers it is less than 1 minute for LTand 10 minutes for HT systems (or according tothe practice of a particular country)V, = system voltage in voltsu = residual voltage, say, 50 V (or 75 V, dependingupon the practice of a country) or a maximum up to10% of the rated voltage for a quick reclosingk = a constant that will depend upon the configurationof the capacitor unit and its discharge resistance, asnoted in Figure 25.8.

ak= 1Rk= 113Making capacitor units and ratings of switching devices 25/823- 30log, 14.67--30 or 11.17seconds2.686and discharge resistanceR, = 11'17 (2 = 93 k12, say, 100 kC2120 x 10-6If the closing period is only 15 seconds, the required dischargeresistance will become 50 kR. The resistance loss in 100 kQ,discharge resistancev,"-- i.e. 4152 or 1.72 wattsR d 100 x 103and 3.44 watts in a 50 kQ discharge resistance. Since theloss is negligible the resistance may be left in the circuitpermanently.k= 125.7.2 The value of a series inductorIn an L and R dampening circuit the capacitor willdischarge according to the following:(25.7)i.e.tL VP= 2 x - R . log, - U seconds(25.8)k- 1 k=3Rk= 1 k=3Figure 25.8 Possible arrangements to provide the dischargeresistances across the capacitor terminals and values of kExample 25.4Consider the scheme of Example 23 4 having an automaticparallel switching If we assume the closing sequence cycleto be 30 seconds, the recommended value of dischargeresistance for each 20 kVAr capacitor bank having acapacitance of 120 pF can be determined as follows'. z =t = 30 secondsVI = 415 voltsu = 40 volts (10% Of V,)k = 1 for D configuration30fi 4151 x log,40wheret =changeover time or required discharge time insecondsL = series inductance in the circuit in henryR = series resistance in the circuit in Q.vp =ziz-.v, = 1 P.U.43u = residual voltage, say, 50 V or 75 V as noted earlier,a maximum up to 10% of the rated voltage for aquick reclosing.Example 25.5Determine the discharge device for the discharge of a threephase6.6 kV, 50 Hz, 1000 kVAr, y-connected capacitor bank,connected in units of 10 x 100 kVAr each, through an automaticp.f. correction relay, having a closing cycle of 10 seconds.Data available from the capacitor manufacturer,C = 30 pF (for each 100 kVAr bank)Alternative 1Consider an open delta transformer for this purpose, havingthe following data:L = 120 henryR=90Qu=75vThe time of discharge with these datatL VIJ= 2 x - x log, - secondsRU120= 2 x - log,906600 x ~7545seconds

25/824 Industrial Power Engineering and Applications Handbook= 2 x -120log, 71.8490= 2 x -120x 4.274 = 11.4 seconds90The transformer should be able to meet the switchingrequirements in view of the following:That the line impedances have been ignored, which wouldfurther dampen the capacitor charge on every switch-off.For a fast changeover, a terminal voltage of more than75 V is permissible, which in the above case after 10 secondswill not exceedm= fi x 6600 ,(-A'lo) (ignoring the line6 impedance)= 126.7 V which is quite low compared to 10% of 6.6 kVand will not damage the equipment.NoteUse of additional resistance or inductance or both in serieswith the VT must be avoided, particularly when the VT is alsobeing used for the purpose of measurement. Any introductionof L or R into the VT circuit would affect its accuracy.Alternative 2The above discharge is also possible through only an extraresistance in each capacitor switching circuit. Using equations(25.5) and (25.6),Considering the capacitors to be connected in y, then k = 1,andR, = lo x11 x 30 x lodf i x 6600log,-75=69k52(6600)'and resistance loss = ~ = 631 W69 x lo3This is a significant loss. The resistance therefore should beconnected during an interruption only, through a switch (orthe no contact of the interrupting device which can easilycarry this short time burden) and should not be left permanentlyin the capacitor circuit.Rd tZFor a normal discharge in 10 minutes a resistance of -t,69x103 ~10x60i.e.10= 4.14 MR (say, 4 MR) would be necessaryThis resistance would normally be fitted across the capacitorterminals by their manufacturers and will cause a permanentloss of--",2R- wattsor 6600 6600 = 10.89 W4x lo6which is reasonably low and the resistance can be leftpermanently in the capacitor circuit.Relevant StandardsIECTitle IS BS60044-31198060056/198760099-4/199860 129/1984Instrument transformers ~ combined transformersHigh voltage alternating current circuit breakersSurge arresters - Metal oxide surge arresters withoutgaps for ax. systemsA.C. disconnectors and earthing switches-13118/19913070-3/19939921 (1 to 5)BS 7628/1993BS 531 1/1996BS EN 6009941993BS EN 60129/199460044- 1 / 1996Specification for current transformersGeneral requirementsMeasuring current transformers2705part- 1/1992part-2/1992BS 7626/199360044-6/199260044-2/1997Protective current transformer for special purpose applicationsApplication guide for voltage transformersSpecification for voltage transformersGeneral requirementspart-3/1992part-4/19924201/1991,4146/19913156- VI99231S6-2/1992BS 762Y1993BS 7729/1995601 86/1995602521199360269-1/199860282- 1/199460289/1988Capacitive voltage transformers. For measurement andprotection. General requirementsA.C motor capacitorsLow voltage fuses ~ General requirementsHigh voltage fuses - Current limiting fusesReactors3156-3/19923156-4/19922993/199813703-1/19939385-1 to 55553 (1 to 8)BS EN 60252/1994BS EN 60269-111994BS EN 60282-1/1996BS EN 60289/1995

60549i 1'176Making capacitor units and ratings of switching devices 25/825Ili~h-\oltnge fuse\ for shunt power capacitors 9402/1992 BS 556419926llh-131 I979 Specification for high-voltage fuse-links for motor -circuit applicationsBS EN 60644199360947-3/1 u9x Specification for LV switchgear and controlgear 13917-3/1993 BS EN 60947-3/1992S\vitche\. disconnectors, switch-disconnectors andluse comhination units611947-4- 111 990 ILow \oltage switchgear and controlgearF.lectroinechanical contactor\ and motor \tarter\Relevant US Standards ANSYNEMA and IEEEANSI!IEEE(' 37.W/IW4N m1,\/I(.S ~2/1')91Y 131 ,W.I'-'I/ I 99'N Ehl A tT .-I / 19x11NEM.\ S G21 I 9') 3Giiitie tot pi-otecuoii of shun1 capacitor banksInchstrial control and \y\teins. controllers. contactors and overload relays. rated not inore than2000 V J.C. or 750 V d.c.Ezrernal f iises for shunt c;ipacitorsLo\\ \ olrage cartridge furesHigh \oltayc I'USC\,V(l/?\I23In the table\ oi'Rclcvanr Standards in this book while the latest editions ofthc standard5 are provided. it i\ possible that revi\cd edition\Iha\t. become a\ailable. With thc advances 01' technology and/or its application, the updating of Standards ih a continuous process bytiilferciir ~t:ii)ti;t~d~ or-ganiiations. 11 is therefore advisable that for more authentic references. reader\ should conhuh the iclevilntor~ani/ati(iii~ 1.o~ the latest \ersion of a standard.Soine ot the BS or IS standards mentioned again\[ IEC may not be identical'The hex noted .iyin\t each Standard may ;d\o refer to the year of its last aniendment and no( nece\?arily the year of publicationList of formulae usedPrecautions against overvoltages on the failureof some units(iii) Double-star capacitor banks% over voltage = 6.N.N) -5N6N2 . N, - 6N. N, i 5Nx lOO(25.4)(i) Single-phase units connected in parallel./(25. I)V, = enhanced voltage across the healthy units on thefailure of one unit. 2'- ,- no. of units in parallel(ii) Single-star capacitor banks3 N N, -2N('1) '4 oven oltage = 3NL N, -3N N, +2N(25.2)N = number of units failedN, = number of units in series per group per phaseN2 = number of units in parallel per group per phase(b) Fault currentDischarge devices(i) The value of discharge resistanceR, = CR, = discharge resistance in 0.C = capacitance in faradT = time constant in seconds(35.5)tr=(35.6)d3. V,k ' log, ~ut = changeover time or required discharge time insecondsV, = system voltage in voltsu = residual voltagek = a constant that will depend upon the confip-ationof the capacitor unit and its discharge resistance(25.3)(ii) The value of a series inductorI, = fault current on failure of one unitf,. = rated current of one unit (25.7)

25/826 Industrial Power Engineering and Applications HandbookL VPt = 2 x - . log, - secondsRUt = discharge time in secondsL = series inductance in the circuit in henryR = series resistance in the circuit in SZa 'vp = -J5v, = 1 p.u.u = residual voltageFurther reading(25.8) AIEE Committee, Report of a Survq on the Connection of ShuntCapacitor Bank.r, Vol. 77, pp. 1452-58, Feb. (1959).Central Board of Irrigation and Power, India, Manual on ShuntPower Capacitors - Operation and Maintenance, Technical ReportNo. 33, Jan. (1984).Longland, T., Hunt, T.W. and Brecknell. W.A., Power CapucitorHandbook, Butterworth, London (1984).

Protect ion,Maintenance andTesting ofCapacitor Units

26.1 Protection and safetyrequirementsAs discussed in the previous chapters, a capacitor mayhave to encounter many unfavourable service conditionswhen in operation. Most of them may lead to its directoverloading. Summing up all such conditions, a capacitorwould need protection against the following:26.1.1 Protection of shunt capacitors1 Overvoltage (Section 23.5.1)Capacitors become overloaded with overvoltages andlong duration of overvoltages may reduce their life. Thepermissible over-voltages and their safe duration as inIEC 6083 I - 1 for LT and IEC 6087 1 -I for HT capacitorunits are indicated in Table 26.1. An overvoltage factorof up to 10% can be considered on account of this. Ininstallations having wide voltage fluctua-tions and whenberies capacitors are used on such systems for improvingtheir regulation then the capacitors must be protectedagainst abnormal overvoltages. Overvoltage protectionmay be provided through an overvoltage relay with inversetime characteristics. For closer monitoring, it is better tohave the relay in small steps of 1%.2 Number of switching operationsIn LTcapacitors controlled through aut0maticp.f. correctionrelay it is recommended that the capacitors are not switchedmore than 1000 times during a year or, say, three or fourswitchings per day. More switchings will mean morefrequent overvoltages and inrush currents which mayreduce the life of a capacitor. In HT capacitor banks,however, such a situation may not be relevant as HTcapacitors may be switched only once or twice a day.3 OvercurrentDue to harmonic effects the capacitors are designed fora continuous overloading of up to 30% (Section 23.5.2)and are accordingly rated for 1.31,. Since the permissiblecapacitive tolerance may enhance the capacitance ratingof a capacitor by 1.05-1.15* times its designed rating(Section 26.3.1 (1)) the maximum permissible current maybe considered up to 1.3 x 1.15 = 1.5 times its ratedcurrent. A capacitor generally is a fixed current device,its rating is greatly influenced by the circuit parameters,particularly the presence of harmonics and fluctuationsin system voltage. Since a capacitor unit is designed for150% of its nominal rating an overload protection cannothe over-emphasized under normal operating conditions,"We have considered the maximum overcurrent factor for illustration.It corresponds to the maximum permissible upper variation in thecapacitance value of a capacitor or a bank. One can, however,consider a lower factor depending upon the maximum permissibleupper variation for a particular type (voltage and VAr rating) ofcapacitor unit or bank. A factor of I .I5 is, however, recommendedfor capacitor duties. See also Section 25.6.Protection, maintenance and testing of capacitor units 26/829particularly for an LT installation. For large HTinstallations, however, experiencing wide voltagefluctuations (which may be rare) and the presence of alarge degree of harmonics, it may be mandatory to provideovercurrent protection, even if series reactors are used tosuppress the harmonic currents. An IDMT (inverse definiteminimum time) relay may be provided with overloadand short-circuit protections and a time delay to by-passthe momentary transient and switching inrush currents.Protection against a short-circuit is essential for both LTand HT capacitor units. On LT systems it may be providedthrough the HRC fuses or the in-built short-circuit releasesof a circuit breaker when a circuit breaker is used toswitch the capacitor banks. On HT systems short-circuitprotection is also provided through HRC fuses or shortcircuitreleases of the breaker.4 Switching currentsThe switching currents must be limited to as low a valueas possible. However, in general, capacitors are designedfor switching currents up to 100Zc (r.m.5.) for milliseconds.If necessary, a series resistance or inductance can beintroduced into the switching circuit to limit them to adesired level (Section 23.1 I).5 Short-circuit and ground fault currentsThese will depend upon the grounding system adopted.The level of fault current would generally he as follows:For grounded star capacitor units: In this case when acapacitor unit develops a short-circuit it will establisha line-to-ground fault with a high fault current. Use ofcurrent limiting HRC fuses will be easy because, forheavy fault currents, the fuses can be selected of aslightly higher rating to avoid an interruption due toswitching surges and transient inrush currents withoutaffecting operation of the fuses on an actual fault.However, a fuse free system (Section 12.1 1) is advisableby using an MCCB.For ungrounded star capacitor units: In this case whenit causes a line to neutral fault the impedance of theother two phases will limit the fault current to almostthree times the rated current. It is still easy to chooseHRC fuses so that they would remain inoperativeTable 26.1durationsLikely overvoltages in service and their permissibleMaximum durcitiun~~ ~~~~~ ~~~ ~ ~- ~ ~~ ~ ~ ~ ~~ ~~ ~ ~For capacitor iinitsfactor up to I000 v ribow 1000 VI .oo1.101.15I .20"1.30"- -ContinuousContinuous8 h per 24 h12 h per 24 h30 rnin per 24 h< 5 min 3I < 1 min >"Not to occur more than 200 times during the total life span of thecapacitor unit.

26/830 Industrial Power Engineering and Applications Handbookup to 150% of the rated current continuously and operatefor a few seconds at 300%. Refer to Figure 2 1.4 fortypical current-time curves of HRC fuses. Use of anIDMT relay with an appropriate setting will also begood practice in such cases.For-delta connected capacitor units: In this case, itwill establish a line-to-line fault with a heavy faultcurrent. Protection by HRC fuses would be moreappropriate than for a grounded star or using an MCCB.For series-parallel connected units: This is applicablein HT capacitor banks comprising a numher of smallunits arranged in series and parallel combinations. Thefault current in such cases for an isolated neutral canbe expressed by3N, NlI,-=-. 3N, -2 I‘(from equation (25.3))Now also an IDMT relay would provide the mostappropriate protective scheme.To summarize, HRC fuses or fuse-free MCCBs for LTand a breaker and IDMT relay arranged for 2 O/C and1 G/F, with a short-circuit unit for HT capacitor banks,will provide a reliable protective scheme for short-circuitand ground fault protections.6 Shell protectionThis is applicable to both LT and HT capacitors. But it ismore important in HT banks, which are relatively muchlarger and are built of a number of single units connectedin series-parallel. These may encounter much higher faultcurrents in the event of a severe internal fault, even inone unit and are thus rendered more vulnerable to suchruptures. This phenomenon is more applicable to unitsthat are externally protected where the intensity of faultmay be more severe, than internally protected units.10t 3’0-. 1.0mEmE 0.30.1Probability curves for case rupture0.1 0.5 0.9I Safe I Yqlt-I30 100 300 1000Fault current, amp (rms)Figure 26.1 NEMA’S case rupture curveProtection with internal fuses is easier, as fuses areprovided for each element which can contain the severityof the fault well within the safe zone in all probability.Some users even recommend capacitor units 250/300kVAr and above with internal fuses only. Figure 26.1shows a typical operating band of the internal fuses foran internally protected unit. It demonstrates a sufficientmargin between the operation of the fuses and the shell’ssafe zone. The fuse characteristics are almost the samefor all manufacturers.Since such explosions are dangerous and may cause afire hazard through the dielectric liquid which may beinflammable, it is imperative that such faults are clearedbefore they may result in bursting of the shell. They willrequire short-circuit protection. Leading manufacturersproducing units suitable for external protection providea pressure-sensitive disconnector which operates andreleases the pressure during a fault when the insidepressure builds up to a preset level. This may be in theform of expandable bellows, which expand on suchexcessive pressures inside and snap open the powerconnections. Safety through expansion bellows is usuallya feature in LT, MPP capacitors. In HT this technology isnot used. NEMA has also provided probability curvesfor the case rupture in the form of 1’ versus 1. The curvesare defined in Figure 26. I , which can be used to selectthe appropriate protection to isolate the unit on a faultbefore a possible rupture. The severity of fault is expressedin terms of the magnitude of explosion. The protectionmust be commensurate with the location and the criticalityof the installation and is categorized in four zones,depending upon the severity of the fault:1 Safe zone: There is only a slight swelling of the shelland no severe damage.2 Zone I: There may be a slight rupture and fluid mayleak. Safe for areas where the leakage will pose nohazard.3 Zone 11: There may be a violent rupture of the shell.4 Hazardous zone: There may be a violent rupture witha blast, which may damage the adjacent units.NoteThe practice adopted by manufacturers, for all voltages and kVArratings in the making of the shell, particularly for its size, materialand thickness, assume almost the same I‘ versus t characteristics,as provided by the NEMA curves (Figure 26.1).For a more accurate selection of a protective scheme itis essential that the manufacturers provide the probabilitycurves of their shell design for each voltage and rating.Generally, the fault must he cleared well within ZoneI and for which the protective scheme must be chosen.As discussed in Section 25.4.2, protection of capacitorunits with external fuses is not easy. It is not practical tocontain a mild internal fault as isolation of the units isnot possible on mild internal faults until the fault currentrises to the level of the fuse’s operating range (Figure26.2 illustrates this). By then enough time will haveelapsed to cause severe damage to the unit.The only solution is to choose fuses of a lower ratingas far as possible, with fast operating characteristics (low1*t) or provide IDMT protection. But the risk of a shell

Protection, maintenance and testing of capacitor units 261831502010t:: 402zK01Safe zone Zone-1 Zone-2 Hazardous zone--I----Minimum rating of fuses = 200% of lci.e. 2 x 6.82 A or 13.64 ASelect from the manufacturer's catalogue 16 A fuses foreach unit, according to the characteristics in Figure 26.2reproduced in Figure 26.3. The fuse characteristics fall wellwithin the safe zone and provide adequate protection to theunits. If series reactors are provided to control the inrushcurrent, a fuse corresponding to 150% of lc, i.e. of 10 A,would be better to protect the unit more closely. We can seefrom Figure 26.2 that in an internally protected unit thecharacteristics of the internal fuses will fall well below that ofan external fuse and provide a more exacting protection on amild fault.7 Voltage transientsThese may be caused by internal switchings or externallightning strikes. HT capacitor units, particularly. maybe protected through surge arresters against such voltagetransients (Chapter 18). The voltage rating of the surgearrester may be chosen on the following basis:For ungrounded systemi = V,For grounded systems = 0.8 V,See also Table 18.20 013-0 0110 30 100 200 300 5001000 3000 1 10000*Fault current (Amp ) rms5000@- Characteristic of a 10A external fuse from Figure 26 30- Characteristic of a 16A external fuse from Figure 26 30- Probability curves for case rapture' Use asymmetrical rms values for transient faults8 Voltage unbalanceAn unbalance in the supply sy\tem causes an okervoltage20 0004 hrs100001000Figure 26.2 NEMA probability curves for a case rupture andselection of internal and external fuses~100rupture is not eliminated. An unbalance protection is abetter choice in such cases, which may isolate the faultyunit much faster than the luses. When it is so, the unitsremain intact. At this stage visual checks can be made tofind all such units that may be excessively hot or havebulged out. The capacitance of the units can also bemeasured to identify the faulty unit(s) or the extent ofdamage to individual units and take corrective steps beforereahitching the units. Example 26.1 illustrates a simpleprocedure to select external HRC fuses. The fuses areprovided separately for each unit.Example 26.1Consider the same units as those in Example 25.2 whereV,, of each unit = 11 kVkVAr = 75N1 = 4N2 = 8I, = 6.82 Aand If = 65.47 A40EE 10c2t!?1001-0 0110 10 100 lo00RMS symmetrical prospective current (Amp )Figure 26.3 Timehrrent characteristics for 11 kV HRCfuses (Courtesy: GEC Alsthom)

261832 Industrial Power Engineering and Applications Handbookacross the capacitor units in phases that have a highervoltage. A partial or total failure of a capacitor unit in alarge capacitor bank will also cause an unbalance and anovervoltage across the healthy units connected in parallel.Such a situation is not desirable, and unless detectedpromptly, may result in a cascade failure of more units.It can be detected by the following methods for an alarm,indication or trip of the entire bank:(i) Voltage unbalanceThis method is applicable to single-star or delta-connectedcapacitor banks. Unbalance can be detected through theuse of an RVT (residual voltage transformer) (Section15.4.3). See Figure 26.4. The theory of operation is thatany unhalance, of the system or the capacitor bank, willshift the neutral and reflect as the residual voltage acrossthe open delta and can be used for the protective scheme.The unbalance voltage across the open delta in the eventof failure of a unit in any series group can be expressedby' vphvu = 3N2. Ni - 3N. N, + 2NUngroundedcapacitor banksITertiary winding i,ITo the relay and ~control panel tomonitor discharge 1conditionsTo the neutral(26.1)Residual voltagetransformer (RVT)Figure 26.4 Detecting an unbalance in a capacitor bankthrough a voltage unbalance method(established by the capacitor manufacturers)whereV,, = unbalance voltage across the open delta. This isfor grounded systems. For isolated systems it willbe three times this (Section 15.4.3).N = number of units failedN, = number of units in series per group per phaseN2 = number of units in parallel per group per phase.A neutral displacement relay (NDR) with a suitable settingmay be used across the open delta for unbalance protection.Example 26.2Referring to Example 25.2,V,, = 44 kVNJ = 4N2 = 8Then the unbalance voltage for a grounded system, on afailure of any one unit (assuming the capacitor units to haveexternal type fuses), according to equation (26.1)1 x 44 x 1000vu =3X4X8-3X1X4+2Xl- 44 000- 86= 511.6 Vc5VJdSurge\darresterSurge21dVT"arresterL1(a) One bank (b) Two banks (c) Three banksFigure 26.5 The use of RVTs with large capacitor banks

~ 15351535Protection, maintenance and testing of capacitor units 261833and for an ungrounded systemVu = 3 x 511.6 J VIf the ratio of VT is 44 kVillO V, then the voltage across theopen deltax 11044000=384Vand suitable NDR may be selected for this voltageFor more than one capacitor bank one RVT for each bankwill be necessary for the correct discharge of each bank,(Figure 26 5) An RVT, if provided with an additional starconnectedtertiary winding rated for 110 V line to line, canalso check the discharge condition and its duration and if itmeets the design requirements This can even be monitoredremotely through a relay and control panelNoteWhen an RVT is being used for the purpose of discharge,also no protective fuses are recommended to ensure a positivedischarge. On a trip, the RVT is programmed to be connectedautomatically to the capacitor units to discharge them.(iij Ciirt.etit iit~brrltrric~ nirthorlThi\ method is applicable to double-star-connectedcapacitor banks (Figure 26.6). In this method the neutralsof the two identical star-connected banks are interconnectedand a protection Neutral CT (NCT) is insertedthrough it. In a balanced state, there will be no currentthrough the neutral. On a partial or a total failure of aunit in either of the banks it will disturb the balance andtaus ;I current through the interconnected neutral. Aground fault relay will be most suitable for such arequirement. To avoid momentary transient trippings. atime delay of a few seconds can be introduced into thetrip circuit. Sensitivity and the setting of the relay willdepend upon the size of the banks and the sensitivity ofthe system, i.e, whether an alarm or indicator will besufficient initially while repairs are undertaken duringoff-peak periods, or whether a trip will be needed. Theamount of unbalance current can be determined by thefollowing formula:(26.2)(established by the capacitor manufacturers)where I, = imbalance current through the interconnectedneutral, andI, = current per capacitor unit.NoteCurrent-sensing protection i\ found to be more accurate than avoltage sensing. It ia therefore recommended that the unit\ arearranged in double $tar as far as possible.(iii) Ampere meter mrthodTo monitor the health of a capacitor uiiit il is advisableto provide an ammeter in each phase. When it i\ found,that there is an unbalance not caused by voltage unbalancethen further investigations may be conducted to replacethe defective unit(s) before a major fad^ occurs.Example 26.3Considering the earlier Example 25.2 with 14400 kVAr banks,connected in double star whereN, = 4N2 = 8TTRBank-I1wL- ~ ~Unbalance relayStar connected capacitor banksBank-ll1Figure 26.6 Detecting an unbalance in double-star identical capacitor banks through a current unbalance method

26/834 Industrial Power Engineering and Applications HandbookThen the overvoltage developed in the event of a failure ofany one unit (assuming the capacitor units to have externaltype of fuses) as determined in the said example= 10.98%This overvoltage is higher than permissible. The fault conditionmust be removed before one full unit fails. Say, we permitonly one half of a unit to fail for an alarm and maximum threequarters for a trip. Then overvoltage when one half of the unithas failed from Example 25.2.= 5.2%and overvoltage when three quarters of the unit has failed6 ~ $ ~ 4 - 5 ~ :x 100 = 8%6x8x4-6x$x4+5x$which is well within the limits for a tripand I, for an alarm from equation (26.2)3X+X8- X 6.82 = 0.45 A6 x 8 x 4 - 6 ~ 3 ~ 4 + 5 ~ 3and I, for a trip- 3Xax8X 6.82 = 0.69 A6 ~ 8 ~ 4 - 6 ~ : ~ 4 + 5 ~ $One can use a 5/1 ACT and provide the desired setting in a1 A relay for alarm and trip.9 Frequency variation and voltage variationWhen capacitors are used to improve the p.f. of a varyingfrequency load, such as for an inductive heating or varyingvoltage loads for improving the regulation of a powersystem, they would be subjected to excessive overloadingdue to frequency variation in the first case, since kVAr 0:f(equation (23.4)), and voltage fluctuations in the latter.Capacitors performing such duties must therefore beprotected, against overloads by an IDMT overloadprotection scheme.10 No-volt protectionOn supply failure capacitors must drop out and shouldnot switch on automatically on resumption of the supplyto avoid an overvoltage as a result of the trapped charge.Even sudden voltage dips may cause the charged capacitorsdischarge into the terminal equipment and damage them.On LT, to achieve the required protection, the capacitorsmay be switched through contactors which have a novoltcoil and drop out on failure of the supply. On HT aninstantaneous undervoltage relay, with a low drop-offvalue (say, 30-60%) may be used with the interruptingdevice, which may be a breaker or a vacuum contactor.Reclosing may be through a lockout relay with a timedelay, for at least the safe discharge time or less, if thecontrol is by a p.f. correction relay, whose time setting islow and the capacitors are provided with suitable dischargedevices.Figures 26.7 and 26.8 illustrate general layoutsrespectively for an 11 kV and 33 kV p.f. improvementsystem with switching and protective devices. Figure26.9 illustrates a typical layout of a switching andprotection scheme for an HT capacitor system, suggestingthe recommended switching and protective devices. Basedon this, an application engineer can plan a more strongerprotective scheme for a particular system.26.1.2 Protection of series capacitorsSeries capacitors are subject to higher voltage variationsas a result of generally fluctuating line current. During afault in the line, such as a short-circuit, the voltage variationacross the capacitors may be so high that it may cause aninstantaneous failure of the capacitors unless the capacitorsare selected for a higher insulation level. IEC 60143-1for series capacitors suggests that the short-time maximumovervoltage across the capacitors should not exceed 2.15times the capacitors’ voltage rating for 10 seconds. Sincean overloading results in an overvoltage, overvoltageprotection becomes even more essential. IEC 601 43- 1permits a temporary overloading as in Table 26.2 basedon which overvoltage protection is provided.The following are a few common methods for theprotection of series capacitors.1 Protection against overloadsFor overloading during normal operation, which may bedue to load fluctuations or failure of a few capacitorelements, normal overload protection will suffice, asdiscussed for shunt capacitors (Section 26.1.1(3)). Thisprotection may be provided in conjunction with the linefault protection scheme, discussed below.2 Protection against line faultsThere may be a number of ways by which the seriescapacitors can be protected against line faults. Commonlyused methods are briefly discussed below.(i) Using a dampening circuitThe theory of protection is based on a rapid voltage riseacross the series capacitors during a line fault. This voltagerise is used to achieve the required high-impedancerequirement of the system during a fault, i.e. restoringthe near-natural impedance of the line as if it were withoutseries capacitors. A simple R-L dampening-cum-dischargecircuit is illustrated in Figure 26.10. This consists of aresistor and reactor combination, which helps to limitthe discharge current and hence the overvoltage acrossthe capacitors during a fault by almost offsetting thecapacitor fault current, IC (fault), with the Inductive current,I,, through the dampening circuit, i.e. by achievingIL IC (fault)This realizes a near-zero or very small fault current throughthe capacitors and protects them besides restoring thefault level of the system to its original level.(ii) Connecting a spark gap across the capacitor banksIf a breaker alone is used to bypass the capacitors duringa line fault it may fail to discharge its required duty due

Lightningarrester -a, Overhead bus barsCable boxsupportingframetransformer (RVT)transformer (RVT)Figure 26.7 Installation of 11 kV capacitor banks and their protective equipmentGantry Isolator Current Circuit Lightning * Series reactor *- Residual voltage Capacitor banktransformers interrupter arrester transformer (RVT)" To limit the inrush and harmonic currents through the capacitor banks*' For unbalance protection and also for capacitor discharge and measurement if requiredFigure 26.8Typical layout of 33 kV power factor improvement system with switchmg and protective equipment

~~ ~~~26/836 Industrial Power Engineering and Applications HandbookFor more groundfault schemesrefer to Ch. 21SurgearrestersILine sideseriesreactor'Ptf'p--Capacitors' discharge V-(open delta)tertiary winding secondary windingI I rVTI11wUngroundedcapacitor banksTo monitor dischargeconditions of thecapacitor banksTo NDRFigure 26.9 Switching and protection scheme for an HT capacitor systemTable 26.2 Permissible momentary overloadings for seriescapacitorsTimes the rated capacitorcurrentDuration110 For 8 h during d 12 h operation135 For 30 min during a 6 h operation150 For 10 min during a 2 h operationNote Provided that the average output in a 24 h operation does notexceed the rated outputto its definite minimum time of closing (Table 19.1).This inherent time delay may be too long to protect thecapacitors during a fault. To overcome this, a spark gapis additionally provided across the capacitors, which willspark-over instantaneously at the preset value and limitthe fault current to the required level during a faultcondition and bypass the capacitors as illustrated in Figure26.11. Extinction of the arc is achieved by the bypassbreaker, which closes soon after (3 to 5 cycles frominitiation of the fault).The excessive line current results in a high voltagedrop across the capacitors (Isc . Xc), which sparks overthe gap generally set at about 2.5 to 3 times the nominalvoltage of the capacitors. This high voltage for anextremely short duration will economize on the cost ofcapacitors. The arc will be sustained only until the faultclears or the line de-energizes or the bypass breaker closes,whichever occurs first. The arc will extinguish at everycurrent zero and attempt to insert the capacitors backinto the line. The arc will reappear until the overvoltageattenuates to less than the spark-over voltage, or the linede-energizes or the bypass breaker closes. The capacitorsmust therefore be suitable to withstand repeated transientvoltages developed during each arcing.A CT is provided in series with the spark gap to senseits operation during a line fault. As soon as there is arcing,it provides an instantaneous command to a short-circuitrelay. The relay, in turn, closes the bypass breaker, within3 to 5 cycles, leaving only the natural line impedance inthe faulty circuit. Now X, = 0, which limits the faultcurrent to the natural level of the system, as if the capacitorswere not connected. The shorting device is restored toits original status as soon as the fault condition is cleared.The device must be capable of interrupting the line fault

~~~~~~~~~ IProtection maintenance and testing of capacitor units 261837/c (fault)@ Series reactor to add toE, line impedanceiving end") Isolators0@ Dampening resistor@ Auxiliary saturating reactor/c(fault)I (fault)An R-L dampening circuit@ Main saturating discharge reactorFigure 26.10Dampening circuit across the series capacitors to limit the fault level

26/838 Industrial Power Engineering and Applications Handbookventilation facilities should also be provided, eitherlouvres, exhaust fans or by providing an expandedmetal enclosure, for the capacitors to ensure a free aircirculation around each capacitor unit (Figure 23.32).Concentration of capacitor banks at one point in asystem may cause amplification of harmonics, selfexcitationof machines, overvoltages due to switchingand unsatisfactory working of audiofrequency remotecontrolapparatus. To overcome this, a concentrationof capacitor banks at one point must be avoided as faras possible. These may be installed at different locationsin the system.A loose contact within the circuit, producing sparks,will generate high-frequency oscillations (harmonics).However small these oscillations are, they may raisethe system voltage and frequency. High-frequencyoscillations will also reduce the impedance of thecapacitors xch a llfh and cause high currents, leadingto premature failure of the capacitors. Loose contactsmust therefore be thoroughly checked. It isrecommended that a capacitor circuit particularly bechecked periodically for loose connections orsparkings.When the reactive control is not automatic, then duringoffpeak periods (such as during the night) it is importantthat some of the banks are dropped manually to avoidan overcompensation and a consequent overvoltage.It is advisable to periodically check the condition (i.e.the capacitance value C) of each capacitor unit todetect failure of one or more of them and the consequentovervoltage across the healthy units. This can bemeasured by any of the methods discussed in Section26.3.1(1). The capacitance value of each unit mustfall within the permissible limits as indicated in thatsection. The failed units, if any, must be replacedpromptly with units of the same rating and brand toavoid any performance variation. If this is not possiblethen the banks must be balanced by readjusting theunits. When the banks are connected through a seriesreactor, the value of the reactor must also be readjustedto match changed parameters. Otherwise the reactormay resonate with the fifth or the seventh harmonic,for which it is designed, and cause an overvoltage.26.2.1 Precautions in handling a capacitor unitwith PCBAlthough the manufacture of PCB-filled capacitor unitshas been discontinued, for older installations some maystill be used. For the benefit of their users we give belowthe precautions that they must take when handling theseunits:Do not inhale the vapour of this chemical.Avoid its contact with eyes and skin.Avoid its contact with birds and animals.Avoid its contact with animal feed.Do not re-use contaminated paper and cloth.In the event of contact with the human body, wash theskin with soap and water and the eyes with clean runningwater.Avoid any spillovers, leakage or vaporization of liquidor disposal of a waste unit. One can use Araldite, solderor any other means to stop a leakage. If this is notpossible, collect carefully all the chemical PCB andcarefully dispose it of.Soaked rags, cloth or papers, must be destroyed in anincineration plant at 1000°C. Disposal at a landfillarea is not advisable for during rain they may be carriedby storm drains into rivers, canals or ponds.Old units should not be sold to a scrap dealer forsimilar reasons.26.3 Test requirementsBelow we discuss briefly recommended tests on a finishedcapacitor unit based on various IEC recommendations.For standards refer to the list provided at the end of thechapter.26.3.1 Routine testsRoutine tests are carried out on each capacitor unit.(I) Capacitance measurementThis is to be determined at rated voltage and frequencyand can be carried out between 0.9 to 1.1 times the ratedvoltage and 0.8 to 1.2 times the rated frequency.Permissible tolerances in capacitance from the declaredvalues can be as follows:For LT capacitor units up to 1000 V For both selfhealingand non-self-healing types:S% to + 10% for units and banks up to 100 kVAr5% to + 5% for units and banks above 100 kVAr.For HT capacitors5% to + 15% for capacitor units or banks containingone unit per phase5% to + 10% for capacitor banks less than 3 MVAr0% to + IO% for capacitor banks from 3 MVAr to 30MVAr0% to + 5% for capacitor banks above 30 MVAr.0 For series capacitorsThese are usually HT capacitor units/banks:f7.5% for capacitor units or banks containing one unitper phase* 5% for capacitor banks less than 30 MVArk3% for capacitor banks 30 MVAr and above.Methods to determine the capacitance of acapacitor unitTwo common methods to determine this are:0 Voltage method and0 Bridge method.In the voltage method the normal voltage is applied acrossthe capacitor terminals and line current, I,, is measured.The value of capacitance, C, can be determined from theequations provided in Section 23.5.2. In the bridge method

DescriptionLT capacitors up to 1000 VRoutine testType testHT capacitors above IO00 VRoutine testType testTest Duration Test Duration Test Duration Test Durationvoltage voltage voltage voltage(i)Voltage test betweenterminals2.15 V: 2s 2.15 V: 10 s 2.15 V: 10s - -(ii) Voltage test betweenterminals and container(a) for rated voltage 5 660 V(b) for rated voltage > 660 V3 kV 10 s 3 kV6 kV 10 s 6 kV 1 minFor lightning impulse withstand voltagerefer to Table 26.4"V, = rated voltageNotes1 For a repeat test, only 75% of the test voltage must be applied.2 The above tests refer to dry power frequency voltage withstand tests for indoor units. Outdoor capacitor units will be subjected to a wettest as in IEC 60060- 1.the capacitance is measured directly through a capacitbcemeter.NoteThe values thus obtained will be accurate when the system is freefrom harmonics. If harmonics are present in the system correctionwill be necessary for both V andf, as in equation (23.5).(2) Measurement of capacitor loss in terms of tan 6Tan 6 is a measure of dielectric loss in a capacitor unitand is represented by the ratio of equivalent seriesresistance and capacitive reactance of a capacitor unit atthe rated voltage and frequency (Figure 9.7) Le.Rtan 6= -xcwhere equivalent series resistance, R, is the virtualresistance which, if connected in series with an idealcapacitor unit of capacitive value equal to that of thecapacitor under reference, will have a power loss equalto the active power dissipated in that capacitor underspecified operating conditions.The test can be performed at voltages and frequencies,as noted in (1) above. The value of tan 6 should notexceed the declared value.(3) Voltage test between terminalsEvery capacitor unit will be subject to an a.c. test at avoltage specified in Table 26.3 at a frequency between15 and 100 Hz, preferably as close to the rated as possible.During the test no permanent puncture or flashover shouldoccur.(4) Voltage test between terminals and containerThe test procedure and test performance should be thesame as noted in (3) above except the test voltages, whichwill be as specified in Table 26.3.(5) Test for the internal discharge deviceThis is required to determine the suitability of theresistance of the discharge device to ensure that itdischarges a charged capacitor unit or bank at fi V, to50 V or less in 1 minute in LT and in 10 minutes in HTshunt capacitor units or banks and in 5 minutes in seriescapacitors. The arrangements of discharge devices areillustrated in Figure 25.8.*(6) Scaling testThis is required to check the leakage of the capacitordielectric during long operations. It is carried out byheating the unit uniformly on all sides, at least 20°Chigher than the maximum ambient temperature of thecapacitor unit, for at least 2 hours. There should be noleakage.26.3.2 Type tests(I) Thermal stability testThis is another name for a heat run test. A voltage sufficientto cause an output of the capacitor unit equal to 1.5times the rated output is applied at the rated frequencyand test conducted for at least 48 hours. During the lastsix hours of the test the temperature rise and tan 6 aremeasured at least four times. The temperature rise duringthe last six hours should not exceed 1"C, otherwise thetest duration must be extended until the temperaturestabilizes to a rise of only 1°C. The final test results*These figures are only indicative, as different countries may specifydifferent discharge voltages and discharge times. For instance, variousIEC Standards mentioned at the end of the chapter specify thedischarge voltage as 75 V, reached in 3 minutes in LT and 10minutes in HT shunt as well as series capacitors. When required forfast switching operations, special discharge devices are sometimesemployed to achieve a faster discharge, as discussed in Section 23.17.

26/840 Industrial Power Engineering and Applications Handbookcompared to the measurement taken at the start of thetest should not vary by more than the following:For capacitance valueFor LT units: 2% of its capacitance valueFor HT units: The two test values should be correctedto the same reference of dielectric temperature, say,20°C or as agreed. The difference between the twotest values should be less than an amount correspondingto either breakdown of an element or operation of aninternal fuqe.For tan 6valueFor LT units: 2 X IO4For HT units: 1 x lo4 or as agreed between themanufacturer and the user.(2) Measurement of capacitor loss in terms of tan 6This is the same as for the routine test(3) Voltage test between terminalsThis is similar to the routine test (4) except the durationof the test, which should be as specified in Table 26.4.(4) Voltage test between terminals and containerThis is similar to the routine test (4) except the durationof the test, which would be as specified in Table 26.4.(5) Lightning impulse voltage test betweenterminals and containerThe test should be conducted generally with a wave of1.2 to 5/50 ps as illustrated below:For LT units: Three impulses should be applied ofpositive polarity followed by three of negative betweenthe terminals joined together and the container.For HT units: Fifteen impulses should be applied ofpositive polarity followed by fifteen of negative betweenthe bushings joined together and the container.The value of the first peak of the test wave is specifiedin Table 26.4. During the test there should be no ruptureor flashover at any part of the unit, which should beverified by a cathode ray oscillograph. An oscillographshould also be used to record the voltage and check thewave shape.(6) Short-circuit discharge testThe unit should be first charged by d.c. and then dischargedthrough a gap situated nearby. The test voltages shouldbe as follows.For LT unitsThe testshould be carried out between any two terminals, leavingthe third open.Delta-connected units Test voltage 2V,. Any twoterminals of the unit should be shorted and the testconducted between the third and the shorted terminal.Star-connected units Test voltage 4 V, /a.For HT unitsThe test voltage should be 2.5 V,, irrespective of star- ordelta-connected units. The capacitor unit should be chargedwith the test voltage and discharged five times within 10minutes. Within 5 minutes after the test the unit may besubjected to voltage test (3) between the terminals. TheHighest vulruge forRated lightning impulse withstundRated power-frequency short-durcition withstand volttrge forDurutionRoutine test 1 Type testup to 0.660.66 to 1.01.22.43.67.21211.524364060759s145101s~.Notes1 For a repeat test only 7S% of the test voltage must be applied.2 The power frequency voltage test refers to indoor units. The outdoor capacitor units must be subjected to a wet test as in IEC 60060-1.3 The choice between list I and list 2 will depend upon the extent of exposure of the units to the internal and external voltage surges andthe amount of surge protection, if provided.4 It will also depend upon the condition of system grounding, i.e. whether effectively grounded. non-effectively grouned or isolated neutraletc. For systems other than effectively grounded, list 2 must be preferred.S For higher voltage systems, refer to the relevant Standards noted in the table of Standards provided at the end of the chapter.

Protection, maintenance and testing of capacitor units 26/841capacitance of the test unit may be measured before thedischarge test and after the voltage test. The capacitancevalues thus measured should not be exceeded by morethan 2% or breakdown of an element or blowing of aninternal fuse.(7) Partial discharge test (applicable to HT units)There are invariably voids that will appear in an insulatingsystem as a matter of process, irrespective of the typeand procedure of impregnation adopted. as discussed inSection 9.6.1. For all practical purposes, a dielectric systemmay therefore be considered to be an imperfect capacitor.All such voids may cause internal discharges and lead toerosion and ultimate failure of the dielectric under theeffect of the applied voltage. The applied voltage maybe fluctuating too. aggravating the situation. This test isconducted to check the condition and suitability of thedielectric used to perform the required duties in servicewithout causing an undue discharge (ionization). It is ameans of detecting defects in the insulating system of acapacitor unit and is complementary to a dielectric test.To perform this test, the rated voltage may be appliedlong enough for thermal equilibrium to be achieved. Thena test voltage of twice the rated voltage is applied onlyonce for one second. The voltage is then reduced to 1.2times the rated voltage and maintained for 10 minutesand then it is raised again to I .5 times the rated voltagefor 30 minutes. During the last 10 minutes, the dischargeunder such conditions should be less than SO pc (picocoulombs).The capacitance should be measured bothbefore and after the test. The change in capacitance shouldnot exceed 2%.One will appreciate that such tests are no guarantee ofthe minimum life of a capacitor unit. Moreover, it is notexpected that such tests can predict the future of thecapacitor units. They can, at best, suggest the complianceof the test requirements which should ensure a reasonablyprolonged operating life of the capacitors, as envisaged.These tests do provide feedback to the manufacturer onthe quality of the dielectric and the process of insulationadopted and if any improvements are required.(8) Endurance test (process test for HT capacitors)This is an alternative to the ageing test and is applicablefor shunt capacitors of 1000 V and above. It is an inhouseprocess test and is carried out on the capacitorelements before they are assembled into a capacitor unitto ascertain their dielectric stability against repeated overvoltages.For the test procedure refer to IEC 6087 1-2.(9) Ageing test (applicable to LT units)This test is conducted to estimate the working life of acapacitor. Briefly. the test can be conducted in three stagesas follows:(a) The capacitor is energized at 1.25 times the ratedvoltage for 750 hours.(b) It is then subjected to 1000 discharges.(c) After test (b). the test in (a) is repeated.On conclusion of the test no permanent breakdown orflashover should be observed. The variation in thecapacitance value as measured before and at the end ofthe test should not vary by more than 3%. The capacitorunit should withstand the voltage test between theterminals and the container as stated in test (4) aboveand the scaling test as in test (6) under routine tests.(IO) Self-healing testThis test is applicable to LT capacitors only. which arethe self-healing type. The capacitor is subjected to avoltage of 2.15 V,, for 10 seconds. If fewer than fivebreakdowns occur during this time. the voltage may beincreased slowly until it reaches 3.5 times its rated value,or five breakdowns, whichever happens first. If fivebreakdowns have still not occurred, the test is continueduntil they do. The capacitance measured before and afterthe test should show no significant change.(1 I) Destruction testThis test is also applicable to LT capacitor units only.The basic objective is to establish that the failure of aunit is within a safe zone and is not accompanied by therisk of a violent explosion or fire.These are only brief descriptions of the recommendedtests and their procedures on a completed capacitor unit.For more details refer to the table of Standards providedat the end of the chapter.26.3.3 Checking field operating conditionsOn larger installations, particularly. it is recommendedthat the actual operating conditions he checked andcompared with the assumed conditions for which thccapacitor units were designed to ensure that thcy do notdiffer by more than the permissible limits. This is becausea capacitor is highly susceptible to some operatingconditions such as fluctuations in voltage and harmoniccontents in the system. An excess of any of these or bothV and V, may shorten the life of the capacitor units andcause frequent failures of individual units. It may alsolead to overvoltages across the units and incrcasc harmoniccontents. All this may cause failure of the capacitor units,leading to even a cascade failure. increasing the unbalancein voltage and kVAr across each phase and destabilizethe cntirc capacitor nctwork.The effect of voltage and voltage fluctuations can hechecked through the capacitor line current. I,, which is ameasure of the capacitance C of the capacitor units. Thecapacitor current must correspond to the designed (rated)current plus the permissible variation. A higher currentthan this would mean a voltage variation or the presenceof harmonics in the system or both. While the voltagemay be measured through a voltmeter and the currentthrough an ammeter in the three lines, the harmonicscontent may be checked by an oscilloscope. If the actualoperating conditions differ from the designed parametersby more than is permissible. suitable measures will beessential to restore the operating conditions within thepermissible limits. say, by providing series reactors orfilter circuits in the system as discussed alrcady.

~~~General~~261842 Industrial Power Engineering and Applications HandbookRelevant StandardsIEC60060- 11198960060-2/ 199460071- 1/199360143-1/199260143-2/199460831- 1/199660831-2/199560871-1/199760871-2/19876093 1 - M99660931-2/1995Title IS RSHigh voltage test techniques ~ definitions 2071-1/1993 BS 923-1/1990and test requirementsHigh voltage test techniques. measuring systems 207 1-2/1991 BS EN 60060-2/1997207 1-3/1991Insulation coordination. Definitions, principles and rules 2 165- 1/1991 BS EN 60071-1/1996Series capacitors for power systems9835/199 1 BS EN 60143-1/1993General performance, testing and rating. Safetyrequirements. Guide for installationSeries capacitors for power systems -BS EN 60143-2/1995Protective equipment for series capacitor banksShunt power capacitors of the self healing type forI3340/1993 BS EN 60831-1/1998a.c. systems having rated voltage up to and including1000 vGeneral performance, testing and rating. Safetyrequirements. Guide for installation and operationShunt power capacitors of the self healing type for13341/1992 BS EN 60831-2/1996a.c. systems having rated voltage up to and including1000 vAgeing test, self-heating test and destruction testShunt capcitori for ax. power system having ratedBS EN 60871-111998)voltage above 1000 V (non-self-healing type) }13925-1/1998General performance, testing and rating ~ Specialrequirements - Guide for invtallation and operationShunt capacitors for a.c. power system rated voltage13925-2/1994 BS 7264-211990above 1000 V (non-self-healing type)Endurance testingShunt power capacitors of Iht: non-\clf healing type13585- 1/1994 BS EN 6093 I - I/1998up to and including 1000 VGeneral, performance, testing and rating. Safetyrequirements. Guide for installation and operationShunt power capacitors of non-self healing type up13585-2/1998 BS EN 60931-2/1996to and including 1000 VAgeing test and destruction testRelevant US Standards ANSUNEMA and IEEEIEEE 4/1995IEEEI 3 13.1/1996ANSUIEEE18/1993ANSHEEEC 37.90.1/1989NEMA-CP1/1992NEMA CP-9/ I992NEMAStandard techniques for HV te\tingInwlation coordination, Definition\, principle\ and rulesStandard for shunt power capacitorsGuide for surge withstand capability tests, for protective relays and relay systemsShunt capacitors, both LT and HTExternal fuses for shunt capacitor5Probability curvesNotcsI In the tables of Relevant Standards in this book while the latest editions of the Standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of Standards is a continuous process bydifferent Standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.23Some of the BS or IS standards mentioned against IEC may not be identicalThe year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

Protection, maintenance and testing of capacitor units 26/843List of formulae usedCurrent unbalanceVoltage unbalanceN' Vph= 3N2. N, - 3 N. N, f 2N(26.1)V,, = Unbalanced voltage across the open delta. This isfor grounded systems. For isolated systems, it willbe three times thisN = number of units failedN, = number of units in series per group per phaseN2 = number of units in parallel per group per phase(26.2)I, = Unbalanced current through the interconnectedneutralI, = current per capacitor unitFurther readingAIEE Committee, Rrport of (2 Sunq on tlw Coniirc.tiorr of ShlrfitCapacitor Banks, Report, Vol. 77, pp. 1452-58, Feb. (1959).Central Board of Irrigation and Power, India, Mcinucrl of! SkirrirPower Capacitors - Operritiorz and Mriintennnc~~. Technical ReportNo. 33, Jan. (1984).Longland, T., Hunt, T.W. and Brecknell. W.A. I'oiivr Cop

Power Reactors271845

Power reactors 27184727.1 IntroductionPower reactors are similar to transformers. However, theyhave only one winding per phase and can be representedas shown in Figure 27.1. They are employed to performa number of functions, primarily to control and regulatethe reactive power of a power system by supplying theinductive and absorbing the capacitive power. Controlcan be achieved in different ways as noted later. Thereactors, depending upon their design and I-@characteristics. can be classified as follows:Single- or three-phase Single-phase reactors areused in the neutral circuit either to limit the groundfault currents or as arc-suppression coils (Section 20.5).Similarly, three-phase reactors are used for three-phaseapplications.Air-cooled dry type and oil-immersed type Thiswill depend upon the size of the reactor and the designof the manufacturer. The latest practice is to use aircooleddry type, which call for lesser maintenanceand are free from any fire hazards.Indoor or outdoor types These may be designedindoor or outdoor types depending upon the application.4 Tap-changing facility Where necessary, the reactanceof the coil can be varied by providing an on- or offloadtap-changing gear with the reactor, similar to a powertransformer.27.2a Selection of power reactorsWhen it is required to limit the inrush current a fixedreactance (linear) reactor is more suitable. A variabletype reactor will be necessary when it is to be used forvoltage regulation or load sharing. In circuits whereharmonics may be present, saturated type reactors maybe preferred.The harmonic content may be measured throughharmonic analysers and expressed as a percentage of thefundamental component. The current and voltage ratingsof the reactors will depend upon their application. Aseries reactor connected permanently in the circuit, forinstance, will be rated continuously and for full systemvoltage, whereas a reactor used in the ground circuitmay be short-time rated and rated for the likely maximumground fault current.27.21, Magnetic characteristicsThe magnetic characteristics of an inductor coil will varywith the type of its configuration as discussed below. Itcan have one of the following shapes:LinearNon-linearSaturated(Figure 27.2(a))(Figures 27.2(bl) and (b2))(Figure 27.2(c))Figure 27.1General representation of a reactorA reactor can be designed to provide any of thesecharacteristics to meet the different reactive powerneeds.Saturationknee-point- Peak current + Peak current Peak\ Non-saturatedregion1 20 1 2 3 0 1 2 3 4 0 1 2 3current + Peak current-(a) Linear (b,) Non-linear (b2) Non-linear (c) Saturated@ Saturation knee-point for curve-I@ Saturation knee-point for curve-2Figure 27.2Magnetic characteristics of a reactor

27/848 Industrial Power Engineering and Applications Handbook27.3 Design criterion and I-+characteristics of differenttypes of reactor27.3.1 Air-cored or coreless type reactorsThese reactors are made of copper winding without anycore, similar to an air-cored solenoid, as shown in Figure27.3. In the absence of an iron core it causes a largeamount of leakage flux in the space, which may alsoinfringe with the metallic tank housing the reactor, andaffect the reactance of the coil, in addition to heating thetank itself. It is therefore important to provide some kindof shielding between the winding of the reactor and thetank. The shielding can be magnetic or non-magnetic asdiscussed later. With shielding, the characteristics of anair core reactor can be altered according to its application.Such reactors provide linear I-$ characteristics in theoperating range, say, up to 150% of the rated current asshown in Figure 27.2(a). In the absence of an iron core,there is no saturation of the core. These reactors aremore useful when they are required to be used as currentlimiting devices. But they reduce the steady-state powertransfer capability (V,’ / Z) of the system, as discussed inSection 24.8.With magnetic shieldingA magnetic circuit develops a stray magnetic field aroundit. A reactor, which is a magnetic circuit, at higher currentssuch as during switching operations or faults will develophigh magnitudes of stray fields around it that may linkmagnetic objects in the vicinity and cause high inducede.m.f.s in them, as in nearby metallic structures andelectrical apparatus. All this may result in high circulating=1,MS tankGroundingterminalMagnetic,’- shieldA Copper woundcoil,- Pressing platey Tie rodt-,- Copper woundcoilM.S tankIFigure 27.4IISectional view of a magnetically shielded reactorFigure 27.3reactorWindingsThree-phase ‘air-cored’ magnetically shieldedcurrents and consequent heating. In addition, it may alsoaffect the working and performance of the measuringand indicating instruments connected on the system. Toreduce such an effect a magnetic frame made of steellaminations and rigidly clamped to suppress vibrationsand noise (magnetostriction effect) is provided aroundthe inductor coil, as illustrated in Figure 27.4. This framewill arrest most of the space magnetic field within theclose vicinity of the reactor. The field produced by thecoil will link the iron frame and would be almost usedup in magnetizing it. The steel frame is called the magneticshield. The self-inductance of the coil, L, is now muchless affected. The flux density of the core is designed sothat it does not saturate up to 150% of the rated current.

Power reactors 271849Any increase in the fundamental value of the currentbeyond 1 50470, or a voltage drop across the coil of morethan 150% of the reactor voltage (this may occur in thepresence of harmonics) may, however, saturate the coreand reduce the reactance of the coil. Magnetically shieldedreactors therefore have limitations when the systemharmonics are high or when linear V-I characteristicsare desirable beyond 150% of the rated fundamentalcurrent.The need for a magnetic shielding is greater in highcurrent reactors than in smaller ratings. For more detailson magnetic shielding see Section 28.2.2 on segregatedphase bus systems.With non-magnetic shieldingIn non-magnetically shielded reactors a cylindrical shieldof non-magnetic material, such as aluminium or copper,is provided around the inductor coil instead of a magneticmaterial (Figure 27.5). Since there is no iron path for themagnetic field, the coil now may not maintain a constantinductive reactance, as in the case of magnetic shielding.Instead. it may become reduced with an increase in thecurrent due to a counterfield generated in the coil by thenon-magnetic shielding.KGrounding terminali:- M.S Tank- Non-magnetic shield- Spacer- Copper wound coil- Pressure plateFor the theory of neutralization of the magnetic effecton the conductor in a non-magnetic shielding, refer tothe continuous enclosures for isolated phase bus systemsdiscussed in Section 31.2.2. As a result of non-magneticshielding there will be no saturation of the iron core andthe V-I characteristic of the reactor will remain almostlinear.These types of reactors can now be used as currentlimiting reactors and also as harmonic suppressors. Theyare also recommended for capacitor application due totheir linear characteristic which will not disturb the tuningof the filter circuit.27.3.2 Gapped iron core or saturatedtype reactorsReactors of this type, as shown in Figure 27.6, tend tosaturate at lower currents (Figure 27.2(c)). The currentdrawn by them is too low, even up to the saturation level,due to high leakage reactance which can increase to 100%.They therefore provide a high inductive impedance initiallywhich becomes stabilized with saturation of the core.After saturation, the I-$ characteristic becomes almostconstant or flat. Such reactors thus have non-linearmagnetizing characteristics and the current drawn by themcontains many odd harmonics. When the reactor is to beconnected on a power system these harmonics must besuppressed as much as possible through filter circuits orby using a multi-limb core arrangement, such as the sixornine-limb zig-zag arrangement illustrated in Figure27.7. Beyond the point of saturation, the rise in currentis rather fast compared to a small rise in the magnetizationor the voltage.Unlike an air core, the inductor coil now has an ironcore that may be provided with air gaps or non-magnetic-Steel/IT laminations, Air gaps,,- M.S. tankNon-magnetic shield/ Note/ It becomes air core whenthere is no iron coreFigure 27.5 Sectional view of non-magnetically shielded reactorFigure 27.6 A gapped iron three-phase reactor

27/850 Industrial Power Engineering and Applications Handbook\\\Figure 27.7 Six- or nine-limb zig-zag arrangement ofwindings to limit the harmonicsLimit the switching surges as discussed in Section23.5.1. But they may affect the steady-state powertransfer capability of the system (V,2/Z). Refer toreactive power control (equation (24.10)).Adjust the steady-state voltage control by supplyingreactive power and compensating the capacitivecontent.Suppress the harmonic contents.Their ratings can be calculated byV2 R=- Wseparators in between to reduce the iron content andhence the induced magnetic field. The 1-4 characteristiccan thus be varied as required by altering the gap, i.e. theiron content in the core. They are suitable as currentlimiters, and can also limit the occurrence of overvoltages.Where required, they can be provided with a tap-changingfacility to regulate their reactances. Likely applicationsare:Voltage stabilization and control of temporary overvoltages.Flicker control in industrial supplies through V =L (dildt)(Section 6.9.4).27.4 ApplicationSome of the applications where a power reactor can beused to provide a reactive support or compensation toimprove the quality of a power system are noted below.27.4.1 Shunt reactors or compensating reactorsThese are meant for parallel connections to absorb thereactive power (capacitive current) of the system and aregenerally used on transmission and large distributionnetworks, as shown in Figure 27.8. They may have afixed or variable reactance, rated continuously, and anyof the magnetic characteristics as illustrated in Figure27.2. Broadly speaking, they can perform the followingfunctions:XL =V2VAr = I* . XL(neglecting R)whereR = reactor resistance - WphaseX, = reactor reactance - WphaseV = rated voltage of reactor - voltsW = reactor loss - wattdphaseVAr = rated output - VAr/phaseI = rated current of reactor - A27.4.2 Current limiting or series reactorsThese are connected in series in a circuit, as shown inFigure 27.9 and are meant to limit the high inrush current,such as during switching of HT capacitor banks (Section23.10). They may also be used to limit the currents underfault conditions by adding to the circuit impedance tomatch with the breaking capacity of the interrupting devicewhen the fault level of the system may exceed this. Somereactor connections are illustrated in Figure 27. IO. Theyare also used for load sharing of two power systems.They are connected in the circuit permanently and mayhave a fixed or variable reactance, rated continuouslyand can be made to have linear (fixed reactance) or nonlinearmagnetic characteristics as required. When theyare required to limit the inrush currents, fixed reactance,linear reactors should be preferred. During a faultcondition, the reactance of the reactor should not diminishdue to the saturation effect. This is an essential requirementto limit the short-time fault currents. Ideally, currentAJ- I I V ILumped or leakage capacitanceFigure 27.8 Use of a shunt reactor to compensate for thereactive powerFigure 27.9 Use of current limiting reactor, (1) to limit thefault current, or (2) to limit inrush current during a capacitorswitching

Power reactors 271851cT(a) Feeder connections(b) Tee-off connectionsFigure 27.10 Reactor connections(c) Busbar sectionalizer connectionslimiting reactors must have no-iron circuit (air core orcoreless type). The iron core type provide non-linearsaturating type characteristics, and at overcurrents havea tendency to diminish their reactance due to the saturationeffect, while the reactors are required to offer highimpedances to limit the fault currents. The coreless typewill provide a near-constant reactance at all currents dueto the absence of an iron core and hence, their preferenceover other types for such applications.Similarly, gapped iron core reactors as shown in Figure27.4, in which the iron core content is reduced by providingan air gap or non-magnetic material between the corelaminations, also raise the saturation level (the coreremaining unsaturated, Figure 27.1 1) to help provide anI0 1 2 3Peak current in the --t4faulty regionFigure 27.11 Typical characteristics of a current limitingreactor (coreless, gapped iron core or magnetically shieldedcore type)Iadequate impedance on fault. They may also be employedas a current limiter, such as shown in Figure 27.9. Itmust, however, be ensured that the voltage available acrossthe load does not fall below the permissible level as aresult of the voltage drop across the reactors. The reactorsfor such applications should be continuously rated.For Figure 27.9 the rating of the series reactor can bedetermined byVAr' V, Ix, = -"' Rlphase43.1whereVAr = rating of the series reactor (Volt. Amp)V, = rated voltage drop of the reactor phase to phasein volts= voltage induced in the reactor when operatingat rated current and rated reactanceI = rated current of the reactor (A)R = resistance of the reactor.. R

27/852 Industrial Power Engineering and Applications HandbookSize of a three phase reactor Size of a single-phase reactorVAr= &xl.Ox650= 1125 kVAr1 x 1000and X,/phase = ~fi x 650= 0.89 Wphase27.4.3 Dampening reactors650 x 1 .OVAr = ~6= 375 kVArThese are meant to limit the inrush currents occurringduring a switching operation of a capacitor. They areconnected in series with the capacitors and may be shorttimerated for the values of the inrush currents andcontinuously rated for normal line currents. They arealmost the same as the series reactors with fixed reactance.27.4.4 Neutral grounding reactorsThese are meant to limit the ground fault current and areused between the neutral of the system and the ground.They are single-phase and may be short-time rated,otherwise they are the same as the current limiting reactors(Figure 27.12). Their ratings can be calculated bywhereX, = reactance - RVI = rated voltage of the system(5-1rated voltage of the reactorIg = current through the ground (rated current ofthe reactor)X, = zero sequence reactance (or impedance) - R= 3.x,27.4.5 Grounding transformer or neutralcouplersThese are meant to provide a neutral to an ungroundedsystem.When the ground transformer neutral is connected toBYthe ground directly or through a current limiting reactorits neutral current may be considered for a short-timeduration only, i.e. until the ground fault exists assumingthat the ground fault protective scheme will isolate thefaulty circuit promptly.But when the neutral is grounded through an arcsuppressioncoil (reactor) the current through the groundedneutral may be of a limited amplitude, say, up to itscontinuous rating (Section 20.5) and it may exist forlonger.These transformers are three-phase and may beconnected for zig-zag or staddelta connections (Section20.9.1). The delta may also be made open type by insertinga resistor across it to help adjust the zero-sequenceimpedance, if required.27.4.6 Arc suppression or Petersen coil (reactor)These are meant to compensate the ground capacitivecurrent on a ground fault in the system, which may begrounded naturally or artificially (Section 20.5). Theyare connected between the neutral of the system and theground and are single-phase and may be short-time orcontinuously rated, depending upon the systemrequirement. If it is being used as a ground fault neutralizerit may have to be continuously rated. It may be of variabletype to help tuning with the system ground capacitance.27.4.7 Tuning or filter reactorsThese are meant to be used with a capacitor to tune afilter circuit, with resonances in the audio frequency rangefor reducing and filtering the harmonics or communicationfrequencies. They provide a near short-circuit for therequired harmonics to filter them out of circuit. Theymay be single-phase or three-phase and connected inseries or parallel of the capacitor circuit and may have afixed or variable reactance, rated continuously withsaturated magnetic characteristics. They may incur heavylosses.27.4.8 Smoothing reactorsThese are meant to provide high impedance to harmoniccurrents and block their entry or reduce their amplitudesand are therefore also known as blocking reactors. Theymay have any of the magnetic characteristics shown inFigure 27.2 and have a fixed reactance, rated continuously.Example 27.2Consider a distribution system as illustrated in Figure 27.1 3(a)being fed by two power sources in tandem:Figure 27.12 Neutral grounding reactor or a Petersen coil Fault MVA = baseZe,A transformer 5000 kVA, 33/11 kV, having a reactance of7.15% and connected to a power grid.A generator 2500 kVA, 11 kV, having a steady-statereactance, xd of 25 %.A three-phase fault somewhere in the bus system, withoutreactive compensation and ignoring the line impedance, canreach a level of(equation(l3.6))

Transformer5000 kVA, 33/11 kVZ,, = 7.15%(Reactive)81T (Reactive)Grid busI IGenerator2500 kVA, 11 kV4, = 25%Z,, = 7.15%(Reactive)Power reactors 27/853z,, = 50%(Reactive)Seriesreactor x%3-phase loads(a) Location of faultf Fault(b) Equivalent circuitFigure 27.13 Application of a series reactorwhere qq = equivalent p.u. impedance of the combined powersystem at the point of fault.Consider the base level as 5000 kVA, 11 kV:. Generator reactance at the new base~ 2 55000 ~ -(equation (13.1))2500= 50%and1=- +-1Z,, 0.0715 0.5or Zes = 0.0625:. Fault level = - 5000 (equation (13.6))0.0625= 80 MVAIf the fault level is required to be limited to 50 MVA, a seriesreactor of reactance, XY0 may be introduced into thetransformer circuit, as illustrated in Figure 27.13(b),The new- 5000eq - 50 x 1000= 0.1 p.u.+- 1and the new reactance - 1=0.1 0.0715 + X 0.5oror0.1 =(0.0715 + X ) (0.5)0.571 5 + XX = 0.0535 p.u. or 5.35 YOat 5000 kVA base and 11 kV:. A series reactor of reactance 5.35% may be used to limitthe fault to the required level.Relevant StandardsZEC Title zs BS60289/1988 Reactors 5553 (1 to 8) BS EN 60289/1995Relevant US Standards ANSVNEMA and IEEEIEEE 62/1995Guide for field testing of electric power apparatus insulation. Part I - oil filled reactorsNotes1 In the tables of Relevant Standards in this book while the latest editions of the Standards are provided, it is possible that revised editionshave become available. With the advances of technology andor its application, the updating of Standards is a continuous process bydifferent Standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.

PART VBus Systems

281857Carrying PowerThrough MetalenclosedBusSystems. -onseauence of theher reading 28/902

Carrying power through metal-enclosed bus systems 28/85928.1 IntroductionIn a power-generating station power is carried from thegenerator to the power transformer, to the unit auxiliarytransformer (UAT) or to the unit auxiliary switchgear asillustrated in Figure 13.21 through solid conductors (HTbus systems). This is due to large capacity of the generators(upto 1000 MW). The transmission of such large amountsof power over long distances is then through overheadlines or underground cables.Similarly, for a distribution system of 3.3, 6.6 or 1 IkV and even higher such as 33 or 66 kV, feeding largecominercial or industrial loads, the distribution of poweron the LT side (Figure 28.1) may be through cables orsolid conductors (LT bus systems), depending upon thesize of the transformer. The HT side of the transformermay also be connected through cables or the HT bussystem as illustrated.For moderate ratings. say. up to 600/800 A, cables arepreferred. while for higher ratings, (1 000 A and above)the practice is to opt for solid conductors (LT bus systems),on the grounds of cost, appearance. safety ease of handlingand maintenance. For larger ratings, more cables maybecome unwieldy and difficult to maintain and maypresent-problems in locating faults. The conductors usedare generally of aluminium. though sometimes the useof copper may be more appropriate in highly corrosiveareas.In humid and corrosive conditions. aluminium erodesfaster than copper. These solid or hollow conductor5connect the supply side to the receiving end and arecalled bus ducts. They may be of the open type, such a sare used to feed a very high current at cery low voltage.A smelter unit IS one such application. But normallythey are housed in a sheet metal enclosure. See Figures28.2(a) and 28.33b).Our main concern here will be dealing with large tovery large currents, rather than voltages. Currents aremore difficult to handle than voltage5 due to mutualinduction between the conductors and between theconductor and the enclosure. Here we briefly discuss thetypcs of metal-enclosed bus systems and their designparameters. to select the correct size and t) pe of aluminiumor copper sections and the bus enclosure for a requiredcurrent rating and voltage system. More emphasis is givento aluminium conductors rather than copper. as they aremore commonly used on grounds of cost.28.2 Types of metal-enclosed bussystemsA bus system can be one ofthe following types. dcpcndingupon its application:Non-segregatedSegreg'ttedIsolated phaseRising main5 (vertical bus system\)Overhead bus (horizontal bus sy\tcm)_I_Bus duct28.2.1 A non-segregated phase bus systemIn this construction all the bus phasex are housed in onemetallic enclosure, with adequate spacings between themand the enclosure but without any barrier5 between thephases (Figure 28.2(a)).ApplicationBeing simple, it is the most widely used construction lora11 types of LT systems.Nominal current ratingsThe preferred current ratings may follow series K- IO ofIEC 60059 and as discussed in Section 13.4.1(3). Theymay increase to 6000 A or so. depending upon theapplication as when required to connect a large LTalternator or the LT side of a large transformer to itsswitchgear. The preferred short-time ratings may be oneof those indicated in Table 13.7.LT power control centre (PCC)* 11 kV breaker for isolation and protection of transformer andinterconnecting cablesFigure 28.1Application of a bus system28.2.2 A segregated phase bus systemIn this construction all the phases are Iiou.secl in onemetallic enclosure as earlier, but with a metallic barrierbetween each phase. as illustrated in Figure 28.2(b). The

28/860 Industrial Power Engineering and Applications HandbookFigure 28.2(a)Low rating LT bus ductNon-magnetic or Barriers (same metal Clamps (same metalmagnetic enclosure as the enclosure) as the enclosure)Figure 28.2(b)A segregated phase bus systemmelallic barriers provide the required magnetic shieldingand isolate the busbars magnetically from each other,rather like an isolated phase bus system (IPB). For moredetails see Section 3 1.2. The enclosure can be of MS oraluminium and the barriers can be of the same metal asthe enclosure. The purpose of providing a metallic barrieris not only to shroud the phases against short-circuits butalso to reduce the effect of proximity of one phase on theother by arresting the electric field produced by the currentcarryingconductors within the barrier itself. It nowoperates like an enclosure with an interleavingarrangement (Section 28.8.4) balancing the fields producedby the conductors to a great extent and also allowingonly a moderate field in the space, as in an IPB system(Section 31.2). The enclosure losses with such anarrangement may fall in the range of 6M5% of conductorsin case and 30-35% in aluminium enclosures for allvoltage systems 3.3-1 1 kV and current ratings above3000A and up to 6000A or so. Only aluminium enclosuresshould be preferred to minimize losses and enclosureheating. The effect of proximity is now almost nullifiedas is an imbalance in the phase reactances. An unbalancein the reactance is otherwise responsible for a voltageunbalance between the three phases as discussed in Section28.8.2 and enhance the electrodynamic forces that maylead to a phase-to-phase fault at higher rated currents.ApplicationThey are generally used for higher ratings, 2000 A andabove, on all voltage systems. They are, however, preferredon an HT rather than an LT system for reasons of cost,such as between a unit auxiliary transformer (UAT) andits switchgears and a station transformer and its switchgearsas in a power-generating station and shown inFigure 13.21.NuteFor such ratings, enclosure of non-magnetic material alone isrecommended due to high iron losses in a magnetic material.Nominal current ratingsThese will depend upon the application. The preferredratings may follow series R-10 of IEC 60059, as describedin Section 13.4.1(4). They may increase to 6000 A or sodepending upon the application.

28.2.3 An isolated phase bus (IPB) systemThe design criteria and construction details of this systemare totally different from those of a non-isolated phasebus system. This type of enclosure is therefore dealtseparately in Chapter 3 1.28.2.4 A rising mains (vertical bus system)For power distribution in a multi-storey buildingThis is another form of a bus system and is used inRigid coupling straightthrough joint-'f-SMC/DMCsupportsOutgoing tap-offplug-in boxCarrying power through metal-enclosed bus systems 28/861vertical formation to supply individual floors of a highrisebuilding (Figures 28.3 (a) and (b)). This is muchneater arrangement than using cables and runninginnumerous lengths to each floor which may not only beunwieldy but also more cumbersome to terminate. Sucha system is the normal practice to distribute power in ahigh-riser. It rises from the bottom of the building andruns to the top floor. To save on cost, the ratings may bein a decreasing order after every three or four floors, asafter every floor the load for that floor will be reduced.rlTo upperfloorsO/G distribution board (DB)to cater for each floorGlasswool fireproof barrier 1Floor thicknessYIFloor slab -,ib *Mounting clampsRising mainsThrust pad andstoppers of rubber -or similar materialBus conductorFlexible coupling -expansion joint(a) Front view with cover removedI/C plug-in boxto feed the busIncoming switchfuse MCCB boxdCable boxHigh riser wall on whichthe bus system runsIncomingcableFigure 28.3 Rising mains mounting arrangementSide view(b) Rising on wallGF

28/862 Industrial Power Engineering and Applications Handbook(say, after every 7.5-10 m) to absorb the expansionof busbars on load.28.2.5 An overhead bus (horizontal bus system)Unlike a high riser, now the overhead bus system runshorizontally, below the ceiling at a convenient height, asshown in Figure 28.4(c) to distribute power to light andsmall load points. A large tool room or a machine shopare installations that would otherwise require a distributionsystem, for short distances, to meet the needs of variousload points and make power distribution unwieldy andcumbersome. Moreover, it would also mean running manycables under the floor to feed each load point. In anoverhead busbar system, the power can be tapped fromany number of points to supply the load points just belowit through a plug-in box similar to that used on a risingmains. The floor can now be left free from cables andtrenches.28.3 Design parameters and serviceconditions for a metal-enclosedbus systemDB with MCCB Typical DB with HCC fusesFigure 28.3(c)The rating can be grouped for three or four floors together,depending upon the total load and the number of floors.A smaller rating of, say, 200-400 A need not be furtherstepped for it may not be of any economic benefit.Special features of a rising mainsThey are manufactured in small standard lengths, say,1.8-2.5 m, and are then joined together at site to fitinto the layout.Wherever the rising mains crosses through a floor ofthe building, fireproof barriers are provided as shownin Figure 28.3(b) to contain the spread of fire to otherfloors.On each floor an opening is provided in the risingmains to receive a plug-in box (Figure 28.3(b)) totap-off the outgoing connections and to meet the loadrequirement of that floor. The plug-in box can normallybe plugged in or withdrawn from the live bus withoutrequiring a shutdown.To take up the vertical dynamic load of busbars andto prevent them from sliding down, two sets of thrustpads are generally provided on the busbars in eachstandard length of the rising mains, as illustrated inFigure 28.3(a).Flexible expansion joints of aluminium or copper areessential after every three or four standard lengths28.3.1 Design parametersA bus system would be designed to fulfil the followingparameters.RatingA bus system, like a switchgear assembly, would beassigned the following ratings:Rated voltage: the same as that assigned to the associatedswitchgear (Section 13.4.1( 1))Rated frequency: the same as that assigned to theassociated switchgear (Section 13.4.1(2))Rated insulation level(i) Power frequency voltage withstand - see Section32.3.2(ii) Impulse voltage withstand - for bus systems of2.4 kV and above see Section 32.3.3Continuous maximum rating (CMR) and permissibletemperature rise: this is the maximum r.m.s. currentthat the bus system can carry continuously withoutexceeding temperature rise limits, as shown in Table32.3. The preferred current ratings of the bus systemwould follow series R-10 of IEC 60059, as shown inSection 13.4.1(4).Rated short-time current rating: this is the same as forthe system to which it is connected, and as assigned tothe associated switchgear (Section 13.4.1(5)). Theeffects of a short-circuit on an electrical system arediscussed below.Rated momentary peak value of the fault current: thesame as assigned to the associated switchgear as inTables 13.11 or 28.1. See also Section 13.4.1(7).Duration of fault: the same as assigned to the associatedswitchgear (Section 13.4.1(6).

Carrying power through metal-enclosed bus systems 28/863Plugged in positionOverhead busbar system in a machine shopWithdrawn positionFigure 28.4(a)Plug-in tap-off boxInstallation of overhead bus system with tap-off boxes in a large assemblyshopFigure 28.4(c)Figure 28.4(b)An overhead bus system shown with tap off boxes (Courtesy: GE Power)

~ ~~~ ~~~~~28/864 Industrial Power Engineering and Applications HandbookTable 28.1 Momentary peak (maximum r.m.s.) current ratings, asymmetrical, for switchgear and metal-enclosed bus systems,based on ANSI-C-37/20CNominal voltage Rated current (Ir)kV(Km.s.) A"Non-begregated phase cjstenz Segregated phase sxstem Icolrited phase osteiiikAh kA" kA0.6 I6000.6 30000.6 4000 to 60004.16-13.8 1200 to 300014.4 I200 to 20 00023-34.5 1200 to 20 00075 - -100 --I50 -~19 to 78 --- 60 to 190 To match with the iatingof the connectedinterrupting device58 60 to 190._a (i) These values are based for a system, pertaining to series I1 and a frequency of 60 Hz.(ii) For systems pertaining to Series I and a frequency of 50 Hr, values furnished in Section 13.4.1(4). would apply.The peak value is a function of fault level Section 13.4.1 (7), ydbk 13.1 I. Which in turn, is a function of size and impedance of the feedingsource, such as a transformer or a generator, Section 13.4. I (S), Table 13.7. The values prescribed in the above table are thus based on theheparameters.28.4 Short-circuit effects(To determine the minimum size of current-carryingconductors and decide on the mounting arrangement).A short-circuit results in an excessive current due to lowimpedance of the faulty circuit between the source ofsupply and the fault. This excessive current causesexcessive heat (= IC', . R) in the current-carrying conductorsand generates electromagnetic effects (electricfield) and electrodynamic forces of attraction and repulsionbetween the conductors and their mounting structure.These forces are distributed uniformly over the length ofconductors and cause shearing forces due to the cantilevereffect as well as compressive and tensile stresses on themounting structure. The effect of a short-circuit thereforerequires these two very vital factors (thermal effects andelectrodynamic forces) to be taken into account whiledesigning the size of the current-carrying conductors andtheir mounting structure. The latter will include mechanicalsupports, type of insulators and type of hardware, besidesthe longitudinal distance between the supports and thegap between phase-to-phase conductors.The electrodynamic forces may exist for only three orfour cycles (Section 13.4.1 (7)), but the mechanical systemmust be designed for these forces. On the other hand, themain current-carrying system is designed for thesymmetrical fault current, I,, (Table 13.7) for one orthree seconds according to the system design. For moredetails refer to Section 13.5.The fault level, which is a function of the size of thefeeding transformer, is generally considered to last foronly one second, as discussed in Section 13.4.1(5), unlessthe system requirements are more stringent. This durationof one second on fault may cause such a temperature rise(not the electrodynamic forces), that unlcss adequate careis taken in selecting the size of the current-carryingconductors, they may melt or soften to a vulnerable levelbefore the fault is interrupted by the protective devices.NuteWhen the circuit is protected through HRC fuses or built-in short-circuit release\ of a current limiting interrupting device the cut-offtime may be extremely low, of the order of less than one quarter ofa cycle, i.e. < 0.005 second (for a SO Hz system) (Section 13.5.1)depending upon the size and the characteristics of the fuse? or theinterrupting device and the intensity of the fault current. Any levelof fault for such a system would be of little consequence, a5 theinterrupting device would isolate the circuit long before the faultcurrent reaches its first peak. This is when the fault is downstreamof the protective device. Refer to Example 28.1 below.Example 28.1Since the heating effect = I$ . ttherefore heating effect of a 50 kA fault current for 0.005second = 50' x 0.005, compared to the heating effect of anequivalent fault current I,, for 1 second, i.e. - . 1or /& = 50' x 0.005i.e I,, = 50 x liooosor 3.5 kA onlyThus to design a system protected through HRC fusesor a current limiting device for a higher fault level thannecessary will only lead to overprotection and the extracost of the current-carrying.system, switching equipmentand power cables. An individual device or componentand its connecting links in such cases may therefore bedesigned for a size commensurate to its current rating.See also Section 13.5.1.Below we discuss the thermal effects and the electrodynamicforces which may develop during a fault todecide on the correct size of the conductor and its supportingsystem.28.4.1 Thermal effectsWith normal interrupting devices the fault current wouldlast for only a few cycles (maximum up to one or threeseconds, depending upon the system design). This timeis too short to allow heat dissipation from the conductorthrough radiation or convection. The total heat generatedon a fault will thus be absorbed by the conductor itself.

~ 625Carrying power through metal-enclosed bus systems 28/865The sir.e of the conductor therefore should be such thatits temperature rise during a fault will maintain its endtemperature below the level where the metal of theconductor will start to soften. Aluminium. the most widelyused metal for power cables. overhead transmission anddistribution lines or the LT and HT switchgear assemblyand bu$ duct applications. starts softening at a temperatureof around 180-200°C. A\ a rule of thumb, on a fault asafe temperature rise of 100°C above the allowable endtemperature of 85°C or 90°C of the conductor duringnorinal service. i.e. up to 185-1 90°C during a faultcondition, is considered safe and taken as the basis todetermine the size of the conductor. The welded" portion,such a\ at the flexible joints. should also be safe up tothi4 temperature. Tin or lead solder starts softening ataround this temperature and should not be used for thispurpose. It is ad\isable to use brass soldering wherehigh-injection pressing is not possible. Welding of edgesis essential to sen1 off flexible ends.T'o determine the minimum size of conductor for arequired fault level. I,,, to account l'or the thermal effectson11 one can use the following forinula to determine theiiiiniinuin size of' conductor for any fault lewl:(28.1 )u here0, = temperature rise (in "C)I,, = syminetrical fault current r.m.s. (in Amps)'4 = cross-sectional area of the conductor (in mm')0c2,, = temperature coefficient of resistance at 20"C/"C,which as in Table 30.1 is 0.00403 for purealuminium and 0.00363 for aluminium alloys and0.00793 for pure copperH = operating temperature of the conductor at whichthe fault occurs (in "C)K = I. 166 for aluminium and 0.52 for copper/ = duration of fault (in seconds)Example 28.2Determine the minimum conductor size for a fault level of50 kA for one second for an aluminium conductor.Assuming the temperature rise to be 100°C and the initialtemperature of the conductor at the instant of the fault 85°Cthenoror100 = 'E x [?I*100xA=- 50000 x t'i1.166x1.342551006 mm2 for pure aluminium(1 + 0.00403 x 85) x 1= 617 6 mm2 for alloys of aluminium(assuming a2,, = 0 00363)'.Welding of flexible ,joints \hould preferably be carried out withliigh-in,jection pressinf (weldin? hy prem heating). eliminating theIIW 01 urlding rod\.The standard size of aluminium flat nearest to this is50.8 mm x 12.7 mm or (2" x 1/2'') or any other equivalent flatsize (Tables 30.4 or 30.5).This formula is also drawn in the form of curves asshown in Figure 28.5, l,c/A x yit. (I,' in kA) versusfinal temperature. From these curves the minimumconductor size can be easily found for any fault level, forboth aluminium and copper conductors and for any desiredend temperature. As in the above ca\eor -AGeneraliLing,vG = IIO44 1.166 x 1.34255 x IO"= 0.0799 (/\c is in kAjI,,- x ?G = 0.0799 for an operating temperatureAat 85°C and temperature at 1XS"C (18.2)Therefore. for the same parameters as in Example 28.2A=-0.0799x & --- 6253 nim'A small difference, if any. between this and that calculatedabove may be due to approximation and interpolationonly.This minimum conductor six will take account of theheating effects only during the fault. irrespective of thecurrent rating of the conductor. The required conductorsize may be more than this. depending upon the continuouscurrent it has to carry, as discussed latcr.Example 28.3If the conductor is of copper then, assuming the sameparameters,1' x100 = 0.52 X (F100= 416 mm2(1 + 0 00393 x 85) Y 1Copper is one two thirds the size of aluminium for thesame parameters and the melting point of copper at almost1083°C (Table 30.1) is approximately I .S times that ofaluminium at 660°C. These melting points are also locatedon the nomograms in Figure 28.6. Refer to nomograms(a) and (bj for aluminium and (c) for copper conductors.The same area can also be obtained from the coppercurves of Figure 28.5. Assuming the sane end temperatureat 185°C. then corresponding to the operating curve of85°C.I,, -- dt = 0.12(28.3)Aand for the same parameters as in Example 28.3.

28/866 Industrial Power Engineering and Applications Handbookgg ggVP 99Mia/ temperature 8 a $j wr- d N‘““Jr- AFigure 28.5Determining the minimum size of conductor for a required fault levelorJi = 0.12A = 416.7 mm2Almost the same size is also determined through the useof nomograms drawn on subsidiary nomogram (c) andthe main nomogram (d).NoteIn case of copper also, the end temperature is considered at 18S’Conly. Although this metal can sustain much higher temperaturethan this, without any adverse change in its mechanical properties,merely as a consideration to Table 32.3, and to safeguard othercomponents, insulations and welded parts etc., used in the samecircuit.Nomograms Figure 28.6(a)-(d) have also been drawnbased on equation (28.1). From these nomograms theminimum conductor size can be extrapolated that wouldbe necessary to sustain a given fault level.The results of these nomograms are also the same asthose from the earlier two methods except for theapproximation and the interpolation.Example 28.4Assume the same parameters as in Example 28.2.ProcedureLocate the initial temperature (85°C) and the finaltemperature (185°C) on the subsidiary nomogram (b).Draw a straight line between these points to obtain theheating function H.Transfer the value of the heating function H to the H scaleon the main nomogram (d).Locate time t as one second on the T scale.Locate the current to be carried, Is,, as 50 000 A on the I,,scale.Draw a straight line through the points on the T and I,,scales to intersect the turning axis X.Draw a straight line through the points on the H scale andon the turning axis X. The point where the line intersectson the A scale will determine the conductor area required.In our case it is 1 squre inch or 645 mm2.

Carrying power through metal-enclosed bus systems 28/867Melting pointMelting pointMelting point 660(solid) 600500Melting point 660(solid) 600500 C8300200oi.---. .10060'5020(a) Subsidiary nomogram for electrolytic grade aluminium 'INDAL'-CISM.(bj Subsidiary nomogram for electrolytic grade aluminium'1NDAL'-D 505 WP2001000tP? 600a!?400-400 $22 ---._85- z --. ...--._50 157.-100100-20 200-(c) Subsidiary nomogram for 100% IACS copperFigure 28.6Use of nomograms

IIIIIIIIE, , I IIII I , IIII I , , I I I I I , , IIII

Carrying power through metal-enclosed bus systems 28/86928.4.2 Electrodynamic effectsThe short-circuit current is generally unsymmetrical andcontains a d.c. component, Zdc, as discussed in Section13.4.1(7). The d.c. component, although it lasts for onlythree or four cycles, creates a sub-transient conditionand causes excessive electrodynamic forces between thecurrent-carrying conductors. The mounting structure,busbar supports and the fasteners are subjected to theseforces. This force is greatest at the instant of fault initiationand is represented by the first major loop of the faultcurrent, as noted in Table 13.1 1. Although this force isonly momentary, it may cause permanent damage to thesecomponents and must be considered when designing thecurrent-carrying system and its mounting structure. Themaximum force in flat busbars may be expressed by16.Z:F,,, =k.- x 104N/m(28.4)SwhereF, = estimated maximum dynamic force thatmay develop in a single- or a three-phasesystem on a fault. This will vary with thenumber of current-carrying conductors andtheir configuration but for ease of applicationand for brevity only the maximumforce that will develop in any configurationis considered in the above equation. It willmake only a marginal difference to thecalculations, but it will be on the safe side.For more details refer to the further readingat the end of the chapter.Zsc = r.m.s. value of the symmetrical fault currentin amperesFactor ofasymmetry = as in Table 13.11, representing the momentarypeak value of the fault current. Thisfactor is considered in the numerical factor16 used in the above equation.k = space factor, which is 1 for circular conductors.For rectangular conductors it canbe found from the space factor graph(Figure 28.7) corresponding toS-a wherea+bS = centre spacing between two phases in mm(Figure 28.8)a = space occupied by the conductors of onephase in mm, andb = width of the conductors in mm.For application of the above equation, refer to Example28.12.28.5 Service conditionsThe performance of a bus system can be affected by thefollowing service conditions:1 Ambient temperature2 Altitude3 Atmospheric conditions and4 Excessive vibrations and seismic effects28.5.1 Ambient temperatureThe ratings as provided in Tables 30.2, 30.4 and 30.5and others refer to an ambient temperature, with a peakof 40°C and an average of 35°C over a period of 24hours. The end temperature for aluminium is consideredsafe at 85-90°C, at which the metal does not deteriorate(oxidize) or change its properties (mechanical strength)over a long period of operation. Figure 28.9 shows theeffect of higher operating temperatures on the mechanicalstrength of aluminium metal. The oxidation andmechanical strength are two vital factors that need beborne in mind when selecting busbar size to ensure itsadequacy during long hours of continuous operation. Table28.2 lists the permissible operating temperatures of thevarious parts of a bus system.For higher ambient temperatures, current capacityshould be suitably reduced to maintain the same endtemperature during continuous operation. Refer to Tables28.3(a) and (b), recommending the derating factors for ahigher ambient temperature or a lower temperature risefor the same end temperature of 85" or 90°C respectively.For intermediate ambient temperatures, see Figure 28.10.Table 28.2Operating temperature of a bus systemMaximum operating temperature(hot spot)Bus conductor with plain connection joints 70°CBus conductor with silver plated orwelded contact surfaces 105°CEnclosureAccessible part 80°CNon-accessible part 110°CTermination at cables with plain connections 70°CTermination at cables withsilver-surfaced or equivalent connections 85OCaOr as specified by the user.Maximumtemperaturelimit as inIEEE- C-3 7-2PNoteFor temperatures above 100°C it is recommended to use epoxyinsulatorslsupports, which can continuously operate up to 125°C.SMC/FRP (fibreglass reinforced plastic) insulatorslsupports maynot withstand 105'C.Operating temperatures of bus conductorsAluminium and copper conductors start oxidizing at about90°C. The oxides of aluminium (A120) and copper (CuO)are poor conductors of electricity, They may adverselyaffect bus conductors, particularly at joints, and reducetheir current-carrying capacity over time, and lead totheir overheating, even to an eventual failure. Universalpractice therefore is to restrict the operating temperature

281870 Industrial Power Engineering and Applications Handbooki P,4/ I / /1 1.2 1.4 1.6 1.8I0Table 28.3Figure 28.7S-a-a+bSpace factor for rectangular conductors (Courtesy: The Copper Development Association)Derating factors on account of higher ambient temperature or restricted temperature rise(a) Operating temperature 85'12Ambient Permissible bur Deruting factortemperarure "C temperarure rise "C303540455055555045403530I .051 .o0.9450.880.8150.75NotesI These data are drawn in the shape of a graph in Figure 28.10.(b) Operating ternlieratwe 90°C_-Ambient Permissible bur Derating factortemperature "C temperature rise "CI 3555 1.05I 4050 I .045 45 0.94550 40 0.8855 35 0.81560 30 0.752 Intermediate values can be obtained by interpolation.

Carrying power through metal-enclosed bus systems 28/871nnnoR Y B N+ Ainin!S > 2a or 26 whichever is moreFigure 28.8 Placement of busbars to minimize the effect ofproximityD$251065°C 85°C125°C' 150°CNE5 15175°C200°Cof the bus conductors to 85-90°C for all ratings, at leastin the medium range, say, up to 3200 A.Silver oxide (Ag20) is a good conductor of electricityand can also be used for welding joints. It can seal theinside surfaces from the atmosphere and can also preventthe contact surfaces from oxidation. If the joints are silverplated or welded, the bus system can be made suitable tooperate at higher temperatures. In aluminium conductors,for instance, they can be operated up to an optimumtemperature of 125"C, until aluminium begins to lose itsmechanical strength (Figure 28.9). Similarly, copperconductors can operate at still higher temperatures. Theentire bus system can now be operated at much highertemperatures than given in Table 28.2.However, operation at such high temperatures mayimpose many other constraints, such as high temperaturein the vicinity which may endanger the operatingpersonnel. It may even become a source of fire hazard.Such a high temperature may also damage componentsmounted inside the enclosure, which may not be able tosustain such high temperatures. It may also causelimitations on gaskets and other hardware of the bussystem to operate at such high temperatures continuously.Accordingly, the maximum operating temperature of abus system aluminium or copper with silver-plated orwelded joints is also permitted up to 105°C only. Theenclosure temperature is still restricted to 80°C or up to110°C at locations that are safe and inaccessible to ahuman body (Table 28.2).0.1 1 100 10 000Heating period (seconds)-+Figure 28.9 Curves showing the tensile strength of lndalD50SWP at higher temperatures12511.051 .oL ,9458 88E ,815Q 756 'EQ50.25@ 30 235 40 45 50 55@ 35 40 45 50 55 60Ambient temperature pC) ---+@ For maximum operating temperature 85°C@ For maximum operating temperature 90°CFigure 28.10 Derating factors for different ambient andmaximum operating temperatures28.5.2 AltitudeThe standard altitude for a metal enclosed bus systemwill remain the same as for a switchgear assembly (Section13.4.2). Higher altitudes would require similar deratingsin dielectric strength and the current ratings as for aswitchgear assembly (Table 13.12), would apply. Toachieve the same level of dielectric strength, the insulationof the bus system may be improved by increasing theclearances and creepage distances to ground and betweenphases, as noted in Tables 28.4 and 28.5. To achieve thesame value of continuous current, the size of the currentcarryingconductors may be increased sufficient to takecare of the derating.NoteIt is also possible to derive almost the same value of deratingby reducing the allowable temperature increase by 1% for every300 m rise in altitude above the prescribed level.Clearances and creepage distancesThe clearances and creepage distances for open andenclosed indoor-type air-insulated busbars, as suggestedby BS 159, are given in Tables 28.4 and 28.5 respectively.These values are considered for an altitude of up to2000 m for LT and 1000 m for HT systems. For higheraltitudes to achieve the same level of dielectric strength,the values of clearances and creepage distances, as givenin Tables 28.4 and 28.5, may be increased by at least 1%for every 100 m rise in altitude.

28/872 Industrial Power Engineering and Applications HandbookTable 28.4Clearances for enclosed, indoor air-insulated busbarsRated voltage Minimum clearance Minimum c1prrranr.rto ground in air between phases in uirkV (Km.s.) mni mmUp to 0.415 160.6 193.3 516.6 6411 7615 10222 14033 2221919518912116524 1356Table 28.5 Creepage distances for enclosed indoor ailinsulated busbars as in BS 159Ruted voltage Minimum creepage Minimum creepagedistance to ground di.mnce between phaseskV (rm,s.) in air mm in uirUp to 0.415 190.6 253.3 516.6 8911 12115 15222 20333 305~. ~1Minimum 50% moreINotesI The above figures are only indicative, and may be considered asa minimum for a bus system that is dry and free from dust or anycontamination, which may influence and reduce the effectivecreepage over time. These creepages may be increased for dampdirty or contaminated locations.2 For clarification and more details refer to BS 159.Common to both tables1 The above clearances and creepage distances are for altitudes ofup to 2000 m for LT and 1000 m for HT systems.2 For higher altitudes than this, these distances should be increasedby at least 1%. for every 100 m rise in altitude.3 Voltages higher .than above, are not applicable in case of bussystems.28.5.3 Atmospheric conditionsThe same conditions would apply as for a switchgearassembly (Section 13.4.2). Unlike a controlgear or aswitchgear assembly, a bus system may be required to bepartly located outdoors. This is true for most installations,as the switchyard is normally located outdoors as is thefeeding transformer, while, the switchgears are locatedindoors, to which the bus system is connected.In such conditions, it is important that adequate careis taken to construct the bus enclosure to weather theoutdoor conditions such as by providing a canopy on thetop and special paint treatment on the outdoor part. It isalso recommended to seal off the indoor from the outdoorpart to prevent the effect of rainwater, dust and temperatureand other weather conditions on the indoor part. Thiscan be achieved by providing seal-off bushings, one oneach phase and neutral, wherever the bus enclosure passesthrough a wall. The bushings may be of SMC/DMC/FRP or porcelain for LT and epoxy compound for HTsystems. They may be fitted at the crossovers so that theindoor bus is sealed off from the outdoor one. The busconductors will pass through the bushings. The HT busconductors may be moulded with the epoxy bushings, asillustrated in Figure 28.1 l(b), similar to bar primary CTs(Figure 15.14) to make the joint airtight. In LT a simplermethod is found by providing glass wool in the part thatpasses through the wall as illustrated in Figure 28.1 l(a).28.5.4 Excessive vibrations and seismic effectsThese will require a more robust enclosure, similar to aswitchgear assembly. For details refer to Section 13.4.2.28.6 Other design considerationsSize of enclosureVoltage dropSkin and proximity effects28.6.1 Size of enclosureThe enclosure of the bus system provides the coolingsurface for heat dissipation. Its size has an importantbearing on the temperature rise of conductors andconsequently their current-carrying capacity. The enclosureeffect and the ventilating conditions of the surroundingsin which the enclosure is installed should thus beconsidered when designing a bus system. The ratio ofthe area of the current-carrying conductors to the area ofthe enclosure will provide the basis to determine theheat dissipation effect. Table 28.6 suggests the approximatedissipation factors that can be considered as likelyderatings for a bus system under different conditions.See also Example 28.12.28.6.2 Voltage dropThe voltage drop across a bus system should be as lowas possible and generally within 1-2% of the rated voltage.This criterion will generally be applicable to a highcurrentLT system. On HT and low LT current-carryingsystems, this drop may be quite low. The length of a bussystem, in most of applications, may not be long enoughto cause a voltage drop, IZ, to be taken into consideration.It may be the connection from the incoming transformerto the main receiving switchgear or the busbars of themain switchgear assembly itself. Applications requiringextra-long current-carrying conductors, however, mayhave large impedance and may cause high voltage drops,of the order of 3-5% and even more. When so, they mayaffect the stability of the system as well as the performanceof the connected load. This is illustrated in Example28.9. To ascertain the voltage drop in such cases it isessential to determine the actual values of the conductor’sown resistance, reactance and the impedance under actualoperating conditions. It may be noted that reactance is

Carrying power through metal-enclosed bus systems 281873Phase segregation(In case of segregatedbus system)Sloping top7ISection XX'Indoor*1XSeal-OFF bushingmounting plateOutdoor- I Seal-OFF bushingInspectionchamberWall frame assemblywith seal-OFF bushingInsulatorSilicagel breather(a) For LT systems (b) For HT systems ElevationFigure 28.11Wall frame assembly with seal-OFF bushingthe main cause of a high voltage drop. Skin and proximityeffects play a vital role in affecting the resistance andreactance of such systems. We discuss these aspects brieflybelow.28.6.3 Skin and proximity effects on a currentcarryingconductorIn a d.c. system the current distribution through the crosssectionof a current-carrying conductor is uniform as itconsists of only the resistance. In an a.c. system theinductive effect caused by the induced-electric field causesskin and proximity effects. These effects play a complexrole in determining the current distribution through thecross-section of a conductor. In an a.c. system, theinductance of- a conductor varies with the depth of theconductor due to the skin effect. This inductance is furtheraffected by the presence of another current-carryingconductor in the vicinity (the proximity effect). Thus,the impedance and the current distribution (density)through the cross-section of the conductor vary. Boththese factors on an ax. system tend to increase the effectiveresistance and the impedance of the conductor, and causea higher lic . R,, loss, and a higher voltage drop I,, . Z.and reduce its current-carrying capacity. An a.c. systemis thus more complex than a d.c. system and requires farmore care when designing it for a particular requirement.While these phenomena may be of little relevance for alow-current system, they assume significance at highercurrents and form an essential parameter to design ahigh current-carrying system say, 1600 A and above.These phenomena are discussed briefly below.28.7 Skin effectA current-carrying conductor produces an electric fieldaround it which induces a back e.m.f. and causes aninductive effect. This e.m.f. is produced in the conductorby its own electric field cutting the conductor. It i\ moredense at the centre and becomes less at the surface. Theconductor thus has a higher inductance at the centre thanat the surface, and causes an uneven distribution of current

~ ~~~28/874 Industrial Power Engineering and Applications HandbookTable 28.6 Heat dissipation factorEnclosure Cro\ \-sectional area Derating.f hutburs - cross juctorsectional areu ojencloureI Outdoors < 1 8 0 95S% 0 9010% 0 852 Indoors, where the < I %enclosure is in a well- 5%ventilated room 10%3 Indoors, where the < 1%enclosure is poorly 5%ventilated and the room 10%temperature is high-0.850.750.650.650.600.50Notes1 Intermediate values can be obtained by interpolation.2 These deratings are meant only for non-magnetic enclosures,where the heat generated is only through induced electric currents(I’R) and hence low. There are no hysteresis or eddy currentlosses.3 For MS enclosures, which will have both hysteresis loss (= E’.‘)and eddy current loss (-B2), a higher derating factor must beconsidered, The fullowing text will clarify this aspect.through its own cross-section. The current tends toconcentrate at the outer surface of the conductor, i.e. its‘skin’, shares more current than the other parts of theconductor and reduces with depth. It is lowest at thenucleus. For more than one conductor per phase all theconductors together may be considered as forming a largeconductor for the purpose of analysing the skin effect.Now the bulk of the current will be shared by the endconductors and only partly by the middle conductors.Figure 28.12 demonstrates an approximate sharing ofcurrent and the heat generated in one, two, three andfour flat sections per phase, placed in vertical disposition,in an almost isolated plane, where they have no or onlya negligible effect of proximity.NoteThese current sharings are only indicative. The actual current sharingwill depend upon the thickness, the width and the configuration ofthe conductors. Refer to Tables 30.2, 30.4 and 30.5.The phenomenon uneven distribution of current withinthe same conductor due to the inductive effect is knownas the ‘skin effect’ and results in an increased effectiveresistance of the conductor. The ratio of a.c. to d.c.resistance, Rac/Rdc, is the measure of the ‘skin effect’ andis known as the ‘skin effect ratio’. Figure 28.13(a)illustrates the skin effect for various types and sizes ofaluminium in flat sections. For easy reference, the skineffects in isolated round (solid or hollow) and channelconductors (in box form) are also shown in Figures28.13(b) and (c) respectively.Tables 30.7,30.8 and 30.9 for rectangular, tubular andchannel sections respectively, give the d.c. resistance andthe reactance values between two aluminium conductorsof small and medium current ratings of any two adjacentphases, with a centre-to-centre spacing of 305 mm ormore, when the proximity effect is considered almostnegligible in these ratings.Since the skin effect results in an increase in the effectiveresistance of the busbar system it directly influences theheating and the voltage drop of the conductor and indirectlyreduces its current-carrying capacity. If R,, is the resistanceas a result of this effect then the heat generated= I:c R,,where I,, is the permissible current-carrying capacity ofthe conductor on an ax. system to keep the same heating2305 mm + 2 305 mm+(a) 100 100 100+ 2 305 mm +(b) 100% 100% 100%2 305 mm 4tit ttt t t t(a) 8 800mm(b) 1:l 1:l 1:l00 mm6 2 305 mm 2 305 mm 4ttttt t t t t t t t t t tmcommmm8 mcummcummc9mmc9m(a) 8 :(b) P 2 $2 -r.- -r.-(a) g % % g(b) m e b mNotet 305 rnm t 305 rnrn 40000mcucumm b b m(a) Current sharing %(b) Heat generated (I‘R) ratio1.tN11. f t t0000m ~ ~ mm b b mEach phase is considered in isolation not influenced byproximity effectFigure 28.12 Skin effect in different bus sections of the samephaseN

~ 1.0Carrying power through metal-enclosed bus systems 28/875effect as on a d.c. system then the reduction in the currentrating due to the skin effect can be deduced by equatingthe two heats, i.e.(28.5)The skin effect can be minimized by employing differentconfigurations and arrangement of busbars, as discussedlater and illustrated in Figure 28.14. It can also beminimized by selecting hollow round or hollow rectangular(channels in box form) conductors, and thus concentratingthe maximum current in the annulus and optimizing metalutilization. For current ratings in round and channelsections, refer to Tables 30.8 and 30.9, respectively.1.51.425 -1.41 1.3-1.275 -5cr".? 1.25 1.182s= -%.1.13--E$ 1.1-=4 8 16 2401055-6.45 -10-. 12.9 4219.35,c+ 20.02"0 25.8-$2 30@(j i3 38.71 -1 504060@ lndal D 50 SWP at 85°C@ lndal ClSM at 85°C@ lndal D 50 SWP at 20°C@ lndal ClSM at 20°CNoteThe lower cross sectional area curves relate to f= 50 Hz. For otherfrequencies must be calculated by multiplying these valuesby ,I%%(a) Flat bus bars@ lndal D 50 SWP at 85°C@ lndal ClSM at 85°C@ lndal D 50 SWP at 20°C@ lndal ClSM at 20°C@ Copper at 85°C@ Copper at 20%NoteThe lower curves apply for f = 50 Hz only. For other frequencies,must be calculated by multiplying these values by ,I%%(b) Channels in box formFigure 28.13(a) Skin effect in isolated rectangular busbars. (neglecting the proximity effect) (Courtesy: I?ndian Aluminium Co.based Alcon of Canada)

~28/876 Industrial Power Engineering and Applications HandbookI=dOInOlOO InInb*@-Jrn N zooooo o o1=015d'GO12d-=010 td1=008d@ lndal D 50 SWP at 85°CI = o6 @ lndal ClSM at 85°Cd@ lndal D 50 SWP at 20°C= 0 04 @) lndal ClSM at 20°Cd@ Copper at 85°C'=OOld @ Copper at 20°CNoteThe lower curves apply for f= 50 Hz only For otherfrequencies must be calculated bymultiplying these values by &Figure 28.13(b) Skin effect in isolated tubular conductorsExample 28.5If there is a rise of 5% in the effective resistance of thebusbars due to the skin effect, then the ax. rating will be118% -125%-- 100%1 1 1 1II II0128%1 2 3 4151%--LA5 6 7Figure 28.14 Ratio of ax. current ratings for differentconfigurations of busbars of the same cross-sectional area(Courtesy: The Copper Development Association, U.K.)= 0.976 or 97.6% that of the d.c. rating28.7.1 Skin effect analysisWhen a number of flat bars are used in parallel theireffective current-carrying capacity is the result of thecumulative effect of the restricted heat dissipation andthe increased content of the skin effect. A stage mayarise when further addition of any more bars may notappreciably increase the overall current-carrying capacityof such a system. Referring to Tables 30.2, 30.4 and30.5, we can observe a wide variation in the currentcarryingcapacity of a conductor when it is added to anexisting system of one, two or three conductors per phase,depending upon the thickness and width of the conductors.Thinner sections of shorter widths provide better metalutilization, compared to a thicker section and larger widths.Use of bars up to four sections per phase is quite commonfor higher current systems (2500 A-3200 A). For stillhigher current ratings, use of more than four bars inparallel is not advisable due to an extremely low utilization

Carrying power through metal-enclosed bus systems 28/877of metal, particularly in larger sections. While largersections would be imperative for such large ratings, theirown rating would fall to a low of 14-1 8% of their normalcurrent capacity. (See Table 30.5 for larger sections,providing current ratings up to six bars in parallel.) Insuch cases it is advisable to arrange the bars in any otherconvenient configuration than in parallel, as illustratedin Figure 28.14 or to use round or channel sections toachieve better results and a higher level of metal utilization.28.7.2 Determining the skin effectAs a result of the electric field around the conductors thefrequency of the system has a very significant bearingon the skin effect. The various curves as establishedthrough experiments and, as reproduced in Figures 28.13(a), (b) and (c) respectively for rectangular tubular andchannel conductors, are thus drawn on the ,/m basis.At 50 Hz, the value of the skin effect, RaclRac, can beread directly from these curves, as the curves for differentcross-sectional areas and conductivity, at 50 Hz, havealso been drawn in the lower part of the figure.Ratings of up to 3200 A are normally required fordistribution purposes such as for interconnecting adistribution transformer to a PCC, or a large PCC toanother large PCC in a sub-station. Common practicefor making such connections is to use rectangular crosssections,which are easy to handle, manoeuvre and makejoints, compared to a channel or a tubular section. Channeland tubular sections require special tools and skilledworkers, particularly when bending or making joints andend terminations. However, suitable fittings and fixtures,some of which are shown in Figure 28.15 (a) and (b), arealso provided by leading aluminium section manufacturersas standard practice to facilitate such connections. Thewelding of such joints will require special weldingequipment and adequate in-house testing facilities to checkthe quality of weld. It is, however, recommended to usesuch sections, for ratings 3200 A and above, for betterutilization of active metal compared to flat sections. Webriefly deal with such sections as follows.(i) Rectangular sectionsExample 28.6Consider a section of 101.6 mm x 6.35 mm of grade EIE-Mas in Figure 28.16. From Table 30.7 for its equivalent gradeCIS-M(i) Rdc = 44.55 mlm at 20°Cor 44.55 x 1000 x 1 O~ R/I 000 mLe. 0.0445 R/lOOO mArea of cross-section = 101.6 x 6.35 x= 6.4516 cm2cm2Since the operating temperature should be considered to be85"C,Rd, at 85°C = Rd@o [1 + azo(& - e,)] (28.6)where= temperature coefficient of resistance for CIS-M grade of aluminium from Table 30.1,= 0.00403 Per "C at 20°CCL

281878 Industrial Power Engineering and Applications HandbookS=lOO 100 7 100IN6.35 Sb= 10.la d ka = 6.35b = 101.60s = 100.00Figure 28.16 Illustration of example 28.6A// dimensions in mm.a = 44.45b = 101.60S = 184.45 (It is recommended to be min. 300)= 0.0445 (1 + 0.26195)= 0.056 CY1000 mNow refer to Figure 28.13(a) to obtain the skin effect ratioRac/Rdc. Consider the cross-sectional curves for EIE-M gradeof flat busbars at an operating temperature of 85°C for across-sectional area of 6.45 cm2 and determine the R,,/Rdcratio on the skin effect curve havingFigure 28.17(a) Illustration of Example 28.7b/a =101.6/6.35 = 16... ' = 1.055RdCLe. an increase of almost 5.5%, due to the skin effect aloneandR,, = 1.055 x 0.056= 0.059 WOO0 m(ii) Skin effect for more than one conductorperphaseIn such cases, the group of busbars in each phase may beconsidered to be one large conductor and outsidedimensions a and b as illustrated in Figure 28.8 measuredfor all calculations.Example 28.7Consider a four-conductor system of section 101.6 mm x6.35 mm in each phase (Figure 28.17(a)) of grade EIE-M forcarrying a current of 2000 A.Rdc : As calculated above for one section of bus= 0.056 ohm/l000 m per conductor of four bus-sectionsin parallelSkin effect ratio R,,IRd, from the graph of Figure 28.13(a), atan operating temperature of 85°C for a cross-sectional areaof 25.8 cm2 (4 x 101.6 x 0.635) for an EIE-M grade of aluminiumhavingR,JRd, 5 1.33:. R,, for the phase = 1 x 1.33 x 0.0564= 18.62 x 10-~ CY1000 mFigure 28.17(b) Minimizing the effect of proximity in thickersections (Section 28.8.4)(iii) Busbar configurations(To improve heat dissipation and minimize the skin effect)The busbars may be arranged in different configurationsas shown in Figure 28.14 to improve heat dissipationand reduce the skin effect as well as the proximity effect.The improvement in the ratings is indicative of the coolingand skin effects with different configurations. When anumber of bars are used in parallel, each bar shields theadjacent bar and reduces its heat dissipation. Moreover,together they form a large conductor and the current willtend to concentrate at the outer surfaces only, due to theskin effect. It will cause inner surfaces to share smallerand the outer surfaces the larger currents. In configurationsother than parallel bars, an attempt is made to improveheat dissipation and reduce the skin effect. It is obviousthat most of the conductors are now sufficientlyindependent of the others and can carry higher currents.28.8 Proximity effectIf there is more than one current-carrying conductor otherthan of the same phase, placed adjacent to each other, sothat the electric field produced by one can link the other,mutual induction will take place. The magnitude of thiswill depend upon the amount of current and the spacingbetween the two. This tends to further distort the selfresistanceof the conductor over and above the distortionalready caused by the skin effect current distribution

Carrying power through metal-enclosed bus systems 28/879through its cross-section. Figure 28.18(a), (b) and (c)illustrate diagrammatically distortion of current flow ina round conductor and also the mechanical forces exertedon the conductors, due to this distorted current distribution.There is always a force between two current-carryingconductors placed adjacent to each other, whether it is ad.c. or an a.c. system. The proximity effect, however,will exist only in an a.c. system due to mutual inductionbetween the two current-carrying conductors. It may beless pronounced in low current systems, say, 1600 A orless, and all HT systems, where the spacings betweenthe phases is considerably more, except their effect onthe enclosure, which is discussed in Chapter 3 1 on isolatedphase bus systems. If the second conductor carries currentin the same direction, such as in a three-phase system(Figure 28.18(b)) the current will flow in the remoteparts of the two conductors. If the current flows in theopposite direction, as in a single-phase system (Figure28.18(c)) the current will flow in the adjacent parts ofthe two conductors.The displacement of current and the forces (equation(28.4)) on the conductors are two different effects. Theeffect of current displacement is to increase the effectiveresistance and the impedance of the conductor on one(a) Uniform current distribution in an isolatedconductor or a d.c. conductorc--.--Attractive force'-_-'(b) Distortion in current distribution when the currents are in thesame direction, like a 3-$ systemElectric(c) Distortion in current distribution when the currents are inopposite directions, like a 1-4 systemNote1. 8 - Direction of current in a conductor looking from top.- Current coming out6 - Current going in.2. Direction of electric field by Cork-Screw rule.Figure 28.18 Current distribution in round conductors,illustrating the effect of proximityside, as illustrated in Figure 28.18(b) and (c) and causea distortion in its heating pattern. This will lead the variousconductors of a particular phase to operate at differenttemperatures and add to Z:c . R, losses. The rating of allthe conductors of one phase must therefore be determinedby the hottest conductor. The distortion of current willalso distort the heat produced. The area having high currentdensity will produce higher heat. The proximity effectthus also causes a derating in the current-cawing capacityof a conductor.In general, the proximity effect is directly proportionalto the magnitude of the current and inversely to the spacingbetween the two conductors. The smaller the phasespacing, the greater will be the effect of proximity aswell as the derating and the greater will be the forcesdeveloped between the adjacent conductors (equation(28.4)). But the reactance of the two phases is directlyproportional to the spacing. Reactance is the main causeof an excessive voltage drop (IZ). The smaller the spacing,the lower will be the reactance, due to the proximityeffect and vice versa. While the requirement of a lowerreactance will require less spacing and will mean higherforces, demanding stronger busbar supports and mountingstructure, requiring a lower effect on current-carryingcapacity would require a larger spacing between thephases, which would result in a higher reactance andconsequently a higher voltage drop. But a high reactancewould help to reduce the level of fault current, I,, andalso forces, Fm, between the conductors.A compromise is therefore struck to meet both needsand obtain a more balanced system or other methodsadopted, as discussed in Section 28.8.4, to reduce theskin and proximity effects.28.8.1 Proximity effect in terms of busbarreactanceReactance, X,, of the conductors plays a significant rolein transmitting the power through a bus system from oneend to the other. For long bus systems, it must beascertained at the design stage whether the voltage dropin the total bus length on account of this will fall withinthe permissible limits, particularly for higher ratings,(2000 A and above) besides the current-carrying capacity.A higher reactance will mean a higher drop. For smallerratings and shorter lengths, as well as HT systems, thisdrop would be too low as a percentage of the rated voltage,to be taken into account. For higher ratings, however, itmay assume a greater significance and precautionarymeasures may become necessary to restrict it withinpermissible limits. To determine X,, proximity effectcurves have been established by conducting tests on themetal and are available for all sections, configurationsand spacings of busbars. We have reproduced* them fora 50 Hz system (for a 60 Hz system, Xa60 = Xaso . 60150or 1.2 Xaso), for rectangular sections as in Figure 28.19(a),tubular sections as in Figure 28.19(b) and channel sectionsin box form as in Figure 28.19(c). A brief procedure todetermine the reactances with the help of these curves isgiven below.

~28/880 Industrial Power Engineering and Applications Handbook2001501005000 .5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5Conductor spacing - 2 -Semi perimeter a + b1. For 3-4 systems read reactance against [ + .26-asb]I ,2. The reactance varies with the ratio $- and therefore there may be a number of possible busbar combinations and the correspondingcurves for different a However, only a few curves have been drawn for the likely minimum and the maximum values of a Since theb ’ b ’variation is not large therefore by interpolation the more pertinent value of reactance can be determined from these curves.Figure 28.19(a) Reactance of rectangular busbars at 50 Hz on account of proximity effect< I-2802602402202006180&1602.$1408 120C,m8 100c?80604020Itm i i i FH i i i i i i i i i i i-i t i i I0 20 40 60 80 100 120Conductor spacing Se between centres (em)Figure 28.19(b) Reactance of tubular busbars for singlephaseor three-phase systems at 50 HzRectangular sections (Figure 28.19(a))Thc rcactancc is drawn as a function ofCentre snacing (‘S?Semi-perimeter (a + b)At lower spacings this value will be influenced by thewidth (b) and the thickness (a) of the conductor. At lowerspacings, therefore, proximity curves are different fordifferent ratios of a/b whereas for larger spacings theyapproach the same curve.When more than one section is used together, to makelarger ratings, all the sections of one phase may beconsidered to be one large section. The dimensions aand b of the whole section are now considered as oneconductor, as illustrated in Figure 28.8.The reactance thus obtained can be doubled for singlephasesystems. For a three-phase system the configurationof the three phases with respect to each other will play asignificant role and the linear centre spacing S has to bemodified to an effective or geometric mean spacing S,,wheres, = (Sa‘ Sb ’ S,)”3 (28.7)For configuration (a) of Figure 28.20sa = Sb = ss, = 2s

Carrying power through metal-enclosed bus systems 28/881Figure 28.19(c) Reactance of channel busbars, two channels per phase in box form, single-phase or three-phase, at 50 Hz.-. S, = s . (21113= 1.26 Sand for configuration (b) of Figure 28.20sa = s, = s, = s:. s, = sFor any configuration, the effective spacing, S,, may thusbe calculated.Tubular sections (Figure 28.19(b))For determining S, in solid or hollow round sections it isessential to first determine the self geometric meandistance, Ds, of the conductors which varies with thethickness t (annulus) of the conductor. Ds approaches itsouter radius, rl, in an infinitely thin conductor and to0.778r1 in a solid bar. This variation, in the form ofDslrl is drawn in Figure 28.21, as a function of r2/rl.For very thin conductors, when r2 = rl, r2/r1 = 1,Dslrl will also approach unity and Ds = rl. Forsolid conductors, when r2 = 0, r2/r1 = 0, Dslrl becomes0.778 and Ds = 0.778 rl etc.After having obtained the value of Ds, value of S, isdetermined as discussed above. The reactance of theconductors can then be obtained from the graphs of Figure28.19(b) drawn for S, versus X,, for varying thicknessesof round conductors. Here also the basic graph willrepresent a single-phase system, having a reactance of2 . X,. Refer to Example 28.8.Channel sections (Figure 28.19(c))These should normally be used in a box form for bettermounting, uniformity and metal utilization. The methodof determining the reactance for single- and three-phasesystems is the same as for rectangular sections (Figure28.20(a)).From the proximity curves it may be noted that X,rises with S. While a higher centre spacing would reducethe effect of proximity on the current-carrying conductorsand which is so much desired, it will increase X,, whichwould mean a lower p.f. for the power being transferredR Y Bsc /R0/ \I \\ Sa// \1(a) Rectangular sections (b) Rectangular sections (c) Tubular sectionsFigure 28.20 Influence of conductor configuration on linear spacing S

28/882 Industrial Power Engineering and Applications HandbookFigure 28.21Graph to determine 0, of a tubular bus section(through the busbars) and a higher voltage drop. In LTsystems the spacings can be adjusted only marginally toreduce X,, as a lower spacing would mean a higherelectrodynamic force, F, (equation (28.4)) and greaterproximity effects, requiring higher busbar deratings. Acompromise may therefore be drawn to economize onboth. In HT systems, however, which require a largerspacing, no such compromise would generally be possibleand they will normally have a high content of X,. But inHT systems, voltage drop plays an insignificant role inview of a lower voltage drop as a percentage of thesystem voltage.28.8.2 Voltage unbalance as a consequence of theproximity effectThe proximity effect does not end here. It still has somefar-reaching consequences in terms of unequal voltagedrops in different phases at the same time. This is moreso on large LT current-carrying, non-isolated bus systemsof 2000 A and above, resulting in an unbalance in thesupply voltage, as discussed below.A three-phase system has three current-carryingconductors in close proximity. While the conductors ofphases R and B will have an almost identical impedance,with the same skin and the proximity effects, the conductorof phase Y is under the cumulative effect of electric fieldsR Y BtInfluence of electric fields of conductors Rand B isoffset in a 3-4 system. The proximity effect in phase Ytherefore gets nullifiedFigure 28.22Influence of proximityof the other two phases, which would offset their proximityeffects (see Figure 28.22). The conductor of phase Ytherefore would carry no distortion beyond the distortionalready caused by the skin effect (Rac/Rdc). The result ofthis would be that in a balanced three-phase system thethree phases will assume different impedances and causean unbalance in the current distribution. The Y phasehaving a smaller impedance, would share more currentcompared to the R and B phases and cause a smallervoltage drop. Such an effect may not be as pronouncedin lower ratings and shorter lengths of current-carryingconductors, as on higher currents, depending upon thespacing between the phases and the length of the system.Consider a feeding line from a transformer to a powerswitchgear through a bus duct. The voltage available atthe distribution end of this feeding line may be unequaland tend to cause a voltage unbalance. Depending uponthe rated current and length of the feeding line, it mayeven cause a voltage unbalance beyond permissible limits(Section 12.2) and render the system unstable and insome cases even unsuitable for an industrial application.For larger current systems, 2000 A and above andlengths of over 50 m, a correct analysis for such aneffect must be made and corrective measures taken toequalize the voltage and current distribution in all thethree phases. Where adequate precautions are not takenat the design stage through phase interleaving ortransposition techniques, as discussed later, the problemcan still be solved by making up for the lost inductancein the Y phase by introducing an external inductance ofan appropriate value in this phase. It is possible to dothis by introducing a reactor core into this phase, asillustrated in Figure 28.23. This inductor will compensatefor the deficient inductance and equalize the impedancesin all three phases, thus making the system balanced andstable. We illustrate briefly later a procedure to determinethe size of a saturable reactor core, when required, tomeet such a need.Example 28.8Consider Example 28.6 to determine the content of proximity;(i) For reactance X, on account of the proximity effect, useFigure 28.16 and the graph of Figure 28.24:

='i(ii)*i. 2l N 7Plan of a 36 and N bus systemFigure 28.23 Balancing of phase currents in a large threephasesystem by introducing a reactor in the middle phaseand1.26100~ 1 .26~a+ b (101.6 + 6.35)100= 1.26 x __ 1.167107.95then from the graphX, = 106 pWm or 106 x x 1000 R/1000 mCarrying power through metal-enclosed bus systems 28/883i.e. 0.106 RilOOO m andlmpedanceZ= ,/m= ,/(0.059* + 0.106') L1/1000 m= 0.12 n/lOOO m(iii) Voltage dropIf we consider the average current-carrying capacity of thissection as 1000 A, after normal deratings (without a derating1235 A from Table 30.4), then the voltage drop duringnormal running, say, for a 150 m length of this section ofbusbar= 1000 A x 0.12 x 150 volts1000= 18.0 Vwhich is around 4.3% of a 415 V system. Such a high voltagedrop, although less than 5% and normally permissible, maynot be advisable in this case, since in addition to this dropthere may be further voltage drops in the connecting cables,resulting in a higher drop than 5% up to the connected load.It is also possible that the voltage at the receiving end itselfwas already a little less than rated, due to drops in the uppernetwork.This example is considered only to emphasize thesignificance of voltage drops in current-carryingconductors, particularly when the system load is highand the end distribution is distant from the receivingend. In normal practice, however, consideration of avoltage drop in a bus system may not be of muchsignificance, due to the generally short lengths of thebus ducts, which may not be more than 3040 in in mostof installations, irrespective of the size of the feedingtransformer. However, if such a situation arises. as in5Figure 28.24 Reactance of rectangular busbars at 50 Hz on account of proximity effect

1.5928/884 Industrial Power Engineering and Applications Handbookthis particular instance, one may reduce the content ofX, by reducing S if permissible, or consider the nexthigher cross-section of busbars. This size of bus section,in this particular instance, may be considered for a currentrating up to 800 A.R+Example 28.9Consider Example 28.7 to determine the effect of proximity:(i) For the configuration of Figure 28.17(a)a - 44.45 - 0.4375b 101.60S 184.45 ~1.26 - = 1.26 xa+b 44.45 + 101.60then X,, due to the proximity affect from the graph of Figure28.24,or= 125Wm0.125C2/1000 m per phase and(ii) impedance, z= d0.0186' + 0.125'= 0.126 R/1000 m per phase and(iii) voltage drop, considering a length of busbars as 40 mand current rating as 2000 A,40= 2000 x 0.126 x ~1000= 10.08 V which is 2.43% for a 415 V systemThe bus system is therefore suitable to carry 2000 A up to alength of 40 m. Beyond this the voltage drop may becomehigher than permissible and the bus rating may call for aderating.NoteThe rating for this section considered here as 2000 A, is hypotheticaland must be checked for the various design parameters as discussedalready, in Sections 28.5 and 28.6, and analysed in Example 28.12.Use of a saturable reactor (choke) to balance alarge unbalanced power distribution systemDetermining the size of reactorConsider a three-phase bus system as shown in Figure28.27. If X, and X are the inductive reactances of eachphase on account of skin and proximity effects respectively,then the impedances of each of the three phases can beexpressed as2,=R+X,+X,=Z(since the Y phase will have no proximity effect)Therefore inductive reactance equal to Xp must beintroduced into the Y phase to equalize the reactancedistribution and make the system balanced (Figure 28.25).If IR, Iy and IB are the currents in the three phases andVph the phase voltage thenFigure 28.25 Distribution of inductive reactance andimpedance of each phase in a three-phase systemI, = I, = I, (say)(Basically these are all phasor quantities but for ease ofillustration absolute values are considered)I, . Vphor Vph -I, X, = -I,orI, ' vphI, . xp = Vph - - I,The values of I, and I, must be known to determine thevalue of the reactor, Xp. Otherwise the reactance, X,, asdetermined earlier, to account for the proximity effectmay be considered as the lost reactance in the Y phaseand for which the reactor may be designed for this phase.Designing a reactorThe self-inductance

-Carrying power through metal-enclosed bus systems 281885A = area of cross-section of core in square metresk = total length of the magnetic circuit in metresorwhereReactor magneticcircuit (length = k)iy - Variable air gap, say, 1.5 rnm to 5.0 rnrnFigure 28.26A typical reactor coretarnpings of0.2-0.5 rnrn thickC.R.G.0 silicon steellaminationswhereL = self-inductance of the choke in henry (H)@ = flux produced in Weber (wb)I, = rated current in AmpsZ, = number of turns in the choke = 1(since it will be used like a bar primary. as shownin Figure 28.26)Total reluctance R, of the magnetic circuit of the choke- MMF - 2, '1,($ 4L ' I ,since I$ = ~z,The total reluctance of an iron choke (Figure 28.26) canalso be expressed by R, = R,,, + R,,,,whcrcandR,,, = reluctance of the air gap21,-__ -Po AR,,,,, = reluctance of the iron path-k-21,PC .Pu, A(28.9)whereI, = length of the air gap in metres (Figure 28.26)pel = permeability* of air (free space) = 47r. lo-' H/mp, = relative permeability of the silicon steel used forthe laminations in H/m*Permeability defines the magnetic property of a material, and is ameasure of how easily it can be magnetized. The greater thepermeability of a material. the easier it is to magnetize.and ,u is a design parameter for the silicon steel and= BIH.whereB = flux density in wb/m'and H = magnetic field strength in A hIt may be observed that the value of permeability is afunction of the flux density being attained to energizethe magnetic circuit. It is therefore not a constant parameterand is measured at a particular flux density.Example 28.10Consider a bus duct having a rated current of 4000 A and anunbalanced current in the middle phase of 4400A. Determinethe size of the reactor to achieve a balanced voltage system.SolutionIf the line voltage = 440 Vthen Vph = 440/&= 254 VL]:. X, = 254 x ~(4dOO 4400- 254x4004000 x 4400= 0.00577 = 2 x K x 50 x L (for a 50 Hz system)or L=0.005772 x 3.14 x 50= 18.38 x HLx l:. Total flux @ = 2i.e. 4 =zc18.38 x lo4 x 4000 wb1= 73.52 x wbTherefore area of cross-section of the core- 7352 (assuming B = 1.1 wb/m')1.1= 66.84 x square metres.If we assume the stacking factor of the core laminates tobe 0.9, the gross cross-sectional area of the core

28/886 Industrial Power Engineering and Applications Handbookf$---cIIIIfIIIIIII2Length = kII1IIItIIIIIII1Air gap'tg' = 2 mrni7 6.35I+6.35Min 250 mm 4Min 250 mmDepth d = 500200Magnetic field sfrength 'H' (Ah) --+Figure 28.27 Design of a reactor for a power distribution bussystem. Illustrating Example 28.10Figure 28.28 A typical magnetic saturation (B-H) curve forCRGO sheetsA=66.84 x 10-30.9or 74.27 x squre metresor 74270 mm2Assuming the width of the laminates to be 150 mm and thedepth of the core as d (Figure 28.27),:.Area of cross-section of the choke = 150 . d = 7427074 270or d=-150= 495 (say, 500 mm):. A = 150 x 500 = 75 000 mm2 or 0.075 m2The size of the opening will depend upon the width andheight required by the current-carrying conductors. Say, forsix numbers 152.4 mm x 6.35 mm busbars, as shown inFigure 28.27, an opening of 170 mm x 250 mm will beadequate.To determine the value of p, the saturation or the B-Hcurve of the silicon steel being used must be available.Assuming a normal flux density for such a core to be 1 .I wb/m2 (see also Section 1.9) and making use of a normal 8-Hcurve as shown in Figure 28.28, the corresponding value ofH for a value of B as 1.1 wb/m' can be read as 200 Nm,. p,=B=l.l= 0.0055 H/mH 200Pand pr = -1 0:. pr= 0.00554nx 10-7= 4378 H/mlength of magnetic circuitk = 2(320 + 400) + 2./g= 2 x 720 mm or 1.44 mz:and R, = - L- 12H18.38 x lo4= 5.44 x 104~:. Air gapR1I, =.pr .A- k2(Pr - 1)-5.44 x lo4 x 4n x lo-' x 4378 x 0.075 - 1.442(4378 - 1)- 22.44 - 1.44 M2 x 4377= 2.4 mm.and weight of the chokeW = Volume x specific gravitywhereVolume = 1440 x 150 x 500 mm3Assuming the specific gravity of the laminates to be = 8.5g/cm3:. w =1440 x 150 x 5001000 1000= 918 kg8.5 kg28.8.3 Derating due to the proximity effectWe will discuss this aspect in two parts, one for the nonisolatedbus systems and the other for the phase-isolatedbus systems as in Chapter 3 1.

Carrying power through metal-enclosed bus systems 28/887Proximity effect on non-isoluted bus systemsDrawing inferences from the literature available on thesubject (see the further reading at the end of the chapter),based on laboratory tests, practical experience and thefield data available, deratings for different configurationsare shown in Table 28.7 which should be sufficient toaccount for the likely proximity effectsProximity eflect on the enclosureThe electric field produced by the current-carryingconductors of each phase also links the metallic busenclosure, its mounting supports, and structures existingin the vicinity, parallel and around the axis of the currentcarryingconductors. It causes induced (parasitic) currentsin such structures and leads to the following:0 Resistance losses (Z'R) andMagnetic losses.Magnetic losses will constitute the following:Eddy current losses (= B2, Section 1.6.2.A-iv) and0 Hysteresis losses (= B'.6, Section 1.6.2.A-iv)The electrodynamic forces between the enclosure andthe conductors will be small because the enclosure, whichis non-continuous, will carry much less current than themain conductors. They therefore need not be consideredseparately, as the metallic structure will have sufficientstrength to bear them.In a non-magnetic enclosure, such as aluminium orstainless steel, there will be only resistance losses. In amagnetic enclosure, such as mild steel (MS), there willalso be hysteresis and eddy current losses in addition toresistance losses. All these losses appear as heat in theenclosure and the metallic structures in the vicinity. Athigher currents say, 2000 A and above, this phenomena,particularly with MS enclosures, may assume such largeproportions that the enclosure, instead of providing aheat-dissipating surface to the heat generated by thecurrent-carrying conductors inside, may add to their heat.Depending upon the current rating, the configuration ofthe busbars and the material of the enclosure should bechosen to minimize these effects as far as possible. It ispossible to do this by adopting one or more of the followingmethods. Since the spacing in an HT system is alreadylarge, an HT system is generally not affected by theproximity effects. The following discussion thereforerelates primarily to an LT system.Table 28.7 Approximate deratings due to proximity effects for different configurations of bus systemsCurrent rating(1) Forflat busbarCentre spacing SApprox. derating(I)For LT systems1 Smaller ratings up to 1600 A2 For 2000-3000 A3 For larger ratings up to around 6500 A, asrequired for medium size turbo-alternators, upto 5 MVA used for captive powergeneration in a process plant, such as asugar mill, mostly utilizing its own surplus orwaste gasedfuel and steam. Also smallgas and hydroelectric power-generatingstations(11) For HT systems 2000-3000 A(2) For channel sectionsFor 2 channels in box form---'In channels in box form a > b as shown in Table 30.9.Normal spacings(i) S 2 4b(ii) S 2 2bS 2 4bGenerally S t 4bs 2 3a"S24aas 2 5aaS 2 6aa5%5%15%15%5%18%11%5%1%

28/888 Industrial Power Engineering and Applications Handbook28.8.4 Minimizing the proximity effectElectric field1 Maintaining greater spacing (S) betweenthe phasesThis can be done by providing adequate clearancesbetween the conductors and the inside of the enclosure(say, 300 mm). Table 30.5 and all the other tables inChapter 30 are based on the fact that the proximity effectis almost negligible at 300 mm from the centre of theCurrent-carrying conductors. But in thicker sections, orwhere a number of smaller sections are used together toform a phase, the current will concentrate at the outersurfaces only (skin) rather than the nucleus. Therefore,to achieve an almost Lero-proximity efrect condition it isdesirable to provide a space of 300 mm and more betweenthe extreme outer surfaces, rather than between the centres,as shown in Figure 28.17(b). The condition of S 2 4b(Table 28.7) will also be almost satisfied by doing so.For still higher currents, this distance must be increasedfurther or a segregated construction adopted (Section28.2.2). But to keep the phase conductors completelyout of the inductive effect of the other phases may requirevery large enclosures, particularly at higher ratings (above3200 A or so), which may not be practical.Below we describe improvised bus systems to limitthe reactance and hence the voltage drop and obtain aninductively balanced system to achieve a balanced voltageand equal load sharing by the three phases at the far end.The systems are0 Phase interleaving and0 Phase transposition2 Phase interleavingThis is a highly efficient and more practical method forlarge ratings and offers a very high metal utilization ofthe conductors. It provides an almost balanced and a lowreactance system. Each phase, consisting of a number ofconductors, is split into two or more groups and eachgroup of conductors is then rearranged into three or threeand a half phases, according to the system requirement,as illustrated in Figures 28.29(a) and (b). It is, however,suggested, to limit the number of groups to only two forconsiderations of size and the cost of enclosure. The twogroups would meet the design requirements in most cases.Therefore if four or more flats are used per phase it isnot always necessary that as many groups be arranged,unless the current rating of the system is too large toeffectively reduce the reactance of the entire system. Infour conductors, for instance, two groups, each with twoconductors per phase, can be arranged as shown in Figure28.29(b) to achieve a low reactance system as a result ofthe smaller spacings between the split phases, on the onehand, and two parallel paths, on the other. The two parallelpaths will further reduce the total reactance to one half.The field produced by each split phase would now becomehalf (@ = I ) and fall out of the inductive region of theother. The arrangement would thus provide a systemwith a low proximity effect.1- 1st group 2nd group -!(a) Interleaving of 2 conductors per phase2 300 rnrn preferablyR Y i B N R Y I B N R Y1- 1st group 2nd group 4(b) Interleaving of 4 conductors per phaseFigure 28.29 Minimizing the effects of skin and proximitythrough phase interleavingItili iSince the conductors in each phase are now arrangedas close as dielectrically possible, the electrodynamicforces on each group in the event of a fault on account ofthe spacing would be high as a result of the smaller spacingbetween the phase conductors (F, ~c I/S equation (28.4)).But the overall forces would become much less comparedto the conventional arrangement because of two or moreparallel current paths, each carrying a reduced amount ofcurrent, depending upon the number of parallel paths soformed. Fc.r two parallel paths, for instance,F,, = 2 . ( I:c /2) or = I; /2. Generalizing, F,,, -, l/n ifn is the number of parallel paths. Moreover, the mountingsupports will also become stronger than before, becausethere are as many mounting supports as the number ofparallel paths, each sharing the total force equally. Themethod of interleaving will therefore require no extrareinforcement of the busbar supports or the mountingstructures. (See also Example 28.1 1). As a result of thelow reactance obtained the arrangement will provide asomewhat inductively balanced system. The reactance ofthe conductors can be calculated on an individual groupbasis and then halved when the conductors are split intotwo halves, or reduced by the number of parallel pathsarranged.The arrangement will also minimize the skin effect toa very great extent, as the current of each phase is nowshared by two or more independent circuits, each of athinner section than a composite phase. It can beconsidered an improvised version of arrangement 2 inFigure 28.14. The thinner sections (smaller nucleus) willprovide a better and more uniform sharing of current byall the conductors. The current rating may now bedetermined by multiplying the individual current ratingof each split phase by the number of parallel circuits. Asthe space between the two groups of the same phase willnow be large, 300 mm or more, they will have nil or onlynegligible influence of skin effect among themselves.

fromCarrying power through metal-enclosed bus systems 281889For instance, a bus system with 4 x 152.4 x 6.35 mmconductors may be arranged into two groups of twoconductors each, according to Figure 28.29(b). Then theimproved rating of this system as in Table 30.4 will be= 2 x 2860= 5720 Aas against4240 A for all conductors put together as in Figure28.331a) or1.18 x 4240, i.e. 5003 A when arranged as inarrangement 2 of Figure 28.14.Thus. in phase interleaving, there will be better utilizationof conductor capacity by 5720/5003 or >14% overarrangement 2 in Figure 28.14.Example 28.1 1Consider Example 28.7 again, using four sections of 101.6 x6.35 mm AI conductors, now interleaved as shown in Figure28.30. To determine the improved reactance and resistanceof this arrangement we can proceed as follows.Proximity effect:a = 19.05 mmS = 94.05 mm:. ~Ra,R dCthe graph in Figure 28.1 3 (a) by interpolation foran EIE-M grade of aluminium = 1.1 3Rdc = 0.056 Q/lOOO m per conductor= 0.056/4 R/1000 m for 4 conductors0.056:. R,, = 1.13 x -4= 0.0158 R/1000 m:. Impedance Z = 1/0.0158' + 0.045'= 0.0477 R/1000 mVoltage dropAccordingly the revised voltage drop for 40 m of bus length= 200040x 0.0477 x ~1000= 3.82 Vwhich is even less than 1% for a 415 V bus systemElectrodynamic forcesFor a system fault level of 50 kA maximum forces on eachgroup,(i) For a conventional arrangement (Figure 28.17(a))S 1.26 xand space factor, 1.26 x -94.05-a+ b 19.05+ 101.6= 0.98X, from graph of Figure 28.24 = 90 pWmand for two parallel circuits = 90/2= 45 pR/m or 0.045 R/1000 mas against 125 @/m with the conventional arrangementcalculated in Example 28.9.Skin effectArea of cross-section per split phase = 2 x 101.6 x 6.35b 101.6- = - = 5.33.a 19.05R6.35+ p-6.35B= 12.9 cm2a = 19.05 mmb = 101.60 mmS = 94.05 mm(Depending upon the current rating, it would be advisable to keep itminimum 300 mm, by increasing the gap between the split phases.)Figure 28.30 Illustrating Example 28.1 1Kfor a space factor of S-a - 184'45 - 44.45 i.e. 0.958a + b 44.45 + 101.6corresponding to a or - 44.45 i.e. 0.4375b 101.6from the graph in Figure 28.7 by interpolation K = 0.96. 0.96 x 16 x 50000' 10-4" rn -184.45= 20 819 N/m(ii) For the improvised interleaving arrangement in Figure 28.30:K for a space factor of 94.05 - 19.05 i.e. 0.62 from the19.05 + 101.6graph in Figure 28.7 corresponding toa of 19.05 i.e. 0.1875 = 0.87b 101.6:. F, on each set of supports:.F,, on both supports = 2 x 9250.40.87 x 16 x [ 502000 )*- x lo494.05= 9250.4 Nlm= 18 500.8 Nim.This is less than the force developed with the conventionalarrangement in Figure 28.17(a).3 Phase transpositionIn this arrangement the three-phase conductors are evenlytransposed in a length of busbars by interchanging their

~ 1/328/890 Industrial Power Engineering and Applications Handbookphysical location so that each phase is under an equallyinductive effect produced by the other two phases. Thearrangement is illustrated in Figure 28.31 This can beperformed by arranging a straight length of a bus intothree equal sections (or in multiples of three), as shown.If x is the reactance of phases R and B, and y that ofphase Y in the first section, then phases B and Y will havereactance x and R a reactance y in the second section. Inthe third section, phases Y and R will have a reactance ofx each and phase B will have a reactance y. Hence, thereactance of each phase, at the end of the three lengths,will be balanced at (2x + y), causing equal load sharingand an equal voltage drop in all three phases. Thisarrangement would thus make the system almost balancedinductively by each phase having equal exposure to theinductive fields produced by the other two phases. Dueto inductive balancing, the transposition equalizes thereactances in each phase and improves the current sharingby all the three phases, besides an equal voltage dropthrough the length of the bus.However, there may not be an appreciable improvementin the proximity effect between each section, unless thetranspositions are increased infinitely, as in the case of astranded three-phase cable which has continuously twistedconductors and represents an ideal transposition. Inaddition, there is no change in the skin effect. Thisarrangement therefore has the purpose primarily ofachieving an inductively balanced system and hence abalanced sharing of load and equal phase voltages at thefar end.It is also cumbersome to arrange a bus system withphase transposition. This technique has therefore not foundmany applications in a bus system. It is more useful indealing with inductive interference in communicationlines (Section 23.5.2(C)).4 Changing the configuration of busbarsBy arranging the busbars into a few more configurationsalong the lines discussed above it is possible to reducethe proximity effect to a great extent. Some of theseconfigurations are illustrated in Figure 28.14. See alsoExample 28.12, illustrating the marked improvement inthe capacity utilization of the busbars by using differentconfigurations.RXXXYYYYYBXXXR+ 1/37 1'(/3 4&=2x+y Where X,, X, and XB are thexy =2x+y reactances of each phasex, =2x+yFigure 28.31transpositionBalancing of reactances through phaseFMS cover \0000000Magneticloopsoc300o0Each loopcauses 000000magneticlossoooooooo(a) Breaking of electrical paths do not diminish the magnetic fieldAluminium top andbottom covers\-MS sidecoversI(b) An economical and low loss enclosureFigure 28.32 Magnetic field in a magnetic material5 Busbar enclosureNon-magnetic enclosure. The proximity effect can alsobe minimized by using a non-magnetic enclosure ofaluminium 01- stainless steel. In magnetic materialsthe field in the enclosure is produced in the form ofsmall magnetic loops. Its effect cannot be mitigatedby breaking the electrical path alone, as illustrated inFigure 28.32. Its effect can be diminished only byreplacing a few parts of the magnetic enclosure itself,such as its top or bottom covers or both, with a nonmagneticmaterial. It is possible to achieve aneconomical and low-loss enclosure by replacing onlyits top and bottom covers with a non-magnetic material.The covers constitute the larger part of the surfacearea of the enclosure.By providing adequate louvres in the enclosure as shownin Figure 28.33(b) or by using a forced-air draughtthrough the length of the enclosure.,I

28.9 Sample calculations fordesigning a 2500 A non-isolatedphase aluminium busbar systemExample 28.12Design parametersSupply system three-phase four-wire 415 V ? IO%,50 HZ ? 3%Fault level45 kADuration of fault1 second870f 1Insulator supporting(All dimensions in mm )MS or aluminiumenclosureFigure 28.33(a) Busbar arrangement and enclosure size forbus duct of ExamDle 28 12Carrying power through metal-enclosed bus systems 28/891Continuous current rating2500 AAmbient temperature 50°CMaximum permissible operating temperature 85°CPermissible final temperature at the end of thefault 185°C(A) Rectangular sections(i) Minimum size of busbars for short-circuit conditionsThe minimum size of busbars for an operating temperatureof 85°C and a final temperature of 185°C can be ascertainedfrom the curves of Figure 28.5, suggestingr. 4 = 0.0799Aor A=-. 45 4i0.0799= 563.2 sq. mmMaximum temperature rise of the busbars at the rated current= 85 - 50= 35°CAssume the temperature of the busbars at the time of fault =85°C and rectangular flats of electrolytic grade E-91 E or itsequivalent. Busbars chosen for each phase - four (152.4mm x 6.35 mm) - which are more than the minimum sizerequired to account for the thermal effects during a shortcircuit conditionfor neutral -two (52.4 mm x 6.35 mm)Size of busbar enclosure - 870 mm x 380 mm (Figure28.33(b))Ground terminalSide viewTop view (without cover)Legend1. M.S. anqle4. Bottom and top covers 2 mm M.S. or 3 mm AI. 7. Gasket2. AI. bus i x 152.4 x 6.35 mm 5. Ground bus 50 x 6 mm 8. Louvres3. Side frames 2 mm M.S. or 3 mm AI. 6. Insulators9. Metallic spacers(All dimensions in mm)Figure 28.33(b)General arrangement of a typical running section of the busduct of Figure 28.33(a)

28/892 Industrial Power Engineering and Applications HandbookMaterial of enclosure - aluminiumBusbar configuration - as Figure 28.33aBusbar support made of SMC or DMC (Section 13.6.1).Distance between two busbar supports: 400 mm (Figure28.33(b))(ii) Data availableAluminium busbars, from Tables 30.4 and 30.6Electrical conductivity = 1Minimum tensile strength = 2050 kgf/cm2 (Table 30.1)Minimum cross-breaking or yield strength = 1650 kgfkm’(Table 30.1)Busbar supports: from Table 13.1 4Mechanical properties DMC SMCin kgf/cn? in kgf/cn?Minimum tensile or shearing 250-500 500-900strengthMinimum compressive 1200-1 800 1600-2000strengthMinimum cross-breaking or 700-1200 1400-1 800flexural strength (bendingstrength)HardwareMechanical propertiesHigh tensile (HT) Ordinary MSfasteners as in (mild steel)IS0 4014 fasteners asand IS0 898, in IS0 4016,grade 8.8 grade 4.6Minimum tensilestrength 8000 kgf/cm2 4000 kgf/cm2Minimum cross-breakingor yield strength 4400 kgflcm’ 2200 kgf/cm2(Iii) Deriving the actual current ratingApplicable deratings:Due to higher ambient temperatureFor 50°C as in Table 28.3 and Figure 28.10 = 0.815Due to altitudeNil, since the installation of the equipment is assumed tobe within 2000 m above the mean sea levelDue to grade of busbarsFor E-91 E or its equivalent, as in Table 30.6 = 1 .ODue to size of enclosure and environmental conditions oflocationThe Enclosure is of nonmagnetic material, therefore it willbe devoid of hysteresis and eddy current losses.Heat dissipation factor- Cross-sectional area of active aluminiumArea of enclosurei.e.(3 x 4 x 152.4 x 6.35) + (2 x 152.4 x 6.35)380 x 870- 14 x 152.4 x 6.35380 x 870= 0.041or 4%As in Table 28.6, by simple interpolation, for condition oflocation against serial number 2, derating = 0.77Due to skin effectNil, as it is already considered in Tables 30.2, 30.4 and30.5, while establishing the basic ratings of the busbars.Due to proximity effectApproximately 20%, since S < 2b. It is recommended tohave the centre spacing S at least 2 x 152.4, Le. 305 mm.If the width of the enclosure poses a limitation, a moreappropriate configuration such as in Figure 28.34 or thetechnique of interleaving as in Figure 28.35 may be adoptedto achieve better utilization of the active metal. In ourcalculations we have considered all these alternatives.Due to voltage variationWe have already discussed the impact of voltage variationon an industrial drive in Section 1.6.2(A-iii). The impact ofthis on a bus system may not be the same. A bus mayhave to supply lighting, heating and other resistive orinductive loads. All such loads except electric motors willperform low at lower voltages and hence draw a lowercurrent. Generally, we can assume that the loading on abus, as enhanced by the industrial drives, will be almostoffset by the decrease of loading by the other loads if weassume industrial drives to be up to 50% of the totalconnected load. Usually therefore no derating will benecessary for a lower voltage, except for large installationswhere the drives may constitute the bulk of the load.Consideration of voltage variation will therefore dependupon the type of installation. In our calculations, however,we are ignoring the impact of this.Frequency variationAt higher frequencies up to 3% skin and proximity effectswould be slightly higher, but can be ignored as their impactwill be only marginal.Considering all these deratings the total derating for theconfiguration of Figure 28.33(a) or (b) will be= 0.815 x 1 x 1 x 0.77 x 0.8= 0.50Basic current rating of 4 x 152.4 mm x 6.35 mm aluminiumbusbars per phase as in Table 30.4= 4240 A:. Effective rating after considering all possible deratings= 4240 x 0.5= 2120 AThese sections of busbars are not adequate for the requiredcurrent rating of 2500 A. The rating of the bus system can,however, be improved by almost 20% and make it suitablefor the required rating by providing the busbars and theinside of the enclosure with a nonmetallic, matt finishblack paint. If voltage variation is also to be considered,then this bus may not be suitable for the required dutyeven after painting.(iv) Voltage dropRdc for E-91 E grade of conductor, from Table 30.7= 32.38 pWm per conductor at 20°C:. Rdc at an operating temperature of 85°C= Rdc20 [l + a20 (02 - el)]= 32.38 [l + 0.00363 (85 - 20)]= 32.38 (1 + 0.236)or Rdc = 40.02 pR/m per conductor andfor the phase = ~40.02 or 10.005 x4Q/1000 m

~wE IIIIIIIIIIirIIII +IIIIIIIIIIIR Y B N6 ilNon-magnetic insulatorsupports of SMC/DMSaluminium orstainless steel etcFigure 28.34 Illustration of Example 28.12R Y Bi.e. 1.344, as in the curve, corresponding to alb, as- i.e. 0.29,X, = 110 pQ/m per phase or 110 xphase.Impedance,Q/1000 m per-- ++I- -li-6 35 a = 19.05 6.35Figure 28.35 For Example 28.1 2(All dimensions in mm)Area of cross-section per phase = 4 x 152.4 x 6.35 mm2= 38.71 cm2For this area of cross-section, the skin effect ratio Ra,/Rd,from Figure 28.13(a) for aluminium grade E-91E at 85"C,having bla = 152.4144.45 = 3.43 measures almost 1.425by approximating the interpolation,:. RacIRdc = 1.425i.e. an increase of almost 42.5%, due to the skin effectalone.:. R, = 10.005 x 10-3 x 1.425= 14.257 x CY1000 m per phaseProximity effectMeasure reactance X, from Figure 28.24 for1.26 x- =1.26x210a+b 44.45 + 152.4Z= 414.257' + 110' x= 0.111 R/1000 m per phaseFor a 50 m length of bus duct this impedance will cause avoltage drop of2500~0.11150x - = 13.9V1000which is 3.3% of the rated voltage and is thereforeacceptable. For higher bus lengths, however, which maybe rare, while the current rating of the system selected willbe suitable the voltage drop may exceed the recommendedlimits. In this case it will be advisable to adopt the alternativeconfiguration of Figure 28.34 or the technique of interleaving(Figure 28.35). A comparison is drawn for these configurationsas in Table 28.8, which reveals that both alternativessignificantly improve the performance of the same bussection.(v) Effect of proximity on the centre phase YFor a length of 50 m the voltage drop in phase Y, due to R,,,and assuming the content of X, = 0= 2500 x 14.257 x 10-3 x - 501000= 1.78 V

28/894 Industrial Power Engineering and Applications Handbook:. Receiving side voltage of phases R and B- 415 13.96= 225.7 V41 5and of phase Y = - 1.78fi= 237.8 VThe imbalance for this length and rating of bus system isnot substantial, yet if we assume that a balanced supplysource is desirable, then we must make up the lost inductancein phase Yby inserting a reactor into this phase, as discussedin Section 28.8.2 of an equal value of X,, i.e.X, =110~10-~ x-=0.0055R501000(vi) Calculation for short-circuit effectsElectrodynamic forcesThese can be determined from equation (28.4),F, =16. l& K. lo4SN/mwhereI,, = r.m.s. value of fault current in Amperes = 45 000 AK = space factor for rectangular conductors, determined fromthe curves of Figure 28.7, corresponding toor 0.84- S- a i,e, 210 - 44.45a+b 44.45 + 152.4corresponding to the curve for alb = 44.45D52.4 or 0.29Assuming the curve for alb = 0.25 with little error, K = 0.93As in Figure 28.33(a):. F,,, =16 x (45 000)' x 0.9321 0= 14 348.6 N/m.. 1 N/m =- kgf/m9.807:. F, - 1463.1 kgflmS =centre spacing between twophases = 210 mma = space occupied by the conductorsof one phase = 44.45 mmb = width of the busbars = 152.4 mmN/mSince the busbar supports are assumed to be at a distanceof 400 mm,:. force on each section of busbars, insulators and themounting fasteners= 1463.1 x 0.4= 585.24 kgWe have drawn a comparison of these forces for the otherbusbar configurations in Table 28.8 for more clarity.Since Figure 28.34 is found to be a better arrangement,we have considered forces as in this arrangement only in allour subsequent calculations.(vii) Mechanical suitability of busbars and theirsupporting systemBelow we analyse the adequacy and the suitability of busbars,fasteners and the insulators supporting the busbars, towithstand the above forces acting differently at differentlocations.Bending stresses on the busbarsBending stress at section x - x = = kg/cm'12.M.N(28.10)whereF, = maximum electrodynamic forces acting on each support,in the event of a fault, as calculated above = 514 kgfI = centre distance between two busbar supports = 40 cmM = sectional modulus of each busbar at section x - x= 1 a . b' in cm36xlx-! 1where for a 152.4 mm x 6.35 mm busbar sectiona = 6.35 mmb = 152.4 mm, M= 1 6.35 x 152.4' cm36 1000= 24.58 cm3 (as indicated in Table 30.7)N = number of busbars per phase = 4:. Bending stress = 514 40 kg/cm212 x 24.58 x 4= 17.43 kg/cm2To calculate the sectional modulus (or moment of resistance)of the four bus sections in parallel we have multiplied thesectional modulus of one bus by 4. This is a simple methodwhen the busbars of each phase are in the same plane andequally spaced as in Figure 28.33(a) with no additional spacersbetween them to hold them together.But when other configurations are adopted as shown inFigure 28.34, this concept may not hold true. In otherconfigurations, however, the sectional modulus will only riseand reduce the bending stress on the busbars.The method adopted to calculate the sectional modulus istherefore simple and on the safe side. However, to calculatethe sectional modulus more accurately or to derive it for anyother section of the support than considered here referencemay be made to a textbook on the strength of materials or amachine handbook.NoteThe sectional modulus, when required, can be increased byproviding spacers between the straight lengths of bus sections,as shown in Figure 28.33(b). The spacers, when provided,can make them more rigid and add to their bending strengthdue to higher sectional modulus. At joints these spacers occurautomatically in the form of overlapping of busbars or fishplates(Figures 29.4 and 29.5). The spacers prevent the bus lengthsfrom being deflected towards each other.

~~Carrying power through metal-enclosed bus systems 281895Table 28.8 Calculation of electrodynamic forcesDescription Arrangement as in Arrangement as in Interleaving as inFigure 28.33(a) Figure 28.34 Figure 28.35Basic current rating without 4240 A 4240 x 1.572 x 2860 (2860 A is the ratingderating = 6657 A of one split phase Table 30.4)= 5720 AEffective current rating 0.5 x 4240 0.5 x 6657 0.5 x 5720considering approximately = 2120 A = 3328 Aa = 2860 Aathe same deratingsProximity effect- -- 44'45 - 0.29 -- --On per split phase basisa19'05 - 0.04419'05 - 0.125b152.4430.8152.4S 1.26 x 210 = 1.344 1.26 x 180 = o,504 1.26 x 99.05 = o.731.26 -a+b44.45 + 152.4 19.05 + 430.819.05 + 152.4Approx. Xa from Figure 28.24 110 @/m 64 pQ/m 74 pnlm per circuit.Since there are two parallelcircuits,:. Combined X, = 37 pWmSkin effectArea of cross-section- ba4 x 152.4 x 6.35= 38.71 cm2-e 152.4 3.4344.454 x 152.4 x 6.35= 38.71 cm2-- 430'8 - 22.619.052 x 152.4 x 6.35= 19.35 cm2-- 152.4 -19.05Approx.R*cfrom Figure 28.13(a)1.4251.31.18Rdc as calculated instep iv:. R,40.02 @/mper conductor- 1.425 x 40.024= 14.257 @m40.02 pnlmper conductor- 1.3 x 40.024= 13.01 pQ/m40.02 pnlmper conductor- 1.18x40.024= 11.81 pnlmImpedance 2dl 10' + 14.2572= 111 m m464' + 13.01*= 65.31 @m437' + 11.81'= 38.84 @/mVoltage dropas % of system voltageElectrodynamic forcesS-aa+b- abK from Figure 28.7F,,, in N/mIn kgfh2500 x 50 x 111 x lo4 2500 x 50 x 65.31 x 2500 x 50 x 38.84 x=13.9.100 = 8.16 V = 4.85 V4.513.9 x 100 = 3.3%4.58.16x100=

28/896 Industrial Power Engineering and Applications HandbookJnferenceThe minimum shearing strength of aluminium is 1650 kg/crn2(Table 30.1) which is much larger than the actual force towhich the busbars will be subject, in the event of a fault.They are thus more than adequate in cross-section andnumbers. Other than bending stress, there is no significanttensile or shearing force acting on the busbars.Suitability of fastenersOn fasteners also, other than cross-breaking or shearingstress, there is no significant compressive or bending force.Fasteners for busbar supportsAs in Figure 28.34, each phase of four aluminium sections issupported on four two-way insulators. Each insulator ismounted on 2 x M8 size of bolts (diameter of bolt shank, 8mm).:. Total number of bolts = 4 x 2 = 8 numbers of size M8But it is possible that the first peak of the force, Fm, may acteither downwards or upwards. In which case only four fastenerswould be sharing the force at a time. Stress area of four bolts= 4 x x 8' sq.mm4= 2.01 cm2Cross-breaking strength of(i) Ordinary MS fasteners as in IS0 4016 of grade 4.6,= 2200 kg/cm2 (minimum):.Total force they can withstand considering a factor ofsafety as 100%(ii) High tensile fasteners as in IS0 4014 and IS0 898 ofgrade 8.8,:.= 4400 kg/cm' (minimum)Total force they can withstandIn this particular instance, due to the large number of fasteners,even an ordinary type of MS fastener will be suitable.Notes1 The above situation may not always be true particularlywhen the current rating is low, say up to 600 A and thesystem fault level is still high. In this case much less crosssectionof aluminium would be used, and the number ofsupports and fasteners would also be less. Then thefasteners will also be of smaller cross-sections. In suchcases the suitability of fasteners will be more relevant.2 For busbar joints, however, use of high tensile fastenersalone is recommended with a view to ensure adequatecontact pressure per unit area, as discussed in Section29.2 and noted in Table 29.1 which an ordinary fastenermay not be able to maintain over long periods.Fasteners for busbar jointsFor busbar joints we have considered 8 Numbers bolts ofsize M-10 (diameter of bolt shank, 10 mm), as in Figure 29.4.As the size of these fasteners is greater than that of thebusbar supports, their suitability is not determined separately.Suitability of insulatorsThe busbar support is the most vulnerable component in acurrent-carrying system. It has to withstand all kinds of stressesdeveloped by the busbars on a fault. From Figure 28.36 wehave identified the following likely vulnerable locations in asupport that may yield on the occurrence of a fault;Finger between the two busbars, section a-a, which mayshear off from its roots.At the bolt mounting holes section y-y.At the wedges marked with hatching.But the insulator is more vulnerable at a-a, than at the wedges,since the shear area at a-a is only 0.6 x 1.5 cm2 comparedto 1.98 x 1.5 cm2 at the wedges. Hence, for brevity, weanalyse possibilities (i) and (ii) only.In Table 28.9 we have evaluated the cross-breaking,shearing and bending (flexural) stresses, that may act atsuch locations, to establish the suitability of the supportsused.NoteFor more clarity on the subject, in the light of note 1 abovewe have also worked out the earlier Example 28.6, for shortcircuitconditions and then analysed the above details for thisarrangement. AssumingandlSc = 45 kAS= 100 rnm (Figure 28.16)K= 0.9 from graph of Figure 28.7 for-- S-a 100-6.35a+b 6.35+101.6= 0.87and alb = 6.351101.6 Le. 0.0625Choosing the curve for d b = 0.1. A slight interpolation in thiscurve will determine K as 0.9 for a/b as 0.062516 x (45000)2 x 0.9.'. F, =100= 29 160 N/mor 2973.4 kglmx lod N/mAssuming the distance between each busbar support of thesame phase to be 400 mm then the force on each section ofbusbars, insulators and the fasteners.= 2973.4 x 0.4 kg= 1189.36 kgSuitability of busbars; for the given parameters:F,,, = 1189.36 kgt = 40 cm= 1 6.35 x 101.6' cm36 1000= 10.925 cm3 andN= 1:. Sending stress =1189.36 x 4012 x 10.925 x 1= 362.89 kg/cm2which is much less than the cross-breaking strength ofaluminium.

FL Bus lengthY A=15a= 10 R YCarrying power through metal-enclosed bus systems 2818972 x 8.54A= 15I I ISide view(All dimensions in mm)I_ 210-15=214 e = 13ElevationFigure 28.36 Mounting arrangement of busbars for Example 28.12Detail of each supportSuitability of fastenersAs in Figure 28.36, although each phase will be supportedon two insulators and each insulator mounted on 2 x M8 sizeof fasteners, it is possible that at the instant of fault, theforces are acting either upwards or downwards. Therefore,assuming forces to be acting only on two fasteners, at theinstant of fault and assuming a factor of safety as 1 OO%,Cross-breaking strength of these fasteners= 1 x 2 x x 82 x 2200 kg (for ordinary MS fasteners)2 4 100= 11 05.28 kgwhich is marginal compared to F,,, of 1189.36 kg.It is advisable to make use of high tensile fasteners only.Suitability of insulatorsWe give in Table 28.9 a comparison of different types offorces that the insulators may have to encounter during afault condition at different locations.(B) Channel sectionsConsider channels in a box form for the same requirementsand operating conditions as for rectangular sections.Choose two channels for the phases and one for the neutralof size 127 mm as in Table 30.9 and let them be arranged asshown in Figure 28.38.2 x 8.54-b@G$ys= 100(All dimensions in mm)0 Electrical conductivity for grade E-91 E = 1 .ODerating for an ambient of 50°C = 0.815Figure 28.37 Mounting arrangement of busbars forExample 28.6

~~~~28/898 Industrial Power Engineering and Applications HandbookTable 28.9 Checking suitability of insulators1 Cross-breaking stress X, atsection a - a1.5 F, . IXb =- A.5'Shared byFactor of safety:. Minimum cross-breaking stress thesupports should be able to withstandInference.~Shearing stress S, at section a - as, = 5A(A = cross-sectional area of section a - a)Shared byFactor of safety:. Min. shearing stress thesupport should be able to withstandInference3 Bending (cantilever)or flexural stress, 6, atF, . LSection y-y, ~MwhereM= 1. a. b26Shared byFactor of safety:. Min. bending stress the supportsmay have to withstandExample 28.12Figures 28.34 and 28.36F,, = 514 kgflcm'I = 1.3 cmA = 1.5 cmB = 0.6 cmx, = 1.5 x 514 x 1.31.5 x 0.6'= 1856 kgflcm'4 fingeres100%- 18564= 928 kgflcm'Only SMC supports must be usedA = 0.6 x 1.5 cm2:. s, =514~0.6 x 1.5= 571 kgflcm24 fingers100%--571 x 24= 285.5 kgWcm'DMC supports may be usedbut limiting factor is cross-breakingstress hence only SMC supportsmust be usedL = 3.9 cma = 1 .O cmb = 6 - 2 x 0.85= 4.3 cm:. M = 1 x 1 .o x (4.3)Z6= 3.08 cm3514 xand B, =3.9~3.08= 650.8 kgflcm'4 supports100%- 650.8 --4= 325.4 kgflcm2Example 28.6Figures 28.16 and 28.37F,, = 1189.36L = 1.3 cmA = 1.5 cmB = 2.24 cm. - 1.5 x 1189.36 x 1.3'. b -1.5 x 2.24'= 308 kgflcm24 wedges100%= m x 24= 154 kgf/cm2DMC supports can be usedA = 2.24 x 1.5 cm2:. s, =1189.36~2.24 x 1.5= 354 kgflcrn'4 wedges1000/0= E x 24= 177 kgfhm'DMC supports can be usedL = 3.9 cma = 1.0 cmb = 5.2 - 2 x 0.85= 3.5 cm:. M = 1 x 1.0 x 3.526= 2.04 cm3and - 1189.36 x 3.9s -2.04= 2273.8 kgf/cm22 supports100%-- 2273.82= 2273.8 kgf/cm2(Contd.)

~~Carrying power through metal-enclosed bus systems 28/899Table 28.9 Contd.InferenceConclusionExample 28.72 Example 28.6Figures 28.34 and 28.36 Figures 28.76 and 28.37DMC supports may be acceptablebut limiting factors is crossbreakingstress, atlocation a - a. Hence only SMCsupports must be usedIn this particular instance, sectiona - a at the fingers is morevulnerable. If the supportsfail, they may fail from here.SMC supports alone must beused with a modified design,to withstand a higher bendingstress or the design of thebus system itself be modifiedas mentioned below undernote 3In this case, section y - y at thebolt mounting holes is morevulnerable If the supportsfail, they may fail from here~~ -~~NotesFactor of safetyIt is possible that during the fault only one of the insulators is subject to the transitory first peak of the fault, as there may be slightmisalignment between the insulators, asymmetry in the busbars, an imperfect bolt fixing and their fastening, or a combination of such factors.To be on the safe side it is advisable to consider each support and its fasteners to be suitable to withstand the forces by themselves We haveassumed a factor of safety of 100% in all the above calculations to account for this.F,,, for all particular fault levels remains the same, irrespective of the current rating of the system (equation (28.4). A low current systememploys fewer and smaller busbars and is supported on less and smaller supports and hardware. Such a system therefore may have towithstand much higher stresses than a higher current-carrying system and is more likely to yield to such stresses unless adequate measuresare taken while selecting the size of supporting hardware or spacing (t) between the adjacent horizontal supports.When a component supporting the current-carrying system is likely to be subject to severe stresses, and its own stress-bearing capacity maybe marginal as in Example 28.6, it is advisable to take a higher size of such a component in its thickness or diameter or a better qualitymaterial. It is also possible to mitigate the stresses by reducing the distance between the two mounting supports of the same phase (i to be

5028/900 Industrial Power Engineering and Applications HandbookProximity effectFrom Figure 28.1 9(c) conductor effective spacingse = 3JSAB ' SBC ' ScASA,= Sac = 302 mmand SC, = 2 . SA, = 604 mm.'. Se = 3d302 x 302 x 604= 380.5 mmand X, from graph of 127 mm channel = 108= 0.108 R/1000 m per phaseImpedanceZ= J0.01195z +0.108'= 0.108 WOO0 mVoltage drop for 50 m length in phases R and B50= 2500 x 0.108 x ~1000= 13.5 Vwhich is 3.25% for a system voltage of 415 V. The systemchosen is good up to a length of 50 m. Beyond this thevoltage drop may exceed the desirable limits and requireeither larger channels or a reduced centre spacing S. Reducedspacing S is possible in this instance due to the sufficientmargin available in the rating of the channel section chosen.(viii) Effect of proximity on the centre phase YLet us assume a length of 50 m. The voltage drop in phaseV, because of Rac, assuming content of X, = 050= 2500 x 0.01195 x ~1000= 1.49 v:. Supply-side voltages of phases R and 6- 415&13.5 = 226 V415and of phase Y = - 1.49 = 238 V43which will provide a reasonably balanced system. To make itmore balanced a reactor can be inserted into the middle- LIFigure 28.39sections1165(A// dimensions m mm)Illustrating Example 28.12(c), with tubular busphase along similar lines to those calculated in Example28.10 for a reactance ofX, = 0.108 x ~ = 5.40 x 10-3 R1000For the rest of the calculations for mechanical suitability ofthe busbar system the procedure for rectangular sectionscan be followed.(C) Tubular sectionsNow consider a tubular section for the same requirements.Choose a tube of size 5" (standard pipe) from Table 30.8having the following dimensions,For phasesOutside diameter (OD) = 141.30 mm = 2 r,Inside diameter (ID) = 128.20 mm = 2 . r2A = 2724 mm2Nominal current rating = 3550 AFor a neutral choose a tube of size 3", havingOD = 88.90 mmand ID = 77.92 mmA = 1439 mm2Arrange them as in Figure 28.39.Likely deratings:For grade of busbars, for E-91 E = 1 .OFor ambient temperature = 0.815For altitude = 1 .OFor enclosure factor- 3 x 2724 + 1 x 1439450 x 1165= 1.83%:. derating = 0.83For proximity = 0.9 (such sections, because of their shape,would have a low proximity effect as noted later):. Total derating = 1 x 0.815 x 1 x 0.83 x 0.9= 0.609:. Actual current rating = 0.609 x 3550= 2162 AThe busbars and the inside of the enclosure may be paintedwith matt finish black paint to make this section suitable for= 1.2 x 2162= 2594 AIf voltage variation is also to be considered, then one mayhave to choose the next higher size. As in Table 30.8, onecan choose an extra-heavy pipe of 5" size.Voltage dropFor skin effect:d.c. resistance ffdc20 = 11.3 pWm from Table 30.8.. RdCa5 = 11.3 x 1.236' (refer to step (iv) above)= 13.97 pQ/mActive area per phase = 2724 rnm2:. skin effect ratio from Figure 28.13(b)

Relevant StandardsCarrying power through metal-enclosed bus systems 28/901IEC Title IS BS IS0600591199960439- 11199260439-2/1987Standard current ratings (Based on Renald seriesR-10 of 60-3)Low voltage switchgear and controlgear assembliesSpecification for type-tested and partially type testedassembliesLow voltage switchgear and controlgear assembliesParticular requirements for busbar trunking systemsHexagon head bolts. Product grade CHexagon head screws. Product grade CHexagon nuts. Product grade CHexagon head bolts. Product grades A and BHexagon head screws. Product grades A and BHexagon nuts, style 1. Product grades A and BHexagon thin nuts (chamfered).Product grades A and BHexagon thin nuts. Product grade BMechanical properties of fastenersInterconnecting busbars for a.c.voltage above 1kV up to and including 36 kVCriteria for earthquake resistant design of structuresPlain washers8623-1119938623-2/19931363- 111 9921363-2119921363-311 9921364-1119921364-211 9921364-3/19921364-4119921364-5/19921367-3,6,208084-19921893/19912016/1996Relevant US Standards ANSWNEMA and IEEEBS EN 60439-111994BS EN 60439-211993BS EN 2401611992BS EN 24018/1992BS EN 24034/1992BS EN 24014/1992BS EN 2401711992BS EN 2403211992BS EN 24035/1992BS EN 2403611992BS EN 20898-1,2,7BS 159/1992DD ENV 1998 (1 to 5)-31197340161198840 1 811 9884034/19864014/1986401 7/19884032/19864035/19864036/1979898-1,2,7NEMA L11/1983ANSYIEEEC37.23/1992Industrial laminated thermosetting productsMetal enclosed bus and guide for calculating losses in isolated phase busNotes1 In the tables of Relevant Standards in this book while the latest editions of the Standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of Standards is a continuous process bydifferent Standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a Standard.2 Some of the BS or IS standards mentioned against IEC may not be identical.3 The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publication.= It300 x 300 x 600= 1.26 x 300= 378 mm:. RaCB5 = Rdd5 = 13.97 pQlm= 0.01397 QAOOO mFor proximity effect:128.2 141.3r, 2 ' 2= 0.907From Figure 28.21 5 = - A -Corresponding to this Ds/r, = 0.967:. D, = 0.967 x -141.3= 68.3 mmand2se = qSAB 'SBC '&A(refer to Figure 28.39)Corresponding to this, the reactance from Figure 28.19(b)can be determined by extrapolation. We have assumed it tobe 0.06 R11000 m per phase.ImpedanceZ= ,/0.01397' + 0.06'= 0.061 6W1000 mVoltage drop for 50 m length in phases R and B50= 2500 x 0.0616 x -1000= 7.7 vwhich is only 1.86% for a system voltage of 415 V, and issatisfactory.

28/902 Industrial Power Engineering and Applications HandbookEffect of proximity on the centre phase YThe voltage drop in this phase, assuming X, = 0= 2500 x 0.01397 x 501000= 1.75 vwhich is only 0.4% of the system voltage. This bus systemwill thus provide a near-balanced system.For the rest of the calculations for the mechanical suitabilityof a busbar system the procedure for rectangular sectionscan be followed.List of formulae usedShort-circuit effects(1) Thermal effectse---. - l;o (2- )* . (1 + a,,0). t (28.1)0, = temperature rise in "CI,, = symmetrical fault current r.m.s. in AmpA = cross-sectional area of the conductor in mm2= temperature coefficient of resistance at 20"C/"C6 = operating temperature of the conductor at whichthe fault occurs in "CK = 1.166 for aluminium and 0.52 for coppert = duration of fault in secondsor x & = 0.0799 for aluminium for an operatingAtemperature at 85°C and end temperatureat 185°C (28.2)I,, JI = 0.12 for copper for an operating tempera-A(2) Electrodynamic effectsture at 85°C and end a temperature at185°C(28.3)16.1;F,, = k .- x 104 N/m(28.4)SF, = maximum dynamic force that may develop on afaultI,, = r.m.s. value of the symmetrical fault current inAmpsk = space factorS = centre spacing between two phases in mmSkin effectEffect on current-carrying capacityI,, = I,,'I,, = permissible current capacity of the systemRdc = d.c. resistanceR,, = a.c. resistanceId, = d.c. currentConductor resistance at higher temperature(28.5)Rdc at 85°C = Rdc20 [1 + az"(62 - e,)] (28.6)azo = temperature coefficient of resistance at20°C per "CRdcZo = d.c. resistance at 20°C13~ = operating temperature = 85°CO1 = since the value of R,, is available at 20°Ctherefore, 01 = 20°CProximity effect in terms of busbar reactances, = (Sa ' s, ' sy3(28.7)S, = effective or geometric mean spacingSa, Sb, and S, = spacing between conductorsUse of saturable reactor to balance a largeunbalanced power distribution systemTo determine sizeXp = lost reactance of Y phaseL = inductance of XpI, = current in R or B phaseIY = current in Y phaseReluctance of the magnetic path(28.8)(28.9)I, = length of the air gap in metrespo = permeability of air (free space) = 4n. H/m= relative permeability of the silicon steel used forthe laminates in H/mA = area of cross-section of core in square metresk = total length of the magnetic circuit in metresCalculating stresses on a faultBending stress on busbars at sectionF, . Ix -x= = kg/cm(28.10)12,M.NF, = maximum electrodynamic forces acting on eachsupport in the event of a fault1 = centre distance between two busbar supportsM = sectional modulus of each busbar at sectionx-x1= -a. b2 in cm36N = number of busbars per phaseFurther reading1 ERDA, 'Study on feasibility of upgrading the operatingtemperature of AI busbars without plating'.2 Golding, E.W., Electrical Measurements and MeasuringInstruments.3 Lynthall, R.T., The J & P Switchgear Book. Buttenvorth, London.4 Thomas, A.G. and Rata, P.J.H., Aluminium Busbar. HutchinsonScientific and Technical for Alcan Industries Ltd.

RecommendedPractices forMounting Busesand Making BusJoints29/903

Recommended practices buses and making bus joints 29190529.1 Precautions in mountinginsulators and conductorsOften a failure on a fault may be due not to the inadequatesize of busbars, fasteners or insulators but to pooralignment of the insulators or to too large a gap betweenthe busbar and the insulator slots. It may he a consequenceof an inappropriate mounting or unequal width of thebusbars or insulator slots. In such cases, load sharingwill be uneven and the weakest section may fail. Thiscan be illustrated as follows:1 As shown in Figure 29.1 (a) as a result of loose fit ofbusbars with an unequal gap, the insulators (shadedin the figure) may fail for the following reasons:Misalignment of insulators may cause an unequaldistribution of forces.A loose fit of busbars inside the slots may causeexcessive vibrations on a fault and may lead toloosening of the fasteners and shearing of the wedgesand/or the edges and the fingers of the insulators.Even the insulator mounting section .Y - x' maybecome vulnerable to failure.3 When one or all of the busbars are shorter in width asshown in Figure 29.l(b) the upper insulator may failat the shaded parts through the wedges or the edges,as they will now encounter relatively higher cantileverforces.Conclusion1 Loose busbars within the slots give rise to vibrationshood Dnod bond(4 (b) (C)(a) Smaller thickness of Busbars 'a'It may cause vibrations within the insulator slots during a fault andmagnify forces acting on the insulators and fasteners(b) Unequal width of busbars ' b (b,, b2)The insulator may shear off at Section X - X or yoke y(e) Proper mountingFigure 29.1Mountings of insulators and busbars2and a humming noise due to magnetic inductance.This may lead to loosening of fasteners and bedetrimental to the performance of the busbar systemin the long run. To lessen the effect of this, thebusbars should be only marginally loose inside theslot for easy movement during expansion orcontraction. This requires accurate size of insulatorsand correct mounting and alignment as shown inFigure 29. I(c). In this case all the load-bearingmembers are equally involved in sharing the forceand make the system stable and strong.One may consider a factor of safety of SO-IOO%,depending upon the criticality of the installation inall the forces that may arise on an actual fault toensure a foolproof system.29.2 Making a jointThis requires special precautions for both alurniniuinand copper, as both metals are highly susceptible tooxidation and corrosion. Oxides of aluminium and copperare poor conductors of heat and electricity and must heavoided. particuhrly at joints rather than in the straightlengths, to ensure proper transmission of current fromone section of the bus to another. Also, aluminium issoft. Making a perfect joint to achieve a longer durabilityis therefore essential. It involves attaining the least contactresistance by ensuring proper contact pressure to eliminateany localized heat. A slightly faulty joint may yield tofaster erosion of the metal and relaxation of the contactpressure. Loose contact pressure will lead to high contactresistance and cause a high localized heat, which mayresult in ultimate failure of the joint. For instance. if theouter diameter, thickness or the hole of the washer is notcommensurate with the diameter of the hole in the busbarthen the washer may gradually sag into the hole, in normalservice, through pressure by the bolt. Gradually it mayloosen its grip at the joint, release the contact pressureand lead to failure of the joint.The contact resistance can be minimized by increasingthe pull of the fasteners. Increasing the area of overlapmay not reduce the contact resistance. unless the numberof fasteners is also increased. It is mandatory to maintaina certain minimum contact pressure per unit area of'thejoint overLap. An average contact pressure at around 40-5.5 kg/cm- IS considered adequate. For the purpose ofeasy application, it is expressed in terms of bolt torque.depending upon the area of overlap and the number offasteners, as specified in Table 29.1.The following are more precautions that are consideredmandatory for making a good joint:1 Before making the joint, clean the surface and applythe contact grease to avoid oxidation. as discussed inSection 13.6.1(iv).2 Make the joint immediately after the above process.3 Make the joint by using the correct size of bolts, nutsand washers. Refer to Table 29.1 for the recommendednumber and size of fasteners for different widths ofbus sections and Table 29.2 for the recommendedsize of washers for different sizes of bolts. See also

29/906 Industrial Power Engineering and Applications HandbookBolt headPlain washer ,-20 + 3-, ShankMin. 2 threadsto protrudeFigure 29.2(a)assemblyCorrect procedure for using a bolt and nutPlain washerBusbar-Figure 29.2(c) Use of manual torque wrench to tighten thefasteners (motorised spanners are employed for faster production)ElevationSide viewFigure 29.2(b) A typical mounting and fixing arrangement ofa conical insulator and a flat busbarFigures 29.2(a) and (b) for a correct fastening method.4 For jointing large bus sections it may be advisable touse pressure plates to avoid excessive local pressure.5 Use a torque wrench to tighten the fasteners to ensurecorrect surface-to-surface contact of the currentcarryingparts (Figure 29.2(c)). The recommendedvalues of bolt torque are given in Table 29.1. A pressurethat is too high may cause relaxation of the joint bycold flow and must be avoided.29.2.1 Straight-through jointsTo join two sections of a bus, fishplates are used asillustrated in Figure 29.5. Slotted holes are usuallyprovided in the fishplate to allow for fixing adjustments.They are not meant to absorb the thermal expansion ofthe busbars on load, for they are supposed to make arigid joint, hence there is no scope for surface movement.For typical sizes of slots, refer to Table 29.1, and forwashers, Table 29.3. Smaller sections and single busbarscan also be joined by simple overlapping as shown inFigure 29.6.iA7Figure 29.3 Busbar bolting for Table 29.1

Recommended practices buses and making bus joints 29/907Table 29.1Recommended busbar overlaps for different sizes and torques of fastenersBar Length of Bolt Dimensions Bolt size Hole Minimum Typical size ofwidth overlap arrangement (Figure 29.3) diameter recommended slots for fishplatesas indicated bolt torque or straightin Figure 29.3through jointsI 2 3 4 5 6 7 8A B Cmm mm mm mm mm rnm kgm" mm25.4 50b 1 12.5 12.5 - M6M8 6.619 1.512.5 -38.1 76b 1 19 19 - M10 11 3.5 11 x 1650.8 76b 1 25 25 - M12 14 5.5 14 x 1876.2 76 2 19 19 - M10 11 3.5 11 x 16101.6 102 2 27 27 - MI2 14 5.5 14 x 18152.4 152 3 32 29 48 MI2 14 5.5 14 x 18203.2 203 4 32 29 48 M12 14 5.5 14x 18aThese torque values will normally require high tensile fasteners.bOverlap for tee joints even up to the width of bar will be adequate. Such as, for a tee joint of 50.8 mm wide bar, with a 101.6 mm straightbar, 50.8 mm overlap will be adequate, refer to Figure 29.7.Table 29.2Recommended sizes of punched washers for hexagonal bolts and screws as in IS 2016 to make a good jointBolts sizedrnm (ma)Dmm (min.)Smm (min)Conical Belliville washersha (ma.) mmM6M8M 10M 126.69111412.51721241.61.62.02.52.02.63.23.95aBased on DIN 6796 for conical spring washers for bolthut assemblies.Notes1 The above are the sizes of washers when the hole in the busbar is circular. If the hole is in the shape of a slot, as recommended in Table29.1, column 8, to facilitate easy jointing of the fishplates or straight-through joints, then the bulk of the washer should span the slotas illustrated in Figure 29.4. A normal size of washer as noted above may lose its efficacy and sag into it (by setting), loosen its gripin the course of time, and lead to failure. In such cases it may be recommended to use either pressure plates or heavy washers. Thewashers may be chosen corresponding to the larger width of the slot, considering this as its hole size d, as noted in Table 29.3, to achievea good rigidity and stability of washers to maintain the required contact pressure over long periods of service.2 Where the contact pressure is of utmost importance, it is recommended to use conical washers (some call them Belliville washers) tocounteract the loosening of a bolt and nut assembly, caused by setting or indentation. The last column of Table 29.2 indicates the vitaldimension h for such washers as in DIN4796.Slot sizemmllx 1614x 18Recommended Sizes of normal washers as in Comect-overlapping of joint is an important parametersize of washer IS 2016 to make a good joint, as well as to allow no excessivefor slot dheat at the joints. Based on the recommendations of thed D S leading aluminium section manufacturers, the desiredmm mm mm rnm overlaps are shown in Table 29.1, and two such jointsare illustrated in Figures 29.4 and 29.6. Laboratory tests16 11 21 2.5and site experience have revealed that a larger overlap is18 14 24 3'15of no additional benefit. For larger sections also, only

29/908 Industrial Power Engineering and Applications HandbookBolts M-10 Bolts M-10Plain washerPlain washerMinimum2 threadsecond section\Gap to allow forany mismatchLkl9 -33 76 -1 5 -i- Slotted hole(11 x16)(All dimensions in rnm.)Figure 29.4 A typical arrangement for jointing two sections of a busbar through a fish plate4 x M12 boltsBolts M-10+- 200-*I Ij I-+i IPlain washer- 50 100(All dimensions in mrn) 25 25Figure 29.5 Another arrangement for jointing two sections ofa busbar through a fish plateFigure 29.6 Jointing of two single busbar sections

-PlainSlotsCu or AI flexible connector(made of foils or braids) ~Recommended practices buses and making bus joints 291909Spring washer ~~washerMain busM Plain washeriRivets to hold the ~~flexible wram 1 r- ~(11 x16)Tap-off links--45 -z= 90L One sectionIt Second section -of bus45 -of bus(All dimensions in mm)Figure 29.7 Tap-off connections from alarge section of busbar showing the overlapas equal to the width of the smaller sectiontap-off linksFigure 29.8A typical flexible expansion jointone row of fasteners, as illustrated in Figure 29.5, isconsidered adequate to provide a reasonably good joint,so long as the recommended contact pressure per unitarea. of40-55 kg/cm' is maintained, as indicated earlier.29.2.2 Tee jointsRefer to Figure 29.7 when making a tee joint with alarger section of bus to tap for the outgoing feeders in aPCC or MCC from the main bus. A smaller overlap up tothe width of the feeder bus section will be sufficient toprovide an adequate contact area and bolting surface.29.2.3 Expansion jointsDuring normal operation busbars undergo elongation asa result of heating. When the busbars are short, as in aPCC or MCC, and have free ends, no provision to accountfor expansion of busbars will be necessary. Expansion ofthe structure on which the busbars are mounted and thefree ends will absorb the small expansion. But for longerlengths and when the end of the bus is to be bolted at arigid end, as at a transformer, expansion joints must beprovided at suitable locations to absorb the linearexpansion of the busbars. For the normal grade ofaluminium in use, the linear temperature coefficient ofexpansion can be considered to be 0.000023 rnm/"C (Table30.1 ). A busbar 25 m long and operating at a temperatureof 85°C having a temperature rise of 40°C above anambient of 45°C will have an expansion of 25 x 1000 x40 x 0.000023 mm. i.e. 23 mm. In such cases, the busbarsmust have free longitudinal movement and must beprovided with suitable expansion joints at reasonableintervals, say, at every 75/10 in. Busbars supported onbolted clamps as shown in Figure 29.2(b) are notrecommended as they block the expansion of the busbars.which may deform the busbars and result in damage tothe insulators and the supports and cause a fault. Fingertypebusbar supports, as shown in Figure 13.3 1 must bepreferred to clamp type supports.The expansion joints may be of aluminium or thincopper sheets (foils 32 gauge and thinner) or even copperbraidedwires to allow easy flexibility on expansion. Figure29.8 illustrates one such flexible joint. The normalprocedure to make a flexible joint is to fold these sheetstogether and press clamps at the ends, as shown in Figure29.9, where it is to be bolted with the bus sections. It isriveted at convenient locations to hold the foils in position.To avoid oxidation at the contact area, when it is open toatmospheric conditions, it is recommended to brace themat the edges where they are to make the joint, as shownin Figures 29.8 and 29.9. Fusion welding, inert gas, metalor tungsten arc welding processes are recommended forthis purpose.29.2.4 Flexible jointsThis is a synonym for an expansion joint.0 A flexible-end joint connects a generator or transformerto a bus system or a bus system to a power panel.An expansion joint connects two straight, normallyaligned sections of the same run of busbars.A flexible joint may also have to connect two nonalignedsections of current-carrying conductors. whichmay also be different in configuration and size (Figure29.10). They may therefore be longer than an expansion

29/91 0 Industrial Power Engineering and Applications HandbookPalmAluminium foilsWelding requiredEither solid or of foils(when in foils, bracethem at the edges)joint to suit the site requirement and help to connectthe two ends.The purpose of a flexible joint is thus to make anelectrical connection and to absorb the busbar's expansionand vibrations of the generator or the transformer and toprevent transmission of these vibrations to the bus systemand mounting structure.Depending on the rating and length of the bus system,a flexible-end joint may be necessary at the panel endalso to absorb expansion of the busbars and also to assistin making the end connections when the configurationsof the bus sections in the equipment are different or havea mismatch, although there are no vibrations at the panelend.It is normal practice to provide a flexible copper jointat the generator or the transformer end as the terminalsare also of copper and usually have a smaller spacingbetween them, where termination of aluminium flexiblesmay present a problem (although the use of aluminiumflexible is not forbidden).Ferrule tube before compression(No bracing required now)Figure 29.9 Making a flexible connector29.2.5 Bimetallic jointsElectric current passing through a metal joint having amoisture content causes electrolysis of water vapour.Copper, being a galvanic metal, forms an electrolyticcircuit with other metals and decomposes the joint.Decomposition is corroding and erodes the aluminiummetal.To PCCBus ductenclosure ,Grounding busbaron both sides\FlexibleLWm+*t-2 SideTransformer terminal- -- .-iview Front viewFigure 29.10 Typical arrangement of a flexible connection between a bus duct and a transformer

Recommended practices buses and making bus joints 29/91 1For making joints between aluminium and copper. careshould be taken that both surfaces are properly cleanedand dried and applied with a thin layer of grease, beforejointing, to eliminate electrolysis between the two metalsin the presence of moisture. It is recommended that wcha surface (particularly of copper) be tin or sliver platedto acoid electrolysis, which may take place with thepassage of time. Such a situation is predominant whenmaking aluminium connections to the main switchgeardevices and components such as breakers. switches, fuses.contactor%. relays and all other current-carrying andswitching devices. The connecting terminals of all thesecornponents are invariably of copper or bronze alloy. Asstandard practice. these terminals, are either silver platedor tin plated to facilitate a direct jointing or connectionwith aluminium links. However, use of grease at everyjoint and precautions, to eliminate the presence of moistureat the joints is mandatory to ensure a good joint.Some application engineers may, however, prefer abimetallic joint (e.g. a Cupal joint) for jointing betweencopper and aluminium. A bimetallic (Cupal) joint hascopper foil on one side and aluminium on the other. Thebasic purpose of such a joint is to climinate electrolysisduring normal operation. It becomes superfluous, whenproper care is taken in making the joint as noted above.It is a misconception that such a joint can deal withdifferential expansions of the two metals.Checking a jointIt is important to check the fitness ofa bus joint made ina factory or at site. Thi\ can be done with the aid of a d.c.millivolt drop or measurement of the joint resistance(mil) test. Such measurements arc taken on a number ofsimilar joints and the results tabulated and compared.Values in the same rangc may be considered good joints,while those with wide variations will be indicative of apoor joint. Such joints may then be investigated andimproved.29.2.6 Silver plating of joints*It I\ wmetinies preterred to sll\er plate aluminiuinjointsto eliminate any possibility of contact oxidation and toensure an almost uniform current distribution throughthe contact area without an excessive heating. Such apractice may, however. be of little advantage to mallerratings in view of cost. It will be worth while only forhigher currents. say, 2500A and above and also for higheroperating temperatures. It is seen that the oxidation ofaluminium or copper starts at about 85-90°C. It is forthis reason that the operating temperature ofa bus systemis limited in this region. as discussed already (Table 28.3).At higher ratings it is therefore recommended either toweld the edges ofthe joints (straight-through or flexibles)to seal the openings and prel'ent any oxidation duringoperation, or to silver plate the joints. Silver oxide is agood conductor of heat and electricity. With the use ofsilver joints or welded joints, a higher operatingtemperature of the busbar\ and the joints is permissibleup to 105°C (Table 2x2) as against 70°C as in Table28.2 (IEEE-C-37-20) or 85-90°C as in Tablc 14.5 (IEC-60439-2) in ordinary joints. With the use of silver platedjoints therefore, the rating of a bus system niay beimproved and the use of metal optimired. See also Section2X.S.l.29.3 Bending of busbarsBending a busbar also requires utmost care. Sniallcrsections niay not matter as much as larger and thickersec ti on s . The me tal s (part icu I ar 1 y a 1 u in i n i u m ) . be i ngbrittle. niay show up cracks. particularly on the outersurfaces. when bent, as a result of excessive tensile forceat this surface. The cracks may reduce the current-carryingcapacity of the busbar at this section. bedes renderingit mechanically weak, to withstand the electrodynamicforces on a fault. Sharp bends are therefore not recommended.When it becomes necessary to haw sharp bendsto meet locational requirements, it is recommended thatthe particular area of the bus section be heated first. tomake the metal somewhat soft and then to bend it gentlywhile the metal is still hot. A more appropriate methodwill be to use a hydraulic bending machine which canexert pressure evenly and more gently to prevent crack$.Siltcr and copper do ti01 make ;I chemical bonding with aluininiunithrough the electrolytic proces\. Alurninium thcreforc cannot besilbcr coated dii-ectly. Thc rccoinincnded procedure is to hr\t applya %n ctrating ( 1 -2 micron\). then a coating ofCu (1-7 microns) andthen ii coating of As (13-15 iiiicrrrns).

29/912 Industrial Power Engineering and Applications HandbookRelevant StandardsIEC Title IS BS IS0Torsional test and minimum torques for 1367-20/1996 BS EN 20898- 898-7/1992bolts and screws with nominal diameters 1 mm to IO mm 7/1995Specification for spring washers forgeneral engineering and automobilepurposes. Metric seriesPlain washers 201 6/1996Relevant German Standards3063/1994 BS 4464/1998 -DIN6796/1987For conical spring washers for bolthut assembliesNotes1 In the tables of Relevant Standards in this book while the latest editions of the Standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of standards is a continuous process bydifferent Standards organizations. It is therefore advisable that for more authentic references. readers should consult the relevantorganizations for the latest version of a Standard.23Some of the BS or IS Standards mentioned against IEC may not be identicalThe year noted against each Standard may also refer to the year of its last amendment and not necessarily the year of publication.

Properties andRatings of CurrentcarryingConductors

Properties and ratings of current-carrying conductors 30/91530.1 Properties and current ratingsfor aluminium and copperconductorsIn Table 30.1 we provide the general properties ofaluminium and copper conductors. The table also makesa general comparison between the two widely used metalsfor the purpose of carrying current.30.1.1 Important definitions of propertiesof a metalFor ease of application of the above table we give belowimportant definitions of the mechanical and electricalproperties of a metal.30.1.2 Physical and mechanical properties1 Specific heat (This is a physical property)The specific heat of a substance is the ratio of theheat required to raise the temperature of a certainweight by 1 "C to that required to raise the temperatureof the same weight of water by 1°C.2 StressesThis is the force per unit area expressed in kgf/mm'and is represented in a number of ways, dependingupon the type of force applied, e.g.Tensile stress: the force that will stretch or lengthenthe material and act at right angles to the areasubjected to such a force.Ultimate tensile strength: the maximum stress valueas obtained on a stress-strain curve (Figure 30.1).Compressive stress: the force that will compressor shorten the material and act at right angles to thearea subjected to such a force.Shearing stress: the force that will shear the materialand act in the plane of the area and at right anglesto the tensile or compressive stress.Modulus of elasticity (E): the ratio of the unitstress to the unit strain within the proportional limitsof a material in tension or compression. Refer toFigure 30.1.- Proportional limit: the point on the stress-straincurve at which will commence the deviation inthe stress-strain relationship from a straight lineto a parabolic curve (Figure 30.1).Elastic limit: the maximum stress a test specimenmay be subjected to and which may return to itsoriginal length when the stress is released.- Yield point: a point on the stress-strain curvethat defines the mechanical strength of a materialunder different stress conditions at which a suddenincrease in strain occurs without a correspondingincrease in the stress (Figure 30.1).- Yield strength or tensile proof stress: themaximum stress that can be applied withoutpermanent deformation of the test specimen. Forthe materials that have an elastic limit (somematerials may not have an elastic region) thismay be expressed as the value of the stress onMI Yield point /Elastic limitProportional limit-Strain* Tensile proof strength = 0.6 to 0.8Figure 30.1 Stress-strain curveUltimate tensilestrengththe stress-strain curve corresponding to a definiteamount of permanent set (elongation) of, say,0.1% or 0.2% of the test specimen.30.1.3 Electrical propertiesResistivity of metal of a current-carryingconductorA metal being used for the purpose of current carryingmust be checked for its conductivity. This is proportionalto its current-carrying capacity. This will ascertain thecorrectness of size and grade of the metal chosen for aparticular duty. It is necessary to avoid overheating ofthe conductor during continuous operation beyond thelimits in Table 28.2. The electrical conductivity of a metalis reciprocal to its resistivity. The resistivity may beexpressed in terms of the following units:Volume resistivity or specific resistance: this is theresistance of a conductor of unit length and unit crosssectionalarea, i.e.R.mm'and lpuR.cm = 10' ~mMass resistivity : this is the resistance of a conductorof unit length and unit mass;i.e. R . gm/m'which is also equal to the volume resistivity multipliedby the density:i.e. (R m) x (gm/rn3) = R . gm/m'Length resistivity: this is the resistance of a conductorper unit length, i.e. R/m.ConductivityTherefore, the electrical conductivity with reference tosay, volume conductivity, can be expressed by

30/916 Industrial Power Engineering and Applications HandbookmR.m2or - etc.i2.mThe resistivity and conductivity of standard annealedcopper and a few recommended aluminium grades beingused widely for electrical applications are given in Table30.1. Their corresponding current-carrying capacities inper cent, with respect to a standard reference (say, 100%IACS) are also provided in the table.30.1.4 Measuring the conductivityFor this purpose, a simple conductivity meter based onthe principle of eddy current may be used for a directreading of conductivity. The meter operates on the basisof relative variance, in the impedance of the test piececompared to the reference standard piece of aluminiumor copper having a conductivity of 100% or 31.9m/Rmm2 for aluminium and 58.0 m/Slmm2 for IACS(International Annealed Copper Standard) in terms ofconductivity unit. The test probes, that sense the impedanceof the test piece, induce an eddy current in the test piece ata fixed frequency. The magnitude of this current is directlyproportional to the conductivity of the metal. This eddycurrent develops an electromagnetic field around the testpiece and varies the impedance of the test probe (skineffect). The conductivity is thus determined by measuringthe corresponding change in the impedance of the probe.Figure 30.2 shows a simple and portable conductivitymeter.30.2 Current-carrying capacity ofcopper and aluminiumconductorsEarlier practice was to use copper in most applicationsFigure 30.2 Conductivity meter (Courtesy: Technofour)in view of its rigidity and high conductivity. With theeasier availability of aluminium and being more viableeconomically, aluminium i, now preferred whereverpossible. It is employed particularly where the metal hasto simply carry power such as for the transmission anddistribution of power at any voltage and as the maincurrent-carrying conductor in power distribution or controlequipment, such as a bus system or a switchgear assembly.Similarly, it is also used to feed high currents to aninduction or a smelting furnace, electroplating plant or arectifier plant. For main current-carrying components,however, as required for switching or interrupting devices(breakers, switches, fuses, contactors and relays) copperand copper alloys are preferred. The alloys are compactin size and are a much harder metal, suitable for makingand breaking contacts frequently and yet retaining theirshape and size over long years of operation. Copper isalso used for low ratings, up to 100 A or so, required forthe internal wiring of power and control circuits in aswitchgear or a controlgear assembly, where the wireshave to bend frequently. Aluminium, being brittle, isunsuitable for such applications. The use of copper isalso recommended for areas that are more humid andchemically aggressive which may corrode aluminiumquickly. As aluminium is highly oxidizing and a verysusceptible metal to such environments it may loosen atthe joints. Typical locations are mines, ships, textile millsand chemical and petrochemical processing units. Butfor such applications also, the latest practice is to instalelectrical equipment and switchgears in separate rooms,away from the affected areas, thus making it possible touse aluminium. In the following text more emphasis islaid on the use of aluminium, as it is the preferred currentcarryingmetal;Below we give the recommended sizes and ratings:Copper wiredcables refer to Table 13.15 for currentratings up to 100 A, as recommended for the internalwiring of a power switchgear or a controlgear assembly.Copper solid conductors: Tables 30.2 and 30.3 forgeneral engineering purposes.Aluminium solid conductors: Tables 30.4, 30.5, 30.7,30.8 and 30.9 for general engineering purposes.The following factors must be taken into account whiledeciding on the most appropriate and economical sectionsof the metal conductors for thc rcquired current rating;For the same thickness, a smaller cross-section willhave a relatively higher heat-dissipating area comparedto a larger cross-section. The latter therefore will have ahigher deration compared to a smaller cross-section onaccount of poorer heat dissipation. This can be illustratedas follows.Consider a 25.4 x 6.35 mm conductor with a crosssectionalarea of 25.4 x 6.35 mm2 and a surface area of2 (25.4 + 6.35) x I (1 being the length, in mm) = 63.51mm2.A conductor with twice the width (i.e. 50.8 x 6.35mm) will have a cross-sectional area of 50.8 x 6.35 mm2,and a surface area of 2 (50.8 + 6.35) x 1 = 114.31 mm2.Thus the larger section having twice the cross-sectionalarea in the same thickness will have 114.3/(2 x 63.5) or

~~ -p~ ~~~~~~~~~~~~ ~~~~~pil~~~~~~~ 11/19~~Notp~~~~ ~Table 30.1 Selected properties (average) of copper and aluminium at 20°C~ ~~~~Relevant standards~ ~--Standard grade\(a) As in BS(b) A\ in BS 2898 (IS0 209-1)(c) Equivalent lndal grade\Physical properties(a) Chemicd composition-~~ ~ -~Copper (Cu)Magnaium (Mg)Silicon (Si)Iron (Fe)Cr, T,. Zn and MnAluminiumStandard 1 opper(IACS)'7 IbC 60028% IiCommrrt rcrl purit\ tilurntnium (for elritrri (11 use)1 2 1' t- IEC 60105I150 209- I .2,-, E1E'- MhHEY ~ WPhE ~i 13506063 A1 CIS-M~50s - WPbCu. Si. Fe ~ inore than 0 5% 0 10 6-0 9% 99 9% pure Mg, Cr, Sn, 7n and Mn = Nil 0 3-0 60 15-0 15% 06I91E'6101AD 50s - WP"0.050.4-0.90.3-0.70.40. I--fSpecific heatgm.cal/"CDensityigdcm')Melting pointi"C)Mechanical propertiesp~~ --~Ultimate tensile strengthhgflmm'22 to 266 517 0 IS 5/23 5 20 5/25Ultimate \hearing strength~ ~-Modulus ot elasticity bkgflmin?kgflmm'0 2% tenvle-proot \trcngth kgflmni'Electrical propertiesVolume resistivity or specific rmihtance (p)(a) in p. (2. cm.. iiim(h)'in ~mI12,00060-804 of thc ,ten\ile wengthI55 I6 5 15~-I7241 2 8731 7241 x IO ' l 2871 x lo-'- ~ ~~7000 6700 67005 16 5/223.113.6 3.1333.1 x IO 'l3.6 x IO ' 3.133 x IO

~ -Parameter.5Standard copper(IACS)aCommercial purity aluminium (for electricul use)Volume conductivityConductivitym/Gmm2% IACS*5810034.8061/60~~ ~32.26127.8 3 1.9050148 55Temperature coefficient of electrical resistanceper "Cu.20 (applicable over a working range of100-200°c)a10 at a particular conductivity= at 100% conductivity x actual conductivityof the metal.Coefficient of linear expansion (thermal), (md'C)(applicable over a working range of 20-200°C)Mass resistivity = volume resistivity x density (~X2.gm/crn2)~-3.93 x 10-11.73 x15.3284.03 X 10-'/3.96 X lo-' 3.3 x 10-'/3.168 x 3.63 x IO-'~~2.3 x 2.3 10-5 2.3 x7.786 83719.72 8.49NotesWhen using the above metals for the purpose of current carrying, their mechanical suitability must be checked with the data provided above to withstand, without permanent deformation, theelectrodynamic forces that may develop during a short-circuit condition (Section 28.4.2).For important definitions, refer to Section 30.1.1.These values are based on the mean values of a number of tests carried out on specimens of standard copper and aluminium conductors.a IACS - International Annealed Copper StandardSuffix M (now F) or WP (now T6) represent the type of tempering.EIE, HE9 and E91E are old designations. They have now been replaced by 1350, 6063 A and 6101 A respectively.

Properties and ratings of current-carrying conductors 30/919Table 30.2 Current ratings for single rectangular copper sectionsConductor size(mm)2.5 x 12.516.020.025.031.540.050.063.04 x 16.020.025.031.540.050.063.080.0100.06.3 x 25.031.540.050.063.080.0100.0125.0160.010 x 50.063.080.0100.0125.0160.0200.0250.016 x 100.0125.0160.0200.0250.0315.0SourceReview._______~Area of crosssection(mm2)3 1.2540.050.062.578.75100.0125.0157.56480100125160200252320400157.50198.45252.0315.0396.9504.0630.0787.51008.050063080010001250160020002500160020002560320040005040Approximate ratingAmp (a.c.)15919523528734742651663025430536744554266080290011854735696938321010122014651755214510601260152518002150262031403710222026403180376045005370The Copper Development Association and the ElectricalNotes1 The ratings are based on a 50°C rise over 35°C ambienttemperature in still but unconfined air.2 AC ratings are based on spacings at which the proximity effectis considered almost negligible (2 300 mm Section 28.8).3 These are the basic maximum ratings, that a current-carryingconductor can carry under ideal operating conditions. The ratingis influenced by the service conditions and other designconsiderations, as discussed in Section 28.5. Apply suitablederating factors to arrive at the actual current ratings of theseconductors under actual operating conditions.4 Ratings may be improved by approximately 20% if the busbarsare painted black with a non-metallic matt finish paint. This isbecause heat dissipation through a surface depends upontemperature, type of surface and colour. A rough surface willdissipate heat more readily than a smooth surface and a blackbody more quickly than a normal surface. Also refer to Section31.4.4 and Table 31.1.5 The above ratings are for single bars. When multiple bars areused, apply the multiplying factors, as recommended in Table30.3. These factors will account for the restricted heat dissipationand additional skin effect due to the larger number of bars.Table 30.3 Multiplying factors for copper sectionsTotal area of cross-section Multiplying factors(mm2) 2 bars 3 bars 4 bars500 1.78 2.45 3.131000.72 2.36 3.001500.65 2.24 2.842000.60 2.16 2.702500.55 2. IO 2.603000.52 2.02 2.523500 1.48 1.98 2.484000 1.44 1.96 2.45Source The Copper Development Association and the ElectricalReview.NoteThe space between the bars is considered to be equal to the thicknessof the bars.only 90% surface area compared to the smaller section,and consequently less heat dissipation in the same ratioand will require a higher derating.Corollary1 Thinner sections will have a relatively higher surfacearea to dissipate heat compared to thicker sections.The thinner the section, the better will be metalutilization and vice versa.2 More bars will reduce the heat dissipation further andwill require yet higher deratings.3 Skin effect - the same theory is usually true for theskin effect. The thinner the surface, the smaller willbe the nucleus resulting in a higher concentration ofcurrent at the surface and better utilization of metal.We can derive the same inference from Tables 30.2,30.4 and 30.5, specifying current ratings for differentcross-sections. The current-carrying capacity varies withthe cross-section not in a linear but in an inconsistentway depending upon the cross-section and the numberof conductors used in parallel. It is not possible to defineaccurately the current rating of a conductor through amathematical expression. This can be established onlyby laboratory tests.Mechanical and electrical data for important rectangular,circular and channel sections are also provided in Tables30.7, 30.8 and 30.9 respectively for reference. For moredetails contact the manufacturer.

~~ ~~ ~~~ ~~ ______.~ ~~~~ .~ ~~________~~~~~30/920 Industrial Power Engineering and Applications HandbookTable 30.4 Current ratings for rectangular aluminium sections, grade E91-E (6101 A)SizemmI bur 2 burs 3 burs 4 bursD. C. 50 Hz D. C. 50 Hz D. C. 50 Hz D. C. 50 HzA. C. A. C. A.C. A.C.Approximate ratings (A)25.4 x 6.35 355 355 710 705 980 970 1120 110038.1 x 6.35 520 520 1030 I020 1380 1350 1585 153550.8 x 6.35 670 670 1315 1290 1765 1705 2050 194063.5 x 6.35 820 810 1550 1510 2100 2000 2430 226076.2 x 6.3.5 970 960 1805 I740 2440 2310 2860 2620101.6 x 6.35 1260 1235 2260 2140 3060 2800 3640 3200127.0 x 6.35 1545 1505 2700 2510 3660 3240 4410 3700152.4 x 6.35 1840 1780 3130 2860 4290 3680 5250 4240~. - ~~50.8 x 9.53 840 830 I560 1500 2090 1970 2460 226076.2 x 9.53 1210 1180 2180 2050 2940 2660 35 10 3030101.6 x 9.53 1550 1495 2710 2480 3660 3150 4400 3560127.0 x 9.53 1940 I860 3290 2930 4450 3660 5400 4200152.4 x 9.53 2260 2120 3770 3340 5 I40 4080 6300 4680203.2 x 9.53 2940 2750 4800 4150 6500 4900 8060 574076.2 x 12.7 I405 1355 2450 2240 3290 2830 4000 3240101.6 x 12.7 1830 1740 3100 2720 4170 3360 5 100 3900127.0 x 12.7 2230 2080 3720 3120 5040 3900 6170 4550152.4 x 12.7 2620 2420 4300 3500 5850 4400 7200 5100203.2 x 12.7 3380 3060 5450 4450 7420 5300 91 10 6150254.0 x 12.7 4080 3640 6500 5000 8860 6000 10900 6850SourceIndalco- ~~~~~~._________________Notes1 The ratings are indicative and based on a 50°C rise over 35°C ambient temperature in still hut unconfined air.2 For a multiple-bar arrangement, the space between the bars is considered to be equal to the thickness of the bar.3 A.C. ratings are based on spacings, at which the proximity effect is considered almost negligible ( 2 300 mm, Section 28.8).4 These are the basic maximum ratings that a current-carrying conductor can carry under ideal operating conditions. They are influencedby the service conditions and other design considerations, as discussed in Section 28.5. Apply suitable derating factors to arrive at theactual current ratings of these conductors under actual operating conditions.5 Ratings may be improved approximately by 20% if the busbars are painted black with a non-metallic matt finish paint. This is becausethe heat dissipation through a surface depends upon its temperature, type of surface and colour. A rough surface dissipates heat morereadily than a smooth surface and a black body more quickly than a normal surface. See also Section 31.4.4 and Table 31.1.6 Other gradea as in BS 1474 and BS 2898, for electrical purposes, and as produced by the leading manufacturers, are provided in Table30.6.7 To obtain the current rating for any other grade of busbar, multiply the above figures by the appropriate factor defined in Table 30.6.~ ~ . ~ . _ _ _ _.~ ~Table 30.5Current ratings for rectangular aluminium sections, Grade EIE-M (1350)Conductor Cro.wsectconu1 I bur 2 barr 3 bars 4 bar7 5 bars 6 barsri;eurea(mm (rnm?) Approximate ratrngs (A)25 x 12 30 1 I8 210 285 360 425 48016 40 151 275 395 490 580 65520 50 183 320 450 575 675 7702.5 62 5 223 390 540 685 800 91030 75 263 480 660 840 990 1 I1540 100 342 610 860 I080 I260 14254x 12 48 156 290 420 536 620 70016 64 198 340 470 600 710 81520 80 238 410 570 720 8.50 95525 100 290 530 755 950 1110 125030 120 339 600 845 1060 I245 140040 I60 434 750 10.50 1320 1550 175050 200 512 905 1260 1575 1825 20356x 12 72 200 350 480 610 720 82516 96 252 450 640 805 960 1075

Properties and ratings of current-carrying conductors 30/921Conductor Cross-sectional I bar 2 bars 3 bars 4 bars 5 bars 6 barssizearea(mm) (mm2) Approximate ratings (A)6 x 20 120 301 550 790 lo00 1170 132025 150 364 640 900 1120 1315 148530 180 424 730 1025 1290 1520 172040 240 545 935 1310 1630 1900 213050 300 660 1130 1580 1950 2255 250560 360 782 1350 1870 2200 2630 288580 480 995 1700 2310 2745 3070 3330100 600 1215 2090 2770 3190 3490 3745120 720 1415 2415 3180 3640 3985 4270160 960 1830 3100 4050 4600 5025 534010 x 40 400 720 1230 1710 2110 2425 267050 500 870 1500 2060 2505 2850 310060 600 1015 1750 2350 2795 3120 338080 800 1250 2215 2940 3355 3675 3950100 1000 1565 2650 3465 3940 4315 4635120 1200 1810 3050 4010 4560 4980 5290160 1600 2310 3940 5170 5870 6300 6620200 2000 2795 4750 6160 - - -250 2500 3365 5720 - - - -16 x 100 1600 1950 3260 4250 4890 5340 5660120 1920 2255 3850 5030 5700 6130 6450160 2560 2840 4830 6260 - - -200 3200 3395 5780 - - - -250 4000 4095 6500 - - - -Source The Aluminium FederationNotes1 Current ratings for E-91E (6101 A) bars are about 3% lower than EIE-M (1350); refer to Table 30.6.2 The ratings are indicative and based on a 50°C rise over 35°C ambient temperature in still but unconfined air.3 For a multiple-bar arrangement the space between the bars is considered to be equal to the thickness of the bar.4 AC ratings are based on spacings at which the proximity effect is considered almost negligible (2 300 mm, Section 28.8).5 These are the basic maximum ratings that a current-canying conductor can carry under ideal operating conditions. They are influencedby the service conditions and other design considerations, as discussed in Section 28.5. Apply suitable derating factors to arrive at theactual current ratings of these conductors under actual operating conditions.6 Ratings may be improved approximately by 20% if the busbars are painted black with a non-metallic matt finish paint. This is because,the heat dissipation through a surface depends upon its temperature, type of surface and colour. A rough surface dissipates heat morereadily than a smooth surface and a black body more quickly than a normal body. See also Section 3 1.4.4 and Table 3 1.1.7 Other grades as BS 1474 and BS 2898, for electrical purposes, and as produced by the leading manufacturers, are provided in Table 30.6.8 To obtain the current rating for any other grade of busbar, multiply the above figures by the appropriate factor defined in Table 30.6.Table 30.6 Grades of aluminium alloys for electrical purposesGrades as the oldGrades as in BS 2898 Equivalent grades of Indian Multiplying factor to ratingsdesignation (IS0 209-1) Aluminium (Indal) alloys of Table 30.4EIE-MEIC-ME91EHE-9-WP1350-6101 A6063 ACIS-M2 S-MD 50 S-WP50 S-WP1.031.021 .oo0.94Notes1 Aluminium conductors for engineering application are produced in commercial grade quality, having an electrical conductivity varyingfrom 70% to 94% (approx.) as well as electrical grade quality having an electrical conductivity of 94% and higher, as noted above,varying slightly from manufacturer to manufacturer. Commercial grade, not below HE-9-WP (6063A), can be used for currentcarrying, say, up to 1000 A, although electrical grade is preferred. For higher currents, however, electrical grade only should bepreferred.2 When the short-circuit forces are likely to be high, say, 1500 kgf or more, per metre run, such as on a main power circuit the electrolyticgrade aluminium of type EIE-M (1350) may not be recommended. It is a soft metal mechanically, as noted in Table 30.1, which willrequire busbar supports at very close spacing, defeating the economics of the selection. The grade type E-91E (6101 A), which has abetter mechanical strength, would be a better choice for all types of power applications. The selection of busbars, shape and grade, isthus governed by mechanical considerations and economics, rather than the purity of the alloy alone.

~~ ~~~ ~ ~~~~ ~~~~~~~~~~~~~ ~~~~~ ~~~ ~~ ~~~ ~- ~~Table 30.7 lndal cis - M and D 50 S - WP rectangular busbars equivalent to EIE - M (1350) and E - 91E (6101 A) as in BS 1474 and BS 2898 (mechanical and electrical data)SizeCrosssecrionulureaWeigh/CISM d.c. D 50 SWP d .~. Reuctunce X,, Moment of inertia Section modulusresis lance resistarice(niux.)(max.)ut 20°C at 20°Cx i xRadius of gjirutionYrnmmni7kghiix-.r(crn4)(cm')Iyy Zl :r(cm')Zy-y25.4 x 6.35 161.2938.1 x 6.35 24 1.3050.8 x 6.35 322.5863.5 x 6.35 403.2376.2 x 6.35 483.870.4350.6520.871I .OXY1.306178.15 194.35I 18.76 129.5689.07 97.187 I .26 77.8259.38 64.80236.22215.22199.14186.35175.850.8742.9146.95 113.56923.4340.0 I2 0.6880.033 1.5400. I25 2.7310.125 4.2610.166 6.1450.1310.2620.3930.4260.5080.734 0. I831.102 0.1831.468 0. I831.836 0.1832.202 0. I83101.6 x 6.35 645.16127.0 x 6.35 806.45152.4 x 6.35 967.7450.8 x 9.53 484.1276.2 x 9.53 726.191.7422. I772.6131.3071.96144.55 48.5935.63 38.8829.69 32.3859.38 64.8039.60 43.18159.45146.00134.5 I195.86174.2155.484108.387187.30410.40635.1300.208 10.9300.291 17.0750.333 24.5810.374 4.0970.54 1 9.2260.6880.8521.0320.7871.1472.936 0.1833.670 0.1834.404 0. I831.466 0.2742.202 0.274101.6 x 9.53 968.25127.0 x 9.53 1210.31152.4 x 9.53 1452.37203.2 x 9.53 1936.5076.2 x 12.70 967.74101.6 x 12.70 1290.32127.0 x 12.70 I6 12.90152.4 x 12.70 1935.48203.2 x 12.70 2580.64254.0 x 12.70 3225.802.6143.2683.92 I5.2282.6133.8484.3555.2266.9688.71029.69 32.3823.75 26.2 I19.82 21.5914.83 16.2129.69 32.3822.28 24.2817 81 19.4214.83 16.2111.12 12.148.89 9.7 I157.48144.36134.18117.45170.93155.84143.04132.54115.8196.7883 246I62 580280 956665 97046 826-I1 1 009216 773174 608887 8221734 4360.749 16.3870.9 I6 25.6131.082 36.871I .457 65.548I .290 12.2901748 21 8442 164 34 1342 622 49 1613 455 87 3764 329 I16 s701.5401.9172.31 I3.0642.0482.7373.4094.0975.4416.8 172.936 0.2743.670 0.2744.404 0.2745.872 0.2742.202 0.3682.936 0.3683.670 0.3684.404 0.3685.872 0.3687.341 0.368Source:Indalco

~~~~~ ~~~~~ ~~~~ ~~~~~~~~~~~~~~~@Table 30.8lndal D50S WP tubular busbars (mechanical and electrical data and current rating)InJideWullthickne.s.sAnwNominal Moment Section Radiw ofweigh/ of inertirr modulus gyrotiorzG M LlOSin.tnmninimmmm-c 111(standard iron pipe sizes)1 33.40I ‘I4 42.161 (II 48.262 60.332‘12 73.0326 6435 0440 9052 5162 713.38 3193.56 4323.68 5153.91 6935.16 I 1000.8611.1661.3921.8712.9703.6348. 10412.89927.70963.6830.8573.8445.3459.18617.45 I1.0681.3711.5811.9992.406I56019912 2912 8781 47598 172 760 745 128 4186.7171.3162.7148.3136.2675 815845 1010960 11601245 14401720 19503 88.903 I/? 101.604 114.304‘12 127.005 141.3077 9290 12102 26114 46128 205.49 14395.74 17296.02 20486.27 23796.55 27243.8844.6675.5296.4237.490125.606199.279301.039434.68363 1.00728.25839.22752.67568.45489.3262.9573.3963.8354.2754.7704.2674.8775.5146.1606.85321 817 415 3I3 211 3123.7115.2107.6100.193.82130 23502510 27502800 30503160 34203550 3810(extra-heavy iron pipe sizes)1 33.40I ‘I4 42.161 ‘I? 48.262 60.332112 73.0324 3032 4618 1049 2559 014.55 4124.85 5685.08 6895.54 9547.01 14541.1131.5351.8612.5753.9264.39510.06416.28336.12580.1832.6324.7746.7481 1.9772 1.9541.0331.3301.5371.9472.3471.52 I1.95 12.2452.8353.41476 055 145 432 921 5188.3172.6164.0149.3137.1745 925975 I I70I125 13351465 I6801955 22303 88.903 Ill 101.604 114.30411: 127.005 141.3071 6685 4497 18I OX 96122 247.62 I9468.08 23748.56 28449.02 33430.53 39475.2546.4107.6789.02710.656162.09326 1.393199.998584.938860.35036.4665 1.4.5569.98992. I 15I2 1.7722.8853.3203.7524. I834.67 I4. I784.8015.43 I6.0456.74616 1I3 2I I 0948 0124.7116.1108.3100.794.52470 27002x50 31603260 35903700 40404 I90 4550So LII-CP: I nd al coNote Thcsc data are indicative and provided for typical standard \i/rs from a manufacturer. Busbar\ larger than the ahovc arc gcncrally not manufiictui-ed in tuhula~- sections hut 111 \cction\and configurations that are convenient by extrusion (Figure 31. 15). By welding such sections. nnc can form any desired hire of tubular or any other conductor shape (hexagonal or octagonal).Such largc wctions are required for isolated phase bu\ (IPB) system\, discussed in Chapter 31.

~~~~ -* 1'' I4 xYSi7e Web Flung,and widthflangethicknessmm mm mmb t BInnerflatsut$acemmC fFilletradiusmmTable 30.9 Channel busbar: mechanical and electrical data and current ratings (Grade E - 91 E)Section Weightareamm2kg/mMoment ofinertiaOne channelSectionmodulusRadius ofgyrationcmKx-xKy-.Distance D C Depthto Resisneutraltanceaxis x (mar)at 20°Cmm pWm mmIbTwo channelsWidthmmaReact- D.C. A.C.ance at current current305 mm rating rutingspacingut 50 Hzat 50 HipWm Amps Ampsxo76.2 6.35 38.10101.6 6.68 41.83127.0 8.00 48.00152.4 8.10 51.6645.08 7.6269.20 9.5292.00 9.50113.98 11.111013 2.7351187 3.2051695 4.5761993 5.38177.42 12.07171.07 17.07375.44 31.63629.34 42.0420.48 4.5933.76 5.7459.16 9.1882.59 10.982.84 1.123.78 1.194.70 1.375.61 1.4511.7 26.4113.2 18.4113.5 15.72101.6 106.4127.0 127.0152.4 152.4114.83 3380 3340106.96 4250 419092.52 5540 544080.38 6500 6340177.8 9.98 58.40203.2 11.84 65.00254.0 13.18 92.07Source: Indalco135.62 11.11154.12 12.70199.06 14.292794 7.5443734 10.0825515 14.89~-~ ~1179.18 74.092033.29 120.294924.43 400.41132.74 17.21200.09 25.07387.72 56.706.50 1.637.39 1.789.45 2.6415.2 11.2317.0 8.4024.1 5.68177.8 177.8203.2 203.2254.0 254.070.54 8200 780062.99 9850 915048.56 12800 11450

~~ ~~ ~-~~ ~~~-Relevant StandardsProperties and ratings of current-carrying conductors 3019251EC7-/t/rISIS06002Xil925lnternational \tdndard of reustance for copper~ 60 IO511 958 Recommendation for commercial purity aluminium busbarmaterialSpecification for wrought aluminium and aluminium alloys for 733/199 I BS 147411987 6362.engineering purposes: bars, extruded round tubes and sections 209-1. 2Specification for wrought aluminium and aluminium alloys for 5082/1998 BS 289811985 209-1. 2electrical purposes: bars. extruded round tube and sectionsAluminium and aluminium alloys. Extruded rodbar, tube andprofilesCopper tor electrical purpo\e$ Rod dnd bar- ~~~128511991 BS EN 7558 partslV(Jr?\I73In the tables of Relevant Standards in this book while the latest editions of the Standards are provided, it Is possible that revised editionsh,ive bccome available. With the advances of technology and/or its application. the updating of Standards is a continuous process bydifferent Standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganization\ for the latest version of a Standard.Some of the BS or IS standards mentioned against IEC may not be identical.The year noted against each standard may also refer to the year of its last amendment and not necessarily the year of publicationFurther reading4 Copper Development Association, Copper Bushar: Pub. No.22 47 .5 Dwight, H.B., Elecrrical Coils and Conducror.s, McGraw-Hill.I Alcoa, A/urnrniuni Bus Conductor Hmd Book. New York (1945).23British Aluminium Co. Ltd, A/uminiirm Busbars, Pub. No. L4.British Aluminium Co. Ltd. Allirnirziurn for hushurs. earthing(id/ifihtniri,q c.o,idrrcforv. Pub. No. M4.6 Indian Aluminium Co., Indal Aluminium Bushar (CollaboratorsAlcan, Canada).

An Isolated PhaseBus System311927

~ TeeAn isolated phase bus system 31/92931.1 An isolated phase bus (IPB)systemIn this construction the conductors of each phase arehoused in a separate non-magnetic metallic enclosure toisolate them completely from each other with the followingadvantages:I It eliminates phase-to-phase faults.2 It eliminates the proximity effect (extra forces andheating) by providing a magnetic shielding to the\upporting and metallic structures in the vicinity.3 It reduces the proximity effects between the maincurrent-carrying conductors of the adjacent phases toalmost zero due to magnetic shielding, on the onehand. and large centre spacing, on the other.4 It provides complete protection for operating personnelt'roin high touch or step voltages (for details on contactvoltages. see Section 22.9) across the enclosure andthe metallic structures caused by parasitic (electromagnetic)currents.5 The bu\ system is are easy to handle, bend and install.This is used for HT systems of very large ratings.Since it is a more expensive arrangement it is normallypreferred only for crirical installations such as a powergenerating station. where it has to carry large to verylarge currents due to high ratings of the generating units.In a thermal power plant it is used between the generatingunit (G) and the generator transformer (GT) and the unitauxiliary transformers (UATs) as well as sometimesbetween the UAT and the unit HT switchgear as illustratedin Figure 13.2 1 (redrawn in Figure 3 1.1 for more clarity).A typical layout of such a bus system in a thermal powerplant is illustrated in Figure 31.2(a) and (b). It maycomprise the following sections.31.1.1Main run and auxiliary feed lines1 From generator phase terminals to the lower voltageside of the GT. This section may partly be indoors,connecting the generator inside the turbine room andpartly outdoors, connecting the GT in the transformeryard.2 A generator neutral bus to form the generdtor starpoint.3 Tap-offs with tee joints from the main run to the twoUATs.4 Sometimes from UAT secondaries to the unit HTswitchgears.5 If the GT is made of three, single-phase transformers,an interconnecting IPB will be required to form thedelta as illustrated in Figure 3 I .?(b).31.1.2 Instrumentation and protectionconnectionsSince an IPB forms an important integral part ofa powergeneratingstation, with it are associated a number ofmetering and protective devices such as CTs, VTs andgenerator grounding system. Below we identifiy interconnectionsthat may be required to connect these CTsand VTs while installing an IPB system.IRemovable links for easy mounting and maintenanceof CTs for metering and protection. typically as follows.For each phase in the main lines between the yeneratorand the GT,1 CT for meteringTo switch yardTransformerGeneratortransformer.Tap-offconnection 1jointsTap-off- Three phase two winding transformer (UAT)Outdoorinstallation- Three phase two windingtransformer with ONload tap changer (GT)Unit bus ductCircuit breakerIndoorinstallationchamberJ H T breaker( 33 kV/ll kV/I 66kVH T breakerFigure 31.1 Application of an IPB

311930 Industrial Power Engineering and Applications Handbook1 CT for protection1 CT for AVR (automatic voltage regulator)1 CT for differential protectionFor each phase on the neutral side, 1 CT for differentialprotection, two sets of three CTs each for metering,in the tap-offs for the two UATs and two sets of threeCTs each for protection in the tap-offs for the twoUATs.2 Tap-offs with a neutral CT, from star point of thegenerator, to the neutral grounding transformer (NGT).3 VTs and surge protection tap-offs for each phase fromthe main run, for metering and surge protection.Any other connections, that may be necessary at siteor additional CTs for protection.31.1.3 Shorting of phase enclosuresFor continuous enclosure IPBs, the following ends ofthe enclosures are shorted to ensure continuous flow ofinduced currents:At the generator end, as near to the generator terminalsas possible.GT ends, as near to the transformer flanges as possible.UATends.Both sides of the neoprene or EPDM rubber or metallicexpansion bellows.Near the generator neutral terminal.Any other section, where it is felt that it is not properlybonded and the induced currents may not have acontinuous path.The enclosure, however, should be insulated at all suchlocations, where the IPB terminates with anotherequipment, such as at the generator, GT, UATs, VTs, andthe NGT to avoid longitudinal currents through suchequipment. These equipments are grounded separately.31.2 Constructional featuresAs discussed later, the enclosure of an IPB may carryinduced currents up to 95% of the current through themain conductors. Accordingly, the enclosure is designedto carry longitudinal parasitic currents up to 90-95% ofthe rated current of the main busbars. The cross-sectionalarea of the enclosure is therefore maintained almost equalto and even more than the main conductors to accountfor the dissipation of heat of the main conductors throughthe enclosure only, unless an additional forced coolingsystem is also adopted. The outdoors part of the enclosureexposed to atmospheric conditions is also subjected tosolar radiation. Provision must be made to dissipate thisadditional heat, from the enclosure.The main bus and its enclosure is normally in a tubularform, in view of the advantages of a tubular section, asdiscussed already (concentration of current in the annulus,$of turbo generatotof turbo generatorGeneratorIflange-+-Indoor OutdoorI I SealmaIII1.-4= TIIIIIIIII%%&NeutralgroundingtransformerGrating floorTurbine roomVT, & surgeprotectioncubicleBellows assemblywith seal offbushingFigure 31.2(a)Layout of 24 kV, 20 kA isolated phase bus duct from generator to generator transformer for NTPC

~~~~ -~-HV side400 kV (y)LV side (D )-R Y BAn isolated phase bus system 31/931Generator transformer (GT)(arrangement illustrates when3,l-9 transformers are used)Metering CTsBendsIPB bends andinter connectionsEnclosure shorting pointsMetering CTsIPB tap-offsUnit switchgearIPB tap-offsCTs forLine side differential -protection_.GeneratorAdCTs for diffe -CTs forAVR'1 "L\ c2a, , ,--b2.. . . . .. . . . . .. .. . Unit switchgearCubicle for VTs' andsurge protection!- Enclosure shorting points- -A little distance away tobe out of IPB proximityarea on a ground faultStator neutral groundingtransformer (NGT)Gen. voltage GFFLi3Neutral groundingtransformer with enclosure'It may be supplemented with a surge capacitor, to suppress the steepness of a t.r.v. and diminish its r.r.r.v, (Section 17.10.1)"TypicalFigure 31.2(b)enclosuresTypical layout of an IPB system in a thermal power station, illustrating enclosure end shortings for continuousSection 28.7). One more advantage of a tubular sectionis that it exerts equal forces at all points of the enclosureand relieves it and the conductor from any undue stresses.Octagonal and hexagonal sections are also used as theyalso have near-symmetry.Insulating medium: Dry air or SF, gas0 Insulators: The conductors are resilient mounted onpost insulators, to hold them in the centre (Figure 31.3)and dampen the forces, during normal operation or onfault. The conductors are held in position to have freemovement axially but they have almost no movementradially. The thermal effects of the conductors are takencare of by the expansion joints (Figure 31.4(a)) andthose of the enclosure by the bellows (Figure 3 1.4(b)and (c)).0 Naminal current ratings: Some typical current ratingsfor the stator voltages considered in Table 13.8 are10 000 A for a 250 MVA generator to 40 000 A for a1000 MVA generator. (See item 3 of the table for thenormal current ratings for different MVA ratings andstator voltages of generators.) The preferred ratingsmay follow series R-10 of IEC 60059 as in Section13.4.1(4).

31/932 Industrial Power Engineering and Applications HandbookEmclosureFigure 31.3 Arrangements of insulators to hold the busbarsTypes of enclosures: These may be of two types, Le.Non-continuous, andContinuous.31.2.1 Non-continuous or insulated enclosuresThese were used until the early 1960s and have sincebeen discontinued in view of the inherent advantages ofa continuous enclosure. In non-continuous enclosuresthe individual sections that are added together to obtainthe required length and configuration of the bus layoutare electrically insulated from each other. They are alsoinsulated from their mounting structures by rubber orfibre-glass or similar insulating pads, as illustrated inFigures 31.5(a) and (b). This is to prevent the longitudinalflow of current from one section of the bus system to theother as well as from one phase enclosure to the other.There is no external return path for the induced currents.But local induced currents do flow through each insulatedsection and may cause nominal step and touch voltages.Each section is grounded at one point (only) to its ownseparate ground bus which in turn is connected to thestation ground bus at one point only to make the inducedcurrent flow in one direction only. The ground bus is ofthe continuous type. Layout of a non-continuous IPBwith grounding arrangement is shown in Figure 3 1.6.This system of insulation and grounding minimizes thestep and touch voltages (Section 22.9) across each sectionof the enclosure. The induced voltage across each sectionis kept as low as possible, preferably below 2 V, whenoperating at the rated current. The ground bus whichmay be of copper or aluminium (only of non-magneticmaterial and not of GI) for each phase is continuouslyrunning and capable of carrying the momentary peakcurrent (Table 28.1) of the main bus system for twoseconds as in ANSI C-37/20C.Such an arrangement, although adequate in somerespects, has some disadvantages as noted below:(a) There are higher losses in the enclosure due to thehigher proximity effect as the induced current is notcontinuous throughout the enclosure.(b) It provides negative magnetic shielding to the outsidemetallic structures in the vicinity. Magnetic shieldingis an extremely important requirement to minimizethe eddy currents (= B2) and hysteresis losses(= B1'6) instructures and to reducethe electrodynamic forces developed between themand the main bus conductors.Stainless steel bandAI strip weFigure 31.4(a) An expansion jointFigure 31.4(b) Rubber or metallic expansion bellows forenclosure jointing and end terminations (Courtesy: Best &Crompton)

An isolated phase bus system 31/933EnclosureflexiblesshortingRubber orbellowsMetallicFigure 31.4(c)24 kV, 20 kA, IPB straight run and tap-offs (Courtesy: Controls and Switchgears)cml?ZG-OE-2& 7Dal DFigure 31.4(d) Terrnlnatlon of an isolated phase bus systemat a transformer (Courtesy: Best and Crompton)531.2.2 Continuous or bonded enclosuresThe system is called electrically continuous when theindividual sections (enclosures) of the IPB joined togetherto obtain the required length and configuration of thebus layout are also electrically bonded to each other.Each enclosure is then cross-connected with the enclosureof the other phases at the extreme ends of the installedbus duct. The bonding permits longitudinal flow of inducedcurrents through the length of the enclosure and returnthrough the enclosure of the other phases, as shown inFigure 31.8. This arrangement provides the requiredmagnetic shielding between the phases, as the inducedcurrents flowing through them adjust among themselvesand make it an almost balanced current system. Themagnitude of such current is almost equal to that of themain conductors while the direction is the opposite.These balanced enclosure currents also induce electricfields into nearby structures, RCC beams and columnsin the same way as the main conductors, and hence nullifymost of the space magnetic fields. These space fields(fields outside the enclosure) are otherwise responsiblefor causing eddy current and hysteresis losses in themetallic (magnetic) structures, RCC beams and columnsin the vicinity. The electrical bonding of enclosures thus

31/934 Industrial Power Engineering and Applications HandbookFibre glass insulating padFigure 31.5(a) Typical insulating practice to isolate enclosure of a discontinuous IPB system (Courtesy: Best & Crompton)Main bus conductorAI. shorting bandSupport insulatorinsulating padGalvanized steelsupport structureFigure 31.5(b) Cross-sectional view of an IPB with insulated(discontinuous) enclosure illustrating busbar-supportingarrangement (Courtesy: Best & Crompton)nullifies the proximity effect and helps to reduce theheating of such structures to a very great extent.The electrodynamic and electromagnetic forces betweenthe conductors and the structures are also reduced toonly 10-15% or even less. These two advantages are notavailable to this extent in a non-continuous enclosure.The induced current causing the magnetic field in thespace is now reduced to only,(31.1)Figures 31.7 and 31.8 illustrate the disposition of theconductor and the enclosure fields in the space and theirnullification, whereI, = conductor currentI, = enclosure currentI, and I, are almost 180" out of phase.therefore, (I,- I,)For better shielding it is essential that bonding is carriedout at the farthest possible locations where the configurationof the IPB changes such as at the bends andtap-offs or where it passes through a wall. Where bondingis not possible, the supports and the enclosure must beadequately reinforced to sustain electrodynamic forces,especially during a fault. Figure 3 1.2(b) illustrates theshorting and grounding locations of an IPB system.To limit the fault level, if it is likely to exceed thedesigned limits, or the breaking capacity of the interruptingdevice, or the associatedequipment and to limit the inducedcirculating currents in the enclosure during normaloperation, to contain enclosure losses in very large ratingsof bus systems, say, 25 OOOA and above (above 500 MW),then unsaturable type series reactors (for details refer toChapter 27) may be provided in the enclosure circuit, asillustrated in Figure 31.8 to limit such high currentsespecially during a fault. The reactors should not saturateunder fault conditions, as they are provided to supplementthe enclosure circuit impedance to limit the high currentsthrough the enclosure, especially during faults, and reducethe forces between the main conductors. For a better flowof circulating currents, it is also grounded only at onepoint. Accordingly, it is supported on non-conductingsupports to keep the ground path continuous and completelyisolated. The enclosure must also be insulated electricallyfrom the rest of the plant by rubber bellows.A continuous enclosure provides a high degree ofmagnetic shielding for metallic objects and structureslocated in the vicinity and generates only nominal eddycurrent or hysteresis losses. Magnetic shielding alsosignificantly reduces the electrodynamic stresses causedby short-circuits on the structures and between theenclosures of the other phases. For details on the thicknessand size of enclosure to provide magnetic shielding upto a required level, refer to the sample calculations inExample 3 1.1 and further reading at the end of this chapter.

~~~ InsulatedAn isolated phase bus system 31/935Aluminium or copperSplit covercontinuous ground bus 77Discontinuousbus enclosure -i-Generator endinterconnections To station ground bus - 'FFigure 31.6Grounding arrangement of a noncontinuous isolated phase bus (IPB) systemDirection ofmain currentDirection of/ enclosure current, Magnetic field withinthe enclosure and itsdirection according tocorkscrew rule31.3 Special features of an IPBsystemIt is almost mandatory for such a bus system to have thefollowing features to make it more reliable and to provideuninterrupted services over long years of strenuarsoperation:Field in the space Field in the spaceas caused by the as caused by themain current I,enclosure currents"Space field nullifying each otherFigure 31.7enclosuresGroundingNullifying of magnetic field in space in continuousThc system is grounded only at one point as discussedabove. It is also insulated from the ground as well as therest of the plant along its length by being supported oninsulated foot mountings. The minimum clearance ofthe bus from the ground will depend upon the voltage ofthe IPB system. The ground bus is not screened throughthe enclosure but is provided some distance away from itto remain unaffected by the fault currents. Otherwise itmay also develop large electric fields and be subjectedto excessive forces on a fault.1 The enclosure is constructed of non-magneticmaterial, generally aluminium, in view of its lowcost and weight as compared to copper. The nonmagneticmaterial eliminates hysteresis and eddycurrent losses in the enclosure. as a result of mutualinduction.2 To contain the proximity effect. in the metallicstructures existing in the vicinity, it is essential thatthe IPB enclosures be at least 300 mm from allstructures existing parallel and 1 SO mm existingacross the enclosures. A distance of 300 rnm issufficient to contain the proximity effect in view ofthe substantially reduced magnetic field in the space.3 The enclosure must be airtight and waterproof.4 Large generators are hydrogen (H,) cooled. H2 formsa highly explosive mixture with air and may leakthrough the generator terminals to the bus enclosure.Suitable sealing-off chambers must therefore beprovided between the generator and the IPB enclosurewith ventilation to allow gases to be vented to theatmosphere in the event of a leakage.S The termination of the conductor at the generator,transformer and the switchgear ends are made throughflexible connections (Figure 31.4(a)). Jointing ofthe enclosure sections is carried out with a rubberor metallic (mostly aluminium thin sheets) expansionbellows assembly, as shown in Figure 31.4(b). Besidesabsorbing the enclosure expansion, the bellowsassembly takes care of minor mismatches betweenthe flanges of the IPB and the generator or the

31/936 Industrial Power Engineering and Applications HandbookUnbonded portions may experienceexcessive forces.Shorting links (bonding) betweenenclosures. It helps to balance the flowof current through the enclosure.Saturable reactors to limit the enclosurecurrents and reduce the fault level andlosses caused by it.Direction of current in the enclosures.Enclosure induced currents can be upto 0.90 to 0.95 I,.Direction of current in the mainconductors.Field produced by the enclosurecurrents nuliifying the main field.Residual field in the space = 1&5% ofthe total field produced by C.Figure 31.8 A continuous IPB system with phase reactorstransformers and also prevents transmission ofgenerator and transformer vibrations through- theenclosure. The expansion bellows, when they aremetallic, are insulated with the housing as they aremade of thin sheets and are incapable of carryingexcessive enclosure currents. They are clamped tothe housing with the help of an insulating pad andmetallic bands as shown in Figure 31.4(b). Tomaintain the continuity of enclosure currents, thebellow assemblies are superimposed with enclosureshorting flefibles, sufficient to carry the full enclosurecurrents (see Figure 3 1.4(c)).However, they should remain insulated whenterminating with an equipment or a device such asat the ends of generators, GTs, UATs or VTs. It isessential to avoid IPB longitudinal currents throughthe terminal equipment. Now the bellows necessarilyshould be of rubber. Figure 31.4(d) shows a rubberbellows but in this small part of the bellows theconductor field will not be nullified and occupy thespace affecting the metallic structures, beams andequip-ment/devices in the vicinity. This needs to betaken into account at site and it should be ensuredthat the nearest structure, beam or equipment is atleast 600 mm away from the IPB enclosure.The terminal ends are provided with terminal palmsor tap-off palms, as required, to bolt them to theother sections.The structures supporting the IPB, at bends and teeoffsparticularly, are adequately reinforced to sustainthe excessive forces that may arise on a fault andtend to shear off such joints and structures. Even ina continuous type of enclosure it may not always bepossible to fully bond all such parts electrically.Should such a situation arise, current limiting reactorsmay be provided at the bends and the tee-off jointsto contain the excessive enclosure currents at suchlocations, particularly during faults.8 The exterior of the conductor and the interior of theenclosure are painted with non-magnetic matt finishblack paint to improve the heat transfer due toradiation.9 (a) Provision of space heaters is made to preventcondensation of vapour during a shutdown.(b) Shorting links may also be provided to shortcircuitthe buses for the purpose of generatordrying when required, after a long shutdown,due to a fall in insulation level.IO Locations having low ambient temperature or highhumidity may adversely affect the insulation levelof the bus system. In this case an ordinary spaceheater may not be adequate and a separate anticondensationarrangement may become imperativeto dry the condensate and to maintain a high dielectricstrength. This may be achieved by:Blowing hot air: A typical arrangement is shownin Figure 3 I .9. The bus duct enclosure is providedwith inlet and outlet valves at suitable locations.Hot air is supplied through the inlet valve untilthe interior of the enclosure is completely dry.The blowing equipment comprises an axial flowfan, a heater unit, an inlet air filter unit, pressure

An isolated phase bus system 31/937I- GeneratorIndoor I Outdoor/ *-ekkPower house wall-. . - Tap-off1- 1- .-to U.A.T- i-Hot airoutlett*, \t7Unit aux.transformerGeneratortransformerFigure 31.9 A typical arrangement illustrating drying of enclosure through hot air blowingvalves. required controls and interlocks and airducts.Pressurizing the bus enclosure: A typical arrangementis shown in Figure 31.10. In this case theenclosure is maintained at slightly positive pressureby continuously supplying it with dry and cleanair or SF, to prevent vapour condensation.Compressed air is supplied at a pressure of 5-7kg/cm' and passed through the air dryer and theflow control valve. The leakage, if any, is recom-mended to be within 5% of the volume of theenclosure per hour. The air dryer removes themoisture present in the compressed air and leavesthe air in a bone dry condition having a dew pointof -25°C. The dry air is then passed into theenclosure through the control valve and the flowindicator. The controls are set to allow the dry-airflow to the enclosure, until the pressure inside theenclosure reaches the required level.2 Pneumatic0 control valve@ Safety valveBus ductI I 0 Pressure sensor(for low pressure alarm,high pressure alarmGenerator Transformer and operation of thesafety valve etc.)A & B -Are sensing pointsC 8. D - Are purge pointsV I(3- -Compressedair supplyFigure 31.10 Typical layout illustrating pressurization system for an IPB bus system

31/938 Industrial Power Engineering and Applications Handbook31.4 Enclosure heatingHeating of enclosures for bus systems in such large ratingsis detrimental to their design. In smaller ratings (say, upto 3200 A) it has been easier to do this by increasing thesize of enclosure, choosing larger bus sections, changingtheir configuration or using a non-magnetic enclosure orby adopting more than one of these measures, as discussedin Chapter 28. But this is not the case in large ratings,where heat generated, particularly in the enclosure, isenormous due to carrying almost the same amount ofinduced currents as the rating of the main conductors.Therefore, dissipating the heat of the conductor and theenclosure is a major task in an IPB system. When designingsuch a system it i s imperative to first check the adequacyof the enclosure to dissipate all the heat generated withinpermissible limits.If the heat generated by induced currents (there are nomagnetic losses) on load on per phase basis isIn the main conductor, due to I, = W, = IF2 . R, ,In the enclosure, due to I, = Wz = If . R,In the outdoor portion of the enclosure, due tosolar radiation = Ws*then the total heat generated= W, + W, + Ws* in watts per unit length(*only for the outdoor parts)(31.2)whereI, = rated current of the main conductors in Amperes(r.m.s.)R, = a.c. resistance of the main conductors per unit lengthat the operating temperature, in RI, = Induced current through the enclosure in Amperes(r.m.s.). This will vary with the type of enclosure,the effectiveness of its electrical bonding and theimpedance of the bonding links. If current limitingreactors are used in the enclosure circuit this currentwould be reduced substantially, as discussed above.Without such current limiters, the current may beassumed to be up to 95% of I,. Experiments havecorroborated this assumption in continuous bussystems.Re = a.c. resistance of the enclosure, per unit length atthe operating temperature in R.To determine R, and R, the same procedure may beadopted as discussed in Section 28.7. For a tubularconductor,(3 1.3)(in MKS or FPS units, depending upon the system adopted)wherep = resistivity of metal at 20°C in Rm or in pRcmd, = mean diameter of the tube in m or cmt = thickness of the tube in m or cmI = length of the conductor in m or cmFrom similar skin effect curves, as in Figure 28.13(b),corresponding to tld the ratio Rac20/Rdc20 can be found.Having determined RaC2", it must be corrected to theoperating temperature by usingRdce = Rdc2[) 1 + a20 (0 - 2011 (3 1.4)whereRdce = d.c. resistance of the conductor at the operatingtemperatureazo = temperature resistance coefficient of the metal at20°C per-"C (Table 30.1)W,, W, and W, ( W,, when considering the outdoor part)can thus be determined. The total heat, W,, so generatedcan be naturally dissipated through the enclosure byradiation and convection. If natural cooling is not adequate,forced cooling can be adopted through forced air or water.But precautions must be taken to ensure that the systemis protected from absorbing dirt, dust or moisture fromthe atmosphere.NoteFor medium ratings as in a non-isolated bus system of up to 3200 Alittle significance is attached to such effects because of acomparativelyweaker magnetic field produced by the conductors of each phase.andcancellation ofapart ofit by the fieldsoftheotherphases (Figure28.22 and Section 28.8). The enclosure currents are now a muchlower percentage of the current of the main conductors and hencecause a low heating of the enclosure. The residual magnetic field insuchcases is alsosmall andis normallyignored.Yetcertaininstallationsmay require magnetic shielding in such ratings also where thesefields may influence the operation ofthe instruments that are installedin the close vicinity.31.4.1 Skin effect in a tubular conductorTubular conductors provide the most efficient system forcurrent carrying, particularly large currents. As discussedabove, the current density is the maximum at the skin(surface) of the conductor and falls rapidly towards thecore. Experiments have been conducted to establish thenormal pattern of current distribution in such conductorsat different depths from the surface (Figure 3 1.11).At a certain depth, the current density reduces to 1/&=0.368 of the value at the surface. This depth is termedthe depth of penetration, SP. Of the total heat generatedin such conductors, almost 80% occurs within this depth(annulus). In other words, around 90% of the total currentheat generated a 0.9' 2- 0.8 is concentrated in this areaalone. This depth can be represented by(31.5)wherep= resistivity of the metal in e.m.u.= 1 O9 x (p in R cm) as in Table 30.1.f'= frequency of the system in Hzp = effective permeability of the medium in which thefield exists (aluminium in the present case), andwill depend upon the electric field induced in theenclosure= I for non-magnetic materials6,will thus vary with the material of the conductor andthe enclosure and the operating temperature. For

An isolated phase bus system 311939100908to/0 $06%c5 0.5k2$04c_m$0302010I\ :Ilie = 0.368penetration)0 1 6P 26P 36PDepth from surface t +Figure 31.11 Skin effect in tubular isolated conductors interms of depth of penetration Spaluminium, grades EIE-M and E-91E, we have workedout the likely values of 6, at an operating temperature of85°C a\ follows:1 For grade EIE-Mpzo = 2.873 x IO-"a cmazo = 4.03 x IO-? per "C:. px4 = plo [I + 4.03 x IO-' x (85-ZO)]= p30 x 1.26195 R cmTherefore depth ot penetration at 20°C= 1.207 cm or 12.07 inmand at 85°CGpgT = 12.07.\/1.26195= 13.56 mm(The suffix refers to the temperature.)2 For grade E-91E:.pZn = 3. I 33 x I OF cma2,] = 3.63 x lo-' per "Cps5 = pzo ( 1 + 3.63 x IO-' x 65)= p20x I 236 !2 cm... Spz,, = -2nand6pp8s3.133 x IO-" x IO"1 x50= 1.26 cm or 12.6 mm= 12.6 1/12.16= 14.01 mmSince most of the current will tlow through SP, ;I thickerconductor will only add to the bulk and cost of the tubewithout proportionately raising its current-carryingcapacity. A greater thickness does not assist in heatdissipation, as the heat is dissipated more quickly fromthe outside than the inside surface of a body.31.4.2 Thickness of enclosureFor a near-total shielding of the field produced by themain conductors (i.e. for I, - I, to be very low), it isessential to have the thickness of the enclosure as near to6, as possible. But this may prove to be a costlyproposition. In addition, a higher induced current in theenclosure will also mean higher losses. This has beenestablished by computing the cost of the enclosure andcapitalizing the cost of losses for minimum losses in theenclosureThickness of tubular conductor ( I jDepth of penetration )= rrI2t = IT 6, (31.6)or-But this will also prove to be a costly proposition.A smaller surface area may require forced cooling andan extra cost for the same while a larger enclosure wouldmean a higher cost for the enclosure itself. Considerationsof cost will thus determine a larger enclosure or a forcedcooling. Two factors have been considered relevant here:Loss factor: This is a function of losses (W, j in themain conductor (W,) and the enclosure (W?). A curvcas illustrated in Figure 31.12 can be established betweenthe losses versus thickness (t) of the enclosure byexperiments.Optimization factor: This is a function of the cost ofenclosure for different thickness. 1, the cost of thecooling system (if cooling is considered necessary)and the capitalized cost of the losses at differentthickness t. A curve as shown in Figure 31.13. rathersimilar to that in Figure 31.12, can be establishedtheoretically between the total cost of the IPB systemversus t.The above two curves will help optimize the thickness,t, of the enclosure for a total rninirriurri ~ o soPlthe system. Enclosure losses may not be lowest at thisthickness as shown in Figure 31.12, but they wouldmaintain the temperature rise of the enclosure withinlimits. The magnetic field in the space. being alreadyvery low, would require no other measure. Moreover.a small field in the space may causc only a smallamount of heat in nearby structures, which may be

31/940 Industrial Power Engineering and Applications Handbookt52DP2-co,zs;, 1Minimum -losses0-Enclosure thickness 't't, = Thickness of enclosure for minimum cost of IPBsystem (Figure 31.13)Figure 31.12 Variation in losses with thickness of enclosuretE2$m9%.1v)8.-.SI!coCost at $4Minimuncost ~Enclosure thickness 'f'--CFigure 31.13 Optimization curve between thickness ofenclosure and the total cost of the IPB systemacceptable. Therefore, to choose a thinner enclosure iscommon practice to economize on the total cost,allowing part of the magnetic field to occupy the space.31.4.3 Proximity effect in an IPB systemThe three phases are now completely isolated andadequately spaced. They are thus hardly under anyinfluence of proximity. The forces are now greater betweenthe conductor and the enclosure rather than between twophases due to an almost negligible or only moderatemagnetic field in the space (Figures 3 1.7 and 3 1.8). Theemphasis is now more on the losses in the enclosure andthe metallic structures outside the enclosure and theirconsequent heating, rather than the derating of the currentcarryingconductors.To arrive at the most appropriate and economical designof the enclosure is a complex subject and requires detailedstudy. For brevity, in our present attempt, we have derivedinferences from the established work in this field byengineers and authors (see the further reading) and haveunderlined briefly the basic approach to design such asystem. For smaller ratings, up to 3200A, the discussionsin Chapter 28 will generally suffice to design a good bussystem. There we have assumed the content of proximityon the conductors and exercised care while selecting thesize and material of the enclosure, spacings between theenclosure and the busbars and between the busbars oftwo adjacent phases etc. In an IPB system, however,when the space occupied by the electric field is largemay cause excessive parasitic currents within the enclosureand in the metallic structures in close proximity outsidethe enclosure and excessive heating in both. Theassumptions made earlier may not suffice for its effecton the enclosure, supports and the structures in the vicinity.To provide magnetic shielding within the enclosurewill require the enclosure to be made of non-magneticmaterial at the first instance to eliminate magnetic lossesand a cross-section sufficient to carry a near full loadcurrent on the other. It is also essential to allow a nil oronly a moderate field in the space outside the enclosureto limit the parasitic currents in the supporting and othermetallic structures so that they do not require any specialtreatment or insulation and protect the operating personnelfrom excessive touch or step voltages. The main emphasisto design such a system is therefore to optimize thethickness of the enclosure and its overall size to obtainthe required magnetic shielding and a temperature risewithin desirable limits.31.4.4 Solar radiationThe part of the bus enclosure installed outdoors is exposedto solar radiation which is a cause of additional heat gainby the enclosure. ANSI C-37.24 has provided a basis todetermine its effect in terms of heat generated as follows:during winter - 670 W/m2during summer - 990 W/m2for tropical areas* - 1100-1 200 W/m2By simulating these effects in a test laboratory it hasbeen established that solar radiation may raise thetemperature of the external surfaces by up to 15"C,depending upon the colour and the condition of the surface.*Tropics are the regions of the earth that lie about 2570 km northand 2570 km south and parallel to the equator (Figure 31.14).These regions signify the areas that generally have a warm to hotclimate throughout the year, as the sun reaches its greatest altitude.

~~0.97~~~~An isolated phase bus system 31/941TropicalregionsFigure 31.14 World map showing Tropical regions (Courtesy: World Book Encyclopaedia)Surfaces with light colours absorb less heat than darksurfaces. But light colours may fade with the passage oftime. The outcr surfaces may also collect soot and dirtand hence lose the benefit of being light-coloured tosome extent. The heat generated by solar radiation perfoot length can be estimated byW, = k,l G . A Wift (31.7)wherek;, = coefficient of absorption. When a canopy is providedon the outdoor part of the enclosure the coefficient,k,, may be reduced.G = global solar radiation= I 100 W/m' (for tropical areas)A = area of the enclosure that is exposed to the sun. Wehave assumed about 50% of the surface area fallingoutdoors which is exposed to direct solar radiation.During heat dissipation by radiation the colour andcondition of the surface plays a similar role. Dark-colouredbodies dissipate more heat than the light-coloured ones.The amount of heat absorption and emission for the samebody may therefore be assumed to be almost the same.Accordingly, Table 3 1. I, for selected colours, may beconsidered for the coefficients of absorption and emissionof heat due to solar radiation and natural radiationrespectively.In our sample calculations (Example 3 1.1) we havechosen the colour of the outdoors surface as light greyand taking the weathering effect into account, haveconsidered the coefficient of both absorption and emissionas 0.65. The manufacturer, depending on the colour andsite conditions, may choose a suitable coefficient. It is,however, advisable to be conservative when deciding thetemperature rise due to solar radiation to be on the safeside.The normal practice of all leading manufacturers is tokeep the same bus and enclosure sections for the outdoorand indoor parts of the bus system. This is to retainsimplicity in design and ease of jointing and interconnections.But this may prove to be a costly arrangement, asnow the indoor section of the bus system will have to beTable 31.1 Coefficients of absorption (of solar radiation) andemission (natural radiation) for a metallic surface locatedoutdoorsWhiteLight greyMedium greyDark greyBlackCoejfirient ofn hs orption, K,, crndemission

31/942 Industrial Power Engineering and Applications HandbookA, = effective surface of radiation per foot of enclosurelength in square inchesT, = absolute temperature of the hot body= (e, + 273)"C (0, = enclosure surface temperaturein "C)T2 = absolute ambient temperature= (0, + 273)"C (0, = ambient temperature in "C)(For conductors, the enclosure temperature will becometheir ambient temperature)By convection (W,) (from the external surface ofthe enclosure): For a hot body installed indoors andreasonably ventilated this can be expressed by (Dwightet nl., 1940)(3 1.9)whereA, = effective surface area per foot of enclosure lengthin square inchesP = air pressure in atmospheres= 1 at sea level0 = temperature rise of the enclosure above theambient temperature in "Ch = width of flat or bar mounted in a vertical plane,or the diameter of a round conductor in inches.By forced convection The factors that can influencethe temperature of the enclosure, installed outdoorsare wind and snow, other than forced cooling. But theireffect on actual cooling may be small. Sometimes thishappens and sometimes not. It is better to ignore thiseffect when estimating various thermal effects. Naturalconvection and radiation will take account of this.31.6 Continuous ratingThe continuous current rating of a bus system can bedefined by the current at which a steady-state thermalcondition can be reached. It is a balance between theenclosure and the conductor's heat gain and heat loss. Ifthis temperature is more than the permissible steadystatethermal limit it must be reduced to the desired levelby increasing the size of the conductor or the enclosureor both, or by adopting forced cooling. Otherwise therating of the bus system will have to be reducedaccordingly.31.7 Forced coolingFor ratings up to 20 000 A natural cooling is found to begenerally adequate to maintain all temperatures withinlimits. Where this is not possible, as when the size of theenclosure cannot be increased for whatever constraints(generator termination being one), to dissipate the requiredheat, then forced cooling may be adopted. The generalpractice is to eliminate forced cooling as far as possible toavoid the piping network and the associated equipmentand accessories and their regular maintenance andupkeep.This is one major factor against forced cooling. A breakdownof cooling system will not result in a shutdown ofthe bus system but will require its deration and lead to anunderutilization of the generating unit. Forced cooling(by air, gas or liquid) is more relevant when the total lengthof the bus system is sufficient to justify initial investmentin the external cooling arrangements and their regularupkeep and also when one can economize on the cost ofconductor and the enclosure at the cost of forced cooling.For 25 000 A and above, heat dissipation through naturalcooling alone may not be sufficient. It may require alarger enclosure. But this may become too large toterminate at the generator end and be a limiting factor,besides becoming more expensive. In this case one canselect forced cooling.31.8 Influence of a space field onmetallic structuresThe influence of a induced field on a metallic (magnetic)structure is in the form of closed magnetic loops, whichcause hysteresis and eddy current losses. These closedloops cannot be broken by insulating magnetic structuresat bends or joints or any other locations. (Refer to Figure28.32 for more clarity.) There is thus no treatment thatcan be applied to such structures or bodies in the vicinityof an IPB to protect them from the magnetic effects ofthe field if present in the space.The IPB system must therefore be designed so that itallows only a moderate field in the space. The field shouldbe incapable of heating structures in the vicinity beyonda reasonable limit, or cause step and touch voltages.These voltages are undesirable and may become dangerousto a human coming into contact with them.31.9 Fault levelAs discussed in Section 31.2.2, in continuous enclosuresthere is a near-magnetic shielding between any two phases.The space field between the enclosures is also of theorder of 10-15% or even less (=(I,- I,)) than that of thefield in the enclosures. As a result, the electrodynamicand electromagnetic forces generated betweenthe enclosures during a fault are also only nominal(F, = (I,' - I,',, ) where I,, is the fault level of thesystem and I,,, the corresponding fault current throughthe enclosure). One therefore need to deal only the forcesbetween the conductor and the enclosure, which willremain as high as in a non-segregated bus system(F, = Is: ) rather than the forces between the enclosures.When such large generating units are connected to apower grid the bus system connecting the generator and

An isolated phase bus system 31/943the generator transformer will be subject to a higher faultlevel commensurate with the fault level of the power gridlimited upto the fault level of GT (Section 13.4.1.5 andFigure 13.1 I ). Similarly. the tap-offs connecting the UATswill be subject to a cumulative fault level. one of thegenerator and the other of the power grid (limited up toGT). Generally, therefore such large ratings ofbus systemsare subject to a high fault level. (For a few large ratingsof bus systems these lehels are shown in Table 13.8).For ratings of 25000 A and above it is possible thatthe fault level of the system may exceed the rupturingcapacity of the available interrupting devices. To reducethe fault level in such cases, current limiting series reactorscan be provided with the bus system, as noted above andillustrated in Figure 3 I .8.31 .I 0 Voltage dropAs di~ussed earlier. this has no relevance in HT systems.Octagonal sectionHexagonal section,A3 !31.11 Forming sections for IPBsystemsIt is not possible to extrude large sections in full circular,hexagonal or octagonal shapes. These are norinallyproduced in smaller extruded sections as illustrated inFigure 3 I . 15. Circular sections may, however, also berolled from sheets in two halves to obtain the desiredsize. Moderate sizes of circular sections are also availablein extruded form. For hexagonal and octagonal shapes.the thickness and lengths of such sections will dependupon the practices adopted by the manufxturer of theextruded sections. The desired size of conductor andenclosure can then be shop fabricated out of these sections,choosing the nearest next available size. to form theminto the required sire of conductor and enclosure.EnclosuresSinall margins are added in enclosures for inspectionand maintenance windows and tap-offs, which may affecttheir current ratings. These sections are welded circumferentiallyto give them the required shape (circular,hexagonal or octagonal). They are welded longitudinallyto shape them into the required length and configuration.Extreme care must be taken when making such joints, tomake them airtight and to ensure maximum continuityof current.ConductorThe conductor may not have any inspcction or maintenancewindows. but may be provided with surface openings, asillustrated in Figure 3 1.15 to dissipate its interior heatinto the enclosure for onward dissipation by the enclosure’swrface to the atmosphere. This facility. however, maynot be available in smaller circular conductors if they areextruded as one piece (which is possible in smallersections).Since the alumini~im sections are formed to individual* Slits are left (- 50 mm wide) to facilitate heat dissipation from theinside surface of the conductors [enclosures are totally closed]NoteNumber of sections would depend upon the dimensions of theconductor or the enclosureFigure 31.15 Forming a conductor or enclosure from theavailable extruded sectionsdesigns, there are no ready reckoners available fromaluminium manufacturers of extruded sections to quicklydetermine their current ratings. This may be establishedonly by conducting shop tests. It is not recommended toassign the rating on the basis of per unit area in view ofa number of factors that may influence the rating. suchas the surface area, the thickness of the enclosure or theconductor sections and ambient conditions. The designof a bus section is a matter of experience of the designengineer. As a rough guide, however. to establis,h thebasic parameters, a rating around 350 to 300 A h - maybe considered which will also takederatings. (Refer to the sample calculations in Example3 1. I .) The normal practice is to use a computer. whichcan provide a number of alternative designs.The cross-sections of the conductor and the enclosureare then checked for their adequacy to dis-ipate the heatgenerated. The cross-section is then adjusted suitably bypermutations and combinations to arrive at the mostappropriate size. keeping in mind the available extrudedsections, accounting for all such factors that may affectthe rating or the fault level of the enclosure and theconductor. such as tap-offs. which are subjected to acuinulative fault level, openings in the conductor for heatdissipation or the enclosure for inspection windows thatcan influence their ratings. A calculation in Example3 I . I will clarify the procedure to establish the sire of the

31/944 Industrial Power Engineering and Applications Handbookconductor and the enclosure for an IPB. The final sizesattained can then be laboratory tested to establish theiraccuracy.31.12 Determining the sectionand size of conductor andenclosureAn isolated phase bus is also a tailored system to suit aparticular layout. Size and area of cross-section will varywith the manufacturer. The insulators also play an importantrole in determining the diameter of the enclosure and itsthickness as well as the diameter and thickness of theconductor. The manufacturer of an IPB may have theinsulators made to specifications (for voltage, height,clearance and creepage distances) or make a choice fromwhat is available in the market and compromise on theenclosure’s diameter. To standardize on the size of conductorand enclosure for such systems is not normal practice. Amanufacturer, however, can do so for his adopted designpractices and standardize on some vital parameters suchas the type of section (round, hexagonal or octagonal),depending upon thecurrentrating as well as the availabilityof the extruded aluminium sections produced by indigenousmanufacturers, design of insulators, area of crosssectionand the system of cooling, etc. Generally, naturalcooling and hollow cylindrical sections are preferred.In smaller bus systems, say, up to 3200 A, standarddata for sizes of busbars and their current ratings havebeen long established, as noted in Section 30.2 and Tables30.4 and 30.5 for different sections and cross-sectionalareas. These data can be used in the design of a particularbus system. But an TPB has to be specially designed.Below we provide brief guidelines for designing such asystem from fundamentals, i.e. by applying the theory ofthermal equilibrium between the heat generated by thesystem (conductor and the enclosure) and that dissipatedby the conductor and the enclosure surfaces.The amount of heat generated will decide the size andshape of the main current-carrying conductor, the ratingand size of the enclosure and the provision of externalcooling, if necessary. As noted earlier, forced cooling, ifit can economize on the cost of the bus system, can beused. The one factor against the forced cooling is theneed for piping network, additional equipment and itsregular upkeep. Breakdown will not result in a totalshutdown but a reduced rating of the bus system and anunderutilization of the generating unit. For larger ratings,however, above 25 000 A, forced cooling may becomemandatory to dissipate the great heat generated in theenclosure. In such cases the size of the enclosure whichis restricted by the size of the generator terminals mayprove insufficient to dissipate the heat generated.We have estimated the likely heat that may be generatedby a particular size of conductor and enclosure for acertain current rating and then have countercheckedwhether the conductor and the enclosure so chosen candissipate this heat by radiation and natural convection,and reach a state of thermal stability within permissiblelimits or we may have to increase the size of the conductoror the enclosure, or both, and/or augment the heatdissipation through forced cooling to meet thermalrequirements. The theoretical design is then put tolaboratory tests to establish its accuracy.Current rating vanes with the surface area of a conductorand its thickness (annulus). In our sample calculation, toestablish the basic parameters of the conductor and theenclosure, we have considered the current density forboth as 400 A/inch2.The outdoor part of the enclosure has to perform moreonerous duties as it has to withstand weather conditionsand also absorb solar radiation. It has also to dissipatethe heat of the conductor in addition to its own. It istherefore possible that the surface area of enclosure sochosen may have to he increased, and this will be revealedduring thermal calculations which are carried out to checkits suitability.With this assumption, the basic cross-section of theconductor and the enclosure can be chosen. It is thencounterchecked whether the size so chosen is adequateto reach a thermal stability. When desired, the t can besuitably modified to reach thermal equilibrium. The sizescan he optimi7ed by plotting a few theoretical graphs:Losses versus t (Figure 31.12) andCost versus t (Figure 3 1.13) to optimize t and hence,the cost of the conductor and the enclosure. For this t,the diameter may be modified to arrive at a moreeconomical design. A higher t will mean a lowerdiameter and vice versa, and may be modified to satisfythe conductor’s current-carrying and the enclosure’sheat-dissipating requirements. Since the size of a hollowconductor, for large to very large current ratings, isnot standardized the cylindrical diameter (6) and thewall thickness (t) can be varied, depending upon therating, the extruded sections available and the coolingsystem adopted by the manufacturer. The exact d andt is then established by trial, and optimized as notedabove. Figures 3 1.12 and 3 1.13 suggest that for a moreeconomical design, the thickness must be less than thedepth of penetration (6,). In enclosures a moreeconomical design is achieved by keeping t in thevicinity of 50-60% of 4. This may mean some fieldin the space, but its severity is already mitigated byarresting most of it by the enclosure. Leadingmanufacturers have established their own data andprogrammed them on computers for routine referenceand designing a bus system. These data are then checkedfor their accuracy by conducting heat run tests on samplelengths (Section 32.3.4). After a few initial designs itis possible to optimize and predefine the vital parametersfor a particular rating.The exercise below is purely theoretical, based on theinformation and the data available on the subject, with aview to providing the basic guidelines to a practisingengineer to design an IPB system.Since ANSI-C-37.23 is available in the FPS system,we have adopted this for ease of corroboration. Also,since continuous enclosures are more often used for theirobvious advantages, we have also considered them in

An isolated phase bus system 311945our sample calculations. For details of non-continuousencl-osures, reference may be made to the work noted inIEE Committee (1968).31.13 Sample calculationsExample 31.1To apply what we have discussed so far we give below abrief outline of a design for an IPB system. The economics ofthis design would be a matter of further investigation of itsperformance versus cost. This will require an optimization onthe thickness of the enclosure as discussed earlier.When optimizing the natural cooling of the enclosure itsdiameter may have to be increased but this would add to thecost of the IPB, on the one hand and make it too large toterminate at the generator end, on the other. The total designis thus a matter of many permutations and combinations.Most manufacturers have optimized their designs, based onexperience and test results, through complex computerprogramming. In this example we will suggest only a basicapproach to establish a design. Its optimization would requirean elaborate cost analysis as noted earlier although itsauthenticity can be verified through laboratory tests.Assume the following parameters of a turbo-alternator (Table13.8):Generator size (MW)500 at 50 HzRated p.f. 0.85Rated MVA -- 500 - 5880.85Stator nominal voltage (kV) 21Stator maximum voltage (kV) 24Stator current (CMR) (A)16 200Rated current of IPB (CMR) (A) 18 000IPB to be designed for (A)20 ooo*One-second short-time rating (kA) 122Type of enclosure for IPBContinuousWe limit our discussion to the main length of the generatorbus. A similar exercise can be carried out to design the neutralshorting star bus, tap-offs to the two UATs, the delta bus tointerconnect the GT when it is made of three single-phaseunits, tap-offs for CTs, VTs and surge protective equipmentand all the auxiliary buses described in Section 31.1.1. Someof these tap-offs, such as for UATs, which are low-rating bussections, can also be made with the construction of segregatedphase bus system (Section 28.2.2) to economize on cost.RequirementsAmbient temperature for enclosure (& in "C)Ambient temperature for conductor (Oat in%),for indoors80 The same as forfor outdoors95 1 the enclosureConductor's maximum allowable temperature (0, in "C)for indoors95 Assuming weldedfor outdoors105 i or silver-plated jointsEnclosure's maximum permissible temperature (0, in "C)for indoors 80for outdoors 95AssumptionsConductivity of conductorThis is a normal safety margin (= 10%) for such duties. Itaccounts for the dip in the grid voltage to which it is connectedwhich may require the generator to operate at a lower voltagebut still cope with the full-load (MVA) demand, resulting in acorresponding rise in its current for a short period.48and enclosure (Yo IACS)60 (at 20"C, Table 30.1)GradeEIESpecific resistance2.873 (at 20°C,(P20 in cm) Table 30.1)= 2.873 x QcmTemperature coefficient ofelectrical resistance (a2,, in per "C) 3.96 x(at 20°C, Table 30.1)Effect of solar radiation on theoutdoor part of the bus section ("C) 1 O+Factor of emissivity, for the purpose of heat dissipation forthe light grey surface of the enclosure noted above (e) 0.65*Cross-sections of conductor and enclosure are to be thesame for indoor and the outdoor parts. This is normal practiceof all manufacturers to achieve simplicity in design and easeof interconnections. We have considered a circular conductor(the enclosure is usually circular).Establishing the size of the conductor20 000Approximate area of cross-section = - 400= 50 square inchesLet us assume the thickness of the conductor to be close tobut less than 4, say, around 11 mm, to reduce its diameterand hence increase the gap between the enclosure and theconductor.:. t = 0.43"and for the conductor's outer diameter d,Area of hollow pipes = E d,? -4 4 dg ]50 = (dj + d2)(dl - d2)4= rr(d- 0.43) x 0.43_._ d = ___ 50 + 0.43rt x 0.43= 37.46and the conductor's outer surface area (no() per foot runA,, = K x 37.46 x 12= 1411.49 square inches(U is the length of conductor in inches)Conductor's inner surface area per foot runA,, = ~[37.46 - 2 x 0.431 x 12= 1379.09 square inchesEstablishing the size of the enclosureAs discussed above, that thinner the section, the fewer will+Assuming the approximate absorption coefficient of solarradiation to be 0.65, for a light-grey external surface, havingcollected soot and dirt over a period of time, as in ANSI-C-37-24 the approximate temperature rise on account of solarradiation*We have considered the emission of heat, from the surfacethrough natural radiation, nearly the same, as its absorptionof heat through solar radiation.

31/946 Industrial Power Engineering and Applications Handbookbe the losses and the more economical will be the design.But we cannot reduce the thickness and increase the diameterof the enclosure indefinitely. A very large diameter may alsobecome unsuitable to match at the generator end. Thegenerator terminal dimensions, in fact, decide the size of theenclosure. For the generator rating under consideration, weare taking the outside diameter of the enclosure as 1500 mm(it should be considered according to the site requirementand adjusted for the support insulators selected). The terminalspacing of the generator between phases will also determinethe centre spacing of the IPB. In our calculations we assumeit to be around 68 inches (which may almost be the case inpractice).Enclosure's approximate area of20 000 x 0.98-cross-section 400= 49 square inches(assuming the enclosure's current to be around 98% of theconductor's currentEnclosure's outer diameter = 1500 mmD = 59 inches:. enclosure thickness t c- 0.27 inch (Figure 31.16)[49 = n(D - t) . t = IT(= 59)tl:. exact area of cross-section =(59' - (59- 0.54)')4= 49.79 square inchesEnclosure surface area per foot run A, = x x 59 x 12= 2223.12 square inchesChecking the suitability of the above sections byheat balancingHeat generated (In the outdoor part)1 By the conductorWI = I: R,, W/ft= I,' Rdc x skin effect ratioand- (2.873 x C2 cm) x (1' x 12 x 2.54 cm)IT x [(37.46- 0.43) x 2.54 cm] (0.43 x 2.54 cm)= 0.271 x Rift%c105* = RdcPO [l + (lo5 - 20)1= 0.271 x 10-6[1 + 3.96 x= 0.271 x [I,33661= 0.3622 x WftSkin effect ratio, from Figure 31.17 for= 371.54and t/d = -0.4337.46= 0.011:. Skin effect ratio = 1.03:. W, = (20000)' x 0.3622 x 1 0-6 x 1.03= 149.22 Wlft2 By the enclosure(a) By currentW, = (0.98 /,)' . RaCx 851The current through the continuous enclosure is assumed tobe at about 98% of that of the main conductor. The actualcurrent may be less than this as a result of low enclosurethickness, which is considered to be less than the depth ofpenetration 4. But it may not cause a high field in the space,leading to heating of steel structures, beams, columns orother equipment installed in the vicinity or step and touchvoltages higher than permissible. This is a subject for furtherinvestigation and experiments need be conducted to establishthat the field in the space is within reasonable limits. Otherwisethe enclosure thickness may need to be increased slightlyand the diameter reduced accordingly.(2.873 x lo4) (1 x 12 x 2.54)Now, R,m =IT x (59- .27) x 2.54 x 0.27 x 2.54= 0.273 x aftand RdcgSt = 0.273 x [l + 3.96 x x (95- 20)]= 0.273 x 10-6[1 .297]= 0.354 x 10-6 aftjwand skin effect ratio from Figure 31.17for or 375.80.27and tld = - 59 = 0.0046Figure 31.16 IPB conductor and enclosure for Example 31.1*We have the conductor to be operating at its maximumpermissible temperature, although it may be operating atless than this and generating less heat than assumed.tHere also we have assumed the maximum operatingtemperature although the enclosure may be operating at lessthan this.

An isolated phase bus system 31/947t/ d0.01 140.015a!0.0080.006I0.00460.0041.0050 50 100 150 200 250 300 350-400 450 500 550 600Jm&, in .u ST per footNote Figure 28.13(b) in RllOOO rn is redrawn in pR/ft.0.002Figure 31.17 Curves for skin effect for isolated tubular conductors:. skin effect ratio = 1.005:. W2 = (0.98 x 20 000)' x 0.354 x 1 0-6 x 1.005= 136.67 W/ft(b) By solar radiation (W,)W, = k, . G . A Wlftwherek, = 0.65G = 1100 W/m2 (for tropical areas)- 1100 WlfP3.28 x 3.28= 102.24 W/ftzA = 0.5 x ___ 2223.1 * ft' (assuming only the upper 50% surface144to be subjected to direct solar radiation)( 0.5 ~~22~23.12.. W, = 0.65 x 102.24 x= 512.98 Wlft]:. Total heat generated (within the enclosure)= 149.22 + 136.67 + 512.98= 798.87 wiftSuitability of the conductorHeat dissipation by the conductor(1) by Radiation36 9W, = -. 10''e.A,[T: - Ti]Wlffwheree = 0.85 (conductor is painted matt black withnonmagnetic paint)A, = 141 1.49 square inchesTI = 105 + 273= 378°CT2 = 95 + 273= 368

311948 Industrial Power Engineering and Applications Handbook... w, = - x 0.85 x 1411.49 [3784 - 36a4]36.91012= 91.92 Wlft2 By convection(i) From the outer surfacePo-5 ' I9=5W, = 0.0022 A, ho,25A,, = 1411.49 square inchesp = 1 (near sea level)e= i05-95=io'ch = 37.46 inches.'. W, = 0.0022 x 141 1.49 x ~= 22.32 Wlft(ii) From the inner surface'Asi = 1379.09 square inchesWlft101.2537.46°.25h = 37.46 - 2 x 0.43 = 36.6 inches101.25W, = 0.0022 x 1379.09 x -36.6°.25= 21.93 W/ftI. Total heat dissipated= 91.92 + 22.32 + 21.93= 136.17 Wlftas against the total heat generated of 149.22 wattlft. Theconductor chosen appears to be very close to the requiredsize. Its diameter or thickness may be slightly modified toachieve thermal equilibrium and a more appropriate size.3 By enclosure(a) By radiation36.9W, =-.e.AS[T,4-T:]W/ft10'2wheree = 0.65 (This is on a conservative side. For dull, nonmagneticmatt finish painted surfaces, which suchenclosures generally are, some manufacturersconsider this factor up to 0.85)A, = 2223.12 square inchesTI = 95 + 273= 368°CT2 = 48 + 273= 321°C:.36.9W, = - x 0.65 x 2223.12 [3684 - 321"]10l2= 411.76 Wlft(b) By convection (W,)PO5 .W, = 0.0022 A, ho,25Wlft'This dissipation of heat will not be applicable when the circularsection is made of circular extruded sections which have nosurface openings.whereA, = 2223.12 square inchesp = 1 (near sea level)e=95-48= 47°Ch = 59 inchesW, = 0.0022 x 2223.12 x -1 X47l.2559.25= 217.17 Wlft:. Total heat dissipated= 411.76 + 217.17= 628.93 Wtftcompared to the total heat generated of 798.87 Wlft.NotesIf we assume the emissivity factor to be 0.85 the shortfallin the heat dissipation is nearly made up. However, itwould be a matter of laboratory testing to establish thesuitability of the enclosure section chosen. Otherwise thethickness of the enclosure may be slightly increased andthe calculations repeated.Add a suitable margin in the conductor's cross-section toaccount for the opening considered to dissipate the insideheat. Similarly, add a suitable margin in the enclosure'scross-section to account for the inspection windows.In fact, it is the solar effect that is causing the maximumheat. The factors considered for the solar effect are alsohighly conservative. Nevertheless, a canopy over theoutdoor part is advisable in the above case. This willensure the same size of enclosure for the outdoor as wellas the indoor parts and also eliminate the requirement fora thicker enclosure or a forced cooling arrangement. Nowthere will be no direct solar radiation over the bus systemand the total solar effect can be eliminated, except forsubstituting the indoor ambient temperature of 48°C withthe maximum outdoor temperature for the outdoor part ofthe bus system.If we can overcome the solar effect, the size of the conductorand the enclosure can be reduced to economize theircosts. Another exercise with reduced dimensions will benecessary for this until the most economical sections areestablished.Once the required sizes of the conductor and the enclosureare established, one can choose the nearest economicalsizes of the extruded sections available in the market, orhave them specially manufactured if possible.A design engineer will be the best judge of the bestand the most economical design and extruded sections.The above example is only for the outdoor part of thebus system. The indoor part, in any case, would be coolerthan the outdoor one and will also provide a heat sink tothe hotter enclosure and the conductor constructedoutdoors. No separate exercise is therefore carried outfor the indoor part of the bus system, for the sake ofbrevity. For a realistic design that would be essential.The above example provides a basic approach to thedcsign of an IPB system. With some permutations andcombinations, a more realistic and economical designcan be achieved. A computer is necessary for this exercise.

An isolated phase bus system 31/949Relevant StandardsIEC Title IS BS6005911999 Standard current ratings (Based on Renald Series R-10 of ISO-3)Relevant US Standards ANSIINEMA and IEEEANSInEEE-C-37-23/1992ANSVIEEE-C37.24/1998Metal enclosed bus and Guide for calculating losses in isolated phase busGuide for evaluating the effect of solar radiation on outdoor metal enclosed switchgeaNotes1 In the tables of Relevant Standards in this book while the latest editions of the Standards are provided, it is possible that revised editionshave become available. With the advances of technology andor its application, the updating of Standards is a continuous process bydifferent Standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a Standard.23Some of the BS or IS Standards mentioned against IEC may not be identical.The year noted against each Standard may also refer to the year of its last amendment and not necessarily the year of publication.List of formulae usedSkin effect in a tubular conductorConstructional featuresContinuous or bonded enclosures@space(I,- I,)= magnetic field in spaceI, = conductor currentI, = enclosure current(31.1)(31.5)dp = depth of penetration in mmp = resistivity of the metal in e.m.u.f = frequency of the system in Hzp = effective permeability of the medium, in whichthe field exists and will depend upon the electricfield induced in the enclosure= 1 for non-magnetic materialsEnclosure heatingTotal heat generated = W, + W2 + W,*(31.2)in watts per unit length. (*only for the outdoor parts)W1 = heat generated in the conductorW, = heat generated in the enclosureW,* = heat generated due to solar radiationFor tubular conductorsRdc20 = P1 in Q/m or pQ/ft (31.3)(in MKS or FPS units, depending upon the system adopted)p = resistivity of metal at 20°C in Qm or in pQ cmd,,, = mean diameter of the tube in m or cmt = thickness of the tube in m or cm1 = length of the conductor in m or cm&ice = Rdc20[1 + a20 (6 - 2011 (31.4)Rdce = d.c. resistance of the conductor at the operatingtemperaturea20 = temperature resistance coefficient of the metal at20°C per "CThickness of enclosureFor minimum losses,Thickness of tubular conductor (t)= 7d2Depth of penetration (6, )orSolar radiationt=-ap n2(31.6)W, = k, . G . A W/ft (3 1.7)k, = coefficient of absorptionG = global solar radiationA = area of the enclosure, that is exposed to the sunNatural cooling of enclosuresBy radiation(31.8)e = coefticient of emissivity, expressed by the conditionof the surface of the hot body

31/950 Industrial Power Engineering and Applications HandbookA, = effective surface of radiation per foot of enclosurelength in square inchesTI = absolute temperature of the hot body= (6, + 273)OC (e, = enclosure surface temperaturein “C)T2 = absolute ambient temperature= (6, + 273)”C (0, = ambient temperature in “C)(For conductors, the enclosure temperature willbecome their ambient temperature).By convectionpO.5 , @1.25h 0.25W, E 0.0022. A, . W/ft(31.9)A, = effective surface area per foot of enclosure length- in square inchesP = air pressure in atmospheres= 1 at sea level6 = temperature rise of the enclosure above the ambienttemperature in “Ch = width of flat or bar mounted in a vertical plane, orthe diameter of a round conductor in inchesFurther reading1 Albright, R.H., Bates, A.C., Conanagla, A. and Owens, J.B.,‘Isolated-phase metal enclosed conductors for large electricgenerators’, Trans. IEEE, 81, February, 1067-72 (1963).2 Ashdown, K.T. and Swerdlow, N., ‘Cantilever-loaded insulatorsfor isolated phase bus’, AIEE paper, April (1954).3 Buchanan, G.E., ‘Lab field test experience with high capacityisolated-phase bus’, Trans. IEEE, 78, October, 925-93 1 (1959).4 Conangla, A.C., ‘Heat losses in isolated phase enclosure’, AIEETrans. June, 309-3 13 (1 963).5 Conangla, A. and White, H.E., ‘Isolated-phase bus enclosureloss factors’, Trans. IEEE, 87, August, 1622-1628 (1968).6 Dwight, H.B., ‘Theory for proximity effects in wires, thintubes and sheaths’, AZEE Trans. 42 (1923).7 Dwight, H.B., ‘Some proximity effects formulae for busenclosures’, Trans. IEEE, 83, December 1167-1 172 (1964).8 Dwight, H.B., ‘Proximity effect formulae for bus enclosures’,Trans. IEEE, 87, August, 1622-1628 (1968).9 Dwight, H.B., Andrews, G.W. and Tileston Jr H.W.,‘Temperature rise of busbars’, Generul Electric Review, May( 1940).10 Elgar, E.C., Rehder, R.H. and Sverdlow, N., ‘Measured lossesin isolated-phase and comparison with calculated values’, Trans.IEEE, 87, August, 1724-1730 (1968).11 Gupta, C.P., and Prdkash, R., Engineering Heat Transfer,Nemchand Bros, Roorkee, India.12 IEEE Committee, ‘Proposed guide for calculating losses inisolated phase bus’, Trans. IEEE, 87,August, 17361737 (1968).13 Jain, M.P., Proximity Effects of Busbars, ME thesis, ElectricalEngineering Department. University of Roorkee, Roorkee, India(1969).14 Jain, M.P. and Ray, L.M., ‘Field pattern and associated lossesin aluminium sheet in presence of strip busbar’, Trans. IEEE,89, Scptcmber/October, 1525-1539 (1970).15 Jain, M.P. and Ray, L.M., ‘An experimental study of proximityeffects of busbars’, Journal of Institution (.$Engineers (India),51, No. 12, Part EM, August, 341-350 (1971).16 Mankoff, L.L., Sverdlow, N. and Wilson, W.R., ‘An analoguemethod for determining losses in isolated-phase bus enclosures’,Trans. IEEE, 82, August, 532-542 (1963).17 Niemoller, A.R., ‘Isolated-phase bus enclosure currents’, Trans.IEEE, 87, August, 1714-1718 (1968).18 Paulus, C.F. and Bellack, J.H., ‘Progress in generator leads:Economical application of forced cooling’, Trans. IEEE, 88,February, 175-180 (1969).19 Poritsky, H. and Jerrard, R.P., ‘Eddy current losses’ Trans.AIEE, Part I, 73, May, 97-108 (1954).20 Sebold, G.K., ‘Induced current, in supporting steel for 10 000Ampere. generator open bus’, Trans. IEEE, 642447 (1960).21 Skeats, W.F. and Sverdlow, N., ‘Minimising the magnetic filedsurrounding isolated phase bus by electrically continuousenclosures’, AZEE Trans. No. 62 (1962).22 Skeats, W.F. and Sverdlow, N., ‘Minimising the magnetic fieldsurrounding isolated phase bus by electrically continuousenclosures’, Truns. IEEE, 81, February, 655-657 (1963).23 Stratton, J.A., Electromagnetic Theory, McGraw-Hill, NewYork(1941).24 Sverdlow, N. and Buchta, M.A., ‘Practical solution of inductionheating problems resulting from high current buses’, Trans.IEEE, 78, 1736-1742 (1959).25 Wilson, W.R. and Mankoff, L.L., ‘Short circuit forces in isolatedphase buses, AIEE Trans., April (1954).

Testing a MetalenclosedBusSystem321951

Testing a metal-enclosed bus system 32/953Testing a bus system is generally along similar lines tothose for a cwitchgear assembly, discussed in Chapter14. In this chapter we discuss additional tests that arespecifically for a metal-enclosed bus system.32.1 Philosophy of quality systemsThis has been covercd in Section 11.1.32.1.1 Quality assuranceTo fulfil the quality requirements, the material inputsgoing into the making of a bus system must be properlychecked as soon as they are received,Aluminium or copper scctions and sheets (for theircross-sectional areas. thickness of sheet, surface finish.bending properties and conductivity etc.)Hardware (for proper size, quality of threads and tensilestrength etc.)Insulators and supports (for sizes and quality of material)Other materials, components or equipmcnt used formaking. inter-connections and bondings of such systemsCooling systems (in large current rating systems).All these items must be properly checked and recordedaccording to the manufacturer’s internal quality checksand formats before they are used in the manufacture of abus system. This will eliminate any inconsistency in amaterial or component at the initial stage. Similarly, stageinspections are necessary during the course of manufacturingto ensure quality at every stage and to eliminateincorrect construction and assembly or poor workmanship.And thus assure a product of desired specifications andquality.32.1.2 Purpose of testingThe purpose of testing of a bus system is to ensure itscompliance with design parameters, material inputs andmanufacturing consistency.32.2 Recommended testsThe following are the recommended tests that may becarried out on a completed bus duct, as in IEC 60298,ANSI C-37/20C. IEC 60694 and BS 159 for both LT andHT systems.32.2.1 Type testsType tests are conducted on the first assembly (bus system)of each voltage, current rating and fault level to demonstratecompliance with the electrical and constructionaldesign parameters. The tests provide a standard referencefor any subsequent assembly with similar ratings andconstructional detailh. The following tests are conducted:of the leakage current, both before and after thedielectric test2 Verification of dielectric properties:(i) Power frequency voltage withstand or HV test(ii) Impulse voltage withstand test for system voltages2.4 kV and above3 Verification of temperature rise limits (for ratedcontinuous current capacity)4 Verification of short-circuit strength:(a) For straight lengths and tap-offs(b) For the tap-offs in a power-generating station,connecting a UAT through the main bus sectionbetween the generator and the generatortransformer.(c) For the ground bus in isolated phase bus (IPB)systems5 Verification of momentary peak or dynamic current6 Verification of protective circuit7 Verification of clearance and creepage distances8 Verification of degree of protection:(a) Enclosure test(b) Watertightness test - for all outdoor parts of anybus system (but for outdoor as well as indoorparts for isolated phase bus (IPB) systems)(c) Air leakage test, for isolated phase bus (IPB)systems9 Measurement of resistance and reactanceIO Endurance of trunking system with trolley type tapofffaci 1 i tiesI I Additional test on an IPB enclosure for radio interferenceas in NEMA-107. The maximum radioinfluence voltage (RIV) should not exceed 100 pV at1000 kHz. For test equipment and test procedure referto the Standard.32.2.2 Routine testsRoutine tests are conducted on each completed bus system,irrespective of voltage, current, fault level and constructionaldetails and whether or not it has undergonetype tests. The following will form routine tests:To check for any human errorGeneral inspection of the bus assemblyInspection of electrical wiring if there is any (such asfor space heaters, cold or hot air blowing, enclosurepressurizing or any other protective circuit)Verification of insulation resistance or measurementof the leakage current, both before and after thedielectric HV testVerification of dielectric properties - limited to powerfrequency voltage withstand test or HV testAdditional tests for an IPB system are:(a) Partial discharge test: On all cast resin componentssuch as instrument transformers, bushings andinsulators to ensure that the insulation is free fromdefects and voids.(b) Checking of welded joints: as in GDCD-198. normsfor aluminium welding (CEGB. UK)*: All shop-1 Verification of insulation resistance or measurement *Central Electricity Generating Board. U K

321954 Industrial Power Engineering and Applications Handbookwelded joints will be subjected to dye penetrationexamination and 10% of butt-welded joints,including joints on flexibles, enclosures andconductors, will be subjected to radiographic (Xray)examination.32.2.3 Seismic disturbancesIn Section 14.6 we have provided a brief account of suchdisturbances as well as the recommended tests andprocedures to verify the suitability of critical enclosuresand bus systems for locations that are earthquakeprone.For this the user is required to provide themanufacturer with the intensity of seismic effects at siteof the installation in the form of response spectra (RS).(See Section 14.6.)32.2.4 Field testsThese are to be conducted aftcr the installation and beforeenergizing the bus assembly at site:Checking for any human errorVisual inspection of the bus assemblyInspection of electrical wiring if there is any (such asfor space heaters, cold or hot air blowing or enclosurepressurizingor any other protective circuit)Verification of insulation resistance or measurementof the leakage current, both before and after thedielectric HV test, if the HV test is to be carried out.Verification of dielectric properties - limited to powerfrequency voltage withstand or HV test (usually notrecommended)Watertightness and air leakage test for isolated phasebus systems.32.3 Procedure for type testsBelow we briefly outline the procedure for conductingtype tests at the manufacturer's works.32.3.1 Verification of insulation resistance ormeasurement of the leakage currentThe procedure and test requirements will remain the sameas that discussed in Section 14.3.2, for metal-enclosedswitchgear and controlgear assemblies.32.3.2 Verification of dielectric propertiesPower frequency voltage withstand or HV testThe procedure will be the same as that discussed forswitchgear assemblies (Section 14.3.3). The test voltagemay be applied for one minute as shown in Tables 32.1 (A)or (B) for series I and Table 32.2 for series I1 voltagesystems. Any disruptive discharge or insulation breakdownduring the application of high voltage will be consideredto be dielectric failure.The tap-offs in a power-generating station connectinga UAT through the main bus section between the generatorand the generator transformer which are under thecumulative influence of two separate power sources, mustbe subjected to a higher withstand voltage as prescribedby IS 8084 and indicated in column 4, Table 32.1(A).32.3.3 Impulse voltage withstand test (for systemvoltages 2.4 kV and above)The procedure for testing will be the same as that discussedfor switchgear assemblies in Section 14.3.4. The impulsetest voltage is applied as in Table 32.1(A) for series Iand Table 32.2, for series I1 voltage systems with a fullwave standard lightning impulse of 1.2J50 (Section 17.6.1).There should be no disruptive discharge or insulationbreakdown.32.3.4 Verification of temperature rise limits (orrated continuous current capacity)(Recommended for systems having a current rating ofmore than 400 A)This will be carried out under similar parameters ofroom condition and the type of test voltage wave tothose for a switchgear assembly (Section 14.3.5). Thecurrent in each phase should be within 2% of the specifiedtest value (rated current).For LT bus systems: The length of the test piece willbe a minimum 6 m as in IEC 60439-2 with at least onejoint. The joints must be both in the conductor and inthe enclosure in each phase. If the total length of the bussection is less than specified, the entire length of thebus system in the fully assembled form will then betested.For HT bus systems: The entire length of the bus systemmanufactured for a particular installation in the fullyassembled form will be tested. Each end of the busenclosure will be properly sealed to eliminate any heatleakage.The bus sample is located 600 mm above floor level.The ambient temperature should be within 10-40"C asmeasured by the average value of at least four thermometers,two placed on each side of the enclosure on thecentre line, at least 300 mm from it, and 600 mm fromthe ends. The bus sample must be a three-phase unit aswell as the test equipment for a three-phase system toaccount for the proximity effect. This effect will not bereflected if a three-phase system is tested on a singlephaseone. Similarly, a single-phase unit and single-phasetest equipment should be used for a single-phase system(such as from the neutral side of the generator up to theneutral grounding transformer). The test will be conductedat the rated current until a near-stable condition is reachedand three successive readings at not less than one-hourintervals for an LT system and 30 minutes for an HTsystem do not show a maximum variation of more than1°C. To shorten the test, the current may be increasedduring the initial period. The test may be conducted at areduced voltage, as the emphasis is on heating due tocurrent. For a successful test, the hottest spot temperatureshould not exceed the values in Table 32.3.

Testing a metal-enclosed bus system 32/955Table 32.1(A)bus systemsFor series I voltage systems: Insulation levels, power frequency and impulse withstand voltages for metal-enclosedNominal systemvoltageRated max. systemvoltageOne-minute power frequency voltage withstandat a frequency between 45 and 65 Hz for LT and25 and 100 Hz for HT systemsStandard lightning impulse,1.2/50 p voltage withstand(phase to ground)1 2 3 4 5kV (r.m.s.)0.4150.63.36.611152233For a switchgearassembly and a bussystem(phase to ground)kV (r.m.s.) kV (r.m.s.) kV (rms.)0.440.663.67.21217.524362.52.5102028385070For the tap-offs in apower-generating station,connecting a UAT throughthe main bus section betweenthe generator and thegenerator transformer(Figure 31.1) (phase toground)--212735-5575(a) We consider here onlyList 11, which is moreprevalent. For List I,refer to Table 13.2.(b) For impulse voltagewithstand across the contactgaps of an interruptingdevice refer toIEC 60694kV (peak)--40607595125170Based on IEC 60439-2 LT, IEC 60694 for HT and IS 8084 for tap-offs.Notes1 Generally, there is no application of a bus system beyond a maximum system voltage of 36 kV.2 Field tests (power frequency voltage withstand) after erection at site, if required, may be conducted at 85% of the values indicated abovein LT and 80% in HT systems.Table 32.1(8)For series I voltage systems: Power frequency voltage withstand levels for metal-enclosed bus systemsNominal system Rated maximum One-minute power frequencyvoltage system voltage withstand voltage at anyv, frequency between 25 and100 Hz (phase to ground)I 2 3Field test after erection at site (phase to ground)A.C. test lfor one minute at anyfrequency between 25 and 100 Hz)4 5D. C tesPkV (mu.) kV (r.m.s.) kV (r.m.s.)0.415 0.44 2.00.6 0.66 2.03.3 3.6 9.56.6 7.2 17.011 1215 17.522 2433 3627.036.052.076.0kV (rms.)2.02.08.615.224.032.046.068.0(2V, + 2) for HT systemskV3.03.05.010.518.025.037.560.0Based on BS 159* In the absence of a.c. test voltage at site, a d.c. test can also be conducted, the duration of which will be 15 minutes.

:(r.mkV~ 350~ kVNominal system Rated maximum One-minute power frequency voltage Standard lightning impulse One-minute d.c.volrage system voltage withstand at afrequency not less than (I.2/50 p) voltage withstand i voltage withstand1the rated, 60 Hz (phase to ground) (phase to ground) I (phase to ground)I -7IkV (rm S kV (rm.s.) ~ s ) Dew ~ (peak)0.64.1613.814.423.034.569.00.6354.7615.0015.5025.8038.0072.501 80.0(1 minute) (I 0 seconds)2.219.036.050.060.0160.015.024304070140i-6095I25' 150For application of isolated phase bus to generator, the following voltage ratings will apply.3.127.050.070.014.4 to 242 _.- 50.0 50.0 i 110 70Table 32.3 Hottest spot temperature rise limits for metal-enclosed bus systemsComponent1 Bus conductors(i) Bolted plain jnints(ii) Bolted joints silver plated(iii) Aluminium welded joints2 Enclosure and support structures(i) Easily accessible(ii) Not accessible3 Bus conductors in contact with insulating materialsLimit of hottest spot temperature rise over an ambient of40°C - in "CAs in IEEEX37.20As in IEC-60439-2(Tuble 28.2) (Table 14.5)30'65b65b5065b65b30'5050*(or higher, as limited by insulation class, Table 14.6e)equivalent connections

vd3Testing a metal-enclosed bus system 32/957Effect of solar radiationThe effect of solar radiation must be considered in theoutdoor part of a bus enclosure exposed to solar heat.This effect may be observed for all LT and HT busenclosures, as all are influenced equally by such radiations,particularly IPB enclosures, as discussed in Section 3 I .4.4.32.3.5 Verification of short-circuit strengthFor straight lengths and tap-offsThe test procedure is generally the same as that discussedfor switchgear assemblies (Section 14.3.6). It is carriedout on a similar section to that used for the temperaturerise test and the power supply can be single-phase orthree-phase. If three-phase, the bus conductors are shortedat one end while the other ends are connected to thepower source. If a single-phase power is used, the circuitshould be so arranged that the current flows through thetwo adjacent phase conductors. The force due to a threephasefault would be approximately 86.6% of this.Therefore, for single-phase tests, the current would bezix6.6 or approximately 93% of that prescribed in Table13.10. The test results should show no sign of unduedeformation. Slight deformation may be acceptable,provided that the clearance and the creepage distancesare complied with and the degree of protection is notimpaired. The insulation and mounting supports, however,must show no signs of deterioration.For the tap-08s in a power-generating station,connecting a UAT through the main bus sectionbetween the generator and the generatortransformerAs discussed in Section 13.4.1(5), these sections are underthe cumulative influence of two power sources and maybe tested for a higher short-time rating, which would bethe algebraic sum of the two fault levels, one of thegenerator and the other of the generator transformer asnoted in Table 13.8. Also refer to Figures 3 1. I and 13.18for more clarity.For the ground busIn an isolated phase bus (IPB) system, the ground busmust be capable of carrying the same short-time currentas for the main conductors for 2 seconds, for bothdiscontinuous and continuous grounding systems.32.3.6 Verification of momentary peak ordynamic currentThe test method and test results will generally be thesame as for a switchgear assembly. discussed in Section14.3.7.For the tap-offs, connecting a UAT through the mainbus section between the generator and the generatortransformer. however, as discussed above, the momentarypeak current will depend upon the short-time rating ofsuch tap-offs. The likely ratings are noted in Table 13.8.32.3.7 Verification of protective circuitsAll the protective circuits will be checked for continuityand operational requirements.32.3.8 Verification of clearance and creepagedistancesThe clearance will be verified as in Table 28.4 while forcreepages Table 28.5 may be followed.32.3.9 Verification of degree of protectionEnclosure testThe types of protection and their degree are generallythe same as those defined for motors in Section 1.15.Tables I. 10 and I . 1 I . The test requirements and methodsof conducting such tests are also almost the same as forthose motors, discussed in Section 1 1 S.3.Watertightness testThis test is applicable to all outdoor parts of any bussystem and on both indoor and outdoor parts of an isolatedphase bus (IPB) system. Each enclosure to be tested(such as the enclosure of one phase in an isolated phasebus system) should be complete in all respects with allits fittings and mounts in place. The length of the testpiecemust be the same as that considered for the temperaturerise test (Section 32.3.4).As ANSI C-37120C and IS 8084 water under a head of1 1 m through a hose of 25 mm diameter held at 3 m fromthe enclosure will be impinged at an angle of at least 45"from the horizontal on all sides and the entire length ofthe enclosure for a period of 5 minutes. No water shouldenter the enclosure.Air leakage testThis test is prescribed by ANSI C-37/20C for isolatedphase bus systems after they have been assembled atsite. The test is conducted by filling the enclosure withair up to a pressure of 15 cm of water (1 500 N/m'). Afterthe air supply is shut off, the pressure must not drop toless than 7.5 cm of water in 15 seconds. All breathersand drain holes must be sealed before the test.32.3.10 Measurement of resistance andreactanceThe mean values of the resistance and reactance per phaseare determined on at least two trunking units. includingthe joints, from a total length of 6 m. If the total lengthof the bus section is less than this, the values may bedctcrmined on the entire length of the bus system.If Vdt, Vd2 and Vd, are the voltage drops in the threephases, then the average voltage drop'dl+Vd = vd2 volts per phase and3I, +I, +I, Amps per phaseaverage current I,, =3

~ ~~~32/958 Industrial Power Engineering and Applications Handbookwhere ZR, Zy, IB are the phase currents in the three phases:. Impedance, Z = -"' i2 per phase/rnIph 'and R=- Q per phaselm3Zih . LX = 4-andR per phase/mwhereP = total power input in wattsL = length of the bus section on which the voltage drophas been measured in metres32.3.11 Endurance of trunking system withtrolley-type tap-off facilitiesThis test is generally applicable to an LT overhead busbarsystem or rising mains which have sliding-type plug-inboxes for outgoing terminations (Figures 28.3 and 28.4(a)).The sliding contacts carrying their rated current at ratedvoltage must be able to operate at least 10 000 times toand fro with the trunking conductors. A successful testshould reveal no mechanical or electrical defect such aspitting, burning or welding of contacts.32.4 Routine testsThe tests against step numbers 1,2 and 3 are of a generalnature and no test procedure is prescribed for these. Theremainder have already been covered under type tests.The procedure for tests and the requirements of the testresults will remain the same as for the type tests,32.5 Field testsThese are generally the same as the routine tests. Nofield test as regards the power frequency voltage withstandor HV test is normally prescribed on a bus system, whichhas already been tested at the manufacturer's works.Repeated application of high voltage may deteriorate theinsulating properties of the insulation system and its lifeunless modification has been carried out at site or wherethe insulation of joints between the busbars can becompleted only after erection at site. In such cases, ifthis test becomes essential it can be conducted,At 85% of the test values in Table 32.1(A) for LTsystems, applied for 1 minute.Or at 80% of the test values for HT systems for1 minute.Or a d.c. voltage as in Table 32.1(B) applied for15 minutes as in IS 8084.Or at 75% of the test values in Table 32.2, for series I1voltage systems according to ANSI C-37/20C.NoteIf it is not possible to carry out the test at 75%, 80% or 85% of thetest values, as prescribed above or for some reasons the test durationmay have to be increased by more than one minute, then it ispermissible for the Lest voltage to be further reduced to suit the siteand instead, the test duration increased The reduced test voltagesfor longer test durations are provided in Table 14.8.Relevant StandardsIEC602981 1990 A.C. metal enclosed switchgear and controlgear for rated voltages 3427/1991 BS ENabove kV and up to and including 52 kV 60298/ 199660439-211987 Specification for low voltage switchgear and controlgear assemblies 8623-2/1993 BS EN 60439-Particular requirements for busbars trunking systems (busways)2/19936069411996 Common specifications for high voltage switchgear and 3427/ 1991 BS ENcontrolgear Standards 60694/1997Specification for high voltage busbars and busbar connections 8084- 1992 BS 159/1992Relevant US Standards ANSVNEMA and IEEENEMA 107/1993ANSMEEE-C-37-23/1992ANSI/IEEE-C37.24/1986Method of measurement of Radio Influence Voltage (RIV) of HV apparatusMetal enclosed bus and Guide for calculating losses in isolated phase bus.Guide for evaluating the effect of solar radiation on outdoor metal enclosed switchgearNotes1 In the tables of Relevant Standards in this book while the latest editions of the Standards are provided, it is possible that revised editionshave become available. With the advances of technology and/or its application, the updating of Standards is a continuous process bydifferent Standards organizations. It is therefore advisable that for more authentic references, readers should consult the relevantorganizations for the latest version of a Standard.2 Some of the BS or IS Standards mentioned against IEC may not be identical3 The year noted against each Standard may also refer to the year of its last amendment and not necessarily the year of publication.

Indexabbreviation\ uwd. 39 1a.c. tlrivcs compari\on. 138-502.c. solenoid braking. 152xcelcrating torque. 47. 48. I89actuator motors. 176-7ageing. winding insulation. 221agriculture motors. 175-6air:blast circuit breakers. 63s-7. 651-Sbreak contactors. 3 17-1 8circuit breaker.;. 634-7. 6S 1-Sconditioning plants, 387-90cored reactor\. 848-9handling units, 387-8leakage tests. 957AI-Cr alloy steel resistors, 86alarm schemes. 507. 508-9altitudc:bus systems. 871-2.induction motors. 16- 17witchgear assemblies. 362aluminium:busbar propertic\, 922-3current ratings. 915-16. 920-1curi-ent-carrying capacity. 9 16-24grades. 92 IINDAL. 867. 922-3properties. 915-16. 922-3rectangular sections. 920- Istandards. 925ambient temperatures:bus systcms. 869-7 Iinduction motors, 15-1 6AMF panels. 507-10ampere meter method. 833-4annunciation schemes. 507. 508-9ANSI standard device fiinctiori number\550- Iarc:ilswtance techniques. 632-3coils. 665-6formation. 636plasma theory. 63 Iquenchins. 610-1re-ignition. 566rntarion. 641-3-suppreshion coils, 665-0yuppre\\ion reactnrs. 852time cnn\tmt drfinition. 56.5arnioured cable\. 534. 541-3HT. 538LT. 535-6thrce-core, 538-9artificial neutral grounding. 668-9;i\semhlies wv switchgear assembliesassessment, surge arresters. 617-23atmospheric conditions, 362, 872auctioneering switching. 768auto-mains failure schemes. 506-7srr dso captive power generationauto-synchronizing schemes. 525. 526.527autniiialic capacitor control panels. 767automatic speed control. 95automation. logic controllers. 528-30autotransformer starting. 73-7auxiliary controls. 3367auxiliary feed lines. 929axial forces, bearings. 2 I2basc sclection, generators, 356-7basic insulation levels. 5934bearings:axial forces. 2 I2checking. 21 1-15cleaning. 236electric motors. 195-2 I7formulae, 217. 243greasing. 238high temperatures. 238-9installation. 2334life expectancy, 214-15load transmission, 195-2 I7load-carrying capacity, 214-15inaintenancc, 235-9I-adial forces. 2 I 1-1 2relubrication. 236-8\tandards, 242-3witability. 195. 21 1-15tcmperaturcs, 238-9. 255. 307thrust loads. 2 12bclt drives, 202-1 1V-belts, 206-10tlat. 20.1-6bending busbars. 9 I Ihi-metallic thermal relays. 284-5hi-phase chopper schemes, 122bimctallic joints. 910-1 1binary switching sequences. 769. 770biolocgical chemical demand (BOD). 415-17bipolar junction transistor control. 1 I2bipolar transi\toi-s (IGBTs). 113-14. I25blowing hot air. 036-7bolts. standards. 9 12honded enclo\ures. 933 -5brake presses. 66braking:electrodynamic. 15.34flux braking. I I Iformulae. 163internal types. 153-6magnetic particle braking. IS?--.?inechanical. 151-3incthod\, I5 1-6motor\. 97, I I I. 137-56regenerative. 155-6types. I5 1-6breakers. 3 13air. 3 I7control switches. 3x3fuse coordination. 316see u/.w circuit breakers: circuitinteri-upters: motor protectioncircuit breaker\brushes, standard\. 232bulk oil circuit breaker\. (7.32. 65 1-5bus joints. 903-1 2.see o h Ih\tenrrh: .jointsbus rystcms. 855-958altitude. 871-2ambient temperatures, 869-7 Iamospheric conditions. 872bar sections. 880-2clearances. 871-2conductor space Factors. 870configuration. 890creepage. 87 1-2derating. 870. 88&7design. 368-7 Iciilculation\. X9 1-900fasteners. 370parameters. 37 Iservice conditions. 370- Iducts. 860. 891electrodynamic effects. 869, 895cnclosurcs. 872. 93 1-12, 944-S. 949.957fasteners. 370. 897littinga, X77formulae. 902hcat disbipation. X73instrumentation. 929-30insulation. 955. 956

960 Indexbus systems (conrd)insulator suitability, 898-9interleaving phases, 888-9isolated phase systems, 861joints, 903-12LT systems, 860metal enclosed, 857-902nomograms, 866-8non-segregated, 859overhead, 862phases, 888-90power frequencies, 955, 956proximity effectq, 878-90, 886-90reactance, 879-82, 957-8reactors, 884-5resistance measurement, 957-8rising mains, 861-2saturated reactors, 884-5segregated, 85940seismic effects, 872service conditions, 370-1, 869-72short-circuit effects, 864-9skin effects, 873-8spacing, 888standards, 901-2thermal effects, 864-8transmission, 857-902transposition, 889-90types, 859-62vertical, 861-2vibration, 872voltage drop, 872-3voltage unbalance, 8824see also isolated phase bus systems;metal enclosed bus systemsbusbars:aluminium properties, 922-3bending, 91 Ichannel properties, 924configuration, 890enclosures, 890fittings, 877metal enclosed switchgear assembliestesting, 425-9mounting precautions, 905mounting systems, 368-9mountings, 897proximity effects, 879-82reactance, 879-84rectangular, 922sections, 877shrouds, 378, 379supports, 934switchgear assemblies testing, 425-9temperature testing, 427tubular properties, 923cables:armoured, 534, 541-3HT, 538LT, 535-6three-core, 538-9copper, 537derating factors, 544HT, 538insulation, 53142comparison, 532-3LT, 310,5354motors, 310multicore skin effects, 547optical fibre, 739paper insulated, 534. 539-41proximity effects, 547rating factors, 545-6selection, 314-17, 53149service conditions, 544short-time rating, 547-8skin effects, 547standards , 548-9technical details, 544temperatures, 545termination, 548three-core armoured, 538-9trenches, 3867voltage drop, 544-7XLPE three-core, 534cage induction motors see squirrel cageinduction motorscapacitative currents, 569-7 1capacitor voltage transformers (CVTs),4644capacitors, 723-854ampere meter method, 833-4application, 727automatic control panels, 767banks, 835HT systems. 780manufacture, 815-1 8behaviour, 725-75charging currents, 750-4charts, 761,762component rating/selection, 818-2 1control panels, 767current unbalance, 833dielectrics, 813direct switching induction motors, 76discharge devices, 8 19-24double star arrangement, 749element manufacture, 81 1-12excessive charging currents, 750-4field operating conditions, 841formulae, 774-5, 825-6, 843frequency variation, 834grounded star units, 729harmonic effects, 732-40HT maintenance, 837-8HT systems, 780induction motors, 760-3inductive interference, 73240inrush current limitation, 754-9installation, 837-8insulation levels, 840internally protected units, 812-14location, 763-5low PF effect, 727-8LT maintenance, 837-8maintenance, 82743manufacture, 809-26metallized polypropylene, 8 1 1mixed dielectric, 81 1-12operation, 72840panel design parameters, 759-60parallel switching, 73 1, 7524partial discharge tests, 841PCB-filled, 838performance, 7334PF improvement, 761,762power factor improvement, 725-75precautions, 838protection, 82743rating, 760-3, 818-21regulation improvement, 779restrike voltage, 730routine tests, 838-9safety, 829-37selection, 818-21self-healing type, 81 1, 8 I4series reactor compensation, 747-8single-capacitors switching, 751-2single-phase motors, 27-8, 30-1special purpose units, 818standards, 773, 824-5. 842star banks, 816-18starting motors, 27-8, 30-1surges, 729-32switching, 725-75, 768-9inrush, 750-4overvoltages, 729-32sequences, 768-9system voltage regulationimprovement, 779test requirements, 83841testing, 82743, 842, 843thermal stability. 839-40type tests, 83941ungrounded units, 730-1unit manufacture, 814-15unit standards, 824-5voltage regulation improvement, 779voltage unbalance, 831-2voltage variation, 834see also power capacitorscaptive power generation, 331-551diesel generators, 499-500driving torque, 516, 520I2 field systems, 502-3load sharing, 526, 527-8load types, 505-6operating parameters, 500-3operation procedures, 521-6operation theory, 503-4parallel operation, 5 15-26programmable logic controllers, 528-30standards, 530synchronizing schemes, 526-7card motors, 167, 168carrier equipment, 738cascade failure, 812cast iron grid resistance, 86CBCT see core-balanced currenttransformerscentral processing units. 339certification:locations, 266-9machines, 266-9certifying authorities, 183channcl busbar properties, 924charging currents. capacitors, 750-4chemical oxygen demand (COD), 417chemical resistance starting, 78chemical vapour classification, 179chokes see saturated reactorscircle diagrams, 18-20, 157circuit breakers, 317air blast, 635-7air type, 634-5bulk oil, 632ground leakage. 679-8 1

minimum oil. 633-4operating conditions. 63 I-?.wlphur hexatlouride gas. 63743cwv-pole. ox Iiiicuutn. 643-6v w ci/.w breakers: circuit interrupters:motor protection circuitbreakersc i rc ti i t i lit crruprer5. 627-5 6i\tance techniques. 642-3arc extinction. 640-1comparisons, 650-5cmxmt c hoppi 118. h4GXdifierent mediums. 632-46inediunis. 632-46puffer technique. 640-1\tandai-ds. 656\witch closure>. 629\witching mediuni\. 629-30t heory. 629-30thermal blast techniques. 642-3circuits \cc pow’er circuitscirculation theory. 711. 742c.las\ificatioii:chcinical \apour. I79pses. 170surge mcsterc. 0 13-1 5I:HV \ystem. 736volatile liquid. 179\ oltagl‘ \y\tem. 730cleaning bcarings. 236clcaning prior to painting. 100 Iclearances. 434. 87 I-?closing coil. 383CMR sev continuoils duty niotorscoaling.;, 400-4. 409COD scv chemical oxygcn demandcolliery motors. 187commissioning motor checks. 234c(jiiiiiiuiiiciiti[)ii service disturbances. 73(i-9commutation. I I6components:electric motors. 3 I8niotors. 3 I?- I7ratings. 3 18sclection for motor\. 3 12- I7short-time rating guidelines. 367condenser\. Fenerators. 501 -2conductivity. 7 IO. 0 I6conductors:electrical propcrtic\. 9 IS- I6iwlated phase bus \ystems, 944-5mechanical properties. 915metals. 701-3ininimum sile. X66niouiiting precautions. 905physical properties. 9 15properties. 905. 913. 915-16proximity effects. X73. 878-00ratings. 9 13 25scctions. 8x0-2. 934-5size determinaiion. 941-5skin effects. 873-8. 938-9space factors. 870,\w cilso aluminium. copper: silverconnection\:currcnt transformer\. 490- Imultispeed motors. 100reactors. 85 Iconnectors ,see jointscon\tant tilled fluid couplings. 19X. 190.200constant torque speed control. 101-3constants, metals. 7 12contact bouncing. 57 1-2, 576-7contact resistance. soil. 7 IOcwitactors:air brcak. 317-18duty cycles. 3 I7motor ctimponent selection. 3 12- I3\;icLIuIll. 31 1-12. 315continuous current capacity. 954continuous current ratings. 345-6. 371-2continuous duty motors (CMR). 5 I . 53-5continuous enclosures see cnclosurescontinuous operating voltage. maximum.593continuous rating, 942control:a\\ernbly types centres, 336niotors. 1-240d.c. link \moothing. 128-9bipolar junction transistors, I 12constant d.c. voltage source. 129current sources. I79cyclo converters. I27Ditrlington transistors. I 12-1 3dircct torquc. 108- I Ienergy conservation, 131-8feedback devices. I I1frequency converters. I27gate turn-off rwitche\. 1 IS. 125induction motors. 81-96insulated gate bipolar transistors.113-14inverter units. I IO. 1 19-25MOSFETs. 113rectifier unit configurations. 120-1regenerative schemes. 127-8shrouding centres, 374-5\ilicon rectifiers, 115-16slip recovery systems. 140-2slip-ring induction motors. 81-96soft starting. I3840\olid-state technology. I 11-16\piice heater\ centres. .374-5speed variation, I45static. 97- I64static drives, 138-45\toppins. 140switchgcar assembly centrcs. 37h-7thyristors. 114-16torque. 108-1 1p;rnels. 337-9. 507-9. 759-60. 767rectiticr unit configuration\. 120-1silicon rectifiers. I 15-1 6slip-ring induction motors, X 1-96solid-state technology. I I 1-165tiitic. 97-164\upply schemes, 3x3. 384, 385switchgcar assemblies diagrams, 387-07thyristors. I 14-16\oltage transformer schemes, 168.sc’c O/S(J motor control centre\control triinsfbrniers. 4.55. 406-9. 49 1-4control type couplitip ICY-20 Icontrolgear awmhlies. 3334 I8drawout chassis centres. 376-7installation. 384--7niaintcnance, 3x4-7painting, 400- I7conversion formulae. 374conversion table%. .329convertcr unit$. I 17.- 19, 132-3cooli tip:air. 255curves, 56-8induction motor\. Ih. 24-7. 5h-Xisolated phase hu\ \! \tenis. 042.;y.;tems. 16. 24-7water teniperaturc. IOcopper:Index 961cable current ratin?. -371-2current ratings. 371-2. 915-lh. 910current-carrying capacity. 9 16-34electrical propertic\. 9 17- I8mechanical properties. 9 I7physical propertic\. 9 I7propertic\. 9 15- I Sctandard\. 02.5core losses. 18core-balanced current transformer\(CBCT). 4x8. 685-8design parameters. 0x54grwnd leakage cui-i-cnts. OX7insulation. 686mounting. 686-Xwiring incthods. 6Xhcoreles5 i-ciictor\. 848 9corrcction:V-belt factors. 2 I Ipower factors. 765-72relay\. 169-72coupling\:carrier equiptneiir. 73Xconstant tilled Iluid. 198. 199. 200construction. 108-202control type. 199-201delaqed action. 107-8dircct. 197eddy current couplings. 202. 20.;environincnt-friend14. 198tlcxible. I97fluid couplings. IOX-201dc. drivc comparisons. 1 -4X-~SOcomparisons. 148-50static drivc comparison. 115-6hydi-aulic, 198-201linc devices. 738magnetic. 202operation principle\. 198-202rigid. 197speed control. 1-45speed \iiriati(iii. 145traction type, 199variable speed fluid couplings. 14s.148-50. 201v~i~i;iblc~tillinp. I9‘)-71)1criiiic ni(otoi-\. l6X-9creepage. 433. 87 I-?critical installations. I83--5CSI .see current sourcc insertersCT operated thermal o\ el-current iclaqs286Curie point\. 303-4

962 Indexcuring paint, 410current chopping, 646-8current limiting reactors, 85&2current polarization, 691-2current ratings:aluminium, 915-16, 920-1copper, 915-16.919copper cables, 371-2isolated phase bus systems, 931PVC insulated cables, 371-2single core cables, 371-2switchgear assemblies, 345-50current release protection, 681-4current source inverters (CSI), 126-7current sources, motor control, 129current toleration, human body, 679current transformcrs, 470-9 1circuit diagrams, 474connection precautions, 490-1core-balanced, 488, 685-8current error, 474design, 471-5.685-6differential protection schemes, 479-83, 487error limits, 475external fault conditions, 482general specification, 471-5ground fault protection schemes, 688high impedance protection schemes,479-83indoor, 472-3insulation, 470-1, 686interposing, 476limits of error, 475measuring, 4754mounting, 686-8outdoor types, 472protection, 471, 477-83, 487, 688PS type, 479-88ratio error, 474relays, 483-4routine tests, 494-6special-purpose, 479-88specifications, 471-5, 492stability, 482summation, 476-7system harmonics, 485type tests, 493-4VA burden, 473-4zone faults, 483current unbalance, capacitors, 833current zero, definition, 565current-carrying capacity, 9 16-24current-carrying conductors, properties/ratings, 913-25current-carrying equipment, ratings, 364-8CVTs see capacitor voltage transformerscyclo converters, 127dampening, 135reactors, 852damping, 446-7Darlington transistors, 112-1 3d.c. drives:fluid coupling comparisons, 148-50motors, 147performance comparisons, 148-50variable speed comparisons, 148-50d.c. link smoothing, 128-9d.c. voltage sources, 129de-rusting, 402, 403deep-sea mining, 175, 176deep-well pump motors, 170-5definitions, 560-5, 915degreasing, 400-1degree of protection verification, 264-5deionization theory, 629-30delay tripping, 364-5delayed action couplings, 197-8delta to star conversion, 758derating:bus systems, 870, 886-7cable factors, 544proximity effects, 886-7design:2500A non-isolated phase bus systems,89 1-900bus systems, 368-71calculations, 89 1-900metal enclosed, 8624, 872-3core-balanced current transformers,685-6current transformers, 471-5, 685-6grounding grids, 7 12-19grounding stations, 706-7metal enclosed bus systems, 862-4,872-3reactors, 848-50switchgear assemblies, 342-64, 37 I-83voltage transformers, 457-69dielectrics:applications, 813capacitors, 811-12loss factors, 260property verification, 259-6 I, 422-4test voltage verification, 260diesel generators, 499-500annunciation schemes, 507, 508-9electrical protection, 507-14metering, 507-14power diagrams, 51 Ipower factors, 504-5protection, 507-14selection, 504-5starting, 506-7differential protection schemes, 479-83,487diodes, 112direct couplings, 197direct on-line starting (DOL), 71-2direct switching induction motors, 762direct torque control (DTC), 108-1 1directional ground fault protection, 690-2discharge current, 595discharge devices:capacitors, 819-24resistance, 822-4series inductors, 823-4discontinuous enclosures see noncontinuousenclosuresdiscrete constant loads, 55-6distribution networks:layouts, 347surge arresters, 606-7, 614-15typical, 727diversity factors, 346, 366-7DOL see direct on-line startingdomestic single-phase system protection,679-8 Idouble dielectric process, 81 1double squirrel cage motors, 3840see also squirrel cage motorsdouble star arrangements, capacitors, 749draw-out components, 376-7, 386draw-out switchgear assemblies, 341-2dripping water ingress protection. 24. 267driving torquc, 516, 520drying enclosures, 936-7ducts, 860, 891, 930dust protection, 265duties, induction motors, 49-68duty cycles, contactors/switches, 3 17duty rating, slip-ring induction motors, 90dynamic currents:metal enclosed switchgear assemblies,432-4testing, 957transformers, 488-90earth. cross-section construction, 437earth leakage circuit breakers see groundleakage circuit.. .earthquakes:causes, 437duration, 439engineering testing methods, 436-52metal enclosed switchgear assemblies,436-52preventative measures, 452recording, 43943response spectrum, 439-41time history, 439, 440TPC a1.w seismic effectseddy current couplings, 202,203effective magnitude, harmonic currents/voltages, 7414effectively grounded systems see solidneutral grounding systemsefficiency, electrical machines, 18-20, 256effluent treatment:painting, 400, 412-17passivation, 4 17physio-chemical analysis, 412-13plant flow diagram, 415plant layout, 416waste water, 412-15EHV systems:communication service disturbances,736-9grounding, 668-9line parameters, 786electric motors see motorselectrical interlocks, 378-83electricaI machines:certification, 266-9degree of protection verification, 264-5dielectric property verification, 25941efficiency, 256hazardous locations, 266-9impulse voltage withstand tests, 261-3load tests, 256-63locked rotor tests, 263-4measurements, 25 1no load tests, 263-6noise level measurement, 259overspeed tests, 257power factors measurement, 257

power frequency voltage tests. 262protection verification, 264-5quality system,, 245-72resistance measurement. 252-3rotor tests, 263-4routine tests, 252seismic disturbances, 252slip measurement, 256-7speed-torque tests. 257-8standards. 242-3, 269-7 Itemperatures, 253-5test tolerances, 266testing. 245. 25C69type tests, 25 1-2\ ibration measurement tests, 258-9boltage tests. 262water ingress protection, 266-8see dso electric motors; HT & LTmotors: induction motors; slipringinduction motorselectrical properties:conductors, 915-18copper. 917- 18electrical protection see protectionelectrode electrolytes. 94-5clcctrodynamic:braking, 153-4effects, 869forces, 895stresses. 818-19electrolyte switching. 78electromagnetic:induced currents, 559relays. 294-5shoe brakes. 152transference. 601voltage transformers, 459-60electronic circuits, 740-1see also thyristorselectroplating fasteners. 370electrostatic:induction, 559, 736-7powder painting. 409-1 2transference, 599embedded temperature detectors (ETD).254. 302. 30.5emergency power generation see captivepower generationenclosures:bonded. 933-5bus systems. 872, 890.93142. 944-5.949.957continuous, 933-5drying, 936-7end terminations, 931-2formulae. 949-50heating, 938-4 1induction motors. 20- Iinsulation. 932-3, 934IPB systems, 93142. 944-5. 949, 957joints. 931-2magnetic fields, 915natural cooling. 941-2NEMA. 21-2non-continuous, 932-3nullifying magnetic fields. 935section determination, 944-5size, 872, 944-5standards. 949tests. 957thickness, 93940end terminations. 931-2energy absorption, inductors. 132energy balancing theory, 648-50energy capability, surge arresters, 612-13.62 1energy conservation, 134-8environment-friendly couplings. 198epicentres, 443ETD see embedded temperature detectorseutectic alloy relays, 286-7Ex ‘i’ safe circuits. 182-3Ex ‘e’ motors, 18CIEx ’n’ motors, 182Ex ‘p’ motors, 181-2expansion joints, 909explosion proofmotors. 180switchgear assemblies, 362-3factor of inertia, 56factor r values, 602fan power requirements, 323fan-cooled motors, 2 Ifast discharge devices. 819-24fast responding relays, 583fasteners, 370bus systems, 897draw-out components, 386torque. 907fault conditions, 279-83. 482fault currents:generators, 355, 673power systems, 348fault levels:generator circuits, 352-7HT systems, 351-62isolated phase bus systems, 942-3LT systems, 350-6power distribution circuit diagrams,366power generating stations, 353feedback devices, control, I I1feeders, interlocks, 379-83Ferranti effect, 790-2, 796-8ferro-resonance effects, 465, 557-8, 662,78 Ifield excitation, 5 18field operating conditions. capacitors, 84 Ifield systems. 502-3field tests, 422metal enclosed bus systems. 954, 958metal enclosed switchgear assemblies,435-6field-oriented control (FOC). 106-8filter circuits. 745-50filter reactors, 852final coats of paint, 405-8first in first out switching. 768first in last out switching, 768fishplates, 908fittings. busbar sections. 877five-step stator-rotor starters, 89fixed type switchgear assemblies, 340-1flame-proof equipment, 180, 362-3flat belt drives. 204-6formulae, 2 17power ratings, 205standards. 216Index 963flexible couplings. I97flexible joints, 909-10floor response spectra, 44 I . 445fluid couplings, 198-201d.c. drive comparisons. 148-50speed variation. 145static drive comparison, 135-6flux-oriented phasor control. 109. I 1 Iflywheels. 65-6FOC see field-oriented controlforced cooling systems. 942forming sections. 943-4formulae:V-belts. 217bearings, 217. 243braking. 163bus systems. YO2capacitors, 825-6. 843electric motors, 48. 3 19-20enclosures. 949-50flat belts, 217ground fault protection. 693grounding practices. 720-1induction motor\. 34, 67-8. 79. 96insulation, 229moment of inertia. 324overvoltages. 674-5painting. 4 I 8power capacitors. 774-5slip-ring induction motors. 96special-purpose motors. I93squirrel cage induction motors. 79static control, 163surge arresters, 624-5switching devices. 825-6system grounding. 674-5system voltage regulation. 808temperature conversion. 324transformers. 496voltage surges. 584-5four-wire three-phase system\. 6x3-4frame sizes, 8-9free suspension vibration measurementtests, 259free vibration test. 448frequency:converters. 127limits, 10-17variations, 10. 13-15, 834friction losses, pipes. 325-8fume control. painting, 412-17function numbers. ANSI, 550-1fuse-free systems. 3 17-1 8fuses:breaker coordination. 3 I6characteristics. 83 1HRC, 288. 289. 310-11. 314. 318HT motors. 310-1 ILT motors, 310selection. 83 Igalvanizing processes, 146-7gapless surge arresters, 590- Icharacteristics. 616protective characteristics. 620selection. 605-1 3station class, 620TOV capability. 605-9gapped iron core reactors, 849-SO

964 Indexgapped surge arresters, 589-90gas:circuit breakers, 637-43, 651-5classification, I79gate bipolar transistors (IGBTs), 113-14,125gate turn-off switches (GTOs), 115, 116,I25generating stations:auxiliary services, 358-9parallel operation theory, 515-21protection schemes, 512-13generators:circuits, 352-7as condensers, 501-2fault currents, 355, 673fault levels, 352-7grounding, 671-4induction, 156-61selection of circuits, 3567stability levels, 788as synchronous motors, 501-2wind-power, 157-61see nlso diesel generatorsGFF see ground fault factorsgland plates, 372, 373grades, aluminium alloys, 921greasing bearings, 238grid conductor sizes, 712, 713, 715grid resistors, 86ground conductors, 705, 715-19ground fault factors (GFF), 666-7ground fault neutralizers, 665-6ground fault protection schemes, 677-93current determination, 710-1 1current sharing, 715-16directional, 690-2formulae, 693hazardous areas, 684-5LT systems, 681-4parameter selection, 669-73relay current settings, 692restricted, 689-90selection parameters, 669-73single current transformers, 688standards, 693three-phase four-wire systems, 6834unrestricted, 690ground faults, voltage transformers, 462,463ground leakage:CBCT currents, 687circuit breakers, 679-8 1HI systems, 685ground potential rise, 705-6ground resistance, 701-2, 709-10grounded systems, 660-5grounding:application of different methods, 668-9arrangements, 375-6artificial, 668-9conductors, 702-3, 704different methods, application, 668-9EHV systems, 668-9electrodes, 697-704generators, 671-4HT systems. 668-9insulation, 667-8isolated phase bus systems, 935LT systems, 668-9necessity, 663neutral, 668-9practices, 553-722systems, 657-75theory, 677-93transformers, 852grounding grids:corrosion factors, 7 12design, 712-19step/touch voltages, 712-14grounding practices, 695-722conductor metals, 702-3formulae, 720-1grid design, 712-19ground fault currents, 710-1 1ground resistance, 709-10industrial installation, 697-704location potential differences, 704method selection, 699-701power generation, 704-19standards, 720voltage gradients, 704-7grounding stations:design parameters, 706-7maintenance, 704safe potential differences, 717-19step/touch voltages, 712-14GTOs see gate turn-off switcheshand-held sprayers, 268hardness tests, 408-9harmonics:appearance, 744-5capacitors, 73240performance, 7334circulation theory, 74 1, 742current magnitude, 7414electronic circuits, 740-1filter circuits, 745-50generation, 740-2networks, 745-50nullifying effects, 748-50order magnitude, 743phase-controlled rectifier units, 130power electronic circuits, 740-1power networks, 745-50protection, 275quantity circulation theory, 741, 742resonance, 74 Istatic device switching circuits, 129-30suppression, 745-50systems, 9-10, 485telephone lines, 734triple quantity circulation, 741, 742voltage magnitude, 7414hazardous locations:certification, 266-9classification, 179electrical machines, 266-9ground fault protection, 684-5motors. 178-9head loss, pipelines, 325-7heat:bus systems, 874dissipation factors, 874losses, 425-6metal enclosed switchgear assemblies,425-6rotors, 279, 281-2run tests, 427sinks. 294-5stators, 278, 281switchgear assemblies, 425-6heating:enclosures, 93841induction motors, 56-8motor heating, 43-6high impedance differential protectionschemes, 479-83high impedance relays, 483-4hold-on values, 468hose test nozzles, 269hot air blowing, 936-7hot dip galvanizing line. 144-5HRC fuses, 288, 310-11, 314, 318, 6814HT motors, 287-8, 308-12failures, 241-2formed coils, 227-8fuses, 310-11impregnation, 2234insulation, 222-3, 241-2protection, 287-8, 308-1 2squirrel cage type, 75-6starters, 309surge impedance, 563switchgear protection, 308-12vacuum pressure impregnation, 223-4HT systems:capacitor banks, 780capacitor protection, 836classification, 736fault levels, 35742ground leakage, 685grounding, 668-9isolators, 309-10resin-rich coils, 26 1switchgear assemblies, 3834terminal box dcsign, 184, 185see also HT motorshuman body:current toleration, 679ground conductor contact, 705leakage current determination, 707-10resistance, 707-9HV systems:classification, 736communication service disturbances,736-9line parameters, 786hydraulic couplings, 198-201hypocentres, 443I-phi characteristics, 848-50IGBTs see insulated gate bipolartransistorsimpedance grounding, 664-5.672-3impregnation, HT motors, 2234impulse definition, 560-1impulse levels, 344-5impulsc voltagc withstand:metal enclosed bus systems, 954, 955,956specification framing of tests, 262-3tests, 261-3, 424inching motors, 161INDAL aluminium, 867, 922-3indoor current transformers, 472-3

induction generators. 156-6 1wind-power. 157-6 Iinduction motors. 3-34altitude. 16-17ambient temperature. IS- 16capacitor rating, 760-3circle diagrams, I 8-20constructional features, 3-31continuous duty. 5 I. 53-5control. 81-96cooling systems. 16. 24-7. 568discrete constant loads. 55-6duties. 49-68C).ClC

966 Indexlimits of error, current transformers, 475lines:characteristics, 787coupling devices, 738lengths, 79C2, 796system voltage regulation, 785-9, 790-2telephone, 734transmission, 785-9, 790-2liquid rotor starters, 95load angles, reactive control, 794-802loads:bearings, 195-217, 214-15effects, 17-1 8power factors, 18sharing, 526, 527-8tests, 256-63torque, 35transfers, 796transmission, 195-217types, 505-6locked rotors, 2634, 291loom motors, 167, 168loss factors, 260LT motors, 308cables, 287, 310fuses, 310insulation systems, 222relays, 310squirrel cage type, 75switches, 310LT systems:air circuit breakers, 635bus systems, 860cables, 535-6current release protection, 681-4fault levels, 350-6ground faults, 6814grounding, 668-9HRC fuse protection, 681-4switching surges, 571see also LT motorsmachines:quality systems, 245-72ratings, 575-6 .testing, 245, 25C69voltage surges, 575-6see also electrical machinesmagnetic characteristics, reactors, 847magnetic couplings, 202magnetic fields, enclosures, 935magnetic particle braking, 152-3maintenance:bearings, 235-9capacitor units, 82743checks on motors, 234-5controlgear assemblies, 384-7electric motors, 231-9grounding stations, 704HT capacitor units, 837-8insulation, 224-7LT capacitor units, 837-8switchgear assemblies, 384-7winding insulation, 219-29manufacture:capacitors, 809-26practices, 372-6switchgear assemblies, 372-6material constants, 712maximum continuous operating voltage,593MCCs see motor control centresmeasurement:electrical machines, 25 Iground resistance, 701-2power factors, 257measuring current transformers, 475-6measuring voltage transformers, 458mechanical:braking, 151-3HT & LT operation verification, 435interlocks, 378-80properties of conductors, 915-17mesh voltage, 705metal enclosed bus systems, 857-902continuous current capacity, 954design, 862-4, 872-3dynamic current testing, 957field tests, 954, 958frequency voltages, 955, 956impulse voltage withstand, 955, 956insulation levels, 955, 956leakage current measurement, 954minimum sizes, 866momentary peaks, 957power frequency voltages, 955, 956protection, 957rated continuous current capacity, 954reactance measurement, 957-8recommended tests, 9534resistance measurement, 957-8routine tests, 953-4, 958seismic disturbances, 954service conditions, 862-4, R69-72short-circuit effects, 864-9solar radiation effects, 957standards, 958temperatures, 954, 956testing, 951-8thermal effects, 864-8trunking system endurance, 958type tests, 953, 954-8types, 859-62see also bus systemsmetal enclosed switchgear assemblies:busbar testing, 425-9dynamic current, 4324earthquake engineering, 436-52field tests, 435-6heat losses, 425-6heat run tests, 427impulse voltage withstand test, 424insulation, 423momentary peaks, 432-4quality systems philosophy, 421recommended test, 421-2routine tests, 421, 435seismic effects, 436-9short-circuit strength, 429-32standards, 452-3temperature rise limits, 424-9testing, 419-54type tests procedure, 422-35see also switchgear assembliesmetal properties, 712, 915metallized polypropylene capacitors, 8 I 1metals:constants, 712grounding conductors, 702-3metering, diesel generators, 507-14microprocessor-based relays, 296-9, 771-2mid-point compensation, 791mine motors, 182minimum oil circuit breakers, 6334, 651-5mixed dielectric capacitors, 811-12moisture, 307-8, 710moment of inertia formula, 324momentary peaks:current, 488-90metal enclosed assemblies, 4324testing, 957monitoring:insulation, 227-8operating conditions, 305-8surge arresters, 618MOS controlled thyristors, 114MOSFETs control, 113motor control centres (MCCs):assembly types, 336draw-out chassis, 376-7short-circuit tests, 433shrouding, 374-5space heaters, 374-5switchgear assemblies, 376-7motors:accelerating torque, 47, 48actuator motors, 176-7agriculture, 175-6bearings, 195-217braking, 97, I1 1, 147-56cable selection, 314-17card motors, 167, 168centrifuges, 169-70collieries, 182commissioning checks, 234components, 312-17, 318control, 1-240d.c. link smoothing, 128-9bipolar junction transistors, 112constant d.c. voltage source, 129current sources, 129cyclo converters. 127Darlington transistors, 112-13direct torque, 108-1 1energy conservation, 134-8feedback dcvices, 111frequency converters, 127gate turn-off switches, 115, 125induction motors, 81-96insulated gate bipolar transistors,113-14inverter units, 119-25MOSFETs, 113panels, 337-9,507-9,75940,767rectifier unit configurations, 120-1regenerative schemes, 127-8silicon rectifiers, 115-16slip recovery systems, 140-2slip-ring induction motors, 81-96soft starting, 13840solid-state technology, 111-16speed variation, 145static, 97- 164static drives, 13845stopping, 140thyristors, 114-16

torque. 108- I Iturn-off switches. I 15. 125voltage sources. I29crane\. 168-9critical installations. I 83-5deep-well pumps. 170-5d.c drives. 147direct torque control, 108-1 1duties. 19-68explo\ion-proof. I80fault condition\. 279-83field-oriented control. 106-8flame-proof, I80tlux braking. I I 1flux-oriented phasor control, I09tnrmulae, 48. 3 19-20generators. I 56-6 Ihaiardous location\. 178-9heating. 43-6inching. 161increased safety. I XO-Iinduction. 3-34. 49-68induction generators, 156-6 Iinstallation, 23 1-9IS specifications. 250logging. 161load torque. 35. 11lockcd rotors. 29 Iloom motors, 167. 16Xmaintenance, 23 1-9rnill centrifuges. 169-70mines. I82monitoring operating conditions, 3058no-load start-up. 441101se. 177-8non-sparking, I82on-load \tart-up. 44opposing torque. 41phasor control. 103- I Ipower stations. 186-9pressurized enclosures, I8 1-2problem\ solved. 239-41protection. 1-240. 273-320circuit hrrakcrs. 1 I 8semiconductor device\. I304witchgear. 308-12pullcy loads, 211pumps, 170-5quarries. I82ratings. 3 1 Xre-acceleration. I 84-5. I 87I-ecovcry voltage. 572-7remedies to problems. 239-41ring frame, 167-8\at+. I 80-1safety devices. 306selection. 1-240\einiconductor device\. I 16-28\ervice factors. I78\haft current\, 235-8\ingle phasing. 280-1. 29 1-2\ingle phasor/vector control, 104-6\lip-ring induction type, 8 1-06\pecial-purpose, 165-94\peed sensors. 11 I\peed-torque curves. 37\pinning frame motor\, 167-8stalled conditions. 45-6\tallins protcction, 29 Istandards. 48. 162. 3 I9start-up. 43-4starting current effects, 40-1starting frequency. 161static control, 97-164\tator currents, 277steep-fronted TRVs. 572-7stopping frequency, 16 Isugar mill centrifuges, 169-70surface-cooled, 176\urge arresters. 615-17testing, 1-240textile. 167-8thermal power stations, 186-9thermal withstand time, 44-8thermistor protection, 302-5time of start-up, 41-4torque. 35-45total protection, 299-302transient recovery voltage. 572-7triacs, I16TRVs. 572-7turbine pumps, 170-1unfavourable conditions. 275-9bector control, 103-1 1vibration, 177-8water-cooled, 307Zone specifications. 180-2see (i[.so electric motors; HT & LTmotors; slip-ring inductionmotors; squirrel cage inductionnintorsmoulding compounds, 369mounting:busbars, 368. 897, 903-1 2conductors. 905current transformers, 686-8induction motors, 20insulation. 905insulators. 905practice\, 903-1 2transformers. 686-8types, 2 Imok ing electrode electrolyte controllers/\tartcs\, 94-5multicore cables, 547multispeed motors connection diagrams.I00National Electrical ManufacturersAssociation (NEMA), 21-2, 37-8.41natural cooling, 941-2natural frequency of vibration. 445-6NEMA see National ElectricalManufacturers Associationnetworks. harmonics. 745-5011eutl-a~ c

968 Indexpeelable coatings, 409per unit voltage, definition, 565performance:d.c. drives, 148-50a.c. drives, 148-50induction motors, 3, 9-17service conditions, 9-17time of start-up, 414tolerances, 269periodic duties, 5 I 4periodic speed change, 54-5Petersen coil reactors, 852PF (or p.f.) see power factorsphase bus systems. segregated, 859-60phase enclosure shorting, 930, 931phase interleaving, 888-9phase spacing, 888phase transposition, 889-90phasor control, 103-1 1diagrams, 459, 460direct torque control, 108-1 1tield-oriented control, 106-8flux braking, 1 I Iflux-oriented. 109motors, 103-1 1rectifier units, 130single vector control, 104-6torque control, 108-1 1philosophy of quality systems, 245,247-9.953phosphating. 4004.417physical properties, conductors, 915, 917physio-chemical analysis, 412-13pickling, 401, 403pipes:friction losses, 323, 325-8grounding, 697-8pitch lengths, V-belts, 213plant:effluent treatment, 415, 416flow diagrams, 415layouts, 416plastics, 369plate grounding, 697polarization index, 225polyphase induction motors. 5-8powder painting, 4W-12power cables see cablespower capacitors see capacitorspower circuits:short-circuit tests, 429switchgear assemblies, 387-97power control centres (PCCs), 335-6power Darlingtons, 112-13power diodes, 1 12power distribution, 335-6circuit diagrams. 358-9fault levels, 366galvanizing processes, 146-7panel short-circuit tests, 360-1panels, 430short-circuit tests, 360-1power factors (p.f.):automatic correction, 765-8benefits of improving, 728capacitor rating improvement, 761,762correction, 765-72diesel generators, 504-5electrical machines, 18, 257improvement system, 835industries, 766load effects, 18measurement, 257power capacitors, 725-75reactive control, 795-6relays, 769-72power frequencies:bus systems, 955, 956metal enclosed bus systems, 955, 956switchgear assemblies, 343tests, 262voltages, 262, 955, 956power generation, 331-55 1circuit diagrams, 358-9grounding practices, 704-19station fault levels, 353surge arresters, 606-7see also generator circuitspower networks, 745-50power ratings, 205, 208power reactors see reactnrspower saving, 138power stations, 241-2, 508-10power system fault currents, 348power transfer, 789-90, 792-802pre-treatment non-ferrous components.404pressure relief facilities, 6 17pressurization systems. 937pressurized enclosures, I 81-2priming paint, 406process monitoring, 336-7process plant operation, 1424programmable logic controllers (PLCs),337-9, 528-30protection:analysis, 62 1-3bus systems, 929-30, 957capacitorsbank, 835units, 827-43contact bouncing, 576-7current transformers, 47 1. 477-83,487, 688diesel generators, 507-14domestic single-phase systems, 679-8 Idust, 265electric motors, 1-240. 273-320equipment, 835-6gapless surge arrester characteristics,620generation stations, 5 12-13grounding faults, 677-93harmonic effects, 275induction motors, 20-1. 224industrial single-phase systems, 67981isolated phase bus systems. 929-30locked rotors, 29 Imargins, 594-5metal enclosed bus systems, 957motors, 1-240, 273-320parameters, 669-73power stations, 508-10relays, 682-4schemes, 669-73selection of schemes, 669-73selection of surge arresters, 609-12series capacitors. 834-7shunt capacitors, 829-34single current transformers. 688single-phase systems, 291-2, 679-8 1stalling, 291surge arresters, 596-605analysis, 621-3levels, 596-605, 609-12margins, 594-5selection, 609-12surge transferences, 576-7switchgear, 308-12switching surges, 576-7transformers, 688unfavourable conditions, 283-7verification, 264-5voltage surges, 292-3voltage transformers, 458voltage unbalance, 291water ingress, 266-8protective circuit verification. 434proximity effects:bus systems, 886-90busbar reactance, 879-82conductors, 873, 878-90derating, 886-7isolated phase bus systems, 940minimizing, 888-Y0multicore cables, 547voltage unbalance, 882-6PS type current transformers, 479-88PTC thermistors, 304, 306-7puffer technique, 640-1pull-out torque tests, 258pull-up torque tests, 258pulleys:drive rotor suitability, 215-16installation, 2334motor loads, 214tolerances, 233-4pulse encoders, I 1 Ipump motors, 170-5pump power requirements. 323punched grid resistors, 86PVC insulated cables. 371-2quality assurance, 42 1, 953quality systems, 953electrical machines, 245-72formation, 247-8IS0 9000,247-9machines, 245-72philosophy, 245, 247-9, 953quarries, 182quenching plates, 636-7radial forces, bearings, 2 1 1-1 2rated continuous current capacity, 954rating:cables, 545-6capacitors, 818-19, 820-1conductors, 9 13-25current-carrying conductors, 9 13-25duty, 90fxtors, 545-6induction motors, 8-9, 59-63short-time motors, 59-63slip-ring induction motors, 90switching devices. 809-26

atings. switchgear assemblies. 345-50,364-8. 371-2re-acceleration of motors, 184-5. 187re-strikes. 566reactance:bus systems, 8x34. 957-8measurement. 957-8reactive banks, 803-7reactive control:line length effect, 794load angles. 794-802power factors. 795-6power transfer optimization. 792 802qymmetrical lines, 798reactive power management. 783-90mid-point compensation, 79 Iobjectives. 784-5power transfer, 789-90series inductance, 757transmission lines, 785-9uncompensated transmission lines.785-9rcactors. 845-53I-phi characteristics, 848-50air cored. 848-9application, 850-3arc suppression, 852connections, 85 1coreless. 848-9couplers. 852current limiting, 850-2dampening, 852design criterion. 848-50gapped iron core type, 849-50grounding transformers, 852magnetic characteristics. 847neutral grounding, 852saturated types. 849-50. 884-5selection, 847series, 85&2shunt, 850smoothing, 852-3standards, 853types, 848-50receiving-end voltage, 799-801recording earthquakes, 43943rectangular busbars. 920-1, 922rectifiers:comparison. 120converter circuits, 120semiconductor devices. 117-19thyristors. 120unit configurations, 120-1reduced voltage starting, 72-8reflections. surge arresters, 621-3regenerative schemes, 127-8, 155-6relays:bi-metallic thermal, 284-5electromagnetic, 294-5fwt-responding, 583heavy duty. 285LT motors. 3 IOmicroprocessor-based, 296-9. 77 1-2overcurrent settings. 285-6overcurrent type, 283,284, 286protection, 293-9, 682-4\ingle-device motor protection, 293-9holid-htatc. 29.5-6thermal overcurrent Iype. 283. 284,286relubrication of bearings, 236-8required response spectrum (RRS), 441-4.450-1residual current circuit breakers seeground leakage ...residual voltage transformers (RVTs).460-4resin-rich coils, 261resistance:bus systems measurement, 957-8discharge devices, 822-4electrical machines measurement, 252-3human body, 707-9insulation tests, 225measurement, 252-3, 957-8resistance temperature detectors (RTD),254, 302, 305. 306-7resistivity of soil. 700-1, 710resonance:effects, 557-8, 604-5harmonic generation, 741surge arresters, 604-5response spectrum of earthquakes. 439-4 1restrike voltage, 730rigid couplings, 197ring frame motors, 167-8rising mains bus systems, 861-2rod grounding. 697-8rotating arc technology, 641-2rotating machines:protection, 576-7surge protection, 578-83switching surges, 578-83see also electric machines; electricmotors; HT & LT motors; slipringinduction motors; squirrelcage induction motorsrotors:designs, 3740double squirrel cage motors. 3840heat, 279, 281-2locking prevention, 291pulley drives, 215-16resistance induction motors, 84-7, 9 1 -3slip-ring induction motors, 84-7stator rubbing, 242tests, 2634round conductor grounding, 698-9routine tests:bus systems, 9534, 958current transformers, 494-6clcctrical machincs, 252metal enclosed bus systems. 9534. 958metal enclosed switchgear assemblies,421,435voltage transformers, 492-3rules of thumb, 321-9RVTs see residual voltage transformerssafety:capacitor units, 829-37devices, 306electric motors, 306Ex ‘i’ circuits. 182-3salurated type reactors, 849-50. 884-5scoop control fluid couplings. 201-3screws. standards, 912Index 969section determination. conductors/enclosures. 941-5segregated phase bus systems. 85940seismic effects:bus systems. 871. 954electrical machines. 252energy released, 437-9metal enclod systems, 436-9. 954switchgear assemblies. 363. 436-9testing, 421-2, 444-7see nlso earthquakesself-healing capacitors, 8 I I. 8 I4semiconductor devices:circuit configurations. 1 16-28converter units. I 17-1 9microprocessor-based relays. 296-9.771-2motor control. I 16-28protection, 130-4rectifiers units. 1 17-1 9relays, 296-9. 771-2sequence impedances. 349series capacitors, 779-82. 799, 800ferro-resonance effects. 78 1protection, 834-7rating, 779-80sub-synchronous resonance, 78 1-2system voltage regulation, 779-80series compensation, 799-802advantages. 780-2syztem voltage regulation. 782-3series inductance. 757. 823-4series reactors, 747-8. 850-2. 853service conditions:bus systems. 8624. 869-72cables, 544metal enclosed systems. 8623performance. 9-17switchgear assemblies. 342. 362-4service factors. 17X. 210sewage submersible pump motors. 173shaft currents, 236-8shell protection, 830-1shielding signals, 339shock loading, 65-6short-circuit effects. 864-9. 930. 93 Ishort-circuit tests:metal enclosed switchgear asseinblies.429-32motor control centre. 433panels, 430power circuits, 429power distribution, 360-1. 430short-time current. 43 1switchgear assemblies. 429-32short-time:current, 43 1duties, 63-5factors, 489induction motors. 63-5ratingscables, 547-8components guidelines, 367switchgear assemblies. 34650,364-8transformers. 48X-90shrouding, 374-5, 378. 379~ e dso r enclosuresshunt capacitors, 799. 800. 829-34shunt reactors. 798-9. 850

970 Indexsignal shielding, 339silicon-controlled rectifiers, 115-16see also semiconductor devicessilver plating joints, 911single core cables, 371-2single current transformer protection, 688single phasing, 280-1, 291-2single phasor (vector) control, 104-6single-capacitors switching, 75 1-2single-device motor protection relays,293-9single-phase motors:capacitor start, 27-8, 30-1induction, 27-33protection, 679-8 1split phase winding, 27, 29universal, 29, 32single-phase systems protection, 679-8 1skin effects:analysis, 876-7bus systems, 873-8conductors, 873-8determination, 877-8isolated conductors, 875-6rectangular bars, 875tubular conductors, 938-9slip measurcmcnt, 2567slip recovery systems, 140-2slip-ring induction motors:automatic speed control, 95control, 81-96duty rating, 90electrode electrolyte starters, 94-5features, 83five-step stator-rotor starter, 89formulae, 96rotor resistance calculation, 91-3rotor resistance selection, 84-7speed control, 934, 95standards, 95starting, 81, 83-91, 94-5step numbers, 89, 90temperature limits, 91slip-ring motor tests, 264smoothing reactors, 852-3soft starting, 77-8, 13840soft stopping, 140soil:conductivity, 710contact resistance, 710moisture, 710performance improvement, 710resistivity, 700-1, 710solar radiation, 940-2, 949, 957solid neutral grounding systems, 664solid-state technology:control, 111-16energy conservation, 134-8process plant operation, 142-4relays, 295-6speed control, 99-101space factors, 870see also proximity effectsspace fields, 942space heaters, 307-8, 378-9motor control centres, 374-5switchgear assemblies, 397special-purpose capacitor units, XI Xspecial-purpose current transformers, 479-88special-purpose motors, 165-94formulae, 193selection, 189-92standards, 192-3specifications:current transformers, 47 1-5, 492electrical motors, 250framing, 262-3impulse voltage withstand tests, 262-3IS, 247-50summary, 492Zones, 180-2speed control:automatic, 95cage motors, 99constant torque, 101-3fluid couplings, 145motor sensors, 111slip-ring induction motors, 934, 95solid-state technology, 99-101squirrel cage motors, 99speed switches, 308speed-current tests, 257-8speed-torque:curves, 37tests, 257-8spinning framc motors, 167-8splashing water ingress protection, 268split phase winding, 27, 29splitter plates, 636-7sprayers, hand-held, 268squaring technique, 297squirrel cage induction motors:autotransformer starting, 73-7direct on-line starting, 7 1-2electrolyte switching, 78formulae, 79HT motors, 75-6LT motors, 75reduced voltage starting, 72-8soft starting, 77-8speed control, 99standards, 79, 162staddelta starting, 72-3, 76starting, 69-80suitability, 190-1switching, 78vertical hollow shaft, 172see also induction motorssquirrel cage motors, 22-3, 26, 99, I62double, 38-40stability, 482, 788stainless steel grid resistance, 86stalled conditions, 45-6stalling protection, 291standard device function numbers, 550-1standards:V-belts, 216aluminium, 925bearings, 242-3bolts, 912brushes, 242bus systems, 901-2, 958cables, 548-9capacitors, 773, 824-5, 842captive power generation, 530circuit interrupters, 656copper, 92.5electric motors, 48, 3 19electrical machines, 242-3, 269-7 1enclosures, 949flat belts, 216ground fault protection schemes, 693grounding practices, 674, 720induction motors, 33, 67, 79, 95, 162insulation, 228, 674IS0 8402,247IS0 9000, 247-9IS0 9001, 247IS0 9002, 248IS0 9003, 248IS0 9004, 248metal enclosed systems, 452-3, 958motors, 162overvoltages, 674painting, 418reactors, 853screws, 912slip-ring induction motors, 95solar radiation, 949special-purpose motors, 192-3squirrel cage induction motors, 79surge arresters, 624switchgear assemblies, 398-9, 452-3switching devices. 824-5system grounding, 674system voltage regulation, 807transformers, 495voltage regulation, 807voltage surges, 584star capacitor units, 729, 730, 816-18star connected stators, 282staddelta starting, 72-3, 76start and brake duties, 52, 53starting:current effects, 40-1diesel generators, 506-7frequency, 161induction motors, 81, 83-91motors, 12control, 13840frequency, 161heating, 43-4induction, 69-81, 83-91slip-ring induction motors, 81, 83-91soft, 77-8, 138-40squirrel cage induction motors, 69-80switchgear assemblies, 388, 392-6starting on reduced voltage, 72-8static control, motors, 97-164static devices, harmonics, 129-30static drives, 13845, 145-6stations see power stationsstator-rotor starters. 89stators:currents, 7, 277heat, 278, 281induction motors, 7star connected, 282steel:core losses, 18punched grid resistors, 86steep-fronted transient recovery voltage,572-7steep-rising surges, 566-72step numbers, 89, 90step voltages:grounding grids, 712-14tolerable, 704stopping motors, 140, 161

Index 971straight-through joints, 906-9strip grounding, 698-9stripping tests. 408-9stroboscopic methods, 257sub-synchronous resonance (SSR), 781-2submersible pump motors, 170-1, 173-6sugar mill centrifuge motors, 169-70sulphur hexaflouride gas circuit breakers,63743. 651-5sumination current transformers. 476-7jupplies, control schemes, 383. 384. 385surface-cooled motors, I76surge arresters:application. 587-625basic insulation levels, 5934classification, 613-15condition assessment. 617-23coordinating current, 595discharge current, 595distribution networks, 606-7, 614-15electromagnetic transference, 601electrostatic transference, 599energy capability. 612-13. 621factor r values, 602formulae. 624-5gapless. 590-1, 605-13, 616. 620gapped, 589-90lrakagc current monitoring. 618-19monitoring conditions, 6 I8motors. 615-17parameters. 603-4, 623power generation, 606-7pressure relief facilities. 617protectionanalysis, 621-3characteristics, 61 1levels, 596-605. 609-12margins. 594-5reflections, 621-3resonance effects, 604-5wlection, 587-625standards, 624station class, 620transferences, 621-3transmission, 606-7travelling wave reflection, 597-9ZnO, 591-3surges:capacitor$. 580-2, 729-32containmenl. 648-50defining, 594diverters see surge arrestersenergy definition. 564frequencies, 343, 561-2HT motors, 563impedance, 574-5definition, 562-4power capacitors. 729-32steep-rising. 566-72suppression. 582-3switching, 567-9transferences, 576-7, 599-604see nlso protection: voltage surgesswitchboards see switchgear assembliesswitches:circuit interrupters, 629closures. 629duty cycles, 3 I7LT motors. 3 IOmotor component selection, 312-13switchgear assemblies, 331-55 1altitude, 362ambient temperatures, 362application, 335atmospheric conditions, 362auxiliary controls, 336-7bus systems design, 368-71construction requirements, 372-6continuous current ratings, 345-6.371-2control scheme diagrams, 387-97controlgear, 333-418conventional, 339-42current ratings, 345-50, 364-8. 371-delay tripping, 364-5design, 34244, 371-83diversity factors, 346, 366-7draw-out construction, 341-2equipment ratings, 364-8explosion proof. 362-3fault levels, 350-6fixed type, 340-1tlame proof, 362-3grounding arrangements, 375-6impulse levels, 344-5installation. 384-7instrumcnt wiring, 388, 397insulation, 344-5, 423LT fault levels, 350-6maintenance, 384-7manufacturing practices, 372-6metal enclosed, 419-54motor control centres, 376-7motor protection, 308-12painting, 400-17power circuits, 387-97power frequencies, 343ratings, 345-50, 364-8, 371-2seismic effects, 363sequence impedances, 349service conditions, 342, 362-4short-time ratings, 346-50, 364-8signal shielding, 339space heaters, 378-9, 397standards, 398-9starters. 388, 392-6surge frequencies, 343temperatures, 345-6. 362, 37 1-2testing, 419-54thermal ratings, 364tripping, 364-5types, 335-9vibration effects, 363wiring, 388, 397see ulso metal enclosed switchgearassembliesswitching:auctioneering, 768binary sequt-ncas, 769, 770capacitors, 725-75, 768-9circuit harmonics, 129-30circuit interrupters, 629-30devices, 809-26equipment, 835-6first in first out sequences, 768first in last out sequences, 768formulae, 8254harmonics. 129-30inrush, 7504interrupters. 629-30LT systems, 57 Imediums, 629-30overvoltages, 729-32power capacitors. 725-75. 768-0protection. 576-7rating, 809-26reactive banks. 803-7sequences, 768-9. 770standards. 824-5surges, 567-9containment. 648-50LT systems. 57 Iprotection, 576-7-2 rotating machines, 578-83transient-free. 758. 805symbols, 391symmetrical lines, 798synchronization, 521-7synchronous motors, 501 -2synchroscopes, 52&5system grounding, 657-7sformulae, 674-5insulation, 667-8standards, 674system harmonics. 9-10? 485see ulso harmonic effectssystem voltage regulation. 777-808capacitors. 779-80. 804compensation. 782-3formulae, 808large reactive banks. 803-7lines, 785-9. 790-2reactive banks. 803-7reactive power management, 783-90receiving-end voltage. 799-80 Iseries capacitors. 779-80series compensation. 782-3stability levels. 802-3standards, 807switching reactive hanks. 803-7thyristors, 804-5transient stability levels, 802-3transmission lines. 785-9. 790-2tachogenerators, 1 1 1. 308tank sizes, 404-5tap off boxes, 863, 933TCRs see thyristor-controlled reactorstectonic pIate boundaries, 43Xsee also earth cross-sectiontee joints, 909TEFC .we totally enclosed fan-cooledmotorstelephone lines, 731temperatures:bearings, 238-9. 255, 307bus systems, 427. 864-7 1. 954.956cables. 545conversion formulae. 324electrical machines, 253-5limits, 91metal enclosed systems. 4249,954. 956slip-ring induction motors, 91switchgear assemblies, 345-6. 362,371-2.424-9windings. 241. 306-7temporary overvoltages. 657-75terminal boxes, 183. 184. 185. 240. 241terminal equipment, 572-7

972 Indextermination, cables, 548test response spectrum (TRS), 447testing:authorities, 183bus systems, 951-8capacitor units, 82743control transformers, 491-4dielectric property verification, 260electric motors, 1-240electrical machines, 245, 250-69enclosures, 957equipment, 449-50impulse voltage withstand, 954instrument transformers, 4914insulation, 225metal enclosed assemblies, 419-54metal enclosed bus systems, 951-8nozzles, 269painting, 408-9purposes, 250-1quality systems philosophy, 421switchgear assemblies, 419-54tolerances, 266transformers, 4914voltage transformers, 491-3voltages, 260surges, 578TETV see totally enclosed tube ventilatedmotorstextile motors, 167-8thermal:blast techniques, 642-3curves, 59-64overcurrent relays, 283, 284, 285-6power stations, 186-9, 241-2ratings, 364stability, 83940stresses, 818-19time constants, 56withstand time, 44-8see also temperaturesthermistors, 302-7thermocouples, 305, 307thermosetting plastics, 369Thevenin’s theorem, 610three-phase four-wire systems, 683-4throttle control, 135, 136, 138through-fault condition, 481thrust loads, bearings, 212thyristor-controlled reactors (TCRs), 804-7thyristor-switched capacitor banks (TSC),804-7thyristors:circuit comparison, 120commutation, 116conduction, 116control, 114-16inverter units, 122-5reactors, 804-7rectifier converters, 120switched capacitor banks, 804-7transistor comparison, 1 18time history of earthquakes, 439,440time of start-up, 414tolerable step voltages, 704tolerable touch voltages, 704-5tolerances, 2334, 269torque:accelerating, 189fasteners, 907induction motors, 8motors, 3545, 1767output, 8Total Quality Management (TQM), 249totally enclosed fan-cooled (TEFC)motors, 21totally enclosed tube ventilated (TETV)motors, 26touch voltages, 704-5, 712-14TOV capability, 608-9determination, 605-7, 608gapless surge arresters, 605-9TQM see Total Quality Managementtraction type couplings, 199transferences, surge arresters, 621-3transformers:application, 455-96control, testing, 4914core-balanced current type, 685-8dynamic currents, 488-90formulac, 496instrument, 455-96momentary peak currents, 488-90mounting, 686-8residual voltage, 4604selection, 455-96short-time factors, 489short-time rating, 488-90standards, 495surge transference, 599-604test requirements, 491-4types, 457see also current transformers; voltagetransformerstransient recovery voltage (TRV), 132,568-9induction motors, 561motors, 572-7steep-fronted, 572-7transient-free switching, 758, 805transients:overhead lines, 558-9severity determination, 577-8stability, 558-9, 802-3voltage surges, 558transistors, 112-14, 118bipolar, 113- 14, 125Darlington. 112-13transmission:belt drives, 202-1 1lines, 785-9, 790-2loads, 195-217network layouts, 347networks, typical, 727surge arresters, 6067system circuit diagrams, 358-9see also individual couplingstravelling wave reflection, 597-9triacs, 116triple harmonic generation, 740, 741, 742tripping, 569switchgear assemblies, 364-5TRS see test response spectrumtrunking system endurance, 958TRV see transient recovery voltageTSC see thyristor-switched capacitorbankstubular busbar properties, 923tubular conductor skin effects, 938-9tuning reactors, 852turbine pumps, 170-1turns per coil, 575two-pole circuit breakers, 681type tests:capacitor units, 83941electrical machines, 251-2metal enclosed bus systems, 953, 954-8metal enclosed switchgear assemblies,422-35transformers. 4934UHV system classification, 736uncompensated transmission lines, 785-9underground systems, 660-3, 730-1unfavourable conditions, 275-9, 283-7see also hazardous locationsunits, 324universal motors, 29, 32unrestricted ground fault protection, 690V-belts, 206-10correction factors, 211cross section selection, 207formulae, 217pitch lengths, 213power ratings, 208service factors, 210standards, 216VA burden:current transformers, 4734voltage transformers, 467-8vacuum circuit breakers, 643-6, 651-5vacuum contactors, 311-12,315vacuum pressure impregnation, 2234vane control, 135variahle speed fluid couplings, 14.5, 148-50, 201-2variable-filling couplings, 199-201vector control, 103-1 1see also phasor controlvelocity of propagation definition, 564-5vertical bus systems, 861-2vertical hollow shaft squirrel cage motors,172vibration:bus systems, 872measurement tests, 258-9, 259motors, 177-8probes, 308switchgear assemblies, 363virtual current chopping, 648volatile liquid classification, 179voltage:applying up to 200 I84dielectric tests, 260drop, 544-7, 872-3gradients, 704-7limits, 10rating, 759-60system classification, 736tests, 262variations, 10-17, 834ZnO surge arresters, 593vollage source iriverlrrr (VSI), 125voltage surges, 553-722causes, 555,559-60

contact bouncing. 57 1-2effects. 555formulae. 584-5insulation, 579-80. 583-4inverter units, 133-4lightning strikes, 559machine ratings. 575-6protection. 292-3. 576-7. 583-4remedies, 555standards, 584.;urge suppressors. 582-3test circuits. 578transient\, 558. 577-8turns per coil. 575voltase transformers, 457-70capacitor. 164-6control. 466-9design. 457-69electromagnetic. 459-60general specification. 457-69gt-ound fx~lts, 462. 463hold-on values, 468inatallation. 469-70measuring. 458phasor diagrams. 459. 160protection. 458residual, 4604routine tests. 492-3specifications, 457-70testing. 49 1-3type tests. 493VA burden, 467-8voltage unbalance:bus systems, 882-6capacitors. 832induction motors. 9-10protection, 291proximity effects, 882-6VSI scc voltagc sourcc inverterswasher sizes, 907waste water treatment, 41 3-15water ingress protection. 24. 266-8water temperature, induction motors, 16water-cooled motors. 307watertightnes\ tests, 957weatherproofing, 2 1-2, 434wet test, windings, 228wind-power. 157-61windings:ageing. 22 Iinsulation. 219-29ageing. 22 Imaintenance, 2 19-29material propertie\. 22 1qystems, 22 1-3wet re\[. 228maintenance. 2 19-29material properties, 22 Imoi5ture condensation. 307-8temperatures, 24 I. 306-7wet test, 228windmills .we wind-pou erwiring method\. 686withdrawable ... .\re draw-out. .XLPE three-corc cable\. 534iero period acceleration (%PA). 443ZnO surge arrester\. 591-3. 624Zone 0 specification. 170. I80Zone 1 \pecification. 179, 180-2Zone 2 specification. 170. I82zone faults, 483ZPA ,YW zero period accelerationIndex 973

ET WTRICAL ENGINEEWCIndustrialPowerEngineeringand ApplicationsHandbookK. C. AgrawdIndustrial Power Engineering and Applications Handbook is designed for the electrical power engineeror electrical engineering student who needs a single source reference guide to every subject containedwithin that discipline. It covers a huge range of topics from electric motors and generators to switchgearand transformers, capacitors, cabling and buses as well as testing and installation requirements. Eachsubject is covered comprehensively from theory to practice with relevant examples and details on allrelevant international standards including ISO, IEEE, ANSI, NEMA, BS-EN, DIN and of course IEC. Thereare also details of design, use and maintenance techniques that will make the book of vital interest toengineers responsible for plant power generation and supply in any industry.Never before has so much ground been covered in a single volume reference source. This book issure to be of great value to students, technicians and practising engineers as well as equipmentdesigners and manufacturers, and should become their one-stop shop for all information needs inthis subject area.Newnes Power Engineering SeriesElectronic Control in Electrical Systems Acha, Agelidis, Anaya & Miller 0-7506-5126-1Industrial Power Engineering and Applications Handbook Agrawal 0-7506-7351-6Power Engineering titles from NewnesElectrical Engineer’s Reference Book Jones, Laughton & Say 0-7506-1202-9Newnes Electrical Engineering Handbook Warne 0-7506-4879-1J&P Transformer Book, 12th ed Heathcote 0-7506-1158-8Transmission and Distribution Electrical Engineering Bayliss 0-7506-4059-6Electrical Power Systems Technology Fardo & Patrick 0-7506-9722-9High Voltage Engineering Fundamentals Kuffel, Zaengl& Kuffel 0-7506-3634-3Industrial Bmhless Servomotors Moreton 0-7506-3931-8High Voltage Test Techniques Kind & Feser 0-7506-5183-0Cover image courtesy of Digital Stock.NewnesAn imprint of Butterworth-Heinemannhttp://w.newnespress.corn