Proceedings of the 2010 Australian Geothermal Energy Conference

G E O S C I E N C E A U S T R A L I AProceedings of the 2010 AustralianGeothermal Energy ConferenceGurgenci, H. and Weber, R. D.Record2010/35GeoCat #71204APPLYING GEOSCIENCE TO AUSTR ALIA’S MOST IMPORTANT CHALLENGES

Proceedings of the 2010 AustralianGeothermal Energy ConferenceGEOSCIENCE AUSTRALIA Record 2010/35Edited byHal Gurgenci 1 and Rikki Weber 21 Queensland Geothermal Energy Centre, The University of Queensland, St Lucia 40722 Geoscience Australia, GPO Box 378, Canberra, ACT 2601

Department of Resources, Energy and TourismMinister for Resources and Energy: The Hon. Martin Ferguson, AM MPSecretary: Mr Drew ClarkeGeoscience AustraliaChief Executive Officer: Dr Chris Pigram© Commonwealth of Australia, 2010This work is copyright. Apart from any fair dealings for the purpose of study, research, criticism,or review, as permitted under the Copyright Act 1968, no part may be reproduced by any processwithout written permission. Copyright is the responsibility of the Chief Executive Officer,Geoscience Australia. Requests and enquiries should be directed to the Chief Executive Officer,Geoscience Australia, GPO Box 378 Canberra ACT 2601.Geoscience Australia has tried to make the information in this product as accurate as possible.However, it does not guarantee that the information is totally accurate or complete. Therefore, youshould not solely rely on this information when making a commercial decision.ISSN 1448-2177 ISBN 978-1-921781-38-4GeoCat # 71204Recommended bibliographic reference: Gurgenci, H. and Weber, R. D. (editors), 2010.Proceedings of the 2010 Australian Geothermal Energy Conference, Geoscience Australia,Record 2010/35.Recommended bibliographic citation: Glikson, A. and Uysal, T.2010. Evidence of impactshock metamorphism in basement granitoids, Cooper Basin, South Australia, in: Gurgenci andWeber (editors), Proceedings of the 2010 Australian Geothermal Energy Conference, GeoscienceAustralia, Record 2010/35.Cover illustration: 3D image of gravity inversion modelling in the area between Mt Isa and theGeorgetown Inlier, Queensland, viewed obliquely from the south. The red bodies enclose regionsof low density interpreted as granites located at depth beneath Millungera Basin sediments, thebase of which is shown as the yellow surface. The interpreted granites are shown projected abovean image of the Bouguer Gravity field, from which the inversion model was generated. Courtesyof Alison Kirkby, Geoscience Australia.

The Convenors would like to thank the following sponsors for their kind support:Gold SponsorsSilver SponsorsBronze SponsorsConference Dinner SponsorHappy Hour Sponsors

Australian Geothermal Conference 2010IntroductionThis volume is a compilation of Extended Abstracts presented at the 2010 AustralianGeothermal Energy Conference, 17-19 November 2010, Adelaide Convention Centre,Adelaide, organised by the Australian Geothermal Energy Association and the AustralianGeothermal Energy Group.As editors of these proceedings, we first would like to thank the Technical Committeewho have been very generous with their time to review the submissions and to workwith the authors of the accepted submissions to improve the quality of the Abstractsand the presentations.We also thank the Conference Chair Tony Hill and the rest of the Organising Committeewho were brave enough to take on the job organising this second event on behalf of theAustralian Geothermal Energy Association and the Australian Geothermal Energy Group.Rob Bulfield and Sapro Conference Management have done an excellent job in assistingthe organisation.We also thank again the sponsoring companies for their generous support.We thank Geoscience Australia for publishing the proceedings.Finally, we thank all delegates without whose participation none of this is worthwhile.Hal GurgenciRikki Weberi

Australian Geothermal Conference 2010Conference Organising CommitteeSusan Jeans, AGEAVicki Cleavland, AGEADr Graeme Beardsmore, Hot Dry Rocks Pty LtdDr Betina Bendall, PIRSADr Anthony Budd, Geoscience AustraliaBarry Goldstein, PIRSAProfessor Hal Gurgenci, University QueenslandChris Matthews, AGEAJonathan Teubner, PetrathermProgram Technical PanelProfessor Hal Gurgenci (chair), Queensland Geothermal Energy Centre of Excellence,Dr Anthony Budd, Geoscience Australia,Delton Chen, Geodynamics Ltd,Gareth Cooper, Hot Dry Rocks Pty LtdPeter Dowd, The University of AdelaideDes Fitzgerald, Intrepid Geophysics,Dr Massimo Gasparon, The University of QueenslandHelen Gibson, GeoIntrepidAlison Kirkby, Geoscience AustraliaAlan Knights, Green Rock Pty LtdChris Matthews, AGEA,Tony Meixner, Geoscience AustraliaProfessor Behdad Moghtaderi, The University of NewcastleProfessor Klaus Regenauer-Lieb, The University of Western AustraliaPeter Reid, Petratherm LtdMichel Rosener, Geodynamics LtdDr Catherine Stafford, Granite Power LtdJonathan Teubner, PetrathermDr Adrian Williams, Geodynamics Ltd DrDoone Wyborn, Geodynamics LtdDr Yung Ngothai, The University of AdelaideConference OrganisersRob Bulfield Sapro Conference Management Ph: 08 8274 6043 Fax: 08 8274 6000www.sapro.com.auii

Australian Geothermal Conference 2010Stream OneA.R. Budd, E.J. Gerner, A.M. Kirkby, R.D. Weber, A.J. Meixner and D.C.ChampionGeoscience Australia’s Onshore Energy Security Program: progress by the GeothermalEnergy ProjectCara Danis and Craig O’Neill3D thermal modelling VS down-hole temperature extrapolation and the implications fortargeting potential geothermal anomalies: a Sydney Basin case studM. Hand and Y. NgothaiSouth Australian Geothermal Research CentreHearst Ghabri Mehran, Robert B. Retallcik, Trent and Killin KevinApplication of magnetotellurics in geothermal reservoir characterizationDavid Jenson, Kerry Parker, Ron Palmer and Bertus de GraffHutton Sandstone – a viable geothermal developmentAshok A. Kaniyal, Jonathan J. Pincus and Graham J. NathanInvestment in common-use network infrastructure to trigger incremental expansion ingeothermal energy generating capital stockIvan Kočiš, Tomáš Krištofič, and Igor KočišUltra deep drilling technologies for geothermal energy productionp1p4p9p11p14p20p21Gideon B. Kuncoro, Yung Ngothai, Brian O’Neill, Allan Pring and Joel BruggerA Preliminary Study on Na-Cl-H 2 O-Rock Interactions of the Hot Fractured Rock GeothermalSystem in Cooper Basin, South Australiap24Adrian Larking and Gary MeyerGeothermal Energy in the Perth Basin, AustraliaComparisons with the Rhine Grabenp30A.D. Long, A.R. Budd, H. Gurgenci, M. Hand, C. Huddlestone-Holmes, B. Moghtaderi and K.Regenauer-LiebAustralia’s Geothermal Research Activitiesp35V. Marshall, J. Van Zyl, S.E. Byron, T. Uysal and M. GasparonComparative petrology & geochemistry of high heat-producing granites in Australia & Europep41Craig McClarren, Massimo Gasparon and I. Tonguç UysalGeothermal Prospection Using Existing Groundwater Geochemical and Thermal Datasets:Identifying Regions of Interest in Queensland from Government Well and Bore Datap48Malcolm B. McInnesGenie Impact Drilling: Synopsis of a New Hard Rock Drilling DevelopmentAlexander W. Middleton, I Tonguç Uysal, Massimo Gasparon and Scott BryanHydrothermal alteration aspects of high heat producing granites in Australia & EuropeThomas Poulet, Soazig Corbel and Michael StegherrA Geothermal Web Catalog Service for the Perth Basinp52p59p64Martin Pujol, Grant Bolton and Fabrice GolfierFlow and heat modelling of a Hot Sedimentary Aquifer (HSA) for direct-use geothermal heatiii

Australian Geothermal Conference 2010production in the Perth urban area, Western Australia (WA)p69Sukanta RoyGeothermal Exploration in Indiap75Behnam Talebi, Mark Maxwell, Sarah Sargent and Steven BowdenQueensland Coastal Geothermal Energy Initiative – An Approach to a Regional Assessmentp81David H. Taylor, Ben Harrison, Michelle Ayling and Kate BassanoGeothermal Exploration in Victoria, AustraliaI. Tonguç Uysal, Alexander W. Middleton, Robert Bolhar and Massimo GasparonGeochemistry of silica rocks in the Drummond Basin as a record of geothermal potentialp87p91Gordon Webb and Ben HarrisonEstimates of the magnitude of topographical effects on surface heat flow from finite elementmodellingp94Doone Wyborn, Robert Hogarth, Delton Chen and Michel RosenerThe next Australian EGS Resevoir – Jolokia 1 stimulationChaoshui XuOptimised Fracture Network Model for Habanero Reservoirp97p98Stream TwoAleks D. Atrens, Hal Gurgenci and Victor RudolphProgress in CO 2 -based EGSp104Peter Barnett and Tim EvansExploration and assessment of Hot Sedimentary Aquifer (HSA) geothermal resources in theOtway Basin Victoriap106J.E. Davey and J.T. KeetleyValidation of Geothermal Exploration Techniques for Analysing Hot Sedimentary Aquifer andEnhanced Geothermal Systemsp112Bertus de Graaf, Ian Reid, Ron Palmer, David Jenson and Kerry ParkerSalamander-1 – a geothermal well based on petroleum exploration resultsp117Ben Clennel, Lionel Esteban, Ludovic Ricard, Matthew Josh, Cameron Huddlestone-Holmes,Jie Liu and Klaus Regenauer-LiebPetrophysical Methods for Characterization of Geothermal Reservoirsp123Farizal and Prashanti Amelia AnisaGeothermal Power Plant Competitiveness in Indonesia: Economic and Fiscal Policy Analysisp130Desmond Fitzgerald and Raymond SeikelA Method For Calibrating Australian Temperature-Depth ModelsAndrew Y. Glikson and I. Tonguç UysalEvidence of impact shock metamorphism in basement granitoids, Cooper Basin, SouthAustraliaT. HarinarayanaGeothermal Energy in India: Past, Present and Future Plansp138p141p148Peter Leary, Peter Malin, Eylon Shalev and Stephen Onachaiv

Australian Geothermal Conference 2010Aquifer heterogeneity – is it properly assessed by wellbore samples and does it matter foraquifer heat extraction?p149Yan Liu, Huilin Xing, Zhenqun Guan and Hongwu ZhangAutomatic Meshing and Construction of a 3D Reservoir System: From Visualization towardsSimulationp157Luke Mortimer, Adnan Aydin and Craig T. SimmonsTargeting Faults of Geothermal Fluid Production: Exploring for Zones of Enhanced Permeabilityp163Grahame J.H. OliverAn EGS/HAS concept for SingaporeJared Peacock, Stephan Thiel, Graham Heinson and Louise McAllisterMonitoring EGS Reservoirs Using Magnetotelluricsp169p175Joshua Koh, Nima Gholizadeh-Doonechaly, Abdul Ravoof and Sheik RahmanReservoir characterisation and numerical modelling to reduce project risk and maximise thechance of success. An example of how to design a stimulation program and assess fluidproduction from the Soultz geothermal field, Francep179P. W. Reid, L. McAllister and M. MesseillerStatus of the Paralana 2 Hydraulic Stimulation ProgramAlvin I. Remoroza, Behdad Moghtaderi and Elham DoroodchiThree Dimensional Reservoir Simulations of Supercritical CO 2 EGSCatherine E. StaffordMeasuring the Success of EGS Projects: An Historical to Present Day Perspectivep203p206p210J.F. Wellmann, F.G. Horowitz, L.P. Ricard and K. Regenauer-LiebEstimates of sustainable pumping in Hot Sedimentary Aquifers: Theoretical considerations,numerical simulations and their application to resource mappingp215Thomas W. Wideman, Jared M. Potter, Brett O’Leary and Adrian WilliamsHydrothermal Spallation for the Treatment of Hydrothermal and EGS Wells: a Cost-EffectiveMethod for Substantially Increasing Reservoir Production Flow Ratesp222Bisheng Wu, Bailin Wu, Xi Zhang and Rob JeffreyThermal effects on the wellbore stabilityMatthew Zengerer, Chris Matthews and Jerome Randabel3D Geological Modelling for Geothermal Exploration in the Torrens Hinge ZoneJ. Zhang and H.L. XingNumerical simulation of geothermal reservoir systems with multiphase fluidsXi Zhang and Robert G. JeffreyDevelopment of Hydraulic Fractures and Conductivity near a Wellborep226p232p243p247Stream ThreeDavid L. Battye, Peter J. Ashman and GUS J. NathanEconomics of geothermal feedwater heating for steam Rankine cyclesGraeme Beardsmore, Ladislaus Rybach, David Blackwell and Charles BaronA Protocol for Estimating and Mapping Global EGS Potentialp253p257v

Australian Geothermal Conference 2010G. Beardsmore, A. Budd, B. Goldstein, F. Holgate, G. Jeffress, A. Larking, J. Lawless, J. Libby,M. Middleton, P. Reid, C. Stafford, M. Ward and A. WilliamsChanges to Geothermal Reporting Code and Guidelines on Company Reporting p263Anna Collyer, Nadia Harrison and Natasha McNamaraGeothermal Energy’s Transition from Resource to Power: Legal Challengesp269Barry Goldstein, Betina Bendall, Alexandra Long, Michael Malavazos, David Dockshell andDavid LoveTrusted Regulation for Geothermal Resource Development (Including Induced SeismicityAssociated with EGS Operations)p273B.A. Goldstein, G. Hiriart, J.Tester, B. Bertani, R. Bromley, L.C.J. Gutierrez-Negrin, E.Huenges, H. Ragnarsson, A. Mongillo, M.A. Muraoka and V.I. ZuiGreat Expectations For Geothermal Energy – Forecast to 2100Hal Gurgenci and Zhiqiang GuanNew Design Concepts for Natural Draft Cooling TowersKamel HoomanQGECE Research on Heat Exchangers and Air-Cooled Condensersp281p290p294Robert McKibbin, Franklin Horowitz, Neville Fowkes and Brenda FlorioEquation-Free Summary of the 2010 Mathematics in Industry Study Group on Geothermal DataAnalysis and Optimizationp298Jim LawlessStatus of Compliance With The Geothermal Reporting CodeJim LawlessApplication of the Australian Geothermal Reporting Code to “Conventional” GeothermalProjectsDavid N. Love, Gary Gibson, Russell Cuthbertson and Wayne PeckSmall Australian seismic events and their effectsSteve Kennedy, Chris Matthews, Damia Ettakadoumi and Cynthia CroweCommunity Consultation Best Practice ManualCraig Morgan and David PaulOptimising ORC use in Australian Geothermal Applicationsp302p305p306p400p402Yung Ngothai, Norio Yanagisawa, Allan Pring, Peter Rose, Brian O’Neill and Joel BruggerMineral scaling in geothermal fields: A reviewp405Donald J.B. Payne and Dezso KiralyDirect GeoExhange Cooling for the Australian Square Kilometer Array Pathfinder – Field Trialp410Klaus Regenauer-Lieb and the WAGCoE team, Peter Malin and the IESE team, Steve Harveyand the CSIRO CESRE team, Edson Nakagawa and CSIRO’s Geothermal & PetroleumPortfolio and Hal Gurgenci and the QGECE teamA Research Well for the Pawsey Geothermal Supercomputer Cooling Projectp416Kyra ReznikovFracture Stimulation – legal risks and liabilitiesAndrew S. Rowlands and Jason CzaplaQGECE Mobile Geothermal Test Plant & ORC Cycle Challengesp421p422vi

Australian Geothermal Conference 2010Jonathan TeubnerAustralian geothermal Industry working with Investors – Defining the key parameterscontributing to the long run marginal cost of geothermal and providing a comparison benchmarkon the key unknownsp426O. Bouchel, T. Hazou, C. Scott and Roger WindersCOMMUNITY TRUST – Engagement & CommunicationMelissa Zillmann and Myria MakrasGeothermal energy legislation and developments in Queenslandp427p434vii

Australian Geothermal Conference 2010Geoscience Australia’s Onshore Energy Security Program: progressby the Geothermal Energy ProjectA.R. Budd,* E.J. Gerner, A.L. Kirkby, R.D. Weber, A.J. Meixner and D.C. ChampionGeoscience Australia. GPO Box 378, Canberra, ACT 2601.*Corresponding author: Anthony.Budd@ga.gov.auAbstractGeoscience Australia’s $58.9M 5-year OnshoreEnergy Security Program began in 2006 andincludes a new Geothermal Energy Project.The project aims to assist the development of ageothermal industry in Australia by: providingprecompetitive geoscience information, includingacquisition of new data; informing the public andgovernment about Australia’s geothermalpotential; providing technical advice togovernment; and partnering with industry ininternational promotional events for the purposeof attracting investment.This abstract gives a brief summation of activitiesundertaken by Geoscience Australia within theOnshore Energy Security Program, principallythose of the Geothermal Energy Project.Keywords: data acquisition, modelling,Geothermal Play SystemsThe Geothermal Energy ProjectFollowing consultation with industry and State andNorthern Territory geological surveys, a numberof activities were identified where GA could eitherfill gaps where no other organisation was able tofulfil, or could complement activities by otherpartners.Advice to governmentGeoscience Australia (GA) is a prescribed agencywithin the Department of Resources, Energy andTourism. GA provides advice to government ongeoscience-related matters, including resources.GA participated in the development of theGeothermal Industry Development Frameworkand Geothermal Industry Technology Roadmap.GA was involved in the program design andsubsequent technical assessment of theGeothermal Drilling Program. GA co-authored(with ABARE) the geothermal chapter of theAustralian Energy Resource Assessment. GA hasbeen involved in the development of theAustralian Code for Reporting of ExplorationResults, Geothermal Resources and GeothermalReserves, and sits on the Joint AGEA-AGEGCode Committee for the purpose of one daycompiling and reporting geothermal resource andreserve estimates in the same way as is done formineral and oil & gas commodities.OZTEMPGA has released OZTEMP, an updated datasetand map of predicted temperatures at 5 km depth,available from the project’s webpage(www.ga.gov.au/minerals/research/national/geothermal). Extensive QA/QC was conducted on thedataset of bottom hole temperatures, and whereavailable new data was included. For the map, theOZSeeBase dataset was used (FrOG Tech,2006), and the Bureau of Meteorology’s MeanAnnual Average Air Temperature (BoM 2010) wasalso used with a correction for surfacetemperature. For the first time, heat flow data wasalso incorporated into the map.Data acquisitionThere is a paucity of temperature-specific data inAustralia, and to address this GA has establisheda capability for measuring surface heat flow viathermal gradient logging and thermal conductivitymeasurement.Thermal gradient loggingWithout the possibility of drilling new holes, GAhas worked with State geological surveys andminerals exploration companies to accessexploration and water bores. Figure 1 shows thedistribution of logged bores as of July 2010.Figure 1. Map showing existing heat flow measurements for theAustralian continent, and the distribution of bores logged fortemperature by Geoscience Australia (black stars).Thermal conductivityGA operates an Anter 2022 Unitherm thermalconductivity meter and associated samplepreparation equipment. Project staff have beeninvolved in establishing operating procedures forthe instrument, a process which has includedinter-lab comparison testing with Torrens Energy,Hot Dry Rocks Pty Ltd, and Southern MethodistUniversity. In addition, GA has engaged Hot Dry1

Australian Geothermal Conference 2010Rocks Pty Ltd to measure two batches ofsamples.Samples have been taken from the majority of thebores measured by Geoscience Australia for theirthermal gradient, as well as other bores for‘stratigraphic’ conductivity values.PromotionUndoubtedly the greatest impediment to thedevelopment of the Australian geothermalindustry since the Global Financial Crisis hasbeen the lack of capital investment. GeoscienceAustralia has participated in “Team AustraliaGeothermal” events at the Geothermal EnergyExpo, Reno 2009, and the World GeothermalCongress, Bali 2010. The collaboration ofindustry, government and associations has beenaimed at portraying Australia as a favourableinvestment destination.Resource assessmentsIn 2007 GA produced an estimate of thecontained heat above 150°C in the top 5 km of theAustralian crust. Understandably this produced avery large estimate of energy. To provide a morerelevant estimate 1% of the thermal resource wasassumed to be accessible and convertible, andthis equates to ~26,000 years of energyconsumption (Budd et al., 2009).The production of the Australian Energy ResourceAssessment highlighted the need for a betterunderstanding of Australia’s geothermalresources potential. This contrasts with otherrenewable energy resources where theknowledge-base is quite advanced. There is aneed for geothermal to be better understood sothat it can be compared more directly to otherenergy sources. With a limitation on theavailability of temperature data, other geosciencedatasets must be used to inform or deriveestimates of geothermal resource potential.Geothermal Play SystemsAs mentioned above, there is a paucity oftemperature-specific data publicly available inAustralia. There is, however, a wealth of highqualitygeoscientific data available throughout thecountry, much of which can be used to makeassessments of geothermal potential from aconceptual view point. Like some companies andother organisations, GA has developed 2.5D and3D methods for processing geological,geophysical and geochemical data to produceresource potential assessments and estimates. A‘systems’ approach has been utilised to enablethese data sets to be used as ‘mappable proxies’to estimate heat production and thermal insulationat regional scales (Budd et al., 2009). Conceptualand empirical methods of developing theGeothermal Play Systems approach have beenpursued, and are still in development.3D thermal modellingThrough a Primary Industries and ResourcesSouth Australia (PIRSA) Australian GeothermalEnergy Group (AGEG) Technical Interest Group 9Tied Grant, a thermal calculation module was builtfor GeoModeller software by Intrepid Geophysics(Siekel et al., 2009). This has been used todevelop thermal models of the Millungerra Basin(unpublished), and the Cooper Basin (such asGibson et al., 2010).We have also undertaken synthetic thermalmodelling tests. This involved the development ofa synthetic grid of granites buried under flat-lyingsediments that was used to model the effects ofchanging thermal and geometric variables (one ata time). Modelled variables include thermalconductivity, heat production and density of thesediments and granite, as well as the thickness,radius and depth extent of the granites and totalsediment thickness. This has resulted in 5,400individual test models, and we are in the processof interpreting these results. These interpretationswill then serve as a guide for a first-passassessment of thermal potential of the wholecontent based on estimates of granite size, basingeometry and composition (Meixner, 2009).North Queensland Energy AssessmentA GIS-based approach was used to qualify theHot Rock and Hot Sedimentary Aquifergeothermal systems of northern Queensland(Huston, 2010).ReferencesBoM, 2010, Climate Data Onlinehttp://www.bom.gov.au/climate/data/Budd, A.R., Barnicoat, A.C., Ayling, B.F., Gerner,E.J., Meixner, A.J. and Kirkby, A.M., 2009, AGeothermal Play Systems approach forexploration Australia, in: Budd and Gurgenci(editors), Proceedings of the 2009 AustralianGeothermal Energy Conference, GeoscienceAustralia, Record 2009/35.FrOG Tech, 2006, OZ SEEBASE ProterozoicBasins Study. PR107.Gibson, H., Seikel, R., and Meixner, T. 2010,Characterising uncertainty when solving for 3Dtemperature: New tools for the Australiangeothermal energy exploration sector. ASEGExtended Abstracts 2010, 1–3.doi:10.1071/ASEG2010ab290.Huston, D.L., 2010, An assessment of theuranium and geothermal potential of northQueensland / edited by Huston, D.L. withcontributions by Ayling, B., Connolly, D., Huston,D.L., Lewis, B., Mernagh, T.P., Schofield, A.,Skirrow, R.S. and van der Wielen, S.E.Geoscience Australia Record 2010/14, Canberra2

Australian Geothermal Conference 2010https://www.ga.gov.au/products/servlet/controller?event=GEOCAT_DETAILS&catno=69711Meixner, T., Johnston, S., Kirkby, A., Gibson, H.,Seikel, R., Stuwe, K., Fitzgerald, D., Horspool, N.Lane, R. and Budd, A., 2009, Establishing hotrock exploration models: from synthetic thermalmodelling to the Cooper Basin 3D geological map.Siekel R., Stuwe K., Gibson H., Bendall B.,McAllister L., Reid P., Budd A., 2009, Forwardprediction of spatial temperature variation from 3Dgeology models. ASEG Extended Abstracts 2009,–. doi:10.1071/ASEG2009ab117.3

Australian Geothermal Conference 20103D thermal modelling VS down-hole temperature extrapolation andthe implications for targeting potential geothermal anomalies: aSydney Basin case studyCara Danis 1* and Craig O’Neill 11 GEMOC Macquarie University, Dept Earth & Planetary Science, NSW 2109, Australia* Corresponding author: cara.danis@mq.edu.auGeothermal exploration programs requireaccurate subsurface temperature information andcurrently this information primarily comes fromtemperature maps created from the extrapolationof shallow down-hole temperature measurements.These extrapolations are often taken frommeasurements made in non-equilibratedboreholes and do not account for variations ingeological structure or thermal conductivity. Herewe present a case study for the Sydney basinwhere we explore temperature maps at 5kmcreated from extrapolated equilibrated and nonequilibratedborehole measurements and frommodelled basin temperatures and the implicationsfor targeting potential geothermal anomalies. Themodelled temperatures are derived from finiteelement models using 3D basin geology anddefined thermal properties.Keywords: Sydney Basin, 3D thermal modelling,temperature extrapolation, geothermal explorationCase Study – Sydney BasinThe Sydney Basin, part of the Sydney-Gunnedah-Bowen Basin system, is a major sedimentarybasin in the east coast of Australia and animportant economic resource. Much attention isgiven to the coal and coal seam gas prospects butrecently the focus has shifted to the basinsthermal structure and geothermal potential.Previous work in the Gunnedah Basin (Danis etal., 2010) has shown that basin architecture andinsulating sediments have a profound impact onthe thermal structure. Heat refracts aroundinsulating coal and sediment layers, into adjacentzones of lower thermal resistance, resulting inlarge lateral variations in the subsurfacetemperature.In order to determine areas for potentialgeothermal resource exploration an assessmentof the temperature at depth is required. Short ofdrilling deep and expensive boreholes, currentmethods extrapolate down-hole temperaturesfrom shallow boreholes to 5km using shallowgeothermal gradients. This method relies on nonequilibratedtemperature information andgeothermal gradients, which fails to account forlateral variations in geology and thermalconductivity. Such extrapolations may differsignificantly to actual temperatures at depth andmy result in false target anomalies. 3D thermalmodelling provides a more representative analysisof the thermal structure, with the ability to setmodel parameters for lateral geological variationand define thermal characteristics.The work presented here compares the results ofextrapolated equilibrated and non equilibratedtemperature information from shallow boreholesto 5km depth and thermal modelling usinggeological models for target anomaly location.There is a significant difference in temperaturestructure at depth between equilibrated and nonequilibratedextrapolated temperatures as well asthermal modelled temperatures. Thesedifferences have significant implications fortargeting resource areas for geothermalexploration.Methods and ResultsDown-hole Temperature Measurements:Equilibrated and Non-equilibratedRoutinely temperature measurements inboreholes are not conducted for the purpose ofgeothermal exploration, instead they often aredesigned to assess groundwater aquifer locationsand cement setting during construction ofgroundwater bores. In exploration drillholestemperatures are recorded at the bottom of thehole during airlift tests or geophysical surveys(other than temperature). These results are takenimmediately following drilling, generally within24hrs, and are therefore considered nonequilibrated.On rare occasions temperatureinformation is collected several months or evenyears after drilling, this is considered equilibrated.To assess the impact of non-equilibrated andequilibrated temperature maps we collected alarge amount of temperature results from bothnon-equilibrated and equilibrated boreholes. Mostequilibrated data is collected in the field usingmethodology outlined in detail in Danis et al.(2010). As Beardsmore & Cull (2001) recommendequilibrated measurements can be recorded afterwaiting 3 times the drilling time length, ourtemperature results were divided accordingly. Ingeneral any measurements taken one month orgreater after drilling were considered asequilibrated results.Geothermal gradients were determined for eachequilibrated and non-equilibrated borehole using4

Australian Geothermal Conference 2010the 1D two layer extrapolation method (outlinedby Chopra & Holgate, 2006) and temperature at5km depth below surface contour map produced(Figure 1). All temperatures were corrected forclimatic variations based on Cull (1979). Whereonly a bottom hole temperature measurementwas recorded a surface temperature average of15°C was used to calculate the geothermalgradient. Where possible only temperatures frombelow 100m ground surface were used, to try andavoid diurnal/seasonal temperature influences.Figure 1: Shallow borehole temperatures extrapolated to 5km(a) equilibrated and (b) non-equilibrated. Location ofboreholes used shown as crosses.(a)(b)The first layer (sediments) is defined from ourSydney Basin geological model and used thecalculated geothermal gradient. The second layer(basement) is also defined from our geologicalmodel and uses the uniform geothermal gradientof 25°C/km so as to be comparable to the currenttemperature at 5km map.Figure 1 shows a distinctive difference betweenthe extrapolated equilibrated and non-equilibratedmeasurements for temperature observed at 5km.The equilibrated boreholes (Figure 1a) show lowtemperatures (60-120°C) in the central part of theSydney Basin, where sediment thickness isgreatest, whilst near the edges of the basin (i.eUlan and Singleton) temperatures areapproaching 200 - 250°C. The non-equilibratedbores also show a similar trend, however on aslightly different scale, with several elevatedanomalies in the generally colder parts of thecentral and southern Sydney Basin.The position of temperature highs (or in somecases lows) will shift depending on whetherequilibrated or non-equilibrated temperatures areused. Three highs appear around the Sydney-Campbelltown-Wollongong region in the nonequilibrateddata which are not prominent in theequilibrated map. The high near Singleton is alsomore localised with the non-equilibratedmeasurements.The pattern in temperature distribution of lowsover the centre and highs on the edge of theSydney Basin, were an expected feature. They tiein with the fact that the coal measures provide athermal insulator, thus temperatures measuredabove coal measures would be expected to becooler than those measure in or below coalmeasures. On the edge of the basin heatrefracting around the coal measures produces theelevated temperatures.Previous thermal modelling of the GunnedahBasin (Danis et al., 2010) showed basinarchitecture and the refraction of heat around thecoal interval to be major controlling factors in thethermal profile. Therefore to better understand theimpact of extrapolating shallow thermalmeasurements to depth we created 15 thermalmodel profiles along the lines shown in Figure 1.Thermal Modelling: UnderworldThermal models were developed using the finiteelement code Underworld. The code solves thenon-steady state heat equation with internal heatsources in two dimensions. Distinct layers fromour 3D geological model are imported as differentmaterials into the code, with constant temperaturetop and bottom boundary conditions. The thermalproperties for each material layer are outlined inTable 1, and are aggregates of measurements oneach unit/rock type. In addition there is one quasi-5

Australian Geothermal Conference 2010material called ‘air’. This top air layer has a largeconductivity and its purpose is to allow directthermal coupling of the varying topographicsurface with the top boundary condition. The sideboundary conditions are reflecting, and an extra10km has been added to either side of the modelprofile to avoid any reflecting edge effects.Rock TypeDensity(kg/m 3 )Table 1Conductivity(W/m-K)HeatProduction(µW/m 3 )Basement 2700 3 2Mafics 2950 3 0.5Sediments 2460 2 1.25Coal Interval 1900 0.3 1.25The main free parameter is the bottomtemperature condition at 12km, which wasextrapolated from the National Temperature at5km map (e.g Budd, 2007) to be ~350°C. Heatproduction in the basement is taken fromrepresentative Lachlan Fold Belt granites in theOZCHEM database. The model considers thermalconduction only, it doesn’t take into accountadvective effects or the effects of varying surfacetemperature conditions.The model boundary conditions were calibratedusing limited available equilibrated temperaturefrom shallow boreholes (Figure 2).equipment, whilst profiles 2, 11 and 13 have beengeophysical logged. The precision of the Hobologging equipment is 0.37°C at 20°C and thegeneral precision of geophysical temperaturelogging equipment is 0.1°C (from AUSLOG). Theuncertainty of the climate correction applied to theequilibrated measurements is approximately ±5°C at the surface, ±3°C at 500m and ± 1°C at1km. We therefore allow an uncertainty buffer of ±10°C in our measured temperatures whencomparing them to the modelled geotherms. It isimportant to note in profile 2 the greenmeasurements are the equilibrated results andare lower than expected due to cleaning of theborehole before logging. The red and bluemeasurements represent non-equilibratedtemperatures.In order to compare the results of extrapolatedtemperature measurements with thermalmodelling, a series of temperature profiles weretaken along a selected thermal model line. Herethe equilibrated and non-equilibrated temperaturemeasurements were extracted at 500m and 5kmdepth and compared to those of the thermalmodel profile at the same depths, as shown inFigure 3.Figure 2: Geotherm calibration checks using shallowequilibrated borehole temperatures for four selected profiles.Ground surface (green line) varies from model depth zerodepending on topography. Gray shaded area represents±10°C of the model geotherm.Figure 3: Thermal model Line 12 temperature output withextrapolated temperatures (green = equilibrated, blue = nonequilibrated)at (a) 500m and (b) 5000m below groundsurface and showing model geology.The equilibrated borehole temperatures (Figure 2)compare well with the model geotherms and arefrom boreholes in the Permian sequence.Equilibrated temperatures in profile 1 are fromfield measurements using Hobo loggingFrom Figure 3 a distinctive difference in anomalystructure can be seen between the 500m and 5kmdepths. In both cases the extrapolated6

Australian Geothermal Conference 2010equilibrated and non-equilibrated measurementsunderestimate the modelled temperatures. Evenat 500m, above the influence of the coalmeasures the extrapolated temperatures are stilllower than the modelled temperatures.At 5km the model temperature also shows verylittle variation as they are predominately in theLachlan Fold Belt basement. At 500m shallowsurface variations in the modelled temperaturesare observed and are most likely related to thegeology and/or influences of shallow groundwateraquifers.In both profiles (Figure 3) the effects of theinsulating coal interval on extrapolatedequilibrated or non-equilibrated versus modelledtemperatures can be seen. It appears that theextrapolation of shallow measurements, withoutany consideration of the geology and/or thermalconductivity, to depth propagates shallow surfacefeatures and produces false anomalies.To further illustrate the difference between theextrapolated and modelled temperatures Figure 4is a contour map of modelled temperatures at5km below ground surface for the Sydney Basin.Figure 4: Modelled temperatures at 5km depth below groundsurface. Thermal model lines are numbered 1 to 15, outlineof the Sydney Basin shown in grey and edge of continentalshelf is the offshore black line.Modelled temperatures (Figure 4) in the SydneyBasin at 5km are hottest where the thickest layersof sediment and coal intervals are. This is mostprominent around Sydney and north towards theHunter/Newcastle Coalfields, where sedimentthickness ranges from 3 to 4km, and a thick layerof basal volcanics are also present. Thetemperature anomaly observed wast of Singletonin the equilibrated extrapolation map appears in asimilar location with the modelled temperatures,but is less pronounced. The high on Line 12 northof Mossvale in the equilibrated extrapolation mapis not apparent in Figure 4. High basementtemperatures offshore of Newcastle in theNewcastle Syncline are a new feature observedwith the modelled temperatures. Where basementis shallow, i.e. the southern parts of the SydneyBasin, modelled temperatures at 5km are lowerthan those further north with greater sedimentcover, predominately because the coal interval isgenerally absent, removing the insulating effect,and sediment thickness is less than 1km.Modelled temperatures show high (210+°C)temperatures under areas of thick sedimentcover, whilst shallow extrapolated temperaturemeasurements often reflect the insulating effect ofthe coal interval (if above the coal interval) withtemperatures ranging from 60°C to 230°C.SummaryThese results show that the temperaturevariations observed in maps created fromextrapolated shallow borehole measurements arelikely to be inaccurate. The simple linearextrapolations shown here of both equilibratedand non-equilibrated show shallow temperaturemeasurements propagate near surface features todepth which, when compared to modelledtemperatures, are not true anomalies. Thereforegiven most of the measurements used in thecreation of temperature at 5km maps are takenfrom non-equilibrated boreholes extreme cautionshould be exercised when considering thetemperatures.However thermal modelling did show thatequilibrated measurements are an excellentcalibration tool. The modelled geotherms fit wellwith the shallow equilibrated measurements. Awell calibrated thermal model will better accountfor basin structure and changes in thermalconductivity than extrapolated temperature maps.Extrapolation doesn’t take into account thethermal effects of basin architectural structure andshould be avoided as a geothermal explorationtool. Not only is there a risk of the targetanomalies being false positives, possible positivetargets may be incorrectly located as a result ofrefracted heat from the basin structure.ReferencesBeardsmore, G.R., Cull, J.P., 2001, Crustal heatflow: a guide to measurement and modelling.Cambridge University Press 324p.Budd, A.R., 2007, Australian radiogenic graniteand sedimentary basin geothermal hot rock7

Australian Geothermal Conference 2010potential map (preliminary edition), 1:5 000 000scale. Geoscience Australia, Canberra.Cull, J.P., 1982, An appraisal of Australian heatflow data. BMR Journal of Australian Geology &Geophysics v.7, p. 11-21.Danis, C., O’Neill, C., and Lackie, M., 2010, 3Darchitecture and upper crustal temperatures of theGunnedah Basin. Australian Journal of EarthSciences 57, 483-505.8

Australian Geothermal Conference 2010South Australian Geothermal Research CentreM. Hand and Y. NgothaiSACGER, IMER, The University of Adelaide, South Australia, 5005.*Corresponding author: martin.hand@adelaide.edu.auThe emergence of the Australian GeothermalSector has focussed attention on the potential ofthe Australia’s non volcanic thermal systems toprovide both direct and indirect energy. Howeverthe challenges are significant, and require acoordinated effort between the private,government and R&D sectors to focus expertiseand bring projects to demonstration. TheDevelopment Framework and TechnologyRoadmap (DRET, 2008) outlines a clear plan forproject development. A key aspect of thisroadmap is consultation between the privatesector and research organisations that prioritisesresearch gaols that directly aid the sector. In thiscontext there are now a number of researchcentres within Australia that are seeking to form acohesive national R&D effort to assist the sector.This paper provides an overview of the SouthAustralian Geothermal Research Centre(SACGER), which is hosted at the University ofAdelaide, and provides a compliment to thenational overview presented by Long et al (thisvolume).Keywords: South Australian GeothermalResearch Centre.South Australian Centre for GeothermalEnergy Research - SACGERThe South Australian Centre for GeothermalEnergy Research is based at the University ofAdelaide within the Institute for Minerals andEnergy Resources (IMER),www.adelaide.edu.au/imer. The centre is fundedfrom the South Australian Renewable EnergyFund, established with a grant of AU$1.6M overtwo years from 1 July 2009. However initiation ofexpenditure was delayed until July 2010. TheUniversity of Adelaide is providing AU$400,000over the same period. Within IMER the SouthAustralian Centre for Geothermal EnergyResearch forms part of a broader R&D portfolioon energy that includes the Centre for EnergyTechnologies which has a focus that includes theenergy conversion and hybrid technologies whoseaim is to optimise the geothermal energydelivered to the surface.The SACGER research program has threeprincipal foci which compliment the researchprograms of other centres around the country andbuild on strengths in subsurface modelling andcharacterisation to foster reliable estimates andefficient development of geothermal resources.Program 1: Development ofElectromagnetic tools to monitor andassess geothermal reservoirbehaviourThis program is designed to the development oftechnologies and methods to make surface-basedmagnetotelluric (MT) surveys useful in (1)predicting the presence of fractured reservoirsahead of the drill bit and (2) monitoring fluid flowsin Engineered Geothermal Systems.Under certain conditions, MT surveys can indicatethe presence of naturally occurring hot fracturedrock targets that are susceptible to reservoirenhancement with fracture stimulation to createEngineered Geothermal Systems. MT surveyscan provide measurements that indicate thenature of changed fracture networks and postreservoir enhancement with hydraulic fracturestimulation. The development of MT technologiesand MT interpretation methods is identified as apriority for research in the AustralianGovernment’s Geothermal Industry DevelopmentFramework and its associated GeothermalTechnology Roadmap as requiring immediateattention. The University of Adelaide isinternationally recognised for its expertise in theuse and development of MT techniques, and iswell positioned to become an international leaderin the development of electromagnetically basedtools for the geothermal industry.The initiation and propagation of the undergroundfluid-reservoir is dependent on severalparameters, including the stress regime, geometryof pre-existing faults, and reservoir lithologies.Currently, measurements of shear-wave splitting(SWS) is deployed to define EGS stress regimes,while refined measurement of seismicity (naturaland induced during EGS field operations) is astandard approach to determine the location ofthe enhanced extent in the subsurface of EGSreservoirs (by locating seismic events associatedwith the creation of fractures during EGSoperations). Because MT surveys can, undercertain conditions, detect fluid flow patterns,advances in MT technologies and interpretationmethods hold promise to foster more accuratedelineation of EGS reservoirs, and models of fluidflows in EGS reservoirs. Locating MTmeasurement stations alongside micro-seismicdetectors is a way to better understand therelationship between the two datasets and9

delineate the location of fracture zones in EGSfields.The proposed research leverages on knowledgeobtained from a pilot project concluded by theUniversity of Adelaide with Petratherm Ltd. Initialstudies and field surveys indicate close sitespacing of the MT sites and high data quality arecrucial for successfully monitoring the reservoirdevelopment.The technologies and approaches developed willbe readily transferable to other potential EGSsites in South Australia and more globally.Program 2: Fluid-rock interactions ingeothermal reservoirsFracture and flowAustralia’s three flagship geothermal projects areall located in South Australia (Geodynamics’Cooper Basin project; Petratherm’s Paralanaproject; and Panax’s Salamander (Penola)project. The Geodynamics and Petrathermprojects represent two of the world’s mostsignificant Engineered Geothermal Systems(EGS) projects. Both projects entail theenhancement of naturally fractured rocks withhydraulic fracture stimulation. Additionallydependent on the reservoir nature, HotSedimentary Aquifer systems such as thosetargeted in Panax’s Salamander project may alsorequire stimulation.Reliably predictive flow modelling of EGSreservoirs is a recognised critical researchchallenge that is driving a peak internationalconsortia (the International Partnership forGeothermal Technologies; the IPGT) to focus onimproving 3-D modelling of reservoir performancethrough the life of a geothermal field life e.g.models that couple heat exchange, pressure,rock-fluid interaction, fluid flow dynamics,geomechanical properties and other geothermalreservoir and geothermal well parameters overlife-cycle production history. The aim of researchin SACGER will be to:(1) Improve 3D models with new informationfrom the EGS projects, and work incollaboration with the R&D expertise inAustralia and internationally. This work willhave implications for the risk management ofinduced seismicity by characterising thegeometry of naturally existing fracturepatterns, and improving knowledge of the insitustresses and potential overpressureregimes.(2) Develop laboratory procedures using coreflooding under stress aimed at predictingformation damage in production/injectionwells and removal of well damage.Australian Geothermal Conference 2010Program 3: Chemical reactivity ingeothermal systemsArea-specific research to mitigate fouling andgeothermal reservoir (pore and fracture) blockingis essential for efficient geothermal energydevelopment projects. This research will build onpilot studies using a hydrothermal cell atrepresentative reservoir temperatures (100-250 o C) to study fluid-rock interactions ingeothermal reservoirs. However, the pilot studycell operated at pressures (roughly 50 bar, thesaturation pressure of fluid) well below pressures(~ 300 bar) typical of EGS reservoirs in SouthAustralia. More realistic reservoir conditions willbe simulated by the creation of a hightemperature, high pressure flow through cell toexperimentally explore the risks of fouling andreservoir blockage at conditions that mimicsconditions in the South Australian EGS projects.The research will also allow the chemicalprocesses in HSA systems to be examined. Thiswill allow prediction of permeability for down-dip(thermally viable) targets HSA targets as well asscaling and fouling.Development and application of tracers andsmart particles to track fluid flow withinreservoirsTracer technologies are a critical factor incalibrating reservoir performance models in allgeothermal projects and can be used todetermine the reactiveConservative tracers for characterizing inter-wellreservoir flow processes are:- Thermally stable to 350 o C- Detectible to low parts per trillion- Environmentally friendly and non-toxicReactive tracers for characterizing fracturesurface areas in injection/backflow tests are:- Thermally decomposing- Reversibly sorbing- Possessing contrasting diffusivitiesIn addition to these research programs is workcurrently underway within the within the Centre forEnergy Technologies that is experimentingverification of underground cooling for efficientthermal cycles and examination of lowtemperature thermal processing using geothermalenergy aimed at optimising geothermal energyinvestments.For more information seewww.adelaide.edu.au/geothermal/or contact the centre: imer@adelaide.edu.au ormartin.hand@adelaide.edu.au10

Australian Geothermal Conference 2010Application of magnetotellurics in geothermal reservoircharacterizationGharibi 1 Mehran, Hearst 1* Robert B. , Retallick 2 Trent and Killin 1 Kevin1 Quantec Geoscience Ltd., 146 Sparks Ave., Toronto, ON, Canada M2H 2S42 Quantec Geoscience Pty. Ltd., Unit 2/77 Araluen Street, Kedron, Qld., 4031*Corresponding author: rhearst@quantecgeoscience.comIn this paper we examine Magnetotelluric (MT)data and analysis from a few different styles ofgeothermal resource. The geothermal resourcesexamined include both shallow (2000m) scenarios. The implicationswith respect to survey and limitations imposed bysurvey design on the interpretation of the resultsare discussed. The results of 1-D, 2-D and 3-Dinversions are compared and discussed in termsof their vertical and spatial resolutions and thereliability of the results in conjunction with thegeology of the study areas. The continuedimprovement in the power and affordability inmulti-core PC-platform computing allows for therelatively rapid inversion of MT data in 3-D. In thepast it had been necessary to invert in 3-D withonly a subset of the original dataset and with alimited number of frequencies (often 2000m) andfor reconnaissance type surveys whereby a largeamount of ground must be covered in a limitedamount of time and hence relatively wide siteintervals are required (> 500m). Detailed profilingcan also be accomplished.The past decade has seen the development ofarray type systems whereby a large number ofMT sites can be deployed in a rapid fashionallowing for a detailed investigation to becompleted in a short period time. Often thesesystems include the ability to acquire othercomplementary geophysical data sets (e.g. DCresistivity, IP chargeability, TEM). An example ofsuch a system is illustrated by Figure 2. Ingeneral, these array type systems can beextremely effective in the delineation of nearsurface (

Australian Geothermal Conference 20101-D and 2-D inversions and the data must bemitigated or removed before the inversion. On thecontrary, the 3-D inversion uses impedance dataand is capable to handle this type of data; makingthe inversion a robust tool to produce a realisticrepresentation of the subsurface.In this discussion we contrast the differencesbetween not only the 1-D, 2-D and 3-D inversionsbut also the effect of variations in MT field sampleintervals and spacing on the effectiveness of thevarious inversion methods on the ability to discernand characterize geothermal reservoirs.Figure 2: Example of an Array Type MT Acquisition SystemInversion MethodologiesInterpretation of the MT data is performed usingthe maps of true resistivity of the subsurface.Inversion algorithms in one-dimension (1-D), twodimension(2-D), and three-dimension (3-D) areused to invert the apparent resistivity and phasedata in to the maps of true resistivity of thesubsurface. A simple layered subsurface structuregenerally can adequately be reproduced using the1-D inversion. In the case of more complex 2-D or3-D structures, the MT response will be affectedby lateral variations in resistivity. Consequently, a2-D or 3-D inversion algorithm is required to allowthe lateral resistivity variations.In 1-D earth assumption, the 1-D inversion of theMT data produces a resistivity-depth profile foreach MT site. The results represent a first orderapproximation of the resistivity variations withdepth using a layered-earth model. Often theseinversion results are presented in pseudo-sectionform as “stitched” 1-D inversion sections.If there are lateral variations in the resistivity ofthe subsurface along one direction only(perpendicular to the strike) then a 2-D inversionand interpretation is required. A cross-section ofthe true resistivity variations perpendicular to theassumed strike direction is created in the 2-Dinversion and is used in interpretation.For more complex geological structures a 3-Dinversion is essential to adequately describe theresistivity variation of the subsurface. This isusually the case when mapping the geologicalsettings hosting a geothermal system. In this caseno simplifying assumption is made in terms ofproperty of the MT data and dimensionality of theunderlying subsurface. In highly heterogeneousenvironments MT phase data often exhibit an outof-phase(phase-wrap) behaviour; caused by thecomplexity of the current paths in the subsurface.Modelling of these data is essential in order toresolve the heterogeneity of the subsurface. Thiskind of data, however, cannot be modelled usingReferencesBaars, R.M., Owens, L., Tobin, H., Norman, D.,Cumming, W., and Hill, G., 2006, Exploration andtargeting of the Socorro, New Mexico direct usegeothermal exploration well, a GRED III Project:Proceedings, Thirty-First Workshop onGeothermal Reservoir Engineering, StanfordUniversity, Stanford, California, January 30-February 1, 2006, SGP-TR-179.Bahr, K., and Simpson, F., 2005, PracticalMagnetotellurics, Cambridge University Press.Brophy, P., 2007, A brief classification ofgeothermal systems: Geophysical Techniques inGeothermal Exploration Workshop, GRC AnnualMeeting, 2007, Sparks, Nevada.Constable, S.C., Parker, R.L., and Constable,C.G., 1987. Occam’s inversion - A practicalalgorithm for generating smooth models fromelectromagnetic sounding data. Geophysics, 52(3), 289-300.de Lugao, P.P., and Wannamaker, P.E., 1996.Calculating the two-dimensional magnetotelluricJacobian in finite elements using reciprocity.Geophysical Journal International, 127, 806-810.Marquardt, D.W., 1963. An algorithm for leastsquaresestimation of non-linear parameters. J.Sot. Ind. Appl. Math., 11, 431–441.Nabighian, M.N., 1987. Electromagnetic Methodsin Applied Geophysics, Volume 2: Application(Parts A and B). Society of ExplorationGeophysicists (SEG), Tulsa.Orange, A.S., 1989. Magnetotelluric explorationfor hydrocarbons. Proceedings of the IEEE, 77,287-317.Rodi, W., and Mackie, R.L., 2001. Nonlinearconjugate gradients algorithm for 2Dmagnetotelluric inversions. Geophysics, 66, 174-187.Siripunvaraporn, W., Egbert, G., Lenbury, Y., andUyeshima, M., 2005. Three-DimensionalMagnetotelluric: Data Space Method. Physics ofthe Earth and Planetary Interiors, 150, 3-14.12

Ussher, G., Harvey, C., Johnstone, R., Anderson,E., 2000, Understanding the resistivities observedin geothermal systems: Proceedings WorldGethermal Congress 2000, Kyushu-Tohoku,Japan, May 28-June 10, 2000, p.1915-1920.Vozoff, K., 1972. The Magnetotelluric method inthe Exploration of Sedimentary basins.Geophysics, 37, 98-141.Wannamaker, P.E., Stodt, J.A., and Rijo, L., 1987.A stable finite-element solution for twodimensionalmagnetotelluric modeling.Geophysical Journal of the Royal AstronomicalSociety, 88, 277-296.Wight, D.E., 1987. MT/EMAP Data InterchangeStandard, Revision 1.0. Society of ExplorationGeophysicists (SEG). (Document available at theSEG web site: www.seg.org).Australian Geothermal Conference 201013

Australian Geothermal Conference 2010Hutton Sandstone – a viable geothermal developmentDavid Jenson*, Kerry Parker, Ron Palmer, Bertus de GraafPanax Geothermal Ltd. PO Box 2142, Milton, Qld 4064*Corresponding author: djenson@panaxgeothermal.com.auPanax Geothermal Limited (Panax) holds fourGeothermal Exploration Licences (GEL’s) in thenorth east of South Australia, namely GEL’s 220,221, 281 and 502. These licences overlap theGreat Artesian Basin and in particular the HuttonSandstone (“Hutton”), a known prolific reservoir.It has been demonstrated in existing wells andextrapolated using seismic data that the Huttonare recognized as a very porous and permeableaquifer and can be found at depths greater than2,000m and temperatures of up to 145 °C.Reviewing core samples in the region suggeststhat there is a high likelihood for transmissibilityup to 90-100 Dm.The Hutton Project is based on producing aninitial artesian flow of hot water sufficient togenerate about 1 MWe. Although greater flowscould be produced, either by artesian flow or bypumping, capable of generating up to 5-6 MWeper production well, a smaller development ismore appropriate whilst the market evolves.Initially, the focus is on replacing diesel or gasgeneration in the local oil fields, small settlementsand pastoral leases.This study is based on small scale generationbased on multiple (between 2 and 5) proprietaryPureCycle 280kW units (580kW – 1400kW grosspower output) receiving flows of artesian hotwater between 135 and 145 °C from a single150mm nominal diameter well. The maximumflow that may be required is about 51 kg/s whichis less than half of the natural artesian flow thatcould be expected.Geothermal potential of Panax’sHutton Sandstone project licencesPanax Geothermal Limited (Panax) holds fourGeothermal Exploration Licences (GEL’s) in theCooper Basin region, north east of SouthAustralia, see Figure 1. GEL’s 220 and 221 are inthe Nappamerri Trough and GEL’s 281 and partof 502 are in the Patchawarra Trough. Both areasare covered by the hydrogeological GreatArtesian Basin.Figure 1. Panax's geothermal licences located in NE SouthAustralia. All four licences are in the hydro-geological GreatArtesian BasinBeardsmore (2009) assessed GEL 281 for itsgeothermal potential and determined a “MeasuredGeothermal Resource” of 11,000PJcomplemented by an additional 30,000 PJ of“Indicated Geothermal Resource”, see tablebelow. This resource classification was inaccordance with the Australian Code forReporting of Exploration Results, GeothermalResources and Geothermal Reserves, 2008edition.The ability to positively assess the significantgeothermal resource potential is due to the longhistory of petroleum exploration with numerouswells and seismic surveys (Figure 2).This highly encouraging resource classification,the third in Australia and second resourceclassification of a Hot Sedimentary Aquiferconfirmed the high potential of geothermal energydevelopment in sedimentary basins.Table 1. Geothermal Resource, GEL281Measured Indicated Inferred Total[PJ] [PJ] [PJ] [PJ]11,000 30,000 0 41,00014

Australian Geothermal Conference 2010Figure 2. GEL281 - outlines of Measured GeothermalResources showing existing petroleum wells.In the following Geothermal SystemsAssessment, Cooper et al. (2009) found that theHutton Sandstone has potential for low enthalpygeothermal energy generation. Cooper points outthat the temperature modeling may have beenunderstating the temperature as the data points atthis stratigraphic level are sparse and that it ispossible that advective (lateral) heat flow mayoccur that has not been included in the modelling.Cooper’s positive assessment of the GreatArtesian Basin in GELs 220 and 221 confirmedtogether with earlier work over GEL 281 thatPanax’s GEL’s in the region represents a largearea with a significant geothermal resourcePanax commissioned Hot Dry Rocks Pty Ltd(“HDRPL”) to review depth and temperaturepotential of GEL’s 220, 221 and 281 (GEL 502was not granted at the time of this review). Walsh(2009A) found in this brief review that preliminarytemperature models using ‘best fit’ indicated thatthe temperature at the top of the Hutton where itis at its deepest was estimated to be in the orderof 142°C (see Figure ?) and that the top of HuttonSandstone in the area of maximum thickness isinterpreted to lie between 2,150 and 2,200mdepth.Walsh (2009B), finally, also assessedpermeability assessments from 246 core samplescollected from 12 petroleum wells in a depthrange of 1,700m to 2,000m in the area andconcluded that the permeability appeared to behigh, with the bulk of the values in theencouraging 100 to 2,000 mD region.Interestingly, though data presented by Walshshowed a wide range of permeabilities, they doindicate that even a less thick horizon of theHutton Sandstone has the ability to sustainreasonable flow rates.Concluding, the Hutton sandstone in GEL 220 ismodeled at a depth in excess of 2,000m and witha temperature at its top of 142 deg C.Permeability estimates from 12 wells with porosityand permeability estimates of the HuttonSandstone indicate a bi-modal population with thelarger population in the vicinity of 50mD to2,000mD. If a conservative value of, say, 250mDis chosen assuming not all parts are productive,then at the estimated thickness of HuttonSandstone of 380m, there is a potential of areservoir with a transmissivity of 90-100 Dm.These highly encouraging estimates suggest thatthe Hutton sandstone represents a strongprospect for developing a low risk hot sedimentaryaquifer, geothermal energy.Great Artesian Basin and HuttonSandstoneThe Eromanga basin was deposited as an uninterruptedsequence from the Early Jurassic toMid-Late Cretaceous, see Figure below. Thebasal units are mainly non-marine to marginalmarine, however, marine influences dominate thesilts and mudstones of the Marree Subgroup.Non-marine sandstones were deposited in thefinal stages (Winton Formation). Of petroleumand geothermal interest are the terrestrialsandstones of the Hutton, Namur, and CadnaowieFormations. These units, in particular theCadna-owie Formation, constitute much of theGreat Artesian Basin.Figure 3. Preliminary 1D temperature prediction for HuttonSandstone based on temperature data from nearby wells15

Australian Geothermal Conference 2010Figure 5. GAB extent and ground water temperatures (afterHabermehl, 2001).supply, as is the case at Birdsville, as well as forstock use. Ground temperatures at the well headsrange from 30deg C to 100deg C. Groundwatertemperatures from 1,880 water wells tapping in tothe Cadna-owie aquifer, above the deeper HuttonSandstone, from are presented in Figure 5.Figure 4. General stratigraphy of the Eromanga and LakeEyre Basin sequences in the Cooper Basin (Gravestock et al,1998.The Jurassic Hutton Sandstone is a sandstonedominatedunit that has been interpreted to havebeen deposited in a continent-scale, braidedstreamenvironment, and to have beenunconfined across a broad alluvial plain extendingover much of Queensland and South Australia.The Canterbury Plain in New Zealand isconsidered as a close modern analogue. TheHutton Sandstone is well-known as a regionalaquifer in a number of fields in both SouthAustralia and Queensland.Habermehl (reference?) notes that the GreatArtesian Basin is a well known aquifer with some4,700 artesian water bores drilled in to it. Most ofthese artesian wells tap in to the Cadna-owieaquifer, the uppermost artesian aquifer of thismulti-layered sequence. Groundwater in theCadna-owie aquifer is of good quality, containingbetween 500mg/l to 1,000mg/l total dissolvedsolids. It is suitable for domestic, town-waterGeothermal developmentThe Hutton Project aims to produce an artesianflow of hot water sufficient to generate about 1MWe. Although greater flows could be produced,either by artesian flow or by pumping, capable ofgenerating up to 5MWe per production well,sufficient demand to justify such a developmentdoes not exist and so the focus is on smallerplants to replace diesel or gas generation in thelocal oil fields, small settlements and pastoralleases.Artesian flows at adequate temperature can beproduced from the Hutton Sandstone at welldepths down to 2,500m. Artesian pressures of atleast 10bar are expected which will be easilysufficient for a hot water flow of about 30 kg/swhich will be required for 1 MWe GeothermalPower Plant.It is intended that the cooled exhaust water will bereinjected back into the same formation. For thesmall flows that are required, it is anticipated thatno or minimal reinjection pumping power will berequired. ie if pressures can be maintained in a16

Top Hutton Sandstone > 2000 mbslclosed circuit, it is expected that a thermosiphonwill be established.The Hutton Sandstone can be found at suitabledepths over relatively wide areas of the GELs andso there is a possibility that a number of standaloneplants based on a production/reinjectionwell pair could be developed.The development scenarios considered are basedon the use of two to five PureCycle © units asmarketed by Pratt and Whitney Power Systems.These units have a nominal 280 kW grosscapacity and are designed to be water cooled. Itis believed that water cooling, as opposed to aircooling will be required because of the highambient dry bulb temperatures encountered in theregion.In order for a small project to be economic, it willbe necessary to adopt water well drilling practicesas opposed petroleum drilling practices.Australian Geothermal Conference 2010Potential marketAlthough extensive research has not beencompleted into available electrical loads that couldbe satisfied by geothermal electricity producedfrom the Hutton Sandstone, the project location isabout half way between the townships of Moombaand Innamincka, where it is known that there is asignificant reliance on diesel and gas firedgeneration. In addition grazing homesteads arealso known to rely on diesel generation. Finally,there are several gas plants and pipeline nodesrequiring energy, also representing a small localmarketAlthough the gas plants and pipeline node pointsare known, maps of these have been removedfrom public access to prevent them from beingtargets of a terrorist act. However, studying thepipeline network in Figure 7, one can indirectlyA 150mm diameter x 1,500m deep water welldrilled at Clifton Hills (outside the license area)flowed at approximately 155 kg/s under artesianpressure and had an overall cost of less than$1m. It is believed that similar technology couldbe utilized to drill to depths of 2-2500m for anoverall cost of less than $2m per well.MoombaHutton Sst > 2000m(red areas)210 0 mbsl2 0 0 mbs lGEL 220(black)Location and InfrastructureGEL 220 has been estimated to be host thehottest Hutton Sandstone in Panax’s geothermallicence portfolio. A strong candidate for the HuttonSandstone Geothermal Well is the south westernpart of GEL 220, as indicated in Figure 6. Thislocation is in a large area of deep Hutton, close toa road and outside the Innamincka RegionalReserve. There are Native Title agreementscovering this are as is the case for all of Panax’sgeothermal licences in Cooper Basin.Figure 7. Pipelines in Cooper Basin. Gas production plantsare inferred at end points of pipelines (blue) and nodes atintersection of pipelines .locate gas plants and pipeline nodes,representing small local electricity markets.Innamincka and Moomba townships are alsoannotated in this figure. It is interesting to note thehuge number of wells that have been drilled in toCooper Basin, represented by black dots on thismap. Each of these wells were preceded byintense studies of seismic, 2D and more recentlyalso 3D seismic. These data, available free ofcharge of PIRSA, represent a huge investmentand to Panax, a significant advantage inassessing the Hutton Sandstone, as has beenreviewed in previous sections.It is believed that a mini 3 phase power grid couldbe established to which users in the vicinity wouldhave access, either with a 3 phase connection ora Single Wire Earth Return(WER)line.Figure 6: GEL 220 (black outline) with the areas whereHutton Sandstone is interpreted from regional data to bedeeper than 2000m (red areas). The estimate of the totalarea of Hutton Sandstone in excess of 2000 m is about 300km2. Light purple indicate pipe lines with Moomba gas Plantto the west of GEL 220.17Development Time FrameIt is envisaged that a small scale plant will be ableto be brought on line twelve to eighteen monthsafter the spudding of the production well.

Australian Geothermal Conference 2010PureCycle © units are built on a production line inthe US and are shipped as a fully assembled skidand therefore a minimum amount of site work isrequired. A separate cooling system will need tobe engineered and installed.The production and injection wells are eachestimated to take about 3 weeks to drill.The overall project schedule is estimated to beabout 24 months.Cost estimates and assumptionsCapital costs have been estimated fromquotations received from plant suppliers anddrilling companies and from detailed in-houseestimates for balance of plant and otherinfrastructure requirements.In respect of operating costs, a small plant of thetype initially envisaged will not be permanentlymanned. Remote or automatic operation will berequired with routine maintenance performed onan as required basis. PureCycle plants are verysimple with plant overhauls required on a semiannualor longer interval basis.Modelling resultsThe investigation of various engineering scenariosfor using the Hutton Sandstones have focused onsmaller scale development (500kW – 1.5 MW)which could be used to replace existing diesel orgas fired generation in the adjacent oil fields,small settlements and pastoral leases.Even smaller sub-optimal developments can stillbe very attractive as small package plant can beused and parasitic pumping power is muchreduced because the artesian pressure togetherwith thermo siphon effects negates the need forproduction or reinjection pumps.PureCycle units as marketed by Pratt & WhitneyPower Systems could be ideal for this application.These are package units with a nominal rating of280 kW gross and multiple units could be installedor added to match electricity demand.Slim (150mm diameter), Deep (1,500m) waterwells drilled into the GAB have yielded significantartesian flow rates (155 l/s) at temperatures ofabout 100 deg C at surface. These wells can bedrilled very economically (

Australian Geothermal Conference 2010“Petroleum geology of South Australia, Volume 4:Cooper Basin, Department of Primary Industriesand Resources, South Australia, Report book,98/9. 1998.19

Australian Geothermal Conference 2010Investment in common-use network infrastructure to triggerincremental expansion in geothermal energy generating capital stockAshok A. Kaniyal 1,3* , Jonathan J. Pincus 2 , Graham (Gus) J. Nathan 1,31 School of Mechanical Engineering, The University of Adelaide SA 50052 School of Economics, The University of Adelaide SA 50053 Centre for Energy Technology, The University of Adelaide SA 5005*Corresponding Author: Ashok A. Kaniyal, ashok.kaniyal@adelaide.edu.auThis paper analyses the potential of a novelapproach to avoid the cost of expensivetransmission infrastructure in connecting remotegeothermal power suppliers to market through acommunity of data centres. A strategy forinvestment in a common-use facility to triggerincremental growth in a complementarycommercial alliance is presented. The commercialalliance consists of investors in the ‘common-use’fibre-optic cable data network, a community ofdata centres, and geothermal energy resources.A real options approach to capital investmentdecision making is used to define the conditionsunder which investment in the common-usefacility (which in economics is a ‘local publicgood’), will be commercially justifiable.Excess capacity in the fibre-optic networkresource is what gives it the attributes of a ‘localpublic good’. A critical assessment shows thatinvestment in such infrastructure could play asignificant role in triggering the initial and lateradditions to installed geothermal energy capitalstock, to generate and supply power to a networkof data centres in a remote area. Theinfrastructure needs of data centres are chosenfor this assessment because these facilities canbe collocated with the geothermal resource, andtheir energy consumption is well suited to ademonstration-scale plant.The real options analysis of resource allocation inthis hypothetical commercial setting isunderpinned by the stochastic representation ofthe future growth in demand for energy from anetwork of data centres, in the micro-gridcommunity described. The results from thisanalysis provide an understanding of the criticalconditions under which investment in the fibreoptic network resource could take place.geothermal energy production capacity can bederived. This derivation provides an optimisedtime series profile of incremental expansion incapital stock, to satisfy the discrete units ofstochastic demand for energy in this micro-gridsetting. The benefits of this assessment thus lie inthe ability to understand the likely time period overwhich significant economies of scale in generatingcapacity can be achieved. This henceforthenables a characterisation of the point at whichinvestment in electricity grid infrastructure (or thelike) may be justified. The process associated withthis stochastic approach to analysing investmentdecisions enables the consideration of thesystematic risks in this much larger but end-goalinvestment horizon.Finally the study presents public policyrecommendations regarding State involvement inestablishing capacity in this network resource, toassist the development of the necessaryeconomies of scale in geothermal energyproduction over time.Keywords: Real options theory, networkinfrastructure, local public good, geothermalenergy.ReferenceDixit, A., Pindyck, R.S., 1994. Investment underuncertainty. Princeton University Press, PrincetonNJ.The real options approach is an appropriate toolto assess investment strategy in this context,because it incorporates uncertainty in consumerdemand, and analyses the resultant flexibility inthe timing of investment decisions. In contrast, thediscounted cash flow method suggests that aninvestment decision is made only once, i.e. at thecommencement of the discounting period.Once a commercially justifiable investmentpathway for the local public good is defined, aprofile of incremental expansion in the installed20

Australian Geothermal Conference 2010Ultra deep drilling technologies for geothermal energy productionDipl. Ing. Ivan Kočiš, PhD., Dipl. Ing. Tomáš Krištofič, Dipl. Ing. Igor KočišGeothermal AnywhereBratislava, Slovak republicAbstractSubstance of the presented research anddevelopment is the technology for realization ofgeothermal boreholes approximately 6 to 10 kmdeep. In such depths, rock temperature reaches200°C to 400°C practically all over the world. Inthis depth, thermal capacity of 1 km 3 is about1017 J (when cooled down by 100°C); this iscomparable with annual demand of Slovakia. Withreal consumption of approximately 20 to 40 MWthe capacity of the same borehole would besufficient for several hundreds of years. Deepgeothermal energy is available anywhere,represents the most serious candidate for a baseload 24/7 electricity production, there are no CO 2emissions and is scalable according to the localneeds. The main obstacle for utilization of the vastenergy reserves is the price rising exponentiallyper 1 km of a borehole depth in present-daydrilling technologies. Costs of such boreholescannot be considered as a basis of repeatedprojects in most territories of the world. Thus,intense research and development of new deepdrilling technologies is necessary.There are more than 20 researched new drillingtechnologies, but none of them gave at presentdesired economic and technical parameters for 6-10 km depths. The technology must becompatible with extreme conditions at such depth(pressure up to 1000 bar and temperatures morethan 400°C) and very hard rock like granite.Presented radically innovative concept of drillingand transport of material is protected by patentsand being a basis of a know-how based uponinnovative utilization of proven technologies. Themethod is suitable primarily for extreme pressureand temperature conditions, and has all prerequisitesto meet the basic requirement – linearprice growth with borehole depth. The conceptwas first time publicly presented duringConference of the 6th EU Framework Programmeproject "Engine", aimed at ultra-deep geothermalenergy and related research within future 7thFramework Program projects. The conceptaroused considerable attention and interest inparticipation in future projects.Keywords: Deep drilling, Bitless drilling, Noncontactdrilling, Pulsed plasma drilling, EGSIntroductionThe renowned MIT (USA) study "The Future ofGeothermal Energy – Impact of EnhancedGeothermal Systems (EGS) on the United Statesin the 21st Century" (2006) points out theessential importance of the developing aneconomical deep geothermal drilling technology.With current drilling technologies, drilling pricebelow 3 km rises exponentially with depth. Thus,finding a drilling technology with which the drillingprice rise would be approximately linear withincreasing well depth is an important challenge.In his presentation, Jefferson Tester, a co-authorof this study, characterizes the key requirementson new fast and ultra-deep drilling technology asfollows: the price of drilling rises linearly with depth, the possibility to make vertical or inclinedboreholes up to 20 km deep, casing formed on site in the well.There are under way more than 20 researchefforts solving innovative drilling technology suchas: laser, spallation, plasma, electron beam,pellets, enhanced rotary, electric spark anddischarge, electric arc, water jet erosion,ultrasonic, chemical, induction, nuclear, forcedflame explosive, turbine, high frequency,microwave, heating/cooling stress, electric currentand several other.No one from these until now proved to beeffective in severe conditions and no one issolving the problem in complex including theenergy and material transport from 6 to 10 kmtechnically and economically. We selected drillingapproach based on integration of pulsed electricalplasma with water jet assisted processes.What is different from current drillingtechnologies?Process of rock disintegration is the mostimportant activity when drilling ultra-deepborehole, but not the only one contributing to theexponential price rise. Therefore a completelynew approach should be considered. That is alsosubstantial difference between GA’s drillingconcept and mentioned drilling technologies.GA drilling approach is based on integration ofpulsed electrically generated plasma with waterjet assisted processes. Advantages of the GA’stechnology are as follows:21

Australian Geothermal Conference 20101. Controllability of the process of disintegrationand efficient generation of media fordisintegration2. Pulsed mode of the media towards thedisintegrating rock3. High heat flow4. Pulsed power technologies - Timecompression5. Optimized cuttings6. Cuttings washout using thermo lift and waterjet interaction7. Casing while drilling – vitrified rock8. Dynamic temperature stress on the rock9. Control and sensoric system in extremeconditions10. Restart of drilling process in extreme highpressure conditionsWe identified nine different physical processesused for disintegration of the rock. Under researchis a synergic exploitation of selected processes,also using supercritical fluids.Total energy costs of drilling represent only afraction of the total cost and our approach is tomelt the rock with a safe temperature tolerance.What is different compared tospallation technology?Spallation process is the closest technology, buthas several issues and drawbacks. We canmention small temperature interval in whichspallation process takes place, dependence of theinterval from rocks and non-effective process ofspallation in some types of rock.Total energy costs of drilling represent only afraction of the total cost. Our approach is to meltthe rock and manage cuttings with optimized size.The presented drilling technology is based on aset of research project, development of criticalaggregates, proving correctness of the conceptand preparation for the industrial developmentstage which should serve the application aspect.In cooperation with several Universities andAcademy of Sciences teams are solved thedifficult functional and material problems forextreme in-field conditions. The purpose of theproject is to research all critical parts up to thelevel of functional laboratory samples. Sufficientbackground data for further industrialdevelopment should be provided. On the basis ofthe results obtained in technological parts of thesolution, the project is assumed to givetechnological proof of the whole concept.How is that possible?The innovation concept is able to integrate theproven technologies compatible with hightemperatures and pressures using thecombination of electrical energy source for both –producing the needed form of the energy and forenabling the thermal lift buoyancy effect fora cutting transport.The rock crushing technologies will be used ina synergy of electrical discharge, pulsed plasmaand water jet in appropriate alternatives which willbe the subject of research and experimental work.This radical innovative method avoids the inherentlimitations of conventional rotary drilling andmaterial transport.Complexity of the problem, harsh environment inwhich the targeted technology shall work, and theradical innovation concept requires gradualapproach to the research, development andrealization of the whole system.Ultra deep drilling technology -geothermal energy revolution enablerThe realization of the Ultra deep drillingtechnology will open the way to thermal energyresources to be used for production of electricenergy without greenhouse gas emissions andwith zero operation costs of fuel. The currentstatus of electric energy production is based uponfossil fuel, it is not compatible with reduction ofemissions, and, moreover, is not resistant againstincrease of prices of crude oil and gas in theworld, and of those of the nuclear fuel, which arederived from them.Deep geothermal systems open the way toreplacing coal and gas capacities by geothermalcapacities within the future 20 years.Figure 1: Extra and ordinary geothermal conditions –difference brings the ultra-deep drilling technology.Such resource of the energy will adequatelysatisfy the demand of energy, which has thefollowing properties: It is local, stable, scalable upon need, with nolosses caused by long distance transfer. It does not produce any emissions, and issafe.22

Australian Geothermal Conference 2010No supply of energy carriers is needed, "fuel"costs are zero and unaffected by world pricesvariation, crisis developments and geopoliticalsituations.Creates numerous secondary applications,primarily for small and medium enterprises;moreover, creation of job opportunities isenhanced.ConclusionThe revolutionary patented concept of the ultradeepdrilling and transport of crushed rockmaterial upwards leads to new implementationpossibilities.It is a radical abandonment of the classic drillingtechnologies with connected tubes in long stringswith the inefficient energy transport to the drillingsite at the bottom of the well and complicated,inefficient pumping of the crushed material to thesurface.This new drilling technology is a result of the longtime effort to solve the problem of theexponentially growing price in relation to theborehole depth and thus enabling the realexploitation of the geothermal energy for theelectrical power generation.ReferencesTester, Jefferson W.; et. al. (2006), The Future ofGeothermal Energy, Impact of EnhancedGeothermal Systems (EGS) on the United Statesin the 21st Century: ISBN 0-615-13438-6,http://geothermal.inel.gov/publications/future_of_geothermal_energy.pdfJefferson Tester et al., US Pat 5771984,“Continuous drilling of vertical boreholes bythermal processes: rock spallation and fusion"Kobayashi et al., US Pat 6870128, “LASERBORING METHOD AND SYSTEM”Hirotoshi et al., 1999, "Pulsed Electric Breakdownand Destruction of Granite", published in Jpn. J.Appl. Phys. Vol.38 (1999), 6502-6505Dipl. Ing. Ivan Kočiš, PhD., Dipl. Ing. TomášKrištofič, Ultra deep drilling technologies forgeothermal energy production, InternationalGeothermal Days Conference, May 2009, CastaPapiernicka, Slovakia23

Australian Geothermal Conference 2010A Preliminary Study on Na-Cl-H 2 O-Rock Interactions of the HotFractured Rock Geothermal System in Cooper Basin, South AustraliaGideon B. Kuncoro 1 , Yung Ngothai 1* , Brian O’Neill 1 , Allan Pring 2 , Joël Brugger 21 The University of Adelaide, School of Chemical Engineering, North Terrace, Adelaide, SA, 50002 South Australian Museum, North Terrace, Adelaide, SA, 5000*Corresponding author: yung.ngothai@adelaide.edu.auA preliminary study was undertaken to observethe fluid-rock interaction in Na-Cl-H 2 O system.Samples of drill cuttings from a borehole 5 kmdeep from Habanero 3 well were contacted with250 ppm sodium chloride solution in athermosyphon induced loop reactor at 250 o C and40-50 bar. The experiments were carried out in aTitanium flow through cell for 1, 7, and 28 days at250 o C and 40 bars). The fluid was replaced every24 hours in order to accelerate the dilution rate toobserve which minerals are reactive (soluble) andto mimic a condition which uses fresh water ortreated water from a precipitation tank that isreinjected to the fracture. Fluid and rock sampleswere analysed prior to, and after circulation toobserve the dissolved minerals and mineralogy.Water analysis was performed using ICP-MS, androck analyses were conducted using an opticalmicroscope, SEM and XRD. The experimentalresults indicated that mineral dissolution wasmore rapid in the early stages of the experiment.This may be a consequence of the dissolution offiner rock particles. SEM observations showedevidence of etching of the mineral surfacesconsistent with partial dissolution. XRD resultsindicated that the feldspars (NaAlSi 3 O 8 andKAlSi 3 O 8 ) in the rock had completely dissolvedand small amount of quartz remained. ICP-MSanalysis on the water sample confirmed that somemineral dissolution has occurred. Theconcentrations of the elements increased withtime.Experiment running for 14 days is currently beingundertaken and the results will be used tocomplement this study. Future work will involvereaction path modelling to predict mineralprecipitation.Keywords: hot fractured rock, fluid-rockinteraction, Na-Cl-H 2 O system, geothermal,Cooper Basin, Habanero, thermosyphonIntroductionA study of the interactions between rock andcirculating fluid is essential to determine thechemical changes and mineral alteration of ageothermal system. This study was undertaken toobserve mineral dissolution due to the interactionwith Na-Cl-H 2 O system. Existing reservoirgranites are currently in equilibrium with thesurrounding ground water. The injection of freshwater to extract thermal energy from the host rockwill alter the ground water chemistry and thus thefluid-rock equilibrium. While recirculating thebrine, partial chemical dissolution or mineralspecies alteration may occur. This may potentiallyincrease the dissolved solids such as silica andother metals in the water. These dissolutionproducts of the components have differentequilibria, which will be a function of temperatureand pressure, thus precipitation or scaling of pipework and closure of fractures in the granite bodyare possible. The saturation of metals in fluid isvolume-dependent, where very small volumes offluid may require slight under-cooling beforeprecipitation occurs. Clearly, characterization offluid geochemistry is important in the evaluation ofthe performance of geothermal systems (Grigsbyet al., 1989). Moreover, understanding thechemical interactions due to the injection of fluidinto hot granite is crucial for problems concerningclogging by precipitation and heat loss caused bydissolution (Azaroual and Fouillac, 1997).The study of fluid-rock interaction will allowdetermination of mineral alteration and dissolutionof minerals due to the circulating water. To date,there are a number of geochemical modellingcodes therefore thermodynamic data bases areavailable and improved to predict equilibriumconditions and dissolution-precipitation rates(such as EQ3/6, TOUGH-REACT, SOLVEQ-CHILLER, SOLMINEQ, GWB, etc). However,fundamental processes associated with mineraldissolution and precipitation, pressuretemperaturegradient remain poorly understood(Marks et al., 2010). Unfortunately, it is impossibleto generalize the actual field experience ofmineral deposition in geothermal systems into oneconsistent theory due to the vast chemical andoperational variation between field sites(Robinson, 1982). Therefore, although there havebeen a number of studies on fluid-rockinteractions for different geothermal sites(Rimstidt and Barnes, 1980; Robinson, 1982;Posey-Dowty et al,. 1986; Savage et al., 1987;Grigsby et al., 1989; Savage et al., 1992;Azaroual and Fouillac, 1997; Yangisawa et al.,2005; Tarcan et al., 2005; Castro et al., 2006;Marks et al., 2010), and studies of rock-chloridesolutions (Ellis, 1968; Dove and Crerar, 1990;Icenhower and Dowe, 2000) these are probablynot directly applicable to the hot granite-basedgeothermal systems in the Cooper Basin.24

Australian Geothermal Conference 2010(a)(b)Figure 1: (a) Diagram of the flow through cell (b) Photograph of the geothermal flow through cellExperimentalSamples of drill cuttings from the Habanero 3 wellwere provided by Geodynamics. The sampleswere ultrasonically cleaned and analysedusing,the scanning electron microscope (SEM) atthe Adelaide Microscopy Centre to observe bothsurface and cross-sectional area. Powder X-raydiffraction (XRD) analysis was carried out byAMDEL to observe the mineralogy. A preliminaryset of experiments to observe the fluid-rockinteraction were performed in batch mode. Thedrill cutting was used as the rock sample. Therock sample was crushed and sieved to give 100– 200 μm size fraction and approximately 0.7grams of rock were used in each batch. Thesample was enclosed in a pre-weighed wirebasket (approx. 50 μm mesh size) made fromstainless steel and placed in the sample holder ofthe cell. The cell was filled through a valve in thebase of the cell with 250 ppm sodium chloridesolution and 90 ml was drained from the cell toallow for fluid expansion upon heating. Thevelocity of the thermosyphon circulating fluid isestimated to be approximately 0.1 m/s. The fluidrockinteraction periods were 1, 7, and 21 days. Atconclusion of this interaction period, the rocksample was dried (105 o C for 48 hours), cooled ina desiccator, and weighed to determine anyweight loss.The rock samples were analysed using SEM andXRD, and the water samples were stored andpreserved with 4% w/w of 1M nitric acid untilanalysis by ICP-MS to identify the dissolvedmetals. The silica content was quantified usingheteropoly blue method (HACH, 2009).Unfortunately, the anions were unable to be-analysed since the concentrations of HCO 3 andSO4 2- in the sample are lower than the detectionlimit of the current analysis.Results and DiscussionPreliminary results from the geothermal cellexperiment illustrate that the dissolution of the25

Australian Geothermal Conference 2010Weight loss of rock (%)12010080604020images are presented in Figure 3. The SEMimages of the rock sample surfaces of feldsparsbefore and after the experiments are shown inFigure 4. It can be seen that pitting has started tooccur in 1 day circulation period. At the conclusionof 7 days circulation, the surface of feldsparsexperienced severe pitting and breakage. By theend of 28 days circulation, the feldspar hascompletely dissolved and some quartz andundissolved minerals remain.00 5 10 15 20 25 30Time (d)H2ONaClFigure 2: Change in weight loss of rock (% w/w) versus timerock (% w/w) increases with time (Figure 2). Itshows that the dissolution appears more rapid inthe early stages of the experiment. This isprobably due to the dissolution of the morereactive minerals (feldspars). It is seen that moreminerals dissolved in the Na-Cl-H 2 O systemcompared to pure water (reverse osmosis treatedwater). The presence of higher electrolyteconcentration (e.g., Cl - ) in Na-Cl-H 2 O systemincreases the dissolution rate of some mineralssuch as quartz (Dove and Crerar, 1990). Thelatter stage of the experiment approached anequilibrium state. However the fluid is notnecessarily saturated with minerals (feldspar), butdue to the exhaustion of the soluble phase(s) andslower dissolution rate of the remaining phase(s).Note that the circulating fluid is replaced every 24hours.Rock AnalysisMineralogical analysis was performed on the drillcuttings from Habanero 3 well using x-rayfluorescence (XRF) and x-ray diffraction (XRD).The major elements and their host minerals of therock are quartz (SiO 2 ), albite (Na,Ca)AlSi 3 O 8 , andmicrocline (KAlSi 3 O 8 ) (Pring, pers. comm. 2009).The XRF analysis was carried out by the Schoolof Earth and Environmental Sciences, AdelaideUniversity (Table 1). The XRD analysis wasperformed by AMDEL (given in %wt) andconverted to mg for better understanding. Theresults show clearly that rapid dissolution hasoccurred. Silicon internal standard was added toobserve any possible amorphous, unfortunatelythe amorphous content were unable to bequantified by Amdel. Hematite, magnetite,anatase, chlorite and goethite may be accessoryminerals due to corrosion of the stainless steelreservoirs and the stainless steel wire basketused in the experiment.SEM images of the surface particles wereobtained using the secondary electron detector(SE) of the XL30 in Adelaide Microscopy. TheTable 1 XRF analysis of the rock sampleStarting Sample % WeightSiO 2 77.27Al 2 O 3 10.62Fe 2 O 3 1.46MnO 0.08MgO 0.09CaO 0.60Na 2 O 2.54K 2 O 4.45TiO 2 0.07P 2 O 5 0.01SO 3 0.15LOI 2.10Total 99.44Table 2 XRD analysis of the rock sampleStartingsampleAfter 1dayAfter 7daysAfter 28daysMineral composition (mg)QuartzSiO 2 273.5 252.2 119.6 7.6Albite(Na,K,Ca)Al (1-2) Si (3-2) O 8 206.9 184.0 138.9 0.0K-FeldsparKAlSi 3 O 8 125.6 115.9 57.9 0.0Muscovite(Na,K,Ca) 2 (Al,Fe,Mg) 4-6(Al,Si) 8 O 20 (OH,F) 4 59.1 54.5 34.7 0.0PyroxeneXY(Si,Al) 2 O 6 14.8 13.6 0.0 0.0Silicon*Si 44.3 47.7 27.0 1.4Chalcopyrite and or CalciteCuFe 2 and or CaCO 3 7.4 6.8 3.9 0.0SideriteFeCO 3 7.4 6.8 0.0 0.0HematiteFe 2 O 3 0.0 0.0 0.0 3.7MagnetiteFe3O4 0.0 0.0 0.0 1.0AnataseTiO 2 0.0 0.0 0.0 0.4Chlorite(Mg,Fe) 3 (Si,Al) 4 O 10 (OH) 2 .(Mg,Fe) 3 (OH) 6 0.0 0.0 0.0 2.3GoethiteFeO(OH) 0.0 0.0 3.9 0.6Amorphous 0.0 0.0 0.0 2.5Total sample 739.1 681.5 385.7 19.4SEM analyses of the cross section of the rocksamples were also conducted. Results are given26

Australian Geothermal Conference 2010in Figure 4. These images were obtained usingthe backscattered electron detector (BSE) andanalysed using the energy dispersive x-rayspectrometer (EDAX). The results suggest thatalbite feldspar is the most reactive (soluble)mineral, followed by microcline and quartz. TheBSE image of the sample from 28 day circulation(Figure 4D) confirms that only quartz and otheraccessory minerals were observed, and nofeldspars were present.ABFigures 5 and 6 show the concentration ofelements versus time from the ICP-MS analysis,and Figure 7 show the silica concentration on thecirculating water after 1, 7, and 28 days of fluidrockinteractions.Cumulative Concentration(ppm)120.00100.0080.0060.0040.0020.000.000 5 10 15 20 25 30time (d)AlKFigure 5: Concentration of Al and K in experimental liquidversus time12000.00CFigure 3: SEM images of the rock sample surface (a) startingrock (b) after 1 day experiment (c) after 7 days experiment (d)after 28 day experimentDCumulative concentration(ppb)10000.008000.006000.004000.002000.000.000 5 10 15 20 25 30time (d)Fe Mg Mn Ca BaFigure 4: BSE images of the rock cross sectional area (a)starting rock (b) after 1 day experiment (c) after 7 daysexperiment (d) after 28 day experimentFigure 6: Concentration of Fe, Mg, Mn, Ca and Ba inexperimental liquid versus timeABAccumulativeConcentration (ppm)1800160014001200100080060040020000 5 10 15 20 25 30time (d)SiO2CWater AnalysisThe initial pH of the water was 5.5. This did notchange significantly in all batch experiments.DFigure 7: Concentration of SiO2 in experimental liquid versustimeWater analysis results showed that there wasenhanced leaching of the minerals, indicated bythe increased element concentration in the liquidphase. The rapid reaction is probably the27

consequence to the dissolution of more reactive(soluble) phases in the rock sample. It is seen thathigh concentration of silica, aluminium andpotassium were dissolved in the liquid due todissolution of albite, microcline and quartz.Unfortunately the sodium concentration could notbe analysed accurately due to the high initialconcentration of sodium in the circulating fluid(250 ppm NaCl).Mass BalanceA simple mass balance was undertaken toobserve the sample loss to the circulating fluid.Since the sodium cannot be analysed accuratelyat this stage, it was assumed that the dissolvedconcentration would be similar to the dissolutionof Albite Feldspars in pure water. Table 3 showsthe system mass balance. A short circulating timewould result in less sample loss. The sample lossmay be due to the replacement of circulating fluid,since at the time of reinjection, the liquid mustoverflow through the sample and exiting the cell toremove excess air (any air bubble may disrupt theflow in the thermosyphon).Table 3 System mass balanceNot including the concentration of the anions mayalso be another cause of the unclosed massbalance.ConclusionDissolved(calc),mgDissolved,(wateranalysis),mgThe mineralogical analysis showed that the majorcomposition of the rock is quartz, albite andmicrocline. The fluid-rock interaction experimentshowed that dissolution reactions occurpredominantly in the early stages (to 7 days) ofthe experiment due to the dissolution of reactiveminerals (feldspars). The dissolution reactionbegins to decrease due to the exhaustion of thereactive mineral and proceeds with the dissolutionof the less reactive phase. Results of XRDshowed that feldspar experienced a clearreduction with time, followed by quartz.Liquid analysis showed higher concentrations ofNa, K, Si, and Al compared to other analysedelements. These results suggest that thedissolution reaction occurs primarily for thefeldspars.The dissolution of feldspar would suggest theimprovement of the permeability (however, notnecessarily improving the fracture where mineralAustralian Geothermal Conference 2010deposits occur) and also suggest of potentialsilica scaling (if circulating fluid is near neutralpH). The concentration of silica 1400 ppm (from0.7g rock sample) was released to the liquidphase throughout the experiment.Future WorktimeInitial,mgRemaining,mg% wt.Loss,mg1 745.2 681.5 63.7 8.55 41.30 22.407 734.4 385.7 348.7 47.48 271.00 77.7028 737.7 19.4 718.3 97.37 427.95 290.3528An experiment is currently being conducted for 14days period to obtain more understanding on thedissolution (Figure 2). Sodium concentrationwould need to be accurately analysed. Thecurrent work has used 250 ppm sodium chloridesolution for the fluid-rock interaction and does notquite represent the actual interaction.Experiments with Increased amount of rocksample (increasing the rock/water ratio) are beingcarried out to observe any changes in equilibriumor saturated condition. A dissolution andprecipitation model will be generated using thedata from previous experiments (pure water) andthe current study. The model will be used topredict the dissolution rate and precipitation rateof the actual fluid-rock interaction.AcknowledgmentsThe research described in this paper has beensupported by The Department of PrimaryIndustries and Resources of South Australia(PIRSA) and Geodynamics. Theauthor would like to thankGeodynamics for supplying the drillcuttings, and the South AustralianMuseum and Adelaide Microscopy forthe equipment access. The authorswould like to thank John Stanley fromthe School of Earth and Environmental Sciences,Adelaide University, Benjamin Wade and AoifeMcFadden for assistance in ICP-MS analysis,Shaun Chan and Kevin Li for their assistance inthe experiments.ReferencesAzaroual, M., and Fouillac, C., 1997,Experimental Study and Modelling of Granite-Distilled Water Interaction at 180 o C and 14 Bars:Applied Geochemistry, V. 12, pp. 55-73.Castro, M. R., Lopez, D. L., Reyes, J. A., Matus,A., Montalvo, F. E., Guerra, C. E., 2006, ExpectedSilica Scaling from Reinjection Waters AfterInstallation of a Binary Cycle Power Station atBerlin Geothermal Field, El Salvador, CentralAmerica: GRC Transactions, V. 30, pp. 487-492Dove, P. M., and Crerar, D. A., 1990, Kinetics ofQuartz Dissolution in Electrolyte Solutions using aHydrothermal Mixed Flow Reactor: Geochimica etCosmochimica Acta, V. 54, pp 955-969Ellis, A. J., 1968, Natural Hydrothermal Systemsand Experimental Hot-Water/Rock Interaction:Reactions with NaCl solutions and trace metalextraction: Geochimica et Cosmochimica Acta, V.32, pp. 1356-1363

Australian Geothermal Conference 2010Grigsby, C. O., Tester, J. W., Trujillo JR., P. E.,and Counce, D. A., 1989, Rock-Water Interactionsin the Fenton Hill, New Mexico, Hot Dry RockGeothermal Systems I, Fluid Mixing and ChemicalGeothermometry: Geothermics, V. 18, pp. 629-656.HACH, 2009, DR/2010 PortableSpectrophotometer Procedures Manual.http://www.hach.com/fmmimghach?/CODE%3A4930022286|1, p 767 – 772, accessed September2008Icenhower, J. P., and Dove, P. M., 2000, TheDissolution Kinetics of Amorphous Silica intoSodium Chloride Solutions: Effects ofTemperature and Ionic Strength: Geochimica etCosmochimica Acta, V. 64, pp. 4193-4203Posey-Dowty, J., Crerar, D., Hellerman, R., andClarence, D. C., 1986, Kinetics of Mineral-WaterReactions: Theory, Design, and Application ofCirculating Hydrothermal Equipment: AmericanMineralogist, V. 71, pp. 85-94.Pring, A., 2009, personal communication.Rimstidt, J. D., and Barnes, H. L., 1980, TheKinetics of Silica-Water Reactions: Geochimica etCosmochimica Acta, V. 44, pp. 1683-1700.Robinson, B. A., 1982, Quartz Dissolution andSilica Deposition in Hot Dry Rock GeothermalSystems, Thesis, Cambridge, MassachusettsInstitute of Technology.Savage, D., Cave, M. R., Milodowski, A. E., andGeorge I., 1987, Hydrothermal Alteration ofGranite by Meteoric Fluid: An Example from theCarnmenellis Granite, United Kingdom:Contributions to Mineralogy and Petrology, V. 96,pp. 391-405.Savage, D., Bateman, K., and Richards, H. G.,1992, Granite-Water Interactions in a FlowthroughExperimental System with Applications tothe Hot Dry Rock Geothermal System atRosemanowes, Cornwall, U.K.: AppliedGeochemistry, V. 7, pp. 223-241.Tarcan, G., 2005, Mineral Saturation and ScalingTendencies of Waters Discharged from Wells(>150 o C) in Geothermal Areas of Turkey: Journalof Volcanology and Geothermal Research, V.142, pp. 263-283Yangisawa, N. Matsunaga, I., Sugita H., Sato, M.,and Okabe T., 2005, Scale Precipitation duringCirculation at the Hijiori HDR Test Field,Yamagata, Japan: Proceedings WorldGeothermal Congress 2005, Antalya, Turkey, 24-29 April 2005.29

Australian Geothermal Conference 2010Geothermal Energy in the Perth Basin, AustraliaComparisons with the Rhine GrabenAdrian Larking 1 , Gary Meyer 21 & 2 Green Rock Energy Limited, 6/38 Colin Street, West Perth, Western Australia 6005, AustraliaAbstractUnlike the Rhine Graben where commercial scalegeothermal power and direct heat projects have beenoperating for some time, geothermal developments inthe Perth Basin are only at the gestation stage. Overthe next two years Green Rock Energy Limited plansto carry out a commercial demonstration project inPerth which is likely to be the first commercial scaledirect heat project of its type in Australia. This has onlybecome possible since July 2009 when Green RockEnergy along with the University of Western Australiawas awarded the first rights to explore for and producegeothermal energy in the state of Western Australia.This followed the introduction of the State’s firstgeothermal legislation in 2008 when the Petroleum Actwas amended to include geothermal energy.This direct heat project is designed to prove theconcept of recoveering geothermal energy from hotsedimenary aquifers in the central Perth Basin todirectly cool and heat buildings in the PerthMetropolitan area. To assist the Company to achievethis objective in December 2009 the AustralianGovernment announced that it was offering GreenRock Energy Limited a grant of $7 million from itsGeothermal Drilling Program to assist the funding ofthe Company‘s project to air-conditon the maincampus at University of Western Australia in the PerthMetropolitan area. The funding agreement was signedin September 2010. One week later the WesternAustralian State Government announced that GreenRock Energy was to be offered a further grant of $5.4million for the same project. Funds from bothGovernments will be provided in stages and eachGovernment requires the Company to provide thebalance of required funds from other sources.Financing of geothermal exploration projects inAustralia may differ from typical funding of projects inGermany. Most funds in Australia are obtained fromshareholders in companies listed on the AustralianSecurities Exchange.Both the Rhine and the Perth basins are favourablylocated near infrastructure. The Rhine Graben is thelocus of a substantial portion of Germany’s populationand industrial muscle and is a major transport hub.The Perth Basin straddles Australia‘s south westerncoast where most of the State of Western Australia’spopulation of 2.3 million and power infrastructure arelocated. This favourable location close to marketstogether with Perth Basin’s geothermal energyresource potential presents the opportunity to supplyenergy for electricity production and direct heat useincluding air-conditioning and desalination of water.Both basins share one key thing in common: theirgeothermal systems have little surface expression atthe surface hence drilling is required to map the subsurfaceheat resources. Knowlege of the extent anddistribution of geothermal energy in the Perth Basinand its exploration and development is only at an earlystage and may be 20 years behind the Rhine Graben.In both grabens the existing knowledge was mainlyderived from the petroleum industry where there is ahistory of petroleum production. Petroleum explorationand production started much earlier in the Rhine, withthe Forst-Weiher Field discovered between 1934 and1936. Production ceased there by the 1960’s butcontinues at present in the Perth Basin. Petroleumdrilling commenced in the Perth Basin in the 1950sand resulted in the discovery of more than 13 oil andgas fields. Data derived from petroleum wells arerather sparsely distributed and focused in limited areaswhere petroleum fields have been discovered, mainlyin the northern Perth Basin. Only 250 petroleum wellshave been drilled and 23,985 km of 2D seismic and2,838 km 2 of 3D seismic acquired in the Perth Basin.In contrast, over 5,000 wells have been drilled in theRhine but around 90% of them were less than 600metres depth.Upper Rhine GrabenPerth BasinBoth the Rhine Graben and Perth Basins are N-Strending sediment filled extensional rifts. The Perthrifting is over 200Ma older and has a much thickersediment fill. The Perth Basin is a 1,000 km long rift orhalf graben containing a thick sequence of sedimentsin places up to 15 kilometres deep. This rifting30

Australian Geothermal Conference 2010produced the series of deep, north-south trendingbasins along the western margin of the Yilgarn Craton.Sediment deposition commenced with extensionduring the Permian (~290 Ma) and continued to theEarly Cretaceous (~138 Ma). The Basin was upliftedduring the final separation of Western Australia andgreater India during the breakup of Gondwana.Rifting in the 300km long Rhine Graben commencedmuch later in the Middle Eocene. Graben sinking andshoulder lifting continues to this day with the maximumsediment fill being around 3.4km in the Upper RhineGraben.There are substantial geological differences betweenthe Rhine Graben and Perth Basin, in particular themuch younger age, shorter graben, thinnersedimentary pile thickness of the former and currenttensional and seismically active tectonic stress regimeof the Rhine Graben compared to the compressionaland relatively seismically inactive regime of the PerthBasin. Even so there are important similarities in theirstructural setting, extensional tectonic origins, thicksedimentary successions which overlie hot crystallinebasements, high heat flows and evidence of theexistence of substantial geothermal energy resourcepotential.now overprinted by an essentially compressionalsetting as the Australian plate migrates NW with arotational component at the relatively fast geologicalrate of nearly 7 cm per year. The contemporary stressregime for the Perth Basin has been interpreted to bea transitional reverse to strike-slip faulting stressregime with an approximate east-west maximumhorizontal compression direction. Stress field data forthe Perth Basin are derived mainly from petroleum wellborehole breakouts and drilling-induced tensilefractures from 34 measurements as shown in theWorld Stress Map. This is important for targetting wellsto tap critically stressed faults and fracture directions.Perth Basin Stratigraphy &Tectonic stagesFinal breakup of Gondwanaleading to dextral strike-slipdeformation (transfer zones)& basin inversionGeological MapsFrom Song & Cawood, 1999From Thomas & Brown 1983Both basins have vertical structural asymetry with thesedimentary fill being geater in the eastern flank thanthe western flank and higher graben shoulders on theeast flank in the case of the Rhine. This is consideredto have important influences on regional ground waterflow in the Rhine. Similarly the hills marking the NStrending Darling Fault forms the eastern boundary faultof the Perth Basin and is also expected to have aninfluence on ground water flow.In the Rhine the current stress regime is strike slip withthe maximum horizontal stress being orientedapproximately NW/SE. In contrast to the Rhine, theearly extensional tectonic regime of the Perth Basin isGeothermal energy has been recovered to powerelectricity generation from both sediments and granitesin the Rhine Graben at Soultz sous Forêts andLandau. Heat flows of up to 150mW/m 2 have beenrecorded in both grabens. Thermal anomalies occuralong the western rim of the Rhine Graben. This is hasbeen attributed to enhancement of a conductive basalheat flow of 70-80mW/m 2 by convection. Some ofthese sites were mapped as early as 1929 where atemperature of 50ºC at depth of 400 metres wasknown at Soultz.At Landau convective hot spots have temperatures of55°C at a depth of 500m & 98°C at a depth of1,000m. There is considerable evidence for convectivewater flows in the Rhine Graben, particularly in thewestern flank where a pattern of adjacent hot and coldspots was found. This is generally considered to be thesurface expression of deep convective geothermalsystems. At Soultz there is a large reservoir with asimilar fluid chemistry. The same geothermal fluid isfound in different wells and in different lithologiesincluding sediments and granite basement.Geothermal water at Soultz has salinities of around100g/l and pH of 5.31

Australian Geothermal Conference 2010sequences in the Perth Basin. It is not known if thisreflects any deeper convection.In the Rhine Graben in recent years the emphasis hasshifted from targeting temperatures of around 200ºC atdepths of 5,000 metres where natural flow rates fromfracture permeability have proven to be inadequate totargetting shallower depths of 2,500m to 3,500 metreswhere tempertures of 150ºC to 160ºC and substantiallyhigher flow rates can be expected from natural faultsand fractures.With its thick sequences of sandstones, the presenceof thermal insulating shales and coals and high heatflows the Perth Basin has potential to housegeothermal energy resources in the form of hotsedimentary aquifers (HSA) and hot rocks. The sourceof the heat flows in the Perth Basin may be radioactivebasement rocks and radiogenic sedimentary horizons.The Perth Basin contains some of the worlds largestmineral sands deposits many of which are rich inradiogenic thorium minerals at shallow depths.Source: From Schwarz 2005 & Lichen 2005The highest known heat flows in the Perth Basin havebeen determined from petroleum wells in or nearproducing oil and gas fields in the northern PerthBasin. Green Rock holds nine Geothermal ExplorationPermits in this region near high voltage power gridsconnected to Perth’s major power markets. Heat flowsexceed 100mW/m 2 within these Permits.Temperatures of around 160°C are expected in hotsedimentary aquifers at depths less than 4,000mwhich is sufficient to generate electricity commerciallyat this location provided that sufficient geothermalwater flow rates can be achieved.Unlike Germany, electricity generated from geothermalenergy in Australia does not benefit from any feed intariff arragement and retail electricity prices in Australiaare substantially lower than Gemany‘s. In general thismeans that higher recovered temperatures or flowrates are required in Westen Australia to ensurecommercial viability. However generators of renewableenergy including geothermal energy are entitled underFederal legislation to sell Renewable EnergyCertificates (REC’s) in addition to selling the electricitythey generate. In the absence of feed in tariffs inWestern Australia electricity is sold to utilities viapower purchase agreements.Landau: Temperatures at 1000m (°C)Grid in kms (Gauss-KrügerSource: Charissé 2007It is early days yet but evidence is emerging ofconvective flows in the shallow sedimentaryWithin Green Rock‘s permits in the North Perth Basinpetroleum wells have proven good primary waterpermeabilities in sandstone reservoirs down to depthsof around 2,700 metres but there is some limitedevidence of substantially lower permeabilities atgreater depths where temperatures above 150ºCwould be expected. Much less is known about thepotential for good water flows from natural fault or32

Australian Geothermal Conference 2010fracture permeability as this has not been targetted bythe petroleum industry there.Drilling Water Well in Perth for Heating Aquatic CentresSource: Hot Dry Rocks Pty Ltd 2008In the central Perth Basin near the city of Perth heatflows are lower and temperatures expected at depthare likely to be insufficient for commercial generationof electricity but should be adequate for direct heatuses such as purification of water by distillation and forair-conditioning of buildings. Viability depends onattaining sufficient permeabilities from the aquifers.Within the Perth city area, where petroleum wells anddeep seismic data are lacking, most of the usefultemperature and permeability data have come fromdeep water bores, the deepest of which extend to onlyabout 1000 metres but have sufficiently high water flowrates from natural matrix or granular permeability. Onthe basis of petroleum wells outside the Perth Permitareas good permeabilities are expected in Perth atleast down to depths of 2,500 metres.Geothermal energy has been recovered from aquifersbeween 750 and 1,000 metres deep in Green Rock’sPermit GEP1 in metropolitan Perth. This geothermalwater is used to heat a number of major swimmingcentres including the Challenge Stadium AquaticCentre where the World Swimming Championshipshave been held twice in the past decade.To develop opportunities in GEP1, Green RockEnergy, plans to carry out a commercial demonstrationproject to recover geothermal energy from hotsedimentary aquifers beneath the main Perth campusat the University of Western Australia (UWA). Thisproject will be the first stepping stone to largercommercial developments in Perth.The demonstration project is designed to replace asignificant portion of the UWA’s electrical powered airconditioningwith geothermal powered absorptionchillers for its air-conditioning and heating needs.Absorption chillers, like the one shown below, arecommercially available in Australia.Commercial viability of this project at the UWA campuswill depend on obtaining adequate geothermaltemperatures and water flow rates from the sandstoneaquifers at depth. Geothermal water temperatures ofbetween 80⁰C and 100⁰C are required for commercialoperation of the absorption chillers which will providechilled water for the campus. Expected temperaturesat the target depths of 2.5km to 3km in Perth should beadequate as indicated by estimated heat flowsdetermined from temperature profiles measured indeep water bores and petroleum wells.33

Australian Geothermal Conference 2010ReferencesAbsorption ChillerFor this first project at the UWA campus, oneproduction and one injection well will be drilled toaround 2.5 to 3 kilometres deep to recover geothermalenergy from sandstone aquifers in the fluvialYarragadee Formation and Cattamarra CoalMeasures. Drilling wells to these depths should notpresent any particular difficulty as there is abundanthistory of petroleum wells drilled without significantproblems in the Perth Basin. Preparatory site work andwell design is underway to enable this drilling to becarried out in 2011. Seismic acquisition will becompleted to optimise well design and orientation priorto drilling.Chopra P.N. and Holgate, F. (Earthinsite Pty Ltd):Geothermal Energy Potential in Selected areas ofWestern Australia, a Report commissioned by WesternAustralian Department of Industry and Resources2007; and also – A GIS Analysis of Temperature in theAustralian Crust; Proceedings of World GeothermalCongress, Antalya, Turkey 2005, pp 7.Clauser, C and Villinger, H, Analysis of conductive andconvective heat transfer in a sedimentary basindemonstrated for the Rheingraben, Geophys. J 100,393-414, 1990Davidson, W.A... Hydrogeology and groundwaterresources of the Perth Region, Western Australia:Western Australia Geological Survey, Bulletin 142, 257pp, 1995Freeman, M.J., and Donaldson, M.J.: Geology of thesouthern Perth Basin and Margaret River wine district,southwestern Western Australia – a field guide:Western Australia Geological Survey, Record 2006/20,2006Genter, A., Baumgärtner, J., Cuenot, N., Graff, J J.,Kolbel, T., Sanjuan, B. The EGS Soultz Case Study:Lessons Learnt After Two Decades of GeothermalResearches. Second European Geothermal Review,Proceedings, Mainz June 21-23, 2010Gori, K.A.R.: Perth Basin’s Geothermal Resources,Australian Geothermal Energy Conference 2008,Extended Abstract. Geological Survey, Department ofIndustry and Resources, 2008Hot Dry Rocks Pty Ltd: Geothermal Energy Potential inSelected Areas of Western Australia, Perth Basin, pp270, 2008Hot Dry Rocks Pty Ltd: Western Perth Geothermalproject, Geothermal Exploration Stage 2, unpublishedreport for Green Rock Energy Limited, pp28, 2009,Mory, A.J., Haig, D.W., Mcloughlin, S., and Hocking,R.M.: Geology of the northern Perth Basin, WesternAustralia – a field guide: Western Australia GeologicalSurvey, Record 2005/9, 2005Mory, R.J, and Lasky, R.P.: Stratigraphy and Structureof the Onshore Northern Perth Basin, Report 46,Geological Survey of Western Australia pp 126, 1996Regenauer-Lieb, Prof. K, Horowitz Dr F, The PerthBasin Geothermal Opportunity, The University ofWestern Australia & CSIRO, p42-44, April 2007Schwarz M, Henk, A, Evolution and structure of theUpper Rhine Graben: insights from three-dimensionalthermomechanical modeling, Int J Earth Sci (GeolRundsch), 94: 732-750, 2005.34

Australian Geothermal Conference 2010Australia’s Geothermal Research ActivitiesA.D. Long 1* , A.R. Budd 2 , H. Gurgenci 3 , M. Hand 4 , C. Huddlestone-Holmes 5 , B. Moghtaderi 6 andRegenauer-Lieb, K. 71 Primary Industries & Resources SA, GPO Box 1671, Adelaide, South Australia, 5001.2 Geoscience Australia, GPO Box 378, Canberra, Australian Capital Territory, 2601.3 QGECE, The University of Queensland, Brisbane, Queensland, 4072.4 SACGER, IMER, The University of Adelaide, South Australia, 5005.5 CSIRO, PO Box 883, Kenmore, Queensland, 4069.6 ATC, The University of Newcastle, Callaghan, NSW, 2308.7 WAGCOE, ARRC, 26 Dick Perry Avenue, Kensington, 6151.*Corresponding author: alexandra.long@sa.gov.auThe challenges and technology needs ofAustralia’s geothermal energy industry aresignificant yet achievable. The sector is organisedand has a clear pathway through the stages ofproject development defined in the DevelopmentFramework and Technology Roadmap (DRET,2008). Collaboration is facilitated by the AustralianGeothermal Energy Group and its TechnicalInterest Groups, where research needs areprioritised by the industry and researchers. Thecapacity and capability to undertake the muchneeded research work will be met by geothermalresearch centres which work competitively yetcollaboratively, with complementary programs, toachieve research outcomes most efficiently. Thispaper will provide a guide to the researchactivities in Australia, outlining the programs andareas of expertise of the key research groupsworking in the geothermal arena; a summary ofcurrent projects; and how to find out moreinformation.Keywords: Geothermal, research, technologydevelopment.Centres of expertiseThe desire to develop Australia’s geothermalenergy resources to supply secure andenvironmentally sustainable energy has led to theimpressive growth of the Australian geothermalsector. The challenges facing the sector areknown, and the research and development needsof the industry as they strive to commercialise theresources are being met by a growing capacity atdedicated research centres, universities andresearch groups.Centres for geothermal energy research haveformed: at the University of Queensland, with theQueensland Geothermal Energy Centre ofExcellence (QGECE); in a joint venture of theUniversity of Western Australia, Curtin Universityand the CSIRO, the Western AustralianGeothermal Centre of Excellence (WAGCOE);and at the University of Adelaide, the SouthAustralian Centre for Geothermal EnergyResearch (SACGER). Geothermal energyresearch is also being supported at manyuniversities around Australia, GeoscienceAustralia and at the CSIRO. The followingsections will describe each of these centres andresearch groups and their capability.Geoscience AustraliaThe Australian Government announced theAU$58.9M Onshore Energy Security Initiative inAugust 2006, and as part of this, GeoscienceAustralia established a geothermal energy project.The project aims to improve the existingknowledge about the type and location ofgeothermal resources in Australia on a nationalscale. It also aims to encourage investment,exploration and exploitation of this energy sourcethrough provision of pre-competitive geosciencedatasets relevant to geothermal energy.To achieve these objectives, the geothermalproject:collects new heat flow data across Australiato better define and locate geothermalresources; uses (heat) source and (thermal) trapmodelling to identify potential Hot Rock andHot Sedimentary Aquifer systems;works to compile national datasets which maybe useful to the geothermal industry includinggroundwater temperatures, boreholetemperatures, rock thermal conductivities andgranite and sediment chemistry;uses these new datasets to produce a revisedestimate of Australia's total containedgeothermal resource; andprovides advice to government on geothermalresource issues, including the AU$50MGeothermal Drilling Program.For more information see www.ga.gov.auor contact the geothermal energy team:geothermal@ga.gov.au.Queensland Geothermal Energy Centre ofExcellence - QGECEThe Queensland Geothermal Energy Centre ofExcellence is based at the University ofQueensland and was established with anAU$15M grant from the Queensland Governmentand AU$3.3M of in-kind support from theuniversity. The centre commenced operations in35

Australian Geothermal Conference 2010January 2009 with a 5 year program. The centre’sresearch programs were developed to fill gaps inthe national and international geothermalresearch effort and have a focus on above groundtechnologies with the aim of quickening the paceof large-scale utilisation of hot rocks geothermalenergy in Australia. QGECE has four researchprograms: Power Conversion: developing technologiesto enable production of 50% more electricityfrom binary plants using the same subsurfaceinvestment;Heat Exchangers: development of naturaldraft dry cooling towers and other coolingsolutions to increase by up to 15% the netoutput of geothermal plants that use aircooledcondensers; Reservoir Geology: establish ageochemical/isotopic and geochronologicaldatabase and improve understanding ofgeothermal resources in Queensland anddevelop routine exploration tools for hot rockgeothermal systems; and,Transmission: research in to electricity gridinteraction with an emphasis on remotegeneration infrastructure.The current major projects include: Design and development of small (5-kWe)supercritical turbines for testing in theQGECE laboratory.Construction of a 100-kWe mobile geothermaltest plant with high-pressure capability to trysupercritical turbines. Characterisation of the heat-producinggranites in Queensland. The effect of ambient dust on air-cooledcondenser performance and design againstdust.Investigation of options for connecting remotegeothermal power generation to theQueensland grid.See also www.uq.edu.au/geothermalor contact the centre Director, Professor HalGurgenci: h.gurgenci@uq.edu.au.Western Australian Geothermal Centre ofExcellence - WAGCOEThe Western Australian Geothermal Centre ofExcellence is an unincorporated joint venturebetween the Commonwealth Science andIndustry Research Organisation (CSIRO), theUniversity of Western Australia, and CurtinUniversity. The centre was established inFebruary 2009 with an AU$2.3M grant over threeyears from the Western Australian governmentand substantial in-kind and cash contributionsfrom the centres’ members. The centre isproviding a scientific focus to the development ofthe geothermal industry in Western Australia,concentrating on the Perth basin, and buildingeducational programs around geothermal energy.The centre’s research focus is on HotSedimentary Aquifers and direct use of the heatproduced from these resources, as summarised inthe centre’s vision statement “To providegeothermal zero-emission desalinated water, airconditioningand power to our cities”.WAGCOE has three research programs: Perth Basin Assessment: Develop a rigorousscientific understanding of the geothermalresource in the Perth Basin;Above Ground Technologies: Identify anddemonstrate innovative applications of HSAgeothermal energy; and, Deep Resources: Provide a scientificframework for the potential exploitation ofdeep geothermal resources.Current major projects include: Research and development activitiesassociated with the direct use geothermalpowered supercomputer cooling system thatis part of CSIRO’s Sustainable Energy forSKA project (see below). This system targetshot sedimentary aquifers of the Perth Basin.Design of an additional sensor equipped deepresearch and monitoring well at the same sitefor research, education, training and longterm monitoring. Research and development activitiesassociated with the direct geothermally drivenMW th scale campus cooling project at UWArun by the leaseholders Green Rock Energyand UWA. Development of a novel desalinationtechnology with 30% yield boost from lowgrade geothermal waters of 65C and less. Acontainerised m 3 /day first generationprototype is sponsored by the AustralianGovernment National Centre of Excellence inDesalination located at Murdoch University.For more information see www.geothermal.orgor contact the centre business manager, SeanWebb: sean.webb@geothermal.org.au.South Australian Centre for GeothermalEnergy Research - SACGERThe South Australian Centre for GeothermalEnergy Research is based at the University ofAdelaide within the Institute for Minerals andEnergy Resources, www.adelaide.edu.au/imer.The centre was announced by the SouthAustralian Government as the first project to befunded from the South Australian RenewableEnergy Fund, established with a grant of36

Australian Geothermal Conference 2010AU$1.6M over two years from 1 July 2009. TheUniversity of Adelaide is providing AU$400,000over the same period.The SACGER research program will focus onsubsurface factors in Hot Rock and EGSresources such as reservoir characterisation andmodelling. This is complementary to the researchprograms of other centres. The research programis under development and will be designed inconsultation with the sector.Current projects at the centre include those whichwere funded by tied grants from PIRSA:Geophysical mapping and monitoring of anenhanced geothermal system usingmagnetotellurics.Rock fracture characterisation for hot dry rockenhanced geothermal systems. Building a regional thermal model for theAdelaide rift complex. Experimental verification of undergroundcooling for efficient thermal cycles. Optimisation of geothermal energyinvestments. Investigate low temperature thermalprocessing using geothermal energy. Reconnaissance thermal mapping foruranium and geothermal exploration.For more information seewww.adelaide.edu.au/geothermal/or contact the centre: imer@adelaide.edu.au.CSIROCSIRO’s research capabilities in the geothermalarena are broad, due to the organisation’sresearch diversity and ability to integratemultidisciplinary skills. The primary focus ofCSIRO’s activities in geothermal has beenthrough its contribution to WAGCOE. CSIRO’scontributions to the centre are mainly in thegeological, geophysical, ground water, andreservoir engineering aspects of the Perth BasinAssessment research program. CSIRO is alsodeploying its research expertise in hydraulicfracturing, reservoir engineering, well borestability, rock petrophysics and microseismicmonitoring to geothermal projectsFunding from the Education Investment Fund forCSIRO’s Sustainable Energy for SKA project wasrecently announced. A significant component ofthis project is a 10 MW th direct use geothermalcooling system for the Pawsey High PerformanceComputing Centre in Perth. The construction ofthis system will start with the drilling of aresearch/monitoring well, followed by a productionand injection doublet. The heat produced will beused in an adsorption chiller to provide cooling forthe supercomputer. CSIRO will work closely withWAGCOE during the development of this project.The geothermal component of the project is led byCSIRO Earth Science and Resource Engineeringin collaboration with the leaseholder, GeothermalPower Pty Ltd.Other projects include: Development of numerical modelling toolsthat couple thermal and poro-elasticprocesses for the assessment of well stability. Development of numerical modelling toolsand procedures for hydraulic stimulation athigh pressures and temperatures.Development of numerical modelling tools forfluid flow in fractures. Evaluating the application of petrophysicallogging techniques to the assessment ofthermal conductivity;Assessment of waveform characterisationtechniques for the interpretation ofmicroseismic monitoring data throughlaboratory based studies (high pressure hightemperature triaxial cell with acousticemissions monitoring) and the analysis offield data.For more information seewww.csiro.au/org/geothermalor contact the CSIRO: geothermal@csiro.au.University of Newcastle (Priority ResearchCentre for Energy – PRCfE)Located at the University of Newcastle, thePriority Research Centre for Energy has beenworking on geothermal energy research projectsfor a number of years through the researchprogram on Renewable Energy Systems. TheUniversity of Newcastle has received AU$30Mfrom the Federal Government through theEducation Investment Fund and AU$30Mmatching funding from other sources to establishthe Newcastle Institute for Energy & Resources(NIER). As part of this initiative, significant fundinghas been allocated for geothermal research, inparticular for the establishment of a state of theart facility for pilot-scale experimental research.The focus of geothermal research at theUniversity of Newcastle is on novel powergeneration cycles and the concept of a CO 2thermo siphon for EGS. The study of powercycles is regarded as one of the key areas formajor technological improvements since many ofthe problems associated with power generationfrom geothermal sources are underpinned byinefficient and often unsuitable heat exchangeprocesses within power cycles. In recognition ofthese shortcomings, the University of Newcastleinitiated a joint R&D program with Granite PowerLtd in 2006 with the goal of establishingalternative and potentially more efficient ways ofgenerating power from geothermal and other low-37

Australian Geothermal Conference 2010grade heat sources, such as industrial waste heat.The result was the creation of the GRANEXRegenerative Supercritical Power Cycle which isnow being commercialised.For more information seewww.newcastle.edu.au/research-centre/energyor contact the Renewable Energy SystemsProgram leader, Professor Behdad Moghtaderi:Behdad.Moghtaderi@newcastle.edu.au.Melbourne Energy InstituteThe Melbourne Energy Institute is located at theUniversity of Melbourne and has a number ofgeothermal projects including the VictorianGeothermal Assessment Report, which intends toaddress critical issues for the successfuldevelopment of geothermal power capability inVictoria.For more information see energy.unimelb.edu.auor contact the institute: mei-info@unimelb.edu.auRMITA significant project linking geothermal energyand desalination is underway at RMIT inpartnership with Greenearth Energy.The project involves the development of aprototype system that combines fresh waterproduction with electricity generation usingentirely renewable sources. The project issupported by an Australian Research CouncilLinkage grant and Greenearth Energy.For more information on this project contact MarkMiller, Managing Director of Greenearth Energy,via email markm@greenearthenergy.com.au ortelephone 03 9620 1566.For the RMIT website see www.rmit.edu.auAGEG Technical Interest GroupsCollaboration amongst all participants in thesector is enabled through the AustralianGeothermal Energy Group (AGEG) and theAGEG Technical Interest Groups (TIGs).Through linkages to the AGEG and its TIGs,Australia is a member of and contributes to thework of both the International Energy AgencyGeothermal Implementing Agreement (IEA-GIA)and the International Partnership for GeothermalTechnologies (IPGT). The 105 organisations thatare members of the AGEG have nominatedresearch topics of the highest priority to theindustry, which are closely aligned with thepriorities of both the GIA and the IPGT.The 12 TIGs are named in Table 1, and describedin more detail in the following section. For moreinformation or to join any of the TIGs: see theAGEG website, ,go to AGEG, Technical Interest Groups; orcontact the TIG leader. Contact details for all TIGleaders are also provided on the website.Table 1. AGEG Technical Interest GroupsTIGTopic1 Water management & environmental sustainability2 Reserves and resources3 Induced seismicity4 Outreach5 Economic modelling and novel use6 Power plants7 Direct use8 Information and data9 Reservoir development and engineering10 Exploration and well log technologies11 Drilling and well construction12 EducationTIG 1 water management and environmentalsustainabilityFollowing discussions and a decision made at theAGEG AGM in February, the name and focus ofTIG 1 has been redefined as water managementand environmental sustainability. The scope of theTIG and definition of outputs from the TIG will bedeveloped by the members and will be availableon the TIG webpage. The TIG leader is StevenKennedy from the Victorian Department ofPrimary Industries.TIG 2 reserves and resourcesThe reserves and resources technical interestgroup is predominantly to provide a forum forAGEG members to contribute to discussion on theAustralian Geothermal Reporting Code. The codeis developed and reviewed by the joint AGEG andAGEA Australian Geothermal Reporting CodeCommittee, chaired by Peter Reid fromPetratherm. Peter is also the TIG 2 leader,allowing feedback to the committee. TheAustralian Code for Reporting of ExplorationResults, Geothermal Resources and GeothermalReserves, The Geothermal Reporting Code(AGRCC, 2008), was launched in August 2008 atthe AGEG and AGEA Australian GeothermalEnergy Conference.TIG 3 induced seismicityHighlighting the importance of this topic and theneed for research in this area, a new technicalinterest group has been created to focus oninduced seismicity. Led by Michael Malavazosand Barry Goldstein from PIRSA, this TIG willbuild on the work done under the previous TIG 1,which included the release of two reports, CooperBasin HDR hazard evaluation: predictivemodelling of local stress changes due to HFRgeothermal energy operations in South Australia(Hunt and Morelli, 2006) and Analysis andmanagement of seismic risks associated withengineered geothermal system operations inSouth Australia (Morelli, 2009).38

Australian Geothermal Conference 2010TIG 4 outreachThe scope of TIG 4 is to assist the developmentof the Australian geothermal industry throughcoordinated geothermal training courses (inconjunction with IGA) an annual conference (inconjunction with AGEA), improvingcommunications through provision of a dedicatedwebsite that provides linkages to geothermalresource material and informing the publicthrough the provision of accessible information.TIG 4 is led by Betina Bendall from PIRSA.TIG 5 economic modelling and novel useTIG 5 has been modified to now cover economicmodelling as well as novel use applications forgeothermal energy including hybrid systems.Stephen Hinchliffe from SKM is the TIG leader.TIG 6 power plantsThe aim of TIG 6 is to achieve substantialimprovements in geothermal power plantefficiency through improvements in, for example,the cycle type, cycle fluids, heat exchangerefficiencies and more efficient cooling processes.Currently the focus is on the following areas:New cycles and cycle fluidsDry cooling Supercritical CO 2 geothermal siphonThe work of this TIG mirrors the IEA GeothermalResearch Annex VI. Hal Gurgenci from theQGECE and Behdad Moghtaderi from theUniversity of Newcastle are co-leaders of TIG 6.TIG 7 direct useDirect use geothermal applications include bothcirculating hot water & geothermal heat pumps.Direct use geothermal is being pursued throughwork in this TIG as a highly desirable substitutefor electricity use due to:Widespread occurrenceShallow depths (i.e. lower cost)High efficiency (no energy conversion) Utilisation of conventional plant andequipment Relatively easy set up Can be small to industrial scaleThe TIG for direct use is led by Klaus Regenauer-Lieb from the WAGCOE and Donald Payne fromthe University of Melbourne.TIG 8 data managementTIG 8, led by Anthony Budd from GeoscienceAustralia, aims to assist the development of theAustralian geothermal industry by simplifying dataavailability, usefulness and exchange throughstandards, database design, content, ongoingenhancements and development of manipulationand interpretive tools. Some initiatives alreadycompleted through the TIG include thedevelopment of an extension to 3D GeoModellerto enable prediction of 3D temperature and heatflow from inputs of heat production and thermalconductivity. Future projects planned are topopulate a heat flow database with new data,liaise with international parties working ongeothermal data management for webinteroperability standards and to build a datasystem within the Onshore Energy and MineralsDivision of Geoscience Australia.TIG 9 geothermal reservoir characterisation andengineeringAgreed terms of reference for this TIG are tofacilitate collaboration and information exchangeon topics including, but not limited to, Reservoir characterisation and modellingincluding 3D stochastic modelling and thesimulation of fracture networksFracture stimulation with emphasis on designand modellingGeochemistry, corrosion and scaling coveringboth the geothermal plant and equipment andthe rock-fluid interaction in the reservoir.Some projects planned by the TIG include tocompile and evaluate a list of available numericalmodels (algorithms) for fracture characterisation,scan for pre-existing geothermal data handbooksand build up to include information learnt fromoverseas projects and to date from Australianprojects, promote international collaborationthrough exchanges and develop 3D stochasticmodelling and simulation of fracture networks andfracture propagation in crystalline rocks. PeterDowd from the University of Adelaide is the leaderof this TIG.TIG 10 exploration technologiesTIG 10 aims to cooperate in studies to advancegeothermal methods and technologies: includingthe indirect detection of subsurface properties todelineate prospective trends; to develop, collect,improve and disseminate geothermal-relatedinformation; to identify opportunities to facilitatethe efficient advancement of geothermal energyprojects; and to disseminate information ongeothermal energy in Australia and internationallyby way of workshops, and open interchange ofexperiences and ideas generated within theAustralian community. The group have held anumber of successful workshops and have beenactively defining the research focus of the TIG. Afew topics that are the focus of the group include: geophysical and geological methods toreduce exploration risks prior to deep drilling39

investigating the possible use of satellite andairborne systems to measure surfaceradiation to detect high heat producinggranites in the sub-surface the use of magneto-telluric survey data toidentify and locate conductive fluids infractured rocks collection and processing of micro-seismicdata around geothermal projects.Des Fitzgerald from Intrepid Geophysics leadsTIG 10.TIG 11 drilling and well constructionTIG 11 has been very active and has regularmeetings to advance the group’s objectives.Topics in scope for the TIG include, but are notlimited to:Lower Cost Drilling: Investigate techniquesand new technologies which may lead to areduction in costs to drill geothermal wells. Zonal Isolation and Packers: Isolationtechniques and equipment need to bedeveloped to isolate zones of interest, atextremely high temperatures and possiblypressures, to allow both stimulation andproduction through different zones.Temporary Sealing of Fractures: Develop anon-damaging means of isolating fractureswhere multiple fractures/ loss zones may beencountered.Cutting Exploration Drilling Costs: A numberof mechanisms can be developed to reducecosts, such as improving ROP, utilisingsmaller rigs, smaller hole sizes, correctselection of materials.The two projects that will be the focus for thecoming year are to progress the slimhole drillinggroup and organise a drilling workshop. TIG 11 isled by Dean Hindle and Melanie Vonthethoff fromGeodynamics.TIG 12 educationTIG 12 was recently formed to work on the topicsof education for the geothermal sector andencompasses topics such as defining theeducation needs for the industry, the developmentof courses either at tertiary or postgraduate levelor short courses for industry. The TIG wouldcoordinate with aligned international programsand facilitate the initiation & promotion ofoccasional society lunches with guest speakers.Australian Geothermal Conference 2010ReferencesDepartment of Resources, Energy and Tourism(DRET), 2008, Australian Geothermal IndustryDevelopment Framework, and, AustralianGeothermal Industry Technology Roadmap,Commonwealth of Australia, Canberra.Australian Geothermal Reporting CodeCommittee (AGRCC), 2008, The GeothermalReporting Code. Australian Geothermal EnergyGroup, Adelaide.Hunt, S.P. and Morelli C.P, 2006, Cooper BasinHDR hazard evaluation: predictive modelling oflocal stress changes due to HFR geothermalenergy operations in South Australia. Report Book2006/16. Department of Primary Industries andResources South Australia, Adelaide.Morelli, C.P., 2009, Analysis and management ofseismic risks associated with engineeredgeothermal system operations in South Australia.Report Book 2009/11. Department of PrimaryIndustries and Resources South Australia,Adelaide.For more information about any of the TechnicalInterest Groups please see:www.geothermal.pir.sa.gov.au/agegor contact pirsa.ageg@sa.gov.au.40

Australian Geothermal Conference 2010Comparative petrology & geochemistry of high heat-producinggranites in Australia & EuropeV Marshall* 1,2 , J Van Zyl 1,2 , SE Bryan 3 , T Uysal 1,2 , M Gasparon 1,21. Queensland Geothermal Energy Centre of Excellence, The University of Queensland, St Lucia, QLD 40722. School of Earth Sciences, The University of Queensland, St Lucia, QLD 40723. Queensland University of Technology, Biogeoscience, GPO Box 2434, Brisbane, QLD 4001To better understand the origin, tectonic settingand generation of high heat-producing granites(HHPG) and provide target criteria for theexploration of granite-hosted EnhancedGeothermal Systems (EGS), Australian examplesare being studied and compared with betterknown European analogues that are being reexamined.Three areas of long-standing interest for EGS inEurope are Cornwall, UK; Soultz-sous-Forêts,France; and the Erzgebirge, Germany. Theseareas have higher than average heat flow (atsurface and depth), and concentrations of themain heat-producing elements Uranium (U),Thorium (Th) and Potassium (K) in granitic rocks.The Big Lake granite suite of the centralAustralian Cooper Basin (Gatehouse et al., 1995),is a pre-eminent HHPG target for EGSdevelopment in Australia. The Big Lake Suitegranites are Australia’s hottest with temperaturesof >250ºC at ~4km depth (Habanero 1, (Baisch etal., 2006). The Soultz-sous-Forêts and CooperBasin systems have in common an insulatingsediment cover of several kilometres.Keywords: high heat-producing granites (HHPG),Cooper Basin, Cornwall, Soultz-sous-Forêts,Erzgebirge, Uranium, Thorium, Permo-Carboniferous, intraplate.BackgroundThe development of EGS in the Cooper Basinwas started by Geodynamics Ltd in 2002 with theaim to be the first Australian Company producingelectricity from EGS sources. Their firstgeothermal well (Habanero 1) encounteredgranite at depth. The granite (Big Lake Suite) issituated in the Nappamerrie trough (Meixner etal., 2000; Sun, 1997) and intruded into theCambrian – Ordovician Warburton Basin(Gatehouse et al., 1995). The Big Lake SuiteGranite is covered by a ~3km thick sedimentarypackage consisting of the Cooper (LateCarboniferous – Middle Triassic), Eromanga(Early Jurassic) and Lake Eyre (Early Tertiary -Holocene) Basins. U/Pb series dating of zircon ofgranite intersected at depth in two wellsestablished a Permo-Carboniferous age rangingbetween 323±5 Ma and 298±4 (Gatehouse et al.,1995).The Cornubian batholith in Cornwall (SWEngland; Fig. 1) is comprised of six plutonsranging in age from 293.1±1.3 Ma to 274.5±1.4Ma, based on U/Pb dating of magmatic monazites(Chen et al., 1993). Two plutons of interest arethe Carnmenellis and Land’s End granites (Fig. 2)as they contain the highest levels of U and Th(Table 1). During the 1970’s and 1980’s theRosemanowes area of the Carnmenellis granitewas a test site for the Department of Energy,European Commission and Cambourne School ofMines investigation into EGS technology as it hasthe highest heat flow in England, over 120 mW/m 2(Tenzer, 2001).Soultz-sous-Forêts (Fig. 1) has been theEuropean test site for EGS since 1987, withtemperatures of 200°C at 5 km depth (Antics andSanner, 2007).The granites of the Erzgebirge region (Fig. 1), likeCornwall, are associated with Tin (Sn), Tungsten(W), Lead (Pb), Zinc (Zn), Fluorine (F), Barium(Ba) and U mineralisation, with the high U and Thcontents and crustal heat flow values givingadded prospectivity as a clean energy resource.Tectonic HistoryLAURASIA200 kmFigure 1: Location of the Hercynian/ Variscan mountainchains in the middle of the Carboniferous period. Red linesare sutures. Yellow line is subduction zone (triangles point indirection of subduction) (Scotese, 2001). Adapted from Matte(2001) and Zeigler (1990).Regionally, the European granites were emplacedin an extensional environment between 334-270Ma following collision of Laurasia and Gondwanato form the supercontinent Pangaea (Willis-Richards and Jackson, 1989). This collisionalevent, known as the Variscan (Hercynian)orogeny, resulted in an extensive rift systemwithin the northern foreland of the Variscan41

Australian Geothermal Conference 2010orogenic belt following termination of orogenicactivity (Wilson et al., 2004).LAND’S ENDCornwallCARNMENELLISFigure 2: Heat flow map of Cornwall, UK, showing thelocation of the Carnmenellis and Land’s End plutons. Heatflow of >120 mW/m 2 (deep red), to 60-70 mW/m 2 (green).Adapted from British Geological Survey GIS heat flow map,2010.The Cornubian batholith was emplaced at the endof the Variscan orogeny terminating a period ofregional, back-arc continental extension duringthe Devonian and Early Carboniferous,(Darbyshire and Shepherd, 1985).This orogenic event caused crustal thickening(Willis-Richards and Jackson, 1989) and hasbeen suspected to have caused anatectic melting(Clark et al., 1993).Figure 1 shows that at the time of emplacementthe Cornubian granites were over 1350 km awayfrom the nearest active plate margin (boundingthe Neo-Tethys), suggesting they are notsubduction related.Soultz-sous-ForêtsThe Soultz-sous-Forêts granites lie within theUpper Rhine Graben, a tectonically reactivatedregion associated with Oligocene crustalextension and thinning (Genter et al., 2003).These granites were emplaced slightly prior tothose in Cornwall following the Variscan orogeny,from 334 +3.8/-3.5 Ma (U/Pb monzogranite) to319.8±0.6 Ma ( 40 Ar/ 39 Ar), (Cocherie et al., 2004and Stussi et al., 2002).ErzgebirgeThe Erzgebirge granites were also intruded fromthe Late Carboniferous to Early Permian duringlarge-scale extension following orogenic collapse(Förster et al., 1999b). In addition, wrenching wasassociated with granite emplacement (Edel et al.,2007).Förster and Förster (2000) concluded that peakorogenic metamorphism occurred 350-340 Ma,followed by rapid uplift and cooling between 340and 326 Ma with the oldest, post-collisional,granites intruded to a depth of 3-6 km at 325 Ma(Förster et al., 1999b).Table 1 is a summary of relevant information forthese areas of HHPG.Granite PetrographyCooper BasinThe samples studied from the Big Lake Suitehave varying degrees of alteration, with alterationincreasing with decreasing depth. Unalteredsamples consist of quartz (31%), K-feldspar(31%) and plagioclase (31%), with minor biotite(5%) and accessory phases (2%). Accessoryphases indentified are magnetite, apatite,thorite/huttonite, zircon, monazite and xenotimewith some fluorite.The altered samples show varying degrees ofsericitisation, ranging from minor replacement ofthe feldspars to complete replacement and theformation of illite. Biotite has altered to chlorite,after which chlorite is replaced by illite. Freshbiotite still exists in the altered samples, but isrestricted to small inclusions in quartz. The illitefrom the Big Lake Suite alteration assemblageoccurs as two main textures: 1) pseudomorphicand 2) cryptocrystalline.AbAnorthoclaseSanidineOrAlbiteOligoclaseAndesineLabradorite BytowniteAnorthiteAnFigure 3: Feldspar classification diagram showing thefeldspar composition. Cooper Basin (•) and Soultz–sous–Forêts (■) (Komninou and Yardley, 1997; Stussi et al.,2002).Electron Probe Microanalysis (EPMA) data showsno compositional differences between thedifferent illite textural types. Although illite doesnot show compositonal differences between thetextural types, there are two compositionallydistinct illites - the first is a pure illite with low(

Australian Geothermal Conference 2010undulose extinction, with some grains exhibitingplanar deformation features.Figure 4: Photomicrographs of two illite textures (top), andplanar deformation features in quartz (bottom) in Big LakeSuite granites.CornwallThe Carnmenellis granites generally containaround 34% quartz, 32% K-feldspar, 22%plagioclase, 6% biotite, 4% muscovite, 1%tourmaline and 1% accessory phases includingandalusite, cordierite, ilmenite, zircon, monaziteand uraninite (Charoy, 1985; Chappell and Hine,2006).The radioactive mineral assemblageimportant for hosting Th and U are predominantlymonazite, apatite, xenotime, zircon and uraninite(hosting ~90% of total U) (Jefferies, 1984; Martelet al., 1990), all of which are commonlyassociated with apatite and ilmenite (Jefferies,1984).Soultz-sous-ForêtsThese granites have a normative composition ofaround 43% plagioclase (37% albite, 6%anorthite), 25% K-feldspar, 22% quartz, 6%biotite, 2% hornblende, and

Australian Geothermal Conference 2010confirmed Tammemagi and Smith’s (1975)observations of abundant K-feldspar megacrystsin places. Some plutons have subhorizontal flowfabrics (e.g., St Austell granite) suggesting somegranitic magma emplacement in sill-like orlaccolithic bodies, and mineralogical and agevariations indicate the batholith has beenincrementally built.Soultz-sous-ForêtsThe Soultz-sous-Forêts granites have beenclassified as S-type according to tectonicdiscrimination diagrams (Fig. 6), however they arenoted to have hornblende (Stussi et al., 2002) andoxidised K-feldspar, suggesting possible I- or A-type classification. They are also peraluminousand porphyritic, with K-feldspar megacrysts up to80 mm in diameter. These granites are thought tohave originated from either crustal partial meltingof enriched metamorphic greywacke, granuliteand amphibolite, or meta-igneous rocks ofappropriate composition; however, a minor mantlecontribution can’t be excluded due to the negativeεNd values (Stussi et al., 2002). Subhorizontalmagma fabric oriented to the NE was also notedsuggesting either a laccolithic body or the roof ofa diapir-like intrusion, in association with postmagmaticfracturing (Stussi et al., 2002).ErzgebirgeThe Erzgebirge granites are more varied, withmildly peraluminous transitional I- to S-types andstrongly peraluminous S-types both present(Förster et al., 1999b, Fig. 5).Förster et al. (1999b) also noted some graniteshave peraluminous A-type chemical affinities,consistent with the intraplate setting of thismagmatism.Förster and Förster (2000) recognised that thesurface heat flow is higher where the granitesoccur, contributing 70-90 mW/m 2 to the surfaceheat flow - two to four times higher than would beestimated from the average composition of thecontinental bulk crust.Table 1: Compositions of targeted HHPG for EGS. Sources: Charoy (1985), Chappell and Hine (2006), Jefferies (1984), Martelet al. (1990), Tammemagi and Smith (1975), Willis-Richards and Jackson (1989), Chen et al. (1993), Müller et al. (2006), FörsterH.J. (1999), Förster and Förster (2000), Förster et al. (1999a), Förster et al. (1999b), Hooijkaas et al. (2006), Stussi et al. (2002),Uysal and van Zyl (unpubl. data), and Taylor and McLennan, (1985).EGSTARGETEDGRANITESCarnmenellisCornwall, UKLand’s EndCornwall, UKSoultz-sous-ForêtsUpper RhineGraben, FranceErzgebirge,GermanyCooper Basin,AustraliaUpperContinentalCrustAGE(Ma)LITHOLOGY295-270 Megacrystic biotite and two-micagranitesSiO2(wt%)K2O(wt%)U(ppm)70-76 4-4.3 4.3-35(12.1mean)Th(ppm)11-25(19.3mean)277-274.5 Early megacrystic biotite granites, 66-73 3.5-6 6.9-38 4.4-46.2younger Li-siderophyllite granites334-319 Porphyritic monzogranite 67-69 3.8-4 6.2-14.1 23-37325-315 Transitional I-, S- and A- type biotitegranites, two-mica granites and S- andA- type Li-mica granites323-298 Coarse grained two feldspar biotitegranite, moderately weathered67-77(biotite)71-76(two-mica)3.8-5.4 9-30.9 10.4-34.373-76(Li-mica)- - 11 - 27 17 - 11766 3.4 2.7 10.744

Australian Geothermal Conference 2010ANK3.02.8 Metaluminous Peraluminous2.62.42.22.01.81.61.41.21.00.8Peralkaline0.60.40.5 1.0 1.5 2.0ACNKFigure 5: ASI index for granites classification plot accordingto Shand (1927), Cooper Basin (); Cocherie et al. (2004),Soultz-sous-Forêts (•); Chappell and Hine (2006) andMüller et al. (2006), Cornwall (•); and Förster et al. (1999b)Erzgebirge (X).100090040080307210Cooper BasinLa Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu100WPGNb10VAG+Syn-COLGORG11 10 100 10002000Figure 6: Tectonic discrimination diagram comparing thestudied granite suites (Pearce et al., 1984).300Y900.0400.0Erzgebirge80.0FeO/MgO20FG30.07.02.0OGT150 600Zr+Nb+Ce+Y0.60.1La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb LuFigure 7: Granite discrimination plot from Whalen et al.(1987) OGT - Other Granite Types (I-, M-, S- type), FG –Fractionated Felsic Granites.Figure 8: REE Spider plots for the four granite intrusions.Normalisation was performed using McDonough and Sun(1995).45

Australian Geothermal Conference 2010Discussion and Comparison Both European and Australian examples ofHHPG are Late Carboniferous to EarlyPermian in age. All HHPG in this study were emplaced well inboard from active plate margins. TheEuropean granites are post-collisional andextension-related, whereas the Cooper Basingranites are also within-plate but may be(temporally) related to intraplate deformation. The Cooper Basin, Soultz-sous-Forêts andthe Erzgebirge show similar abundances ofquartz, plagioclase and K-feldspar, withplagioclase being the more abundant phasein Soultz-sous-Forêts. The Cooper Basin andErzgebirge are equigranular, whereas Soultzsous-Forêtsand the Carnmenellis aredistinctly megacrystic granites. It is difficult to use the I- and S- typeclassification for fractionated granites, but thisstudy suggests all the granites areperaluminous, with the European granitesdominantly S-type, but Cooper Basin andErzgebirge also showing I- and A- typeaffinities suggesting that crustal sourcecomposition may not be a prime control onHHPG generation. Literature and field observations indicate thatat least part of the Cornubian and Soultzsous-Forêtsgranites have sill-like intrusivecomponents but importantly, that all thebatholiths discussed here have beenincrementally built by a number of intrusions. Trace element geochemistry and rare earthelement abundances of the Cooper Basingranites show a range from less fractionatedto highly fractionated (Eu/Eu* =0.04 to 0.17),whereas Soultz-sous-Forêts show moreuniform REE patterns, and a wide range ofREE patterns have been described for theErzgebirge granites. The Carnmenellis has anarrower range in REE than the Erzgebirge.The Cooper Basin, Carnmenellis andErzgebirge all show strong Eu anomaliesindicating fractional crystallisation wasimportant in the production of these granites(Fig. 7). Isotope data suggests that all examples havecrustal signatures. The Cooper Basin granites are more enrichedin Th, but there is similar U enrichment for allexamples. Currently, the reason for U and Thenrichment is unknown, except for the normalfractional crystallisation processes.The current conclusion from this study is that theperaluminous, highly fractionated chemistrieshave promoted the U and Th enrichment ofHHPG.ReferencesAntics, M., and Sanner, B., 2007, Status ofgeothermal energy use and resources in Europe,Proceedings European Geothermal Congress, 2Baisch, S., Weidler, R., Vörös, R., Wyborn, D. andde Graaf, L., 2006. Induced Seismicity during theStimulation of a Geothermal HFR Reservoir in theCooper Basin, Australia. Bulletin of theSeismological Society of America, 96(6): 15.Bartier, D. et al., 2008. Hydrothermal alteration ofthe Soultz-sous-Forêts granite (Hot FracturedRock geothermal exchanger) into a tosudite andillite assemblage. European Journal ofMineralogy, 20: 12Charoy, B., 1985, The genesis of the CornubianBatholith (south-west England): the example ofthe Carnmenellis Pluton, Journal of Petrology, 27,Part 3, 571-604Chappel, B.W. and Hine, 2006, The CornubianBatholith: An Example of Magmatic Fractionationon Crustal Scale, Resource Geology, 56, 203-244Chappell, B.W. and White, A.J.R., 2001. TwoContrasting granite types: 25 years later.Australian Journal of Earth Sciences, 48: 12.Chen Y., Clark, A.H., Farrar, E., Wasteneys,H.A.H.P., Hodgson, M.J. and Bromley, A.V.,1993, Diachronous and independent histories ofplutonism and mineralisation in the CornubianBatholith, southwest England, Journal of theGeological Society, 150, 1183-1191Clark, A.H., Chen,Y., Farrar, E., Wasteneys,H.A.H.P, Stimac, J.A., Hodgson, M.J., Willis-Richards, J., and Bromley, A.V., 1993, TheCornubian Sn-Cu (-As, W) metallogeneticprovince: product of a 30 M.Y. history of discreteand concomitant anatectic, intrusive andhydrothermal events, Proceedings of the UssherSociety, 8, 112-116Cocherie, A., Guerrot, C., Fanning, C.M. andGenter, A., 2004, Datation U-Pb des deux facièsdu granite de Soultz (Fossé rhénan, France),Géochimie (Géochronologie), 336, 775-787Darbyshire, D.P.F., and Shepherd, T.J., 1985,Chronology of granite magmatism and associatedmineralisation, SW England, Journal of theGeological Society, 142, 1159-1177Edel, J.B., Schulmann, K. and Rotstein, Y., 2007,The Variscan tectonic inheritance of the UpperRhine Graben: evidence of reactivations in theLias, Late Eocene-Oligocene up to the recent,International Journal of Earth Science, 96, 305-325Förster H.J., 1999, The chemical composition ofuraninite in Variscan granites of the Erzgebirge,Germany, Mineralogical Magazine, 63, 2, 239-25246

Australian Geothermal Conference 2010Förster, H.J., Davis, J.C., Tischendorf, G., andSeltmann, R., 1999a, Multivariate analyses ofErzgebirge granite and rhyolite composition:implications for classification of granites and theirgenetic relations, Computers and Geosciences,25, 533-546Förster H.J.,Tischendorf, R.B., Trumbull, R.B. andGottesmann, B., 1999b, Late-collisional granitesin the Variscan Erzgebirge, Germany, Journal ofPetrology, 40, 11, 1613-1645Förster, A. and Förster H.J., 2000, Crustalcomposition and mantle heat flow: Implicationsfrom surface heat flow and radiogenic heatproduction in the Variscan Erzgebirge (Germany),Journal of Geophysical Research, 105, B12,27,917-27,938Gatehouse, C.G., Fanning, C.M. and Flint, R.B.,1995. Geochronology of the Big Lake Suite,Warburton Basin, northeastern South Australia. .Quarterly Geological Notes of the GeologicalSurvey of South Australia, 128: 7.Genter, A., Guillou-Frottier, L., Feybesse, J-L.,Nicol, N., Dezayes, C. and Schwartz, S., 2003,Typology of potential hot fractured rock resourcesin Europe, Geothermics, 32, 701-710Ghosh, P.K., 1934, The Carnmenellis Granite: itspetrology, metamorphism and tectonics, QuarterlyJournal of the Geological Society, 90, 240-273Hooijkaas, G.R., Genter, A. and Dezayes, C.,2006, Deep-seated geology of the graniteintrusions at the Soultz EGS site based on datafrom 5 km-deep boreholes, Geothermics, 35, 484-506Komninou, A. and Yardley, B.W.D., 1997. Fluidrockinteractions in the Rhine Graben: Athermodynamic model of the hydrothermalalteration observed in deep drilling. Geochimica etCosmochimica Acta, 61(3): 17Jefferies, N.L., 1984, The radioactive mineralassemblage of the Carnmenellis Granite,Cornwall, Proceedings of the Ussher Society, 6,35-40Martel, D.J., O’Nions, R.K., Hilton, D.R. andOxburgh, E.R., 1990, The role of elementdistribution in production and release ofradiogenic helium: the Carnmenellis Granite,southwest England, Chemical Geology, 88, 207-221Matte, P., 2001, The Variscan collage andorogeny (480 – 290 Ma) and the tectonicdefinition of the Armorica microplate; a review,Terra Nova, 13, 122-128McLaren, S. and Dunlap, W.J., 2006. Use of40Ar/39Ar K-Feldspar thermochronology in thebasin thermal history reconstruction: an examplefrom the Big Lake Suite Granites, WarburtonBasin, South Australia. Basin Research, 18: 15.Meixner, T.J. et al., 2000. The Nature of theBasement to the Cooper Basin region, SouthAustralia. Exploration Geophysics, 31: 9.Müller, A., Seltmann, R., Halls, C., Siebel, W.,Dulski, P., Jeffries, T., Spratt, J. and Kronz, A.,2006, The magmatic evolution of the Land’s Endpluton, Cornwall, and associated pre-enrichmentof metals, Ore Geology Reviews, 28, 329-367Pearce, J.A., Harris, N.B.W. and Tindle, A.G.,1984. Trace Element Discrimination Diagrams forthe Tectonic Interpretation of Granitic Rocks.Journal of Petrology, 25(4): 28.Scotese, C.R., 2001, Late Carboniferous: Atlas ofEarth History, 1, Paleogeography, PALEOMAPProject, 52Stussi, J.M., Cheilletz, A., Royer, J.J.,Chevremont, P. and Feraud, G., 2002. TheHidden monzogranite of Soultz-sous- Forêts(Rhine Graben, France). Géologie de laFrance(1): 19.Sun, X., 1997. Structural style of the WarburtonBasin and control on the Cooper and EromangaBasins, South Australia. Exploration Geophysics,28: 6Tammemagi, H,Y. and Smith, N.L., 1975, Aradiogeologic study of the granites of SW.England, Journal of the Geological Society, 131.415-427Taylor, S.R. and McLennan, S.M., 1985, TheContinental Crust: its Composition and Evolution,Blackwell Scientific Publications, 46Tenzer, H., 2001, Development of hot dry rocktechnology, GHC Bulletin, 18Whalen, J.B., Currie, K.L. and Chappell, B.W.,1987. A-Type granites: Geochemicalcharacterisitcs, discrimination and pretrogenesis.Contributions to Mineralogy and Petrology, 95: 13Willis-Richards, J. and Jackson, N.J., 1989,Evolution of the Cornubian ore field, southwestEngland: Part I. Batholith modelling and oredistribution, Economic Geology, 84, 1078-1100Wilson, M., Neumann, E. R., Davies, G. R.,Timmerman, M. J., Heeremans, M., and Larsen,B. T., 2004, Permo-Carboniferous magmatismand rifting in Europe: introduction, The GeologicalSociety, London, Special Publication, 223, 1-10Veevers, J.J., 2009, Mid-Carboniferous Centralianuplift linked by U-Pb zircon chronology to theonset of Australian glaciation and glacio-eustasy,Australian Journal of Earth Sciences, 56, 711–717.Zeigler, P.A., 1990, Geological Atlas of Westernand Central Europe, Shell International PetroleumMaatschappij BV (2 nd edition)47

Australian Geothermal Conference 2010Geothermal Prospection Using Existing Groundwater Geochemicaland Thermal Datasets: Identifying Regions of Interest in Queenslandfrom Government Well and Bore Data* +Craig McClarren 1,2 , Massimo Gasparon 1,2,3 , and I. Tonguç Uysal 1,21 School of Earth Sciences, The University of Queensland, Brisbane, QLD. 40722 Queensland Geothermal Energy Centre of Excellence, The University of Queensland, Brisbane, QLD. 40723 National Centre for Groundwater Research and Training, The University of Queensland, Brisbane, QLD. 4072Geochemical analysis of well and bore waters hasthe potential to provide valuable clues inidentifying regions of high geothermal potential.Chemical and isotopic markers indicative of thepresence of high heat producing granites or otherheat producing bodies may be mobilized byflowing groundwater and may make their way todeep wells and bores. Though likely highlydiluted, these markers may also be detectable inshallow wells and bores in some regionsdepending on local geological and hydrologicalconditions. A significant quantity of data of thistype already exists in publicly availablegovernment databases. These data are oftencoarse and of frequently unknown quality, butmay prove to be a useful tool in first-order largescalegeothermal prospection. We have obtainedsuch a dataset from the Queensland Departmentof Environment and Resource Management, andhave extensively analysed the data with the aimof identifying potential areas of interest ingeothermal prospection. This research representsthe first stage of a large multi-stage project ingroundwater geochemistry aimed at geothermalprospection.Keywords: Geothermal exploration, bore water,trace element, isotope, QueenslandThe Potential for GroundwaterGeochemistry in GeothermalProspectingGeochemical analysis of groundwater may holdsignificant potential for identification ofunrecognized regions of high geothermalpotential. Geothermal source rocks, whethersedimentary, radiogenic igneous, or youngigneous, produce mineralogical, chemical, andisotopic markers which can be used to identifythem (Marini, L.). Many of these markers arewater soluble and can thus be moved from depthtoward the surface (Barbier, et. al, 1983). Duringthe process, these markers may suffer significantdilution, but may still be detectable in groundwatersamples taken at or near these thermallysignificant regions (Smedley, P., 1991) .Thousands of ground water samples from acrossQueensland have already been taken andanalysed for a limited range of water qualityparameters; this data can be obtained from theState Government for research purposes. Byproviding chemical analysis results from almost28,000 water wells and bores across the state,the Queensland Department of Environment andResource Management (DERM) water boreholedatabase is a locally unparalleled resource forconducting a large-scale first-order typeinvestigation and identification of geothermallyprospective regions within the state.Geochemical toolsWhile there are dozens of mineralogical,chemical, and isotopic markers that can be usedto identify potential geothermal targets, in usingthe data available from the DERM water boreholedatabase, one is restricted to the limited set ofparameters measured. The quality andcompleteness of the chemical data can bedescribed as highly variable, with results collectedand analysed by many different parties usingunknown procedures over the course of severaldecades. Moreover, the results themselves areinconsistent between analyses, withmeasurements for elements such as boron andphosphorous making occasional appearanceswhile largely remaining absent from the rest of thedataset.While the quality of measurements may be calledinto question, the abundance of analyses, as wellas their geographic population density in manyregions, suggests that the data may still becautiously used on a large scale (typically a fewthousand square kilometres at best) to identifysources of heat at depth. From the database,initially a large number of known indicators ofgeothermal potential, used extensivelyinternationally, were selected. These include theelements and compounds Na, K, Cl, F, Cu, SO 4 ,Zn, PO 4 , and B. While it was readilyacknowledged that many of these elements maybe associated to non-geothermally indicativesources, the rationale for this analysis is thatwhere a large number of notably highconcentration values for these elementsoverlapped, these areas would be most worthy offurther investigation.48

Australian Geothermal Conference 2010Figure 1: Early map of regions to contain potentialgeothermal resources based on unfiltered geochemical data.The image was composed by colouring regions with thehighest potential for each selected element or compound andthen overlaying all of the layers. Diagonal hashes indicateslightly promising regions while boxed X’s are the mostpromising regions based on this early analysis.Refining the dataWith this preliminary analysis of data completed,maps of individual elements were more closelystudied to assess their suitability as potentialindicators of associated geothermal systems . Naand K were first discarded because of thetremendous abundance of possible alternativesources. Cl was next discarded because of itsclose relationship with K and Na. PO 4 was nextremoved as data on this element throughout thedataset was extremely sparse and because thehighest values tended to be in wetter, moretropical, parts of Queensland which may simplyreflect the infiltration of organic phosphate fromthe surface into a shallow watertable. Zn wasremoved both for its scarcity of data as well as fornot showing any distinguishable trends or patternswhen concentrations were mapped. While Cuconcentration showed possibly promising results,again the data was relatively sparse. SO 4 did notshow clear trends and has many possiblesources, so it was initially removed. One potentialsource of sulphate, however, are coal measureswhich are known to be effective cap units togeothermal reservoirs. Thus, sulphate data maybe revisited in the future to analyse the suitabilityof a geothermally interesting site to futureexploitation.It was finally decided to focus primarily on boron(B) and fluoride (F - ), which are both relativelysoluble and abundant in felsic igneous bodies.Boron data, though somewhat sparse, showseasily distinguished trends (non-randomgeographical and/or geological distributions) whenmapped. Fluoride data is abundant and alsoshows easily distinguished trends when mapped.It was additionally decided that F - would be mostindicative of buried high heat producing granites(HHPGs) at depth, as there are fewer likelyalternative sources for F - than for B.Filtering the dataThe aim of this research is two-fold. First, it is toidentify potential regions of geothermal interestand secondly to identify regions on which to focusfuture sampling work for later stages of thisresearch project (the details of which will bediscussed elsewhere). The first filter applied wasthe most fundamental: many of the bores in thedatabase were no longer existing, but werecategorized as “Abandoned and Destroyed.”These entries were removed as they were of nopractical use to us, in that the results could not beverified by future sampling. Wells that wereclassified as “Existing” or “Abandoned butUseable” were left in the set. The next filterapplied was for F - concentration; after someconsultation and discussion within the group, avalue of 1.4 ppm was decided upon as a lowerlimit for water chemistries we deemed“interesting”. The choice of this value significantlyreduced the dataset, but left behind more than3500 values showing easily distinguishable trendsand patterns when mapped across Queensland.High values around known igneous regions, suchas the Stanthorpe Granite Belt and the morerecently active Cairns region, strongly suggeststhat this could be an extremely valuable tool inidentifying igneous bodies at depth elsewhere.High values in regions not already known tocontain intrusive bodies will be investigated.The next filter applied was depth as it is likely thatwith greater depth comes a lower likelihood ofsignificant dilution of circulating or simply flowinggroundwater by local and relatively recentmeteoric waters. Entries with bore depths of lessthan 10 meters were removed; however, becausemany entries do not have a recorded depth (andthus receive a value of zero through intermediateprocessing steps), values of zero were left so thatwells of unrecorded depth, of which there aremany, would remain. Next, because at or nearHHPGs we expect to find elevated Bconcentrations in groundwaters, the sites with thelowest B values were removed from the remainingdataset. A cut-off of 0.3 ppm B was chosen to49

Australian Geothermal Conference 2010remove the sites least likely to show evidence of anearby HHPG; similar to the filter for depth,values of zero were left as most of the sitessampled for F - content unfortunately do not showB results.Finally, with purely the aim of sample access inmind the category of “owner” needed to befiltered. Location and ownership information isprovided with the data; however access to themonitoring bores is restricted to DERM staff. Withthese data filtered out, most of the prospectivesampling sites along coastal Queensland areremoved; the majority of the inland bores, whichare usually privately owned, remain, allowing usto extensively investigate most of thegeochemically interesting regions within the state.Coastal regions, with close proximity to populationcentres, however, are thereby restricted.prove useful when considered together withgeochemical data in the future.In addition to water bore and well data fromDERM, there are also publically available datafrom exploratory coal, oil, and gas drilling in theQueensland Petroleum Exploration Database(QPED) from the Queensland Department ofMines and Energy (DME). These data includetemperature results as well as geochemicalanalyses; because of their extremely limitedgeographic distribution, however, this dataset isunsuitable to state-wide geothermal prospection.A high concentration of data points in the GreatArtesian Basin in Queensland’s Southwest cornermay prove useful to individuals or groupsinvestigating this known region of high geothermalpotential. In a far more local context, QPED wellcompletion reports may prove valuable as anearly investigatory step in areas where suchdrilling has been conducted.Figure 2: Water well and bore data across Queenslandfiltered for Fluorine and Boron content, depth, and “owner”.The colour scale is for Fluorine content in ppm.Hydrothermal Data and Results fromOther DatasetsAs well as geochemical analyses of water boresthroughout the state, the DERM dataset alsocontains water temperature data from many ofthese same bores. These data, however, are oflimited value due to the limitations of bore depth;water temperatures are low throughout coastalQueensland and rise rather predictably withdistance inland, likely reflecting the increasingdepth to the water table rather than a significantgeothermal gradient. These data, however, mayFigure 3: Unfiltered water well and bore thermal data acrossQueensland. The map shows maximum temperaturesrecorded at each site; the colour scale is for watertemperature in degrees Celsius.SummaryAn extensive dataset comprising thermal andgeochemical analyses of public and privatelyowned water wells and bores from across thestate of Queensland is publicly available from theQueensland Department of Environment andResource Management. While the thermal datahave limited potential in geothermal prospectiondue to the depth-dependent nature of themeasurements, the geochemical dataset may be50

Australian Geothermal Conference 2010of tremendous value. Chemical markers indicativeof HHPGs or other heat producing bodies may bemobilized by flowing groundwater, making theirway to deep wells and bores. Though likelydiluted, these markers may also be detectable inshallow wells and bores depending on localgeological and hydrological conditions.We have analysed the available dataset, choosingto focus primarily on fluoride and boronconcentrations, as a first-order exploratory tool toidentify regions potentially containing geothermalresources and to identify sites and regions forfuture sampling and geochemical analysistowards that end. Several filters were put in place,reducing the extensive dataset to a moremanageable and consistent and size and nature.The Queensland Petroleum and ExplorationDatabase publically available from theDepartment of Mines and Energy is not suitablefor state-wide geothermal prospection due to thegeographic distribution of sampling sites, but maybe of significant local value in SouthwestQueensland.(including accuracy, reliability, completeness orsuitability) and accepts no liability (includingwithout limitation, liability in negligence) for anyloss, damage or costs (including consequentialdamage) relating to any use of the dataReferencesBarbier, E., Fanelli, M., Gonfiantini, R.,1983,Isotopes in Geothermal Energy Exploration, IAEABulletin, v. 25, no. 2, pp. 31-36Marini, L., Geochemical Techniques for theExploration and Exploitation of GeothermalEnergy, Dipartimento per lo Studio del Territorio edelle sue Risorse, Università degli Studi diGenova, Genova, ItalySmedley, P., 1991, The Geochemistry of RareEarth Elements in Groundwater from theCarnmenellis Area, Southwest England,Geochimica et Cosmochimica Acta, v. 55, pp.2767-2779* Based on or contains data provided by the Stateof Queensland (Department of Environment andResource Management) [2009]. In considerationof the State permitting use of this data youacknowledge and agree that the State gives nowarranty in relation to the data (includingaccuracy, reliability, completeness, currency orsuitability) and accepts no liability (includingwithout limitation, liability in negligence) for anyloss, damage or costs (including consequentialdamage) relating to any use of the data. Datamust not be used for direct marketing or be usedin breach of the privacy laws.+Based on or contains data provided by the Stateof Queensland (Department of Employment,Economic Development and Innovation) [2009]which gives no warranty in relation to the data51

Australian Geothermal Conference 2010Genie Impact DrillsSynopsis of a New Hard Rock Drilling DevelopmentMalcolm B McInnes FIEAust. CPEng MESc BE ARMIT Mech Eng MAIE SPESpecialised Drilling Services Australia Pty LtdPO BOX 297, Fullarton, South Australia 5063, Australia.AbstractIn 2004, a lean new Australian company wasestablished to independently focus on thedevelopment and commercialisation of newdrilling technology for the geothermal, petroleum& scientific industries. It began with assetsincluding know-how, intellectual property,documentation and some equipment retainedfrom pioneering work undertaken in the parentcompany between 1992 and 2002. It hasproduced a new commercial product, the Genie185, a drilling liquid powered impact drill, the firstin a range of new, high performance tools whichoffer higher Rates Of Penetration and simpler,lighter drill strings for hard formation drilling.PurposeTo build a profitable commercial enterprise basedon the development and provision of superiorquality drilling services & products to thepetroleum, geothermal and scientific industries.Geothermal power marketAustralian geothermal wells are being drilled todepths of around 4,500m but financial viability ofhard fractured rock reservoirs is threatened by lowRates Of Penetration and problems associatedwith traditional rotary drilling systems in theseenvironments.TechnologyHigh ROP, drilling liquid powered impact drillingsystems for harder formations.ProductRented, on-site supported, fully serviced &equipped Genie impact drills and bit sales. .Genie impact drillConceived and designed as a flexibly sized tool tocreate a range rather than a series of successiveindividually designed tools. The design isdisciplined and imbedded in the drawings of thedefinitive, mid-range Genie 185 (for bits between8-1/2" & 9-7/8"). See Fig. 1. There are fewer partsthan in predecessor tools as well as enhanced bitalignment, new hard surface designs andtreatments, corrosion protection, simple andsecure jet nozzle mounting, longer tong grip zoneand new assembly system for easy, fast and safemaintenance and set-up. A general mathematicalmodel generates hydraulic & dynamicperformance simulations which are used with awell model for application checks & set-up.FIG. 1 – TALISMAN ENERGY INC WITH GENIE 185 ONNABORS RIG – BC, CANADAGeneral design & manufacturingPart designs are geometrically and parametricallylinked to each other and to a single scalingparameter which sets tool size. This savesrepletion and improves drafting quality. The midrangeGenie 185 has a scaling parameter of 1.Each Genie model has an associated range of bitsizes. Bit head design geometry is linked to asecond independent scaling parameter, the bitgauge diameter. A custom bit of any size betweenthe specified minimum and maximum size can beproduced. Design changes are more easilymanaged. Manufacturing is more flexible andsimpler with improved quality and lead time as theresult of feature commonality through the range.Model selection & model numberThe housing diameter in each Genie model ismatched to BHA collar sizes whilst keeping thetarget bit gauge within range. The tool housingdiameter in mm (rounded to nearest 5) providesthe model number. The model number is notlinked to any particular bit size and bit size rangescan overlap between some models.Size & transportGenie tools were designed with consideration forcarrying in pick-ups and helicopters. They havean integral twin check valve. The Genie 185 is52

Australian Geothermal Conference 2010under 2m long, including the integral jet sub andis half the size and weight of predecessor tools.Connecting threadsGenie tools have pin down API female topthreads for protection during handling and forreduced length & mass. Genie bits are retained ina “pin up”, splined drive sub (chuck) which screwsinto the Genie tool using a new design thread.Model rangeCurrent planned models include Genie130 (6” to6⅞” bits), 160 (7½” to 8⅝”), 185, 200 (9⅜” to 11”),260 (12” to 14¼”) & 320 (14⅝” to 17½”). The 320& 260 are next to be made. Smaller tools involvemore manufacturing challenges and are expectedto follow the bigger tools. A Genie 100 coring toolhas been considered.HydraulicsThe flow rate of fluid to power a new Genie toolover its operating range is generally lower thanthe flow rate used to clean the hole. The Genietool therefore has an integral jet sub with a dualcheck valve. It enables jet nozzles to be fitted asmay be required to provide for any additional flowrate at the same differential pressure. As adynamic mechanism, the tool will be subject togradual wear through its life at a rate dependingon the amount of sand or other abrasive materialsin the drilling fluid. It is important for economicreasons to keep the drilling fluid as free of cuttingsas practically possible. Wear may becompensated for by increasing the pump rate. It isrecommended that a working flow rate range beprovided to maximise tool operating flexibilitythrough its life. The Genie tool stops impactcycling when lifted off bottom and additional flowpaths open within the tool to provide a more directand lower flow resistant passage through theexhaust ports in the bit face. This enables higherflow rate bottoms-up circulation at lowerdifferential pressure. It also facilitates passage ofLCM. Fluid can become more abrasive under lostcirculation control conditions.Differential pressure & flow rateThe flow rate to run the tool from low to highpower, and the resulting differential pressure,depend on the fluid. As an example and using adrilling fluid of 1.05 SG with a Newtoniankinematic viscosity of 5.55 mm²/s (~35 secondMarsh funnel), the flow rate is approximately 0.61to 0.87 m³/min for a new tool and about 0.89 to1.33 m³/min for a half-worn tool. Thecorresponding range of differential pressure isabout 6.21 to 11.03 MPa. This is much lower than12.76 to 17.24 MPa of the predecessor tools.Weight On Bit (WOB)A low WOB of 1.1 to 4.5 kdaN is required for theGenie 185 with an 8½” (215.9 mm) drill bit. TheWOB for a rotary drilling system of the same sizemay exceed 20 tonnes for hard formations. WOBoutside the Genie limits will cause a reduction inimpact energy delivered to the bit face, a loss ofROP and an increase in drill string torque.Bottom Hole Assembly (BHA)A for ease of control, the lower BHA is isolatedfrom the rest of the drill string by a fully pressurebalancedheave compensator type stroke sub.The applied WOB is thus limited to the weight ofthe items below it. See Fig. 2 for an example.TARGET WOB = 1.1 to 4.5 kdaNLengths & weights in example estimatedonly. Detailed drill string & BHA design toachieve target WOB is the responsibilityof the client.StabilizerUnder gauge, roller reamerrecommended. Spiral stabilizer moreprone to getting hooked on ledges.X-OverFully pressure balanced stroke subUpper partTelescopic joint (only items below thisjoint apply WOB)eg: Smiths HE 6½” Hydra-Stroke AEBBfully pressure balanced, or equivalentNOT A THRUSTER.Lower partPony collar or CollarGenie 185 integral jet sub & checkvalve systemGenie 185Genie 185 bitFIG. 2 – GENIE 185 BOTTOM HOLE ASSEMBLY53

Australian Geothermal Conference 2010Project statusApproaching sign-off on the Genie 185 tool withits hard surfacing developments.Relevance to Geothermal projects SUITABLE for HARD formations(metalliferous mining origins) HIGH TEMPERATURE RATING – 204Cstandard & 260C high temp. version HIGH ROP – up to 3 times offset rotary LOW WOB (1.1 to 4.5 kdaN for 8½” hole) LOW rotational speed (typically 25 rpm) LOW DRILL STRING STRESS LIGHT BHA – faster trips with less rig power API top thread & housing suits collar size LOW VIBRATION NO SEALED BEARINGS SIMPLE – only 3 moving parts plus bit ONE PART CUSTOMISED to mud DIRECTIONALLY STABLE EASY TO USE – manual or auto drill LOWER OFF–BOTTOM CIRCULATIONPRESSURE RELIABLE – has always started on bottom RISK MANAGEMENT – no fishing & no bitslost on bottom (since 1993 start) AVAILABLE – designed & manufactured inAustralia & based in Adelaide COST EFFECTIVE – cheaper drill string, nomud motor, less string wear, light & compactfor transport, integral check valve & jet sub,potential to complete drilling in less than50% of rotary drilling time for earlyproduction.Abbreviated development chronology1992 A system was conceived by Fred Moir,founder of S.D.S. Pty Ltd to drill 5¼" blast holes of20 m depth in hard metalliferous mines usingcoaxial drill rods and a down-hole hammerpowered by clean, high pressure, recirculatedwater. He sought the benefits of top-holehydraulic hammers, down-hole air hammers and abenign, dustless drilling fluid. Used water returnedto the surface in drill rods for re-use. A controlledportion was exhausted through the bit for holecleaning. Malcolm McInnes was contracted toundertake system R&D. Trialled in undergroundmines with Western Mining and Pasminco withspeed and dust suppression advantages.1993 Performance diminished with depth dueto “water hammer” vibration, the result of unevencyclic flow resistance. Novel down-hole hydraulicaccumulator systems had limited success.Managing the effect was replaced with managingthe cause. A new concept impact drill (see Fig. 3)with progressive, staged pistons, cyclic flowresistance levelling, anti-cavitation impact facesand a new constant-exhaust bit was invented anddesigned by Malcolm McInnes who then won a$500,000 competitive Government R&D grant andcompleted the project to budget. Extensivemathematical system modelling was undertaken.FIG. 3 – 1993 HYDRAULIC PERCUSSIVE DRILLThe compact new, patented design providedsmoother, more powerful operation (see Fig 4)independently of depth. Recirculation through drillrods was abandoned for hole cleaning.FIG. 4 – WATER HAMMER PROTOTYPE TESTING WITHGRANITE & UNDERGROUND1994 Further underground testing with Mt.IsaMines. Total distance drilled 1000m.1995 Attracted to petroleum industry with moresuitable pump & fluid infrastructure and morecommercial potential. Developed tool coatings forcontaminated fluids. Sub project to accommodate54

Australian Geothermal Conference 2010denser drilling fluids put on hold. Mixed results inoil field trial with partner Santos. Partly related tocontrol & instrumentation issues and a packer.casing. Development progressed in offshore trialswith the ODP and sea water drilling fluid.1996 Breakthrough field trial of 6" WaterHammer with Shell in Canadian Rockies foothillswith water. Drilled in a pipe stand in an 8½” holeat ~4.5 times ROP of roller cone using manualand automatic drilling.FIG. 7 – FLUID HAMMER 260 WITH ODPFIG. 5 – FRED MOIR & MALCOLM McINNES AT WATERHAMMER RUN, CANADA 1996.1998 Ran 12¼" tool off shore from jack-up rigthrough short hard layer with ARCO Indonesia(see Fig. 8). Tool put in at 1,350 m & drilled 11hours at higher than offset ROP (up to 5 x). Thiswas despite some damage incurred to the bit dueto tungsten carbide inserts left in the hole by thepreceding roller cone.Began development of Water Hammer for 12¼"bit to suit same commercial application (water).FIG. 8 – WATER HAMMER & ARCO INDONESIAFIG. 6 – CLUSTER WATER HAMMERDeveloped and successfully trialled 32” holecluster drilling tool based on 4 water hammers forJ C McDermott & Sons. See Fig. 6. Potential for26” conductors & sub-sea anchorages.Manufacturing & drafting moved from SDS Pty Ltdin Adelaide to SDS Digger Tools in WA.1997 New 12¼" hole size water hammerdesigned & made. Immediately worked inCanada. 30 hour running time with new hardcoating technology and flock water.Developed case-while-drilling systems for OceanDrilling Program. Achieved fast, unsupportedspudding on hard, smooth, 30° sloping granite.Quarry tested CWD insertion and retrieval ofContinued drilling 650 m surface holes with 14¼"bits with Talisman Energy Inc at ROP consistently2 to 3 times faster than conventional rock bits inhard formations. See Figs. 9 & 10. A majordownturn in the international mining industrythreatens the internally funded project.FIG. 9 – WATER HAMMER – CHERT CUTTINGS55

Australian Geothermal Conference 2010Depth (m)SDS FLUID HAMMER 260 - CASE STUDIES in the CANADIAN ROCKIES FOOTHILLSTime (days)0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 240Black = Air hammer drills100Red = Roller cone bitsBlue = SDS Fluid Hammer200300400500SDS Corporation Ltd commissions anindependent technical review of the design. Theengineering and technology is validated.2000 Drilled with PDVSA in Venezuela at 4,355m depth (2 to 3 x offset instantaneous ROP) withSDS water hammer 260 (12-1/4" bit). The fluidwas a mineral oil of about 1.03 SG. WOB difficultto control. See Fg. 13.600700Talisman E/A Highhat d-78-K/93-P-5 Phillips Sukunka c-54-F/93-P-5 PC Gwillim B-78-H/93-P-5OEI Petrorep Sukunka a-39-F/93-P-5 Talisman Sukunka A-18-K/93-P-5 Talisman Sukunka B-11-L/93-P-5Shell Esso Gwillim a-51-H/93-P-5 Talisman POCO Highhat a-54-K/93-P-5 Talisman Highhat d-83-L/93-P-5FIG. 10 – COMPARATIVE PERFORMANCEShoreline approach hole opening job at StatoilMongstad refinery, Norway with 12¼” holeopening bit on slope rig run by Halliburton. Fig 11.FIG. 14 STEERING HYDRA 185. FIG. 13 PDVSAFirst steering run using unfinished prototypeHydra 185 (8½”) at Sperry-Sun Nisku test facility.The tool built a turn of 5° using a 1° bent sub and9°using a 1.5° bent sub, in 18.3 m of concretewhich it drilled at 100 m/hr (Fig. 14). Successfullydemonstrated steering and operation in a wellwith Measurement While Drilling System (Hydra185 not developed for fluid SG over about 1.05).FIG. 11 – WATER HAMMER 260 & STATOIL1999 Began development of new 8½” hole toolwith fewer parts, new features and for range ofdrilling fluids (not completed), as new series basemodel. Performance testing undertaken in OrangeGrove quarry, WA. See Fig. 12. Drilling with 12¼”tool with Northstar energy - Recorded continuousROP 4-7 Mt/Hr Vs 2 Mt/Hr tricone. Hammer lifeincreased to 40 hours.2001 May. Controlled surface benchmarktesting of SDS Fluid Hammer 185 at TerraTektesting facility (see Fig. 15) in Salt Lake City,Utah, USA, separately and in conjunction withconsortium lead by US Department of Energy.Facility simulates deep, overbalanced drilling withstandard rock samples and drilling muds. Despitetool not being designed for 1.2 and 1.8 SG muds,Extensive testing confirmed the soundness of theconcept, verified operating characteristics andinspired development to suit denser fluids andexpand application versatility.FIG. 12 – QUARRY TEST 8½” PROTOTYPESFIG. 15 – TERRATEK & FLUID HAMMER 18556

Australian Geothermal Conference 2010Fluid Hammer 260 tool (12¼” bit) was trialled withPDVSA in well MCA-2X in Venezuela at 1,498 m.A new drilling technique had been developed intheory and was successfully used to control thetool and Weight On Bit for optimum ROP, a majorbreakthrough that also protected the bit. Workeddespite failure of hook feed indicator. Aninstantaneous ROP 2 to 3 times off-set roller conewas demonstrated. Project investment andfinance options were pursued for the board.2002 June. Verification field tested prototypeSDS Fluid Hammer 185 (9⅞” bit) with TalismanEnergy Inc. & Precision Drilling. See Fig. 16.Used new design Bottom Hole Assembly. BHAprovided required low Weight On Bit. Identifiedtorque as a key WOB maximum & minimumindicator. System worked and overcame majorhurdles in applying the technology commercially.FIG. 17 – FLUID HAMMER 260 & CWD & IODP JOIDESRESOLUTION2005 Successfully completed IODP jobJanuary, 2005. 2 tools ran a total of over 92 hrs insea water without significant performance loss.The revenue supported the Genie project for thefirst 18 months. See Fig. 18.FIG. 18 – IODP RE-ENTRY CONE ON FLUID HAMMER 260& CWD INSERTED CASINGFIG. 16 – PROTOTYPE 185 & BHA. TALISMAN ENERGYINC & PRECISION DRILLINGFluid Hammer project suspended.2004 Project and its assets moved from SDSDigger Tools into new Adelaide based entityIMPACT DRILLING INTERNATIONAL Pty Ltd. tofocus on development and commercialisation ofthe technology for the petroleum & geothermalindustries. Genie project announced. New design,engineering, manufacturing and supply, testing,operations, quality systems, branding and fundingto develop new range of tools for industry drillingfluids and with lower operating differentialpressure to maximise versatility. Managed byMalcolm McInnes. Concurrently with new tooldevelopment, the company won a commercialcontract with Integrated Ocean Drilling Program touse two 1997 design 260 tools with IODP HardRock Re-entry Systems (up to 15” bits) to insertcasing with re-entry cones 30m into AtlanticOcean floor, (1.5 to 2.5 km depth) north of theAzores, without a template. See Figures 17 & 18.Designed & built in-house, impact drilldynamometer test rig for new Genie range R&D(see Fig. 19) with mud cooling, mixing,instrumentation, prime pump & twin HT400pumps. Many components were reclaimed scrap.FIG. 19 – IDI GENIE DYNAMOMETER TEST RIGSuccessfully tested two different prototypes (seeFig 20) on water and standard muds 1.2 & 1.8 SG(as used in USA benchmark testing), withoutsignificant hydraulic vibration and at lowerdifferential pressures. Different valvessuccessfully used for different fluids. The partssupplier was late and the order incomplete.Enough parts provided for 2 complete Genieprototypes and one spare set of internals.57

Australian Geothermal Conference 2010FIG. 20 – A PROTOTYPE GENIE 185All part drawings parametrically linked and alldrawings of the models in the product rangelinked to simplify and speed-up drafting, improvequality, simplify and standardise manufacturing,reduce lead times & costs. Check valve systemredesigned for high solids mud. Safer, easier &faster assembly procedure and tools developed.2006 SDS Corporation Limited sold to SandvikMining & Construction, Adelaide.IMPACT DRILLING INTERNATIONAL Pty Ltdacquired by Fred Moir & minor shareholders.2007 Prototype Genie 185 design &manufacturing completed. Business namechanged to SPECIALISED DRILLING SERVICESAUSTRALIA Pty Ltd.2008 Prototype Genie 185 tools run with fieldtrial partner Talisman Energy Inc in CanadianRockies foothills, 5 to 9 January 2008. Fig. 21.The new Genie 185 system demonstrated: Directional stability. Deviation no more than ¾° Low rig mechanical & hydraulic vibration. High ROP, average 3 times offset tool ROP. Easy auto & manual drilling control. Reliable & predictable starting & operation. Automatically regulated, low WOB withlightweight BHA. Different but easy to learn to use. Can pass loss control material.R&D program to increase durability. Prototypesdid not employ hard overlays as used inpredecessor tools. Accelerated wear wasobserved during lost circulation management dueto increased sand level in drilling fluid. Newdesign features, materials and pioneeringmanufacturing processes developed andincorporated into the Genie tools. A new dualcheck valve system was developed.FIG. 21 – GENIE 185 TRIAL WELL, CANADA2009 R&D and prototype manufacture of newhard overlaid tools completed in second half ofyear. New production systems developed for highvolumes and low lead time. Additional field trialpartners sought after follow-up drilling programdelay in Canada due to financial crisis in USA.Local opportunity provided with contract won inlate 2009.2010 Tools mobilized to rig. Returned unusedafter anticipated hard formation was notencountered. Tools on stand-by for new project.Contract won for trial managed pressure drillingrun in Asian project during last quarter of year.Development of below-the-hook lifting systems,maintenance procedures, manuals, drawings andother documentation continues.ManagementChairman of the Board of Directors and principalshareholder is Mr Fred Moir. Fred is aninternational business leader with extensivedrilling, service provision and manufacturingexperience in the mining industry. Fred wasfounder of SDS Pty Ltd and instigated the originalproject to develop a water powered percussiondrilling tool in 1992. Impact Drilling InternationalPty Ltd was established under his Chairmanship.The business is managed by Malcolm McInnes,FIEAust. CPEng MESc BE ARMIT Mech Eng MAIE SPE withover 30 years experience in manufacturing,product development, R&D, marketing, sales &management and has run his own business since1992. He is the designer and inventor of theGenie tools and inventor of the of theinternationally patented SDS water/fluid hammerin 1993. Malcolm has 17 years experience in thedrilling industry and became an employee andshareholder of SDSA in 2006.ReferencesMcInnes, M.B. - Internal reports & photographs &company documents & photographs 1992 - 2010.The Author acknowledges the contributions of allpeople involved in the project since its inception58

Australian Geothermal Conference 2010Hydrothermal alteration aspects of high heat producing granites inAustralia & EuropeAlexander W. Middleton* 1 , I. Tonguç Uysal 1 , Massimo Gasparon 2 , and Scott Bryan 3Queensland Geothermal Energy Centre of Excellence, The University of Queensland, Brisbane, QLD. 4072* 1School of Earth Sciences, The University of Queensland, Brisbane, QLD. 4072 2Queensland University of Technology, BiogeoScience, GPO Box 2434, Brisbane, QLD. 4001 3In order to fully understand the relativelyunstudied high heat producing granites (HHPG) ofthe Australian continent, this project has studiedthe European analogues of Cornwall, (UnitedKingdom), Soultz-sous-Forêts (France) and theErzgebirge (Germany). HHPGs are characterisedby their anomalous radiogenic element (U, Th andK) contents (Table 1), which is higher thanaverage continental crust values. The Big LakeSuite in the Cooper Basin is the most investigatedAustralian enhanced geothermal system (EGS);Gatehouse et al., 1995).Alteration mineralogy is a widely used feature toexplore for potential ore deposits; however, thismethod has not yet been deployed for identifyingand characterising EGS. In the current project, weare investigating alteration mineralogy of HHPG,with particular emphasis on trace element andstable isotope geochemistry. If successful, thiscan offer an additional approach to geothermalexploration and resource characterisation.phase disequilibrium with the fluid. The alterationassemblage produced depends on factors suchas wall-rock and fluid composition, temperature,salinity and water/rock ratio. The variation inalteration assemblage may control theconcentration of incompatible elements, includingthe radiogenic elements as certain alterationreactions are associated with either enrichment ordepletion of these elements. A pertinent exampleof this was noted to have occurred in someCarnmenellis (Cornwall, UK) monazites(Poitrasson et al., 1996), whereby Th wasenriched through brabantitic (Förster, 1998) cationexchange during chloritisation of biotites:2REE 3+ Th 4+ + Ca 2+Incompatible elements (HFSE, REE, LILE, and Betc., including the radiogenic elements) have alow partition coefficient (K D ) (White, 2007), andhence will preferentially stay in fluid phases duringTable 1: Typical compositions for high heat producing granites for EGS. Sourced from T Uysal, J Van Zyl, S Bryan, unpubl.data; Charoy, 1986; Stussi et al., 2002; Forster et al., 1999; Chappell & Hine, 2006EGS Target Composition Age SiO 2 K 2 O U (ppm) Th (ppm)(Ma) (wt%) (wt%)Cooper Basin granites Biotite granites ~Mid-Carb.- - 10.0-30 28-144Carnmenellis granites(Cornwall, UK)K-feldspar megacrysticbiotite granites293-27469-76 4.6- 13-20 15-205.9 (4-38) (5-45)Soultz-sous-Forêts(Rhine Graben, France)Erzgebirge (Germany)Keywords: Geothermal, alteration mineralogy,Cooper Basin, stable isotopes, REE, Soultz-sous-ForêtsBackgroundK-feldspar porphyriticmonzograniteBiotite, 2 mica & LimicagranitesDuring emplacement of granitic bodies, advectiveheat flow can occur due to hydrothermalcirculation of surrounding waters. Several studieshave noted that igneous rocks enriched inradioactive elements such as U, Th and K canhave a similar effect and promote hydrothermalcirculation long after the intrusion has cooled(Allman-Ward, 1985, Pirajno, 2005). Oncirculation of these hydrothermal fluids, variablealteration products will result from original mineral~330 67-69 ~4.5 6.2-14.1 24-37325-31870-77 4.5-5.3crystallisation of melt and hydrothermal fluidmovement. .Hydrothermal fluids affecting surrounding wallrockscan be classified (based on their origin), intometeoric, magmatic, marine or connate. Oneaspect of this project is to deduce the origin andevolution of the hydrothermal fluids involved in thealteration of HHPG and surrounding sedimentaryrocks.Analytical Work8.0-30 15-37The project results will be attained viamultifaceted petrographic, geochemical andgeochronological analyses. The geochronologywill use Rb-Sr isochron dating of alteration-relatedillitic clay minerals, whereas the geochemicaltechniques will include both stable isotope59

Australian Geothermal Conference 2010analysis (δ18O and δD) for hydrothermal fluidevolution and ICP-MS analysis for trace elementquantification.Qualitative and quantitative petrographic analyseswill be achieved with transmitted and reflectivemicroscopy, as well as EDS (via JEOL XL 30SEM) and EPMA. The EPMA will be used todetermine quantitatively which mineral phasesencompass the elements of interest; such asincompatibles and radiogenics.Queensland EGS targets and CooperBasinQueenslandFollowing estimated subcrustal temperaturemapping and the increasing need for renewableenergy sources, Queensland has been targetedby the state government as a prospective area forgeothermal energy. The principal areas of interestfor EGS are the Galilee Basin, Innot Hot Springsregion, Hodgkinson Province, Styx Basin,Maryborough Basin and North d’Aguillar Block,Wandilla Province. These targeted areas havebeen based on the presence or likely presence ofHHPG at depth ranging in age from the LateDevonian to Cretaceous. Many of the granitebodies emplaced within Queensland are linked toan extensional tectonic origin (Glen, 2005).Analysis of the intrusions, overlying sedimentsand their associated alteration zones will beperformed as part of this project.Cooper BasinThe Carboniferous Big Lake Suite granite in theCooper Basin’ represents one of the world’shottest EGS (Gatehouse et al., 1996). Accordingto our recently acquired ICP-MS trace elementdata, U and Th contents in these granites reachlevels of up to 144ppm, reflecting enrichmentlevels up to 13 times that of the upper continentalcrust (UCC( McLennan, 2001)). Cores taken fromthe granite and overlying sediments show varyingdegrees of alteration, with a range of incompatibleelement enrichment, such as U and Th. Thehighly altered zones have a predominant greisenstylesericite (illite) and re-precipitated quartzassemblage. We believe that this alteration maywell have caused the localised enrichment inradiogenic element-bearing minerals such as illite,K-feldspar, and some accessory minerals (e.g.,thorite ).The fluid history of the Cooper Basin can bededuced with the use of both crystallinity andstable isotope analyses of the illite. Illitecrystallinity is a useful indicator of the temperaturegradient in active geothermal systems and forlocating fossil hydrothermal systems associatedwith ore deposition (Ji and Browne, 2000).Depth500100015002000250030003500400045000.2 0.4 0.6 0.8 1Illite Crystallinity Δ2θMB 1graniteBL 1 graniteMcL 1 graniteJol graniteHab 1 graniteMB 1‐2‐7 sedimentBL 1‐2‐31‐57 sedimentMcL 1 sedimentTW 1sedimentWT 1sedimentFig 1: Graph of illite crystallinity against depth from CooperBasin coresIllite crystallinity is controlled by crystallisationtemperature, water/rock ratio, and time availablefor crystallization (Arkai, 2002; Ji and Browne,2000; Merrriman and Frey, 1999). Betterdevelopedcrystalline illites show narrower 001basal illite peaks and have lower IC values. Suchillites were formed at higher temperatures orduring prolonged heating events. Higher ICvalues (wider peaks), on the other hand, indicatelower crystallisation temperatures and/or rapidprecipitation during hydrothermal processes. Illitecrystallinities are seen to progressively increasewith increasing core depth, suggesting a highercrystallisation temperature and hencehydrothermal fluid temperature at depth. Thegranite intersected in Jolokia and McLeod 1seems to have experienced highest reservoirtemperatures.According to our present isotopic studies ofalteration-related illite within the granite andsedimentary cover, oxygen and hydrogen isotopecompositions range from δ18O = -1.8‰ to+2.7‰;δD = -99‰ to -121‰ and δ18O = +2.3‰to +9.7‰, δD = -78‰ to -119‰, respectively.Such values are much lower than those reportedfor many deeply buried sedimentary basins(Clauer and Chaudhuri, 1995). The calculatedoxygen and hydrogen isotope compositions of60

Australian Geothermal Conference 2010fluids in equilibrium with the illites are depleted in18 O and deuterium, comparable to those of watersreported for most high-latitude sedimentarybasins. Hence, stable isotope data of alterationminerals in the granite and the overlyingsedimentary rocks suggest the operation of ahydrothermal system involving high latitudemeteoric waters during Permo-Carboniferousextensional tectonism in the Cooper Basin region.EuropeThe European HHPG examples have all beenconsidered to be approximately syn-genetic withextension following the late PalaeozoicHercynian/Variscan orogeny between Gondwanaand Laurasia.extension (Schleicher, 1998). Petrographicstudies of alteration mineralogy by Bartier et al.(2008) identified intense propylitically alteredgranite with newly equilibrated mineralassemblages of tosudite (chlorite-smectite), illite,chlorite and carbonates, whilst K-feldspar appearsto be largely preserved.Following further transmitted and scanningelectronmicroscopy performed at The Universityof Queensland, we identified several keyalteration-related mineral phases. Altered sphene,for example, found in sample K177 (2000m –EPS1), is acting as a principal mineral phase thatcontains various rare earth elements (REE).Estimations of composition from semi-quantitativeEDS have detected Ce, Y, Nd, La, Th and Ca withpossible PO 4 or SiO 4 anion complexes.***Sphene*Cornwall*Soultz-sous-Forêts*Erzgebirge064Fig 2: Possible tectonic suture configuration of Hercynianorogeny from Matte (2001)Erzgebirge, GermanyLocated at the north-western edge of theBohemian Massif (Förster and Förster, 2000), theErzgebirge consist of highly-evolved voluminousgranites intruded into highly crystallinemetamorphic rocks (Tischendorf et al., 1965).These granites are highly enriched in U and Th(Table 1) with principal radiogenic mineral phasesof uraninite, U-bearing micas, xenotime,monazite-(Ce)-brabantite solid solution seriesminerals, and thorite (Förster, 1998).As of yet, our analyses (Rb – Sr dating and stableisotope) have not been undertaken on Erzgebirgesamples, but samples will be analysed incollaboration with Dr. Hans-Jurgen Förster(Potsdam, GeoForschungsZentrum).Soultz-sous-Forêts, FranceLocated south-west of the Erzgebirge outcrops,Soultz-sous-Forêts was emplaced debatablyeither as an I-type subduction-related or latestageextensional granite body (Flötmann andOncken, 1992). Hydrothermal fluid circulation andalteration was promoted by E-W lithosphericFig 3: Altered sphene compassing several REE-elementbearingmineral phases30 µmThis may indicate a potentially high fluid/rock ratioallowing for incompatible element transport duringalteration of the granitic protolith. Monazite hasalso been found proximal to subhedral hematitegrains. A possible mechanism for this mineralassociation can be traced back to the formation ofa redox interface with co-precipitation of U 4+ andFe 3+ .Counts0642700Ca Ti2400210018001500CeCe1200 OYLaLa CeY Th TiCe Ce 900C Ce Si Th CaLaYLa600 Fe Al Th KLa Fe30000.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00keVFig 6: EDS for altered sphene found in sample K177exhibiting intense propylitic alteration.Cornwall, United KingdomCornwall marks the far western district of theEuro-Hercynian granites with 5 major outcrops;Dartmoor, Bodmin Moor, St. Austell, Carnmenellis61

Australian Geothermal Conference 2010and Lands End. Multiple studies by Ball et al.,(1979) and Charoy (1986) have found theCornwall granites to be similarly enriched inradiogenic elements (Fig. 1) as the middle-European examples. Alteration systematics are ofgreat importance within this system as they havebeen directly linked to the leaching andmineralisation of uranium. Allman-Ward (1985)noted 4 predominant alteration styles affecting theSt. Austell granite; greisenisation,tourmalinisation, kaolinisation and haematisation.Uranium is likely to be transported in solutionduring hydrothermal alteration, as oxidising fluidswill cause a loss of electrons (U 4+ to U 6+ ) andhence increase mobility. The most likely locationof the displaced uranium will be in a distalalteration zone possibly accompanied byhaematite mineralisation. An example of this canbe seen in Wheal Remfry where quartz-haematiteveins show uranium values of up to 50ppm(Allman-Ward, 1985).SummaryHighly altered granites in the Cooper Basin aresubstantially enriched in lithophile elements,particularly in Cs, Rb, Be, Th, U relative to theupper continental crust. U and Th contents are 10and 13 times higher than those of the UCC. Someenrichment of heat-producing elements waspromoted by a regional hydrothermal eventleading to the precipitation of U and Th- bearingminerals such as illite, K-feldspar and accessoryminerals (other than zircon). Oxygen andhydrogen isotope compositions of fluids aredepleted in 18 O and deuterium, comparable tothose of waters reported for most high-latitudesedimentary basins. European examples ofenhanced geothermal systems show similarredistribution and enrichment patterns of rareearth and radiogenic elements as a result ofhydrothermal alteration. This can be seen inmonazites proximal to chloritised biotites in theCarnmenellis granite of Cornwall. Unlike theCooper Basin, stable isotope data has not yetbeen acquired from European samples. Howeveronce performed, hydrothermal fluid origin andevolution can be deduced for these EGS.Alteration mineralogy and geochemistry ofrelatively shallow sedimentary sections is apotentially valuable tool to evaluate the presenceof a concealed geothermal heat source in thebasement of sedimentary basins.ReferencesAllman-Ward, P., 1985. Distribution of uraniumand thorium in the western lobe of the St. Austellgranite and the effects of alteration processes,High heat production granites, hydrothermalcirculation and ore genesis conference, TheInstitute of Mining and Metallurgy, 437-458.Aquilina, L., Pauwels, H., Genter, A., and Fouillac,C., 1997. Water-Rock interaction processes in theTriassic sandstone and granitic basement of theRhine Graben : Geochemical investigation of ageothermal resevoir. Geochimica etCosmochimica Acta, 6, 4281-4295.Arkai, P., 2002. Phyllosilicates in very low-grademetamorphism: Transformation to micas. Micas:Crystal Chemistry and Metamorphic Petrology,46: 463-478.Bartier, D. et al., 2008. Hydrothermal alteration ofthe Soultz-sous-Forêts granite (Hot FracturedRock geothermal exchanger) into tosudite andillite assemblage. European Journal ofMineralogy, 20, 131-142.Bryan, S. E., Holcombe, R. J. and Fielding, C. R,2001. Yarrol Terrane of the northen New EnglandFold Belt: forearc or backarc? Australian Journalof Earth Sciences, 48 296-316Clauer, N. and Chaudhuri, S., 1995. Clays inCrustal Environments. Isotope Dating andTracing. Springer Verlag, Heidelberg, 359.Flötmann, T. and Oncken, O., 1992. Constraintson the evolution of the Mid Crystalline Rise –astudy of outcrops west of the river Rhine.Geologische Rundschau, 81, 2, 515-543.Förster, H-J., 1998. The chemical composition ofREE-Y-Th-U-rich accesspry minerals inperalumnous granites of the Erzgebirge-Fichtelgebirge region, Germany, Part 1: Theminazite-(Ce)-brabantite solid solution series.American Mineralogist, 83, 259-272.Förster, H-J., 1998. The chemical composition ofREE-Y-Th-U-rich accesspry minerals inperalumnous granites of the Erzgebirge-Fichtelgebirge region, Germany, Part 2:Xenotime. American Mineralogist, 83, 1302-1315.Glen, R. A., 2005. The Tasminides of EasternAustralia, from: Vaughan, A. P. M., Leat, P. T. andPankhurst, R. J., 2005. Terrane Processes at theMargins of Gondwana. Geological Society,London, Special Publications, 246, 23-96.Ji, J. and Browne, P.R.L., 2000. Relationshipbetween illite crystallinity and temperature inactive geothermal systems of New Zealand. Claysand Clay Minerals, 48: 139–144.Ledésert, B. et al., 2009. Fractures, hydrothermalalterations and permeability in the Soultz EnhanceGeothermal System. C. R. GeoscienceMATTE, P., 2001. The Variscan collage andorogeny (480±290 Ma) and the tectonic definitionof the Armorica microplate: a review, Terra Nova13, 122-128.McLennan, S. M., 2001. Relationships betweenthe trace element composition of sedimentaryrocks and upper continental crust. Geochem.Geophys. Geosyst, 2.62

Australian Geothermal Conference 2010Merrriman, R.J. and Frey, M., 1999. Patterns ofvery low-grade metamorphism in metapeliticrocks. In: M. Frey and D. Robinson (Editors), LowGrade Metamorphism. Blackwell Science,Cambridge, pp. 61-107.Murray, C. G., 2003. Granites of the New EnglandOrogen., from: Blevin, P. L., Jones, M. andChappell, B., Magmas to Mineralisation: TheIshihara Symposium. Geoscience AustraliaRecord, 2003/14, 101-108.Pirajno, F., 2005. Hydrothermal Processes andMineral Systems. Springer Link, GeologicalSurvey of Australia, 73-157, 990-995.Schleicher, A. M., Warr, L. N., Kober, B., Laverret,E. and Clauer, N., 2006. Episodic mineralisationof hydrothermal illite in the Soultz-sous-Forêtsgranite (Upper Rhine Graben, France). ContribMineral Petrol, 152, 349-364.Stussi, J-M. et al., 2002. The hiddenmonzogranite of Soultz-sous-Forêts (RhineGraben, France): Mineralogy, petrology andgenesis. Géologie de la France, 1, 45-64.Taylor, H. P., 1997. Oxygen and HydrogenIsotope Relationships in Hydorthermal MineralDeposits, from: Barnes, H. L., 1997.Geochemistry of Hydrothermal Ore Deposits.John Wiley and Sons, Inc, 229-288.White, W. M., 2007. Geochemistry: TraceElements in Igneous Process. John HopkinsUniversity Press, 258-312.63

Australian Geothermal Conference 2010A Geothermal Web Catalog Service for the Perth BasinThomas Poulet 1,2,3 , Soazig Corbel 1,2 , Michael Stegherr 1,41CSIRO Earth Science and Resource Engineering, P.O. Box 1130, Bentley, WA 6102, Australia.2Western Australian Geothermal Centre of Excellence, P.O. Box 1130, Bentley, WA 6102, Australia.3The School of Earth and Environment, University of Western Australia, WA 6009, Australia.4 School of Science, Curtin University, GPO Box U1987, Perth, WA 6845, AustraliaThe Western Australian Geothermal Centre ofExcellence (WAGCOE) is a science collaborationbetween Western Australia’s leading innovativeresearch institutions – CSIRO, The University ofWestern Australia and Curtin University ofTechnology. One of its key research initiatives isto assess geological and geophysical data fromthe Perth Basin to identify geothermal targets.This geothermal exploration task involves thecollection and analysis of all datasets available,which can be a challenging and sometimesdaunting exercise. This paper addresses thedifficulties encountered in gathering data forgeothermal exploration and presents our webcatalog service, the application we use to manageour spatially referenced resources.Keywords: Western Australia, Perth Basin,geothermal exploration, GeoNetwork opensourceFinding and using geo-referenced dataGeothermal exploration is a data intensive activitywhich requires access to all relevant informationavailable in the region of interest, such asstratigraphic sections and horizons, downholetemperature measurements, rock properties, andhydrogeologic parameters including water quality.Because so many types of sciences are involved(geophysics, geology, geochemistry...), data maybe derived from many different potential sourcessuch as seismic surveys, petroleum well logs,mechanical property laboratories, or groundwatermineralogical assays. Data really is anindispensable element for exploration and yet it isstill so difficult and expensive to discover, search,access and use; those four simple steps cansometimes represent a real obstacle course forany first-time user. Although most public datasetsare nowadays accessible through companies’ orgovernmental agencies’ websites, any potentialuser still needs to know all key data holders in thearea in order to access essential information.Once the appropriate data custodian for aparticular type of data is identified, the secondtask is usually to search for data relevant to thelocation of interest, only to discover that datasetsare rarely fully searchable. The third step is thento access this data, which can sometimes involvegoing physically to an organisation to copyseveral computer hard drives. The last step isgenerally to understand the format used, whichmay be non-standard and can lead to the mostsignificant inefficiencies in a given project. Datamanipulation can include for example largeamounts of preliminary data conversion,deciphering any implicit assumption about units ofmeasures and geographical references, or eventrying to infer the creation process of the data inorder to obtain one piece of missing information. Ifdata is interrelated, this investigative process canloop on itself, creating workflow problems. Anyuser who has spent a long time going through thisprocess can indeed wonder legitimately if newerdata is now available. Our web catalog service isan attempt to eliminate the painful aspects ofPerth Basin data discovery.Using GeoNetwork opensourceGeoNetwork opensource 1 is a discovery serviceand catalog application, which is useful formanaging spatially referenced resources andspecially built to facilitate the connection betweenspatial information communities and their data. Itis a Standardised Geospatial InformationManagement System based on internationalstandards, with a large community of severalhundreds of users and developers participatingactively to its growth. A more complete descriptioncan be found in Ticheler & Hielkema (2007). Incollaboration with AuScope Grid (Woodcock2008) we decided to use GeoNetwork as ourcatalog application since it addresses mostpotential problems mentioned above, of datadiscovery, search, access and usage.GeoNetwork is accessible to users through a webinterface and therefore only requires a webbrowser on any computer platform. Its searchfunctionalities make data easily discoverable andFigure 1: Workflow diagram showing the architecture ofWAGCOE’s geothermal catalog, using GeoNetwork as thecatalog application linking to a petabyte datastore holding thecorresponding data. Both components use iVEC’sinfrastructure and provide an efficient and secure system thatcan connect to other web services1 http://geonetwork-opensource.org64

Australian Geothermal Conference 2010searchable. Since we are using GeoNetwork as acataloguing application, the granularity of thesearch is limited to the metadata and not the datacontent itself. This library function is different froma web portal connected to a full database whichwould make the entire data fully searchable.However, the search facility allows users to findinformation easily and access the correspondingmetadata, including a link to download theassociated data itself from a Petabyte store.Figure presents the workflow of any useraccessing WAGCOE’s web catalog service. Boththe web service and Petabyte store are hosted atiVEC 2 , the hub of advanced computing inWestern Australia, and all connections aresecure. The web catalog service points directly tothe data store, and can also be connected toother web catalog services from other institutions.Using that data is also greatly facilitated asGeoNetwork organises, documents and publishesdata collections in a standardised and consistentway, both at the metadata and data levels,following international standards from the OpenGeospatial Consortium 3 . This fully descriptivemetadata in XML format improves accessibilityand removes any possibility of datamisinterpretation based on assumptions.As Ticheler & Hielkema (2007) point out,GeoNetwork has also been developed using Web2.0 techniques to allow for more interactive andintuitive use of the system and to offer buildingblocks for future web services. This interactionwith other web services is a central feature to helpreduce duplication and working towards a singlepoint of truth paradigm. It is an essentialcomponent of WAGCOE’s data strategy to link toall data custodians’ databases as defined by theAustralian Geothermal Energy Group (AGEG)Technical Interest Group focusing on datamanagement. Such a framework can definitelyimprove geo-referenced data sharing within andbetween organisations, hence developing bettercollaboration. It can also open opportunities fordata comparisons which can empower consumersFigure 2: screenshot of WAGCOE’s GeoNetwork web interface. All data are searchable through metadata, including keywordsand geographical location. Results are displayed with all corresponding metadata including links for download of the underlyingreal data.2 http://www.ivec.org3 http://www.opengeospatial.org65

Australian Geothermal Conference 2010(Rezabakhsh et al., 2006). In our case the firstexternal web service we are planning tointerconnect is the web interface to CSIRO’sPressurePlot 4 , a query tool linked to thePressureDB database. That database containsrock properties such as porosity, permeability andthermal conductivity measurements, as well asstratigraphic and downhole pressure data formore than 1700 wells across Australia.WAGCOE’s web catalog serviceFigure shows a screenshot of WAGCOE’sGeoNetwork web interface 5 with three entriesvisible. Each entry is represented by its title andabstract, as well as two small pictures. Athumbnail on the right hand side represents thecorresponding data and a logo on the left handside indicates visually the origin of the data. Allcurrent entries appear with WAGCOE’s logo asGeoNetwork is not connected to any otherexternal service yet. Part of the data we store wasproduced by WAGCOE and includesmeasurements (see Figure ) as well as resultsfrom numerical models. Most of the underlyingdata however was not originally created byWAGCOE and comes from other organisationssuch as the Western Australian Department ofMines and Petroleum, Geoscience Australia, orother private companies, research andgovernmental organisations. This local hostingdemonstrates the temporary nature of thoseentries as we are working toward the single pointof truth principle enunciated previously. We areonly hosting other organisations’ data with theirpermission as they agree for us to do whilebuilding their own web services. We are alsoplanning to host data for geothermal companieswho choose to use our system to manage part oftheir data.As shown on Figure , our infrastructure isdeployed at iVEC and benefits therefore from theirprofessional services. All connections toGeoNetwork are encrypted and secure. A customgroup management function then provides thegranularity required to open some data to thepublic while restricting the access to protectedcompany data, for example.The data stored in our catalog covers all majorscientific areas relevant to geothermal explorationand GeoNetwork’s advanced options allow usersto search for entries using the following two levelsof categories and sub-categories:• remote sensing: thermal infra-red, landsat,spots, light detection and ranging (LIDAR),Figure 3: Example of data accessible through our catalog web service. This map shows Bouguer gravity from regionalGeoscience Australia data combined with new survey data acquired by WA Geothermal Centre of Excellence.4 http://www.pressureplot.com5 https://wagcoe.ivec.org/geonetwork66

Australian Geothermal Conference 2010• geophysics: seismic, velocities, magnetics,gravity, electomagnetics, radiometrics,geophysical logs, stress measurements,seismology data, magnetotellurics,• geology: geological maps, palynology data,sedimentology data, thin sections, grainsizeanalysis,• hydrogeology: hydraulic heads, water levels,water flows, rainfall, evapotranspiration data,recharge data, specific yield, storativity,• geochemistry: water quality, rock geochemistry,thermochronology,• rock physics: permeability, porosity, thermalconductivities, heat flux.While all mechanisms have been implemented forus to manage these data types, populating thecatalog with data references is a work in progress;not all sub-categories have data available yet.WAGCOE’s catalog management toolsWe have developed a suite of administrative toolsin order to build and manage our system asshown in Figure . We can therefore easily enforceconsistency between the metadata onGeoNetwork and the corresponding data on thepetabyte store. The user’s group management inboth places is also greatly simplified, although notcompletely automatic. The main feature of ouradministrative toolbox is a converter betweenGeoNetworks’ metadata XML file format and anin-house extension of it allowing us to handlespecific metadata information which is notsupported in GeoNetwork by default and is notpart of our customised version yet.Figure presents a workflow diagram of allconversions occurring, between our WAGCOEmetadata XML file format and GeoNetwork’soriginal one, through an internal Python structurewhich allows easy data handling and processing,including checking validity, updating time stampsor creating unique identifiers. We can then storeour extended metadata along with the originaldata and still retain the ability to efficiently evolveFigure 4: workflow diagram describing WAGCOE’s metadatamanagementour GeoNetwork implementation through officialupdates or in-house customisations. All metadatainformation is stored in the most appropriateformat for the particular data type, even thoughthe additional information is displayed currentlyonly as part of a “supplemental information”paragraph for each entry. Any future enhanceddisplay handling in GeoNetwork can be easilyaccommodated with a script updating allGeoNetwork entries from our original extendedmetadata files.Future workWAGCOE’s geothermal catalog is currently underdevelopment and even though nearly all majortools have been developed, some major work stillremains to populate the catalog. The last majorupgrade to our administrative tools remains aGraphical User Interface (GUI) to enter metadatamore easily, as shown on Figure , instead ofdirectly editing metadata files or internal Pythonstructures. There are also several featuresplanned for the future which will come fromWAGCOE’s strong collaboration with theAuScope Grid project (Woodcock 2008). The firstmajor enhancement will be to connect our catalogto other web services, starting with CSIRO’sPressurePlot web service. This linkage willprovide access to a real pressure andtemperature database allowing full contentsearches (whereas current searches on ourcatalog are limited to metadata). We are alsoplanning to register our data collection to theAustralian National Data Service (ANDS) RegisterMy Data service 6 which will then automaticallyharvest all metadata and provide greaterexposure to our public entries. Finally, we aremonitoring closely the AuScope Grid portaldevelopment as it could provide an even moreuser friendly interface to our catalog. Van Oort(2009) pointed the cost of building such a portal,which only emphasises the importance ofAuScope Grid’s generic and open source toolsdevelopment.ConclusionIn the last few years, web services such asGeoNetwork have redefined the way users candiscover, search, access and use geo-referenceddata. This powerful tool is currently seeing amajor uptake as it allows a dramatic change ofpractise for users who can now spend their timeusing data rather than searching for it.WAGCOE’s GeoNetwork catalog brings thisenormous advantage to the geothermalexploration industry and provides an easy,remote, and secure access to various geothermaldata sources through an intuitive web interface. Ithelps to reduce the risk for geothermal explorationin the Perth Basin and could also help some6 http://ands.org.au/services/register-my-data.html67

Australian Geothermal Conference 2010companies to store and access their data. We arehoping to connect it as soon as possible to otherweb services from various geological surveys,companies, research and governmentalorganisations, thanks to some of the technologydeveloped by the AuScope Grid projectReferencesvan Oort, P.; Kuyper, M.; Bregt, A. & Crompvoets,J. (2009), Geoportals: An Internet MarketingPerspective. Data Science Journal, 8, 162-181.Rezabakhsh, B., Bornemann, D., Hansen, U., &Schrader, U. (2006), Consumer Power: AComparison of the Old Economy and the InternetEconomy. Journal of Consumer Policy V29, 3-36.Ticheler, J. & Hielkema, J. U. (2007), GeoNetworkopensource, OSGeo Journal vol. 2, sep 2007.Woodcock, R., (2008), AuScope: Australian EarthScience Research Information Infrastructure,Geological Society of Australia and the AustralianInstitute of Geoscientists. Australian EarthSceinces Convention (AESC) 2008. NewGeneration Advances in Geoscience. AbstractsNo 89 of the 19 th Australian GeologicalConverntion, Perth Convention and ExhibitionCentre, Perth WA, Australia. July 20-24, 276pages.68

Australian Geothermal Conference 2010Flow and heat modelling of a Hot Sedimentary Aquifer (HSA) fordirect-use geothermal heat production in the Perth urban area,Western Australia (WA)Martin Pujol 1* , Grant Bolton 1 and Fabrice Golfier 2 .1 Rockwater Pty Ltd. 76 Jersey Street, Jolimont WA2 Laboratoire Environnement Géomécanique et Ouvrages. Nancy Universités, Rue du doyen Marcel Roubault,Vandoeuvre-lès-Nancy BP40, 54501FRANCE* Corresponding author: mpujol@rockwater.com.auThe deep confined aquifers of the Perth basinhave been explored for water supply and heatproduction since the beginning of the 20thcentury. The availability of warm water was anadditional asset and was very popular withlaundries and bathing services. Currently, sixgeothermal bores use warm water from theYarragadee confined aquifer for heatingswimming pools and buildings and several newbores are proposed to be drilled. The cooledformation groundwater is injected into the sameaquifer for environmental reasons.Following the recent success of the release ofgeothermal acreage for geothermal exploration inWA, HSA direct-use is getting more recognitionand support, and the number of projects is likelyto increase because the technology is now moreadvanced, low-risk, and has relatively low C0 2emissions and at low cost. Typical savings after20 years of production can be as high as$6,000,000, with payback in about 5 years andC0 2 savings of up to 800 tonnes/annum.A key concern for HSA future development inPerth is sustainability and management. This willbe achieved through a more comprehensiveassessment of HSA geothermal resources. Thelatter will also require a good estimation of thelongevity of existing and future direct-use HSAsystems.In this paper we discuss how simple numericalmodelling of temperature and fluid flow usingSEAWAT software and the “equivalent solute”approach allows for more accurate evaluations ofgeothermal resources and sustainability. This isan interesting management tool for existing andfuture HSA projects in Perth and is an alternativeapproach compared to the classical interpolationof measured temperatures between bores. Aboveall, the scope of this paper is to raise questionsand discuss the future management of existinggeothermal bores in the Perth area and to assistgeothermal explorers to develop shallow HSAdirect-use projects in the Perth area.Keywords: Western Australia, Perth, HotSedimentary Aquifer, HSA, geothermal, injection,direct-use, heat transport modelling, SEAWAT.HSA direct-use in the central PerthbasinGeological settingThe Perth urban area is located on the SwanCoastal Plain and is underlain by the Perthsedimentary Basin. The basin is comprised of aseries of sub-basins, troughs, shelves and ridgescontaining predominantly Early Permian to LateCretaceous sedimentary sequences that are up to15 km thick.The central Perth basin comprises a thicksequence of sedimentary rock, of which the upper3,000 m of Quaternary to Jurassic age arerelevant for geothermal projects targeting lowtemperature HSA systems.Existing geothermal bores have been drilled todepths up to 1,000 m in Perth; which havetargeted aquifers within the YarragadeeFormation and/or the overlying Gage Formation.The Yarragadee Formation consists of laterallydiscontinuous interbedded sandstones, siltstonesand shales and is inferred to be about 1,350 to1,500 m thick in the Perth urban area.HydrogeologyThe interbedded sandstones of the YarragadeeFormation form the Yarragadee confined aquiferand are hydraulically connected with interbeddedsandstone aquifers of the Gage Formation.The hydraulic properties of the Yarragadeeaquifer vary with location. In the study area, thediscontinuous nature of the sandstone beds haslowered the average horizontal hydraulicconductivity to about 3 m/day. Hydraulicconductivity can locally be higher as indicated bypumping tests.The average rate of groundwater flow through theYarragadee aquifer is about 0.9 m/year,confirming the very slow rate of flow indicated bythe 14 C dating of the groundwater (Thorpe andDavidson, 1991). It is likely that most of thegroundwater flow occurs in the top part of theaquifer, in about the top 500 m. Beneath thisdepth, the groundwater flow is likely to be loweras indicated by higher salinities (Davidson andYu, 2005).69

Australian Geothermal Conference 2010HSA Resource Assessment: Temperature,Heat in Place, Bore Deliverability andRecoverable HeatUsing a typical “stored heat” method as applied byBeardsmore et al (2009), the HSA geothermalresource for a specific direct-use application canbe estimated. However, the authors of this paperemphasize that the Stored Heat calculated belowis given as an element of comparison and is not tobe used for other purposes.For the purpose of reservoir volume estimations,we assume a minimum cut-off temperature of40°C for direct-use projects, and an injectiontemperature of 30°C. The base of the Yarragadeeaquifer is inferred to be at about 70.6°C. Thisyields a reservoir volume of 126 km 3 below the100km 2 modelled area where most of thegeothermal bores operate. The average reservoirtemperature is 55.3°C. Using calculated values ofheat capacity for the fluid and solid, an averageporosity of 0.15, we estimate 7,926 PetaJoules(PJ=10 15 Joules) of heat in the modelledreservoir. Should other uses be considered, thecut-off temperature, injection temperature andinferred resource may differ.The deliverable thermal energy for a typical directusegeothermal bore is a function of sourcetemperature, heat exchanger efficiency, flow-rateand injection temperature. Typical values of 25L/s and 12°C temperature drop can provide adeliverable thermal power of 1,113 kilowatt (kW th )and recoverable heat energy of 0.04 PJ/annum or9747 Megawatts hours/annum (MWh th ).Hence, in first estimation, the Yarragadee aquiferappears to be able to sustain a generalised use ofHSA for heating buildings and swimming poolsamong other uses in the Perth urban area formany years. However, this may only be possible ifthere is no (or very little) thermal contaminationbetween the geothermal and injection bores. Thiscan be predicted using a numerical model ofgroundwater flow and heat transport. This is thegeneral purpose of this study.HSA modelling: Initial temperaturedistribution in the aquifer (Conductivemodel)3D structural model of a selected area of thePerth basinData from Davidson and Yu (2005), andproprietary data from Rockwater Pty Ltd havebeen used to create a detailed structural model ofan area of about 100 km 2 of the central PerthBasin where most of the geothermal boresoperate. The model comprises seven geologicalformations: superficial sediments (TQ), Tertiaryalluvial (Tk), Cretaceous sediments (Kco, Kwl,Kws and Kwg) and Jurassic sediments (Jy).Geostatistical methods have been used to inferformation top and bottom surfaces. Additionally,formation coverages specifying the extent of eachgeological formation (derived from contours inDavidson and Yu (2005) and modified in areaswhere new data had become available) havebeen used to constrain the model.Purely conductive heat flow 1D modellingA common practice in the geothermal industry,when modelling steady state temperatureconditions of HSA geothermal reservoirs, is toassume purely conductive heat transfer andconstant heat conductivity within each of thegeological layers. This is referred as conductiveheat flow 1D modelling and is considered moreaccurate than the classical approach of averagegradient as it accounts for thermal resistancevariations within the lithological column. This hasbeen demonstrated in several studies includingCooper and Beardsmore (1998).The conductive heat flow 1D assumption failswhen heat convection occurs such as in areas ofhigh groundwater velocity such as faults, bores orwhen significant heterogeneity occurs. However,purely conductive models have proven tosatisfactorily represent temperature conditions inthe Perth Basin and will be used to provide initialtemperatures for a more comprehensiveconductive and convective numerical model.In a conductive heat regime the temperature (T)at the bottom of a geological layer, is equal to thetemperature at the top of the layer (T 0 ) plus theproduct of Heat Flow (Q) and thermal resistanceof the geological layer (R) (where the thermalresistance equals the thickness of the layerdivided by the average thermal conductivity).Consequently, the occurrence of prospectivegeothermally warmed groundwater in sedimentaryaquifers results from sufficiently low conductivity(high thermal resistance) of the sedimentary covercombined with high flow of heat from the centre ofthe Earth. Heat Flow (Q) is a function of the heatgenerated within the crust by the decay ofradiogenic minerals plus heat conducted from themantle. A commonly accepted value, derived fromnearby temperature logs for the modelled area is74 mW/m 2 .Conceptual modelIn order to represent the temperature distributionin the basin, several physical properties andboundary conditions have to be estimated and aso called conceptual model must be constructed.For each modelled stratigraphic layer, it isassumed that the lithology and physical propertiesare the same throughout. Measured thermalconductivities are available for the formations ofthe Perth basin (Chopra and Holgate, 2008) andhave been modified for the purpose of thenumerical model by a classical trial and error70

method during the calibration of the model. Formost of the geological formations, the modelledvalue is close to the calculated value. However,modelled and measured heat conductivity valueswere found to differ for some geologicalformations. It is believed that it is due tolithological variations within those formations andvariation of the physical properties with depth.Table 1: Heat conductivity values (W/mC) of geologicalformations in the Perth BasinGeological Formation Modelled Measuredat 30°CTQ 1.5 1.42Tk 2.20 No dataKco 2.30 to 2.50 No dataKwl 1.70 to 2.50 2.56Kws 1.50 1.71 to 1.72Kwg 2.55 to 2.60 1.71 to 2.20Jy 3.05 to 3.20 2.30 to 4.31Method and boundary conditionsThe thickness of each geological formation atgiven coordinates have been extracted and usedtogether with heat conductivity properties forthese layers, to calculate thermal resistances.Assuming a constant surface temperature of19.5°C at the upper boundary (taken as real meanair value measured at the Perth airport plus 1 Cto account for thermal insulation of rocks) andusing the thermal resistance values calculatedabove and the assumed constant heat flow of 74mW/m 2 , it was possible to predict the temperaturedistribution within the Yarragadee aquifer.Fig. 1: simplified 3D structural model of the modelled areawith geological formations considered in this study andapproximate locations of geothermal production boresCS: Challenge Stadium; TOC:Town Of Claremont;CCGS: ChristChurch Grammar School;SHG: St Hilda Geothermal, BPC: Bicton Polo Club.Australian Geothermal Conference 2010HSA modelling: Initial temperaturedistribution in the aquifer (uncoupledConductive and Convective model)3D structural model of a selected area of thePerth basinThe same data as before are used but thestructural model is limited to the Yarragadeeaquifer (Fig. 1).All layers overlying the Gage Formation are notmodelled, apart from the portion of Kings ParkFormation that has eroded the Gage andYarragadee Formations and which is likely toinfluence groundwater flow and heat transport.Heat transport modelling using the“equivalent solute” approach and thenumerical code SEAWATSEAWAT is a standard finite-difference solutecode included in the state-of-the-art modellingsoftware Visual Modflow Pro.Due to the similarities between heat and solutetransport, standard solute codes such asSEAWAT can be used to represent heat transportand variables for SEAWAT solute transportsimulator can be reinterpreted for heat transport.This has been demonstrated in several studiesincluding Langevin et al (2008). More detailedinformation on using solute transport simulationfor heat transport modelling can be found inHecht-Mendez et al (2009). Additional informationspecific to the use of the SEAWAT code for heattransport is available in Ma and Zheng (2010).In addition, for this study, SEAWAT has beenevaluated against analytical results developed forgeothermal and injection bores (doublet) and theresults were comparable. The analytical solutionhas been developed by Gringarten and Sauty(1976) to predict the temperature evolution atHSA geothermal production bores used forgeothermal urban heating in the Paris basin.Conceptual modelAs groundwater flow needs to be consideredwhen undertaking conductive and convectivegeothermal modelling, a numerical code had to beused. For the present work, SEAWAT is used.The previously created 3D structural model isimported into SEAWAT and extrapolated to afinite difference 3D grid where the flow and heattransport equations are solved.Horizontal cell size varies from 140 m to 5 m nearthe bores and is about 25 m in vertical. Attentionhas been given to keep aspect ratio (ratiobetween cell size along x and z and y and zrespectively) less than 6.The calibrated thermal conductivity for KingsPark, Gage and Yarragadee Formations areassigned to the corresponding cells and setconstant for each formation.71

Australian Geothermal Conference 2010Hydrogeological parameters are available for theformations of the Perth basin (Davidson and Yu(2005) and proprietary data from Rockwater PtyLtd) and have been modified for the purpose ofthe numerical model by a classical trial and errormethod during the calibration of the model.Table 2: Modelled hydrogeological parameters of geologicalformations in the Perth BasinGeological Kh Kh/Kv SFm. (m/day) (-)(-)Tk

Australian Geothermal Conference 2010borehead has been recorded since the bore wascommissioned. Considering the aboveassumptions, run 2 (Fig. 3) shows an acceptableagreement with pumped temperature increasingby 0.1°C after 10 years.Note that a more accurate evolution oftemperature at the bore could be obtained using alocal model with a higher spatial resolution and anexplicit numerical solution for advection.Fig. 3: CCGS1 modelled pumped temperaturesIn addition, steady state modelled drawdown is10.3 m and agrees with the aquifer losses of 8 mmeasured at the end of the 48 hours constant-ratepumping test.Results: environmental impact of the injectionFollowing the calibration, the model was run fromJanuary 2000 to January 2050 to predict theevolution of temperatures within the basin.The model aims are (i) to determine the generalevolution of temperature of the Yarragadeeaquifer, (ii) to give a first estimation of the lifetimeof existing geothermal installations and (iii) toidentify areas where there is an impact of injectedwater on aquifer temperatures.almost 50 years of continuous operation as shownin Fig. 4.The modelled temperature decline at CCGS1 inJanuary 2050 is 0.39°C after 48 years, and0.12°C at TOC 1 after 45 years whereas thetemperature at SHG1 has increased by 0.14°Cafter 40 years. This is likely to have little impacton the geothermal installation efficiencies andsubsequently the lifetime of all three geothermalinstallations; estimated to be more than 40 years.DiscussionFor all three bores, the temperature evolution canbe described as follows:• Stage 1: Increased temperature of thepumped water (increase is higher when flow-rateis higher) provoked by the inflow of deeper andwarmer groundwater in the bores. This is furtherfacilitated by the presence of upward heads.• Stage 2: As the cooled groundwater isinjected and travels through the aquifer in thedirection of the production bore, the rate oftemperature increase diminishes and eventuallystabilises (this happens earlier when the verticaldistance between injection and productionscreened section is small).• Stage 3: Pumped temperatures startdeclining and eventually decline at a linear rate. Itis calculated that bores CCGS1 and TOC1 willreach Stage 3 in year 2050 because of thesmaller vertical separation between productionand injection screened sections, whereas SHG1is likely to still be in Stage 2.Fig. 4: Predicted evolution of temperature at CCGS1, TOC1and SHG1The initial observation is that there is littletemperature decline at the production bores afterFig. 5: Calculated temperature distribution (°C) in selectedlayers: layer 6 from -520 to -676 m AHD and layer 9 from -651 to -767m AHDThe modelled results show that the cooler injectedwater has a limited impact on the pumpedgroundwater temperature because of themoderately low vertical hydraulic conductivities,upward heads in the deeper geothermal bores,and the vertical distances between production andinjection bore screens.73

Australian Geothermal Conference 2010The cooled groundwater plume is calculated toextend 850 m in a circular pattern from TOC1(Claremont AC) by 2050 (Fig. 5) indicating thatnatural groundwater flow has little influence on theshape of the groundwater plume. Conversely, thegroundwater plume generated by CS1 (ChallengeStadium) has a very distinctive tear-drop pattern(Fig. 5, layer 9) indicating that a portion of theinjected water is recirculated and that thermalcontamination is occurring.Conclusion: HSA direct-usesustainability and future researchobjectivesAlthough the presented simulation is decoupled(water density is independent of temperature) andmay not be accurate where density effectsdominate, the results show that pumpedgroundwater temperatures are unlikely to changesignificantly over the next 40 years. This supportsthe notion that HSA direct-use is a cost-effectivesolution (payback is about 5 years) for heatingbuildings or swimming pools for example.However, temperature depletion seems to extendhorizontally and although no visible interferencebetween bores has been recorded, it is advisableto avoid pumping from the same depth as water isinjected. Therefore, it is recommended that futureproduction bores should be sited at least 500 mfrom one another and at different depths (i.e. 100to 150 m deeper than the nearest injectionscreens).To increase the accuracy of the model and to beable to guarantee the sustainability of HSA directuseprojects, additional work could be performed:• Create local, high resolution models foreach geothermal installation.• Perform temperature logging periodically.• Model heterogeneity patterns of theYarragadee aquifer.• Refine the structural model byconsidering geological members within theOsborne and Leederville formations.• Refine the calibration of the model usinga transient constant-rate pumping test.• Perform hydraulic head versus depthmeasurements.• Correlate the stratigraphy (siltstone andsandstone beds) between production andinjection bores to increase vertical accuracy.• Monitor geothermal installationsperiodically to obtain monthly data of injectiontemperatures, pumped water temperatures, flowratesand injection pressures.• Evaluate the impact of density forces(forces driving the formation of convection cells)on initial temperature distribution.• Consider heat flow variations over themodelled domain.• Refine the calibration using recenttemperature logs of artesian monitoring bores.ReferencesBeardsmore, G. R., et al, 2009, Statement ofestimated Geothermal Resources MetropolitanPerth Geothermal Play, 31/07-8 GEP, 27 July2009.Chopra, P.N, and F., Holgate., 2008, Geothermalenergy potential in selected areas of WesternAustralia; a consultancy report by Hot Dry RocksPty Ltd for Geological Survey of WesternAustralia: Geological Survey of Western Australia,G31888 A2.Cooper, G. T., and Beardsmore, G. R.,Geothermal systems assessment: understandingrisks in geothermal exploration in Australia: PESAEastern Australasian Basins Symposium IIISydney, 14–17 September, 2008.Davidson, W.A., and Yu, X., 2005. Perth regionalaquifer modelling system (PRAMS) modeldevelopment: Hydrogeology and groundwatermodelling : Western Australia Department ofWater, Hydrogeological record series HG 20.Gringarten, A.C., and Sauty, J.P., 1976, Theeffect of reinjection on the temperature of ageothermal reservoir used for urban heating:BRGM, Orleans. 12p.Hecht-Mendez, J. et al., 2009, EvaluatingMT3DMS for Heat Transport Simulation of ClosedGeothermal Systems: Wiley-Blackwell. 16p.Langevin, C.D., Thorne, D.T. Jr., Dausman, A.M.,Sukop, M.C, and Guo, Weixing, 2008, SEAWATVersion 4: A Computer Program for Simulation ofMulti-Species Solute and Head Transport: U.S.Geological Survey Techniques and Methods Book6, Chapter A22, 39 p.Loague, K., and R.E., Green, 1991, Statistical andgraphical methods for evaluating solute transportmodels: Overview and application: Journal ofContaminant Hydrology 7, no. 1–2: 51–73.Ma, R., and Zheng, C., 2010, Effects of densityand viscosity in modelling heat as a groundwatertracer: Groundwater, Volume 48 Issue3, pp. 380-389.Thorpe, P.M., and Davidson, W.A., 1991,Groundwater age and hydrodynamics of theconfined aquifers, Perth, Western Australia:Proceedings of the International Conference onGroundwater in Large Sedimentary Basins, Perth,Western Australia, 1990. Australian WaterResources Council Series, no. 20, p. 420-436.74

In most Precambrian terrains including India,moderate-to-low temperature hot spring systemsrepresent the potential conventional geothermalenergy resources. This scenario is in contrast togeothermal fields under production in other partsof the world, which are located in Quaternaryvolcanic / magmatic settings (Gupta and Roy,2006). This paper outlines (i) the major hot springoccurrences in India, (ii) the exploration effortsundertaken so far, (iii) possible geothermalmodels in the light of regional heat flow and heatproduction datasets, (iv) a few critical informationgaps that need to be covered to make realisticassessment of the geothermal energy potential,and (v) perspectives for development andutilisation of geothermal energy in the country,both for electric power generation as well as fordirect uses.Geothermal Exploration in IndiaSukanta RoyNational Geophysical Research Institute(Council of Scientific and Industrial Research)Uppal Road, Hyderabad 500 606, IndiaCorresponding author: sukantaroy@yahoo.comAustralian Geothermal Conference 2010Keywords: India, geothermal energy, heat flow,radiogenic heat productionExploration of Geothermal EnergyResourcesThe major groups of hot springs in India occur inManikaran, Puga-Chhumathang valley andTapoban in the Himalaya, a near N-S trendinglinear belt in the west coast of Maharashtra, theSon-Narmada-Tapti lineament zone in centralIndia, Tattapani in Chattisgarh, and Rajgir-Monghyr, Surajkund and Bakreshwar in easternIndia. The distribution of major groups of hotsprings is shown in Figure 1. Rao (1997) providesa brief summary of the historical development ofstudies on the Indian geothermal resources,starting with the first compilations of 99 hotsprings in India and the adjacent countries bySchlagintweit (1865). The locations, geologicalsettings and temperatures of hot springs locatedin different geologic provinces in India have beencompiled by previous workers (for example,Oldham, 1882; Ghosh, 1954; Gupta, 1974; Guha,1986; GSI, 1991, 2002 and others). A number ofthose geothermal springs have been used forbalneological purposes. However, India is yet toproduce electric power from a geothermal field.Among the most notable achievements during thepast five decades have been the assessment ofgeothermal fields by the Government of India in1966 and publication of a comprehensive report in1968 recommending preliminary prospecting ofFigure 1. Outline of India showing the distribution of majorgroups of warm and hot springs (modified after Krishnaswamy,1975). Temperatures of the hot spring waters are indicatedusing symbols (see legend). Shaded regions (exaggeratedscale) show the major clusters of hot springs.the Puga and Manikaran geothermal fields in theHimalaya (Hot Springs Committee, 1968). Amajor, systematic, multi-disciplinary, multi-Institutional programme (including drilling up to385 m) covering the Puga-Chumathang field inLadakh was mounted during 1972-74 under thestewardship of V.S. Krishnaswamy of theGeological Survey of India (GSI). The subsurfacefeatures were delineated in considerable detailand the results were presented at the SecondUnited Nations Symposium on the Developmentand Use of Geothermal Resources in 1975. Themost significant outcome of the effort was aproposal to set up a 1 MWe binary-cycle powerplant on a pilot scale basis, which has not beenimplemented so far. Attempts to revisit thegeothermal exploration in the area include anumber of geochemical studies (GSI, 1996) andrecent magnetotelluric studies (Abdul Azeez andHarinarayana, 2007). An expert group set up in2008 by the Ministry of New and RenewableEnergy, Government of India made strongrecommendations to install pilot-scale plant by75

Australian Geothermal Conference 2010drilling exploration cum demonstration wells in thearea (MNRE, 2008). This would be useful not onlyfor monitoring the hot water discharge andtemperatures over a period of time but alsostudying the shallow reservoir characteristics.Another major initiative, directed towards the hotsprings of the West Coast belt and the Son-Narmada-Tapti belt was taken up by the GSI withUNDP assistance during 1976-77, and wasextended for a few years on its own, includingdeep drilling up to depth of 500 m. The resultswere published in the Records of the GSI (1987).The Tattapani hot springs of Chattisgarh districtwas identified for trials with regard to powerproduction using a 300 KW e binary-cycle powerplant. Summaries of various explorationprogrammes undertaken in the country during theperiod 1970-1990 are given in Gupta et al. (1973,1974); Shanker et al. (1976); Singh et al. (1983);Krishnaswamy and Ravi Shanker (1980); GSI(1983, 1991, 1996, 2002); Thussu et al. (1987);Moon and Dharam (1988) and Gupta (1992). Thesalient results in the case of the two potentialgeothermal fields, Puga and Tattapani, aresummarized below.Puga Valley Hot SpringsIn the case of Puga valley springs, geothermalevidence gathered so far and geochemicalindicators including the occurrence of cesiumdeposits around the springs have suggested thepossibility of high temperature hydrothermalcirculation in the subsurface (Absar et al., 1996).A shallow reservoir in the top few hundred meterswas inferred from geophysical and geothermalinvestigations carried out during the 1970s.Recent magnetotelluric studies indicate thepresence of an anomalous conductive feature (~5Ohm m) below a depth of ~2 km in the area of thethermal manifestations (Harinarayana et al., 2006;Abdul Azeez and Harinarayana, 2007). Althoughbroad correlations between high electricalconductivity and high temperatures have beenobserved in some geothermal areas, thecalibration of electrical conductivity anomalies totemperature anomalies at depth is notestablished. Further efforts are necessary to fillexisting gaps in knowledge and verify thehypothesis regarding the occurrence ofsubsurface magma chambers or young intrusivegranites in the region. Such heat sources alonecan sustain power generation on a reasonablescale (Rao et al., 2003). Determination ofbackground heat flow outside the hot springszone, helium-isotopic measurements on thethermal waters, detailed petrographic andgeochronological studies on young graniteintrusives, and tritium dating of the waters wouldbe useful for testing a magmatic heat source.Tattapani Hot SpringsIn the case of the Tattapani hot springs, thetemperatures of the issuing waters is ~110 o C, thehighest recorded so far in the shield. The springwaters are meteoric in origin as indicated byoxygen and helium isotopic data, and the age ofthese waters indicated by tritium dating is ~40years (Thussu et al., 1987; Sharma et al., 1996;Minnisale et al., 2000). Repeat well testing carriedout by GSI in 1995 and 1999 indicate nosignificant fall in temperatures and pressuresduring the intervening period (Sarolkar andSharma, 2002). Recent magnetotelluric studies inthe area have delineated a deep, anomalousconductive zone possibly indicating thesubsurface extent of the reservoir (Harinarayanaet al., 2004). Lack of evidence for Quaternarymagmatism in the region as well as the meteoricnature of the hot spring waters indicate that thehot springs are controlled by forced convectiondue to peizometric gradient existing between therecharge area and the hot springs, and that thesprings could be simply “mining” the normal heatflow (Roy and Rao, 1996; Rao et al., 2003).However, no heat flow measurements outside thelocalized hot springs zone have been made.There is therefore a clear need to establish theregional thermal conditions by systematicgeothermal measurements because they wouldcontain information about subsurface flow andlocation of recharge area also (Lachenbruch etal., 1976).Other areasA large number of warm to hot springs occur inwestern, central and eastern parts of the Indianshield. The temperatures of these springs arelower than those at Tattapani and vary between35 o and 80 o C. It is likely that a geothermal modelsimilar to that at Tattapani could explain theoccurrence of these hot springs.Heat Flow and Heat ProductionHeat flow determinations made through precisetemperature measurements in boreholes andthermal conductivity measurements onrepresentative rock formations have been thecentral theme of the heat flow studies programmeat the National Geophysical Research Institutesince its inception in 1961. Although boreholes ofopportunity have been used for the majority ofmeasurements, a selection has been made on thefollowing criteria to make the heat flowdeterminations useful for characterizion of thethermal state of the lithosphere: (1) depths greaterthan 150 m in hard rock areas and severalhundred metres in sedimentary basins (2) sitesaway from hot spring manifestations, tectonicallyactive regions, and rugged topography (3)temperature profiles with no characteristicperturbations such as those of heat refraction andgroundwater movement. The principal constraintin the data acquisition programme has been theavailability of suitable boreholes for making76

Australian Geothermal Conference 2010temperature measurements. This constraint hasbeen addressed, to some extent, through thedrilling of 14 dedicated “heat flow” boreholes todepths of up to 500 m at carefully chosen sites insouth India. The careful siting of the boreholes inareas of relative lithological homogeneity and lowgroundwater yields facilitated the acquisition ofundisturbed temperature-depth profiles fromwhich gradients could be estimated with a greatdegree of confidence. In geologic provinces ofnorthern Indian shield, the heat flow coverage isvariable and several gaps exist over largesegments. In the Himalaya, regional heat flowdata are not available. Borehole temperaturemeasurements in close proximity of hot springsexist at a few locations only. However, thesemeasurements are affected by convection due tohydrothermal circulation in the near surface zoneand are not representative of deep crustalconditions. Thermal conductivity data for all majorrock formations in the southern Indian shield andseveral rock formations from other parts of Indiaconstitute a very extensive database for modellingthe thermal structure in the upper few kilometresof the crust.Heat Flow and Geothermal GradientsFigure 2. Heat flow map of the Indian Shield. Heat flow sitesare shown by filled triangles. Heat flow determinations havebeen made on the basis of temperature measurements inboreholes and thermal conductivity of rock formations. Overlarge regions in northern India which are not covered with heatflow measurements, the contours should be treated withcaution. [Sources of data: Roy and Rao (2000) and originalreferences therein; Ray et al. (2003); Rao et al. (2003); Roy(2008); Roy et al. (2008)].A heat flow map for the Indian shield is shown inFigure 2. Heat flow data are now available frommeasurements made in ~210 boreholesdistributed over the major geological provinces inthe shield. The dataset shown here is confined tosites where complete information on geothermalgradient and thermal conductivity are available.Heat flow determined using other techniques suchas chemical composition of water, andcorrelations with P-wave velocity and age, areexcluded. A detailed discussion of heat flowcharacteristics of different provinces is given inthe papers by Roy and Rao (2000) and Rao et al.(2003). The salient features of the heat flowspectrum in the Indian shield are brieflymentioned below.1. The southern Indian shield comprising theArchaean Dharwar greenstone-granitegneissprovince and gneiss-granuliteprovince, is characterized by low heat flow,generally ranging from 25 to 50 mW m -2 .Geothermal gradients measured in the topfew hundred meters in boreholes and up to2150 m in a deep mine are in the range 12--115 mK m .2. The low heat flow regime of south Indiaextends northward beneath the DeccanTraps in central India, which indicate theabsence of thermal transients related to the~65 Ma Deccan volcanism in the present-dayheat flow. Measurements at several localitiesdistributed in Deccan Traps province indicatean average gradient of ~25 mK m -1 .However, beneath the Traps which rangefrom a few meters in the east to 2-3 km in thewest, the gradient in the Precambrian graniticbasement drops to ~15 mK m -1 due to higherthermal conductivity of granitic rocks relativeto basalts.3. The Precambrian provinces in northernIndian shield show a contrasting heat flowregime, with values ranging from 50 to 96mW m -2 . Previous studies have attributed theenhanced heat flow to high levels ofradiogenic heat production in the upper crust.Temperature gradients vary between 12 and-130 mK m across different rock formations.4. The Gondwana sedimentary basins (UpperCarboniferous to Lower Cretaceous) have agenerally high but variable heat flow in therange 46 to 107 mW m -2 . Temperaturegradients show large variability, and reachpeak values of up to ~50 o Ckm -1 in theo -1Damodar Valley basins and ~40 Ckm inGodavari Valley basin.5. The Cambay sedimentary basin of Tertiaryage in western India shows consistent highheat flow, 75-96 mW m -2 in the northern partsand a lower heat flow, 55-67 mW m -2 , in thesouthern parts. Very high gradients in therange 37 to 56 o C km -1 were computed fromtemperature measurements to depths >1000m in the northern part.77

Australian Geothermal Conference 2010Radiogenic Heat ProductionHeat production due to radioactive decay of longlivedisotopes of U, Th and K (namely, 235 U, 238 U,232 Th, and 40 K) in rock formations constituting thecontinental crust plays a key role in interpretationof heat flow data. With this primary objective, alow-level counting gamma-ray spectrometricfacility using a single-channel analyser wasestablished at the NGRI for analysis of rocks forradioelements U and Th (at ppm levels) and K(Rao, 1974; Rao and Rao, 1979). The facility hasbeen progressively upgraded using a multichannelanalyser and later, a PC-based multichannelanalysing card that provides spectrumstabilization (Roy and Rao, 1999, 2003). Severalhundred samples covering major rock types in theIndian shield have been analysed for U, Th and K.The salient features of the dataset are describedin Roy (2008).Over the last two decades, laboratory analyseshave been complemented by in-situ gamma-rayspectrometric analysis in several areas in theIndian shield (Roy, 1997; Roy and Rao, 2000,2003; Ray et al., 2003, 2008). A field-portable,four-channel, spectrum stabilized, gamma-rayspectrometer and a large crystal (6” high and 4”diameter) detector have been used. Analysis forU, Th and K are made by placing the detectordirectly over fresh rock outcrops. In this case, thedetector senses a much larger volume of rockmass, typically a circle of investigation of ~40 cmand a depth of ~12 cm, compared to thelaboratory method. In areas abounding with freshoutcrops, this method results in faster coverage.Significant observations include delineation ofpockets of very high heat production in graniticrocks, for example in Tattapani area, and younggranite intrusives in northwestern parts of India,detection of granulite facies rocks with lowest heatproduction reported in literature, and a range ofrock formations with intermediate heat productionvalues.Perspectives for Development ofGeothermal EnergyRe-assessment of Energy Potential ofConventional Geothermal ResourcesIn view of growing energy demands and theemphasis on renewable energy in India, a reassessmentof geothermal energy potential ofPuga Valley hot springs in Ladakh and Tattapanihot springs in Chatttisgarh should be carried outby covering some critical gaps in informationthrough acquisition of new data, combinedinterpretation of geothermal datasets and existinggeological, hydrological, geochemical andgeophysical datasets to throw light on the natureof the heat source of the hot springs and theirsustainability for power production, undertakingdrilling and setting up of pilot-scale binary-cyclepower plants.Geothermal resources vary widely from onelocation to another, depending on the temperatureand depth of the resource, the rock chemistry,and the abundance of groundwater. The type ofgeothermal resource determines the method of itsutilization. Variants of binary cycles appropriate tooptimum utilization of geothermal heat from hotsprings in non-volcanic settings such as those inIndia need to be developed.Direct Heat UsesThe heat extracted from warm-to-hot watersemerging from other hot spring systems in thecountry can be gainfully employed for a number ofdirect uses such as development of tourist spasfor bathing, swimming and balneology,greenhouse cultivation in cold climates, extractionof borax and rare materials such as cesium, andagricultural product processing. The significanteconomic and environmental benefits of usingmoderate-to-low enthalpy geothermal waters toreplace even small quantities of conventionalfuels for direct uses cannot be ignored today inview of the steep increase in costs of fossil fuelsand associated greenhouse gas emissions.Exploration for Enhanced GeothermalSystemsA second category of geothermal resourcetraditionally referred to as “hot dry rock” and morerecently as “enhanced geothermal systemsEGS)”, has not yet been explored in India. Theprimary requirement for such a resource is theoccurrence of high temperatures (typicallyupwards of 150 o C) at economically viable depths(typically the top 1-4 km of the Earth’s crust).Areas of anomalous high heat flow, high-heatproducinggranites and other silicic igneousintrusives having a depth extent of a fewkilometers, could be possible targets of futureexploration efforts in the country (Roy, 2008).These considerations reinforce the need forcarrying out systematic heat flow as well asradiogenic heat production investigations on acountry-wide scale.Geothermal Heat PumpsThe viability of geothermal heat pumps for heatinginside buildings should be explored in the statesof Jammu and Kashmir, Himachal Pradesh andparts of Uttarakhand which experience severewinter conditions for long periods. Space coolingrequirements in most parts of India have grownseveral fold in the recent years with the growth ineconomy. There is enormous scope fordeveloping the capabilities in geothermal coolingof buildings by modifying existing technologies tosuit Indian conditions. A proper assessment of thetechnology for application to different climaticenvironments existing in the region, and its78

Australian Geothermal Conference 2010exploitation by integrating it with building designsshould be encouraged.SummaryModerate-to-low enthalpy hot spring systemsprimarily represent the known geothermal energyresources in India. These resources aredistributed in diverse physiographic and tectonicsettings, viz., the Himalayan belt and thePrecambrian shield. Detailed geological andgeochemical exploration followed by limitedgeophysical exploration and shallow drillinginvestigations up to a few hundred meters haveresulted in first-order geothermal models for themajor hot spring zones in the country. However,development of the geothermal resources hasremained at a very low level mainly due toinadequate characterization of the deeper thermalregime leading to low confidence in proposedreservoir models and sustainability of the heatsource. There is therefore an urgent need to carryout a reassessment of the geothermal energypotential of hot springs by employing newgeophysical probing tools and computationaltechniques available today, both for electric powergeneration as well as for direct uses. Efficientexploitation technologies appropriate to nonvolcanicareas need to be developed. Systematicheat flow and heat production investigations needto be carried out for the identification of areaswhere high temperatures in the top few kilometersbelow the ground surface indicate potential for“hot sedimentary aquifers” as well as “enhancedgeothermal systems”. The vast potential forgeothermal heat pumps is yet to be tapped. Theexisting technology must be made accessible toindividuals and small communities as a low-costalternative for their space heating and coolingneeds.ReferencesAbdul Azeez, K.K., and Harinarayana, T., 2007,Magnetotelluric evidence of potential geothermalresource in Puga, Ladakh, NW Himalaya : Curr.Sci., v. 93, p. 323-329.Absar, A., Jangi, B.L., Sinha, A.K., Rai, V., andRamteke, R., 1996, Concentration of rare alkaliesin the geothermal fluids at Puga, Ladakh, J and K– Significance and possibilities of exploitation, inPitale U.L. and Padhi R.N., eds., GeothermalEnergy in India : Geological Survey of India,Special Publication 45, p. 163-167.G.S.I., 1983, Status note on geothermalexploration in India by Geological Survey of Indiaand a profile of the exploration strategy:Geological Survey of India, Kolkata.G.S.I., 1987, Records of the Geological Survey ofIndia, v. 118, Pt. 6.G.S.I., 1991, Geothermal Atlas of India:Geological Survey of India, Special Publication19, 144 p.G.S.I., 1993, Geological Map of India, scale1:5,00,000 : Geological Survey of India.G.S.I., 1996, Geothermal Energy in India, U.L.Pitale and R.N. Padhi, eds., Geological Survey ofIndia, Special Publication 45.G.S.I., 1998, Geological Map of India, 7th Edition,scale 1:2,000,000. Geological Survey of India.G.S.I., 2002, Geothermal Energy Resources ofIndia. Geological Survey of India, SpecialPublication 69, p. 12-29.Ghosh, P.K., 1954, Mineral springs of India :Records, Geological Survey of India, v.80, p.541–558.Guha, S.K., 1986, Status of geothermalexploration for geothermal resources in India :Geothermics, v. 15, p. 665–675.Gupta, H.K., and Roy, S., 2006, GeothermalEnergy: An alternative resource for the 21 stCentury : Elsevier, Amsterdam, 292 p.Gupta, M.L., 1974, Geothermal resources ofsome Himalayan hot spring areas : HimalayanGeology, v.4, p. 492–515.Gupta, M.L., 1981, Surface heat flow and igneousintrusion in the Cambay basin, India : J. Volcanol.Geotherm. Res., v.10, p.279-292.Gupta, M.L., 1992, Geothermal resources ofIndia: geotectonic associations, present statusand perspectives, in Gupta, M.L., Yamano, M.,eds., Terrestrial Heat Flow and GeothermalEnergy in Asia : Oxford and IBH, Delhi, p. 311–329.Gupta, M.L., Rao, G.V., and Hari Narain, 1974,Geothermal investigations in the Puga Valley hotspring region, Ladakh, India : GeophysicalResearch Bulletin, v.12, p.119–136.Gupta, M.L., Rao, G.V., Singh, S.B., and Rao,K.M.L., 1973, Report on thermal studies andresistivity surveys in the Puga Valley geothermalfield, Ladakh district, J&K. NGRI Technical Report73–79, National Geophysical Research Institute,Hyderabad, India.Harinarayana, T., Abdul Azeez, K.K., Murthy,D.N., Veeraswamy, K., Rao, S.P.E., Manoj, C.,and Naganjaneyulu, K., 2006, Exploration ofgeothermal structure in Puga geothermal field,Ladakh Himalayas, India by magnetotelluricstudies : J. App. Geophys., v.58, p. 280-295.Harinarayana, T., Someswara Rao, M., Sarma,M.V.C., Veeraswamy, K., Lingaiah, A., Rao,S.P.E., Virupakshi, G., Murthy, D.N. and Sarma,S.V.S., 2004, Delineation of a deep geothermalzone in Tattapani region, Surguja district, M.P.:EM-INDIA Seminar, Hyderabad, India, Abstracts.79

Australian Geothermal Conference 2010Hot Springs Committee, 1968, Report of HotSprings Committee: Government of India,unpublished.Krishnaswamy, V.S., 1975, Geothermal energyresources of India: Symposium on EnergyResources in India, Indian Science Congress,Abstracts.Krishnaswamy, V.S., Ravi Shanker, 1980, Scopeof development, exploration and preliminaryassessment of the geothermal resource potentialsof India : Records, Geological Survey of India,111, Pt. II, p. 17–40.Lachenbruch, A.H., Sorey, M.L., Lewis, R.E., andSass, J.H., 1976, The near-surface hydrothermalregime of Long Valley Caldera : J. Geophys. Res.,v.81, p. 763-768.Minnisale A., Vaselli O., Chandrasekharam D.,Magro G., Tassi F., and Casiglia, A., 2000, Originand evolution of intracratonic thermal fluids fromcentral-western peninsular India : Earth Plan. Sci.Lett., v.181, p. 377-394.Moon, B.R. and Dharam, P., 1988, Geothermalenergy in India – Present status and futureprospects : Geothermics, v.17, 2/3, p. 439-449.MNRE, 2008, Report of Expert Group on PowerGeneration from Geothermal Energy at Puga,Jammu and Kashmir, India : Ministry of New andRenewable Energy, Govt. of India.Oldham, R.D., 1882, The Thermal Springs ofIndia : Geological Survey of India, Memoir, vol.19.Rao R.U.M., Roy, S., and Srinivasan, R., 2003,Heat Flow Researches in India – Results andPerspectives : Geol. Soc. India Memoir, v. 53, p.347-391.Rao, R.U.M., 1974, Gamma-ray spectrometricset-up at NGRI for analysis of U, Th and K inrocks : Geophys. Res. Bull., v.12, p.91-101.Rao, R.U.M., and Rao, G.V., 1979, Radiogeologicstudies at NGRI – A Review : Geophys. Res.Bull., v. 17, p. 163-173.Rao, R.U.M., 1997, Book review – Geothermalenergy in India, Pitale, U.L. and Padhi, R.N., eds.: J. Geol. Soc. India, v.49, p.746–748Roy, S., 2008, Heat Flow Studies in India duringthe past five decades: Geological Society of IndiaMemoir 68, p. 89-122. .Roy, S. and Rao, R.U.M., 1996, Regional heatflow and the perspective for the origin of hotsprings in the Indian shield, in Pitale, U.L., Padhi,R.N., eds., Geothermal Energy in India:Geological Survey of India, Special Publication45, p. 39–40.Roy, S. and Rao, R.U.M., 2000, Heat flow in theIndian shield: J. Geophys. Res., v.105, p.25,587-25,604.Roy, S. and Rao, R.U.M., 2003, Towards a crustalthermal model for the Archaean Dharwar Craton,southern India: Physics and Chemistry of theEarth, v.28, p.361-373.Roy, S., Ray L., Kumar, P.S., Reddy, G.K., andSrinivasan, R., 2003, Heat Flow and HeatProduction in the Precambrian Gneiss-GranuliteProvince of Southern India, in M. Ramakrishnan,ed., Tectonics of Southern Granulite Terrain –Kuppam-Palani Geotransect: Geological Societyof India, Memoir 50, p.177-191.Roy, S., Ray, L., Bhattacharya, and Srinivasan,R., 2008, Heat flow and crustal thermal structurein the Late Archaean Closepet Granite batholith,south India : Int. J. Earth Sci., v.97, doi:10.1007/s00531-007-0239-2, p.245-256.Sarolkar, P.B., and Sharma, S.K., 2002,Geothermal well testing at Tattapani geothermalfield, district Surguja, India: 27th Workshop onGeothermal Reservoir Engineering, StanfordUniversity, Stanford, California, Abstracts, SGP-TR-171Schlagintweit, R. de., 1865, Enumeration of HotSprings of India and High Asia : Journal of theAsiatic Society of Bengal, v. 33 (2), p. 51-73.Shanker R., Padhi, R.N., Arora, C.L., Prakash, G.,Thussu, J.L., and Dua, K.J.S., 1976, Geothermalexploration of Puga and Chumathang geothermalfields, Ladakh, India : Second U.N. Symposiumon the development and use of geothermalresources, San Francisco, Abstracts, p. 245-258.Sharma S., Nair, A.R., Kulkarni, U., Navada, S.V.,and Pitale, U.L., 1996, Isotope and geochemicalstudies in Tattapani geothermal area, districtSurguja, Madhya Pradesh, in Pitale UL and PadhiRN, eds, Geothermal Energy in India: GeologicalSurvey of India , Special Publication 45, p. 233-242.Singh, S.B., Drolia, R.K., Sharma, S.R., andGupta, M.L., 1983, Application of resistivitysurveying to geothermal exploration in the Pugavalley, India: Geoexploration, v. 21, p. 1-11.Thussu, J.L., Rao, S.M., Prasad, J.M., Navada,S.V., and Kulkarni, U., 1987, Isotope compositionof thermal and non-thermal waters of Tattapanihot spring area, district Surguja, Madhya Pradesh:Geological Survey of India, Records, v.115, Pt.6,p. 56-65.80

Australian Geothermal Conference 2010Queensland Coastal Geothermal Energy Initiative – An Approach to aRegional AssessmentBehnam Talebi 1 *, Mark Maxwell 1 , Sarah Sargent 1 and Steven Bowden 11 Mines and Energy, 80 Meiers Road, Indooroopilly QLD 4068.*Corresponding author: behnam.talebi@deedi.qld.gov.auThe best-known potential geothermal resources inQueensland are located beneath the Cooper-Eromanga Basins in the south-west of the state.The depth to the resources, and their distancefrom potential markets and the existing nationalelectricity grid, are the main challenges for theirdevelopment for power generation in the nearterm. The $5 million Coastal Geothermal EnergyInitiative (CGEI) is the Queensland Government’sprogram to implement the commitment made inthe ClimateSmart 2050 strategy through theQueensland Renewable Energy Fund toinvestigate additional sources of hot rocks forgeothermal energy close to existing transmissionlines and potential markets. The initiative is acooperative project between Office of CleanEnergy and Geological Survey of Queensland(GSQ) and comprises a structured drillingprogram of shallow boreholes for the collection ofnew datasets to identify heat-flow anomaliesalong the east coast of Queensland. Target areashave been identified based on geological andgeophysical characteristics and have been rankedbased on their potential for a geothermalresource. The CGEI will reduce exploration risksand assist potential explorers to explore for anddevelop this source of clean energy inQueensland.Keywords: Queensland, geothermal exploration,drilling, heat flow, inversion modelling.ObjectivesThe main objectives of the CGEI are to identifyprospective areas along the east coast ofQueensland and collect additional datasetsthrough a shallow drilling program. The purposeof this is firstly to increase knowledge of thecrustal temperatures along the coast andsecondly to provide an enhanced assessment ofgeothermal resource potential by generating basicdatasets. This new data should assist geothermalexploration and development programs inQueensland.Heat Flow InvestigationA precise crustal heat-flow determination is thepreferred method of identifying the geothermalprospectivity of a target area. Heat flow is theproduct of the temperature gradient and thethermal conductivity of rocks in the earth’s crust,and can be determined through the sampling andlogging of cored drill holes. A heat-flowinvestigation program has been planned as part ofthe CGEI to evaluate the geothermal prospectivityof selected geological provinces along the eastcoast of Queensland.Drill Target SelectionThe potential occurrence of suitable heat sourceswithin basement was evaluated, based onavailable geophysical datasets and regionalgeological knowledge for each area. The maingeothermal source/reservoir targets are felsiccrystalline basement rocks such as granitoids andrhyolite with high thermal conductivity values(generally greater than 4 Watt per meter Kelvin(W/mK)) and high radiogenic heat productionability (greater than 5 micro Watt per cubic meter(μW/m 3 )).Figure 1a: Location of CGEI drill targets in Central andSouthern Queensland81

Australian Geothermal Conference 2010Priority in the CGEI was given to geologicalprovinces where geothermal source/reservoirunits are overlain by a thick succession of thermalblanketing sediments with harmonic mean thermalconductivity values generally less than 3 W/mK.The assessment process involved a desktopanalysis of each geological province. Theprovinces were selected as they were consideredlikely to have a high heat-flow or evidence ofprevious elevated temperatures. Existingtemperature data from available petroleum wellsin or adjacent to each geological province wereconsidered to infer subsurface temperaturegradients. The main obstacle to this approach hasbeen the variable data quality and availability ineastern Queensland where petroleum drilling isrelatively rare. The temperature gradients wereintegrated with the available geological andgeophysical data with the aim of identifyingpossible sources of heat at depth. Specifictargets were then identified that would test theinterpretation that this part of the geologicalprovince has geothermal potential.To date, forty-seven targets generally within100km of the existing national electricity grid havebeen identified with thirty-two being selected fordrilling. These targets are located from nearCairns in the north to the border of New SouthWales (Figures 1a, b).Figure 1b: Location of CGEI drill targets in NorthernQueenslandTechnical criteria were used to rank targets inorder of likelihood of the identification of ageothermal resource. The criteria were based onthe nature of the basement and the thermalblanketing sedimentary cover. All targets wereranked against these criteria in order to establisha drilling priority. The final location of each drillhole was determined after consultation withlandholders and consideration of terrain condition,environmental and cultural heritage issues. Thespatial distribution of the holes effectively requireseach site to be considered as a drilling project inits own entity. Consequently, a substantialamount of extra work has been required to obtainthe necessary clearances. Phase 1 drilling of the10 highest priority targets is scheduled tocommence in late 2010 (pending contractconfirmation).Geological SettingTraditionally, rifting margins have been identifiedas more prospective for geothermal energyexploitation. However, preliminary geologicalassessment of the geothermal potential of coastalQueensland shows a variety of tectonic settingsmay host high heat producing (HHP) granitesunder insulating cover sequences. Theassessment process resulted in the identificationof five different geological settings which aresummarised in Table 1.Table 1: Geological Setting of CGEI TargetsAGE Geological Setting No. of CGEITargetsProterozoic Intercontinental rifting 5Ordovician-Silurian Thomson Orogen 4Early Carboniferous– Tasman Orogenic Zone 13Early PermianLate Permian – Mid Hunter Bowen Orogen 7TriassicLate Cretaceous -TertiaryCretaceous Rifting 3Intercontinental Rifting: The Proterozoic was atime of elevated heat flow causing theemplacement of voluminous granites andassociated volcanics in northern Queensland(McLaren et al, 2003). These granites wereemplaced during multi-stage intracratonic riftingevents and in some areas they are overlain bysedimentary basins which include the MillungeraBasin (no younger than Triassic, Figure 2) and theJurassic - Late Cretaceous Carpentaria Basin.The heat production values of 5-9 µW/m 3 (GSQGeochemistry database, 2010) from intrusives ofthe Croydon Province and the Georgetown andMt Isa Inliers (including the Esmeralda, Forsaythand Williams Supersuites) delineate a prospectiveheat source at depth in these areas. A number ofgeophysical tools including gravity, magnetic,magnetotelluric and radiometric ternary images aswell as sporadic drilling data were used to identify82

Australian Geothermal Conference 2010five targets which are considered worthy of furtherinvestigation. These targets are found beneaththe insulating units of the Karumba, Carpentariaand Millungera Basins and the Georgetown Inlier(Figure 1b).Figure 2: Millungera Basin CGEI rationale, showing gravityresponse (Bouguer), location and extent of outcroppingProterozoic intrusives and high heat producing rocksThomson Orogen: Another region of outcroppingpotentially HHP granites is located within thebasement packages of the Ordovician-SilurianThomson Orogen. The heat production values ofI-type and S-type granites are generally greaterthan 5 µW/m 3 . If properly insulated, these graniteswould provide a geothermal heat source at depth.A geophysical and geological assessmentidentified the western and southern Bowen Basin,eastern Drummond Basin,and the Burdekin Basin,as target areas, as inferred HHP granites arepresent in the basement complexes of thesebasins with Tertiary volcanics, which couldindicate an additional, younger contributing heatsource.Tasman Orogenic Zone: The Carboniferous-Permian intrusions of the Tasman Orogenic Zone(which encompasses the Northern TasmanOrogenic Zone and the New England Orogen)contain medium to high concentrations of theradiogenic elements uranium, thorium andpotassium and should be prospective forgeothermal energy potential. Within the NorthernTasman Orogenic Zone, the Carboniferous–Permian Wypalla Supersuite and the PurkinGranite have heat production values between 3.5-8.55 µW/m 3 (GSQ Geochemistry database,2010). The heat production potential of theseintrusive units at depth in conjunction with a goodinsulating capacity and thickness of overlyingcover sequences have identified three targetswithin the Hodgkinson Province, northernEromanga Basin and Camel Creek Province. Theexistence of anomalous heat flow within theHodgkinson Province in particular can be inferredby the Innot hot spring which has a surfacetemperature of 71°C (Lottermoser & Cleverley,2007). This spring is located in the southern partof the province.The New England Orogen contains numerous I-type and S-type intrusions and associated felsicvolcanics. Whole rock geochemical analyses ofthese units indicate medium concentrations ofpotassium, thorium and uranium. As a result thegeological and geophysical assessmentsundertaken delineated where these intrusivesextend at depth underneath insulating basins. Theoverlying insulating sedimentary packages includeoil shale units of the Duaringa Basin andextensive coal measures of the Bowen, Clarence-Moreton and Styx Basins. Nine sites wereselected for heat flow determinations including theBulgonnuna Volcanics, Lizzie Creek Volcanics,Bowen Basin, Roma Granite, Styx Basin,Duaringa Basin, western Clarence Moreton Basin(Jandowae) and Inglewood (Figure 1a).Hunter Bowen Orogen: The onset of the Permian-Triassic Hunter Bowen Orogen initiated thewidespread emplacement of granites along theNew England Fold Belt. Elevated heat is likely asthe granitic outcrops generally have heatproduction values greater than 5 µW/m 3 . Gravity,seismic and magnetotelluric datasets suggestthese potentially HHP granites extend at depthunderneath the cover of the Tarong, Mulgildie,Clarence-Moreton, Galilee and MaryboroughBasins, which all contain coal measures. A moredetailed geological and geophysical assessmentof these basins identified seven target sites asprospective for elevated heat flow namely theThunderbolt Granite, Tarong Basin, easternClarence-Moreton Basin, Mulgildie Basin,Stanthorpe Granite, Barcaldine and northernMaryborough Basin (Figures 1a and 3).Cretaceous Rifting: The Cretaceous saw thecessation of the active convergent margin alongthe east of the Australian continent and theinitiation of rifting of the Tasman and Coral Seas.Cretaceous granites lining the western edge ofthe Maryborough Basin (Figure 3) may be aprospective geothermal target. The sedimentthickness of the overlying Maryborough Basin isup to 3400m with two significant coal sequences,Tiaro Coal Measures and Burrum Coal Measures,forming an excellent insulator to any potentialheat producing intrusive at depth.The Hillsborough Basin (Figure 4) and NarrowsGraben sites primarily target the more northernCretaceous intrusions. These two sites may alsobenefit from additional heat sources associatedwith intrusions and related volcanics of the Late83

Australian Geothermal Conference 2010Cretaceous Whitsunday Large Igneous Province.In addition, volcanics resulting from Tertiary riftingand intraplate volcanism suggest that there stillmay be significant heat production in this area.The insulating capacity of the Hillsborough Basinand Narrows Graben sequences is considered tobe good due to the presence of significantintervals of oil shale.Sampling is to be undertaken at the drill siteimmediately after the core has been logged.Each sample is enclosed in plastic wrap topreserve their in-situ fluid saturation levels prior todispatch to a laboratory. Visual assessment of thedownhole geology (lithological log) against thetemperature log will also be used as a tool forfurther sampling in case additional thermalconductivity analyses are required at a later stage(Figure 5).Figure 3: Maryborough Basin CGEI rationale, showinggravity response (Bouguer), location and extent of Jurassic-Cretaceous and Permian-Triassic outcropping intrusivesDrilling Program and Data CollectionAn effective heat-flow determination processrequires a shallow (300–500m) drill hole, suitablycompleted to enable a temperature log to beacquired that accurately records the temperaturein lithological units intersected. In order to get anaccurate temperature log the hole needs toremain undisturbed for sufficient time (typically 5–8 weeks) to minimise any remaining drillinginducedtemperature disturbances and essentiallyensure that the downhole temperature isadequately re-equilibrated. Each well is thenthermally logged from the surface to the bottom ofthe hole by measuring a temperature at specificintervals (typically every 1 meter). From thesemeasurements, temperature gradients can becalculated for the corresponding intervals downthe hole.The hole should also be continuously cored fromthe base of unconsolidated units to total depth(~320m) to recover at least 200-250 meters ofcontinuous core enabling samples to be takenand analysed for thermal conductivity propertiesin order to establish a thermal conductivity profilefor the hole. Core samples are usually taken atregular intervals, typically every 10 metres.Figure 4: Hillsborough Basin CGEI rationale, showing gravityresponse (Bouguer), ternary radiometric image, location andextent of outcropping Carboniferous and CretaceousintrusivesThe CGEI drilling program is likely to consist of upto 32 HQ size shallow boreholes to a nominaldepth of ~320 metres cased with continuous PVC.The hole design and completion have been84

Australian Geothermal Conference 2010optimised to ensure that a high quality datacollection process is attained throughout theprogram (Figure 6). A call for tender was releasedin June 2010 for the selection of drillingcontractors. Geophysical downhole logs such asResistivity, Spontaneous Potential (SP) andGamma Ray are being run on an “as–required”basis with the aim of determining possible aquiferleakage into the hole that may interfere with thetemperature profile and also for data qualitychecking purposes.a purely conductive regime and therefore aconstant heat flow across all lithological unitsintersected. From the global heat-flow database,the mean surface heat flow of the Australiancontinent is approximately 65 mW/m 2 , so areaswith heat-flow values greater than this mayindicate promising areas for further geothermalinvestigation and exploration activities providedthat the thermal resistance of the overlyinglithological units is sufficient to support presenceof economical temperature at depth.Considering the spatial coverage of the CGEI,one dimensional inversion modelling of theconductive heat-flow regime in the vicinity of eachborehole is to be undertaken using the newtemperature and thermal conductivity datasetscollected (Figure 7). Modelled data will then beextrapolated to 5 km, the economic drilling depthfor most geothermal resources, to predicttemperatures at this depth range. Generally,temperatures of greater than 200°C are expectedat such depths for Engineered GeothermalSystem (EGS) to be commercially viable forelectricity generation.Figure 5: Example of a temperature log versus lithological logHeat-Flow ModellingMathematically, the average vertical conductiveheat flow at the earth’s surface, q (in W/m 2 ), canbe calculated using the Fourier expression:q k.dTdzWhere (k) is the rock thermal conductivity (inW/mK) at depth z, and (dT/dz) is thecorresponding vertical temperature gradient (inK/m) over the same interval. As shown by theequation, the temperature gradient has an inverserelationship with thermal conductivity (onedecreases as the other increases). This assumesFigure 6: Schematic borehole designIn addition, depth predictions of the 100, 150 and200°C isotherms may be approached dependingon data availability for estimating the formationsintersected at those isotherm depths. This willassist in the assessment of the prospectivity ofeach target area for either Hot SedimentaryAquifer (HSA) or Engineered Geothermal System(EGS) development in the future.CollaborationCollaboration with other government agencies,research centres and industry is an important part85

Australian Geothermal Conference 2010of this initiative. In this respect, GeoscienceAustralia will provide technical support to theCGEI by providing down-hole temperature loggingand laboratory analysis services through anagreement under the National GeoscienceAccord. Furthermore, additional research over theCGEI target areas has been discussed throughcollaboration with the Queensland GeothermalEnergy Centre of Excellence (QGECE) based atthe University of Queensland through postgraduateresearch studies.Figure 7: Example of 1D inversion heat flow modellingFuture WorkFuture work that would contribute to a betterunderstanding of geothermal prospectivity ineastern Queensland would include the collectionof new heat-flow data at a closer spacing forprospective areas identified from the CGEI. Anassessment of current regional and local in-situstress fields in those areas could also beundertaken to evaluate the susceptibility ofpotential reservoir rock to the artificial permeabilityenhancement process.SummaryThe Coastal Geothermal Energy Initiative is theQueensland Government’s program for identifyingsources of hot rocks for geothermal energy closeto existing national electricity grid and potentialmarkets along the east coast of Queensland. Theinitiative will be implemented through a structureddrilling program of up to 32 shallow boreholes witha nominal depth of ~320 metres to collect newtemperature and heat-flow datasets. The majoraim of the initiative will be to increase knowledgeof the crustal temperatures in selected geologicalsettings along the coast and provide backgrounddata for industry that would consequentlystimulate exploration activities for geothermalenergy in Queensland. It is anticipated that finalresults of the initiative will be publically availableby the middle of year 2012.AcknowledgementsThe authors wish to acknowledge staff from otherGSQ units who are strongly supporting theimplementation of the Coastal Geothermal EnergyInitiative by any means and in every aspect.ReferencesBeardsmore, G. R., and Cull, J. P., 2001, CrustalHeat Flow: A Guide to Measurement andModelling, Cambridge University Press, 324pp.Cull, J. P. and Beardsmore, G. R., 1992,Statistical methods for estimates of heat flow inAustralia, Journal of Exploration Geophysics 23,p. 83-86.Denaro, T. and Dhnaram, C. (Compilers), 2009,Queensland Minerals 2009. A Summary of majormineral resources, mines and projects.Department of Mines and Energy, 87pp.Fitzell, M., Hamilton, S., Beeston, J., Cranfield, L.,Nelson, K., Xavier, C. and Green P., 2009,Approaches for identifying geothermal energyresources in coastal Queensland, Proceedings ofthe 2009 Australian Geothermal EnergyConference, 6pp.McLaren, S., Sandiford, M., hand, M., Neumann,N., Wyborn, L. and Bastrakova, I., 2003, The hotsouthern continent: heat flow and heat productionin Australian Proterozoic terranes, In: Hillis R.R. &Müller R.D. eds. Evolution and Dynamics of theAustralian Plate, Geological Society of AustraliaSpecial Publication 22.Murray, C., 2003. Granites of the northern NewEngland Orogen, in BLEVIN, P., JONES, M., &CHAPPELL, B., editors, Magmas toMineralisation: The Ishihara Symposium,Geoscience Australia, Record 2003/14, 19-24.Rybach, L., 1989, Heat flow techniques ingeothermal exploration, First Break Vol. 7, No.1,p. 9-16.Somerville, M., Wyborn, D., Chopra, P., Rahman,S., Estrella, D. and Van der Meulen, T., 1994, HotDry Rocks Feasibility Study, Energy Researchand Development Corporation, Report 243,133pp.Webb, A. W. and McDougall, Ian, 1968, Thegeochronology of the igneous rocks of EasternQueensland, Australian Journal of EarthSciences, 15: 2, 313 — 346.86

Australian Geothermal Conference 2010Geothermal Exploration in Victoria, AustraliaDavid H Taylor, Ben Harrison, Michelle Ayling, Kate BassanoGeoScience Victoria, GPO Box 4440, Melbourne, Victoria, AustraliaCorresponding author: david.taylor@dpi.vic.gov.auGeothermal energy is keenly sought as asustainable, low emissions energy source forVictoria. Exploration companies have been activesince 2006 looking for geothermal systems wheretemperatures of greater than 150 C would allowelectricity generation. Initial exploration waslargely guided by legacy temperature data ingroundwater and petroleum bores but lots ofdedicated heat flow data is now being acquired,including a State-wide program by thegovernment. Active volcanic systems are notpresent but there is good potential for hotsedimentary aquifers already known from thelegacy data. Other heat sources such asPalaeozoic radioactive granites or residualmagmatic heat from recent basaltic volcanism arepossible but data is very limited. As explorationprogresses, a number of different geothermal playtypes across the various heat sources are beinginvestigated. A number of inferred resourceshave already been declared, in preparation fordeep appraisal drilling.Keywords: Hot Sedimentary Aquifer, Heat Flow,geothermalVictorian geothermal backgroundThe oil shocks of the 1970’s piqued interest intoalternate energy sources. Town water suppliesbeing drawn from thick Cretaceous-Tertiarybasins along the Victorian coast at that time,showed temperatures of about 50-90 C at 1-2 kmdepth. The potential of direct use of this hot waterwas investigated (King et al., 1987). Petroleumexploration in the same area also occasionallyintersected deeper basin aquifers withtemperatures of about 130-150 C at 3.5-4 kmdepth (Woollands and Wong, 2001).Low oil prices throughout the 1990’s suppressedinterest in alternate energy sources but the returnof higher prices and concerns around greenhousegas emissions have again sparked interest ingeothermal energy. Dedicated legislation was putin place in 2005 and a grid of large explorationpermit blocks was established across the State.Exploration companies took up many of thepermits in two rounds of tenders over 2006 and2008. About two thirds of the State is now underexploration for the next 3-4 years by sevencompanies with work plan commitments ofapproximately $365M.Victorian Geothermal DataTo support the tender bidding process for permits,a preliminary review of geothermal prospectivitywas published by the State government geologyorganisation (Driscoll, 2006). This report includeda compilation of about 300 temperature datapoints from pre-existing groundwater andpetroleum bores but the accuracy and precision ofmuch of the petroleum temperature data forgeothermal assessments is questionable.A handful of heat flow calculations from precisiontemperature logging and rock conductivityanalysis already existed as a result of academicstudies since the 1970s. Many of the geothermalexploration companies have started collectingmuch more of this dedicated type of geothermaldata. To accelerate this type of work the Stategovernment has undertaken a data collectionprogram for heat flow.The Victorian Heat Flow MapIn 2009-10 the State government commissioned aone year $500,000 project to generate a Statewideheat flow map. Over the year, this projecthas compiled the existing academic and companydata, plus collected new data, to generate over100 heat flow measurements that give goodcoverage over much of the State. Data points aregenerally close enough together (ideally closertogether than the crust is thick: about 40 km) sothat interpolation of the point data into a mapshould still have regional scale meaning.GeoScience Victoria directly collected boreholetemperature profiles in the southern Murray Basinand across the Otway Basin using its owntemperature logging unit. As part of the NationalGeothermal Energy Project (Budd et al., 2009)GeoScience Australia collected temperature datafrom the northern Murray Basin using theirtemperature logging unit. GeoScience Victoria isalso working collaboratively with MelbourneUniversity who have collect data from theGippsland basin. The exploration companies inthe Gippsland Basin have also collected a fairamount of new temperature data in that regionand made it available. Several recentunpublished honours thesis and some additionaltemperature logging for GeoScience Victoriaundertaken by Hot Dry Rocks add further data tothe previous handful of heatflow measurementsthat was all that was available several years ago.87

About 200 conductivity analyses have beenperformed by Hot Dry Rocks on samples from theboreholes measured for temperature. This dataallows reasonably precise heat flow calculationsto be generated. Heat flow modelling for datafrom completely new holes was undertaken byHot Dry Rocks. The various pre-existing honoursthesis and company work was remodelled byGeoScience Victoria using consistently appliedmethodology and better informed conductivitiesstemming from the new sampling.This regional scale appraisal of heat flow, whencombined with analysis of thickness of basincover, will help show ‘where it is hot and where itis not’. The map will help companies direct theirpermit scale efforts, as well as informinggovernment policy on geothermal resourcepotential.This work is still in progress and thus currentassessments of Victoria’s geothermal potentialstill rely on looking at geological factors.Australian Geothermal Conference 2010Geological Framework for geothermalpotentialGeothermal potential depends on the interactionof a number of geological factors. The bestresources are likely to exist in regions where highheat flow passes through rocks of low conductivity(good insulation) to create high temperature atshallow depth (Duffield & Sass, 2003).The recent publication of a more completegeothermal systems assessment framework(Cooper & Beardsmore, 2008) outlines that inaddition to (1) the heat producing basement (2) aninsulating blanket; there needs to be (3) a fluidavailable to extract and move the heat; and (4) areservoir to accommodate the fluid.Applying this four factor analysis to the majorgeological provinces of Victoria gives some ideaof their relative geothermal prospectivity. Thebroad diversity in the age and types of rocksacross Victoria (Figure 1) gives some potential forall three types of geothermal systems: Hot Rock;Hot Sedimentary Aquifer; and Magmatic.Figure 1: Geological geothermal province map88

Three broad geological/geothermal provinces canbe delineated across Victoria: (1) The Palaeozoicbedrock (2) the onshore Otway and GippslandBasins and (3) the Murray Basin.The Palaeozoic bedrock consists predominantlyof Cambrian to Devonian deep marine muddysiliciclastics that have been tightly folded andcleaved at various times in the Palaeozoic (blues,browns and purples of Figure 1). This bedrockalso underlies the other provinces and thus givesgood insight into the potential of their heatproducing basement. Numerous granites intrudethis bedrock (red blobs on Figure 1). In west andcentral Victoria some of them have felsic,fractionated geochemistry with mild enrichment inheat producing elements, so Hot Rock plays maybe possible where granites lie buried within thebedrock or beneath the adjacent basins. In thesouthwest, an extensive province of youngbasalts has erupted onto the bedrock in the lastfew million to tens of thousands of years (the lightpink in Figure 1). Teleseismic data suggests amantle hot spot still underlies this recentvolcanism (Graeber et al., 2002). In addition to anincreased mantle heat flow contribution acrossthis region, there may be some residual magmaticheat in the crust from the volcanic activity to allowMagmatic geothermal plays.The Onshore Otway and Gippsland Basins areCretaceous rift basins associated with Gondwanabreak-up (greens and yellow along coast in Figure1). These basins are well characterised thanks toa long history of petroleum exploration with manydeep wells. They contain several kilometres ofmuddy sediments overlying a coarser grainedbasal unit. This basal unit provides a naturalreservoir already charged with hot water beneathan insulating blanket to create an attractivefairway for Hot Sedimentary Aquifer plays. In theOtways this basal unit has been intersected uponthe basin floor but in the Gippsland basin this unithas yet to be tested away from the margins, at thedepths necessary for a geothermal play. Beneaththese basins there is also potential for Hot Rockplays in places where they are floored by granites.The Murray Basin is a Tertiary intracratonic sagbasin (light greens and yellow inland in Figure 1).It generally contains only a few hundred metres ofsandy marly cover. This is insufficient cover toact as a heat blanket but in some places deepertroughs – such as the Wentworth, Netherby,Numurkah and Ovens - exist and can contain upto a couple of kilometres of sediment that is poorlyknown since there has been limited petroleumexploration here. Alignment of favourable factorsover the deeper troughs may allow Hot Rock orHot Sedimentary Aquifer plays but thefavourability of the geothermal factors in thisprovince has yet to be validated.Australian Geothermal Conference 2010Pathways to Resource DevelopmentIn the absence of detailed, dedicated geothermaldata, most of Victoria can be viewed as ‘blue sky’or perhaps ‘green fields’ at best, in those areaswhere some legacy petroleum data exists. Thenew heat flow map and better characterisation ofrock thermal insulation (conductivities) will providea new level of interpretation on prospectivity.Most of the Victorian geothermal explorers aresmall companies with limited amounts of capitaland cash-flow to fund their 5 year explorationprograms. Ideally, as these companies collectinformation and decrease risks and unknowns,they could call for more capital through either debtor equity raisings until it becomes probable that amajor backer would farm-in for development. TheGlobal Financial Crisis has badly affected thistraditional venture capital pathway through toresource development.At the national level the Federal government hascommitted a significant amount of funds into ageoscience investigation program and also put upmoney for co-funding deep appraisal drilling. Ifthese drilling appraisals lead to early success,then perceived risks around geothermal energymay be reduced and allow easier funding for thewhole industry from the more traditional pathway.The State government has also offeredsubstantial funds towards industry assistance forshallow and deep drilling plus contingent moneyfor demonstration power plants.Company announcements around InferredResources and/or conceptual targets suggest thatseveral thousand MW of electricity generationmay be possible but it is still early days for thegeothermal industry in Victoria.ReferencesBudd, A.R., Holgate, F.L., Gerner, E.J., Ayling,B.F. & Barnicoat, A.C., 2009, Pre-competitivegeoscience for geothermal exploration anddevelopment in Australia: GeoscienceAustralia's Onshore Energy Security Program.Proceedings, Australian Geothermal EnergyConference, Brisbane, Australia.Cooper, G.T. & Beardsmore, G.R., 2008,Geothermal systems assessment:understanding the risks in geothermalexploration in Australia. Proceedings, EasternAustralian Basins Symposium III, Sydney,Australia.Driscoll, J., 2006, Geothermal propsectivity ofonshore Victoria, Australia. VIMP Report 85.GeoScience Victoria.Duffield, W.A. & Sass, J.H., 2003, Geothermalenergy: Clean power from the earth's heat.Circular, U. S. Geological Survey 1249.Graeber, F.M., Houseman, G.A. & Greenhalgh,S.A., 2002, Regional teleseismic tomographyof the western Lachlan Orogen and the Newer89

Australian Geothermal Conference 2010Volcanic Province, southeast Australia.Geophysical Journal International, v. 149,249-266.King, R.L., Ford A.J., Stanley, D.R., Kenley, P.R.,& Cecil, M.K., 1987, Geothermal Resources ofVictoria, Department of Industry, Technologyand Resources and Victorian Solar EnergyCouncil, Melbourne.Woollands, M.A., & Wong, D., (eds.) 2001,Petroleum Atlas of Victoria, Australia,Department of Natural Resources andEnvironment.90

Australian Geothermal Conference 2010Geochemistry of silica rocks in the Drummond Basin as a record ofgeothermal potentialI. Tonguç Uysal* 1 , Alexander W. Middleton 1 , Robert Bolhar 2 , and Massimo Gasparon 2Queensland Geothermal Energy Centre of Excellence, The University of Queensland, Brisbane, QLD. 4072* 1Department of Earth Sciences, The University of Queensland, Brisbane, QLD. 4072 2Silica is the most abundant mineral in the Earth’supper crust; yet its trace element composition hasnot been utilised routinely in the exploration ofnatural resources. Siliceous sinters are pristinerocks that precipitate at the surface from springwaters heated by deep-seated magma chambers.They can be useful indicators of hot radioactivemagmatic basements in sedimentary basins.However, distinguishing silica sinters from othersiliceous lithologies is often ambiguous becausesiliceous lithologies of various origins occurubiquitously on the ground surface as rock chipsor soil material eroded from deeper geologicalunits. In this study, we characterise trace elementcompositions of silica sinters to providegeochemical criteria for exploration of geothermalenergy sources. We investigated silica sinters andthe associated hydrothermal quartz veins,volcanic rocks, silicified hydrothermal breccias,alteration minerals, and some soil samples fromthe Paleozoic Drummond Basin, Australia by ICP-MS trace element analysis.Keywords: Sinter, quartz, geochemistry,geothermal, Drummond Basin, Galilee BasinSamplingSamples for this study were collected from theTwin Hills area hosting high-grade epithermal goldresource, 309 and Lone Sister (BMA componyreport). Location 309 mine is a fault-boundedarea characterised by the occurrence of a silicifiedbreccia system capped by sinter deposits,whereas Lone Sister represents a rhyolite dome.The textual features of the sinters at 309 are verysimilar to those of sinters from other locations inthe Drummond Basin (Cunneen and Sillitoe,1989; White et al., 1989). They are welllaminated(from white through gray to orangebrown),consisting of dense and vitreouschalcedony.Results and implicationsThe Drummond Basin sinters and quartz veinsare unique in having anomalously enrichedincompatible element (Cs, Li, Be, W, U, Th andrare earth elements) concentrations incomparison to hydrothermal quartz veins fromvarious granitic-pegmatitic systems elsewhere inthe world (Figure 1). Bonding factors such as ionicsize and charge are the main factors controllingthe distribution of elements in precipitatingcrystals. Large-ion lithophile elements (LILE, e.g.Rb, Cs), large highly charged cations (e.g. W 6+ ,U 4+ ) and small variable charged cations (e.g. Be,Li ) are usually prevented from being incorporatedinto crystalline phases during early magmaticprocesses (Strong, 1981). Hence theseincompatible elements tend to concentrate inresidual melts with high concentrations of fluidsand volatiles, which commonly occurs duringpegmatitic stages of granite magmatism. It haslong been recognised that that Cs, Li, Rb, Be (thealkali and alkaline earth elements) can be quitemobile during water/rock interaction (Nesbitt andMarkovics, 1997; Nesbitt and Young, 1989).Tracing such elements can thus be used as ageochemical tool for locating areas ofhydrothermal convection cells interacting withdeep heat-producing granitic rocks. Alkali mobileelement concentrations are divided byconcentration of most immobile elements such asLu to compare alkali element mobility (Fig. 2). Themajority of sinter and quartz samples within ornear the highly mineralised area (309 mine) aresignificantly more mobile for Cs, Li, Rb and Bethan the sinters and silica deposits from areasdistal to 309 (except the quartz vein from OffLimits) and the volcanic rocks (Fig. 2). Plotting alldata set, a good correlation (R 2 =0.70, powerregression) is evident between Cs/Lu and Be/Lu,with the highest ratios being for sinter samplesfrom 309 (Fig. 2A). Considering only silicadeposits in Fig. 8A (black filled symbols) asignificantly better correlation (R 2 =0.79, powerregression) would be obtained. Sinter and quartzsamples from 309 have also high Rb/Lu and Li/Luratios, (Fig. 2B). A well-developed correlation (R 2=0.77) between Rb/Lu and Li/Lu is evident amongsilica samples from the entire Twin Hills area (Fig.2B). The interaction of hydrothermal waters withthe late Carboniferous alkaline igneous rocksresulted in the breakdown of K-Feldspar releasingalkali elements, which were subsequentlymobilised in these fluids. This process led to thegeneration of an evolved, fertile hydrothermal fluidsystem that carried precious metals (Au, Ag) aswas the case at 309 and Lone Sister in the TwinHills area.U and Th content of silica and clay samples fromthe surface can also be used in evaluating hot dryrock geothermal potential of the igneous rockassociation buried by heat-insulating sediments inadjacent sedimentary basins. Compared topegmatite quartz compiled from the literature(Götze et al., 2004), U and Th concentrations in91

Australian Geothermal Conference 2010Figure 1: Upper continental crust (UCC) –normalised trace element concentration of the silica deposits from the DrummondBasin (Twin Hills area) in comparison to quartz from various granitic/pegmatitic environments.sinter and quartz deposits from the DrummondBasin are significantly elevated (Fig. 3A-B). U andTh contents of the illitic clay minerals that formedin response to the interaction of hydrothermalfluids with host rocks are similarly high as thosefrom the Cooper Basin. The latter represent aradiometric heat production from the basementgranite (Middleton, 1979), whereas the clayconcentrations for the Bowen Basin inQueensland and Paris Basin in Europe are similarto those of the upper continental crust (Fig. 6A-B).Geothermal potential in the Galilee BasinIgneous rocks of the Drummond Basin occur asthe basement of the adjacent Galilee Basin ofUpper Carboniferous to Middle Triassic age(Evans, 1980). The late Permian coal depositsand carbonaceous pelitic rocks in the GalileeBasin are ideal heat insulating sediments (Nunand Li, 2002) that would store the radioactive heatgeneration in the basement. Indeed,temperatures of about 80C at 1000 m weremeasured in coal seam gas drilling boreholes inthe Galilee Basin (pers. commun. with severalcoal seam gas companies) that indicate high heatflux from the basement. Furthermore, the vitrinitereflectance values measured on the coal coresthroughout most of the Galilee Basin rangebetween 0.63 and 0.73 %R o max (Holland andApplegate, 2008), translating into reservoirtemperatures of about 95C - 110C (Barker andPawlewicz, 1986).Figure 2. Alkali element mobility in the Twin Hills area.92

Australian Geothermal Conference 2010Figure 3. U – Th abundances of sinter and clay deposits fromthe Drummond Basin in comparison to those of othersedimentary basins. Pegmatite quartz from Götze et al.(2004).ConclusionThe results of the pilot study suggest that traceelement geochemistry of silica samples and theassociated clay alteration can serve as a powerfultool in 1) discriminating between sinter and quartzveins from hydrothermal systems driven bygranitic heat sources and other siliceous rocktypes (barren quartz/silica deposits) and 2) to beused in reconnaissance soil sampling to identifyhydrothermal alteration zones associated withheat-producing granite plutons. Further workaimed at investigating similar samples fromvarious settings with different ages will expandexisting knowledge of how these techniques canbe used to effectively discriminate areas with highgeothermal potential from barren systems on aroutine basis.Evans, P.R., 1980. Geology of the Galilee Basin.In: R.A. Henderson and P.J. Stephensen(Editors), Geology and Geophysics ofNortheastern Australia. Geological Society ofAustralia, Queensland Division, pp. 299–305.Götze, J., Plotze, M., Graupner, T., Hallbauer,D.K. and Bray, C.J., 2004. Trace elementincorporation into quartz: A combined study byICP-MS, electron spin resonance,cathodoluminescence, capillary ion analysis, andgas chromatography. Geochimica EtCosmochimica Acta, 68(18): 3741-3759Holland, J.R. and Applegate, J.K., 2008. RecentCoalbed Methane Exploration in the GalileeBasin, Queensland, Australia. 2008 InternationalCoalbved & Shale Gas Symposium.Middleton, M.F., 1979. Heat flow in the Moomba,Big lake gas fields of the Cooper Basin andimplications for hydrocarbon maturation. Bulletinof the Australian Society of ExplorationGeophysics, 10: 149-155.Nesbitt, H.W. and Markovics, G., 1997.Weathering of granodioritic crust, long-termstorage of elements in weathering profiles, andpetrogenesis of siliciclastic sediments.Geochimica Et Cosmochimica Acta, 61(8): 1653-1670.Nesbitt, H.W. and Young, G.M., 1989. Formationand diagenesis of weathering profiles. Journal ofGeology, 97(2): 129-147.Nunn, J.A. and Lin, G., 2002. Insulating effect ofcoals and organic rich shales: implications fortopography-driven fluid flow, heat transport, andgenesis of ore deposits in the Arkoma Basin andOzark Plateau. Basin Research(14): 129-145.Strong, D.F., 1981. Ore deposit models – 5. Amodel for Granophile mineral deposits. GeoscieCanada, 8: 155–161White, N.C., Wood, D.G. and Lee, M.C., 1989.Epithermal sinters of Paleozoic age in northQueensland, Australia. Geology, 17(8): 718-722.ReferencesBarker, C.E., Pawlewicz, 1986. The correlation ofvitrinite reflectance with maximum temperature inhumic organic matter. In: Buntebarth, G. andStegena, L. (Eds.), Paleogeothermics, New York,Springer-Verlag, p. 79-83.Cunneen, R. and Sillitoe, R.H., 1989. Paleozoichot-spring sinter in the Drummond Basin,Queensland, Australia. Economic Geology, 84(1):135-142.93

Australian Geothermal Conference 2010Estimates of the magnitude of topographical effects on surface heatflow from finite element modellingGordon Webb 1 * and Ben Harrison 21 Northern Territory Geological Survey, PO Box 8760, Alice Springs NT 0871, Australia2 School of Earth Sciences, The University of Melbourne, Victoria 3010, Australia* Corresponding author: gordon.webb@nt.gov.auIntroductionSurface heat flow calculations are employed bygeothermal exploration companies to estimate thegeothermal gradient down to depths of 3 km to6 km. Surface heat flow is typically calculatedfrom coincident thermal conductivity data orestimates, and temperatures obtained by drilling aslim-line hole down to depths of 300 m to 1000 m.At such shallow depths the effect of a variablesurface topography needs to be considered whenperforming temperature extrapolations to greaterdepth. This is because, for a given crustal volumewith an undulating surface, heat will tend to flowtowards topographic lows and away fromtopographic highs as it migrates from the base tothe ground surface. A linear approximationmethod was developed by Lees (1910) to correctfor topography under idealised geometricmountain ranges.This study presents results from a series of finiteelement method (FEM) models that predict themagnitude of variation of surface heat flow at thesurface of a homogenous two-dimensional slice ofcrust with a uniformly curved surface. The resultsfrom this simple geometry are then compared toirregular topographic surfaces derived from realworld examples.The main advantage of using the FEM method,described in this study, - rather than the methoddescribed by Lees (1910) - is that any regular orirregular surface topography may be used.Keywords: surface heat flow, temperature,numerical modelling, topography.Model ParametersA series of two-dimensional model geometrieswere constructed with a flat base and sides and asinusoidal top edge representing the topographicsurface (Figure 1). The models were all 10 kmdeep (measured from the midway point of thesurface curvature to the base) to minimise nearfieldeffects in fixed boundary conditions. Thesurface topography has a sinusoidal form with awavelength denoted by λ and maximum elevationdifference denoted Δh.The width of the models was scaled with respectto the x-axis to generate different values for theratio Δh/λ. Fifteen model geometries wereconstructed with values of Δh/λ varying between0.01 (representing subdued, long wavelengthtopography) and 0.9 (representing steep, shortwavelength topography). Δh had a constant valueof 412 m for each of the models. The lower Δh/λvalues might represent the gently undulatingterrains encountered in areas such as the DarlingRanges near Perth, while Δh/λ values of about0.07 might represent areas of moderatetopographic expression, such as the FlindersRanges and Mt Lofty Ranges in South Australia.The model volume was defined as “nonradiogenicgranite” with an isotropic thermalconductivity (k) of 3.2 W.m -1 .K -1 , a heat capacity(C) of 8500 J.kg -1 .K -1 and a density (ρ) of2600 kg.m -3 .Figure 1: Diagram showing the generic geometry of themodels.A constant temperature of 20 degrees Celsiuswas maintained at the top surface of the models.The sides of the model were defined to besymmetrical and reflective to remove edgeeffects.A uniform heat flux (Q) was applied to the bottomsurface of each model. A basal heat flux of60 mW.m -2 was applied to each of the 15 models.Ten additional models were run with a basal heatflux of 80 mW.m -2 for comparison. The modelswere meshed with a uniform Lagrange linearmesh with greater than 10,000 nodes per modelspace. The models were solved iteratively usingthe time-varying relationship,until the model residuals approach zero and asteady-state solution was attained.ResultsFigure 2 shows the percentage decrease orincrease in the modelled surface heat flow with94

Australian Geothermal Conference 2010respect to basal heat flux at the highest andlowest topographic points, respectively. Forrelatively short wavelength topography (Δh/λ >0.2) the modelled surface heat flow differs fromthe basal heat flux by greater than 50%. Themodelled surface heat flow is within 5% of thebasal heat flux value for very long wavelength,subdued topography (Δh/λ < 0.02).wavelength - low relief topography the observeddifference will be less but may still be significant.These models also predict that the depth ofthermal perturbation is related to the ratio of therelief height and the wavelength of surfacetopography and that this relationship is not linear.Figure 2: Plot of the percentage difference between themodelled surface heat flow with respect to the basal heat fluxmeasured at the highest (peak; negative change) and lowest(valley; positive change) points on the surface topography.Curves were generated below the highest andlowest topographic points that describe thevertical variation between the modelled surfaceheat flow and the basal heat flux (figure 3). Eachpair of curves is a characteristic of the specificgeometry of the model surface. An arbitrary pointat which the curves have decayed to within 5% ofthe basal heat flux value can be usefullycompared in each model to assess the depth atwhich the thermal perturbation is significantlydiminished (figure 4). The depth of the thermalperturbation gradually increases with increasingwavelength (decreasing Δh/λ) until a maximumpenetration is reached at a Δh/λ value of ~0.04(figure 4).DiscussionThe model solutions presented here represent aninitial attempt to assess the magnitude of thevariation of surface heat flow measurements withrespect to regional heat flux due to the effect oftopography using the finite element method. Anumber of other factors also affect surface heatflow measurement including (but not limited to)internal heat production, ground water flux,altitude/climate effects, and lateral thermalconductivity contrasts (Beardsmore and Cull,2001 and references therein). The approachtaken in this study deliberately ignores these othereffects to specifically investigate the effect oftopography in isolation.These models predict that, for relatively shortwavelength - high relief topography, surface heatflow measurements, will be significantly differentfrom the regional, “deep” heat flux. For longerFigure 3: Profile of the modelled vertical heat flux below thepeak (highest surface elevation) and the valley (lowestsurface elevation) for model hft010 (Δh/λ = 0.082, λ = 5 km,Δh = 412 m). The decay of the thermal perturbationincreases with depth until the basal heat flux value isreached. The 5% cut-off values used to generate figure 4 areshown.Figure 4: Plot of the depth at which the model heat flow iswithin 5% of the basal heat flux measured below the highest(peak) and lowest (valley) points on the surface topography.Note that the depth of perturbation attains a maximum valueand then diminishes as a function of Δh/λ.This study highlights a potentially significantsource of uncertainty that must be factoredin/corrected for when extrapolating heat flow datato depths of 3 km to 6 km. Furthermore, theresults of this study suggest that the practice ofestimating resource temperatures by calculatingone-dimensional thermal models may only bevalid for areas where surface topography isrelatively subdued. In most other cases it ispreferable to construct well parameterised twoand/orthree dimensional thermal models.95

ReferencesBeardsmore, G. R. & Cull, J. P. 2001 Crustal heatflow: a guide to measurement and modelling.Cambridge University Press.Lees, C. H. (1910) On the Shapes of theIsogeotherms under Mountain Ranges in Radio-Active Districts. Proceedings of the Royal Societyof London. Vol. 83, No. 563, pp. 339-346Australian Geothermal Conference 201096

Australian Geothermal Conference 2010The next Australian EGS Reservoir – Jolokia 1 stimulationDoone Wyborn*, Robert Hogarth, Delton Chen and Michel RosenerGeodynamics Limited.PO Box 2046, Milton, Queensland, 4064AbstractGeodynamics drilled the Jolokia 1 well in 2008 toa depth of 4,911m in the Innamincka granite at alocation 9.5km west of the Habanero EGS field innorth eastern South Australia. The reservoirstimulation program was delayed until 2010.Initially the delay was related to the need toundertake further evaluation of the well andimplementation of a technical solution to allowmulti-fracture stimulation. The stimulation wasfurther delayed when the casing in Habanero 3well failed in April 2009.The aims of the stimulation are to:(i) Confirm the capacity to create heatexchange reservoirs at locationsspread across the Innamincka graniteresource.(ii) Develop two separate reservoirs in theone well, both deeper and hotter thanthe reservoir operated at theHabanero field(iii) Demonstrate injectivity increasescompared to the Habanero field.Prior to the stimulation the well is to be loggedwith a high temperature imaging tool andcompleted with a liner, tubing and packer. Thestimulation consists of injecting more than 20,000cubic metres of clean water into the well atpressures up to 69MPa. A conservative benignchemical tracer is included in the injected water. Itis expected that a single reservoir will be createdat a depth below the liner set to 4,350m. Thereservoir development is monitored by amicroseismic network of seven triaxial geophonestations up to 5 km from the well and 100 to 200mdeep. The stimulation will be carried out over aperiod of 10 days. A pumping regime has beendetermined based on (i) the understanding thathas come from the stimulations of the Habanerofield, and (ii) the best information frominternational experts who have carried out suchoperations in other projects. Currently theoperation is due to take place in August 2010. Acomprehensive site-specific microseismic riskanalysis was needed before the South Australianregulators could give approval to proceed.Existing understanding of the stimulation processindicates that once a fracture zone begins tostimulate its transmissibility increases by orders ofmagnitude so that that zone dominates thecontinuance of reservoir growth with time. At theHabanero field the reservoir grew outwards at arelatively constant rate of 1 km 2 for every 12,000m 3 of water pumped after an initially more rapidgrowth within 200m radius of the wellbore. Basedon transient well-test analyses the transmissibilityof the Habanero reservoir increased from a prestimulated50 milliDarcy metres to a poststimulation 2000 milliDarcy metres. This increaseis less than has been observed in the lower stressconditions of other projects. The increase isconsidered permanent because the roughness ofthe fracture surface gives rise to self-propping ofthe fracture once it has slipped by an amount inthe order of the dimensions of the grain size of therock. This permanency was vindicated atHabanero since stimulation in 2005 proceededaerially beyond what had been achieved by thefirst stimulation in 2003, with virtually no overlap indevelopment within 1 km of the injection well.Only one reservoir can be developed at a time.Given the stress conditions of the Innaminckagranite determined at the Habanero field there isample potential for the stacking of multiplereservoirs in a vertical interval of well-boreprovided these can be grown separately. TheJolokia reservoir program envisages one reservoirbuilt from Jolokia 1 and a second built from a laterwell Jolokia 2 at a different depth interval. Futureuse of open hole high temperature packers shouldallow for such multiple reservoir development tobe achieved in a single well.The Jolokia 1 stimulation is to be carried out inAugust 2010 and measurement of injectivity andmapping of reservoir growth based onmicroseismic returns will allow comparisons withthe Habanero reservoir. The stimulation programincludes pressure transient fall-off analysis usingdown-hole gauges and pressure-temperaturespinnerlogs to evaluate the distribution ofpermeable fractures.This paper reports on the results of the Jolokia 1stimulation and the implications for furtherdevelopment of the vast Innamincka granitegeothermal resource. The Jolokia stimulation isthe key to proving that EGS development can beeffected at virtually any location in granite bodies.The prize of large scale economic production ofzero-emission EGS electricity will then depend onachieving increased heat extraction rates andreduced development costs.Keywords:Stimulation, EGS, microseismic monitoring, multifracturereservoir97

Australian Geothermal Conference 2010Optimised Fracture Network Model for Habanero ReservoirChaoshui Xu*, School of Civil, Environmental and Mining Engineering, University of AdelaidePeter Dowd, Faculty of Engineering, Computer and Mathematical Sciences, University of AdelaideDoone Wyborn, Geodynamics LimitedAbstract:Fracture networks and their connectivity are theprincipal factors affecting fluid flow in hot dry rock(HDR) geothermal reservoirs. Largely because ofthe complexity of the problem models of HDRreservoirs tend to be over-simplified using either avery limited number of fractures or an equivalentporous media approach. This paper describes aMarkov Chain Monte Carlo (MCMC) conditioningtechnique for reservoir fracture modelling bytaking into account the seismic events collectedduring the fracture stimulation process. Using thetechnique, the fracture model “evolves” during thesimulation process and eventually converges to apredefined optimal criterion. The proposedmethod is tested using seismic data collectedduring the hydraulic fracture stimulationprocesses of the Habanero wells in Geodynamics’Cooper Basin project.Keywords: Fracture network, seismic events,Markov chain Monte Carlo.IntroductionThe technical and commercial viability of HDRgeothermal energy depends on the creation ofartificial reservoirs, or Enhanced GeothermalSystems (EGS), in the rock mass by stimulatingand creating fractures (generally by hydrofracturing)to enable geothermal flow. Theartificial reservoir forms the critical component inan EGS: the fracture network that connects theinjection and production wells and acts as theheat exchange chamber for the system. HDRproductivity depends crucially on theconnectivity/permeability of the reservoir fracturenetwork and a realistic fracture model, such asthat described in this paper, is the key toassessing reservoir performance and designing asuitable heat exchange chamber for the EGS.The characterisation of rock fracture networks is avery difficult problem not least because accuratefield measurement of a single fracture is difficultand measurement of all fractures is impossible.Thus, in practice, the whole fracture system is notobservable on any meaningful scale and the onlyrealistic approach is via a stochastic modelinformed by sparse data and/or by analogues. InHDR applications, a realistic solution is even moredifficult as the only reference data related to thefracture system are either from geophysicalborehole logs and/or sparse seismic eventskilometres beneath the surface and detectedduring the hydraulic stimulation process. Forthese reasons current models of fracture systemsused for HDR flow modelling are oversimplifiedrepresentations of reality. They either use anequivalent porous media approach (e.g., Xing etal. 2009), single fracture representation (e.g.,Zhang et al., 2009) or a combination of both.Stochastic fracture modelling is the generalapproach in which locations, size, orientation andother properties of fractures are treated asrandom variables with inferred probabilitydistributions. In the simplest case, once theparameters of the distributions are inferred, therock fracture model is constructed by Monte Carlosimulation. First, the fracture locations aregenerated, usually by a Poisson distribution inwhich fracture intensity for a particular area iseither assumed to be constant or is derived fromgeostatistical estimation or simulation. Secondly,the orientation of each fracture is generated, mostcommonly from a Fisher distribution. Finally, thesize of each fracture is generated from a specifieddistribution, the most common being exponential,lognormal or gamma. Other fracture properties,such as aperture width and fracture strength, canthen be added into the network by additionalMonte Carlo steps. Options for fractureintersections and fracture termination can also beincorporated. Simulated fracture models areusually validated by sampling the model (usingscan lines or areas) and assessing the extent towhich the sampled values conform to thestatistical models specified (Kulatilake et al.2003). Such models are certainly very useful indescribing the statistical behaviours of the fracturesystem. However, in order to make the generatedmodels more realistic, some forms of dataconditioning must be built into the fracturegeneration process. Data conditioning in twodimensionalfracture trace simulation is generallyconsidered a simple matter but, probably becauseof the complexity involved, there are very fewpublications of algorithms and methods for dataconditioning in three-dimensional stochasticfracture modelling. Mardia et. al (2007) describedan attempt to condition a fracture model byborehole intersection data using Markov ChainMonte Carlo (MCMC) simulation. In this paper,however, the conditioning data are the seismicevents observed during the fracture stimulationprocess of the HDR reservoir and they are not,therefore, confined to any known order (e.g.,boreholes).During hydraulic stimulation, fractureslips/initiations/propagations produce micro-98

Australian Geothermal Conference 2010seismic events that can be monitored by anetwork of geophones and analysed to obtain theevent locations. To date, these “point cloud” datahave been used to estimate the volume of areservoir that is connected by wells. We believethese micro-seismic events not only identify thelocations of fracture slip, initiation and/orpropagation but also contain vital informationabout the structure of existing fractures andfracture networks. Successful extraction of thisinformation will significantly improve the reliabilityof fracture models so that a more realistic fracturerepresentation of the HDR reservoir can beobtained. Moriya et al. (2002) use microseismicmultiplet analysis to derive the fracture plane onthe assumption that seismic events from the samefracture will produce similar microseismicities.This approach, however, only accounts for a smallproportion of the total seismic events (17% inMoriya et al. 2002) and the fundamentalassumption behind the approach is questionable.AssumptionIt is generally acknowledged that during thehydraulic fracture stimulation process, theeffective normal compressive stress acting acrosstwo fracture planes of a fracture is reduced due tothe hydraulic pressure and that will cause the twoplanes to slip against each other. This isgenerally considered to be the key mechanism increating a permeable HDR reservoir as shearslipswill result in mis-alignment of fracturesurface topographies which will cause a lateraldilation and thereby enhance fracture aperturesand significantly increase the permeability of thefracture network. The shear slips betweenfracture planes will produce small-scale seismicitywhose seismic waves can be captured andanalysed by a network of geophones to derive thelocation of the event (Baisch et al. 2006). Theeffect of hydraulic pressure also makes it possiblefor existing fractures to propagate in the reservoirand new fractures to be initiated. The finaloutcome of the process will be a permeablereservoir connected to the wells through acomplex fracture network.Based on this conceptual description of thefracture stimulation process, it is reasonable toassume that seismic events only occur on fractureplanes. The criterion of a realistic fracture modelis then directly related to the overall closeness ofthe seismic events to fracture planes fitted in themodel. In this context, fracture simulationessentially becomes a stochastic geometryreconstruction problem given a set of pointclouds. Reconstruction of a surface from randompoint clouds is computationally and algorithmicallychallenging and is an active research area incomputer and mathematical sciences (Bercovieret al. 2002). The success of current practice,however, depends critically on close samplingpoints on the surface, which is usually not anissue as the point clouds are generally obtainedfrom laser scanning or some form of digitizing.For seismic point clouds in geothermalapplications, however, samples are very sparse.For a given fracture, only a few points areavailable, which indicate either the propagationfront of the fracture or a point on the fracturesurface where shear slip occurs at the time theevents are detected. Current methodology is thusnot directly applicable to fracture modelling.The most common approach in stochastic fracturemodelling is to use a 3D plane to represent afracture, which could be bounded (e.g., ellipticalplane, polygonal plane) or unbounded (e.g.,Poisson plane). The fracture model (network)then becomes a series of connected fractureplanes, F i , i=1,2…n, where n is the total numberof fractures. A seismic event point, P j , j=1,2..m (mis the total number of seismic event points) canthen be associated with a fracture F i with distanced j:i .m2d j : ij1can then be used as a simplecriterion to quantify the goodness of fit of the fittedfracture model.MCMC modelPlanar polygons are used to represent fractures inthis research. For each fracture polygon F i , thelocation is described by the coordinates of itscentre point (x i , y i , z i ), the orientations aredescribed by three angles: dip direction i , dipangle i and rotation angle i and the sizes of thefractures are described by a major axis a i and aminor axis b i of an ellipse containing the polygon(Xu & Dowd, 2010). In other words, weparameterize fracture planes with parameters (x i ,y i , z i , i , i , i , a i , b i ), i=1,2…n. Fracture plane F ican also be expressed in functional form as x y z ( i)( i)( i)( i)x y z( i)( i)( i)x, y, z)( whereis the unit vector normal to thefracture plane and can be calculated from ( i , i , i ). Given a point P j (x j , y j , z j ), not necessarilylying on the plane, the signed orthogonal distanceto fracture F i is defined by:dj:iwhereii( ) x ( ) yx( FP ijjyotherwise)ji ( ) zzji ()ifP( Fi )j F(1)is the projection point of P j onfracture plane F i . A matching function(j){1,2,…,n} is used to associate each point P jwith one and only one fracture polygon. We shalli99

Australian Geothermal Conference 2010impose a simple criterion of minimum distance forthis association and therefore by writing d j:i wemean point P j is associated with fracture polygonF i as its distance calculated by Equation (1) is theminimum when compared with distances to anyother fracture plane.d={d j:i }j is then the complete set of projectiondistances. Since seismic event points that lie onthe same fracture will not be exactly co-planar, wemust allow for statistical variation and treat afracture as a distorted version of an idealizedplane. A simple Gaussian noise model can thenbe adopted:d j i2:~ N(0, )(2)to represent the distortion in the data from theidealized plane. In other words, the orthogonaldistances are identically and independentlynormally distributed with mean zero and variance 2 . Therefore, the likelihood function for the set ofseismic event points P={P j } can be defined as:L2m ( d j:i )m 22( P;F,ej1) 1 2 (3)given a set of fractures F={F i } and the matchingfunction (.). The product of this likelihood withpriors of F gives the posterior distribution for Fand the attention now is to estimate the posteriordistributions of F and hence the parameters of thefracture set. A Markov chain can be used togenerate samples from the posterior distributionwhich is commonly constructed by the Metropolis-Hastings algorithm using the Monte Carloacceptance/rejection technique imposed by theHastings’ ratio:( t )( t ) ( P)q(F | P) F , P min1,( t )( t) (4) ( F ) q(P | F ) where (.) is the posterior (target) distribution andq(.) is a transition kernel which is usually chosenso that it is easy to sample from.A more detailed description of this model can befound in Mardia et al (2007) where a similar modelwas developed to generate fracture modelsconditional on intersection points betweenfractures and boreholes drilled on a regular grid.The Habanero Point CloudThe point cloud used in this study is the Q-Conprocessed dataset of locations of seismic eventsrecorded during the hydraulic fracture stimulationof Habanero 1 between November 6 andDecember 22, 2003 (Weidler, 2005). A total of23,232 seismic events are recorded in thisdataset which covers an area roughly of 2.5 km 2 .Figure 1 shows the absolute hypocentre locationsof these events.Figure 1 Absolute hypocenter locationsof the seismic eventsClearly an ideal sub-horizontal reservoir has beenformed by the stimulation process which is gentlydipping in the South-west direction. Earlyanalysis has revealed that the major part of thereservoir is confined within a sub-horizontal layerof approximately 30m (Baisch et al. 2006). Figure2 shows ratios of increments in the volume of thereservoir, the geographical extents in thehorizontal, East-west and North-south verticalplanes covered by the seismic events during thestimulation period. Note that the increments areplotted on a relative scale where the ordinaterepresents the ratios against the volume orgeographical extents of the reservoir covered byseismic events on November 6, 2003.250200150100500HorizontalVolumeNSEWDays after Nov. 6 20030 10 20 30 40 50Figure 2 Increments in volume and geographicalextents covered by seismic eventsHabanero Fracture ModelA total of 20052 fractures were initially generatedbased on a non-homogeneous point process witha non-parametric density model estimated by thepoint cloud shown in Figure 1. Fracture locations(x i , y i , z i ), i=1,…,20052, are generated. Thefollowing parameters are used for initial fracturegeneration:100

Fracture orientation: Fisher distribution with=1. Orientation parameters ( i , i , i ) arethen calculated.Fracture size: lognormal distribution withmean = 80m and variance = 12,000 m 2 .Australian Geothermal Conference 2010 100064MCMC method described was then applied toupdate the fracture network. The followingtransition kernels are used in the MCMC process: Fracture orientation ( i , i , i ): normalproposal with standard deviation of 0.1 (inradian). Fracture location (x i , y i , z i ): normal proposalwith standard deviation of 1.0.Fracture size is not optimised in the current trial.Fractures without any association with any pointafter 100,000 iterations were removed from thesystem. After 2M iterations, the fracture modelobtained is shown in Figure 3.20864-5 -3 -1 1 3 5(a) initial model 100020-0.4 -0.24 -0.08 0.08 0.24 0.4(b) optimised modelFigure 4 Histogram of dj:i for intial and optimised fracturemodelsFigure 3 Habanero fracture modelafter 2M MCMC stepsThis fracture model includes 10,995 fractures.Figure 4 shows the distribution of d j:i before andafter the application of the MCMC updatingprocess. Comparison of statistics is given inTable 1. Clearly the reliability of the model hasbeen significantly improved.Orientations of fractures have also changedsignificantly. There is no clear indication oforientation preference in the original fracturemodel (see the lower hemispherical projections ofpoles of fracture planes in Figure 5a). Thefracture pole plot after optimisation (Figure 5b)clearly demonstrates a great proportion of subhorizontalfractures. This is encouraging as itagrees well with the propagation pattern of theseismic event point cloud observed duringstimulation (Baisch et al. 2006).Table 1 Comparison of statistics of dj:ibefore and after MCMC optimisationInitialmodelOptimisedmodelNumber of fractures 20052 10995Minimum d j:i -116 -0.35Maximum d j:i 303 0.39Mean value of d j:i 0 0Variance of d j:i 12.5 0.0028d 2 j:i291437 6699% of d j:i within range [-11, 34] [-0.17, 0.17]101

Australian Geothermal Conference 2010signals (Baisch et al. 2009). The hydraulicconductance between wells can then beestimated.(a) initial model(b) optimised modelFigure 5 Hemispherical projections of poles of fractureplanesConclusionsThis is a preliminary trial to demonstrate theapplication of MCMC in fitting a fracture modelusing a point cloud dataset. Clearly the method iseffective, however there are still many remainingchallenging issues. Fracture join/split (Mardia etal. 2007) is not yet considered in this work. Theeffect of the initial point process model has notbeen assessed and this will affect the totalnumber of fractures retained in the final fracturenetwork. The illustrated fracture model is by nomeans the ultimate optimal model for theHabanero fractured reservoir and furtherinvestigation is needed to achieve such a model.The next stage of the process is to assess theconnectivity of the fracture network between wells(e.g., between Habanero 1 and 3). Hydraulicapertures of fractures can be estimated from thedegree of seismicity recorded in the seismicReferencesBaisch S., Weidler, R., Vörös R., Wyborn D. Andde Graaf, L. (2006) Induced seismicity during thestimulation of a geothermal HFR reservoir in theCooper Basin, Australia, Bulletin of theSeismological Society of America, 96, No. 6,2242–2256.Baisch S., Vörös R., Weidler R. and Wyborn D.(2009) Investigation of fault mechanisms duringgeothermal reservoir stimulation experiments inthe Cooper Basin, Australia, Bulletin of theSeismological Society of America, 99, No. 1, 148–158.Bercovier, M., Luzon, M., Pavlov, E. (2002)Detecting planar patches in an unorganized set ofpoints in space, Advances in ComputationalMathematics, 17, 153 – 166.Dowd, P.A., Martin, J.A., Xu, C., Fowell, R.J. andMardia, K.V. (2009) A three-dimensional data setof the fracture network of a granite. InternationalJournal of Rock Mechanics and Mining Sciences,46, 5, pp.811-818.Dowd, P. A., Xu, C., Mardia, K. V. and Fowell, R.J. (2007) A comparison of methods for thesimulation of rock fractures, MathematicalGeology, 39, 697 – 714.Kulatilake, P. H. S. W., Um, J., Wang, M.,Escandon, R. F. and Narvaiz, J. (2003),Stochastic fracture geometry modelling in 3Dincluding validations for part of Arrowhead EastTunnel, California, USA, Engineering Geology,70, 131-155.Mardia, K. V., Nyirongo, V. B., Walder, A. N., Xu,C., Dowd, P. A., Fowell, R. J., Kent, J. T. (2007)Markov Chain Monte Carlo implementation of rockfracture modelling, Mathematical Geology, 39,355 – 381.Mardia, K. V., Walder, A. N., Xu, C., Dowd, P.A.,Fowell, R. J., Nyirongo, V.B., Kent, J. T. (2007a),A line finding assignment problem and rockfracture modelling. Bayesian Statistics and itsApplications, Eds. Upadhaya, S.K., Singh, U.,Dey, D.K. Anamaya Pubs, New Delhi; ISBN 1081-88342-63-7; ISBN 13 978-81-88342-63-1. 319-330.Moriya, H., Nakazato, K., Niitsuma, H. And Baria,R. (2002) Detailed fracture system of the Soultzsous-ForêtsHDR field evaluated usingmicroseismic multiplet analysis, Pure and AppliedGeophysics, 159, 17-541.Weidler, R. (2005), The Cooper Basin HFRproject 2003/2004: findings, achievements andimplications, Technical report, Geodynamics Ltd.Xing, H., Zhang, J., Liu, Y. and Mulhaus, H.(University of Queensland, 2009) Enhancedgeothermal reservoir simulation, Proceedings ofthe Australian Geothermal Energy Conference2009, Brisbane.102

Australian Geothermal Conference 2010Xu, C. and Dowd, P. A. (2010), A new computercode for discrete fracture network modelling,Comps & Geosc., 36, 292-301.Zhang, X. Jeffrey, R. and Wu, B. (CSIRO, 2009)Two basic problems for hot dry rock reservoirstimulation and production, Proceedings of theAustralian Geothermal Energy Conference 2009,Brisbane.103

Progress in CO 2 -based EGSAustralian Geothermal Conference 2010Aleks D. Atrens 1 *, Hal Gurgenci 1 , Victor Rudolph 11The Queensland Geothermal Energy centre of Excellence.Mansergh Shaw Building (#45),The School of Mechanical & Mining Engineering, The University of Queensland, QLD 4072The concept of Engine ering Geothermal Systems(EGS) usin g CO 2 a s a he at extractio n fluid ha sbeen exp anded sin ce it was first p roposed. Abetter the oretical un derstanding is av ailable fo rthe be haviour in the reservoir, process de sign,corrosion & co ndensation i ssues, and ro ckgeochemistry, although knowl edge of som e ofthese systems i s i ncomplete. S ufficientunderstanding is now available to provide a powerplant design given site and reservoir parameters,and viability ca n be e stimated. A numbe r ofchallenges re main, h owever, a nd th ese mu st b eaddressed for the concept to be realisable. Thosechallenges a re contextua lised h ere to provid efocus for the most urgent issues to be addressed.Keywords: Carbon dioxide, CO 2 , EGS, enhancedgeothermal systems, thermosiphon, hot dry rock,hot wet rockIntroductionCO 2 -based EGS have been discussed previously(Brown 2000; Pruess 2 006; Pruess an d Azaroual2006; Gurgenci et al. 2008; Pruess 20 08; Atrenset al. 20 09a; Atren s et al. 2009 b; Atrens et al.2010). A full discussio n of the technical andeconomic issues related to CO 2 -based EGS havebeen reported (Gu rgenci 2009). Many of thoseissues a re broadly applicable to all E GS; thosewill not be discussed here. There are a number ofissues that are unique to CO 2 -based EGS. T hesehave prim arily been la ck of unde rstanding of theprocess design, perfo rmance in th e reservoi r,CO 2 -H 2 O interactions, CO 2 -H 2 O-rock interactions,thermodynamic performance, etc. Som e of theseissues have been ad dressed in varyin g levels ofdetail.Many tech nical challe nges for the concept toovercome re main. Th ese ca n b e g rouped undertwo bro ad ca tegories: gen eral challe nges, whi chare overarching an d ap ply to any u se of CO 2 inEGS, and reservoir challenges, which are specificto the reservoir type.General ChallengesThere a re a number of challenges tha t must b eaddressed to make CO 2 -based EGS viable. Eachof these is effectivel y sh ow-stopping if th eassociated issues are unresolvableThe ove rarching challe nges that must b eaddressed are: CO2 Sourcing CO2 turbomachinery Solubility and re activity of roc kcompounds and elements in pure or nearpureCO 2Legal/regulatory issues with sub surfaceCO 2 .Reservoir ChallengesBeyond the overa rching issue s, the re a re anumber of issue s that depen d on th e type ofreservoir b eing exami ned. The individualchallenges f or the different re servoir types arediscussed below.Un-dryable ReservoirsThis term is used to group reservoirs that are wellhydraulically connected, a nd initially sa turated oroversaturated with water. This g rouping wouldinclude h ot sedimenta ry a quifer systems (HSA),and EGS project s su ch as in the Cooper Ba sinwhere hig h pre ssure implies l ong-distanceconnectivity. These type of systems have anassociated challenge of extreme di fficulty increating a d ry CO 2 -rich operating zone within thereservoir, eit her due to t he large energy inp utnecessary to displa ce wa ter, or the difficulty incontaining CO 2 within the desired region.Dryable ReservoirsDryable reservoirs signify a re servoir grouping ofthose which are under-saturated with water at theoutset, or have a limited initial volume of waterthat is ea sily displa ceable. Examples of thiscategory are the Fenton Hill reservoir, which wasacknowledged a s b eing un connected an drelatively lacking in water content, o r the Hijorireservoir system, which experie nced signifi cantlosses of injected fluid into the reservoir formation(implying an ease in displ acing existing reservoirfluids). The challenges associated with systems ofthis type are the time period req uired to dry thereservoir, an d the p rofile over time o f fluid los sinto the reservoir formation.Wet ReservoirThis category is used for systems that are have aliquid water pha se coexi sting with CO 2 -rich fluidnear the reservoir o perating zo ne. CO 2 -rich fluidmay be prod uced in this ca se, allowing a single -loop, but there are important issues of both waterand carbon dioxide being present in the reservoir.The challen ges a ssociated with this type of104

system are related to the water-CO 2 relationships:time re quired until sufficie ntly pure CO 2 can berecovered, CO 2 -H 2 O dyn amics i n the reservoi r(horizontally and verti cally), rock-water-CO 2interactions including mi neral t ransformation /dissolution / deposition (permeability change) andgeo-sequestration, and the thermodynamic effectsof high water content in the CO 2 -rich phase.CO 2 -Rich ReservoirThere is also the po ssibility of operatin g a CO 2 -based EGS system i n a reservoir th at is al readyCO 2 -rich. Th is co uld b e a naturally occurringreservoir, or artificial as the result of a rtificial CO 2injection for rea sons o ther than CO 2 -basedgeothermal power ge neration. The challen geassociated with this re servoir type is id entifying areservoir that contains or may in the future containCO 2 , and is also suitable for geothermal powe rgeneration.DiscussionThere is a continuum between the three form erreservoir types, an d so me of the challengesrelating to CO 2-H 2 O interactio ns overla p.Separating reservoirs into the catego riesdiscussed a bove cl arifies the range of theseinteractions that are vital to co nsider for differentsituations. Th e implicatio n of this is that there isan opportunity to target th e proposed CO 2 -basedEGS technol ogy towards situation s whe re th echallenges are more readil y overcome. This maybe useful where there are particular issues relatedto certai n systems that may not be feasibl yanswered i n the fore seeable future, o r that maybe m ost suitably be resolved by field trials in areservoir of a different type.It is recom mended th at cu rrent goals to wardsimplementation of a CO 2-based EG S p owersystem fo cus on re servoirs that ca n fe asibly bedried, as th ese have a sm aller nu mber ofassociated challenges. Reservoirs of this type arealso le ss li kely to be targeted for use for wate r-based EGS, so the re is le ss risk of co mpetitivelydisadvantaging one technology over another.Australian Geothermal Conference 2010ReferencesAtrens, A. D. , H. Gurge nci, et al., 2009a., "CO2thermosiphon for competiti ve powe r ge neration."Energy & Fuels 23(1): 553-557.Atrens, A. D., H. Gurgenci, et al., 2 009b., Exergyanalysis of a CO2 the rmosiphon. T hirty-fourthworkshop on geoth ermal re servoir e ngineering.Stanford University, Stanford, California.Atrens, A. D., H. Gurgenci, et a l., 2010. ,"Electricity g eneration u sing a carbon-dioxidethermosiphon." Geothermics.Brown, D. W., 2000., A h ot dry rock g eothermalenergy concept utilizing supercritical CO2 insteadof water. Twenty-Fifth Wo rkshop on G eothermalReservoir Enginee ring, Stanford University,Stanford, California.Gurgenci, H., 2009., Electricity generation using asupercritical CO2 geothermal siphon. Thirty-fourthworkshop on geoth ermal re servoir e ngineering.Stanford University, Stanford, California.Gurgenci, H., V. Rudolph, et al., 2008.,Challenges for ge othermal en ergy utilisation .Thirty-Third Workshop on Geothe rmal Reservoi rEngineering. Stanford Un iversity, Stanford,California.Pruess, K. 2 006. "Enhanced geothermal systems(EGS) using CO 2 a s working fluid —A nove lapproach for gene rating renewable e nergy withsimultaneous seq uestration of ca rbon."Geothermics 35: 351–367.Pruess, K., 2008., "On production be haviour ofenhanced g eothermal systems with CO2 a sworking fluid." Energ y Conversi on andManagement 49: 1446-1454.Pruess, K. and M. A zaroual, 2006., On th efeasibility of usi ng supercritical CO2 as heattransmission fluid in an e ngineered ho t dry roc kgeothermal system. Thi rty-first workshop o ngeothermal reservoir engine ering. StanfordUniversity, Stanford, California.105

Australian Geothermal Conference 2010Exploration and assessment of Hot Sedimentary Aquifer (HSA)geothermal resources in the Otway Basin VictoriaPeter Barnett and Tim EvansHot Rock Limited, GPO Box 216, Brisbane QLD 4001peter.barnett@hotrockltd.com and tim.evans@hotrockltd.comThe Otway Basin contains a n umber of Ho tSedimentary Aquifer (HSA) geothe rmal resourceswith p otential for commercial development. T hemethodology Hot Rock Limited has u sed toassess the se resource s i s detailed in t his pa per.The re sulting stored he at estimate s for the HSAresources of the Otway Basin i s larg e, even byworld geothermal standards. A total of 550,000PJ of HSA resou rces have been so fa r estimatedand declared by three companies.While there is confid ence in the estimation of inplaceheat in the Otway Basin HSA resources andthe rate that heat in geothermal water recoveredto surface can be converted to electricity, there isless ce rtainty as yet on the amount of heat thatcan b e re covered a nd the rate it can berecovered. Both of these variables are determinedby permeability in the geothermal reservoir.By comp arison with bot h HSA a nd volcani cgeothermal systems el sewhere, and a detail edassessment of the structural geology at the OtwayBasin tenements, HRL c onsiders th at the HSAresources i n the Otway Basin are do minated bysecondary fracture perm eability, with primarypermeability being subordinate. HRL i sproceeding to drill in early 201 0 two pro of ofconcept w ells in to its flag ship H SA project a tKoroit. Th ese wells a re specifically targeted onboth primary and secondary permeability, with thefirm expectat ion that se condary pe rmeability willprove to be dominant and that comm ercial gradetemperatures and well flow rates will be obtained.Keywords:Geothermal, Otway Basin, Hot S edimentaryAquifer, HSA, resource assessment, Monte Carlo,proof of concept, gen eration potential, welltargeting.IntroductionHot Rock Limited (HRL) holds a large geothermaltenement position in Vi ctoria of 2 7,500km 2 in fivepermits. Some 15,000km 2 of this holding containsrocks of t he Crayfish S ubgroup i n th e on shoreportion of the Otway Ba sin within Vi ctoria. Theseare predominantly sandstones of Creta ceous ageand have known geothermal potential.They occur in a series of individual depositionarycentres (shown coloured in both ligh t and da rkgreen in Fig ure 1 ), dep osited o nto the tops o fsubsiding basement fault blocks developed alongthe northe rn margin of the rift zone whi ch led toAustralia sep arating fro m Antarctica durin g thebreakup of Gondwanaland in late Cretaceoustimes.The crustal t hinning prod uced in thi s extensionaltectonic environment allows for elevated heat flowthrough the Palaeo zoic ba sement into th eoverlying sediments over most of the Otway Basinwhere geothermal gradients of 35 to 45 o C/km aremeasured. This combination of re gional sh allowheat and larg e areal exten t, thickness and hencevolume of geothermal re servoir ro ck make s th eOtway Basin a significant terrain for potential HSAgeothermal develo pment, even by worldstandards.Figure 1: Series of fault bound rift basins containing thickaccumulations of Crayfish subgroup rocks developed alongthe northern margin of the onshore Otway Basin.Exploration of HSA resourcesAlthough the HSA geothermal plays in t he OtwayBasins a re esse ntially blind – i .e. withou tobservable discharge of h eat or ma ss at surface,HRL h as be en abl e to define the se resource s atdepth from t he exten sive geoscientific data thathave been acquired over the past 40 t o 50 yea rsby parti cularly the petroleum a nd g roundwaterindustries in Vic toria. Th ese d ata have b eensystematically archived by State an d Fe deralgovernment agencies such as the Department ofPrimary Industries and Geosciences Australia andare readily accessibleExisting data includ e g ravity, magnetics an dstratigraphy whi ch are applicable at relativelycoarse scal e to regio nal scale studies. Mo re106

Australian Geothermal Conference 2010detailed data, of value for prospect scale studies,include seismic refle ction data, geol ogical log sand co res from existing petrole um a nd groundwater wells, and mea surements of temperaturesand other data in these wells.In additio n t o u se of ex isting data, HRL h asapplied to it s HSA expl oration stu dies in theOtway Ba sin a n umber of exploratio n meth odsthat are rout inely use d in the exploration an ddelineation of volcanic geothermal systems. Themost significant of th ese ha s been t he u se ofmagneto-telluric (MT) resistivity surveying and theapplication of ge othermal geochemistryinterpretative methods t o grou nd water a ndpetroleum well fluids analyses.Key deliverables from the HRL exploration studiesinclude the following:Starting Point: An initial review of geothermalgradients a s a tool for prioriti zing areas fordetailed geothermal assessment.End Point: A rang e of 3-D mod els gridded attypically 100 m cente rs ov er ge othermal resourceareas of typically 3 00 to 500 km 2 areal extentwhich include: Stratigraphic model s defining fo rmation top s,bottoms, vert ical (cro ss sections) and hori zontalsections (slices in plan) Structural geological models Thermal m odels showing di stribution o fisotherms in 3-D, i sothermal sectio ns in 2-Dcombined with geology, estimates of temperaturein each grid block at any depth Multi discipli ne, integ rated, co nceptual hydrogeological models which form the basi s fo rsubsequent resource assessm ent, an dformulation of exp loration drilling anddevelopment strategy.Assessment of HSA ResourcesThe Otway Basin HSA resources have beenassessed in terms of in-place stored heatcontained in the Crayfish subgroup reservoirrocks between assigned upper and lower depthlimits. This involves initially considerations ofresource thickness, lateral extent and hencevolume, followed by considerations of porosityand temperature from which heat contained inboth the rocks in the resource volume and hotwater contained in pores and factures in thereservoir rocks can be estimated.The upper depth limit to the resource is referred toas the resource temperature cut off (orabandonment temperature) and the lower depthlimit is an assumed depth limit to whichgeothermal fluids can be reasonably recoveredfrom the resource, and is taken to be a practicaldrilling depth limitation.Estimates of temperature values throughout theresource volume have been obtained from 3Dthermal modelling. Porosity has been assessedconservatively at 10% throughout and althoughthis in important parameter in considering theamount and rate at which hot water can berecovered from the geothermal resource, theamount of heat stored in the rock and the water isrelatively insensitive to porosity.The values used for the above parameters inHRL’s assessment of HSA resources are given inTable 1 where they are compared with a range ofvalues used by other developers working in theHSA environment. Relative to these othervalues, those used by HRL are conservative.Table 1: Comparison of parameters used by HRL forestimating the thickness, volume and in-place heat of HSAgeothermal resources.ParameterMinimum resourcetemperaturePower plant rejectiontemperatureResource base: i.e.practical depth limit forrecovery of geothermalfluidsValues used byHRL130 o C (GEP6 and 23) 125 0 C140 o C (GEP-8)70 o C 70 o C4500mValues used byothers in OB5000mPorosity 10% 5 % to 15%Minimum reservoir 500mNil to 250 mthicknessThe computation of in-place heat has beenundertaken by HRL on a probabilistic basis usinga purpose developed Monte Carlo simulationmodel in which probability distributions are usedto characterise the key resource assessmentparameters. The stored heat is summed overindividual resource blocks for which formationtops, top and bottom resource limits and modelledtemperature distribution have been gridded at1000m intervals. The Monte Carlo model is run2,000 times to obtain the frequency distributionsand cumulative probability distributions for thestored heat contained in the geothermal resourcevolume.The results of the resource assessments arereported in accordance with the AustraliaGeothermal Reserve Reporting Code (1st edition,2008) in terms of in-place stored heat. At thisstage in the exploration of HRL’s Otway BasinHSA resources, two types of geothermalresources have been determined: “Inferred resource” – where tem peratures a renot directly measured, and107

Australian Geothermal Conference 2010 “Indicated resource” wh ere tempe ratures aremeasured in-situ within the geothermal resource.Typically th e form er is declared wheretemperatures are measured in petroleum and / ordeep ground water wells that have not penetratedas d eep a s the geothe rmal reservo ir in th eCrayfish Subgroup and the latter is where existingwells have been drilled into the Crayfish thu sallowing for temperatures to be measured directlywithin the geothermal resource. If a nd when wellsdrilled i nto the geot hermal re source a resuccessfully f low t ested then the “Inferredresource” estimate can be upgraded to “Measuredresource”.Results of Resource AssessmentsHRL has completed r esource assessments at thefollowing Crayfish Subgroup depostionary centresin the Victorian portion of the Otway Ba sin. Thesecentres are shown outlined in pink in Figure 1 andinclude (from west to east): GEP-23: Penola Trough GEP-6: Penola Trough GEP-6: Tantanoola Trough GEP-8: Koroit Trough GEP-8:Ross Creek Trough GEP-8/9: Elingamite TroughFrom these assessments a number of geothermalresources have been declared, as summarised inTable 2.Table 2: Results of estimations of Inferred and IndicatedHSA geothermal resources identified and declared by HRLwithin GEP-6, GEP-8 and GEP-23 in the Otway Basin.Resource Tenement ResourceAreaResourceVolumeIndicated InferredResource ResourceReportDatekm2 km3 PJ PJKoroit GEP‐8 50 47 7,600Koroit GEP‐8 400 340 59,000Koroit GEP‐8 450 387 7,600 59,000 1‐Oct‐09Penola GEP‐23 20 24 5,000Penola GEP‐23 120 150 29,000Penola GEP‐6 8 9 1,700Penola GEP‐6 290 306 55,000Penola GEP23 & 6 440 490 6,700 84,000 27‐Jul‐10Tantanoola GEP‐6 180 130 22,000 27‐Jul‐10Totals 1,070 1,010 14,300 165,000 180,000Power Generation PotentialThe estim ated geothe rmal resource s decla redtodate by HRL in the Otway Basin total 15,000 PJfor Indi cated re sources and P1 65,000 PJ fo rInferred re sources, at the P50 level. Thesefigures re present very la rge sto re of geothermalheat, however, not all of this h eat is av ailable forconversion to electricity. Only a small amount ofthe geothermal heat will be recoverable from theresource an d only a sma ll amount of the heatrecovered t o surface will be converted t oelectricity in a po wer plant due to t hermal tomechanical energy losses.Figure 2: “Inferred” and “Indicated” geothermal resourcesdeclared in the Koroit area, GEP-8 (Source: HRL, 2009).Representative value s for HSA geoth ermalresource ar e 5 % a nd 12%, r espectively, w hichmeans that o nly about 1% of the in-pl ace heat inthe geoth ermal re servoir will be converted t oelectricity. On this ba sis, only som e 1,000 PJ o fthe total value of 180,000 PJ of heat contained inthe Koroit g eothermal resource, would be able tobe extracte d and co nverted to electri city.Nonetheless, 1000 PJ of energy in electri calenergy term s is still a la rge value, equivalent toabout 300,000GWh. To achieve generation at thislevel a power plant with a gene ration capacity ofabout 1300MWe would be required to operate ata capacity factor of 90% for a period of 30 years.Figure 3: “Inferred” and “Indicated” geothermal resourcesdeclared in the Penola and Tantanoola Troughs in GEP-23and GEP-6 (Source: HRL, 2010)In addition to the 1 80,000 P J geothe rmalresources estimated by HRL, two othergeothermal develop ers have also deline ated,108

Australian Geothermal Conference 2010assessed and de clared HSA geothe rmalresources in the western and ea stern ends of theOtway Basin, on eithe r side of HRL’ s tenements.The total de claration of geothermal resources todate from th e three com panies for th e onshoreOtway Basin in both Vi ctoria and S outh Australiatotals some 550,000 PJ (Table 3 ). On the sam ebasis a s the cal culations above for Koroit, thi samount of t hermal en ergy has th e potential togenerate so me 9 00,000 GWh of electricity an dthis wo uld require a po wer plant g eneratingcapacity of some 3,500 MWe for 30 years.Table 3: Summary of all declared HSA geothermal resourceassessments over the greater Otway Basin4. The pin k line i s a regression fit f or channelsands in the Pretty Hill and the red li ne is thesame for bar sand s in the Pretty Hi ll. Clea nchannel sands are the prime target for geothermaldevelopment being more coarse grained and withboth highe r poro sity and perme ability at anydepth.As well as variation in pri mary permeability in theOtway Ba sin with d epth t here is also significantvariation in t he character and q uality of the rCrayfish Subgroup reservoir facies, depending onprovenance of sou rce m aterial a nd e nvironmentof deposition.Trough inOtway BasinTenement DeveloperHSAInferredResourcePJHSAIndicatedResourcePJHSAMeasuredResourcePJHSAEstimatedTotalPJTantanoola GEP‐6 HRL 77,000 1,700 78,700Koroit GEP‐8 HRL 59,000 7,600 66,600Penola GEP‐23 HRL 29,000 5,000 34,000Anglesea GEP10 GreenEarth40,000 40,000Tantanoola GEL‐170, Panax171, 172130,000 130,000Rendelsham GEL‐170, Panax184, 21217,000 17,000Rivoli St Clair GEL‐173 Panax 53,000 53,000Penola GEL‐223 Panax 89,000 32,000 11,000 132,000Totals 494,000 46,300 11,000 551,300The in-place heat stored in the HSA resources inthe Otway Basin a nd t he pote ntial gene ratingcapacity of these a re larg e, eve n by worldgeothermal standa rds con sidering that thegeothermal industry ha s a glo bal install edcapacity of 11,000 MWe.The confidence in the e stimation of h eat in-placein the Otw ay Bas in H SA r esources is r elativelyhigh, as i s the rate at which the heat ingeothermal water recove red to surfa ce can b econverted to electricity. Of lesser certai nty is theamount of h eat that can be re covered from theHSA re source and the rate at whi ch i t can berecovered - both of the se are dete rmined bypermeability in the geothermal reservoir.PermeabilityPermeability in Otway Ba sin HSA reservoirs canbe of two types (1) primary permeability,associated with natu ral porosity in th e CrayfishSubgroup sa ndstone re servoir ro cks, and/or (2)secondary permeability associated with faults andfractures in the reservoir rocks.Primary permeabilityThe petrol eum indu stry has generatedconsiderable data on poro sity and permeabilitythroughout th e Otway Basin from geolo gical an delectrical logging and a nalysis of core s andcuttings from petroleum wells. Poro-perm data forthe Pretty Hil l sandstone from petroleum wells i nboth Victo ria and S outh Au stralia (butpredominantly from the latter) are shown in FigureFigure 4: Summary of declared geothermal resourceassessments in the greater Otway Basin with HSAgeothermal development potential Source: 3DGeo (2009)From a d etailed analy sis of seismi c data andpetroleum well data, th e Koroit g eothermalresource is assessed to have pa rticularly go odreservoir facie s cha racteristics hav ing thic kaccumulations of cle an coa rse grained san ds,with mini mal lithic m aterial, espe cially in theLower Prett y Hill Form ation (3 DGeo 2009 ).Based on statistical analysis of th e poro-permdata in Figure 3, estimate s have be en of poro sityand permeability potential with de pth in the Koroitgeothermal resource (3D Geo, 200 9). The seestimates are for bette r t han average poro p ermvalues than occur for the Otway Basin.The variation in poro pe rm with de pth in Figure 4is instructive as it confirms: porosity in the Pretty Hi ll Formatio n to bereducing with depth, from 20% at 2500 m to 14%at 3500m, and permeability decreasing by two orders ofmagnitude between 2500 and 3500m.In spite of t hese re ductions i n po ro-perm withdepth, the evidence is that permeability within thePretty Hill at Koroit will still be sufficient to giv ecommercial geothermal well flows from prima ryreservoir p ermeability alone. Fo r example,assuming, as a conservative case, that the lowervalue in the range of the permeabilities predictedat each the 3 depth ranges in Figure 4 apply, anda net to gro ss ratio of 50%, then a tra nsmissivity109

Australian Geothermal Conference 2010of 13 Dm would be expected over the depth rangeof 2500 to 3 500m, with an expected temperaturerange of 130 to 17 0 o C For t his t ransmissivityvalue, a standard size geothermal well completion(with a 9-5/8 inch diameter open production hole)can be expected to yield a flow rate of 90 kg/sec,and for a large diameter well completion (with 12-1/2 inch di ameter production hole) to yield a flo wrate of 160 kg/sec (SKM, 2009).Secondary permeabilityThe importance of secondary permeability in HSAgeothermal system s se ems not to be wellappreciated in the Australia n geothe rmalcommunity. Few if any geothermal systems in theworld, of a ny type, prod uce from on ly prima rypermeability. For example, for the fol lowing twokey oversea s analo gues for HSA geothermalresources in Australia: Heber a nd East Me sa fields in S outhernCalifornia: secondary pe rmeability controls anupflow of h ot geothe rmal fluids from basementinto overlyin g fluvial sed iments where prima rypermeability i s hi gh. Ge othermal well s producefrom b oth f ractures around th e u pflow a ndprimary permeability in the sediments Southern Germany: high flow rates ( up to 140kg/sec) a re o btained from deep inte rsections (atdepths of u p to 4500m) of well bores withfractures in Malm lime stone with little if anyevidence of primary permeabilityIn co nventional vol canic geoth ermal syste msdeveloped in sedimentary reservoir rocks, fracturepermeability is al so t he major control ongeothermal fl uid hydrol ogy and well p roductivity.For example: The large and very well documented Geysersgeothermal f ield in California produces steamfrom fra ctured greywa cke baseme nt rocks whichhave low primary porosities and permeabilties. The Oh aaki geothe rmal fi eld in Ne w Z ealandhas go od produ ction well flows from fracture dgreywacke b asement rocks which under lie lavasand volcaniclastics.In the Otway Basin there are well developed faultsand structures at both large and small scale whichaffect both t he Pala eozoic ba sement rocks a ndthe overly ing sedim entary su ccessions,particularly the Crayfish Su b group. The presentdayst ress re gime i n the Otway Ba sin is on theboundary between o blique-slip an d reversefaulting. The NW–SE maximum ho rizontal stre ssorientation is cal culated at ~140 d egrees N.Analysis of structural ge ologic data has beeninterpreted t o sho w that faults and fractu resorientated approximately WNW-ESE to NE-S Ware consistent with the present day stress regimeand structures o n th ese o rientations areconsidered t o be m echanically o pen, with g oodpotential for providing secondary permeabilitychannels for geothe rmal fluid flow (3D-G eo,2009). The potential f or second ary fracturepermeability is therefore considered to be good inthe HSA resources of the Otway Basin.HRL is proceeding in early 2010 to drill two proofof concept wells at its fla gship Koroit geothermalproject. These two wells are specifically targetingmajor faults at depth as primary well targets. Asubordinate target is po rosity in the upper level sin the op en production h ole where p orosity andpermeability prosects are best. Having thehighest geot hermal gradi ent in the Ot way Basin(of up to 46o C/km) at the Koroit re source, aminimum production temperature of 130 o C can betargeted at t he top of th e ge othermal re servoirwithin the Crayfish Forma tion at depths of only2500m. At this comparatively shallow depth, poropermp rosects are m uch better than at greate rdepth, particularly belo w 3500m as evident fromFigure 4.HRL’s “Proof o f C oncept w ells” ar e th ereforedesigned to intersect an o ptimum combination ofprimary and secondary p ermeability and HRL i soptimistic of su ccessful outcomes in term s ofgood pe rmeability and transmi ssivity result s andcommercial grade tempe ratures a nd well flowrates.Figure 5: Geological section, Koroit HSA resource, showingproof of concept well targeted on (1) deep intercept withmajor fault target and (2) primary porosity target (upper andlower Pretty Hill Formations from 2500m) Source: 3DGeo(2009)Summary and ConclusionsThe Otway Basin contains a nu mber of HSAgeothermal resources wit h good p otential forcommercial development.These resources exi st as a result of th efavourable combination of: elevated geothermal g radients of t ypicallyaround 40 o C/km, but as high as 46 o C/km in primeresource areas such as Koroit in GEP-8. Thermal gradients in this range are sufficientto obtain com mercial grade geothe rmaltemperatures within practical drilling depths110

Australian Geothermal Conference 2010 The fortuitous occurrence of thick stratigraphicsequences of quart z ri ch sand stones in theCrayfish Subgroup reservoir rock within the samedepths that comm ercial temperatures occur an dwithin practical drilling reachThe stored heat contained in the HSA resourcesof the Otwa y Basin a re larg e, even by wo rldgeothermal standards. A total of 550,000 PJ o fHSA re sources have b een so far estimated a nddeclared by t hree comp anies. Altho ugh currentexpectations are that o nly some 0.5% of this inplaceheat can be recovered from the geothermalresources and converte d to ele ctricity, thegeneration p otential on e ven this ba sis i s la rge,requiring some 3,500MWe of g eneration capacityfor 30 years.There is confidence in the estimation of heat inplacein the Otway Basin HSA resources and therate that he at in geothe rmal wate r re covered tosurface can be converted to electricity. However,there i s less certainty yet on the amo unt of heatthat can b e recovered fro m the HSA resourcesand the rate at which this can be recovered - bothof these variables are determined by permeabilityin the geothermal reservoir.HSA resou rces se em to be viewe d by theAustralian g eothermal co mmunity as having o nlyprimary permeability. B y comparison with bothHSA and volcanic geothermal systems elsewhere,and a detai led a ssessment of the structuralgeology of its Otway Basin tenements, HRL viewsthe Otway Basin resources to be do minated bysecondary fracture perm eability, with primarypermeability being subordinate to this.HRL is proceeding in early 2010 to drill two proofof con cept wells at the Koroit HSA resource ,specifically targeting b oth primary and secondarypermeability with the firm expectation thatsecondary permeability will prove to be dominantand that commercial grade temperatures and wellflow rates will be obtained.Statement of Competent PersonThe inform ation in this pape r that relates t oExploration Results, Ge othermal Re sources orGeothermal Reserves i s based on i nformationcompiled by Peter Barn ett who is a full timeemployee of Hot Rock Limited. He has sufficientexperience which is relevant to the style and typeof geothermal play under consideration and to theactivity which he is un dertaking to q ualify as aCompetent Person as defined in the Code. In thiswork he ha s drawn fre ely from reports o n HSAgeothermal resources i n the Otway Basinprepared u nder his supe rvision, by b oth staff ofHot Ro ck Li mited and b y external consultants,notably 3D-Geo of Melbourne. The assessment ofstored heat and levels of Indicated an d Inferre dResources and the repo rting of these have beenundertaken solely by Pete r Barnett. Peter Barnettconsents in writing to the inclusion in the report ofthe matters based o n hi s i nformation in the fo rmand context in which it appears.AcknowledgmentsThis paper draws freely on graphical materialprepared by 3D Geo of Melbourne under contractto HRL, particularly the power point presentationfor this paper to be given at the conference.References3D Ge o Pty Limited, 200 9, Otway Basin Proj ect.Koroit Area WP2 Re port. Unpu blished re portprepared u nder contract to Hot Rock Limited2009.AGEA and AGEG, 200 8, Australi an Code fo rReporting of Exploration Re sults, G eothermalResource a nd Geoth ermal Re serves, 200 8Edition. Prep ared by the Australi an G eothermalCode Committee, AGEA and AGEG,Hot Ro ck Li mited, 2009, Statement of EstimatedGeothermal Resources for the K oroit GeothermalPower Proje ct, Victoria, Geothermal ExplorationPermit GEP- 8. Hot Rock Internal report, preparedby Peter Barnett, 30 September, 2009. A full copyis available at:http://www.hotrockltd.com/IRM/Company/ShowPage.aspx?CPID=1224&EID=16267720Hot Ro ck Limited, 2010, Characte rization andAssessment of Geothermal Resources in GEP-23& GEP-6, Ho t Rock Inte rnal re port, p repared byPeter Barnett, June 2010. A full copy available at:http://www.hotrockltd.com/IRM/Company/ShowPage.aspx?CPID=1227&EID=77703804Sinclair Knig ht Merz (SKM), 2009, several wellbore flo w mo delling stu dies. Unpubli shed rep ortprepared u nder contract to Hot Rock Limited2009.111

Australian Geothermal Conference 2010Validation of Geothermal Exploration Techniques for Analysing HotSedimentary Aquifer and Enhanced Geothermal Systems.J.E. Davey & J.T. Keetley3D-GEOLevel 5 / 114 Flinders Street, Melbourne, 3000.Validated g eothermal re source e stimates a rebased o n ge ological mo dels that combine th egeology fro m the surface to b asementincorporating prope rties such as lithology,structure, fluids, comp action, po rosity andpermeability with therm al models consisting ofheat flow, condu ctivity and therm al gradi ents.These models must extend beyond single pointwell an alysis, and provide a bal anced andgeologically rea sonable estimate of the size,structure, a ccessibility and m aturity of asubsurface t hermal system. To achieve ageologically rea sonable model, all availabl egeological and ge ophysical d ata m ust beconsidered. H owever, d ata se ts an dmeasurements mu st be analysed, interp retedand weighted in a manne r that reduces error forthermal models.This pa per focuse s o n building validatedgeothermal resource evaluations from geologicaland g eophysical i nterpretations f or HotSedimentary Aquifer (HSA) and EnhancedGeothermal Systems (E GS). Thi s p aper al sofocuses on limitations of some d ata sets, andthis can be realized when modeling.Limitations in Thermal ModelingFor a ny given geolo gical data set (field, well,seismic, rem ote sensing, temperature ) virtuallyan infinite numb er of 2D an d/or 3 Dinterpretations can be made. However, the set ofphysically possible geological interpretations arefar more lim ited. The tests for such physicalplausibility derive from our knowledge of theprocesses which co ntrol sedim ent deposition,structural ev olution, and geothe rmal system s.Interpretations which can pass thes e tes ts maynot be correct, but they are more li kely to be so,and can b e labeled with that highl y soughtappellation: Validated.Limitations occur with thermal modeli ng that arenot as preva lent in other f orms of su bsurfacemodeling. The main limitation i s the abundanceand re presentation of temperature th roughoutthe system. Tempe rature mea surements a retypically limited to single point measurem entsobtained from wells and bores previously drilled.While heat is the basis of all g eothermalsystems, temperature is only ever measuredfrom a singl e point. This provide s the verticalcomponent of a temperat ure g radient at thatpoint source ma king it a n isotro pic a nalysis oftemperature. While this is an ideal measurementfor volcanic and plutonic rocks where heatflow isprimarily in the vertical domain, it d oes notaccount for the a nisotropic ‘radia l’ heatdistribution t hat occu rs in se dimentary an dmetamorphic ro cks (Clau ser & Hu enges 1 995).The tempe rature-depth p rofile from a well cantherefore be thought of a s the vertical summingof the geoth ermal g radients e stablished withinthe individu al layers withi n the sedime nt pile atthat location. Such a profile is ra rely linear withdepth.However, when extra polating temp erature fromthe well p rofiles, the ani sotropic component ofheat distribution within a sedimentary basin mustbe a ccounted for. Be cause of thi s, th e therm alcomponent of the model is highly de pendent onincorporating ot her f actors such a s structure,lithological distrib ution, fluids, porosity andpermeability, hydrocarbo ns and fracturing tocontrol the d istribution of heat thro ughout themodel. Because of this, a geological model mustbe created first. This will help to generate the112

Australian Geothermal Conference 2010properties needed to consider for the anisotropicextrapolation of tempe rature. The rmal modelingwill be revisited later in this paper.Subsurface Geothermal ModelingUtilizing all t he availabl e data to construct ageological model will aid in validating ageothermal resource. In t errains currently beingexplored with Australia, available data includes:wells, 2D-3 D remotely sensed imag ery, 2D-3 Dseismic, potential field dat a sets (namely gravityand m agnetics), petro physical l ogs andoccasionally magnetotelluric data sets.WellsWells provi de geolo gical truthing to anysubsurface model. Ba sic informatio n su ch aslithology, formation de pth, fluid pre sence andtemperature are recorded.Because te mperature is currently o nlymeasurable in situ, wells provide the basis of thethermal mo deling. Do wn hole tools deployedduring drilli ng and after drilling a re used torecord temperature at different stag es within thewell. T his inform ation is co mbined withestimates or meas urements of conduc tivity togenerate information on heatflow. This is in turnused a s the ba sis for extra polating atemperature model across the area.Wells often h ave vital information colle cted fromdown hole logging tools employed during drillingor j ust after. These to ols collect info rmation o ntemperature, pressure, flow rate s, densities andeven fra cturing. A seld om use d, bu t highlybeneficial te chnique for E GS an d HSA is do wnhole vertical con ductivity and horizontalresistivity measurements. These techniques aredesigned to i dentify fractu re o rientation as wellas what orientation of fracture is open (that is, apotential flui d con duit) in the well. Bothmeasurements are o btained whil e drilling.Horizontal resistivity measures the resistivity of aformation by dispe rsing a cu rrent flowing in ahorizontal plane. Vertical cond uctivity ismeasured in a simil ar m anner, ho wever in th evertical plane. The se mea surements a reobtained by using wireline logging tools such aslaterologs a nd pro pagation logs. T wo surveytypes a re u sed, a s single o rientation pla nesurveys (i.e. vertical or horizontal) cannot takemeasurements in the plane parallel to t he planewhere th e e lectromagnetic current is flowing.Vertical cond uctivity is u sed for imagi ng hig hangle f racturing (f ractures with di p 20 -70 o ) an dsub-horizontal fracturi ng (dip 70 o ). Th e su rveys a re analysedand plotted as frequency diagrams indicating theprimary orientation of the open fracturing.Information about the contemporary stress fieldsin the area can be obtai ned f rom well data .Borehole bre akouts reflect a measu rement ofthe orientation of disintegration of the well duringdrilling as a result of com pressive stress failure.The orientation of the breakout is parallel to theorientation o f the minimum hori zontal stre ss.This o rientation is p erpendicular to theorientation of the d rilling ind uced tensil efractures, the orientation in whi ch the t angentialstress is below tensil e ro ck strength. Thisorientation is parallel to th e maximum horizontalstress. Information a bout co ntemporary stresscan assist with drilling and also with determiningthe nature a nd ori entation of any stru cture th atmay be active or ‘open’ within the system.Surficial G eology and Remotely SensedImageryGeothermal exploration i s greatly aid ing fromsurficial geology and re motely sen sed image ryas it p rovides approximate groun d truths tostructure an d morphol ogy interpreted in thesubsurface. Not all subsurface ge ology has asurficial rep resentation, a significant portiondoes. Actual exposure of these features helps tounderstand the magnitude and style of stru ctureand the physical properties of rocks at depth. Animportant co mponent of this is developing a nunderstanding of the cover rocks and how theymay aid in assisting or impeding the success ofthe geothermal resource. Regolith mapping andremotely sensed im agery such a s radi ometricsare being u sed to find a reas of surface fluid113

Australian Geothermal Conference 2010leakages a nd pote ntial conduit fa ulting in sealanalysis and reservoir integrity studies.Magnetotellurics and Potential Field SurveysMagnetotellurics (MT) and potential field surveyscan be u sed to d efine the ba sic li mits of areservoir. MT has been an industry standard forover 20 years for defin ing geothermal reservoi rlimits and boun daries. The ‘fuzzy’ imagingproduced b y MT can pick up resistivityanomalies typically a ssociated with flui d bodies,large scale structures, a nd cap rocks. Thi stechnique h as bee n used within Australia withvarying su ccess, whi ch is often a produ ct ofsurvey design as opposed to the reliability of thetechnique. M T modeli ng requires trial and e rrorof forwa rd m odeling (Ara ngo et al. 200 9). Thisprocess generates different models that could fitthe data, and trie s to a ssign th e m ost li kelymodel. Reliability of the MT is compromised as aresult of this . Ac cessory data r efines this err or;such the te chnique is better used in conjunctionwith other modeling techni ques. The mostimportant result from MT models is the ability toidentify the p resence of geothermal flu ids. Thi sis particularly useful in the identification of HSAreservoir systems. However, the recognition ofthe fluid is still limited to the erro rs definedabove.MT is comm only being combine d with potentialfield modelin g. Magnetic and gravity survey shave be en used with geothermal p rojects todefine large scale stru ctures an d al so as thebasis for d eterministic lithological varianceidentification sub surface. Potential fieldinversions are often co upled with p etrophysicallogs su ch as sonic and d ensity logs to create‘layer cake’ model s th at accumul ate to thepotential field signatu res produ ced by thesurveys. These te chniques provid eapproximately the same resolution of data as theMT tech niques do; h owever ca n be cheaper,and manipulated in different ways to account fordifferent geological scenarios.Seismic InterpretationsSeismic interpretation is not commonly used asa tool for EGS and HSA resou rce mod elevaluation. Seismic is typically used as a tool inthe petroleum industry to extrapolate li thologicalboundaries across l arge area s with theassistance of well s. F eatures in cluded o nseismic are: formation and st ratigraphicboundaries, sedimentary features such as onlapand down la p patterns, f acies and sequencestratigraphy, stru cture, an d tectono stratigraphicrelationships. These features are all essential forcreating balanced geological models.A benefit of seis mic interpretation is that 2D and3D seismic surveys a re availabl e from mo stbasin and foldbelt terrains in Australia as a resultof previou s petroleu m or acade mic wo rkconducted.Basic int erpretations can create t he basis of ageological m odel, an d can be t ruthed to othe raspects su ch as wells and pote ntial fieldsurveys.Advanced seismi c interpretation s can beconducted by pro cession of the seismic d atasuch that th e me chanical sig natures of theseismic t races a re tran sformed in particularways to hig hlight specific features (Nourollah etal. 2010 ). An example of this te chnique that isbeneficial in HSA syste ms is a p rocess calledNeural Network Inversio n Modeling (NNIM ).NNIM attemp ts to builds relation ships betwe endata sets si milar to the way the hu man b rainperceives an d analy ses d ata. The me chanicalsignature of a seismic wa velet is compared to aknown source, such as a well log, and particularfeatures characterized in the log are m atched totheir co rresponding sei smic sig nature. Oncerelationships are built b etween well log s an dseismic traces using well defined data sets, thismodel can b e applied to other sei smic tracesaway from th e known well locations to gene ratetarget log s from sei smic sections. This processcan be used to define sand bodies in a potentialreservoir, fluids such a s gas and water an dassociated m igration pathways, p otential se alsand different lithological patterns that were not114

Australian Geothermal Conference 2010obvious in the initial seismic interpretation and tohighlight structures. Oth er fo rms of sei smicprocessing can produce similar results.Integration of Geological InterpretationsThe integ ration of interp retations in geothermalsystems i s rare. E xploration techni ques areoften con sidered to be unique so urces ofinformation, rathe r than part of a dynamicsystem. However, this is far from the truth.The integrating of th e full suite of available dataconstraints, inclu ding 2 D/3D rem ote sen sing,field, 2D-3D seismic, an d welllogs/picks/production v alidates g eologicalmodels and provides the basis for a constrainedHSA and EGS model.Techniques such a s g ravity and magneti cinversion which are co nsidered to be lowresolution interpretations c an have theirresolution increased by using petrophysical logs,seismic interpretation s, a nd surface data tobetter constrain the layers in the model. Seismiccan also beti de to surfa ce feature s, pa rticularlystructure to help defin e attitudes an d confirmstructure. Th e combi nation of these te chniquesincreases the validation of the model.Revisiting Temperature modelingThermal ConductivityThermal conductivity of a ro ck determines th eefficiency (o r oth erwise) with which heat i spropagated through it. It is controlled by lithologyand state o f the ro ck (that is co mpaction,fracturing and meta morphic a ttributes).Accessory factors such as the presence of fluidssuch as water and gas also play a big role in this(Clauser & Hueng es 1995; Vasseur et al. 1995;Beardsmore 2004).Two primary methods are employed with respectto measuring conductivity: in situ measurementsconducted wi thin the sub surface usi ng loggingtools; a nd, la boratory m easurements preformedon small sections of rock. In situ measurementsare considered to be the best form of measuringconductivity as they are done so within theactual surroundings that the ro ck exist s within.This i ncorporates f actors such as f luids,fracturing, pressure and compaction better thana sa mple done in a l ab. In situ me asurementsalso represent an ave rage over a la rge area o frock (Clauser & Huenges 1995).Lab measured samples have a higher precision,because fa ctors inhibiti ng to the in situmeasurements su ch a s e quipment difficultie s,drilling inte rference a nd general control issue sare negated. This ma kes the measurementmore precise, however, not more accurate.In places wh ere th ere is no data avail able fo rthis modeli ng, default thermal con ductivities forend-member lithologie s that can bemathematically averag ed to reflect li thologicalmixtures as necessary, and so do not rely on thelaboratory m easurement of co nductivities o ndown-hole samples can be used. This has beenrecognized as a useful a nd reasonably accuratetool as it do es not hav e t he large extrapolatio nand e rrors th at are a ssociated with the pre ciselaboratory m easurements (Cla user & Huenges1995). Estimating the conductivity of a rock canoccasionally miss fact ors such as mi norlithological changes, small fractu re zo nes an dpressure ch anges (Va sseur et al. 1995).However th ese tent to have minima l impactwhen avera ged ove r th e syste m, e speciallywhen used within a detailed geological model.Temperature ModelingModeling temperature a cross a large are arequires the measured te mperatures t o be fedinto a constructed geological mo del an dpropagated according to the physical constraintsinterpreted. Come rcially available software ca nthen b e u sed to construct a m ore a nisotropicthermal model. Burial hi story modeling softwarethat is routinely used in petroleum exploration forthe prediction of the degree and timing of sourcerock m aturation, re servoir tempe rature i sfavored a s it can inco rporate the physicalproperties o f rocks as well a s structure.Lithostratigraphic data from a well, or fromseismic interpretation, can be input to the burialhistorymodeling software package and modeled115

Australian Geothermal Conference 2010Salamander-1 – a geothermal well based on petroleum explorationresultsBertus de Graaf*, Ian Reid, Ron Palmer, David Jenson, Kerry ParkerPanax Geothermal Ltd, PO Box 2142, Milton, Qld 4064, Australia.Corresponding author: bdegraaf@panaxgeothermal.com.auSalamander-1 and the PenolaGeothermal ProjectSedimentary Aquifer (HSA) system of theLimestone Coast Geothermal Project.Panax Geothermal Ltd (“Panax”) has securedseveral Geothermal Exploration Licences (GELs),within troughs or grabens in the Otway Basin insoutheast South Australia (Fig 1). Together thelicences cover an area of more than 3,000 km 2 .The initial development of the Limestone CoastGeothermal Project is focused on the PenolaTrough (GEL 223 and GEL 484; Fig. 2) -promoted by an existing comprehensive databaseof historical informaton from extensive oil and gasexploration and production.The Penola Trough has been subjected tointensive oil and gas exploration, including 27deep petroleum wells with wireline logging andconventional core measurements of reservoirporosity and permeability. In addition, 321 km 2 of3D seismic and a significant amount of 2Dseismic data have been acquired in this area. Allof this data is available as part of a public openfile database.Figure 1. Panax's geothermal licences in the LimestoneCoast Geothermal Project – demonstrating proximity toinfrastructureGeological settingThe Penola Geothermal Project lies within thePenola Trough, which is a Cretaceous-agedextensional basin formed on top of Palaeozoicbasement. It is covered in South Australia by GEL223 and GEL 484. The Penola Trough forms partof the Otway Basin, a rift basin initiated during theCretaceous, on what is now the southern marginof the Australian mainland. The trough contains athick accumulation of Cretaceous sediments andis bounded to the northeast by the PadthawayRidge, and to the southwest by the KalangadooHigh. Well and seismic data indicate that thesediment pile is six kilometres thick or more, inthe deepest sections of the trough, including over1,000m thick sections of cretaceous Pretty HillFormation (sandstones).The Penola Trough area of the Limestone CoastProject bears many similarities with the ImperialValley area of California which includes the EastMesa and Heber geothermal areas (Bergosh etal., 1982). The East Mesa reservoir sandstonesare similar to the Pretty Hill Formation sandstone,which is the target reservoir of the propsed HotFigure 2. Panax's geothermal licences in the LimestoneCoast Geothermal Project – outlining four major troughzones, including the Penola Trough and the PenolaGeothermal ProjectThe Pretty Hill Sandstone was deposited by a lowsinuosity, high energy, sand-rich river system(Scholefield et al, 1996). Its good reservoir qualityhas been known from petroleum activities for117

Australian Geothermal Conference 2010some time. Significant permeability is preserved inthe formation at depths of up to 3,500m.3D seismic data has been particularly useful inmapping the distribution of the porous sandstonesof the Pretty Hill Formation. The seismic responseevident on 3D data represents to some extent theporosity distribution within the Pretty HillFormation (Boult & Donley, 2001). The loweracoustic impedance of the more poroussandstones can be identified on the basis of thedata (with varying degrees of success), and hasbeen used to map the distribution of individualsandstone bodies.TroughLimestone Coast Geothermal Resources 1.Measured(PJ)Indicated(PJ)Inferred(PJ)Total(PJ)Report Da tePenola 11 ,000 32,000 89,000 132,000 18/02/2009Rivoli & St. Clair 53,000 53,000 28/01/2009Rendelsham 17 ,000 17,000 28/01/2009Tantanoola 130,000 130,000 31/03/2009Total 11,000 32,000 289,000 332,0001.by Dr. Graeme Beardsmore of Hot Dry Rocks Pty Ltd, Melbourne, AustraliaTable 1 - Reproduced from the 2008 Penola Trough Reportby Dr. Graeme Beardsmore (Hot Dry Rocks Pty Ltd)Temperature data, and rationale forSalamander-1A numerical model was generated for the PenolaTrough to compute, in three dimensions, thedistribution of temperature which best matchesthe observed surface heat flow distribution, whilerespecting the laws of conductive heat transfer,and the thermal properties of the geologicalstrata. The model predicted that the temperatureis relatively constant around 160°C at 4,000 mwithin most of the Penola Trough. Thistemperature modeling and a knowledge of thereservoir quality provided the rationale for thedrilling of the first geothermal well, Salamander-1,in the Penola Trough in the first quarter of 2010.Additionally, Beardsmore (2009), assessed theLimestone Coast Geothermal Project for itsgeothermal potential and determined a “MeasuredGeothermal Resource” of 11,000PJcomplemented by an additional 32,000 PJ of“Indicated Geothermal Resource”, see Table 1below. This resource classification was inaccordance with the Australian Code forReporting of Exploration Results, GeothermalResources and Geothermal Reserves, 2008edition.This highly encouraging resource classification,the third in Australia and second-ever resourceclassification of a HSA, confirmed the highpotential of geothermal energy development insedimentary basins.The ability to directly assess this significantgeothermal resource potential is courtesy of thelong history of petroleum exploration withnumerous wells and seismic surveys, see figure 3below.Figure 3. GEL 223 – Penola Geothermal Project showingoutline of the Measured Geothermal Resource, andinformation relating to prior petroleum wells drilled in thearea.Salamander-1 drilling objectivesSalamander-1 is the first geothermal well in theproject area. It was drilled in Panax’s GeothermalExploration Licence 223 (GEL223), in the firstquarter of 2010.The drilling objective for Salamander 1 was tointersect the Pretty Hill Formation (reservoir) at adepth of 3,000m and continue to 4,000m tosecure an intersection of 1,000m of the reservoir.To minimise the risk of intersecting gas, the wellwas sited 5km away from the closest gasproducing well (now depleted) and at least 500mvertically down dip of any gas producing well.Also, a further drilling requirement was that nosignificant faults should exist between the closestgas producing well, and Salamander 1 - that couldpotentially be a gas seal.To minimise friction losses in a future productionenvironment, the casing scheme was designed aswide as possible, although standard casing118

Australian Geothermal Conference 2010dimensions and safety issues took precedence.Figure 4 below presents the casing scheme forSalamander 1.DrillingSalamander was drilled with Weatherford’s Rig828 and the design and management of the wellwas contracted to AGR.Penetration (ROP), Panax elected to set 13-3/8”casing at this depth. The 13-3/8” casing was runand cemented at 1764m.The 12 ¼” hole was drilled using Schlumberger’sPower V system, an auto-vertical seekingdirectional drilling tool, to maintain verticality. TheROP in the 12-1/4” section was extremely goodand intersected the top of Pretty Hill Formation at2,900m, some 150m shallower than the originalprognosis.Figure 5. Salamander-1 Well Completion SchematicFigure 4. Schematic illustration of the well design and wellcasing design – Salamander-1WDI Rig 828 was handed over to Panax at therig’s previous engagement at 20:00 on 12December, 2009. The rig move to Salamander 1was disrupted due to poor weather restrictingtruck access to the rig, in particular the largertransporters. Salamander-1 was spudded at 23:45on 31 January 2010.The 24” hole was drilled from the base of the 26”conductor to 503m, and 18-5/8” casing was runand cemented at 499m (Figure 4).The 17-1/2” hole was drilled from the 24” casingshoe to 1,766m. Due to lower Rates ofA cement plug was set from 2,800 – 2,900m tostabilise this zone, and the 9 5/8” liner was runand cemented at 2,896m.The 8-1/2” hole section was drilled to 4,025m witha bit change made at 3,670m. An EMS directionalsurvey was recorded at TD and Open Hole logswere run. A 7” perforated liner was run and set tobottom.Following displacement of the drilling fluid withbrine, and with the wellhead and Christmas treeinstalled, the rig was released at 1500 hrs 26March, 2010.The well was drilled to a total depth of 4,025metres in 45 days (inclusive of 3 rig repair days)119

Australian Geothermal Conference 2010using only 6 drilling bits, setting a new drillingrecord for the Australian geothermal sector.Importantly, the Salamander-1 drilling programmewas concluded with zero Lost Time IncidentsWireline loggingThe wire-line logging services were provided bySchlumberger, and four trips were planned. Themain objectives of the wire-line logging were to: i)determine formation acoustic velocities to tie inthe well with 3D surveys, ii) observe pressureregimes to determine the absence of anyhydrocarbons, and iii) collect information aboutformation density and porosity. The wire-lineprogramme is presented in Table 2 below.LogXY Cal-GRDLL MSFLLDT CNL-GRHNGSSonicScannerMDT-GRVSI-4Interval2890 – 1750m4021.5 – 2898m3738 – 28984025 – 29892900 – surface. Sonicscanner switch modes toCement Bond Log37 points (could not pass3845m)3810 to 99mTable 2 – Wireline Logs ObtainedSalamander-1 Well ResultsThe recently announced 15.4°C increase in themeasured bottom hole temperature (“BHT”) of theSalamander-1 well, from 156°C to 171.4°C, whichis likely to have been caused by breakthroughfrom the well bore to the reservoir, has confirmedpredictions of the regional 3D temperaturemodelling by Hot Dry Rocks Pty Ltd (HDR Report,2008). Figure 6 below shows that the 3Dmodelled temperature at 4,025m for theSalamander-1 well of 172°C is very close to therecently measured BHT at 4,000m of 171.4°C.This figure also shows that projectedtemperatures at 4,000m increase to the south andsouth-east into Panax’s GEL484, reaching 180°Cwithin 4km of Salamander-1.The new BHT of Salamander-1 combined with thenew 3D temperature modelling is likely toincrease the geothermal resources of GEL223.More importantly, it upgrades the overall regionalpotential of the Penola Geothermal Project, withhigher temperatures allowing the proposed poweroutputs to be achieved with lower flow rates.Figure 6 - Penola Trough, 3D conductive heat flow modelling(HDR, 2008) - predicted 172°C @ 4,025m, while the actualmeasured temperature at Salamander-1 for this depth was171.4°C – recently measured.A borehole seismic survey (vertical seismic profileor VSP) was recorded by Schlumberger in thenear vertical Salamander-1, on 19 March 2010.This VSP allowed a good correlation of theSalamander-1 penetrated geology with the 3Dseismic data. Examination of the seismic datacorrelated to Salamander-1 located towards thesoutheast of the GEL, shows an excellentcorrelation, with a more reflective seismicpackage in the upper section of the Pretty HillFormation. Seismic attributes of reflectionstrength, and volume attributes of RMS reflectionstrength were used to determine the arealdistribution of these more porous sandstones. Asingle reservoir body is interpreted extending fromthe Ladbroke Grove gasfield area, through theSalamander-1 well, and 6.5km to the south east -covering an area of approximately 22.5km 2 . Thiszone is interpreted as having similar reservoirproperties to those in the well.Full analysis of the petrophysical review arecurrently in progress and will be available at thetime that the paper is presented in November2010.A review of the wireline logging data of the openhole section of the Salamander-1 well (from2,900m to 4,025m) for estimating the reservoirquality of the target reservoir rocks (the Pretty HillSandstone) has been completed by Down UnderGeosolutions (“DUGEO”) based in Perth. Wehave been advised that the combination of the“Lin-Log and MDT methods” are the mostappropriate to estimate the “transmissivities” (≈120

Australian Geothermal Conference 2010reservoir quality) of the intersected targetreservoir rocks.DUGEO has advised that the transmissivities inDarcy metres (Dm) of the Pretty Hill Sandstone inthe open hole section of Salamander-1 well rangefrom 6.7 Dm to 13.5 Dm, as set out in Table 3below:MostLikelyCaseMaximumCaseMethodAverageLin-LogrelationshipOptimisticLin-LogrelationshipTransmissivity(Dm)6.713.5Table 3. Transmissivity RangeA number of depleted gas wells withinapproximately a 5km radius of Salamander-1were checked for injection potential. It isestimated that 4 of these wells could accept morethan 100 l/s in their current configuration andmore than 200 l/s after workover, with furtherimprovement possible with additional perforation.The costs of working over these wells are in therange $2-3m each, which is a much moreeconomic proposition than drilling a purpose builtreinjection well.Salamander-1 Reservoir Testing Program –Step 1 :Transmissivity is loosely described as thecapacity of the reservoir to flow water. The “MostLikely Case” of 6.7 Dm is lower than the minimumrequirements of 10 Dm for a flow rate of 175 l/sec@ 145°C. However, with higher geothermaltemperatures now apparent (from the most BHTmeasurement in Salamander-1, and consistentwith the 3D temperature modelling), flow ratescould be reduced as these two inputs areinversely proportional. Also, the “Maximum Case”estimate provides the upside.Salamander-1 Well TestingSalamander-1 was discharged using air lift to a10,000m 3 capacity HDPE lined pond. Steam wasvented to atmosphere from a flash tank and waterflow was measured using a weir plate.The reservoir fluids collected in the pond will beused for the injection test which is the last test tobe carried out on the well.a. Flow well using air lift;b. Measure pressure, temperature and flow(surface), down hole logging tools in thetarget reservoir measure pressure,temperature and flow.Salamander-1 Reservoir Testing Program –Step 2 :Analysis of downhole pressure and temperaturedata obtained during short cleanup flows,although indicating higher than expectedtemperatures, up to 171 deg C also indicatedinitial mud damage that arose during drilling.The well was acidized using 15% acetic acid inorder to remove the skin damage caused bycalcium carbonate and flowed again usingnitrogen gas lift. Results of this operation areawaited.Results of production and injection testing arepending. Based on the results of these tests, thefeasibility of installing a pilot plant will beinvestigated.121

Australian Geothermal Conference 2010c. Re-inject produced geothermal brineusing filtration unit and high pressurepump;d. Down hole logging tools in the targetreservoir measure pressure, temperatureand flow.Sand81-7008, Terra Tec Inc under contract toSandia National Laboratories.Boult, P.J., and Donley, J., 2001, VolumetricCalculations Using 3D Seismic Calibrated AgainstPorosity Logs — Pretty Hill Formation Reservoirs,Onshore Otway Basin. Eastern Australian BasinsSymposium, 2001Scholefield, T., North, C.P., and Parvar, H.L.,1996, Reservoir characterisation of a lowresistivity gas field – Otway Basin, SouthAustralia, APPEA Journal 1996.Figure 7. Salamander-1 Well clean-up flow test, May 2010.Salamander-1 – Interpretation AndAnalysis Of Well Testing ResultsAt the time of writing this Abtract, this work iscurrently being undertaken.This will be available for presentation at the timethat the paper is presented in November 2010.ReferencesBeardsmore, G. R., 2008, Limestone CoastProject: Penola Geothermal Play. Statement ofGeothermal Resources. Hot Dry Rocks Pty Ltd,2008.Bergosh et al 1982, Mechanisms of FormationDamage in Matrix Permeability Geothermal Wells,122

Australian Geothermal Conference 2010Petrophysical Methods for Characterization of Geothermal ReservoirsBen Clennell 1 , Lionel Esteban* 1 , Ludovic Ricard 1,3 , Matthew Josh 1 , Cameron Huddlestne-Holmes 1 , JieLiu 2 , Klaus Regenauer-Lieb 1,2,31 CSIRO Earth Science & Resources Engineering, ARRC, 26 Dick Perry Avenue, Kensington2 University of Western Australia, M004, 35 Stirling Av., 6009 Crawley3 Western Australian Geothermal Centre of Excellence, ARRC, 26 Dick Perry Avenue, KensingtonAbstract: Australia has a unique wealth of nonvolcanicgeothermal resources. To tap these forlocal and distributed energy needs, optimaltargeting and development strategies for thereservoirs are needed. In the case of HotSedimentary Aquifers (HSA), many tools forformation evaluation developed in the oil and gasindustry for High Pressure/ High Temperatureenvironments could be applied directly; thisapplies both to downhole tools and laboratorycharacterization methods. The standard suite ofpetrophysical methods is adequate fordetermining basic matrix and fluid characteristics,but for full characterization of flow andgeomechanical parameters advanced acousticand nuclear magnetic resonance tools could finda place, at least for high value resources. In bothcrystalline rock and HSA plays, image logs are avital tool, not least for enabling improved core-tologworkflows. Understanding of the thermalproperties of rock and how that relates to fieldscale thermal structure is another challengeunfamiliar to most petrophysicists. Modelling frompore/grain through to formations scale is vital forassessing both thermal and flow performance.CSIRO and the Western Australian GeothermalCentre of Excellence have brought togetherexperience and capabilities in rock physical andthermal properties and in this paper review whatpetrophysical tools and methods are available tothe geothermal industry and illustrate some of theresearch currently underway.Keywords: Petrophysics, Reservoir CharacterizationThermal Conductivity, Heat Capacity,Porosity, Permeability, HP/HT.IntroductionAustralia’s immense geothermal energy resourceshave the potential to contribute significantly to ournation’s demand for affordable and sustainableclean energy. However, for the efficientexploitation of geothermal energy a goodcharacterization of the reservoir is a prerequisite.Petrophysics is the systematic investigation ofrock and fluid physical properties to understandthe rock type and fluid flow performance ofunderground formations. Petrophysicaltechniques measure the physical and chemicalproperties of the rock either in situ (well logging)or in a lab (core and chips). The data generatedfrom these analyses can be used to characterizethe reservoir rock’s composition, structure,porosity and permeability, temperature, pressureand mechanical properties, leading to a betteroverall understanding of the resource.Petrophysical methods for the characterization ofgeothermal reservoirs have progressedsignificantly overseas particularly for the extremeenvironments such as HR and volcanic sources.In this paper we describe CSIRO’s expertise inthe Petroleum sector, which is directly applicableto HSA. It is our intention to grow this expertisefurther into the HR domain.Our petrophysical characterization of geothermalreservoirs forms part of a larger collaborativeinitiative of CSIRO and Universities working withindustry and government with the aim to reducethe risks of implementing geothermal technologythrough computationally assisted guidance ofdrilling programs to target locations and depths.The core of the required scientific innovation willbe to develop new data assimilation inversiontools, which ensure that all available information,including geologic, geophysical, geomechanics,fluid dynamics, rock physics, geochemistry andeconomic data is integrated into the subsurfaceand above-ground geothermal design. In thiscontribution we focus on the petrophysicalmethods for characterization of geothermalreservoirs. These comprise the characterization ofrock mechanical, thermal, geochemical and fluidpermeability properties.Standard petrophysical tools andcharacterization workflowsSince the Schlumberger brothers invented the firstdownhole electric log in the 1920s, the oil and gasindustry has provided the platform to develop awide range of petrophysical logging tools, (seeEllis and Singer 2007). The range of physicalproperties that can be sensed today is impressive(Table 1). While the uptake of down-holetechnologies in water, mining and environmentalsectors has largely been limited by cost to a basicsuite, certain slim-hole and extreme environmenttools have been developed outside of oil and gas,most notably for borehole imaging.Petrophysical characterization serves three mainpurposes, each with its own workflow to: (1)define the lithostratigraphy, and therefore help ingeological correlation between wells and definitionof reservoir structure; (2) quantify matrix, fluid and123

Australian Geothermal Conference 2010Table 1Type of tool Physics employedwhatis measuredResistivitygalvanicResistivityinductionGamma raystandardSpontaneouspotential (SP)Density logPhotoelectriceffectNeutron logSonic LogGamma rayspectralMagneticsusceptibilityGeochemicalLogging ToolElementalCaptureSpectroscopyDielectric logNuclearmagneticresonanceWaveformsonicDirect currentinjected into theformation, usuallyfocused electrodearray.Coil magnetic fieldtransmitter inducesground loops information, measuresconductivity offormationScintillator detectstotal radioactivity offormationElectrochemicallyand electrokineticallyinduced potentialsUses backscatteredgamma rays from aradioactive source,typical 137 Cs.Measured fromrelative energyabsorption from thedensity sourceDetects neutronabsorption fromhydrogen nucleiGeothermalapplicationRelations betweenporosity, watersaturation and watersalinity.Relations betweenporosity, watersaturation and watersalinity.Shaliness, heatproductionPore water salinity.Indicator of clay layersvs sand layers,permeability of thelatter.Sensitive to formationbulk density, based onatomic numbercontrasts.Indicates mineralelectron densitySensitive to watercontent including claybound waterPorosity and strengthCompressional wavevelocity(stiffness) of formationEnergy sensitivity Shaliness, rockdetectors distinguish geochemistry heatU, Th and K productionMeasures magneticpermeability of theformation byinductive coupling orother means.Measures spectra ofgamma absorptionre-emissionMeasures gammaspectra from nucleiactivated by neutroncapturePermittivity fromattenuation andphase of propagatingelectromagneticwaveTransverserelaxation time ofhydrogen nucleiCompressional andshear wavesCan provide goodlithologicaldiscrimination, even ofvery thin layers, and incrystalline rocksIdentifies a wide rangeof specific elements,especially heavy onesIdentifies a narrowrange of specificelements, especiallyheavy onesPorosity/saturationindependent of salinity.Shale and textureindicatorFluids content, poresize and permeabilityestimateGeomechanicsflow properties of the reservoir and (3) assessstresses and geomechanical properties in situ.1. For definition of lithology and structure one islooking to use tools, such as gamma ray,resistivity and sonic, which show substantialcontrasts between different lithologies. At thisstage, quantitative relationships are less importantthan the signatures that mark boundaries betweencorrelated rock units. In crystalline rock situationsmagnetic susceptibility (not generally used in oiland gas exploration) is valuable to identify certaingeological markers. At the next level of detail werequire identification of features such as tiltedbeds, coal seams, igneous dykes, faults andfractures. Dip logs and image logs are most usefulhere. The best type of imaging tool for ageothermal hole (borehole televiewer for air orclean water filled holes, ultrasonic scanning tool)may be different from that used mostly in oil andgas (formation resistivity array imager). Incrystalline rocks, flow properties begin and endwith fractures, so image logs are especially vital.Resistivity images have the advantage that openand flowing fractures may have different contrastto closed or non-flowing ones.In the laboratory we gain complimentarylithological information from petrographic analysisof cores and cuttings samples, usually aided byoptical and electron microscopy. Structurallogging of core can define fault and fracture sets,determine relative timing, and differentiate naturalfrom drilling-induced fractures, which is notalways possible from borehole images alone.2. For quantitative formation evaluation for a HSAtarget reservoir, combining the basic “oilfield”tools performs a vital function. Density andneutron tools enable porosity to be delineated,along with some understanding of matrixmineralogy and especially clay content. Neutronporosityto density porosity ratio can indicatezones not completely saturated with water, andwith gamma ray, indicates clay content. Clayfraction has a first order effect on permeability,even if porosity is constant. Sonic logs canreinforce the interpretation of porosity, andindicate zones that are more or less fractured.Resistivity and spontaneous potential can beanalysed together with an understanding of theclay content and porosity, to deduce pore watercontent, saturation and salinity. This informationalso helps integrate downhole observations withsurface electromagnetic geophysical surveyssuch as deep resistivity or magnetotellurics (MT).Wireline-conveyed formation testing (e.g. MDT*)is the final technology used in a standard “oilfield”petrophysical evaluation. Drawdown and build-uptests define permeability and can also becombined with fluid sampling. Wireline testersoffer a much more versatile solution than largerscale packer tests or drill stem tests, though the124

Australian Geothermal Conference 2010latter are required in crystralline rocks orotherwise low-matrix-permeability settings.Laboratory measurements on core substantiallyreduce the uncertainty in formation evaluation. Atthe CSIRO petrophysics laboratory for example,we routinely measure porosity and permeability ofcore samples under overburden conditions,electrical properties at a wide range offrequencies corresponding to different loggingtools, and sonic compressional and shear wavevelocities under conditions of (P, T) in situ stress.3. Stresses in the Earth are complex, and definingstress conditions in a borehole can be a difficultprocess. Generally it is based on the analysis ofbreak-out and fracture patterns observed in imagelogs, complimented by formation pressure testsdesigned to open natural fractures and/or inducenew fractures. To interpret these results andmake predictions of reservoir performance,“fracture-ability”, well stability in the reservoir andoverburden, it is desirable to have core suitablefor geo-mechanical testing. From core testsunconfined compressive stress, cohesive strengthand frictional strength parameters can bedetermined directly and used to construct aMechanical Earth Model. At CSIRO we conductsuch tests routinely, and have recently obtained aHigh Pressure (to 150 MPa confining, >500 MPaaxial) High Temperature (to 200 C) triaxial rig andwith 20 channel acoustic emission to monitor in3D the development and coalescence offractures. Sonic logs are used widely forgeomechanical interpretation, and coremeasurement of P and S velocities and theiranisotropy is very helpful for defining rock physicsand mechanical property models. This also closesthe loop for seismic-to-log-to-core workflows.Smarter petrophysics: geothermalapplications of non-standard tools.The limited scope of standard tools has led tomore exotic physics being applied to formationevaluation in oil and gas. Some special tools haveobvious HSA applications but require cost/benefitto be proven. Extreme condition versions of someadvanced oilfield tools do not even exist yet.An obvious “upgrade” is to use a spectral gammaray tool that quantifies the U, Th and K content ofthe formation. This enables quantification of in situheat production, and is also useful for rock typing.Natural gamma tools can be cross-calibrated withlaboratory spectral gamma ray detectioninstruments used for recovered cores or chips.Geochemical Logging Tool (GLT* ) and neutronactivation (e.g. ECS* or FLEX*) logs can beemployed to quantify a subset of the chemicalelements in the formation. The GLT has been* MDT, GLT and ECS are marks of Schlumberger.FLEX is a mark of Baker Hughesused in both sedimentary and crystalline rocks,including hydrothermal zones, in the OceanDrilling Program, with excellent results. The GLTis complex, having active neutron and gammasources coupled with filtered detectors and apassive natural gamma ray spectroscopy module.Hot sources raise obvious issues of HSEpermitting. The ECS approach can be engineeredto use a non-chemical neutron source called aminitron, essentially a small particle accelerator inthe tool. However, the number of elementsdetected by ECS is limited, as is its track recordfor hard rock situations.Another oilfield technology of potential value inhot sedimentary aquifers is Nuclear MagneticResonance (NMR) spectroscopy. NMR toolsmeasure the magnetic spin relaxation of hydrogennuclei, and offer the only way to determine rockpore size distribution and thereby estimateformation permeability in a continuous way. Thisis an excellent way to interpolate pointmeasurements from permeability tests on core,and compare with formation tests from MDT.NMR also provides a salinity-independent totalporosity and an estimate of clay content. CSIROis currently comissioning an elevated P/T NMRcore testing capability which will help to assessthe value proposition of NMR petrophysics forHSA geothermal applications.Temperature limits on measurementsSome technologies such as NMR, which dependson magnets, and some spectroscopic tools withcrystal detectors have inherent temperaturelimitation making their application to hottergeothermal settings unlikely. On the other hand,various logging tools have been developedspecifically for geothermal boreholes: the extremeenvironment televiewer for example. Theexploitation of ultra-deep (>5000 m) and HighPressure High Temperature (HPHT, or >150 C >70 MPa) oil fields has also driven thedevelopment of sophisticated petrophysical toolsthat can perform in extreme environments,leading to a convergence of technology towardsthe needs of the burgeoning geothermal market.Major players like Schlumberger, Baker Hughesand Weatherford compete in this market withsmall and specialised tool suppliers, including agrowing number from Australia. Standard toolsare available up to at least 260 C and 200 MPa.We therefore see expanding options to deploysuch oilfield tools—or cheaper versions of them—into the geothermal marketplace.In the laboratory also, standard oilfield analysisequipment for porosity, permeability, resistivityetc. is not designed for the extremes ofgeothermal conditions. A typical operating rangeis 70 MPa and 100C. At least in a researchcontext, systems have been engineered enablingall of these parameters to be obtained under in125

Australian Geothermal Conference 2010situ formation conditions and throughoutprolonged experiments (e.g. Milsch et al. 2008).Understanding the thermal structureWhether it is a Hot Sedimentary Aquifer (HSA) orEnhanced Geothermal System (EGS), appraisaland exploitation of a geothermal reservoirrequires a detailed characterization of thermalproperties and heat flows. The temperaturedistribution on the underground is driven by thesurface climate, the characteristics of the rocks toconduct, produce and accumulate the heat andalso the fluid circulation into the rocks and thebasal (or mantle) heat flow. When estimating thetemperature at depth, we have to consider themain mechanism driving the thermal flow andtemperature distribution according to thecontrolling equation.T C V. grad T div .grad T Az0tHere ρ is mass density, C the specific heatcapacity, T the temperature, V the fluid velocityvector, lambda the thermal conductivity, and A thespecific heat production per unit rock volume. Inthe absence of fluid flow, rock thermalconductivity estimation is adequate for thecharacterization of the thermal conduction regime.An estimation of the specific heat of the rock willprovide direct insights on the stored heat. Thisparameter will also play an important role duringthe production of the reservoir as the rockreleases the stored energy to the working fluid.The radiogenic heat production rate of the rocksin combination with stratigraphic information willenable us to estimate the local heat flow value. Ifadvection is significant, in the system, thenpermeabilities, porosity and fracturecharacterization are needed to quantify V andthus the contribution of the advective flow to thetemperature distribution.In terms of fluid transport characterization, the keyparameters to identify are the permeabilities(absolute for water only and relative, when steamand liquid are both present in the reservoir),porosity and the thickness of the targetedreservoir. Initial and time lapse values from repeatlogging passes will be of interest to design andassess any well stimulation processes and toquantify the contribution of the local hydrogeologyto the temperature distribution.Figure 1. Isogeotherms and formation temperature/ depthprofile in a hypothetical rock sequence (adapted Jorden andCampbell, 1984).These parameter estimations are important at theexploration stage to evaluate the geothermalpotential and must be refined at the productionstage where the underground geothermal systemwill be significantly affected by the production andre-injection of a working fluid.Determining thermal propertiesMost of the research on thermal properties hasfocused on modeling of the thermal propagation ingeological reservoirs. But the needs inpetrophysical data, particularly in geothermalreservoirs, have never been so important to verifyand validate models. Two main approaches areavailable to provide hard data on thermalstructure and properties downhole logging andlaboratory measurements. Both methods havetheir advantages and disadvantages:• Logging data give access to a continuousrecords with depth but with high uncertaintiesbecause of some chemico-physical processesthat actively occurs during the data recording(heat flow, fluid flow, mudcakes, fracturations...)or simply because of the limited sensitivity anddepth resolution of the tools (Beck et al., 1971).Three general logging methods are available: (1)relaxation methods (Wilhelm, 1990), (2) directthermal measurements downhole with no controlat all on chemico-physical influences and (3) theclassical correlation methods which combinelogging parameters such as sonic traveltime,hydrogen index, density, porosity, lithology andtemperature (Brigaud et al., 1989; Demongodin etal., 1991; Griffiths et al., 1992). All of thesemethods use empirical models and some form ofaveraging or effective medium theory to predictthe thermal properties. To calculate the thermalconductivity and diffusivity accurately from suchmethods, the tools being employed must becapable of distinguishing the different mineralsand fluids sensitively.126

Australian Geothermal Conference 2010• Lab measurements enable control of all theparameters (mineralogy, porosity, water content,fractures) but the sampling (except for drillcuttings)is never continuous and the in-situreservoir conditions are difficult to reproduce inthe lab. Moreover, lab equipment and protocolsare not infallible. Poor sample quality,representativeness, coupling, effects of fracturesetc., limit typical accuracy to around 5% or so. It israre that specific information is gathered onanisotropy: this requires measurement of thetensor components of thermal conductivity anddiffusivity. At the highest temperatures, non-linearproperties and especially thermal expansion makegood measurements a challenge.One new technology that can help overcomesome of these issues is Optical Thermal Scanning(Fig. 2). OTS is a non-destructive non-contactmeasurement of the thermal conductivity andthermal diffusivity on rocks and minerals at roomconditions (Popov et al., 1999). OTS is able toscan a sample surface with 3 temperaturesensors in combination with a focused mobile andcontinuously operated constant heat source. Theheat source and sensors move at the same speedrelative to the sample and are calibrated beforeand after each measurement with rock standards,which leads to high accuracy quoted as 1.5%, asampling size from 1 cm to 70 cm long having anyFig. 2. Optical Thermal Scanner instrument for rock thermalproperty measurements on full cores (modified from Popov etal., 2010). Optical Head: sensor (1) measure the rocktemperature at room condition and sensors (2 and 3)measure the rock temperature after heating from the heatsource (4).shape, short time measurements (10-30s),contactless tool and without time and cost ofspecial sample preparation. Hence the OTC offersthe opportunity to investigate direct thermalconductivity, diffusivity, pore space and geometryon full cores and anisotropic thermal properties ofsamples under various conditions, and with awide range of form factor, size, and prevailingconditions (wet/dry, elevated temperature, etc).The combination of the two petrophysicalapproaches (lab and downhole) can be verypowerful and removes most of their respectivedisadvantages. Lab measurements allowparameter calibrations and better control on thelogging interpretations. Proper integration ofpetrophysical data from lab and downholesources enables us to test, correct and enhancethe existing theoretical models used to predictsteady state and dynamic thermal behaviour.Petrophysical property modellingThe characterisation of geothermal reservoirs is atypical multi-scale problem. CSIRO andWACOGS have developed a number of workflowsto model and upscale rock physical propertiesappropriate for geothermal problems. An excitingrecent advance with wide uptake in the petroleumindustry is the incorporation of so-called DigitalRock or Digital Core methods whereby highresolution 3D images of real rocks form the basisfor physical property calculations. In the exampleworkflow presented here we analyse small scalerock properties in 3D based on micro-computedtomography (micro-CT), and scale up forhydraulic properties by using percolation theory.Firstly, the original images of microtomographyare converted to binary images, and 3-D binarymodels are built up from the tomography slices.Image processing and segmentation is crucialwhich recognises target phases (pores, differentminerals) according to their greyscale values inthe images. Fig. 3 shows the pore-structure of asynthetic sandstone sample of size 1.3*1.3*1.17mm 3 after segmentation in which the poreinteriors are “painted” light blue. Micro-scalecharacterisations are analysed based on 3-Dbinary models, including volume fraction (porosityfor pores of porous media), percolation (orconnectivity), specific surface area (SSA),tortuosity, and anisotropy.Stochastic analyses of all phases are the secondstep of this multi-scale characterisation. Itconducts scale-dependent probabilities and thesize of representative volume element (RVE). Ourstochastic analysis uses the moving windowmethod, in which a cubic sub-volume with variableside-length L moves all over the model. In thisway, the probabilities of volume fraction,percolation and anisotropy of different scales areobtained. The size of the RVE is determined whenthese probabilities are convergent with theincreasing sub-volume-size L (Liu et al. 2009).127

Australian Geothermal Conference 2010Fig. 3. Pore-structure of a synthetic sandstone sample, lightblue denotes inside of pores. Model size is 1000*1000*900voxels, resolution of voxels is 1.3 μm.Once at the RVE size or larger, the estimatedproperties are reliable for application to thatparticular rock type. Meanwhile, the probabilitydistribution functions can be used to create “digitalrocks” at any size which have certain commoncharacteristics, but vary in their details. Forexample, one may vary porosity, connectivity, orfracture density and see the effects.Physical properties of geothermal reservoirs, inparticular, permeability, elastic parameters, andthermal expansion coefficient are calculated onthe digital rock elements using numericalsimulations. Permeability can be simulated byusing the finite difference Stokes equation solverPermsolver and based on micro-scale porestructures. Fig. 4 is a RVE of size 1 mm 3 of asynthetic sandstone sample. The simulatedpermeability is 1763 millidarcy. Thermomechanicalproperties are computed by using afinite element method based on the mineralproperties and structure. Fig. 5(a) is a RVE of adigital rock which includes two weak minerals ofdifferent shapes. Fig. 5(b) gives the relationshipsof stress and strain components of the RVE andmatrix, respectively. Although matrix and weakinclusive minerals are all isotropic, themicrostructure makes the upscaled properties ofthe RVE remarkably anisotropic, see Fig.5(b), inwhich red and black lines correspond to x and z-directions, respectively. As stress and strainvalues in Fig. 5(b) are volumetric means onelement sets, the relationships for the matrix showa slight anisotropy over all elements in the RVE(see blue and green lines in Fig.5b), which isreasonable. With these simulated results, thermomechanicalproperties of the representativevolume element can be computed.To extrapolate properties from microscale tomacroscale it is necessary to combine microscaleproperties with scaling laws. We usepercolation theory to extract the main criticalparameters including fractal dimension, criticalFig.4 A RVE of a synthetic sandstone sample forpermeability simulation.(a)(b)Fig.5 (a) RVE of a digital rock with two weak minerals ofdifferent shapes, matrix is not shown; (b) relationships ofstress and strain in x and z-directions for matrix and RVE,respectively. Oriented microstructure makes the RVE beremarkably anisotropic, although minerals of matrix andinclusions are all isotropic.exponent of correlation length, percolationthreshold, and crossover length, for the purposeof upscaling.For the synthetic sandstone sample shown in Fig.4, the critical exponent of correlation length andthe fractal dimension extracted from microtomographyare close to the theoretical values.The percolation threshold of this structuredmedium is 3.94%, while the sample has a porosityof 24.4%, which is far above the percolationthreshold. In addition, the crossover length of thecritical model is smaller than 50 μm. Based uponthese observations, we can directly use thetransport properties obtained from the micro-scaleanalysis for upscale modelling in the macro-scale.128

Australian Geothermal Conference 2010The simulated permeability of 1763 md in 1 mmscale is close to the experimental testpermeability on a plug sample of 2567 md.Results show that the detection of the scaledependence of permeability is accurate from thecrossover length of 50 micron to the cm laboratoryscale (Liu et al. 2010).These upscaling methods based on statisticalphysics on small rock samples (potentially evencuttings) can be combined with more geologicallyinformed methods of geostatistics, based on coreandlog based lithofacies or flow units, to populatereservoir models at the larger scale.SummaryThe characterization of geothermal reservoirsrequires the integration of field, laboratory andcomputational petrophysical methods. Theformulation and parameterization of rock propertymodels is based on proper combination ofremotely sensed (e.g. seismic) downhole andcore measurements. These methods combininginterpretation and modelling provide critical inputsto the understanding of the geothermal reservoir.From early exploration, hard petrophysical dataand good physical understanding will help toconstrain and develop regional and reservoirmodels of the geothermal resources not only toevaluate the resources but also to plan itsproduction. At the development stage, these datacontribute to reduce the risk on the well design,well stimulation, hydraulic fracturing but also thefield sustainability during production.CSIRO, with WAGCOE and other partners aroundAustralia, is investing in applied R&D and criticalinfrastructure to improve the application ofpetrophysics to the geothermal energy sector.Demonstration projects now underway in WA willprovide an excellent testbed for several of themethods described here.ReferencesBaker Hughes 2010. Logging services web pagehttp://www.bakerhughes.com/products-andservices/drilling-and-evaluation/formationevaluation-/wirelineservices/petrophysics/nautilus-ultra-hphtBeck, A.E., Anglin, F.M. and Sass, J.H., 1971.Analysis of heat flow data-in situ thermalconductivity measurments. Canadian J. Earth Sc.,8, 1-19.Brigaud, F., Vasseur, G. and Caillet, G., 1989.Use of well log data for predicting detailed in situthermal conductivity profiles at well sites andestimation of lateral changes in main sedimentaryunits at basin scale. Rock at great depth, Eds.Maury and Fourmaintraux, Rotterdam, A.A.Balkema. ISBN 90 6191 975 4, Vol.1, 403-409.Demongodin, L., Pinoteau, B., Vasseur, G. andGable, R., 1991. Thermal conductivity and welllogs: a case study in the Paris basin. Geophys. J.Int., 105, 675-691.Ellis, D.V. & Singer, J.M., 2007. Well Logging forEarth Scientists. Second Edition, Springer, 2007.Griffiths, C.M., Brereton, N.R., Beausillon, R. andCastillo, D., 1992. Thermal conductivity predictionfrom petrophysical data: a case study. GeologicalApplications of Wireline Logs II. GeologicalSociety Special Publication, 65, 299-315.Jorden, J.R. and Campbell, F.L., 1984. WellLogging I – Rock Properties, BoreholeEnvironment, Mud and Temeprature logging. SPEof AIME, N.Y., Dallas, 131-146.Liu J., Regenauer-Lieb K., Hines C., Liu K.,Gaede O., and Squelch A., 2009. Improvedestimates of percolation and anisotropicpermeability from 3-D X-ray microtomographicmodel using stochastic analyses andvisualization. Geochem., Geophys. Geosyst.,10,Q05010, doi: 10.1029/2008GC002358.Liu J. and Regenauer-Lieb K., 2010. Applicationof percolation theory to microtomography ofstructured media: percolation threshold, criticalexponents and upscaling, submitted to PhysicsReview E.Ocean Drilling Program Downhole Logging Guide.http://www.ldeo.columbia.edu/BRG/ODP/LOGGING/TOOLS/geochem.htmlPopov, Y., Pribnow, D.F.C., Sass, J.H., Williams,C.F. and Burkhardt, H., 1999. Characterization ofrock thermal conductivity by high-resolutionoptical scanning. Geothermics, 28, 253-276.Popov, Y., Miklashevskiy, D., Romushkevich, R.,Safonov, S. and Novikov, S., 2010. Advancedtechnique for reservoir thermal propertiesdetermination and pore space characterization.Proceedings World Geothermal Congress, Bali,Indonesia, 25-29 April 2010.Schlumberger 2010. Logging services web pagehttp://www.slb.com/en/services/additional/geothermal.aspxWilhelm, H., 1990. A new approach to theborehole temperature relaxation method,Geophysical Journal International, 103, 469-481.129

Australian Geothermal Conference 2010GEOTHERMAL POWER PLANT COMPETITIVENESS IN INDONESIA:ECONOMIC AND FISCAL POLICY ANALYSISFarizal 1,* and Prashanti Amelia Anisa 21, 2 Department of Industrial Engineering. The University of Indonesia. Depok, 16424, Indonesia* Corresponding author: farizal@ie.ui.ac.idThis paper assesses ge othermal p ower pla nt(GPP) co mpetitiveness through eval uating itsproduction costs a nd the appli cation ofgovernment incentive s, i.e. the Cle anDevelopment Mecha nism (CDM) and theimplementation of ca rbon tax. Then the GPPcompetitiveness is compared to coal power plant.The results show that the effect of duty free, valueadded tax free (VA T), implem entation o finvestment tax credit, and pre survey governmentincentive ca n de crease the ge othermal sellin gprice by US 0,75 ¢/KWh, 0,91 ¢/KWh, 0,23 ¢/KWhand 0,6 9 ¢/K Wh, respe ctively. Implementation ofCDM and carbon tax o n coal also improves GPPcompetitiveness. Geothe rmal can co mpete wit hcoal in the year 4.Keywords: geothe rmal electri city price, fiscalpolicy, carbon tax, competitivenessElectricity demand in Indonesia increases with therate 5.9%. T his rate i s bi gger than th e ele ctricitysupply gro wth. This situation turn s o ut to be anelectricity cri sis. Sched uled bla ck out occurs inmany places including at the industrial areas.Depending on notorious fossil b ased fuels to fulfilelectric energy is not a sustaina ble policy due tothe fuels are not re newable, means their depositis dimini shing by time, a nd they a re also notenvironmentally friendly. The go vernmentproposed National Energy Mix 2025 to reduce thedependency on fossil fuel s and at the same timeto promote t he utilizatio n of rene wable energyincluding g eothermal (Fi gure 2 ). It is projecte dthat geothermal will contribute 5% to total ene rgysource at the time of 2025.IntroductionIndonesia i s ran ked th e fourth, after USA,Philippines, and Mexi co, among co untries thatutilizes geothermal energy to generat e electricityeventhough she has the large st p otency in theworld, i.e. 27.169 MW or approximately 30 to 40%of worl d g eothermal p otency. The potencyspreads along the volcani c belt from th e island ofSumatra to Timor (Figure 1). Of 27,169 MW, only1052 M W or less than 5 % of the capacity havebeen u sed. One reaso n why geothe rmalutilization has not been used opt imally isIndonesia stil l dep ends heavily on fo ssil b asedfuels a s prim ary ene rgy reso urces for electri citygeneration and the competitiveness of geothermalplant.Figure 2. National Energy Mix 2025As addition to its sca rcity and environmenta limpact, another issue is the price of fossil fuels isnot sta ble. T he p rice vol atility genera tes d oubtabout l ong term stability of the energy in thefuture. In contrast, geoth ermal energy is pollutionfree a nd environmental fri endly. Its fu el cost i srelatively stable during the whole period of a GPP.The i ssue related to its utilization i s explorationand drilling costs whi ch are quite high. Could thelow and stable pri ce of GPP set off its highexploration and dr illing costs? What is the impactof taxation schem es to GPP com petitiveness?How competitive is the GPP compare to coal firedpower plant? Those are some questions that willbe addressed here.Figure 1. Indonesia Geothermal PotencyResearch MethodologyThis study is based on the profitabilit y indicatortheory. Profitability indicators indicate the level oftaking a n investment de cision fro m the investo rstandpoint. Investment de cision is a d ecision t oprocure fixed assets which include resources andfunds at the present time to obtain series of longtermp rofitability in the fu ture. Th ree i ndicators,130

Australian Geothermal Conference 2010net pre sent value, intern al rate of re turn, an dpayback period are used on this study.Net pre sent value (NPV ) of a proje ct is the totalcash flo w p er ea ch unit of time that has beencharged to the pre sent value of the investmentthat have b een cashed. NPV is calculated byadding up all the ca sh flows occur from period ofzero, so-called as investment, to the last period ofthe project.where:I: investmentr : rate of returnAn: cash flow / proceedn: economic value of investmentsVn :salvage value of i nvestments at t he e nd ofeconomic periodIf NPV turns out to be positive, then the project isfeasible to run. Po sitive NPV in dicates th einvestment has achieved favourable condition.Internal rate of return (IRR) is a percent increasein the value of money c ontained in the cu rrentcash flow. IRR ca n be in terpreted al so as thediscount ra te that produ ces zero NPV .In gen eral, i nvestment d ecision ba sed on NPVand IRR will give a consistent result which mean ifan inve stment propo sal is considered fea siblebased on the NPV, the p roposal assessed basedon IRR is also feasible.Payback period (PBP) is the time needed to fullyrecover the costs and li abilities incurred in aproject.Where:PBP: Payback Period, yearsm: Year of the CCF negative after a positive CCFm+1: Year of the CCF p ositive after a neg ativeCCFCCF m : Cumulative Cash Flow in m (0)Although P BP doe s not refle ct profitabilityindicator of an invest ment pro posal and thecalculation d oes not co nsider time value ofmoney, but it is often used to compl ement thefeasibility analysis of investment proposal s. PBPmay reflect the liquidity of an investment proposal.Profitability indicators al so consider investorrequired rate of return (RRR). RRR is influen cedby two main factors, ri sk free rate and ris kpremium project with rel ationship as RRR = riskfree rate + ri sk p remium project. RRR is definedby the investor. The hig her the ri sk of a proje ct,the higher the RRR is. Geothermal electric pricesare considered fea sible if they can generate anIRR more than RRR . In this study, RRR is set as16%. In Indo nesia, gove rnment regulations li mitthe sale p rice of geot hermal ele ctricity notexceeding US 9.7 ¢/KWh . Geothe rmal ele ctricityprice i s con sidered competitive if the price do esnot exceed 9.7 ¢/KWh.Input FactorsGovernment poli cy on tax and incentive sinfluence th e competiti veness of GPP. InIndonesia, g overnment p olicy on tax inclu desinvestment tax credit, duty-free im port, free valu eadded tax, a nd g overnment initial surv ey. Whileincentives include clean development mechanismand carbon tax.Investment tax creditBusiness tax rate for GPP is 30%. Investment taxcredit is assumed to reduce the basic tax rate to acertain extent, up to 5%, over six year period.Duty-Free ImportThrough import duty policy, the govern ment freesthe import d uty that used to be 5 percent to 0percent.Free Value Added Tax (VAT)PMK 24/PMK.011/201 0 policy states thatgeothermal exploration activities are borne by thegovernment. With his p olicy, the governm entbears the V AT payable on the impo rtation ofgoods that are used for geothermal e xplorationactivities.Government Initial surveyPreliminary survey con ducted by the governme ntin the ea rly stages of G PP develop ment is aneffective way to red uce the ri sk of d evelopinggeothermal resource s. The su rvey include ssurface surv eys and drilli ng of two test wells. Allsurvey results are t ransferred to privatedevelopers at no cost.Implementation of Clean DevelopmentMechanismClean development mechanism (CDM) scheme isa scheme in which developed countries are ableto use the amou nt of ca rbon di oxide (CO 2 )reduction as a result of joint proje ct betwee ndeveloped a nd devel oping count ries. Certifie d131

Australian Geothermal Conference 2010Emission Redu ction Credit (CER) is issu eddepending on the amo unt of green house g asesreduction. CER credit can be traded on marketthus can be used to improve the profitability of theproject. CER value is assumed to be US$ 10 perton CO 2 and the emission factor is 0.819 ton CO 2per MWh.Application of Carbon TaxCarbon tax a pplied to fossil fuel plants will makesome changes in the structure of en ergy p ricesthus will ma ke geothermal more attra ctive (Table1).Assumptions on GPP project used in this study:• Type of project: Th e total proje ct (proj ectdownstream + upstream)• Scale of project: 110 MW which consisted of twounits of 55 MW• Period of contract: 35 years (5 yea rs for pr eproduction and the rest for production)• Capacity Factor: 90%• Success Ratio:• exploration wells: 50%, production wells: 80%• Decline Rate 3% per year• Steam Production• exploration wells: 8 MW, production wells: 12MW• Income Tax Rate: 30%• D epreciation Me thod: Declining Ba lance w ith 8years• Investment Tax Credit 5% per year for 6 Years• Free Value Added Tax: 10%• Duty-Free ImportSummary of GPP costs is shown in Figure 3:Figure 3. Geothermal Power Plant costsThe number of wells drilled in the pre productio nis 20 wells. Among them nine make up wells for a30-year production periods (year of p roduction 2,5, 9.12, 15, 18, 21, 24, 27).Result and DiscussionThe calculation result for 110 MW GPP found thatusing the p rice of US 9.7 0 ¢/KWh, the I RR, NPV,and PBP are 15.16%, US$ 17,570, and 10 years,respectively. The IRR is slightly less thandeveloper’s RRR, which is 16%. Meanwhile, whenusing developer’s RRR the electricity price will bea bit highe r than 9.70 ¢/KWh, i.e. 10.0 7 ¢/ KWh.the NPV is US$ 30,578 and nine years PBP.A 10% reduction on capacity factor and electricityprice ca uses a decrea se in IRR of 1.6 8%, whileon the other side a 10% addition of capacity factorand the price of ele ctricity increa ses 1.22% ofIRR. Reduction of investment costs by 10% leadsto an increase of IRR by 1.51%, while the additioncost of inve stment with the same amount causesa decrease in 1.28% IRR (Figure 4)132

Australian Geothermal Conference 2010Figure 4. Capacity Factor, Investment Costs and Price ofElectricity; Sensitivity Analysis on IRRToward the NPV, a 1 0% red uction of ca pacityfactor and electri city price ca uses the NPVdecreases o f $27.83, while the 10 % additioncauses a n increase of $19.32. Reduction ofinvestment costs by 10% lead s to an increa seNPV of $ 1 9.63, whil e the ad dition decreasesNPV of $ 19.59 (Figure 5).Figure 7. IRR Profile at Different Capacity FactorAt government electricity price of US 9.70 ¢/KWh,the IRR may be more than 16% if the in vestmentcost ca n be pre ssed n ot less than 1 0%. If theinvestment cost red uces by 10%, the IRR wo uldincrease by 1.18%, 1.51 % and 1.76 % on thegeothermal electricity prices of US 7 ¢/KWh, 9.70¢/KWh and 12 ¢/KWh, resp ectively. On thecontrary, If t he inve stment co sts in creases by10%, the IRR will drop by 1%, 1.28% and 1.51%on the geothermal prices of 7 ¢/KWh, 9.70 ¢/KWhand 12 ¢/KWh, re spectively. If the investmentcosts d ecreases by 1 0%, then the feasi blegeothermal electricity price at 16% IRR will not beless than $ 9.5 ¢/KWh (Figure 8).Figure 5. Capacity Factor, Investment Cost and Price ofElectricity; Sensitivity Analysis on NPVFor the PBP, a 10% reduc tion in capacity fac torand electricity price causes a sl ower PBP withone yea r, while the 10 % addition causes thepayback pe riod a year earlier. Reduction ofinvestment cos ts by 10% c auses PBP a yearearlier, while the addition causes a slower PBP byone year (Figure 6).Figure 8. Profile of IRR at Different Investment CostsThe m ost i nfluential in centive im proving th eproject I RR is the 10% VAT-Free incentivefollowed b y import duty-free ince ntive,government prelimi nary survey, and investmenttax credit (Figure 9).Figure 6. Capacity Factor, Investment Cost and Price ofElectricity; Sensitivity Analysis on PBPWith the governments’ geothermal electricity priceof US 9.70 ¢ /KWh, with a capacity factor of 90%(base case, CF 90%), th e IRR will be less than16%. If the capa city factor redu ces to 80%, withgeothermal electricity prices of US 7 ¢/KWh, 9.70¢/KWh and 12 ¢/KWh, the IRR will fall by 1.55%,1.87% and 2. 13%, respectively. In the base case(CF 90% ) with IRR 16%, geoth ermal electricityprice is fe asible if it is ab ove 10.2 ¢/K Wh (Figure7).Figure 9. Analysis of the effect of Government Incentiveson IRR and Electricity Prices133

Australian Geothermal Conference 2010The effect of free d uty, free VAT, theimplementation of invest ment tax credit, and thepreliminary survey i ncentive re duces th egeothermal electricity price by $ 0.75 ¢/KWh, 0.91¢/KWh, $ 0 .23 ¢/KWh and $ 0.69 ¢/KWh,respectively (Table 2).In scena rio 1, at the el ectricity pri ce of 9.7 0¢/KWh, the I RR only reaches 14.6 6%. Electricityprices that fe asible at IRR 16% i s 10.5 4 ¢/KWh(Figure 11).Table 2. Effect of Government Incentives on Electricity PricesIncentivesElectricity Price At IRR=16%No Incentives 12Investment Tax Credit 11.77Initial Survey by the Government 11.25Free Duty 11.31Free VAT 11.09The effect of the impleme ntation of CDM redu cesthe electricity price by 0.82 ¢/KWh and increasesthe IRR to 16.53% whi ch is a bit above thedesired RRR (Figure 10).Figure 11. Analysis of the effect of Scenario 1 on IRR andPriceResults of scen arios 1 t o 6 are sum marized atTable 4 bellow:Figure 10. Analysis of the effect of CDM on IRR andElectricity PriceTable 3 b elow lists 6 sce narios of imp lementingcombination of government tax and incentives.Scenario123456Table 3. List of ScenariosFiscal PolicyDuty FreeVAT FreeITC 5% for 3 yearsNo Survey IncentivesNo CDM SchemeDuty FreeVAT FreeITC 5% for 8 yearsNo Survey IncentivesNo CDM SchemeDuty FreeVAT FreeITC 5% for 5 yearsSurvey IncentivesNo CDM SchemeDuty FreeVAT FreeITC 5% for 5 yearsNo Survey IncentivesCDM SchemeDuty FreeVAT FreeITC 5% for 5 yearsSurvey IncentivesCDM SchemeDuty 5%VAT 10%No ITC 5% for 5 yearsNo Survey IncentivesNo CDM SchemeTable 4. IRR and Geothermal Electricity PriceIRR (%)Scenariowhen electricityprice is set as 9.70cents/KWhElectricity Price(cents/KWh)when IRR is set by16%1 14.66 10.542 15.29 10.233 16.02 9.64 16.53 9.265 18.72 8.686 14.53 10.5According to National Energy Mix 202 5, coal willstill play as primary energy source, but it role willbe re duced. As su bstitution, geoth ermal willcontribute 5% of total energy sou rce (Figure 2). Itis im perative to check GPP competitivenesstoward coal power plantThe a ssumptions used i n cal culating t he cost ofcoal fired power plant are as follow:• Capacity: 300 MW• Capital Cost: US$ 2,500/MW• O & M Cost: US$ 88/KW• Operation time: 7000 hours a year• Carbon Tax: Rp 60/KWh• Fuel Consumption: 0.439 Kg/KWhThe fact that geotherm al fuel cost is quite stabl eduring the lif e time of GP P, while on t he contrarycoal fuel cost is in creasing by time. Using US $24.76/MWh for geoth ermal energy cost for 3 0years, g eothermal cost will be less than co al inyear 1 4. At the time coal fuel cost i s US$ 2 4.78(Figure 12).134

Australian Geothermal Conference 2010promote GP P. Without i mplementation of ca rbontax, the cost of geothermal electricity can competewith coal in t he year 14. The impl ementation ofcarbon tax to coal plant improve s thecompetitiveness of G PP. Geothermal cancompete with coal in the year 4.Figure 12. Levelized Cost Comparison BetweenGeothermal and CoalMeanwhile, when applying carbon tax, geothermalwill outperform even earlier, i.e. in fourth year,where the co st of coal is $ 24.88. Application ofcarbon tax adds to the cost of coal Power Plant to$ 1.26/MWh (Figure 13).Figure 13. Levelized Cost Comparison BetweenGeothermal and Coal with Carbon TaxConclusionThe current policy, refe rs to PM KNo. 21/PMK.011/2010 with the highestgeothermal price US 9.70 cent/KWh, is not able toprovide the investor desired IRR, i.e.16%. To fulfilthe investo r IRR with the government elect ricityprice, the investment costs must be reduced to atleast 10% or increase the capacity factor by 10%.Free VAT a pparently ha s the mo st significantinfluence on the I RR and th e g eothermalelectricity price followed by free import duty, initialsurvey by go vernment in centive, and i nvestmenttax credit. The effect of th ose poli cies can lo werthe selling price by US 0.91 cent/KWh, 0.75 cent/KWh, 0.69 cent/KWh, and 0.23 cent/KWh,respectively. The effect a nd the impl ementationof CDM reduces the pri ce by US 0.82 cent/KWhand increases the IRR to 16.53%.From th e six scena rios studied, o nly scen ario 3(combination of Duty-Free, Free of VAT, the ITCis 5%, a nd government pre su rvey), scena rio 4(Duty-Free, Free of VAT, the ITC is 5%,application of the CDM ) a nd scenario 5 that ca nenhance GPP competitivene ss . Amo ng thesethree sce narios, scen ario 5 (Duty Fre e, free ofVAT, ITC i s 5%, gove rnment survey and theapplication of the CDM) i s the be st scenario toReferencesAkmal, F, H. Darn ell, and P. Sugiharto (2000 ).The geothe rmal power busi ness in Java Bali:trading m echanism, com petitiveness with fossilfuel and challenges to Its dev elopment.Proceedings World Geothermal Congress pp 777-783. Kyush u-Tohoku, Jap an, May 28 – Jun e 10,2000.Ministry of Ener gy and M ineral Resour ces(2006). Blu eprint Natio nal Ener gy P olicy 200 6-2025, Th e Rep ublic of I ndonesia. J akarta,IndonesiaCanada, J.R., W.G. Su llivan, and J.A. White(1996). Capital Invest ment Anal ysis fo rEngineering and M anagement, 2 nd ed ., PrenticeHall, Inc. Upper Saddle River.Danar, A.S. (201 0). Geothermal ele ctricity pri cedetermination based on investme nt deci sion.World Geot hermal Co ngress. Nu sa Dua,Indonesia (in Indonesia language)Dardak, E (20 07). Inco rporating en ergycommodity price vol atility in economi c cost ofsupply analysis for elect ricity expansion plan. The10 th International Confe rence on QIR (Quality inResearch). Jaka rta, Indonesia, 4 – 6 De cember2007.Darnell, H. and A.R.T. Artono (200 1). An analysison the g eothermal electricity competitiveness. theAnnual S cientific Confe rence of In donesianGeothermal Association. Jogya karta, Indonesia,February 2001.135

Australian Geothermal Conference 2010A METHOD FOR CALIBRATING AUSTRALIAN TEMPERATURE-DEPTHMODELSDesmond Fitzgerald 1* and Raymond Seikel 11 Intrepid Geophysics Unit 2, 1 Male Street, Brighton Victoria, Australia 3186.*Corresponding author: des@intrepid-geophysics.comThe Au stralian ge othermal indu stry i s movin grapidly, an d i n that pr ocess requi res a lot fromgeophysics t o aid in ch aracterising regio nalprospectivity for exploitable heat resources.Various grou ps are usi ng hybrid me thods toestimate ‘Curie point’ te mperatures at depth, o ralternatively, the temperature at 5 kilometresbelow the surface. D eep drilling observationsand airborne magnetic compilations are the ke ycomponents, together with a basem ent geologyinterpretation. Several ge nerations of this workare already published with more to come.A method to test the se maps an d al so helpcharacterise uncertainty is proposed based upona deep 3D continental model scale, extending tothe lithosph ere. Variable surfa ce tem peratureand heat flow grids, based upon remote sensingare u sed, to gether with a simple lith osphereboundary condition. Th e heat diffu sion is th enemployed to test the temperature -depth map s.Progress on applying this method to Au stralia isreported.Key words: Sensin g Heat Flow Anoma lies, EMspectrum, Hi gh Energy photon s, Co ntinentalscale 3d Geology modelsIntroductionNot enough attention is paid to the influ ences ofheat, both on-going and its history, in the field ofexploration g eoscience. Scant rega rd i s paid toheat alterati on mineral prod ucts, unl ess it isobviously a primary indicator of an economicallyviable re source. A more holistic a pproach tocreating 3 D earth m odels that em brace al laspects of heat an d its i nfluence on rocks i srequired. Th e ch allenge (Stein) i s th e classicalone of usi ng the mea sured tem perature a ndtemperature gradient at an obje ct’s su rface toinfer the te mperature field at d epth, T(x,t), afunction of position x and time t. Near the earth’ssurface, th e t emperature g radient is e ssentiallyvertical, so t he out ward heat flow q s is theproduct of the vertical gradient of th etemperature T(z), whi ch is most eve rywherepositive downwards (temperature increases withdepth z), and the me asured o r estimate dthermal conductivity of the material, k.Australian developmentsThe syste matic gath ering of boreh oleobservations of downhol e temperatu res, ro ckheat cond uctivities and tempe rature gradientswith depth below su rface were the basis ofproducing th e first g enerations of p redictedtemperatures 5km below the su rface for theAustralian continent (Chopra & Holgate, 2005).,see figure 1.Figure 1: Map of estimated crustal temperatures at a depthof five kilometres.Maus, 1997 develo ped the id ea thatcompilations of the ma gnetic anomaly mapcould form the basis for estimating the depth tothe Curie point. This requires a technique to findthe bottom s of ca usative bodi es. A movingwindow method using a modified Spector-Granttype algorith m was initia lly propo sed for thistask.In re cent ti mes, Geoscience Au stralia ha sproduced the 5th gene ration mag netic anomalygrid an d with this edition there i s much greaterconfidence that the longe r wavelengthanomalies (g reater than 50 km ) are faithfullyreproduced. Thi s follo ws from the AWAGSsurvey, fund ed by the "Securi ng Australi a'sEnergy" initiative. Also , more theoretical workhas been done on automatic depth to basementtechniques, such that the ability to separatetop/bottom and centres of causative bodies. (seeStavrev and Reid 2007).Several groups have forme d to ta ckle thi sproblem and the recently reported progress fromthe USGS team (Bouli gand et al 2009) isencouraging. In this work, a v ariety ofapproaches were applied to th eCalifornia/Utah/Nevada re gion whe re o ver 25 %of all g eothermal po wer station capacity for the136

Australian Geothermal Conference 2010world i s lo cated. The se heat sources a relabelled "conventional" as they follow f rom nearsurface v olcanic sources as sociated with t heRocky Mountains and Yellowstone.In Australia, a signifi cant initiative is beingmounted a round the so -called South Australi aHeat Flow anomaly. This was first recognized bySandiford 19 98, as th e burial of a basementsequence en riched i n h eat pro ducing e lementsduring thermal subsidence following rif ting. Thismajor rift in the Aust ralian continent fromAdelaide north to the Coope r Ba sin, has amassive granitic intrusion which is Uranium rich.Naturally o ccurring lo w grade th ermo-nuclearheating of th is gra nite pe aks at aro und 4 kmdeep. This type of play i s termed an EnhancedGeothermal System.GeochemistryIt is mai nly t he oil ind ustry that h as come torealise ju st how im portant the heat history ofrocks and sedimenta ry packa ges is, in theprospectively of basins for oil or gas finds.Observationally, the key technique i s definingthe mine ral alteration produ cts and f romexperiment, just wh at must have bee n the heathistory to have left t his l egacy. Diam ondexploration has also traditionally lo oked fo rindicator minerals that im ply the ne cessary heathistory.Structural geologyMost stru ctural geology phen omena have anunderlying genesis where he at ha s played amajor role.1. A deep pipe from the ma ntle to the surfac ecan form as nature “vents" heat in explosiveevents. Kimb erlite pi pes retain a re cord o fheat altered minerals that can b e interpretedfor temperatures during formation.2. Granite intrusions in general.3. All tectonic a nd plate scal e movement s ar ethought to be manifestation s of hea tcirculation effects in the mantle.4. Volcanic a ctivity is most ly asso ciated wit hcrustal faulti ng, often on a plate b oundary.Basalt flows and flood pl ains, and sills areremnants of this activity.5. Rifting or failed rifts.Observational geophysicsAll the m easurable ge ophysical phenomenasuch as gravity, magnetics, deep crustal seismicand radiometrics can i ndirectly be u sed to hel pinterpret stru ctural geology and defi ne heatprovince boundaries.Missing f rom these lowe r cost, bul kobservational methods, is a viable methodologyto either dire ctly or i ndirectly obse rve heat ortemperature. Thi s would lead to near surfacetemperature and/or heat flow maps th atsignificantly improve upon Figure 1. So what i sthe physics and why do we stru ggle with thischallenge?Electro-magnetic spectrumThermal radi ation is emitted by all su bstancesabove a bsolute ze ro and inclu des visible an dinfrared radiation and some ultra-violet radiation.Thermal radi ation o ccurs in soli ds, liqu ids, an dgases, at the speed o f light and ha s noattenuation in a va cuum. Thermal radiation canoccur between two bodies with a colder mediumin bet ween. Actually, the ele ctromagneticspectrum can be expressed in te rms of energy,wavelength, or freq uency. Each way of thinki ngabout the EM spectrum is related to the others ina precise mathematical way.Wavelength(m)Frequency(Hz)Energy (J)Radio > 1 x 10 -1 < 3 x 10 9 < 2 x 10 -24Microwave 1 x 10 -3 - 1 x 10 -1 3 x 10 9 - 3 x10 11 2 x 10-24 - 2 x 10 -22InfraredOptical7 x 10 -7 - 1 x 10 -3 3 x 1 011 - 4 x 2 x 10 -22 - 3 x 10 -10 14 194 x 10 -7 - 7 x 10 -7 4 x 10 14 - 7. 5 x 3 x 10 -19 - 5 x 10 -10 14 19UV 1 x 10 -8 - 4 x 10 -7 7.5x10 14 -3x10 16 5 x 10 -19 - 2 x 10 -17X-ray1 x 10 -11 - 1 x 10 -83 x 1016 -3 x 2 x 10 -17 - 2 x 10 -10 19 14Gamma-ray < 1 x 10 -11 > 3 x 10 19 > 2 x 10 -14Table 1 : shows these relationships..So why do we have three ways of describingthings, each with a diff erent set of physi calunits? After a ll, frequency is measured in cyclesper second, wavelength is measured in meters,and energy is measured in electron volts and i sinversely proportional to the wavelength.By conventi on, ga mma-rays are reported i nelectron volts, with the hig hest en ergy going tothe right hand side. Figure 2 below has the orderreversed, an d sh ows th e EM s pectrumwavelengths increasing t o the rig ht, in what i sthe conve ntional man ner for visible light, heatetc.137

Australian Geothermal Conference 2010This tra ditional view also i s some whatmisleading with the lab el "thermal radiation".This type of heat i s a rguably the l east wellunderstood. While at th e sam e time, mostpeople are very familiar with the wea ther an dsensing radi ant heat. The body is rece ptive,largely due to the i nteraction of parts of the EMspectrum with water and to a lesser extent, otherorganic molecules. With this in mind, it is normalto associate heat with infra-red ba nd, terra-hertzband a nd th e ultra-viol et band s (sun -burn). Wehave evolve d to se nse visible lig ht whichradiates from the sun due to thermal black bodyradiation at 5800K. What this is really showing isthe range of Plank's law (see Figure 3) and howparts of the spe ctrum relate to "blac k body"radiation fro m the Eart h. As can be se en,microwaves are not con sidered to be thermalradiation, yet clea rly that have a ve ry effectivecapacity to "heat". At the hi gh end of the energyscale, gamma rays are part of the mix in nucl earpower plants that, of course, gene rate heat priorto conversion to electricity.From a g eophysical poi nt of view, any EMemission fro m the Ea rth could b e h arnessed t oextract heat. The ch allenge of how to do this i slargely an engineering problem.Figure 3 Plank's law and Black Body radiation. At normalsurface temperatures of the Earth, the range ofwavelengths still emitting radiation is large, while theoverall energy (the area under the curve), falls offdramatically.There are two types of radiation categories Volumetric p henomenon – ra diation e mittedor a bsorbed throu ghout gases, tran sparentsolids, some fluids Surface ph enomenon – radiatio n t o/fromsolid or liquid surface.Initially, heat is emitted fro m the surface of th eearth and then it is "attenuated" and absorbed inthe atmosphere.The ma gnitude of the radiation va ries withwavelength – it’s spectral. The wavelength of theradiation is a major factor in what its effects wil lbe. This is illustrated in Figure 4.Figure 2. Radiation Spectrum, showing wavelengths. Note,it is not usual to think of thermal radiation extending into thehigher energy parts of the spectrum e.g. X Rays.Radiation is made up o f a continuo us, non -uniform distribution of m onochromatic (sin glewavelength)components.The m agnitude a nd spectral di stribution (howthe radiation varies with wavelength) varies withtemperature and type of emitting surf ace. Weare inte rested in the continental crust an dregional va riations due to the a ge an dcomposition of the rocks.138

Australian Geothermal Conference 2010Attenuation of signalFigure 4. Shows varying emissivity with wavelengthsRadioactive (crustal) heat productionThe contine ntal crust cont ains a relatively highdensity of radioactive isotopes, primarily those ofuranium, thorium, and potassium.Hence, within a region the heat flow depends on(1) radioactivity in the crust,(2) tectonic setting, and(3) heat flux from the mantle below.For a given area, te rmed a heat -flow province,the me asured he at flow " q" varie s li nearly withthe near surface radioactive heat production.The tech nology to obse rve ra diometricsemissions f rom the ea rth o r g amma-rays h astraditionally a lways ig nored the "low channels",due to hi gh cou nts a nd diurn al effe cts, oftenlabelled "skyshine". Very fe w observations usingthese instruments have ever been done at night,when these diurnal effect s are mini mised. Withthe ne wer hardware now availa ble and th esensitivities and counting capacities increased, a66 litre system, de signed for 1 024 channelsand/or a Ge Li purpo se b uilt system, would beable to ob serve a con siderable po rtion of the"thermal radiation" tail. The first channel of aradiometrics instrument cove rs mo st of th econventional EM sp ectrum, cove ring f rom 0 to12 keV. Figure 5 recasts the EM spectra into theform favoured for measuring gamma rays.Counts/sec/chan8060402000 1 2Energy (MeV)Figure 5 Typical Gamma Ray spectrum view, observedremotely within 100 m. of ground.3Radiant ene rgy from the earth i s readilyabsorbed a nd attenuat ed in the lowe ratmosphere. At lower energies, the photoelectriceffect is mo stly responsible for this. Radiometricairborne surveys are conducted at flying heightsof aroun d 100m, to ensure emissi onscharacteristic of particul ar mine rals can beobserved bef ore thi s attenuation sm ooths th echaracteristic spe ctral p eaks. Thi s requiredobservation distan ce reduces as the energy ofthe photons reduces, if in dividual characteristicsare to b e se en. If one re-examines thegeneralized " Black Body" diffuse curve at 3 00degrees Kelv in, it is ve ry smooth an d covers abroad ra nge of energie s, with no p ossibility ofpeaks from good emitting minerals.This appears to be an a rea that req uires eithermore research, or pr actical ob servations indiffering te rrains and na tural lig ht condition s.Towards th e cla ssic heat po rtion of thespectrum, bo lometers exis t th at h ave b eendeployed at large di stances, to ob serve thatportion of th e sp ectrum. (TIMS, ASTER et c.)The rea son these sho w any signal, is tha twavelengths are chosen that minimi zeabsorption with water vapour.Modelling construtsThe o ceanic lithosphere is relatively un iform i ncomposition, and little heat is gene rated within itby radioactivity, oceanic heat flow i s essentiallya si mple functio n of age. In contrast, acontinental cru st is hetero geneous incomposition, due to its much lo nger tectoni chistory. M oreover, th e heat flo w dep endscritically on radioa ctive heat prod uction in thecrust. The two prim ary effects are thus thatcontinental heat flow i s prop ortional to thesurface crustal radio activity in a given region,and decreases with the time since the last majortectonic event.A rational way to con struct a 3 D continentalscale earth model that is to serve the purpose ofconstraining the complexity while still honouringthe impo rtant facto rs i s to use t he n ewradiometrics map of Australia, to d efine tectonicand cratonic provin ces characterised by theradiogenic rock content. The big unknown is thedepth extent of each of these u nits. It is alsohere th at ge ophysics is n eeded to pla y a part.To date, se e Figure 1, no geological influencesare ta ken int o acco unt when inte rpolating thetemperatures away from observation points.At the same time, great strides are being madeto construct large scale 3D solid geology modelsthat are co nsistent wi th st ructural geol ogyobservations and all the geophysical datasets.139

Australian Geothermal Conference 2010In Australia, we have 3 D models for T asmania,most of Vict oria, and ab out on e thi rd of Sout hAustralia. (Prelimina ry 3D geolo gical models o fthe Curnamo na b asin a nd Ga wler craton). I naddition, some pu rpose built 3 D he at/geologymodels for the Coop er ba sin h ave bee nconstructed with a vie w to ch aracterising theheat flow domains and defining uncertainties onthis resource.Numerical/computationalcomplexitiesA current challenge is thus to deal with all thesecomplexities, deciding on the 4 or 5 mostimportant a spects of the ge oscience, an dconstruct m ore reali stic and theref ore u sefulmodels. It turns out there is a ba rrier with thenumerical m ethods u sed to solve t he h eatdiffusion equ ation. The m ethods do n ot scalewell. F or s imple s ituations, a ll is w ell, but w henthe re quirements of 3 D, structu ral geolo gy,material in homogeneity and continent al scal eare stated, n o really viabl e tech niques pre sentthemselves. For thi s reason, lea ding ed genumerical methods, u sing F ast Fouri ertransforms etc., are bei ng follo wed up. T wosignificant re cent publi cations point the wayhere. Carat ori 200 9, elegantly sho ws ho wgravity and magnetic forward mo delling in 3 Dcan be rapidly achieved using 3D FFT methods.Li, 2007, loo ks at faste r methods for the heatdiffusion equation, but ju st the h omogenouscase.ConclusionsCraton scale and larger 3D nume rical models,with st ructural geol ogy constraints, p ropertiesand bou ndary condition s of measu red surfa ceheat flows, and more fu zzy "Curie Point" depths,can b e integ rated to cre ate heat flow anomalymaps th at wi ll be a vast improvement on theexisting practise.with a fra ctal model for crustal magnetization, J.OF Geophysical Research, VOL. 114.Caratori T ontini F., Cocch i L., Carmi sciano C.,2009, Rapid 3-D fo rward model of potentia lfields with application to the Palinu ro Seamountmagnetic an omaly (south ern Tyrrhe nian Sea,Italy) J. OF Geophysical Research, VOL. 114.Chopra, P., and Holgate, F., (200 5) A GISanalysis of tempe rature in the Austral ian crust,Proceedings of the Worl d Geothermal Congress2005, Antalya, Turkey, 24-29 April 2005.Li J-R.,Greengard L., 2007, On the numericalsolution of the heat equation I:Fast solvers in f ree spa ce. Journal ofComputational Physics 226.Maus S., Go rdon D., Fairhead D. 199 7, Curie -temperature depth estimation using a self-similarmagentization mod el, G eophysics J. Int 129,163-168.Sandiford M., Hand M., McLa ren S., 19 98, Highgeothermal gradie nt metamorphi sm durin gthermal subsidence, Earth and Pl anetaryScience Letters 163, 149–165.Stavrev P., Reid A., 2007, Degrees ofhomogeneity of Potential fields and structuralindices of Euler deconvolution, Geophysics,Vol.72, NO. 1.Stein, C.A. 1995, Gl obal Earth P hysics AHandbook of Physical Con stants AGUReference Shelf 1.The EM s pectra and its measurement continueto provide challen ges a nd insight s in to wholeearth processes. The abil ity to interpre t the EMspectra f rom the sta ndpoint of extractin ginformation a bout cru stal heat de serves mo rethought and development.A better un derstanding of how Pla nk's Law andthe physi cs i mplied by it woul d be im portant i ninterpreting non-solar inf luenced ea rth heatemissions.ReferencesBouligand C., Glen J., Blakely R., 2009,Mapping Curie temperature depth in the westernUnited States140

Australian Geothermal Conference 2010Evidence of impact shock metamorphism in basement granitoids,Cooper Basin, South AustraliaAndrew Y. Glikson 1* and I.Tonguç Uysal 21 Australian National University2 Queensland Geothermal Energy Centre of Excellence, The University of Queensland*Corresponding author: Andrew.Glikson@anu.edu.auThe o riginal observation of parallel clo selyspaced (micron s to te ns of micro ns) plan arfeatures in quartz g rains from basementgranitoid samples of Cooper Basin drill holes byI.T. Uysal, followed by Universa l stage,scanning el ectron mi croscope (SE M) a ndenergy disp ersive spe ctrometry (EDS) tests byA.Y. Glikso n, identify intracrystalli ne plan ardeformation features (P DF) in qu artz whichcorrespond to unique Miller indices diagnostic ofshock meta morphism. U-Stage mea surementsof angl es be tween the C-optic axis o f qua rtz(C qz ) an d pol es to plana r deformation features(P pdf ) in the same qu artz grains (87 p lanar s etsin quartz 54 grains) define intracrystalline planarorientations dominated b y {10 - 13} a nd {10 - 12}Miller indices, correl ated with shock pressuresabove 1 0 G pa and ab ove 20 G pa, respectively(Engelhard a nd Stoffler, 1968; F rench, 1968;Stoffler, 197 4; Stoffler an d Lan genhorst, 1994;Grieve et al., 1996; F rench, 1998). These Millerindices are cha racteristic of large impactstructures, i ncluding Chesapeake B ay (D~85km), Woodl eigh (D~1 20 km) and othe r impactstructures. S EM/EDS studies of core sample sindicate fe atures consistent with sh ockmetamorphism, including recrystallizedpseudotachylite veins an d micro breccia vein sinjected into resorbed quartz g rains. E xtensivealteration of feldspar to seri cite and illiterepresent hydrothermal effects. NegativeBouguer anomalies of c. -20 µm/s 2 below partsof the Cooper Basi n reflect lo w den sitybasement sectors (2.64-2.7 6 gr/ cm 3 ) a scompared to adjacent hi gh den sity basem entterrains (2.9-2.99 gr/cm 3 ), consistent with impactfracturing of target rocks. L ow magneticanomalies of c. -2 00 n T asso ciated with theMoomba st ructural dom e and similar structu resto the northea st ma y repre sent resetmagnetisation su ch a s is comm only a ssociatedwith impact structures. An extension of theBouguer a nd magneti c a nomalies ov er a >80km-long belt defines the minimum extent of theshock meta morphosed basement, consi stentwith mea sured sho ck p ressures of a bove 20GPa. A regional altered zone at the top of thebasement ap proximately 10.000 km 2 large an dup to 524 m eters de ep (Boucher, 2001) maycorrespond t o the impa ct aureole. Domalseismic stru ctural elem ents a nd overlyingunconformities in ov erlying sedimentarysequence of the Cooper Basin may be related topost-impact i sostatic uplift of impa ct-fracturedlow-density basement sectors. The evidence forimpact bears potential im plications for the o riginof K-U-Th e nrichment in t he basement in term sof an imp act-triggered hydrothe rmal cell,mobilization and re concentration of radiogenicelements.Fig. 1. Planar deformation features (PDF) in quartzdeflected along a post-impact fracture, McLoed-1 3745 m.141

Identification of planar deformationfeatures (PDF)Penetrative intracrystalline plana r d eformationfeatures (P DF) in q uartz po ssess sp ecificcrystallographic Miller indi ces with orientationsdiagnostic of shock metamorphism at pressures inexcess of levels asso ciated with e ndogenicprocesses such a s s eismic shock, v olcanicexplosions a nd shear pro cesses which p roducemetamorphic deformation lamella (MDL) in quartz(Engelhard and Stoffler, 1968; Fre nch, 196 8;Stoffler, 1974; Stoffler and Lan genhorst, 19 94;Grieve et al. , 1996; Fren ch, 199 8; French andKoeberl, 20 10) (Figs 1 - 5 ). PDFs are (1)restricted to individual grains, i.e. they do notcross grain boundaries; (2) are often multiple, withN sets per grain; (3) have a strict planar characterthough they can be deflected along shear zo nesand fractu res and in some insta nces fa n out; (4)are typi cally 2-4 mi cron wide; (5) are closelyspaced, typ ically 1 0-20 micron; (6) consist ofglassy or recrystallized quartz mosaic lamellae, orare m arked by arrays of fluid in clusions; (7)possess unique crystallographic orientations.Shock metamorphic features are cla ssified in thefollowing terms (French, 1998):1. Mineral fracturing {0001} and {10 - 11} inquartz: 5–7 GPa2. Basal Brazil twins {0001}: 8-10 GPa3. PDF in quartz {10 - 13}: >10 GPa4. Transformation of quartz to stishovite: 12-15GPa5. PDF in quartz {10 - 12}: >20 GPa6. Transformation of quartz to coesite - >30 GpaThe results of universal st age measurements arepresented in Figs 4 and 5. U-stage analysis o fPDF o rientations i s complicated by a n umber offactors, ari sing from d ifferential d eformationrelated to the attenuation of shock of quartz grainsby envelopin g hydro us phases, a s well as p ostshockdeformation (Figs 1, 2) and recrystallizationof quart z gra ins). Thi s in cludes su perposition ofmultiple set s of planar feature s, incl uding welldefined pa rallel micro n-scale to tens of micron -scale PDF sets (Fig. 3) a nd less well -defined topoorly d efined pa rallel or u ndulating pla narfeatures, in cluding defo rmed P DF and pl anarfeatures (PF ) formed by p re-impact an d/or po stimpactd eformation (Fi gs 1, 2). By contra st toPDFs planar feature s m ay sho w de grees ofundulation and wavin ess and may beaccompanied by fluid inclu sions an dcryptocrystalline clou ding (Fig. 3). Rarely areAustralian Geothermal Conference 2010PDFs completely expressed and a high proportionof grains display only one or two PDFs.Planar features displayed by quart z grains of theMoomba-1 and McLoed-1 sam ples includegenuine PDFs defined by:1. Penetrative planar parallel sets with spacingson the scale of few microns to few tens ofmicrons (Figs 1 - 3).2. Differential development of PDFs betweengrains and within grains.3. Common association of PDFs with less-welldefined inclusion-marked somewhat curvedfracture sets (Fig. 3).4. Post-impact deformation indicated by dragfoldfeatures along younger microfractures(Fig. 1) and wavy deformation of PDFs withingrains (Fig. 2).In the pre sent study PDFs were me asured onplanar sets displ aying high degree of parallelorientation at a few mi crons to few tens of micronintervals. PDFs m easured in q uartz grains of th eMoomba-1 and McLoed-1 samples are plotted inFigs 4 an d 5 in terms of percent fre quency ofPDFs which co rrespond within me asurementaccuracy of ±3 o to specific Miller indices of {10 -13}, {10 - 12}, {11 - 22}, {11 - 21}, {10 - 11} and {51 - 61}.The plots demonstrate a prevalence of PDFs withMiller indices of {10 - 13} and {10 - 12} in Moomba-1core samples and of Mille r index of {10 - 12} in th eMcLoed-1 co re sa mples (Figs 4, 5), sugge stingshock pre ssures of >20 GPa in both cores i naccord with criteria indi cated in studi es of sh ockdeformation of qua rtz (E ngelhard an d Stoffler,1968; F rench, 1968; Robertson et al., 1968;Stoffler, 1974; Stoffler and Lan genhorst, 19 94;Grieve et al. , 1996; Fren ch, 199 8; French andKoeberl, 201 0). Othe r m easured Mill er indi cesdiagnostic of sho ck m etamorphic ef fects in clude{11 - 21}, {11 - 22}, {10 - 10}, {10 - 11} and {51 - 61}.PDF indices correlated wi th sho ck p ressures of>20 GPa are found in la rge to ve ry la rge i mpactstructures, including Yarrabubba impact structure,Western Au stralia (M acdonald et al., 2003 ),Woodleigh i mpact struct ure, Western Au stralia(D=120 km; age ~359 Ma) (Glikson et al., 2005A,B), Chesapeake Bay imp act structure (D=85 km;age ~35 Ma) a nd l arge Ca nadian impa ctstructures (Type D) (French and Koeberl, 2010).142

Australian Geothermal Conference 2010Fig. 2. Three quartz grains bearing planar deformation features (PDF): Note the wavy planar structure of the right-hand grain,representing post-impact plastic deformation of the quartz grain. Moomba-1 2857.4 m PPL.Fig. 3. Planar deformation features (PDF) (NNW-trending at lower and right-hand part of image) and inclusion-dotted planarfractures (at left side of image). Moomba-1 2857.4 m, PPL.143

Australian Geothermal Conference 2010Fig. 4. Percent frequency distribution of the Miller-indexed angles between C-optic axis of quartz grains (CQzOA) and the pole toplanar deformation features (PPDF) in 62 planar sets in 35 quartz grains from Moomba-1 granitoid core samples. Planes areindexed within measurement accuracy of ±3% of Miller indices.Fig. 5. Percent frequency distribution of Miller-indexed angles between C-optic axis of quartz grains (CQzOA) and the pole toplanar deformation features (PPDF) in 22 planar sets in 19 quartz grains from McLoed-1 granitoid core samples. Planes areindexed within measurement accuracy of ±3% of Miller indices.144

Australian Geothermal Conference 2010Scanning electron microscopyand energy dispersivespectrometryThree sampl es were a nalysed by scan ningelectron mi croscopy (SEM) and EnergyDispersive S pectrometry (EDS) o n a Jeol-6400electron microscop e at the Research School ofBiological Studies, Australian National University(analyst: A. Glikson; su pervisors: F. Brink, H.Chen). SEM/EDS analytical procedures includedpoint a nalyses of elem ent ab undances atdetection levels of >2000 ppm. Using a 15 KeVaccelerating voltage, spot analyses were carriedwith a 1 micro n-size be am, with a spe ctrumcollection time of 80 se conds (120 seconds realtime) at 8 000 cps. A ccuracy an d p recision arebased on referen ce standards by AstimexScientific Li mited MINM 25-53 (Se rial Numb er95-050) u sing standard olivine, diopside,almandine garnet, albite and barite.Studied samples include:Moomba-1 2857.4 m (Fig. 6): Altered coarsegrainedgranitoid consisting of quartz withresorbed grained boundaries injected bymicrobreccia veins which consist of micron totens of micro-scale quartz, sericite/illite, sulphideand minor magnetite.McLoed-1 3745.9 m: Altered coarse-grainedgranitoid containing K-feldspar, albite andsericite/illite, injected by microbreccia veins.Hydrothermal alteration is manifested by partlycorroded/resorbed quartz grains and magnetitegrains.Big Lake -1 3057m : K-feldspa r-albite-quartzgranitoid consi sting o f reso rbed quart zfragments injected by microbreccia.EDS analyses indicate an abundance of sericiteand occa sional pre servation of K-feldspa r an dalbite. Ob servations relev ant to the search fo rshock metamorphic features include:1. Planar d eformation featu res ide ntified byoptical microscopy do not display on eitherelectron ba ckscatter mod e or se condaryelectron (SE1) sca nning mode. T hisindicates PDF lam ina consist ofcryptocrystalline qua rtz of similar ele ctrondensity as the host quartz.2. The inje ction of microb reccia veins in toquartz grai ns is con sistent with, althoughdoes not prove, shock effects.3. The co rroded boun daries of quartz a ndmagnetite g rains (Fi g. 6) an d theabundance of se ricite and illite a reconsistent wi th hydrothe rmal alteratio n ofthe granitoids.The extre me fine g rained texture of themicrobreccia veins su ggests their possiblederivation by recrystallization of pse udotachyliteveins con sisting of impact -generatedcomminuted and fluidized materi al simila r toveins found arou nd expo sed impact structu res(Spray, 1995; Spray and T hompson, 199 5;Reimold, 1995, 1998).145

Australian Geothermal Conference 2010Fig. 6. Moomba-1 2857.4 m: SEM image of hydrothermally altered microbreccia-injected granitoid. (A) backscatter image ofmicrobreccia veins injected into quartz grains; (B) SE1 image of frame A; (C) secondary magnetite within microbreccia veins; (D)magnified view of microbreccia vein consisting of quartz, sericite and magnetite.Implications for structure andthe origin of geothermalanomaliesBoucher (1996, 2001 ), o n the ba sis of wireli nelog sign atures and d rill core s, rep orted analtered zone up to 524 meters-thick at the top ofthe top of e arly to mi d Palaeozoic b asementunderlying th e Coope r B asin. Bo ucher (200 1)considered t he o rigin of this zone in terms ofeither weathering o r hyd rothermal alte ration, o rmerely as a wireline logging anomaly.The prese nce of si gnatures of sho ckmetamorphism within the altered top basementzone may sug gest extensive hyd rothermalactivity trigge red by a la rge aste roid im pact, a shas been documented in large impact structures(Allen et al., 1982; Naumov, 2002; Pirajno, 2005;Glikson et al., 2005; Fren ch, 1998; Uysal et al.,2001, 2002; French and Koeberl, 2010). Basedon a ssociation of hyd rothermal alte ration withimpact effects, the extent of the impact aureolemay be outlined by the altered zo ne, whichcovers a n a rea large r tha n 10,00 0 km 2 in theCooper Basin (Boucher, 2001; Fig. A8).Shock m etamorphism of ba sement sectors isconsistent with ge ophysical an d structuralevidence fo r low de nsity (2.64 –2.76 g/cm 3 )basement sectors underlying the Moomba-1 andMcLoed-1 holes, as compared to hig her densitybodies (2.8 – 3.0 g/cm 3 ) in adjacent terrain(Meixner et al., 2000 ). The lo w-densitybasement sectors likel y represents intensehydration and fracturing related to the impact, asis the ca se with the Woodl eigh impact structure(Glikson et al., 2005 a, 2005b ), Gnargooprobable imp act structure (Iasky a nd Glikson,2005) and Mount A shmore proba ble impa ctstructure (G likson et al., 2010). T he lo wmagnetic an omalies a ssociated with th e gravitylows (Meixne r et al, 2000) are con sistent withdemagnetisation of impa cted ba sement zon es,observed in a num ber of impa ct structures(Grieve and Pilkington, 1996).The evid ence fo r im pact b ears potentialimplications for the origin of K-U-Th enrichmentin the Coo per Basi n in terms of a n impacttriggeredhy drothermal cell an d a ssociatedmobilization and reco ncentration of radiog enicelements. Tempe ratures over 22 5 o C at 5 kmdepth (T 5km ) occur over a n area about 79,000km 2 large under an insulating sedimentary coverabout 3.5 – 4.5 km-thi ck at the Na ppamerriTrough between Moomba dome and Innaminckawhere geothermal gradie nts a s hi gh as 55-60 o C/km are measure d (Middleton, 1979;Wyborn et al., 2004; Rad ke, 2009). Th e highest146

Australian Geothermal Conference 2010temperatures occu r n ear Innamin cka in theproximity of McLo ed-1, where maximum sho ckpressures of > 20 G pa are me asured (Fig. 5) .Extreme total heat flo w of 7.5-1 0.3 mWm -3originate from enrichment of the Big L ake Suitegranites in ra diogenic heat-producing elements,including Uranium (13.7, 16.5 pp m), Thori um(46, 74 p pm) and Potassi um (5.2, 6.0 % K2O)(Middleton, 1979; Sandiford and McLaren, 2002;Chopra, 2003; McLaren and Dunlap, 2006). Thepresence of a highly radiogenic basement within3-4 km of the surface may in part refle ct upwardmigration an d re -concentration of la rge ionlithophile el ements a ssociated with a n impactgeneratedhy drothermal cell, as is the case insome imp act structu res, including Woodlei gh(Glikson et al. 20 05), Shoem aker imp actstructure an d Yarrab ubba imp act structure(Pirajno, 2005).ReferencesAllen, C. C., Gooding, J. L., Keil, K. 1982.Journal of Geophysical Research 82, 10083 –10101.Boucher, R.K., 1996. MESA Journal, 21, 30-32.Boucher, R.K., 2001. Nature and origin of thealtered zone at the base Cooper Basinunconformity, South Australia. South Australia.Department of Primary Industries andResources, Report Book 2001/012.Chopra, P., 2003. Preview Australian Society ofExploration Geophysics, 107, 34-36.Engelhard, W.V., Stöffler, D., 1968. in: ShockMetamorphism of Natural Materials. MonoBook Corp, Baltimore, MD, 159–168.French, B.M., 1968. In: Shock Metamorphism ofNatural Materials. Mono Book Corp, Baltimore,MD, 1–17.French, B.M., 1998. Traces of catastrophe: ahandbook of shock-metamorphic effects interrestrial meteorite impact craters. Lunar andPlanetary Institute, Houston, TX, ContributionCB-954. 120 pp.French, B.M., Koeberl, C., 2010. The convincingidentification of terrestrial meteorite impactstructures: What works, what doesn't, andwhy. Earth-Science Reviews, 98, 123–170.Glikson, A. Y., Eggins, S., Golding, S., Haines,P. W., Iasky, R. P., Mernagh, T. P., Mory, A.J., Pirajno, F. Uysal, I. T., 2005. AustralianJournal of Earth Sciences, 52, 555-573.Glikson, A. Y., Mory, A. J., Iasky, R. P., Pirajno.F., Golding, S. D., Uysal, I. T. 2005. AustralianJournal of Earth Sciences, 52, 545–553.Glikson, A. Y., Jablonsky, D., Westlake, S.,2010. Australian Journal of Earth Sciences,57, 411–430.Grieve, R.A.F., Pilkington, M., 1996. AustralianGeological Survey Organisation Journal ofAustralian Geology and Geophysics, 16, No.4, 399-420.Grieve, R.A.F., Langenhorst, F., Stöffler, D.,1996. Meteoritics and Planetary Science, 31,6–35.Iasky, R. P., Glikson, A. Y. 2005. AustralianJournal of Earth Sciences, 52. 575–586Macdonald, F.A., Bunting, J.A., Cina, S., 2003.Earth and Planetary Science Letters, 213,235-247McLaren, S., Dunlap, W.J., 2006. BasinResearch, 18, 189–203.Meixner, T.J., Gunn, P.J., Boucher, R.K., Yeats,A.N., Murray, L., Yeats, T.N., Richardson,L.M., Freares, R.A., 2000. South AustraliaExploration Geophysics, 31, 024-032Middleton, M.F. (1979) Bulletin of the AustralianSociety of Exploration Geophysics.Naumov, M. V. 2002. In: Impacts in PrecambrianShields, pp. 117 – 173. Springer Verlag,Berlin.Pirajno, F., 2005. Australian Journal EarthScience, 52, 4/5, 587-606.Radke, B., 2009. Geoscience Australia Record2009/25.Reimold, W.U., 1995. Earth Science Reviews,39, 247–265.Reimold, W.U., 1998. Earth Science Reviews,43, 45–47.Robertson, P.B., Dence, M.R., Vos, M.A., 1968.Shock Metamorphism of Natural Materials.Mono Book Corp, Baltimore, MD, pp. 433–452.Sandiford, M., McLaren, S., 2002. EarthPlanetary Science Letters, 204, 133-150.Spray, J.S., 1995. Geology, 23, 1119–1122.Spray, J.G., Thompson, L.M., 1995. Nature,373, 130–132.Stoffler, D., 1974. Fortschritte der Mineralogie,51, 256–289.Stoffler, D., Langenhorst, F., 1994. Meteoritics,29, 155–181.Uysal, T., Golding, S. D., Glikson A. Y., Mory A.J., Glikson, M., 2001. Earth and PlanetaryScience Letters, 192, 281 – 289.Uysal, T, Golding, S. D., Glikson, A. Y., Mory, A.J., Glikson, M., Iasky, R. P., Pirajno, 2002.Earth and Planetary Science Letters, 201,253-260.Wyborn, D., de Graaf, L., Davidson, S. andHann, S., 2004. Eastern Australian BasinsSymposium II, PESA Special Publication, 423-430.147

Australian Geothermal Conference 2010Geothermal Energy in India: Past, Present and Future PlansHarinarayana, T., National Geophysical Research Institute, Hyderabad (Council of scientific andindustrial research)AbstractIt is well known that a mong th e vari ousnew a nd rene wable e nergy sou rces,geothermal energy is known to be one ofthe cl ean en ergy without smoke an d a lsowithout e nvironmental ha zards. Altho ughit’s importa nce is realized long b ack inother countri es, it’s exploi tation is still faraway in Ind ia, mainly due to lack ofknowledge on the de ep sub surfacestructure and lack of ded icated p rogramon dee p dril ling in high pre ssure, h ightemperature conditions. Geological Surveyof India (GS I) and National Geo physicalResearch In stitute (NG RI) have madeconcerted efforts in i dentifying the seresources in different parts of our countryfor po ssible exploitati on. In thisdirection, different ge othermal regi onshave bee n investigate d in recent yearsusing dee p electroma gnetic geop hysicaltechnique, n amely the Magnetotell urics.The g eothermal regions inve stigatedinclude Pu ga in higher Himalaya s ofJammu and Kashmir, Tattapani of Su rgujadistrict in Chattisgarh, Surajkund ofJharkhand, Tapovan -Vishnugad, Loh ari-Nag-Pala, Badrin ath re gions ofUttarakhand, Kullu-Ma nali regi ons ofHimachal Prade sh. The se investigati onshave provided the quant itative dimensionsof the anom alous su bsurface st ructuresrelated to geothermal he at source. In t histalk, an ove rview of the d etails of p resentday knowl edge of g eothermal ene rgy atdifferent locations with estimated potentialbased on th e ge ophysical stu dies a ndshallow b oreholes tempe rature log s arediscussed. Con crete pl ans a nd futuredirections are also provided to develop thisuntapped resource.148

Australian Geothermal Conference 2010Aquifer heterogeneity – is it properly assessed by wellbore samplesand does it matter for aquifer heat extraction?Peter Leary*, Peter Malin, Eylon Shalev & Stephen OnachaInstitute of Earth Sciences and Engineering, University of Auckland58 Symonds Street, Auckland 1142, New Zealandp.leary@auckland.ac.nzAbstractOil field reservoir formations are sampled forporosity and permeability on the tacit assumptionthat small scale well-log and well-core data arerepresentative of the formation flow properties atarbitrary distances from the wellbore. In formalterms, this statistical assumption is valid only ifthe formation properties are adequatelycharacterised by a mean and standard deviation,or, equivalently, if variations in formationproperties are spatially uncorrelated on all scalelengths. This statistical validity condition is,however, violated by crustal rock; well-log andwell-core data are spatially correlated over a widerange of scale lengths. It is, therefore, formallywrong to assume that small scale sample meansand standard deviations adequately representlarge-scale variation of aquifer reservoir/formationproperties.As a practical matter, the formal failure of oil fieldwell-log and well-core sampling to adequatelyestimate large-scale formation flow propertyvariation is buffered by (i) the high energy densityof hydrocarbons, (ii) lack of need for largedrainage flow rates, (iii) ability to drill infill wells ifde facto well drainage volumes are too small, and(iv) ability of time-lapse seismic imaging to detectfluid substitution volumes to determine large-scaleformation flow structures that are not inferred fromsmall-scale formation sampling strategies.As an equally practical matter, however, theabove caveats do not apply to producing hotaquifer fluids: (i) geothermal energy density is farsmaller than hydrocarbon energy content; (ii) highflow rates are essential to geothermal powerproduction; (iii) infill wells are at high risk to notintersect large drainage volumes unless guided byreliable auxiliary information; (iv) time-lapse hotaquifer imaging has no fluid-substitution signal.An alternative strategy to aquifer production wellsitingbased on small-scale wellbore sampling ofthe aquifer focuses on measuring large-scaleaquifer fracture-structures. Experience withmagnetotelluric (MT) detection of in situ fracturevolumes in geothermal fields suggests that MTsurveys can form the basis for physically accuratesampling of large-scale aquifer fracture/flowstructure.Keywords: fractures, faults, porosity, permeabilityIntroduction – Treating aquifers as oilfield reservoir formationsThe following statement, made at the Bali 2010World Geothermal Congress, succinctly describesan approach to hot aquifer energy productionbased on oil/gas reservoir formationcharacterisation using wellbore samples.As part of the drilling of the petroleum wells, asignificant amount of wireline logging, coresampling and resulting petrophysical evaluationwere undertaken…... The porosity of the target….section was determined…based on wirelinelogs calibrated to porosity samples fromconventional cores and sidewall cores. The coreporosities were calibrated to measuredpermeabilities using all the cores from a largerdatabase…... Several studies…..provide insightsinto the petrophysical evaluation….and itscalibration of porosity to permeability. Using thecalibration of porosity to permeability, and thecalibrated porosity derived from wireline loggingand cores, it is thus possible to determine thepermeability….sandstone section and integratethis across the borehole to get the transmissivityor permeability metres. (de Graaf et al 2010).Parallel statements were made at the WGC2010by Clauser et al (2010) and Vogt et al (2010).The working assumption is that formation wellboredata recorded by geophysical logging tools and/orrecovered in well core adequately samples theformation properties at all relevant scales. Whileindisputably the wellbore data sample specificgeological formations, it does not follow that withina geological formation any or all importantgeophysical properties conform to a small-scalesample mean throughout the formation, or thatimportant geophysical property variations withinthe formation are confined to the formation.Rather the evidence from well-log datasystematics is precisely the opposite: variation ofgeophysical properties within a formation can besubstantial and these variations can be connectedto the enclosing crustal volumes outside theformation. Well-log systematics thus indicate thatnear-wellbore samples do not accurately assessthe degree of large-scale spatial variationexpected for in situ formation properties, and thatthe spatial distributions of formation variationscannot be adequately estimated from small-scalesampling.149

Australian Geothermal Conference 2010These general statements are illustrated by welllogand well-core data for Perth Basin formationsencountered by the 3km-deep Cockburn1 well.10 10 SONIC EXP: -1.05 -1.34Well-log and well-core sample data forPerth Basin sedimentary formationsPerth Basin formations were drilled, logged andcore-sampled by the 3km-deep Cockburn1 oilexploration well on the coast 18km southwest ofPerth (Smith 1967). Well-log data in general, andfor the 1200m thick Yarragadee aquifer inparticular, conform to well-log in situ geophysicalproperty variations observed worldwide. Figure 1shows the well-log systematics for specific aquiferformations in the Cockburn1 well sequence.Well-log power-law scaling systematicsThe Fourier power-spectra of in situ spatialvariations of rock properties measured by welllogs worldwide closely conform to a specificpower-law scaling form (Leary 2002):S(k) 1/k 1 , (1)where k is spatial frequency and S is the well-logfluctuation power at scale length k. Dependingupon the well log, the spectral scale-length rangek tends to ~3 decades in the overall 5-decadescale range ~1cycle/cm to ~1cycle/km. Highspatial frequency data at ~1cycle/cm are recordedby formation microscanner tools measuringelectric resistivity with mm-scale electrodes. Kmlongwell logs of gamma activity, acoustic velocity,neutron density, electron density and electricalresistivity logs routinely return low spatialfluctuation power data at ~1cycle/km.Well-log spectral form (1) is important for threereasons: It is power-law over all scale lengthsrelevant to reservoir performance andcrustal deformation processes;The power-law exponent is the same foressentially all in situ properties, rocktypes, and geological settings; The non-zero power-law exponentdestroys the basis for standard statisticalinferences from standard sampling.Power-law scaling of well-log spatial fluctuationsover the five-decade cm-km scale range isindisputable evidence that something beyondgeology is at work in the brittle crust. A power-lawscaling exponent that is essentially the same for arange of geologic media and settings is evidencethat power-law scaling derives from fundamentalphysical properties of rock with secondary regardto geological details at all scale lengths. Spatiallyfluctuating grain-scale fracture density is a likelycandidate for the fundamental parametercontrolling how in situ physical properties of rockvary both vertically and horizontally (Leary 2002).10 810 610 0 10 2SPATIAL FREQUENCY (CYCLES/KM)10 11 GAMMA EXP: -1.1 -1.0410 910 710 0 10 2SPATIAL FREQUENCY (CYCLES/KM)10 10 RESISTIVITY EXP: -1.23 -1.2810 810 610 0 10 2SPATIAL FREQUENCY (CYCLES/KM)Figure 1: Well-log fluctuation power-spectra for Cockburn1geological section (blue) and Yarragadee formation (black) fitto power-law trends for sonic velocity, gamma activity andresistivity data. Red line fit to entire section, green line toYarragadee section. Spectral exponent ~1.17 0.13 is nonzero,showing that fluctuations of in situ rock physicalproperties of the Cockburn1 drill site and the Yarragadeeaquifer in particular are spatially correlated rather thanspatially uncorrelated over the m-km scale range.The non-zero power-law scaling exponent in (1)means that in situ spatial fluctuations ingeophysical properties are spatially correlated atall scale lengths and hence systematically violatethe necessary condition of the central-limittheorem upon which standard geostatisticalinferences are commonly based.150

Australian Geothermal Conference 2010The general idea that small-scale sample meansand standard deviations reasonably representlarge-scale property variations within an ensembleis valid only if the ensemble property variationsare spatially uncorrelated. Fluctuations are, inturn, uncorrelated only if the associatedfluctuation power-law spectrum is ‘white’,S(k) 1/k 0 ~ constant, (2)NORMALISED FLUCTUATIONPOROSITY (blu) -- LOG(PERMEABILITY) (red)2.521.510.50-0.5-1-1.5-20 50 100 150 200SAMPLE NUMBER250at all relevant scale lengths. Figure 1 testsfluctuation power condition (2) for well-logacoustic velocity, gamma activity, and electricalresistivity data over the entire Cockburn1 section(blue) and the Yarragadee formation (black).Since the power-law exponent of each spectrumis ~1 instead of 0 over the 3 decade m-km scalerange, condition (2) for spatially uncorrelatedfluctuations in the Cockburn1 well geologicalsection is mathematically untenable. Whateverproperties of in situ rock are responsible for thevariations in well-log readings, it cannot belogically maintained that the mean and standarddeviation of small-scale sample data accuratelypredicts the scale of variations in those propertiesat arbitrary distances from the wellbore.Well-core poroperm fluctuation systematics –percolation via grain-scale fracturesAn underlying connection between in situfractures and S(k) 1/k power-spectra isplausible since spatial variations in gammaactivity of soluble radiogenic minerals, acousticvelocity and electrical resistivity are naturallyrelated to spatial variations in fracture density.That is, crustal volumes with a greater number offractures tend to have greater gamma activity,lower resistivity and lower seismic velocity.However, physically more immediate evidence forspatially variable fracture density is availablethrough the systematics of well-core porositypermeability(poroperm) spatial fluctuationsmeasured in numerous oil/gas field reservoirformations (Leary & Walter 2008).Figure 2 graphically illustrates the systematics ofporoperm spatial fluctuations for well-core datafrom tight gas reservoir formations in the CooperBasin, South Australia. The blue trace tracksvariations in well-core porosity φ and the red tracetracks variations in the logarithm of well-corepermeability as the core sequence moves alongthe well. The Figure 2 spatial fluctuation relationbetween porosity φ and logarithm of permeability can be written,φ ~ log(), (3)where φ and log() denote respectivelynormalised spatial variations in well-core values ofporosity and log(permeability) over a well-coresequence.Figure 2: Overlay of poroperm fluctuation data for tightsandstone formations in the Cooper Basin, South Australia.The blue and red traces denote zero-mean/unit-variancefluctuations in, respectively, well-core porosity and thelogarithm of well-core permeability. Cross-correlation of thetwo traces is 85% at zero-lag.High degrees of spatial cross-correlation (3) arecommon in the abundant well-core poropermsequences acquired for clastic reservoir sections.The cross-correlations have a natural explanationin terms of fluid percolation at grain-scalefractures. Consider a core-sized rock volume of Ngrain-grain contacts with intact cement bondingand no fluid percolation. Within the core,however, a number n

Australian Geothermal Conference 2010Thus, if n is the number of percolating defects in aunit volume of rock and n! is proportional to thepercolation permeability of the rock sample with npercolation defects, then empirical relation (3) iseffectively a mathematical identity, log(n!) =n(log(n) - 1). The close equivalence of (3) and (4)argues that in situ permeability is a percolationprocess in rock volumes whose physicalproperties on all scale lengths are internallydefined by spatially fluctuating populations ofgrain-scale defects consistent with power-lawscaling of well-log spectra (1).LOG(PERMEABILITY) (MDARCY)43.532.521.510.500 10 20 30 40POROSITY (%)210-1-20 20 40 60210-1-20 20 40 60 80SAMPLE NUMBER210-1-20 20 40 60210-1-20 5 10SAMPLE NUMBERFigure 3: (Upper) Composite poroperm data as traditionallypresented in oil and gas literature; a sequence ofporoperm traces are sorted by grain size from coarser onleft to finer on right. (Lower) Same poroperm datarendered in Figure 2 format; spatial fluctuation correlationsbetween porosity and log(permeability) masked in upperdisplay emerge in agreement with spatial fluctuationrelations (3) and (4).It may be useful to useful at this point to contrastthe multi-scale-length spatially-correlated fracturephenomenology of well-log fluctuations (1) andwell-core empirical fluctuations (3) and fracturefluctuationpercolation interpretation (4) with thestandard treatment of poroperm data in the oil/gasindustry literature. With reference to empiricalrelation (3), the upper plot in Figure 3 showsstandard industry presentation of poroperm datafor a sequence of size-graded well-core (coarsergrain samples on the left grade into finer grainsamples on the right). Inherent in this poropermdata presentation is the expectation that eachsample is integral into itself, with no reason tosuppose that the sample could be systematicallyrelated to neighbouring samples on any particularscale length (except, of course, by ‘random’happenstance). The lower subplots of Figure 3show, however, that latent in the poroperm data isthe empirical poroperm spatial correlation (3).The subplots render four of the five upper-plotgrain-size-graded poroperm data trends in thezero-mean/unit-variance sequence normalisationformat of Figure 2. Spatial correlation (3)between porosity and log(permeability) emergesdirectly from the obscurity of the standardporoperm data presentation.Again in line with the industry assumption thatrock samples are only ‘randomly’ related to theirneighbours, common oil industry practice uses ageneric permeability dependence on porositysuch as the Carman-Kozeny cubic expression(e.g., Dvorkin 2009; Cox et al 2001; Mavko & Nur1997), ~ φ 3 . (5)Poroperm dependency (5) is derived fromestimates of tubular flow through clusters of porespace without reference to grain-scale fractures orfracture connectivity at any scale. Suchformulations with reference only to the smallestscale lengths are consistent with spatiallyuncorrelated rock property heterogeneity (2) but,of course, make no contact with the essentiallyuniversal well-log observation (1) that in situ rockproperty heterogeneity is spatially correlated overfive decades of scale length.Well-log and well-core data thus provide clearlines of evidence that1. small-scale (wellbore) sampling ofpermeability does not accurately assesslarge scale in situ permeability variability;2. in situ fractures and fracture-controlledpermeability on all scale lengths are anessential ingredient of crustal rockheterogeneity;3. large amplitude in situ permeabilityheterogeneity is expected at large scalelengths.Yarragadee well-core poroperm fluctuationsApplying the above argument to the Cockburn1well data, Figure 4 shows the poroperm spatialfluctuation data for the complete Cockburn1 wellcoresuite in the Figure 2 format. Dotted datapoints in Figure 4 mark poroperm data for theYarragadee aquifer within the Cockburn1 wellsequence. In contrast with typical oil fieldreservoir well-core sample data tightly confined toshort intervals of oil-bearing sands, many of theCockburn1 well-core samples were taken at 100mto 200m intervals over which formation propertieschange significantly.152

Australian Geothermal Conference 2010COCKBURN1 POROPERM (DOTS=YARRAGADEE)3BLUE = | RED = LOG()210-1-20 5 10 15 20 25 30SAMPLE NOFigure 4: Cockburn1 well-core poroperm data sequencein Figure 2 format (blue = porosity, red = log(permeability)normalised to zero-mean/unit-variance). Compared withstandard oil field reservoir poroperm fluctuations in, say,Figure 2, departures from close spatial correlation aredue to well-core samples being taken at 100m intervals invarying formations.Despite the far more variable nature of theCockburn1 well rock-type and formation-typeporoperm sampling, Figure 5 shows 60% zero-lagspatial correlation (red trace) between variationsin Cockburn1-well sample porosity and samplelog(permeability). The blue trace indicates thetypical 20% level of cross-correlation excursion ofspatially-uncorrelated fluctuation sequences withspectral content of the Cockburn1 poroperm data.The 60% Cockburn1 data cross-correlation peakat zero-lag is statistically significant.XCORR(,LOG()COCKBURN1 WELL POROPERM XCORR0.80.60.40.20-0.2-0.4-150 -100 -50 0 50 100 150LAGFigure 5: (Red) Cross-correlation of resampledCockburn1 well-core poroperm data sequences. (Blue)Cross-correlation of uncorrelated random sequences withfrequency content of well-core poroperm data; 60%correlation between poroperm sequences at a specific lag(here zero) is seen to be statistically significant.Fracture heterogeneity and hot aquiferenergy productionThe foregoing discussion challenges the oil/gasindustry reservoir characterisation assumptionthat spatially sparse wellbore samples more orless represent geological formation properties atall larger scale lengths. Plentiful well-log andwell-core data instead point to in situ percolationflow processes controlled by spatially-correlatedrandom fracture networks on all scale lengths.Such random fracture networks are spatiallyerratic and effectively unpredictable from smallscalesparse sampling, leading to a degree ofgeofluid flow spatial heterogeneity consistent withthe statistics of production well drilling success.The following quote assesses the success rate ofdrilling geothermal wells at one half the successrate of drilling oil/gas wildcat wells:Given the extremely high degree of uncertaintyinvolved in well siting and design, hydrothermalexploration success rates are around 25%,estimates the GEA. That compares with aworldwide oil wildcat success rate of 45% in 2003,according to IHS Energy, a consultancy.(Petroleum Economist 2009),While a number of factors affect both productionwell success/failure rates, two factors stand out: oil and gas are far more energy rich thanhot water; to be profitable oil and gas do not have tocome of out the ground at high flow ratesbut hot water does.With the chemical energy of oil about 50MJ perlitre, and long-term oil field production averagerate ~1 litre per 4 seconds (~15 barrels of oil perday for ~5x10 5 US wells for years 1954-2006,www.eia.doe.gov/aer/txt/ptb0502.html), wellheadpower production for a typical oil well is of order12MW. In contrast, a geothermal well dischargingN litres per second of water with T o C excesstemperature produces about 4xNxT/1000 MW ofthermal power. For T = 100 o C, it requires N ~25 litre/s to produce 10MW of thermal power. Forequal wellhead power production, a geothermalwell flow rate must thus be of order 100 greaterthan an oil well.Translating geofluid flow rate into dollar-rate tocover drilling costs, and taking into account thedifferent efficiencies of electrical powerproduction, a geothermal well must flow on order300 times greater rate to produce an incomeequivalent to pure oil recovery. Allowing forproduction of water as well as oil, 90% water cutrequires a geothermal well to flow effectively 30times the rate of its oil equivalent for comparableincome to cover drilling costs.153

Australian Geothermal Conference 2010These contrasting order-of-magnitude well-flowratenumbers for hydrocarbon and geothermalpower production make it clear that effectivegeothermal production well siting demandsunderstanding the potential for flow heterogeneityof the target formation. It is not surprising that theglobal success rate for geothermal production wellsuccess is one half that of hydrocarbon wildcatwells. Within a developed hydrocarbon reservoir,the rate of infill drilling success is probablysubstantially higher than wildcat well success,giving all the more reason to be cautious aboutadopting oil field practices regarding aquiferpermeability distributions.To make aquifer energy production commerciallyviable, physical logic and practical experienceindicate that close attention needs to be paid tofinding aquifer volumes of sufficient size andfracture density that production wells can covertheir cost. To that end, we discuss severalsurveys of producing geothermal fields in which: MT data identified reservoir volumes ofsignificant aligned fracture density; production wells drilled in the MTidentifiedaligned-fracture reservoirvolumes had flow rates far exceedingthe field average.An MT approach to sounding for largescaleaquifer fracture structuresMagnetotellurics (MT) is the practice of measuringthe natural magnetic field fluctuations of theearth’s atmosphere as they reach the earth’ssurface, and at the same time and placemeasuring electric (telluric) currents induced bythe travelling magnetic fields. By measuring thenatural magnetic and induced electric fields over awide range of temporal frequencies (as high as 1-10kHz to as low as 0.1mHz), the electricalconductivity of the earth at a specific site can beinferred as a function of electromagnetic wavedepth penetration.An important aspect of MT data is that themeasurements can register systematic aamplitude differences in electrical currentsrunning along fracture trends versus electricalcurrents running across fracture trends. Sincecurrents move more easily along fracture trends,the earth appears more conductive (less resistive)along fracture trends, and less conductive (moreresistive) across fracture trends. MT surveys,thus, are sensitive to a dual phenomenologyclosely relevant to fluids and fractures: over a range of (x,y) coordinates asequences of MT stations can seek outzones in which electrical currents flowbetter in one direction than they do in theorthogonal direction; the same MT data can reveal theapproximate depth to the current-flowdirectional anomalies by noting at whichMT field frequencies the anomalies firstoccur.Figure 6: Summary of 5km MT traverse of Krafla, Iceland,geothermal field. Central figure is deduced resistivityprofile beneath the survey traverse. Peripheral plots areresistivity depth profile data in the form of measured MTfield resistivity versus MT field wavelength. Red curvesdenote data for electric currents along the traverse, bluecurves data for electric currents across the traverse. Leftends of resistivity curves are for shorter wavelengths(shallower depths), right ends for longer wavelengths(deeper depths). Divergence of blue/red curvesinterpreted as evidence that current-carrying alignedfractures run in/out of the section plane, with fracturesconcentrated in the volume denoted by the red oval at thebase of the crustal section.Figures 6-7 illustrate the dual phenomenology offracture-related MT surveys over a 5km crustalvolume enclosing the Krafla geothermal field ofcentral Iceland (Malin et al 2009; Onacha et al2010). The Krafla field sits astride the NE-SWtrendingmid-Atlantic rift system as it passesthrough central Iceland. Figures 6-7 show thatthe geothermal activity along the rift system isgreatest where a shallow NW-SE trendingtectonic fault system intersects the NE-SW rifttrend.The MT surveyed Krafla crustal volume partitionsinto more resistive rock represented as coldcolours and more conductive rock represented aswarm colours. MT surveys across the volumereturn paired sequences of resistivity versusdepth profile displayed as blue and red curves foreach MT station. The red curve measures theresistivity profile parallel to the MT traverse; theblue curve measures resistivity normal to thetraverse.Figure 6 summarises a rift-parallel SW-NW (leftright)MT traverse of the Krafla crustal volume.The leftmost survey station resistivity profile isshown in the lower-left panel. The blue and redMT profile curves do not diverge significantly asMT wavelength increases from left to right(shallow data register at left end of curve, deepdata register at right end of curve). A lack ofsystematic divergence between blue and redresistivity profile curves persists over the nextthree MT stations moving clockwise from the154

Australian Geothermal Conference 2010lower-left panel. When, however, the MT surveymoves to the vicinity of the deep conductivityanomaly represented as the red oval at depth inthe crustal section, the red and blue resistivitycurves begin to diverge with increasing MT signalwavelength. Relative to the electrical currentflowing in the plane of the traverse (red curve),the current flow in/out of the traverse planeincreases, hence the effective resistivity drops(blue curve). Except for the next station(presumably affected by the near-surface lowresistivity red zone), the blue-curve-lower-thanred-curveresistivity profile relation persiststhrough the succeeding four survey stations, thusestablishing the existence of the buried red ovalhigh conductivity structure. This structure, givenby the blue resistivity profiles, defines a NE-SWtrending fracture/fault system intersecting the riftorientedNE-SW traverse plane.Figure 7: Map MT station resistivity depth profiledistribution for the Krafla, Iceland, geothermal field as inFigure 6. The upper-left and right-hand resistivity profilesshow that electrical currents travelling NW-SW are strongalong the MT station sequence line, while the lower-leftresistivity profile shows that away from the NW-SE linethe NW-SE electrical currents are reduced. Byinference, the 3-station line locates a local fracture/fraulttrend; in the Krafla geothermal field this line collocateswith a known tectonic fracture trend.Figure 7 displays the Figure 6 resistivityphenomenology for a map distribution of MTstations. The NW-SE trend of 3 upper-left + righthandresistivity profile panels defines the fracturetrend seen as the red high-conductivity feature atdepth in Figure 6. For each of the 3 MT stations,the blue resistivity curve diverges strongly fromthe red curve, indicating enhanced ability to carrycurrent along the NW-SE trend. In contrast, thelower-left MT station sees significantly smallerresistivity profile divergence, implying that atdepth at this location there is much reducedfracture alignment to carry electrical currents atdepth. Geologically, the 3-station NW-SE MTstation trend in Figure 7 collocates with a knownregional fault system normal to the NE-SWtrending rift system.Figure 8: (Upper) Seismicity at the Olkaria, Kenya,geothermal field; the NE-SW trends lie along the EastAfrica Rift; the NW-SE trends lie along a local tectonicfault feature; (centre) MT resistivity distribution summaryin which red/yellow tints denote areas of high/lowelectrical conductivity for the NW-SE aligned fractures;(lower) production well history along NE-SW rift trend,with lefthand well in yellow tint zone a poor producer andrighthand wells in red tint zone well above averageproducers.The three parts of Figure 8 summarise a similarfault-intersection phenomenology observed in theOlkaria, Kenya, East Aftrica rift systemgeothermal field (Simiyu & Malin 2000; Onacha etal 2009). The upper plot of Figure 8 seismicallydefines two fault trends, the NE-SW trend alongthe rift, and the NW-SE trend along a locallydefined tectonic fault. The centre plot is a155

Australian Geothermal Conference 2010composite rendering of MT survey data in whichred tints mark MT sites for which NW-SE riftnormalelectrical currents are strong while yellowtints mark MT sites where such electrical currentsare weak. The lower plot summarises theproduction well history: the lefthand well drilled inthe yellow-tinted MT zone is a poor producer,while the two righthand wells drilled in the redtintedMT zone produce at double to triple the fieldaverage production rate.Summary/ConclusionsWe argue:well-log and well-core systematics showthat fracture systems are important,perhaps (probably?) crucial, conduits forgeofluid flow on all scales in all rock, withparticular reference to aquifer rockcurrently targeted for heat extraction; power-law scaling well-log spectraindicate that small-scale rock samplesfundamentally do not represent the rangeof rock property fluctuations likely tooccur at large scale lengths; in absence of utility from small-scalesampling of rock properties, and in lightof essentially unlimited fluctuations inrock properties on large scales, it islogical to consider large scale samplingof rock formations for information on insitu permeability; large scale measurements of fracturedistributions are particularly relevantwhere geofluid flow rates are essential todrill hole success;in geothermal fields, where fractures arealmost universally acknowledged tocontrol geofluid flow, MT resistivityprofiles indicate that geofluid flow isgreatest where known fracture/faulttrends intersect; enhanced productionwell flow has validated MT data as ageophysical guide to well siting.We conclude from these arguments that MTsurveys of potentially exploitable hot aquifers area plausible investment in advance of costlydrilling.ReferencesClauser C, Marquart G, Rath V & MeProRiskWorking Group (2010) MeProRisk – a Tool Boxfor Evaluating and Reducing Risks inExploration, Development, and Operation ofGeothermal Reservoirs, World GeothermalCongress, Bali April25-30 2010.Crude Oil Production and Crude Oil WellProductivity, 1954-2008www.eia.doe.gov/aer/txt/ptb0502.htmlde Graaf L, Palmer R & Reid I (2010) TheLimestone Coast Geothermal Project, SouthAustralia: a Unique Hot Sedimentary AquiferDevelopment, World Geothermal Congress, BaliApril25-30 2010.Dvorkin J (2009) Kozeny-Carman equationrevisited,pangea.stanford.edu/~jack/KC_2009_JD.pdfLeary P (2002) Fractures and physicalheterogeneity in crustal rock, in Heterogeneityof the Crust and Upper Mantle – Nature, Scalingand Seismic Properties, J. A. Goff, & K. Holliger(eds.), Kluwer Academic/Plenum Publishers,New York, 155-186.Leary PC & Walter LA (2008) Crosswell seismicapplications to highly heterogeneous tight-gasreservoirs, Special Topic on Tight Gas, FirstBreak 26, 33-39.Malin P, Leary P, Shalev E & Onacha S (2009)Joint Geophysical Imaging of PoropermDistributions in Fractured Reservoirs, GRCTransactions, Vol. 33, 4-7October 2009Mavko, G., and Nur, A., 1997, The effect of apercolation threshold in the Kozeny-Carmanrelation, Geophysics 62, 1480-1482.Onacha S, Shalev E, Malin P, & Leary P (2009)Joint Geophysical Imaging of Fluid-FilledFracture Zones in Geothermal Fields in theKenya Rift Valley, GRC Transactions Vol. 33, 4-7October 2009Onacha S, Shalev E, Malin P, Leary P &Bookman L (2010) Interpreted FractureAnomalies: Joint Imaging of GeophysicalSignals from Fluid-Filled Fracture Zones inGeothermal Fields, Proceedings WorldGeothermal Congress, Bali, Indonesia, 25-29April 2010Petroleum Economist (2009) Unlocking the valueof the world's geothermal resources,www.petroleum-economist.comSimiyu S & Malin P (2000) A “volcano seismic”approach to geothermal exploration andreservoir monitoring: Olkaria, Kenya and CasaDiablo, U.S.A. World Geothermal Congress2000, Kyoto-Tohoku Japan, 1759-1763.Smith DN (1967) Well Completion ReportCockburn No. 1 Well, Western Australia, WestAustralian Petroleum Pty. LtdVogt C, Mottaghy D, Rath V, Wolf A, Pechnig R &Clauser C (2010) Quantifying Uncertainty inGeothermal Reservoir Modeling, WorldGeothermal Congress, Bali April25-30 2010.156

Australian Geothermal Conference 2010Automatic Meshing and Construction of a 3D Reservoir System: FromVisualization towards SimulationYan Liu 1 , Huilin Xing 1,* , Zhenqun Guan 2 and Hongwu Zhang 21Earth Systems Science Computational Centre (ESSCC), The University of Queensland, St. Lucia, QLD4072, Australia;2 Department of Engineering Mechanics, State Key Laboratory of Structural Analysis of IndustrialEquipment, Dalian University of Technology, Dalian, China*Corresponding author: xing@esscc.uq.edu.auFinite Eleme nt Method (FEM) Geo computing i sincreasingly important in analysing and evaluatingenhanced g eothermal rese rvoirs. Due to thecomplexity of geologi cal objects, it is difficult tobuild a reasonable m esh fo r F EM. Thedevelopment of digital 3 D geolo gical modellingtools and ap plications (such as Ge oModeller andGoCad) makes it possible to establish geometricmodels for 3D geolo gical objects. Our wo rkfocuses on constructing a geoth ermal reservoi rsystem which transforms geometric models usedin 3D geological visualization into FEM models. A3D Geol ogical model for visuali zation i s use d asinput here, a nd the tran sformation processes areimplemented fo llowing: ( 1) g enerating atetrahedral mesh by 3D Delaun ay triangul ationmethods; (2) abstra cting boundaries o f differentgeological ob jects; (3) buil ding a well model bytetrahedral mesh refinement, which will be used infuture FEM-based simulations.Keywords: G eoModels; Delaunay Tria ngulation;3D geological modelling; FEM1. IntroductionIn the field of geoscie nce, modelling of geologicalobjects i s of great inte rest in visu alization an dsimulation. With the development of digital 3Dgeological modelling a pplications (su ch asGeoModeller and GoCad), 3D ge ological obje ctsincluding geological units, faults and ore depo sitscan b e ea sily built into GeoMo dels. Howeverthese GeoModels a re mainly use d fo rvisualization, and lack topological relations whichare crucial in simulation. Topological query [1-3] isa co mmon i ssue which depends on topologicalrelations bet ween differe nt geologi cal object s.Furthermore, in the field of analysing, topolo gicalmodels which are suitable for simulating ar eurgently requ ested. Th e Finite Eleme nt Method(FEM), whi ch is base d on the FEM mesh, is a neffective mea ns of geo computing. Wh enever 3 Dgeological o bjects a re tra nsformed int o an FEMmesh such as a tetrahedral or hexahe dral grid ,FEM can b e employed to present simulations forfuture analy sing and eva luating. 3D Delaunaytriangulation (DT) [4, 5] is one well studied 3 Dtetrahedral m esh gen eration method, whi ch canbe utilized to transfo rm geometric geologicalmodels into FEM ones. Recently, regarding to DT,scientists [1, 6, 7] have wo rked o n modellin ggeological o bjects innovatively. Ledoux[6] usedthe dual of DT name d Voron oi Diag ram (VD) torepresent an d analyze o ceanographic dataset s.Ledoux [7] also compared the VT-b ased metho dwith geo graphical info rmation syste ms (GIS) an doutlined shortcomings of GIS and merits of the VDapproach. Pouliot [1] used DT to build t opologicalrelationships between geological objects.Our work focuse s on constructing a g eothermalreservoir sy stem by transfo rming geomet ricgeological models into a FEM mesh which will beutilized in simulation by FEM. Generally speaking,the output of geologi cal modelling tools can b eformed into a s et of triangles which respect thegeometric profiles of g eological obje cts. Ta kingthese tri angular surfaces as i nput, ourtransformation pro cess is outlined a s below: I nsection 2, au tomatic DT mesh g eneration base don 3D ge ological mod elling data, includin gboundary a bstraction o f different geologicalobjects. In section 3, one of the geological objectsis chosen to illustrate how to build a well modelbased o n te trahedral m esh refinem ent, and aboundary condition will be introduced for FEM i nfuture analy sing. Finally, con clusion a nd furthe rwork are discussed in section 4.2. Topological modelling for 3Dgeological objects2.1 3D geological objects data input andrefinementIn this paper, we choose outputs of software suchas GeoModeller and GoCad as i nput data. It is aset of geological obje cts, including two fault sthrough different geolo gical units, as illu strated infigure 1. The model data are in the format of a setof triangula r facets which respe ct bou ndaries ofdifferent geol ogical obje cts. At the beginnin g ofthis file, node coordinate s are provided. And thenboundary tria ngular facets are pre sented by theindexes of nodes.157

Australian Geothermal Conference 2010To demon strate the effectivene ss of refiningboundaries o f the geol ogical m odel, a geologicalunit with larg e area but small thickn ess from ourgeological model is chosen to present ourapproach, a s illustrated in figure 3 (a). Afterperforming boundary re finement on it, largetriangular facets are b roken into smal l piece s;meanwhile, nodes o n its boun daries are denseenough to re present its g eometric feat ure in 3 Dgeological space, as shown in figure 3 (b).(a)Figure 1: Geological model from Geomodeller: (a)the whole model; (b) two faults in geological spaceAs our a pproach to extracting b oundariesbetween diffe rent g eological obje cts i s based o ndensity of bounda ry nodes, the geolo gical modelshould have a set of reasonable discretized nodeson its bo undaries. In othe r words, nodes used torepresent b oundaries ofthe geol ogical mod elshould have a specified distance between eachother. This distance is defined as the p recisionof our transforming process.The Edg e-splitting meth od, whi ch breaks atriangular e dge into two subo rdinate ones, isemployed to refine the i nitial geologi cal mod elboundaries. It is simple but effective, as illustratedin figure 2. The middl e point P of Edge AB isintroduced to split AB into two edges AP and PB.And then tria ngles linke d with AB are sep aratedinto two triangles respectively.Figure 3: Refinement of geological objectboundaries: (a) before refining; (b) after refining.2.2 3D Delaunay triangulationAfter refining the bound aries of the geologicalmodel, nod es on bo undaries app ear with areasonable density due to a spe cified . In orderto distingui sh different parts of the geolo gicalmodel and construct a topological model for FEM,a tetrahed ral generatio n method should bepresented on the cu rrent boun dary n odes of thegeological m odel. Beca use node s that are clo seto each othe r (whi ch re spect model bo undaries)are exp ected to con struct tetrahe drons, then theDelaunay triangul ation method is adopte d togenerate the tetrahe dral mesh for the geolo gicalmodel.First of all, th e definition of Delaun ay triangulationis introduced below:Figure 2: Edge-splitting method: (a) beforesplitting; (b) after splitting158

Australian Geothermal Conference 2010{ PLet k}dbe a set of points in R . The Voronoi cellPofkVis kdefined byVk { p : p Pk p Pj, j k}{ V. k}formsthe well-known Voronoi Tessellation of the entire{ Pdomain co ncerning k}. The Del aunay{ Ptriangulation of k}is defined as the du al of theVoronoi Tessellation; it wa s presented in 1930 byDelaunay, who found th e pro perty of emptycircumsphere crite rion. In the 3D d omain, theempty circumsphere criterion can be presented asbelow: for every tetrahedron there are no verticesin its circum sphere except for the four vertices ofits own. Then in 1981, Bo wyer[4] and Watson [5]proposed a method of this alg orithm, which i ssimple and efficient. The kernel of thi s method i sinserting points step by step. In order to k eep theempty circumsphere criterion, correct the topologyof the mesh after inserti ng a point. Algorithm 1shows the ba sic procedures of this method usedin this paper.Algorithm 1: Classical 3D Delaunay triangulationStep 1: Construct a super tetrahedron, whichcontains all points prepared to beinserted. This super tetrahedron boundsa convex domain.Step 2: Insert a point into this convex domain, andextract all the tetrahedrons violating theempty circumsphere criterion.Step 3: Delete tetrahedrons extracted in Step 2,which leads to construction of a convexcavity. Use the inserted point to generatetetrahedrons with the surface of thiscavity.Step 4:Repeat Steps 2-3 until all the points areinserted.Step 2: Allocate each geological object auniversal ‘colour’, and calculate its volumeby its components (tetrahedral elements).Step 3: Sort geological objects by their volumes inan increasing order.Step 4: Extend geological objects to theirneighbours one by one, merging the gapgenerated by boundary facets.Step 5: Abstract facets shared by tetrahedronswith different colours.In our geolo gical sample, due to two faults, afterabstraction the wh ole m odel (a s illu strated i nfigure 4 (a)) is se parated into 25 com ponents (infigure 4 (b)), and colours of geological objects areupdated, so they appea r with colou rs differe ntfrom o riginal ones. Boun daries b etween differe ntcomponents have a bstracted i n th e form oftriangular f acets which co ntain topologi calrelationship informatio n. Becau se th e numbe r ofboundary tri angular el ements i s to o huge tovisualize cle arly, only one of the 25 compone ntswhich is a corne r of the geologi cal model i sillustrated in the format of triang ular boundary, asshown in figure 4 (c). Furthe rmore, two faultsbreak the g eological mo del into three blocks, asshown in figure 4 (d) (e) and (f), which is useful forfurther simulation and a nalysing. With the help ofthe topologi cal model gen erated a bove, differentgeological objects can be easily identified and thegenerated tetrahedral mesh is an ideal model forFEM-based simulation.2.3 Abstra ction bou ndaries from differentgeological objectsAs 3 D DT tend s to ge nerate tetrahe drons bynodes close to each other, most of the triangulargeological surfaces will construct naturally duringthe process of DT. Mark facets which do not haveany edg es l onger tha n (which i s specified insection 2.1) boundaries. And then use Algorithm 2to abstract boun daries of different geolo gicalobjects.Algorithm 2: Abstraction of geological objectboundariesStep 1: Roughly abstract geological objects in theform of tetrahedrons by facets which aremarked as boundaries.159

Australian Geothermal Conference 2010Figure 4: Topological geological model afterabstraction based on DT: (a) the whole model; (b)components of geological model; (c) a componentvisualized in the form of grid; (d) left block ofgeological model; (e) middle block of geologicalmodel; (f) right block of geological model.3. Refining geological object andbuilding well modelIn the field of geothermal exploration, wells areessential facilities. Simulation based on FEMmethods re quests elem ents with small sizearound wells, which can enhance the accuracy ofanalysing. To meet this need, first of all, acomponent which i s suitable for drilling is chosenby experie nced en gineers (ta ke a componentwhich is far from faults for example, as i llustratedin figure 5 (a)), and then the 3D geological modelis refined in the following steps: (1 ) choo se aparticular position which is prepared to create thewell model, as shown in figure 5 (b); (2) refine thegeological object no rmally, as sho wn in figure 5(c); (3) ac cording to a spe cified precision, restrictthe element size aro und the well; (4 ) refine thewell in th e g eological m odel ba sed o n eleme ntsize restriction, as shown in figure 5 (d). Finally, adesired tetra hedral mesh whi ch is u sed in futureFEM simul ation an d an alysing i s obta ined, an dthe local tetrahedral mesh is illustrated in figure 5(e)(a)160

Australian Geothermal Conference 20104. Conclusion and future workIn this paper, automatic meshing and constructionof a ge othermal re servoir system is proposed,which focuses on tran sforming geometric modelsinto an FEM mesh. In order to make a genera linterface bet ween geolo gical softwares (su ch a sGeomodeller and GoCa d) and the geothe rmalreservoir syst em, the input data are de signed inthe form of a set of triangl es which pre sents thegeometric features of the geological model. Thenin the process of tr ansforming, in put dat arefinement, tetrahedral mesh gen eration andgeological objects bou ndaries a bstraction a reperformed step by ste p, which focu ses o ngenerating a rea sonable FEM mesh. Finally, thewell model use d in geothermal expl oration i sintroduced; automatic me sh gen eration for it isdemonstrated in detail. Our app roach presents away of using geological data for visualization toconstruct an F EM me sh for ge othermalsimulation.Further works shoul d be done to enhance th eapproach proposed in this paper: in the first place,instead of a spe cified precision , bo undaryabstraction sho uld be i mproved to adaptivelydetect small features on geological models; in thesecond pla ce, as geol ogical mod els have bee ndivided into several pa rts, both parallel meshgeneration and FEM analysis should beconsidered.Figure 5: Building well model on particular geologicalobject for FEM analysing: (a) choosing geological object;(b) well location; (c) refining geological object; (d) buildingwell model; (e) local tetrahedral elements for well.AcknowledgementsSupport is g ratefully ac knowledged by AustralianResearch Council through ARC Discovery projectDP0666203, ARC Linkage Internatio nalLX0989423 and ARC Link age project LP0560932with colla boration of G eodynamics Li mited. Theauthors are grateful to Helen Gibson of IntrepidGeoscience for creatin g the 3D ge ological modelwith Geomodeller which used as the input data inthis paper.ReferencesPouliot, J., et al., Reasoning ab out geologi calspace: Coupling 3 D G eoModels a nd topologicalqueries a s an aid to spatial data selectio n.Computers & Geoscien ces, 2008. 34(5): p. 529-541.Apel, M., From 3d geomodelling systems towards3d geo science informatio n system s: Da ta model,query fu nctionality, and data ma nagement.Computers & Geoscien ces, 2006. 32(2): p. 222-229.James, W., et al., ADVISER: Immersive ScientificVisualization Applied to Mars Research a ndExploration. Photogra mmetric En gineering &Remote Sensing, 2005. 71(10): p. 1219-1225.161

Australian Geothermal Conference 2010Bowyer, A., Computing Dirichlet tessellations. TheComputer Journal, 1981. 24(2): p. 162-166.Watson, D. F., Computi ng the n-d imensionalDelaunay tessellatio n with application to Vorono ipolytopes. The Computer Journal, 1981. 24(2): p.167-172.Ledoux, H. and C.M. Gold . A Voronoi-based mapalgebra. in Spatial Data Ha ndling—12thInternational Symposiu m on Spatial DataHandling. 2006: Springer.Ledoux, H. and C.M. Gold, Modellin g threedimensionalgeo scientific fields with the Voronoidiagram an d its dual. Internation al Journ al ofGeographical Information Science, 2008. 22(4-5):p. 547-574.162

Australian Geothermal Conference 2010Targeting Faults for Geothermal Fluid Production: Exploring for Zonesof Enhanced PermeabilityLuke Mortimer 1,2 , Adnan Aydin 3 and Craig T. Simmons 21 Hot Dry Rocks Pty Ltd, PO Box 251, South Yarra, Victoria 31412 National Centre for Groundwater Research and Training, Flinders University, GPO Box 251, Adelaide, SouthAustralia 51003 Department of Geology and Geological Engineering, University of Mississippi, PO Box 1848, MS 38677, USAluke.mortimer@hotdryrocks.comFault structures can potentially deliver increasedgeothermal fluid production by boosting the bulkpermeability and fluid storage of a productionzone. However, the hydromechanical propertiesof faults are inherently heterogeneous andanisotropic, thereby, making it challenging todistinguish between permeable and impermeablefaults. This discussion paper outlines the keyfeatures that determine fault permeability and howthe probability of locating zones of enhanced faultpermeability can be derived from preliminary faultstress state modelling. It is proposed thatpreliminary fault stress state modelling for earlystage exploration projects or in areas of unknownor complex geology can reduce the uncertaintyand risk of exploring for fault-related geothermaltargets.Keywords: geothermal, exploration, fault,permeability, in situ stress, hydromechanicalmodelling.IntroductionPermeable fault structures present an attractivegeothermal exploration target as faults have beenproven to boost substantially reservoirpermeability and fluid production at some existinggeothermal energy operations (e.g. Dixie Valley,USA; Landau, Germany). However, not all faultsare permeable. Their hydraulic properties areinherently heterogeneous and difficult tocharacterize, making it challenging to distinguishbetween faults acting as fluid conduits and fluidbarriers. This challenge is compounded when afault has no surface expression because toachieve sufficiently high production fluidtemperatures a fault typically must be intersectedat some depth where the exact hydrauliccharacter of the fault is unknown prior to drilling.Therefore, if exploring for permeable faultstructures the key questions are: (1) which fault orfault segment has the highest probability of beingpermeable?; and (2) what information is requiredto reduce the uncertainty and exploration risk priorto drill testing? To help answer these questionsthe main objectives of this discussion paper are todescribe the key features that determine faultpermeability and how the probability of locatingzones of enhanced fault permeability may bederived from preliminary fault stress statemodelling. For illustrative purposes, hypotheticalexamples of preliminary fault stress state modelsare provided.Targeting fault structures involves exploring atdepth for zones of enhanced natural in situfracture porosity and permeability to maximisegeothermal fluid production from a prospectivearea. A permeable fault target can be viewed aseither: (1) having sufficient natural in situ porosityand permeability with natural fluid rechargeoccurring at the optimal fluid temperature (e.g.Dixie Valley, USA); or (2) as a zone of enhancedporosity and permeability that may still requiresome degree of reservoir stimulation (e.g.Landau, Germany).Two critical elements considered positive for faultpermeability targets are:(1) Favourable fault orientation with respect to thein situ stress field (i.e. ‘critically stressed’); and(2) Hydraulic contact with a significant volume ofporous/fractured and permeable reservoir (i.e.fluid mass storage).For the latter, the fluid mass storage may beprovided by the fault structure itself or anadjoining and hydraulically connected rock unitsuch as thick, porous sandstone. Fluidoverpressures associated with the fault may alsobe beneficial for fluid advection along thestructure from a connected, deeper andpotentially hotter reservoir and for lowering theeffective stress state of the fault.Fault Architecture and HydrogeologyThe architecture of a fault structure can varygreatly in form from simple faults where strain isaccommodated along a narrow plane to morecomplex structures where strain is distributed overa composite zone that may include numerousfaults, small fractures, veins, breccias andcataclastic gouge. Generally, fault structures aresubdivided into two simple components being thefault core and the fault damage zone both ofwhich may vary over widths ranging fromcentimetres to hundreds of metres (Figure 1;Caine and Forster, 1999; Gudmundsson et al.,2009). These two components are distinguishableby their distinct mechanical and hydrogeologicalproperties although these are inherentlyheterogeneous and can vary significantly along afault. The fault core refers to the main fault planethat takes up most of the displacement and where163

Australian Geothermal Conference 2010the original lithology has been altered throughfault-related processes such as grain sizereduction, hydrothermal alteration and mineralprecipitation in response to mechanical and fluidflow processes (Caine and Forster, 1999). Thecharacter of the fault core zone is stronglydependent on its protolithology (Caine andForster, 1999). Fault damage zones are definedas the adjacent network of subsidiary structuresincluding small faults, fractures, veins, cleavage,pressure solution seams and folds that laterallydecrease in density away from the fault core zone(Figure 1; Caine and Forster, 1999;Gudmundsson et al., 2009)Figure 1. Illustrative schematic diagram of a fault including itsfault core and damage zone. From Gudmundsson et al.(2009).The permeability of faults can vary considerablyranging from impermeable flow barriers tosignificant flow conduits with a high degree ofspatial heterogeneity and anisotropy. Generally,faulting in low porosity, competent rock isexpected to result in an increase in fault zonepermeability whilst faulting in high porositysedimentary rocks may lead to a generaldecrease in fault zone permeability throughcommunition and porosity reduction processessuch grain-size reduction and the formation ofclay-rich fault gouge and deformation bands(Zoback, 2007; Wong and Zhu, 1999). Commonly,fault cores are of a relatively lower permeabilitythan their associated damage zones, which isattributed to porosity reducing processesoccurring within the cores (Caine et al., 2010). Forexample, at the Mirrors site, Dixie Valleygeothermal field, measured core plugpermeabilities ranged from 10 -8 m 2 (10 7 mD) infault damage zones to 10 -20 m 2 (10 -8 mD) in faultcores, however, the bulk permeability of the faultzone is on the order of 10 -12 m 2 (1000 mD) (Caineand Forster, 1997; Seront et al., 1998). In terms ofanisotropy, the permeability tensor is expected tobe at a maximum parallel to the fault, intermediatedown dip of the fault plane and at a minimumperpendicular to the fault, which allows for bothvertical and lateral flow but may limit cross-faultflow (Ferrill et al., 2004). Faults of sufficiently highpermeability can also contribute significantly toboth local and regional scale coupledgroundwater flow and heat transport viaadvection/convection processes (e.g. Bachler etal., 2003).Stress-Dependent Fault PermeabilityStress acting on a fault plane can be resolved intonormal and shear stresses, which are thecomponents of stress that act normal and parallelto a plane, respectively. In nature, these stressesare highly coupled and can cause faults toundergo reactivation and deform. The linkbetween stress, fracture deformation andpermeability is such that as fracture voidgeometries and the connectivity of a flow networkchange in response to changing in situ stress, thestorage, permeability and flow pattern is alsoexpected to change in magnitude, heterogeneityand/or anisotropy. Fracture deformation can resultin significant changes in permeability and storagebecause the ability of a fracture to transmit a fluidis extremely sensitive to its aperture. Forexample, the transmissivity of an individualfracture (T f ) idealised as an equivalent parallelplate opening can be expressed as:3(2b) gT f =(1)12Where 2b is the fracture aperture width (m), isthe fluid density (kg.m -3 ), g is gravitationalacceleration (m.s -2 ) and μ is the dynamic viscosityof the fluid (kg.m -2 .s).One key aspect of exploring for permeable faultsis the theory of stress-dependant fracturepermeability in deep-seated, fractured rocks. Thistheory is supported by studies relating tohydrocarbon and geothermal reservoirs andstudies of potential nuclear waste repository sites(e.g. Finkbeiner et al., 1997; Gentier et al., 2000;Hudson et al., 2005). The theory is that in situstress fields exert a significant control on fluid flowpatterns in fractured rocks, particularly, for rocksof low matrix permeability. For example, in a keystudy of deep (>1.7 km) boreholes, Barton et al.(1995) found that permeability manifests itself asfluid flow focused along fractures favourablyaligned within the in situ stress field, and that iffractures are critically stressed this can impart asignificant anisotropy to the permeability of afractured rock mass. Critically stressed fracturesare defined as fractures that are close to frictionalfailure within the in situ stress field (Barton et al.,1995). Specifically, the theory of stress-dependentfracture permeability predicts preferential flowoccurring along fractures that are orientedorthogonal to the minimum principal stress (σ 3 )direction (due to low normal stress), or inclined164

Australian Geothermal Conference 2010~30° to the maximum principal stress (σ 1 )direction (due to shear dilation).Frictional sliding along a plane of weakness suchas a fault occurs when the ratio of shear (τ) to theeffective normal stress (σ’ n ) equals or exceeds thefrictional sliding resistance. It is based uponAmonton’s Law, which governs fault reactivation:τ = μ.σ’ n (2)Where μ is the coefficient of friction (a rockmaterial property) and σ’ n is equal to the totalapplied normal stress resolved onto the planeminus the pore fluid pressure (i.e. σ n – P P ). Thevalue of μ has been found to typically rangebetween 0.6 and 1.0 (Byerlee, 1978; Zoback,2007). This relationship also shows that pore fluidpressure can have a significant impact as itdetermines the effective stress acting on a planeand that increasing pore pressures can destabilisea fault surface by increasing the ratio ofshear to normal stress. The coupling of thesehydromechanical (HM) processes means that fluidpressure and flow within faults is linked to tectonicstress and deformation through changes inpermeability and storage whilst tectonic stressand deformation is linked to fluid flow throughchanges in fluid pressure and effective stress(NRC, 1996).How exactly a fault will behave under an appliedstress regime depends upon many factors,however, investigating stress-dependent faultpermeability based solely on fault alignment withrespect to the in situ stress field is anoversimplification. Just as important are thegeomechanical properties of the host rock and itscontained faults. Important intact rock materialproperties include parameters such as density,bulk moduli, uniaxial compressive strength, tensilestrength, cohesion and friction angle that aretypically estimated from laboratory tests. Fracturestiffness is a function of both fracture wall surfacecontact (i.e. fracture roughness profile) and theelastic properties of the intact rock material (i.e.bulk rock moduli) where for the same givenalignment within an in situ stress field relativelystiff fractures deform less than weaker fractures.Fracture normal and shear stiffness are measuresof resistance to deformation perpendicular andparallel to fracture walls, respectively, and bothincrease with increasing effective normal stress.In general, faults tend to exhibit high stiffness ifformed within hard, competent rocks or if theybecome locked open by earlier deformationepisodes (e.g. shear dislocation) or mineral infilland cementation, and may even become stressinsensitive even if subjected to high effectivenormal stresses (Hillis, 1998; Laubach et al.,2004). In contrast, low stiffness faults can exhibita wide range of shear and closure behaviour astheir alignment with respect to σ 1 changes (Hillis,1998). Ultimately, estimates of fracture stiffnessattempt to account for more realistic fractureheterogeneity, asperity contact, deformation andtortuous fluid flow. Equations 3 & 4 belowdescribe the simplified relationship betweenfracture stiffness and fracture deformation(Rutqvist and Stephannson, 2002):Δμ n = jk n Δσ’ n (3)Δμ s = jk s Δσ s (4)Which states (a) that fracture normal deformation(Δμ n ) occurs in response to changes in effectivenormal stress (Δσ’ n ) with the magnitude ofopening or closure dependent upon fracturenormal stiffness (jk n ); and (b) that the magnitudeof shear mode displacement (Δμ s ) depends uponthe shear stiffness (jk s ) and changes in shearstress (Δσ s ).Structural permeability within faults is likely to bea transient effect as faults often become modifiedby porosity reducing processes such ashydrothermal mineralisation, hence, faultdeformation processes compete with permeabilityreduction caused by fluid flow (Sibson, 1996;Zoback, 2007). Active fault slip is typicallyepisodic and can temporarily increase thepermeability of a fault zone by as much as manyorders of magnitude (Gudmundsson, 2000).Therefore, for faults to remain effective permeablestructural conduits fault deformation processesmust be at least intermittent to continual. Forexample, in the Dixie Valley geothermal field fluidproduction is sourced from high permeabilityfaults and fractures that are favourably alignedand critically stressed whilst it is inferred that theformation of fault permeability associated withactive deformation out competes permeabilitydestroying hydrothermal quartz precipitationprocesses (Zoback, 2007).Preliminary Modelling of Fault StressStatesThe numerical modelling of fault stress states haspreviously been employed by several researchersto identify zones of potential fault enhancedpermeability and fluid flux (e.g. Ferrill et al., 2004;Gudmundsson, 2000; Moeck et al., 2009; Zhangand Sanderson, 1996). In a similar methodology,this study uses the Universal Distinct ElementCode (UDEC) to simulate the coupled HMresponse of deformable faulted rock massesunder an applied in situ stress field to derivepreliminary indications of fault stress states and,by corollary, their potential permeability. UDECrepresents a rock mass as an assembly ofdiscrete rigid or deformable, impermeable blocksseparated by discontinuities (faults, joints etc) andcan reproduce fully coupled HM behaviour(Itasca, 2004). Fluid pressure and fractureconductivity is dependent upon mechanicaldeformation whilst simultaneously fluid pressures165

Australian Geothermal Conference 2010modify the mechanical behaviour of the fractures(for a comprehensive review of the UDECgoverning equations see Itasca, 2004). The 2.5DUDEC models describe a geometricalreconstruction that consist of 2D horizontal planaror vertical slices of the conceptual faulted rockmass model and incorporates the effects of the3D stress field (i.e. V , H , and h ). That is, themodels perform plane strain analyses, whichassumes that the model continues indefinitely(and uniformly) out of the plane of analysis withcomputations performed for a slice that is one unitthick (Itasca, 2004). The ultimate aim of this typeof modelling is to distinguish which fault or faultsegments in a specified area are criticallystressed and, therefore, a potential explorationdrill target. These models are designed to assistexplorers in areas of unknown or complexgeology prior to drilling, however, large modelparameter uncertainties means that the resultsare preliminary indications only i.e. a ‘probabilistic’representation of potential fault stress states.To illustrate this methodology, three hypotheticalgeological models are presented based upon thenorthern Perth Basin as an example setting. Thisinvolves a strike-slip faulting stress regime, stresstensor H > V > h equivalent to 1.25 : 1.0 : 0.75and an east-west principle horizontal stress ( H )orientation (King et al., 2008; van Ruth, 2006). Forthe purposes of this illustrative exercise, rockmass parameters (e.g. density, bulk modulus etc)were sourced from the UDEC rock propertydatabase although, where possible, these shouldbe based upon measured representative fieldsamples or at the very least global averagevalues. The most difficult part of this process isassigning fault stiffness values, particularly, asthey are expected to vary with host lithology. Asfracture stiffness is a function of wall contact area,the jk n for smooth planar surfaces canapproximate the value of the Young’s Modulus (E)whereas the jk s , for perfectly matching roughsurfaces, can approximate the value of the ShearModulus (G). At shallow depths, estimates can bederived based upon jk n ranging from 1/2 (smooth)to 1/10 (rough) the value of E and jk s ranging from1/2 (rough) to 1/10 (smooth) the value of G, whichare compatible with published data and thosederived from empirical relationships (Kulhawy,1978; Norlund et al., 1995). However, prior to drilltesting the true nature of the fault at depth isunknown. As the aim is to attempt to evaluaterelative fault stress states, possibly acrossmultiple faults and lithologies, a ‘smooth’ faultstiffness for each respective lithology was chosenalong with zero tensile strength, cohesion anddilation angle values. In theory, thesegeomechanical properties replicate the behaviourof a ‘weak’ fault plane, which allows each faultsegment to potentially deform. This is areasonable approach as most active fault zonesare inferred to be weak (Gudmundsson et al.2001; Gudmundsson et al. 2009).The three hypothetical geological model examplesare:Model 1: 4km x 4km horizontal planar model setat -3.5 km depth below the surface comprising ofa single fault with jog hosted within sandstone(Figure 2).Model 2: 5km x 5km cross-section of a listric faulthosted with a sedimentary sequence comprisingof limestone (surface-1km depth), siltstone (1km-2.5km depth), shale (2.5km-3.5km depth),sandstone (3.5km-4km depth) and granite (4km–5km depth) (Figure 3).Model 3: 4km x 4km horizontal planar model setat -3.5 km depth below the surface comprising ofa central, circular, granite batholith of 1km radiushosted within a weak shale unit plus three crosscuttingfaults of differing orientation (Figure 4).In these models, rock mass deformation wasdefined by the Mohr-Coulomb model, which is theconventional model used to represent shearfailure in rocks and soils whilst fracture behaviourwas defined by the Coulomb-Slip criterion, whichassigns elastic stiffness, tensile strength,frictional, cohesive and dilational characteristics toa fracture (Itasca, 2004). Mechanical boundarieswere defined as fixed velocity (displacement)boundaries and initial in situ and boundary fluidpore pressures are assumed hydrostatic.SandstoneFigure 2. Model 1: 4km x 4km horizontal planar contour mapof x-direction stress magnitudes highlighting theconcentration of high and low stress zones into quadrantsalong the fault jog. This highlights that although faultalignment maybe favourable stress-dependent faultpermeability can be segmented and localised. Note that theprinciple stress (σH) direction is east-west (right-left).Legend: purple, red, brown, green and yellow coloursrepresent a decreasing range of high to low stressmagnitudes, respectively.166

Australian Geothermal Conference 2010SandstoneGraniteLimestoneSiltstoneShaleFigure 3. Model 2: 5km x 5km vertical cross-section profile ofcontoured y-direction stress magnitudes highlighting alocalised low stress field perturbation closely associated withthe trace of the listric fault. This indicates low fault planestress and potentially enhanced fault permeability. Note thatthe principle stress (σH) direction is east-west (right-left).Legend: purple, red, brown, green and yellow coloursrepresent a decreasing range of high to low stressmagnitudes, respectively.Granite(circle)ShaleFigure 4. Model 3: 4km x 4km horizontal planar contour mapof x,y-direction stress magnitudes highlighting a criticallystressed fault intersection (red) which potentially representsthe location of enhanced fault permeability. This intersectionis also coincident with a low stress anomaly in the x- and y-direction. Note that the principle stress (σH) direction is eastwest(right-left). Legend: purple, red, brown, green andyellow colours represent a decreasing range of high to lowstress magnitudes, respectively.ConclusionThe risks of targeting permeable faults asgeothermal reservoirs include: (1) a relatively highpermeability structure may result in fluid pathwayshort-circuiting and accelerated rates of reservoirthermal drawdown; (2) multiple fault structureswith varying amounts of displacement maytruncate and compartmentalise a reservoirthereby reducing accessible reservoir volume;and (3) the targeted fault structure may still behydraulically sealed and/or stress insensitive as aresult of other competing natural processes.These risks can be partially mitigated throughinterpretation of good quality seismic reflectiondata and with direct drill testing. The fault stressstate models shown in this discussion paper aredeliberately simplistic but specifically designed todemonstrate how this method can be used toidentify zones of potentially enhanced faultpermeability prior to any drill testing in areas ofunknown or complex geology. This method maybe of benefit to the Australian geothermalexploration sector as target depths are typically inexcess of 3 km depth below the surface wherenatural in situ porosity and permeability aretypically low and there is a general paucity ofdata.Nevertheless, it is important to note that theresults of such preliminary models can onlyindicate the probability of encountering enhancedfault permeability and that these 2D modelssimplify the 3D reality. It could be argued thatsimply evaluating targets on structural alignmentrelationships within the in situ stress field alonemight be sufficient, however, in complex areasthis would neglect the influence of features suchas multiple rock competency and fault stiffnesscontrasts and fault intersections on perturbing thelocal stress field. The results of these numericalmodels are only as accurate as the quality of theinput data and this particular methodology caninclude a significant amount of model parameteruncertainty (e.g. fault stiffness, estimated orinferred stress field etc). Therefore, thismethodology should be viewed as just one toolthat can form part of a broader exploration riskmanagement strategy.ReferencesBachler, D., Kohl, T., and Rybach, L., 2003,Impact of graben-parallel faults on hydrothermalconvection-Rhine Graben case study: Physicsand Chemistry of the Earth, 28, 431-441.Barton, C.A., Zoback, M.D. and Moos, D., 1995,Fluid flow along potentially active faults incrystalline rock: Geology, 23, 683-683.Byerlee, J.D., 1978, Friction of rocks: Pure andApplied Geophysics, 116, 615-626.167

Australian Geothermal Conference 2010Caine, J.S. and Forster, C.B., 1997, Fault zonearchitecture and fluid flow. An example from DixieValley, Nevada: Proceedings 22 nd Workshop onGeothermal Reservoir Engineering, StanfordUniversity, Stanford, California.Caine, J.S. and Forster, C.B., 1999, Fault zonearchitecture and fluid flow: insights from field dataand numerical modelling, Faults and SubsurfaceFluid Flow in the Shallow Crust: AmericanGeophysical Union, Geophysical Monograph 113,101-127.Caine, J.S., Bruhn, R.L. and Forster, C.B., 2010,Internal structure, fault rocks and inferencesregarding deformation, fluid flow, andmineralisation in the seismogenic Stillwaternormal fault, Dixie Valley, Nevada: Journal ofStructural Geology. In Press.Ferrill, D.A., Sims, D.W., Waiting, J.W. Morris,A.P., Franklin, N.M. and Schultz, A.L., 2004,Structural framework of the Edwards Aquiferrecharge zone in south-central Texas: GSABulletin, 116 (3/4), 407-418.Finkbeiner, T., Barton, C.A. and Zoback, M.D.,1997, Relationships among in-situ stress,fractures and faults, and fluid flow: MontereyFormation, Santa Maria Basin, California: AAPGBulletin, 81 (12), 1975-1999.Gentier, S., Hopkins, D., and Riss, J., 2000. Roleof Fracture Geometry in the Evolution of FlowPaths under Stress: In Faybishenko, B.,Witherspoon, P.A and Benson, S.M. (eds),Dynamics of Fluids in Fractured Rocks. AGUGeophysical Monograph 122, 169-184.Gudmundsson, A., 2000, Active faults andgroundwater flow: Geophysical Research Letters,27 (18), 2993-2996.Gudmundsson, A., Berg, S.S, Lyslo, K.B. andSkurtveit, E., 2001, Fracture networks and fluidtransport in active fault zones: Journal ofStructural Geology, 23, 343-353.Gudmundsson, A., Simmenes, T.H., Larsen, B.and Philipp, S.L., 2009, Effects of internalstructure and local stresses on fracturepropagation, deflection and arrest in fault zones:Journal of Structural Geology, In Press.Hillis, R.R., 1998. The influence of fracturestiffness and the in situ stress field on the closureof natural fractures: Petroleum Geoscience, 4, 57-65.Hudson, J. A., Stephansson, O., and Anderson,J., 2005. Guidance on numerical modelling ofthermo-hydro-mechanical coupled processes forperformance assessment of radioactive wasterepositories: International Journal of RockMechanics and Mining Sciences, 42, 850-870.Itasca, 2004. UDEC 4.0 Theory and Background.Itasca Consulting Group Inc., Minnesota.King, R.C., Hillis, R.R., and Reynolds, S.D.,2008, In situ stresses and natural fractures in theNorthern Perth Basin, Australia: AustralianJournal of Earth Sciences, 55 (5), 685-701.Kulhawy, F. H. and Goodman, R. E., 1980,Design of Foundations on Discontinuous Rock:Proceedings of the International Conference onStructural Foundations on Rock. InternationalSociety for Rock Mechanics, 1, 209-220.Laubach, S. E., Olsen, J. E., and Gale, J. F. W.,2004. Are open fractures necessarily aligned withthe maximum horizontal stress?: Earth andPlanetary Science Letters, 222, 191-195.Moeck, I., Kwiatek, G. and Zimmermann, G.,2009, Slip tendency analysis, fault reactivationpotential and induced seismicity in a deepgeothermal reservoir: Journal of StructuralGeology, 31, 1174-1182.National Research Council (NRC), 1996. RockFractures and Fluid Flow. ContemporaryUnderstanding and Applications. NationalAcademy of Sciences, Washington D.C.Nordlund E, Radberg G, Jing L., 1995,Determination of failure modes in jointed pillars bynumerical modelling: In Fractured and jointed rockmasses. Balkema, Rotterdam, pp 345–350Rutqvist, J. and Stephansson, O., 2003. The roleof hydromechanical coupling in fractured rockengineering: Hydrogeology Journal, 11, 7-40.Seront, B., Wong, T-F., Caine, J.S., Forster, C.B,Bruhn, R.L., and Fredrich, J.T., 1998, Laboratorycharacterisation of hydromechanical properties ofa seismogenic normal fault system: Journal ofStructural Geology, 20, 865-881.Sibson, R.H., 1996, Structural permeability offluid-driven fault-fracture meshes: Journal ofStructural Geology, 18 (8), 1031-1042.van Ruth PJ, 2006, Geomechanics: Vlaming Sub-Basin, Western Australia: CO2CRC ReportNumber RPT06-0043.Wong, T-f. and Zhu, W., 1999, Brittle faulting andpermeability evolution: hydromechanicalmeasurement, microstructural observation, andnetwork modeling, Faults and Subsurface FluidFlow in the Shallow Crust: American GeophysicalUnion, Geophysical Monograph 113, 83-99.Zoback, M. D., 2007. Reservoir Geomechanics.Cambridge University Press, New York, pp.449.Zhang, X. and Sanderson, D.J., 1996, Numericalmodelling of the effects of fault slip on fluid flowaround extensional faults: Journal of StructuralGeology, 18 (1), 109-119.AcknowledgementsThe authors gratefully acknowledge the fundingand logistical support received for this researchfrom the National Centre for GroundwaterResearch and Training, the Department of Water,Land and Biodiversity Conservation SouthAustralia and the University of Hong Kong.168

Australian Geothermal Conference 2010An EGS/HSA concept for SingaporeGrahame J.H. OliverDepartment of Civil Engineering, National University of Singapore.cvegjho@nus.edu.sgAbstractHot springs are a good indicator for geothermalresources i n Singap ore. An Enginee redGeothermal System (EG S) for com mercialpower gene ration would requi re 3 km deepdirectional wells in h ot sedimenta ry a quifers(HSA) or in hot, wet, fract ured granite and thegeneration o f electricity from +1 50 o C hotwater through binary c ycle turbines with the‘waste’ wate r being recy cled do wn i njectionwells. Proof of concept for a 50 MW powe rstation might cost US$ 19 million. Developmentcosts (US$ 200 million) could be written off in6.4 years after produ ction sta rted. Largecorporations and the military could benefit fromautonomous geothe rmal power sou rces (e.g.electricity, h eat proces sing, dist rict cooling).Several n eighbourhood 5 0 MW geothermalpower statio ns could provide p art of thenational b ase load with ‘rene wable’, clean,green po wer g eneration of st rategicimportance for a country that is vie wed ashaving no natural resources.Keywords: Concept, EGS, Singapore.IntroductionSixty million peopl e living on plate boundariesaround the world already obtain their electricityfrom hydrothermal sources in young m agmaticrocks. Many commentators see the return ofthe US$100-plus barrel of crude oil with futuregas prices tracking that rise. Eighty perc ent ofSingapore’s electri city is gene rated fromimported nat ural ga s. Ge othermal exp lorationof bu ried ho t gra nite te rrains i n continentalinteriors is now attrac tive: in Aust ralia andAlaska, EGS exploration has now passed intothe develo pment pha se. Australia n StateGovernments have commi tted US$ 90 millionfor re search and demo nstration a nd anotherUS$ 7 50 mill ion h as been allo cated to wo rksprograms for the perio d 2002 a nd 20 13 (1) . InMay 2009, t he US G overnment anno unced aUS$ 350 million stim ulus boost for USgeothermal energy (2) .The main heat releasing isotopes in rocks are235 U, 238 U, 232 Th and 40 K. Granites usuallyhave more of these el ements than m ost otherrocks. Granites with more than 1 0 ppm U canbe classified as “h ot” and provided the granitehas b een allowed to hea t itself up under athermal blanket of overlying rocks, signi ficantlyhigh temperatures can build up over millions ofyears. New tech nology mean s th at boilinggeothermal water or steam is not required. Thecommercial binary cycle Chena Power Stationin Alaska bo ils R-134a refrige rant with 74 o Cgeothermal water extracted from Me sozoic hotgranite, and pro duces 2 -300 kWh at US 5cents/kWh (3) . This is i n contrast to th e US 30cents/kWh cost of diesel generators previouslyused at Ch ena (3) . The Geysers re gion inCalifornia gene rates electri city at 5cents/kWh (4) . US 5 cents/kWh i s competitiveagainst all forms of po wer gene ration exceptfor coal.Geology and hot springs ofSingaporeSingapore l ies in side the stable Asia ncontinental plate called Sundaland. The islandis co mposed mainly of Middle Tri assic I-Typegranite an d minor ga bbro, intrude d int o a kmthick bla nket of contem porary a cid v olcanicsand p artly covered by Upp er T riassic andLower Jurassic se diments (Fig. 1). Th ere a rethree confirmed hot sprin gs situated at or ne arthe coasts (although the precise locality of theone on the SW side of Pulau Te kong isuncertain b ecause of land reclamation).Seeping “steam” (sic) has been reported to mebut not conf irmed from another loca tion onSembawang Singapore Air Force base (Fig. 1).The best known hot spring at Sembawang hasbeen d rilled down to 100 m into a 50 m widefault zone in granite: temperatures of 7 0.2 o Cwere m easured (Zha o et al. 2002 ). Chemicalanalyses cl assify it a s a p otable neutralchloride spring with total dissolved solids (TDS)measured at 914 mg/l and a Cl content of 43 1mg/l (Zhao et al. 2002). I have applied variousgeochemical thermom eters (li sted in Bowen1986) using Si, Na, K, Ca concentrations listed169

Australian Geothermal Conference 2010Geology of Singapore with hot and cold springsSingaporeterritory10 kmBukit Temahalt 164 mSteamseep?CC ColdSprings24-26 o CKR+UWC+ MFSI+TanjongLokosSembawang70 o CBoreholeSpringsHotColdP. Tekong,~70 o C?Alt 47 mLineamenttrend = 35 o EGeologyGraniteGabbroVolcs &sedsJurongFmOldAlluviumYoungAlluviumLineament trend (35 o E) drawn from geology maps andsatellite images. SI = Sentosa Island. UWC = United WorldCollege. CC = Central Catchment. MF = Mt Faber. KR =Kent RidgeN = 93Figure. 1: Geology map of Singapore (after Lee and Zhou 2009)in Zhao et al. (200 2) which i ndicateunderground rese rvoir te mperatures between122 a nd 209 o C: Na/K thermometers givetemperatures of ~160 o C.Granite be drock is not n ormally con sidered tobe permeable unless it is well fractured, jointedand faulted. I have investigate d the granitequarries a round Bu kit Te mah an d Pul au Ubinand close spaced joi nting is common. Coldsprings occur around Bukit Tema, United WorldCollege and Sentosa Island (Fig. 2).DairyFarmLittleGuilinRose diagrams of vertical joint andlineament orientationsN N12 o E35 N25 o Eo EDairyFarmLoc A,n = 41Loc An = 27LittleGuilinLocLoc B,n = 43Loc Bn = 70Figure 2: Structural data from quarriesDairy FarmSingaporelineamentmap N = 93105 o EStrongly jointed graniteThe average maximum horizontal stress (sigma1) in ~100 m boreholes in central Singaporehas a 1 3 o E orie ntation (Zh ou 2001). Myanalysis of t he stress ve ctors in the Sumatrasection of th e Aust ralian Plate colli sion in th eSumatra-Java tren ch and sub duction zone tothe SW of Singapore, sugge sts t hat themaximum hor izontal stress in the ha ngingwallof the su bduction zone is oriente d 40 o E. Thepresent day stress map of SE Asia (5) , based onearthquake fault sol utions a nd boreholebreakouts and fra ctures, sho ws th at themaximum horizontal stress is orientated 40 o E inneighbouring Sumatra. Furthe rmore, mylineament map of Si ngapore, based ontopography, geolo gical m ap and satell ite datashows a very strong NE/ SW trend (3 5 o E, seeFigs. 1 and 2).The important implication is thatany joints (of whatever age) that are orientatedNE/SW could be open in this stress regime andwould b e the first to op en at de pth d uring anHF-acid fracture stimulation.I have m easured joints and fault o rientations invarious Sin gaporian granite an d gabb roquarries and there is a NE/SW correlation withthe lineament map (see Fig. 2). Th e predictionis that g ranite with a strong NE/SW orientationof ope n join t sets will preferentially channelground water in a NE/SW dire ction away fromthe wate rsheds. Fig ure 1 sho ws that theconfirmed hot springs in Singapore are indeed170

Australian Geothermal Conference 2010NE of their associated watershed maxima.Heat flow and geothermal gradientThere are no direct measurements of heat flow,thermal conductivity of rocks o r g eothermalgradient from Singapore territory. Hall & Morely(2003) estimate that base d on an extra polationof oil and ga s well data from central S umatra,the Malaysian Peninsular and the Mala y Basin,the heat flow values for Singapore are between110 mW/m 2 in the east an d 130 mW/m 2 in thewest of the island (Fig. 3).SumatraMalayBasinSingaporeFigure 3: Contoured heat flow map of part of SE Asia.After Hall and Morley (2003).Applying the Fourier Law of thermal conductionand a ssuming an ave rage heat flo w value of120 m W/m 2 and a n avera ge therm alconductivity for g ranite of 3.48 W/mK gives ageothermal gradi ent of 34.5o C/ km. Thisassumes that the earth’s crust below Singaporeis made from granite.There is a permanent spring at an alti tude of120 m at Jungle Fall s on the N si de BukitTemah, the highest hill in Singapore (164m) inthe Central Catchment A rea, indi cating a highwater ta ble (Fig. 1). The temperature of thisspring i s 24.0 o C (D. Higgitt pers . comm.). At60 m a sl o n the NE si de of Bukit Temah,another sp ring issue s out of the granit e alongthe Wallace Trail. The te mperature is 26.0 o C,an in crease of 2.0 o C o ver a 60 m drop inaltitude from Jun gle F alls, whi ch co uld beinterpreted a s bei ng cau sed by a ge othermalgradient of 3.3 o C per 100 m (or 33 o C / km).Ground water model for SingaporeThe USGS grou ndwater model (6) for islandswith an un confined a quifer surrounded byseawater can be applie d to Singapo re (Barlow2005). Assu ming seawater h as a de nsity ( Ps)of 1.025 g/cc compared with fresh water (Pf) of1.0 g/cc, then according to the Gyben-Herzbergrelation:z = P f . hP s - P fa head (h) of 120m ab ove sea lev el in thecentre of the island will d rive cold fre sh wate rdown 4.8 km b elow sea l evel ( z). T hepermanent (24 o C) spring at an altitude of 120m on Bukit Temah, (164m), indicates that sucha high water table is present. Assuming that theaverage geothermal gradient for the Si ngaporeregion is 35 o C/km (Mazlan et al. 19 99, andsee discussion above), ground water at 4.8 kmdepth will rea ch 168 + 24 = 192 o C. Beca useof the high rainfall (2.4 m /year) and the 120 mhead, the hot ground water will be driven alongthe fre sh wa ter/seawater transition a nd up tothe su rface at th e coast (F ig. 4 ). As describedbefore, this hot water is li kely to bepreferentially chann eled along NE/SWorientated joints and fractures.slCold sea waterGround water model for SingaporeWater tableRainfall = 2.4 m / yrSembawanghot spring 70 o CslCold fresh waterCold sea waterWarm waterHot waterTemperature at 4 .8 km depth = 192 o CInterfaceFigure 4: Ground water model predicts 192 o C water at4.8 km depth.The Sembawang hot spring is 3.7 m above sealevel whi ch was at sea l evel du ring t he warminterglacial periods at ~80 ka and ~125 ka(Kopp 2009). The Pulau Tekong hot spring is at0.5 m above sea level, and from photographs, itlooks to be i n the man grove tran sition, whi chexplains the high Cl a nd Mg contents (Lee andZhou 2009).171

Australian Geothermal Conference 2010Exploration of geothermal prospectsFigure 5 illustrates the th ree prospects in theSingapore geothermal pl ay: Pump te sting of100 m deep wells into a fault zone i n the BukitTemah granit e at the Se mbawang ho t sp ringprospect produced up to 400 l/min at a constant70 o C for many days (Zhao et al. 2002 ).However, thi s rate i s n ot high eno ugh tosupport a commercial p ower plant. Deeperproduction wells are required to intercept hotterwater a nd these nee d to be co upled withinjection wells to supplement the artesian flow.Cross Section Across the Singapore Geothermal PlayThree geothermal prospects (1) Sembawang Hot Spring (HGR), (2) P. TekongHot Springs (HGR, HVR), (3) Jurong (HSA, HGR)Jurong Formation(seds)WSWHGRYoungAlluviumOldAlluviumBasement granite heats up undercarapace of volcanics and sedimentsPulau TekongvolcanicsSembawang(3) ? T MF CC (1) 70 o C (2) 70 o C?HSAHGR +150 o C?v v v v v vVv v v v HVRvHGR 100 o C?Figure 5: Geothermal prospects in Singapore. Arrowsindicate a model for groundwater flow. CC = CentralCatchment Area, MF = Mount Faber recharge area, HGR= hot granite rock, HVR = hot volcanic rock, HAS =hotsedimentary aquifer.The Pulau Teko ng pro spect ha s two hotsprings, one of which is estimated to be 70 o C,the other which has be en lo st due to landreclamation. The heat source is unlikely to be inthe Triassic volcanic/sedimentary carapace butrather in the fractured granite underneath. TheJurong Fo rmation p rospect is composed ofconglomerates, s andstones, mu dstones,limestones and min or co al measures,unconformable lying o r thrust on to p of thegranite ba sement (Lee and Zh ou 2009). If theJurong is thick enough (3 – 4 km?) then it couldhave acte d a s a the rmal i nsulator to t he ‘hot’granite ba sement. If the Jurong i s p ermeableenough, then any aquifers in conta ct with the‘hot’ granite basement could have been heatedto form a hot sedimentary aquifer. Cold springson Santosa Island and fro m near United WorldCollege mi ght be recharged from the hi ghvENEground of M t Faber an d Kent Ridge. Strongjointing in open folded conglomerates atTanjong L okos (Fig. 1) i ndicates goodpermeability.Geophysical survey s: e. g. gravity, MT, TEM,3D seismics, and heat flow su rveys (i.e. a gridof 10 km sp aced 300 m deep bo re h oles) arerequired to locate 2 - 3 km de ep ex plorationwells that will test for high heat flow,geothermal gradients an d st ress ori entation.These dee p well s might intercept sig nificantlyhot artesi an water (HS A) or hot dry ro cksuitable for EGS. Age d ating of spring waterswould be u seful to model artesi an flo w rate s:i.e. 3 H/ 3 He, SF 6 and CFC’s.Proof of ConceptFor the purpose of this study, it is assumed thato150 C geothermal water will be used in abinary generation system . Obviously, viabilitywill be increased with a hotter source but thatwould requi re drilli ng deeper, m ore ex pensivewells.Proof of con cept for an EGS 50 MW powerstation, providing po wer for 50,00 0 ho mes bytapping +150 o C water at 2 km depth might costUS$ 19 milli on: i.e. two 3 km long L-shapedwells at US$ 3 million/km, plus US$ 1 million fora 1 km HF acid-fracture job. The horizontal partof the wells should be 1 km long and orientatedNW/SE s o as to maximiz e intersec tions ofNE/SW trending vertical joint and fracture sets(Fig.6).Head= 120mArtesian waterat 190 o CProof of concept = USD 19mWater table 24 o Cat surfaceProductionof hot waterGeothermal waterat 150 0 CInjection ofcold waterHot springat 70 o C2 km500 mFigure 6: Proof of concept requires sufficient connectivit ybetween injector and production wells.If this was p roved to be succe ssful, then otherpairs of L-sh aped wells could be d rilled from172

Australian Geothermal Conference 2010the same dri lling p ad until the requisite flowrates were acquired.Development CostsDevelopment cost s for a 1 MW dem onstrationpower station can be estimated by comparisonwith the 40 0 kW bina ry gene ration systemconstructed at Chena, Alaska. This uses 74 o Chot sp ring water from shallow wells (~20 0 mdeep) in granite and generates electricity usingmassed p roduced ref rigeration components.Development cost s we re US$130 0/kW (3) , i.e.US$ 1.3 million/MW. Production costs at 5 UScents/kWh and selling at a domestic rate of 12US cents/ kWh wo uld g enerate a p rofit of 7cents/kWh, i.e. US$70/ MWh. Base d on these2007 figu res, and ad ding a notional US$ 1million for deeper drilling than Alas ka, in oneyear a proof of concept 1 MW power station (for1000 homes) might make 70 x 1 x 365 x 24 =US$ 0.613 million “profit” before tax. Thedevelopment costs could b e written off in 2.3 /0.613 = 3.75 years a nd save the e quivalent of10,000 barrels of oil/yea r (with crude oil atUS$100 oil/b arrel thi s i s a saving of US$ 1million/year in carbon credits).The cost of actually drilling 3 km deepdirectional wells in grani te in Singa pore i sunknown th erefore it i s difficult to e stimatecosts and times to break even for a commercialEGS 50 MW power station. Acco rding to theU.S. DoE (2 006) the initi al co st for l arger fieldand po wer pl ants was aro und US $ 2.5 millionper installed MW in the U.S. At 5% infla tion peryear, this is US$ 3.0 milli on per M W at 2010prices. Sany o et al. (2007) estimate that EGSwould be US$ 4.0 milli on per M W at 2007prices, i.e. US$ 4.6 million at 2 010 pri ces.Estimates fo r Australi a are equivale nt to US$4.2 per M W at 2010 p rices (see App endix 1 inCooper et a l. 2010 ). An avera ge of theseestimates is US$ 3.93 million per MW.A 50 MW EG S geothermal power station mighttherefore cost 50 x 3. 93 = US$ 197 million.Assuming production costs at 5 cent s/kWh andselling at a domestic rate of 12 cents/kWhgenerates a profit of 7 cents/kWh, i.e.US$70/MWh. In one yea r a 50 M W powe rstation might make 70 x 50 x 36 5 x 2 4 = US$30.6m “profit” ex cluding taxes a nd i nterestpayments. Write off would take 197/30.6 = 6.4years: an EGS is assumed to last 30 ye ars. 50MW saves the equivalent of 0.25 million barrelsof oil/year which at US$100/ bbl = US$ 25million/year worth of carbon credits.MarketsThere is a domestic market of 5 million peopleand a so phisticated infra structure of transport(electric Mass Rapi d Tran sport system,airports), ind ustrial (coal and bioma ss firedpower stations, oil refine ries, ship and oil rigconstruction), and military installations. Each ofthese could benefit from an autonomous supplyof electri city. Alternatively, hot wate r could b eused directly in proces s industries or to powerdistrict cooling projects using absorption chillertechnology. Neighbourhood 50 MW g eothermalpower statio ns (co sting US$ 200 millio n each)could b e di stributed a round Sin gapore o nreclaimed or Government land or concentratedon Pulau Tekong, to provide b ase lo adelectricity. Th e gene rating co sts would remai nstatic whilst the co st of i mported nat ural g asvaries.Environmental ImpactAn environm ental impa ct and ri sk a ssessmentwould be a h igh priority. Geothermal energy isviewed as re newable, clean a nd g reen with asmall carbon footprint du ring the con structionphase. No doubt, a rig capable of drilli ng 3 kmlong directional wells would b e di sruptive inSingapore’s mainly urb an environm ent. All th edrilling for a power station could be conductedfrom one noise-proofed drilling pad.The surface geolo gy in dicates that there i sample pe rmeability in the strongly jointedgranite an d Jurong F ormation and it m ight bethat perme ability stimulation at depth is notrequired. Ho wever, any HF a cid-fracture j obwould create micro -seismicity. Sin gaporeexperiences micro-sei smicity fro m theJava/Sumatra su bduction zo ne and thepopulation is seismically aware. Consequently,any fra c-job would require mo nitoring with aseismic array.The inf ra-structure fo r a 1 M W ‘ proof ofconcept’ po wer station would fit into 2 or 3tennis court s. A 50 MW co mmercial po werstation could perh aps be located und ergroundon an area the si ze of three or fo ur footballfields. Pipe work a nd high tension transmissionlines would also be placed underground.Water su pply in Singap ore i s an i ssue a ndthere would be an ini tial requi rement toaugment the workin g hydrothe rmal f racturesystem with f resh or sto rm water, but not sea173

Australian Geothermal Conference 2010water which coul d cause scalin g. O nce th esystem was pre ssured up a nd if theinjector/production well conn ections we reefficient, the requi rement for augm ented waterwould drop. Air rather tha n water cooli ng mightbe install ed in tower bl ocks o n top of theunderground gene rating halls to conde nseturbine vapo ur for re cycling. Semba wang h otspring has virtually no sm ell but Pulau Tekonghot spring is reported to b e H 2 S-rich (Lee andZhou 2009); however, bina ry ge neratingsystems d o not rel ease fl uids or ga ses to th eatmosphere.SummaryHot springs i n Singa pore are go od i ndicatorsfor a g eothermal p ower resource. A 1 M Wdemonstration ge othermal po wer station in ahot spri ng area of Singapore could becommissioned for US$ 2.3 million. Constructionof a comm ercial ge othermal po wer g enerationwould involv e the drilli ng of 3 km deepdirectional b oreholes an d the gene ration ofelectricity fro m +150 o C hot water t hroughbinary cycle turbine s with the coole d ‘wa ste’water being recycl ed down injection wells.Proof of con cept for an EGS 50 MW powerstation might cost US$ 19 million. Developmentcosts (US$ 200 million) could be written off in6.4 yea rs a fter produ ction started. Severalneighbourhood EGS 50 MW geoth ermal powerstations could supply base load electricity to thenational g rid. This is ‘ renewable’, clea n, greenpower generation of strategic im portance for acountry that is viewed as h aving n o natu ralresources.AcknowledgementsThis research wa s fu nded by an internal NUSgrant to P rofessor A. Palmer, Depa rtment ofCivil Engin eering, Nati onal University ofSingapore, who is than ked for many insig htfuldiscussions.ReferencesBarlow P.M., 2005, Ground water in fresh water- salt water environments of the Atlanticcoast, United States Geological SurveyCircular, 1262, 26 pp.Bowen, R., 1986, Groun dwater, 2 nd edition,Elsevier, 427 pp.Cooper, G.T ., Beard smore, G. R., Waini ng,B.S., Pollington, N. and Driscoll, J.P., 2010,The Relative Co sts of Engi neeredGeothermal Sys tem Exploration andDevelopment in Australia. Proce edings ofthe World Geothermal Congress 2010, BaliIndonesia, 25-29 April 2010.Hall, R. and Morley C.K., 2003. Sundala ndbasins. In: Clift, P. D., Wang, P., Kuhnt, W.,Hall, R. an d Tad a R. (eds.) Continent-Ocean Inte raction within East A sianMarginal Se as. Ge ophysical Mo nographSeries 1949, 55 – 85.Kopp, R.E., Simons, F .J., Mitrovica, J.X.,Maloof, A.C. and Op penheimer, M., 2009,Probabalistic asse ssment of sea l evelduring the la st intergl acial stag e. Natu re,462, 863-867.Lee, K.W. a nd Zh ou, Y., 2009, G eology ofSingapore, 2 nd Edition, Defence Scie nceand Technology Agency, Singapore, 90pp.Mazlan, M., Abolin s, P., Mohamm ad Jamaaland Mansar, A.,1999, In : PETRONAS. ThePetroleum Geology an d Resource s ofMalaysia. Kuala Lumpar 171-217.Sanyal, S.K ., Morrow, J .W., Butler, S.J .,Robertson-Tait, A., 2007, Co st of electricityfrom e nhanced ge othermal systems.Proceedings Thirty-Seco nd Wo rkshop onGeothermal Re servoir Engi neering,Stanford.U.S. Depa rtment of Energy. (200 6).http://www1.eere.energy.gov/geothermal/m/faqs.htmlZhao, J., Ch en, C.N. and Cai, J.G., 2002, Ahydrological study of the Semba wang hotspring. Bull etin Engin eering G eologyEnvironment, 61, 59-71.Zhou, Y., 2001, Engineering Geology and RockMass P roperties of th e Bu kit Ti mahGranite. Underground S ingapore, 29 -30Proceedings, 308-314.Web sources were accessed in May, 2010.(1)http://www.searchanddiscovery.com/documents/2008/08181goldstein/index.html.(2)http://useconomy.about.com/od/candidatesandtheeconomy/a/Obama_(3)http://www.chenahotsprings.com/geothermal-power/.(4)www.geyser.com.(5)h ttp://wwwwsm.physik.unikarlsruhe.de/pub/stress_data_frame.html(6)http://pubs.usgs.gov/circ/2003/circ1262/#heading156057192.174

Australian Geothermal Conference 2010Monitoring EGS Reservoirs Using MagnetotelluricsJared Peacock 1 *, Stephan Thiel 1 , Graham Heinson 1 , Louise McAllister 21 TRaX, School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia2 Petratherm Ltd, Adelaide, Unley, SA 5061, Australia*Corresponding author: jared.peacock@adelaide.edu.auA novel time-lapse mag netotellurics (MT) methodis proposed to monitor and estimate a real extentof enha nced geothe rmal systems (E GS). Beingdirectly se nsitive to sub surface ele ctricalconductivity, MT has t he capability t o measurefluids at de pth which a re both thermally andelectrically conductive. MT measurements before,during a nd after fluid inj ection are co mpared toestimate the lateral extent of the induced reservoirat Paral ana, South A ustralia. 3 D forwardmodelling shows the resi dual MT responsebetween a system with and without con ductivefluids to be small, of the o rder of a few degree s inMT ph ase. It is crucial to obtain a ccurate andprecise MT resp onses through out the survey s inorder to produce a reliable model of the changingsubsurface. Presented a re MT re sponses befo refluid inje ction and afte r the first fluid injectio n.Furthermore, time series mea surement of theelectric an d magneti c fields du ring th e fractu ringprocess may allo w studies of EM di sturbancesgenerated by the seismi c waves, also known asthe sei smoelectric effect , which ca n provideinformation about pore fluid inclusion.Keywords: Magnetotelluric, EGS, MonitoringIntroductionEGS has th e potential to sup plement energyproduction a s the world slo wly and re sistantlyshifts to wards rene wable energy. On e importa ntconstraint of EGS is moni toring where the fluidsgo once injecte d into the hot lithology.Traditionally, reservoirs a re mo nitored with mi croseismometer a rrays, which me asure se ismicevents generated by frac tures (P hillips et al.,2002). To mography can be a pplied t o e stimatelocation of th e fracture (House, 1987), and shearwavesplitting ca n be u sed to e stimate si ze andorientation of the fracture (Elkibbi and Rial, 2005).Unfortunately, these measurements do n otprovide information about fluid inclusion.Magnetotellurics (MT)MT is a pa ssive volum etric mea surement thatexploits F araday’s la w of magn etic i nduction bymeasuring orthogon al co mponents of the Earth’ snatural m agnetic field and its sub sequentelectrical re sponse. M T is a diffusiv e metho d,which means source energy spreads as a functionof time and resolution i s prop ortional to sou rceperiod, as defined by the skin depth equation. It issensitive to variation s in e lectrical cond uctivity ofthe subsurface, which makes it a prime techniqueto monitor changes a s elect rically con ductivefluids are introduced. The fluids a re assumed tobe co nductive due to thermal an d dissolved io nenhancement and will provide a larg e contrastwith resistive ho st r ocks (Spichak and Man zella,2009).Test site: Paralana, South AustraliaThe te st site is in the northern pa rt of SouthAustralia, ab out 60 0km due north of Adelaid e,near the edge of the Flinders Range s an dParalana hot springs. Here, the thermal activity isgenerated by an innat ely large density ofradiogenic elements, where residual heat is a byproductof millions year s of decay (Brugger,2005). The 1590M a crystalline litholo gy of theuplifted Flinder Ran ges has an estim ated heatflow on the orde r of 60mW/m2 in (Neumann,2000), compared to a typical basement rock heatflow of abou t 10mW/m2. To the East of theFlinders Ranges, the hot crystalline rocks areunconformably overlaid by about 4 500m ofAdelaidian sediment s, wh ich a cts a s a lithologi cblanket keeping heat from escaping. As a resultthe he at flo w i s estimated to be 1 12mW/m2,making the area a prime location for EGS.Figure 1: Survey design centred on Paralana 2 drill hole,overlying elevation (m). Two orthogonal lines parallel toassumed principle stress directions provide dense 2Destimation with 22 stations per line. Spacing varies from200m near the centre to 1.5km near the extremities. Two offdiagonal lines with variable spacing of 250m, 500m, 1kmprovide dimensionality constraints with 6 stations each. Onesurvey encompasses a total of 58 stations, 56 plus tworemote reference stations.175

Australian Geothermal Conference 2010Petratherm and joint venture p artners Be achEnergy Ltd and Tru energy Geothermal Pty Ltdlease a 500km2 tenement where Paralana 2 hasbeen drilled to a de pth o f 4012m and ca sed to3725m. The bottom hole temperatu re has bee nestimated at 190 C and fl uids were en counteredat a depth of 3860m. Note the proposed EGS willnot be l ocated in the ba sement rock, but in thesediment p ackage directl y above, which has adominant horizontal stress field. It is proposed thatinducing permeability and po rosity in thissediment pa ckage via m echanical fra cturing willbe a chieved more e asily than i n the cry stallinebasement without losi ng much in re servoirtemperature. The propose d reservoir si ze will beabout 1 km i n the ea st-west directio n and a bout2km in th e north-south direction and about 400mthick at aro und 36 00m d epth. Perforat ion of thewell casing is pr oposed to co mmence in Aug ust,2010, followed by two separate fracturing periodsoccurring in September, 2010 and January, 2011.ProcedureA preli minary su rvey was coll ected to co mpareany changes in other surveys collected during andafter the two fracturi ng pe riods. For al l surveys,the basic layout, centred on Paralana 2, containsan ortho gonal pair of NNE-SSW and WNW-ESElines employing 22 statio ns e ach, Figure 1. T helines h ave variabl e sp acing of ~20 0m near thecentre increasing to 1.5km near the extremities toget maximal coverage at depths of inv estigation.Two off diagonal lines containing 6 stations each,again centre d on Paral ana 2, with variablespacing of 250m, 500m, 1 km, aid in est imating 3-dimensionality. Each survey will employ 56stations in to tal plus 2 remote refe rence statio ns60km and 80 km so uth of Parala na. T he su rveydesign i s based o n fo rward mod elling a ndlogistics. T he goal i s to estimate reservoir extentas a function of time by measuring changes in theMT responses.Survey 1: Base surveyA base survey was collected in Ma rch, 2010 withAuScope instrument s run by the University ofAdelaide. The sampling rate was 500Hz, with 50mdipole len gths, m easuring h orizontal magn eticfields Bx a nd By, along with horizontal ele ctricfields Ex and Ey, where the x comp onentmeasuring the field relative to magneti c north andy measurin g the field relative to mag netic east.The ve rtical magn etic co mponent Bz wasmeasured for 6 station s. Stainless st eel sta keswere used as dipoles for 51 sites, the other 5 sitesemployed Cu-CuSO4 pot s. Induction coil s wereused as m agnetometers. The average d atacollection time was 22 hours. MT responses werecalculated using BIRRP (Chave, 2004) and an inhouse interfa ce written in Python. It was foundthat 30 sites were not pre cise enou gh due to anunknown source of n oise resembling a n onperiodiccharge-recharge pattern in th e ele ctricchannels, Figure 2. The n oise does not appear tocorrelate between stations and does not appear tohave any lo cation objectivity. Different filteringtechniques were employ ed with little effect ontransfer fu nction a ccuracy. The refore, those 30sites have bee n redone using Cu -CuSO4electrodes, which seem to prod uce mo re sta bleand relia ble electri c field measurement s. Mostdata h ave lo w coherency and larger error barsaround 1-10 se conds, Fi gure 5 which is a wellknown d ead band. Also, the appa rent resi stivityand pha se p olarizations sugge st a do minant 1Dgeoelectric structure to a period of a few seconds,then a multidimensional structure after periods ofa few se conds. The site i s near a larg e workinguranium mine, Beverly in Fi gure 1, whichproduces 50Hz (and harmonics) EM signal. Thissignal is observed i n the time-freque ncyspectrogram, and i n estimate d appa rentresistivities and ph ases below p eriods of .1seconds, where the estimations are less smooth.Figure 2: Plots of time series data for two stations plotting,Bx (counts), By (counts), Ex (mV/m), and Ey (mV/m).Bottom axis is time in hours. Top: station pb44, 5km west ofParalana 2, showing a sporadic charge-recharge pattern inthe electric channels giving irregular estimates of the MTresponse. Bottom: station pb 24, 500m east of Paralana 2,showing smooth response in electric channels giving smoothestimates of the MT response.176

Australian Geothermal Conference 2010A second base survey was collected in May, 2010to redo the stations from the first survey that werenot pre cise e nough. The same parameters andlocations were u sed, but Cu-CuSO 4 electrodeswere used for all station s and no B z signals wererecorded d ue to i nstrument constraints.Unfortunately, at the same time thi s survey wasbeing collected the Epi c Pipleline, which runsparallel to th e NNE-SSW line ab out 2 km we st ofParalana 2, was being tested for corrosion. To dothis an electric 3 second squ are wave is sentdown the pi peline every 12 second s, Figure 3.Figure3: Left: time series of Bx (counts), By (counts), Ex(mV/m), Ey(mV/m) for station pb37 from second survey,200m near the pipeline. Notice the dominating 3 secondsquare wave in the electric channels and the correspondingmagnetic response.Figure 4: Plot demonstrating a method to remove periodicnoise by subtracting the average wave form for a 12 secondwindow from the data for pb42 Ex channel. Top: Ex (mV/m)with periodic noise removed plotted against time (s). Middle:filter in red compared to the raw data plotted against time(s).Bottom: Plot of the power spectrum of the raw data in blackand the filtered data in red. Notice the large notches at 12seconds and its harmonics are removed.Figure 5: MT responses of two orthogonal modes Ex/By andEy/Bx. Top are log-log plots of apparent resistivity versusperiod. Bottom plots are phase in degrees versus log period.Top: pb24 from first base survey displaying a relative smoothMT response with minimal error bars. Bottom: pb42 fromsecond survey with periodic noise removed. Again notice theplots display relative smooth MT responses with minimalerror bars, but still have a coherent biased in periods of 1-10sec. Notice in all plots the error bars are larger around 10seconds which is where a change in the MT response isexpected to occur when fluids are injected. This is also theknown dead band where there is naturally low signal power.177Because thi s si gnal i s distinctly periodic andregular, si mple time se ries analy sis can rem ovethis from the data. By taki ng an average of a 1 2second win dow the average wave fo rm can b e

Australian Geothermal Conference 2010calculated. That wave fo rm is convolved with aseries of del ta function s every 12 seconds. Alinear trend calculated fro m the data is add ed tothe filter time series to fit the data better. The filteris then subtracted from the ra w data, Figure 4.This method works fine for station s at least 1 kmaway fro m the Epic Pi peline, but b reaks d owncloser due to the larg e amplitudes of th e periodicsignal. Along with t his meth od, remotereferencing i s u sed to constrain th e tran sferfunctions, Fi gure 5. Unfortunately, with theperiodic noise remove some of the signal is al soremoved which ads to larger erro r bars around 3-12 se conds. It was deci ded to redo the inner 30stations to get the best estimate of the MTresponse as possible and reduce error bars. Thissurvey will be collected at the end of August, 2010in conjunction with the perforation test.AnalysisThe MT re sponses com monly suffer from nearsurfacegalv anic hete rogeneities that disto rt thecalculated i mpedance tensor Z. Given theexistence of a sub set with 1D characteristics i nthe data it is po ssible to fully calculate thedistortion-free region al im pedance ten sor (Bibbyet al., 2005). In the curren t case, the p resence ofthe 1D re sistivity distri bution of the sedi mentstrata all ows a relia ble e stimate of th e regi onalimpedance t ensor. Kn owledge of the distortionfreeregi onal impedance tenso r ensures a moreaccurate e stimation of th e sub surface re sistivity.The regional impedance tensor is inverted in a 2Dsense to obtain the re sistivity distri butionunderneath the profiles (Rodi and Mackie, 2001).This has been applied toConclusionsA novel method of monitori ng enha ncedgeothermal system s usi ng the magnetotelluricmethod has bee n p roposed be cause M T i ssensitive to t he bul k el ectrical condu ctivity of thesubsurface and therefore directly to the highconductivity of the i njected fluid s. The EGSsystem at Paralana wil l be indu ced in thesediment package above the hot basement rocks.The relatively high conductivity of the sedimentarybasin nea r the su rface requires hi gh quality MTresponses to ensure detection of small changes inthe MT responses before and after the injection offluids. Keep in mind the base survey to which allchanges are going to be compared to needs to beprecise and accurate. Therefore it is important tocollect the b est data p ossible eve n if it mean srepeating su rveys. Learning i s abo ut ma kingmistakes, therefore the m ore you ma ke the moreyou learn.AcknowledgmentsThe autho rs would like to thank Petrat herm an djoint venture partne rs B each Petrol eum andTruEnergy f or support o f this proje ct. GorenBoren, Jonat hon Ross, Hami sh Ada m, Trista nWurst, and Kiat Low for field work assista nce.IMER and PirSa for financial support.ReferencesBibby, H.M., Caldwell, T.G. and Brown, C.,Determinable and no n-determinable parametersof galvani c disto rtion in mag netotellurics,Geophysical Journal International, 1 63, pp 91 5-930.Brugger, J.; Long, N.; McPha il, D. C. & Plimer, I.,An active amagmati c hydrothe rmal system: theParalana hot spri ngs, No rthern Flin ders Range s,South Austra lia, Chemical Geology, 2005, 222,35-64Chave, A. D. & Th omson, D. J., Bound edinfluence magnetotelluric respon se fun ctionestimation, Geophysical Journal International,2004, 157, 988-1006Elkibbi, M. & Rial, J. A., The G eysers geothermalfield: results from shea r-wave splitting analysis ina fra ctured re servoir, Geophysical JournalInternational, 2005, 162, 1024-1035House, L. S., Locating mi croearthquakes inducedby hydrauli c fra cturing in crystal line ro ckGeophysical Research Letters, 1987, 14, 919-92Neumann, N.; Sandiford, M. & Foden, J.Regional geochemistry and continental heat flow:implications for the origin of the South Australianheat flow anomaly, Earth and Planetary ScienceLetters, 2000, 183, 107-120Phillips, W. S.; Rutledge, J. T.; Hou se, L. S. &Fehler, M. C., Induced mi croearthquake pattern sin hydro carbon an d geot hermal re servoirs: sixcase st udies, Pure and Applied Geophysics,Seismic Research Center, Los Alamos NationalLaboratory, 2002, 159, 345-369Rodi, W. a nd Mackie, R. L. Nonlin ear conjugategradients algorithm fo r 2-D mag netotelluricinversion, Geophysics, 66, pp 174-187.Spichak, V .V., and Manzella, A, 2009,Electromagnetic Soundings of Geothermal Zones,Journal of Applied Geophysics, 68, pp459-478178

Australian Geothermal Conference 2010Reservoir characterisation and numerical modelling to reduceproject risk and maximise the chance of success. An example ofhow to design a stimulation program and assess fluid productionfrom the Soultz geothermal field, FranceJoshua Koh 1 , Nima Gholizadeh-Doonechaly 1 , Abdul Ravoof 1 and Sheik S Rahman* 1 , Luke Mortimer 21 School of Petroleum Engineering, University of New South Wales, Sydney2 Hot Dry Rocks Pty Ltd, 1A Yarra Street, South Yarra, VictoriaThis paper describes a predictive, fully coupled poro-thermo-elastic numerical model that isspecifically developed to evaluate and optimise geothermal reservoir development strategies. Thispaper illustrates this via an investigation into the role of stimulation to enhance the permeability foreconomic hot water production from the deep geothermal site at Soultz-sous-Foréts (France). Thisstudy integrates four key elements: a natural fracture characterization model, a fluid flow simulationmodel, a heat transfer model and a stimulation model. Fluid flow is simulated using a finite elementbased model wherein stresses and temperature are fully coupled to the fluid flow equations. The heattransfer model is based on conductive heat transfer within the reservoir rock, convective (includingconduction) heat transfer through the reservoir fluid and time dependent thermal equilibrium betweenthe rock and fluid. The stimulation model integrates shear dilation mechanisms and stress dependentpermeabilities. Effective permeability tensors (based on discrete fracture network) are estimateddynamically and are iteratively used to determine the pore pressure, stress and thermal state withinthe reservoir.A numerical reservoir model of the Soultz-sous-Foréts field was constructed based on the dataavailable in the open literature. The reservoir was subjected to different pressure cycles to determinethe degree of permeability enhancement that could be achieved. Flow rate and correspondingpressure losses between injector and producers as well as heat recovery of the produced water werecalculated to assess the geothermal potential of the reservoir. Results of this study show that effectivetensile normal stresses from the injected cold fluid tend to increase fracture apertures, and hence,increase fracture permeability within the zone of cooling. It was also shown that for a nominalstimulation bottomhole pressure of 10,000 psi, a maximum number of shear dilation events isobserved after a stimulation period of 42 weeks for the generated Soultz fracture system and thislevel of stimulation was sufficient to obtain an economic flow rate of 100 L/s (at an acceptablepressure loss of 2000psi). Our experience also shows that prior to stimulation, flow rates at anacceptable pressure loss between the injector and producers are not sufficient for economic heatextraction. Hence these reservoirs require either stimulation to enhance reservoir permeability orinstallation of down-hole pumps to lift hot water.In addition, it has been observed that a major factor responsible for heat recovery as well as thermaldrawdown is the degree of interconnection of the fractured reservoir. The produced fluid temperaturefor the more weakly connected well GPK4 remains very high (~195 0 C) after a production period of 14years. During the same period the average matrix temperature drawdown is in the order of 20 0 C foran initial reservoir temperature of 200 0 C. For the case of a highly interconnected GPK2 well, the179

Australian Geothermal Conference 2010temperature of the produced fluid falls from an initial temperature of 200 0 C to about 130 0 C after 14years of production.IntroductionUnderstanding of the hydraulic flow and heat transfer within crystalline rocks is of significantimportance in the utilization of deep earth (at least 4 km) hot rock (HR) for geothermal reservoirsystems (Baria et al. 1999; O'Sullivan et al. 2001). The characteristics of fluid flow and heat transfer incrystalline rocks are dominated by the fracture systems found within them and in-situ stress whichaffect the properties of the fracture system significantly. Modelling of coupled hydraulic, thermal andmechanical processes can help in the understanding of the effects of in-situ stresses, induced fluidpressure and heat transfer on fluid flow and thermal drawdown of naturally fractured geothermalsystems. The long-term response of fracture systems to changes in effective stresses has beeninvestigated by several authors (McDermott and Kolditz, 2006). The change in effective stresses canbe induced by hydraulic and thermo-elastic stress alterations resulting from heat extraction from thefractured geothermal systems (rocks). Circulation of cold fluid induces temperature differences withinthe rock matrix due to heat transfer which leads to large amounts of stress energy release. In anexperimental study, O'Sullivan et al (2001) observed some degree of geo-mechanical effects ofthermal stresses on geothermal reservoir characteristics (fracture apertures and fracture permeability)and suggested a need for a long term investigation of the thermal stress effect. In particular, themechanisms that cause phenomena, such as variation of injectivity/productivity with thermaldrawdown and occurrence of seismicity need to be understood.The response of a fracture network to stimulation by injection of cool fluid is time dependent and is afunction of discrete fracture behaviour (McDermott et al. 2006; Beeler and Hickman 2003). Theindividual response of the discrete fractures within the fracture system are determined by interactionof the injected fluid and its physical characteristics such as viscosity, heat capacity and temperature.Other factors, such as the elastic response of the rock matrix and pervasive in-situ conditions,including pore pressure and rock temperature have a significant bearing on the closure and shearstate of a fracture. The effective stress profiles along fracture surfaces are functions of the far fieldtectonic stresses, induced hydraulic stresses and thermal stress release (McDermott et al. 2006)during heat extraction. Induced tensile stresses normal to the fracture surfaces from cold waterinjection can increase fracture aperture and induce new fractures (Zhou et al., 2009). Therefore,alterations of fracture system parameters, such as effective permeability need to be coupled tochanges in the pervasive conditions.Recently, Rutqvist and Stephanson (2003) provided a generic overview of models used forhydromechanical coupling at an aquifer depth and a number of well-known empirical approaches tochanges in the normal stress across fractured medium. Most notable, is the study by Beeler et al(2000) who applied elastic solutions for elliptical cracks to define the deformation responses ofelliptical fractures under tectonic loading. The effect on the surrounding rock mass however, was notincluded.180

Australian Geothermal Conference 2010Few authors have investigated the effects of induced fluid pressure on naturally fractured rock on areservoir scale. Rahman et al. (2002) developed a steady state fluid flow model to study the sheardilation effect due to fluid pressure on permeability enhancement. Shaik et al. (2009) consideredporoelastic stresses in the stimulation of a discrete fracture network by induced fluid pressure. In theaforementioned studies, however, thermal stresses were not considered.Kumar and Ghassemi (2005) developed a numerical model to study fluid pressure and permeabilitychanges in a rock mass by taking into account the effects of thermal stresses and silicaprecipitation/dissolution. The numerical model incorporated a single planar fracture connecting aninjection well and a production well as an idealized reservoir-matrix system. The temporal variation offracture aperture in response to the individual and combined effects of thermo-elastic stress and silicadissolution/precipitation were examined, however, changes in fracture aperture and pressure inresponse to non-uniform cooling were not addressed.Poroelastic and thermoelastic effects of cold-water injection into a single pre-existing fracture in a hotrock matrix was investigated by Ghassemi et al. (2007). Assumptions of plane fracture geometry wereused to derive expressions for changes in fracture aperture caused by cooling and fluid leak-off intothe matrix. The variations in aperture and pressure occurring under non-isothermal flow conditionswhen the fracture is subject to injection and heat extraction were examined. The problem wasanalytically solved for the cases pertaining to a constant fluid injection rate with a constant leak-offrate using a partially coupled formulation.As part of this study, four computationally efficient models are presented in an effort to evaluate andoptimise reservoir development strategies: a natural fracture characterization model, a fluid flowsimulation model, a heat transfer model and a stimulation model. The fully coupled thermo-poroelasticnumerical model is applied to the Soultz geothermal reservoir, France, where several deep wells havebeen drilled and stimulated in the Rhine graben (Soultz, France) to evaluate the geothermal Hot DryRock potential of the deep fractured granite reservoir. Three main boreholes, GPK2, GPK3 and GPK4which reach 5080, 5100 and 5270m depth respectively, intersect a crystalline basement overlain by1400m of Cenozoic and Mesozoic sediments. The layout of the Soultz-sous-Forets EGS site and thewell traces and locations are presented in Fig. 1 . The novelty of this approach lies in its dynamictreatment of the characteristic properties of individual fractures in simulating fluid flow and thepervasive response of the natural fracture system to cold fluid injection. The coupled numericalalgorithm is presented in Fig. 2.Modelling of the Soultz fracture networkVariably oriented and intersected fractures with different sizes and irregular patterns (i.e. all fracturecomplexities) are characterized with the use of two integrated methodologies (Tran, Chen et al. 2007).In the hybrid neuro stochastic methodology, fracture properties are generated based on theircharacterized statistical distributions. A selective random sampling scheme is used for this purpose.Continuum maps of fracture density and multifractal dimensions are integrated in the discrete model.Hierarchical simulation technique is used to deal with different degrees of heterogeneity in the181

Australian Geothermal Conference 2010reservoir. In the second methodology, hybrid neuro stochastic methodology is upgraded to handlemore complex measurements. In this methodology, an object based conditional global optimizationconcept is applied. It combines continuum distributions with statistical distributions and correlations ofdiscrete fracture properties, thus representing a great improvement over other object based stochasticsimulation as well as grid based fracture models.Based on the observed data recorded, a specified number (say 1000 fractures) of unbiased dataspecimens are generated by re-sampling the weighting field using a normalized probability method fordifferent fracture classes. Once at least 1000 fractures are created, then the fractures are sorted withrespect to their radii. The sorting is performed based on the input value for the smallest and thelargest fractures to be modeled. Once the fracture network with specific characteristics is generated,the next step is to define the fracture density in a square volume of rock of interest. The fracturedensity is the representation of the fracture area per unit volume of rock mass observed in the field.The fracture network is stochastically realized within this volume using a pseudo-random numbergenerator. This is accomplished by repeated re-sampling of earlier fracture records and placing thefractures randomly in the volume of interest until the required fracture density is reached.In this study, the statistical analysis of the fracture networks in the Soultz geothermal reservoir asperformed by Gentier et al. (2010) was used. Table 1 presents the four main fracture sets F1, F2, F3and F4, with a fifth set F5 defined for fractures not falling within the initial four main sets as“background noise” (Gentier et al., 2010). Two ~N-S striking sets are distinctly recognised,respectively W-dipping (F1) and ENE-dipping (F2). In addition, two NE-SW and NW-SE striking setsare also observed, respectively NW-dipping (F3) and SW-dipping (F4).The fracture data was used to stochastically resample the fracture network around the wells GPK2,GPK3 and GPK4 in the abovementioned manner. The generated fracture network is presented in Fig.3 for a parallelepiped volume of 2.5 km (east west) by 3.0 km (north south) by 3.5 km (vertical, 2500m– 6000m).Numerical flow simulationA finite element based poro-thermo-elastic reservoir model is developed for evaluating hot waterproduction. This model fully couples fluid flow, temperature and geomechanics. The governingequations of non-isothermal thermo-poroelasticity is expressed as follows (Kurashige, 1989;Ghassemi et al., 2002): G K ( u) G 3 2k tu p T 0(0.1)21Rp(0.2) ii Q' p T2(1.3)Tf2T (J T ) c T 0(1.4)ffff182

Australian Geothermal Conference 2010T R cTR2TR 0(1.5)where, K is the bulk modulus (MPa), G is the shear modulus (MPa), u is the displacement vector, (ux,uy) (m), α is Biot's coefficient, γ 1 is the thermal expansion coefficient of solid (K -1 ), γ 2 is the thermalexpansion coefficient of fluid (K -1 ), T R and T f are the rock and fluid temperatures respectively (K), p isthe pore pressure (Pa). Superconvergent patch recovery (Zienkiewicz et al. 1992) is used to evaluatenodal stress tensors from the numerical results of nodal displacement, u.Fracture response to stimulationThe naturally fractured medium is pressurised by injecting fluid with a view to stimulate the reservoir.In practice, the fluid pressure develops gradually and some of the fractures may start propagatingbefore the fluid pressure reaches a threshold level at which shear slippage may takes place for alarge number of fractures.The shear slippage criterion similar to Rahman et.al (2000) is applied to identify fractures that arefavorable for shear slippage and dilation. When the fracture pressure becomes higher than the givenin-situ stress effective normal stress becomes negative and fractures open and surfaces no longerremain in contact. The shear displacement, U s in such condition can be expressed as:U s (1.6)nK swhere K s is the fracture shear stiffness and nis the access shear stress.The increase in fracture aperture due to shear dilation, a s can be expressed as (Willis-Richards,Watanabe et al. 1996):aseff U s tan( dil )(1.7)The total stimulated aperture can be expressed as (Muskat 1937):aa 19eff0nref as ares(1.8)where a res is the residual aperture that usually exists at high effective stress and is considered to bezero in this study. The normal effective stress, σ eff , is the resultant normal stress arising from theeffects of induced pressure and thermal contraction at the fracture surface. The total stimulatedaperture, a can finally be expressed by mathematical manipulation as:aa0Us tan( dil) (1.9)19eff nrefThe above equation is used in the evaluation of permeability enhancement by numerical simulation offluid flow within the stimulated fractured network.183

Australian Geothermal Conference 2010Numerical algorithmThe algorithm of the design methodology is shown in Fig. 2. Initially, real field data, such as thefracture orientation, size and other fracture parameters that influence reservoir performance areacquired from the open literature.An analysis of hot water productivity in terms of well placement is performed for three wellconfigurations. Well location and distance between wells are chosen based on the Soultz EGSsystem and used to investigate common problems associated with development of EGS, such asshort circuiting and high pressure loss between injector and producer wells, and at the same time aperform a hydraulic stimulation program to provide sufficient permeability enhancement.For the configuration, the reservoir permeability is calculated based on the discrete fracture networkdata and fluid flow simulation. Then the injector is pressurised and the pore pressure, temperatureand stress tensor across the reservoir at each time step are evaluated. Shear slippage and dilationthat may occur as a result of change in local stress and pore pressure are determined and evaluated.Numerical results of stress and pore pressure are used to calculate current total aperture (accountingfor the contributions of fracture opening (mode I), shear dilation (mode II)) for every individual naturalfracture and update the permeability of the reservoir.Using these results, reservoir simulation is carried out to calculate flow rates at the producers fordiscrete values of injector-producer pressure drops as well as temperature profiles of both the rockand fluid within the reservoir.Results and DiscussionIn the current thermo-poroelastic numerical model, flow between the injection well and productionwells are assumed to be planar and is approximated through the open hole interval (3500m to 5100m)through a statistically representative trace. This is achieved by alteration of the fractal dimension,which is used to spatially distribute the fractures within a defined volume. A statisticallyrepresentative trace of the Soultz fracture network between a depth of 2500 m and 6000 m ispresented in Fig. 4 . The well placements on the plane are taken as their separation at a depth of4430m, which is midway through their depth-averaged open hole sections. The reservoir ispressurized by injecting fluid through the injection well (GPK2). To increase the injectivity, a pair of coplanarfractures of half length of 30m, is placed at the injection well. The pressurization was carriedout over a period of one year and the results recorded at different stages (8 weeks, 24 weeks, 40weeks and 52 weeks). During the pressurization, the change in fracture width for each individualnatural fracture and the resulting permeability tensor were calculated. Following stimulation of thereservoir, a flow test was carried out over a period of 14 years. During the flow test, changes infracture apertures due to thermo-poro-elastic stresses and the consequent changes in permeabilitieswere determined. Also estimated was the thermal drawdown of the Soultz EGS.184

Australian Geothermal Conference 2010Effect of stimulation time on shear dilationResults of shear dilation are presented as average percentage increase in fracture aperture anddilation events with time (see Figs. 5 and 6a-6d). From Fig. 5, it can be seen that there exists threedistinct aperture histories: 0-28 weeks, 28-42 weeks and greater than 42 weeks. Until about 28weeks, the rate of occurrence of dilation events due to induced fluid pressure of 10,000 psi(bottomhole) remains fairly constant. Following this time, the rate of occurrence increases sharplyuntil about 42 weeks, after which, no significant dilation events can be observed (a plateau of eventsis reached). This infers that for every set of reservoir and stress parameters as well as injectionschedule, a maximum level of shear dilation can be achieved.In Figs. 6a, 6b, 6c and 6d, the events of shear dilation at different stimulation times are presented.From these figures, it can be seen that by 8 weeks, distribution of shear dilation events is relativelynon-uniform. This can be attributed to rapid propagation of pressure through well interconnectedfractures.As pressure continues to build up, more fractures are dilated in the vicinity of the active shear dilationfront (see Figs. 6c and 6d). In Fig. 7, permeability enhancement in the form of the root mean square(RMS) permeability tensor, at the end of 16 weeks and 42 weeks of stimulation, respectively, arepresented. The low initial permeabilities can be attributed to high in situ closure stresses in thereservoir. After 42 weeks of stimulation, reservoir permeability has been significantly increased, bothdue to pressure induced inflation (temporary) and shear dilation (retainable) of the fracture network.The RMS permeabilities local to the major fractures are roughly an order of magnitude greater after42 weeks of stimulation than after 16 weeks.Effect of well placement on reservoir stimulation and hot water productionAs shown in Fig. 4, the two production wells, GPK2 and GPK4 are separated from the injection well,GPK3, by roughly 480m and 440m respectively. During stimulation, the production wells are keptclosed. Once a desired stimulation is achieved at the end of the stimulation period (42 weeks) thefluid was circulated over a period of 14 years. During this circulation period, production rates fromeach well, fluid velocities throughout the reservoir produced fluid temperatures and average matrixtemperature drawdown were estimated. The production parameters are listed in Table 2. Results ofhot water production over the 14 year circulation period after an initial 42 weeks of stimulation arepresented in Figs. 8 through 13.In Fig. 8, the RMS fluid velocities after 10 years of production are presented. It can be observed fromthis figure that there exists distinct fluid flow paths connecting the injector and the producers. The flowpaths connecting the injector and GPK4 however, are not as prominent (not well connected) as that ofGPK2. In addition, it can be observed that the fluid velocities are much greater for the 42 weekstimulation case than the 16 week stimulation case. Fluid velocities up to 1 x 10 -3 m/s through themajor connected pathways is reached after 42 weeks which is three times higher than that after 16weeks of stimulation. In Fig. 9 , the reservoir temperature profile after 10 years of production ispresented. It is apparent from the figure that the high flow rates experienced between the injector and185

Australian Geothermal Conference 2010producers (see Fig. 8) resulted in significant cooling of the reservoir rock along these flow paths.Once thermal breakthrough has taken place through to GPK2, the temperature of the produced fluidbegins to decline steadily. This thermal breakthrough takes place at about 8 years of production (seeFig. 10 ). From Fig. 10, it can also be observed that the produced fluid temperature drops to about140 o C in well GPK2, while in well GPK4, the produced fluid temperature remains high after the sameperiod. The variation of produced fluid temperature from different wells is caused by several factors,namely, the number of hydraulically active flow paths between the injector and producer, the hydrauliclength of these flow paths and the temperature difference between invading fluid and rock matrixalong these flow paths. In general, a high number of flow paths originating from the injector, whichspan considerable distances before intersecting the producers, tends to delay the onset of thermalbreakthrough (due a high heat-sweep efficiency). It may be noted from Fig. 10 that thermalbreakthrough begins to occur at GPK4 at around 11 years.Effect of thermally induced stresses on hot water productionIn Figs. 11a and 11b, the estimated reservoir impedances for two different degrees of stimulation arepresented. Stage I stimulation (Fig. 11a) is over 26 weeks and stage II stimulation (Fig. 11b) over 42weeks. As shown in Fig. 11b, the production rates at GPK2 increase rapidly for roughly one year andcontinue to increase but at a slower rate until about 8 years. Following this, a quasi steady stateproduction with fluctuations is reached. The initial production period of one year is characterized by asteep increase in production rate as pressure builds up through the well connected network ofstimulated fractures. The steady increase in production after one year can be attributed to the coolingof the rock matrix which induces thermal stresses (see Figs. 12a and 12b). These thermal stressesconsequently increase permeability of the fractured network, leading to changes in the pressuredistribution and hence, the flow rates.The effective stresses in the reservoir rock at early production time (3 years) and late production time(10 years) are presented in Figs. 12a (x direction) and 12b (y direction). It should be noted again thata geomechanics sign convention is adopted for stresses (positive for compression). From both plots,it is apparent that the effective stresses in reservoir at late time (10 years) are significantly lesscompressive than those at early time. The decreases in effective stresses causes fracture dilation andtherefore permeability enhancement. The continuous decrease in matrix temperature at an increasingrate can be observed in Fig. 13.Effect of level of stimulation on reservoir impedanceThe production flow rates at GPK2 were recorded for injector-producer pressure drops of 2000 psiand 1500 psi, for both stage I and II of stimulation. In Fig. 11a for the case of stage I stimulation, theproduction rates rise with time, until roughly 1.5 years, when a quasi-steady state pressure profile isattained. At this point, a 2000 psi pressure drop results in a production rate of roughly 30L/s while a1500 psi pressure drop, roughly 16 L/s. Following this, the production rates continue to rise but at amuch slower rate which was explained in the previous section. Peak production of 43 L/s is observedwith a pressure drop of 2000 psi and 23 L/s with a pressure drop of 1500 psi.186

Australian Geothermal Conference 2010In Fig. 11b, the production rate as a function of time after 42 weeks of stimulation is presented. It canbe seen that the production rates with injector-producer pressure drops of 2000 psi and 1500 psi beara similar trend to the case of stage I stimulation. The production rates begin to reach quasi-steadystate (as observed in Fig. 11b) at around 1 year production time. Similarly, thermally induced stressesincrease the production rates after this period for both cases of pressure drop. The production ratewith a pressure drop of 2000 psi peaks around 100 L/s and around 63 L/s with a pressure drop of1500 psi. The increased production rates were achieved by increasing the stimulation time from 26weeks to 42 weeks.ConclusionsIn this paper, a poro-thermo-elastic reservoir model is developed and used to study the effects ofinduced fluid pressure and thermal stresses (cooling effect) on reservoir permeability and consequentincrease in hot water production. The model is applied to the Soultz geothermal reservoir, France.The paper also investigates the effect of well placement on hot water production and thermaldrawdown. From the results of this study, the following conclusions can be drawn.It has been shown that for every geothermal system there exists an optimum injection schedule(injection pressure and duration). Further increases in stimulation effort, i.e. stimulation time for agiven stimulation pressure, does not enhance any further reservoir permeability. In this study, for abottomhole stimulation pressure of 10,000 psi, a stimulation time of 42 weeks has been foundoptimum.Fracture connectivity has a significant bearing on thermal breakthrough, produced hot watertemperature, temperature drawdown and the hot water production rate. Also demonstrated, was thenon-uniform spatial distribution of shear event occurrences with time.Thermal stresses induced during the circulation of cold water have a significant effect on the longterm production rate. As thermal drawdown of the rock matrix takes place, tensile thermal stressesare induced which allow residing fractures to dilate and enhance permeability. This graduallyincreases the fluid velocities between the injector and producer, yielding increasing production rateswith time.ReferencesBaria, R., Baumg¨artner, J., G´erard, A., Jung, R., Garnish, J., 1999. European HDR researchprogramme at Soultz-sous-Forˆets (France) 1987–1996. Geothermics 28, 655–669.Beeler, N. M., Wong, T. F. & Hickman, S. H., 2003. On the expected relationships among apparentstress, static stress drop, effective shear fracture energy, and efficiency. Bull. Seismol. Soc.Am. 93, 1381-1389.Beeler, N.M., Simpson, R.W., Hickman, S.H. and Lockner, D.A., 2000. Pore fluid pressure, apparentfriction, and Coulomb failure, J. Geophys. Res. 105, 25533– 25542.Gentier, S., Rachez, X., Ngoc, T. D. T., Peter-Borie, M. and Souque, C. (2010). "3D Flow Modelling ofthe Medium-Term Circulation Test Performed in the Deep Geothermal Site of Soultz-Sousforets(France). Proceedings Worlf Geothermal Congress 2010 - Bali, Indonesia, 25-29 April2010.187

Australian Geothermal Conference 2010Ghassemi, A. and G. Suresh Kumar (2007). "Changes in fracture aperture and fluid pressure due tothermal stress and silica dissolution/precipitation induced by heat extraction from subsurfacerocks." Geothermics 36(2): 115-140.Ghassemi, A., Nygren, A., Cheng, A., 2008. Effects of heat extraction on fracture aperture:A poro–thermoelastic analysis. Geothermics 37, 525–539.Jing, Z., J. Willis-Richards, et al. (2000). "A three-dimensional stochastic rock mechanics model ofengineered geothermal systems in fractured crystalline rock." J. Geophys. Res. 105.Kumar, G.S. and Ghassemi, A., 2005. Numerical modeling of non-isothermal quartzdissolution/precipitation in a coupled fracture–matrix system. Geothermics 34 (2005) 411–439.McDermott, C.I. and Kolditz, O., 2006. Geomechanical model for fracture deformation underhydraulic, mechanical and thermal loads. Hydrogeol. J. 14, 487–498.Muskat, M. (1937). The Flow of Homogeneous Fluids Through Porous Media, Edwards Publisher.Narayan, S. P., Z. Yang, et al. (1998). HDR reservoir development by fluid induced shear dilation: Anumerical study of the Soultz and the Cooper Basin granite rock. International Hot Dry Rock -Forum, Strasbourg, France.O’Sullivan, M.J., Pruess, K., Lippmann, M.J., 2001. Geothermal reservoir simulation. The state ofpractice and emerging trends. Geothermics 30 (4), 395–429.Rahman, M. K., M. M. Hossain, et al. (2000). "An analytical method for mixed-mode propagation ofpressurized fractures in remotely compressed rocks." International Journal of Fracture 103(3):243-258.Rahman, M. K., M. M. Hossain, et al. (2002). "A Shear-Dilation-Based Model for Evaluation ofHydraulically Stimulated Naturally Fractured Reservoirs." Int. J. for Numerical and AnalyticalMethods in Geomech. 26(5).Sanyal, S. K., E. E. Granados, et al. (2005). An Alternative and Modular Approach to EnhancedGeothermal Systems. World Geothermal Congress 2005, Antalya, Turkey.Sausse, J., Desayes, C. and Genter, A. (2007). "From geological interpretation and 3D modelling tothe characterization of the deep seated EGS reservoir of Soultz (France)." ProceedingsEuropean Geothermal Congress 2007, Unterhaching, Germany, May 30-June 1, 2007.Shaik, A. R., M. A. Aghighi, et al. (2008). "An Innovative reservoir simulator can Help Evaluate HotWater Production for Economic Development of Australian Geothermal reservoirs." GRCTransactions 32: 97-102.Shaik, A.R., Koh, J., Rahman, S. S., Aghighi, M. A. and Tran, N. H., 2009. Design and evaluation ofwell placement and hydraulic stimulation for economical heat recovery from enhancedgeothermal systems. GRC Annual Meeting October 4–7, 2009, conference proceedings.Tran, N., Z. Chen, et al. (2007). "Characterizing and modelling of fractured reservoirs with objectorientedglobal optimization." Journal of Canadian Petroleum Technology 46(3): 39-45.Willis-Richards, J., K. Watanabe, et al. (1996). "Progress toward a stochastic rock mechanics modelof engineered geothermal systems." J. Geophys. Res. 101.Zhou, X. X., Ghassemi, A. and Cheng, A. H.-D., 2009. A three-dimensional integral equationmodel for calculating poro- and thermoelastic stresses induced by cold water injection intoa geothermal reservoir. Int. J. Numer. Anal. Meth. Geomech. 2009; 33,1613–1640188

Australian Geothermal Conference 2010Figures and TablesFig. 1. Schematic geological map of the Rhine Graben and location of the Soultz-sous-Forets EGS site (a and b). Locationand traces of the Soultz deep geothermal wells; solid lines correspond to well traces (b and c). (Sausse et al., 2006)189

Australian Geothermal Conference 2010Fig 2. Flow chart describing the numerical procedure and integration between the various models.Fig. 3. Generated fracture network of the deep geothermal Soultz reservoir based on statistical data from Gentier et al.(2010).190

Australian Geothermal Conference 2010Fig. 4. A Soultz natural fracture network trace wells GPK2, GPK3 and GPK4 open holes at 4430 m. Trace plotted forfractures with radius greater than 60 m.191

Australian Geothermal Conference 2010Fig. 5. Increase in average fracture aperture (retainable) with stimulation time. Injection pressure is 10,000 psi.Fig. 6a. Cumulative shear dilation events in the fracture network after 8 weeks of stimulation (10,000 psi injection pressure)in a strike-slip stress regime. Marker sizes are proportional to the log of the event magnitudes.192

Australian Geothermal Conference 2010Fig. 6b. Cumulative shear dilation events in the fracture network after 24 weeks of stimulation (10,000 psi injection BHFP) ina strike-slip stress regime. Marker sizes are proportional to the log of the event magnitudes.Fig. 6c. Cumulative shear dilation events in the fracture network after 40 weeks of stimulation (10,000 psi injection BHFP) ina strike-slip stress regime. Marker sizes are proportional to the log of the event magnitudes.193

Australian Geothermal Conference 2010Fig. 6d. Cumulative shear dilation events in the fracture network after 52 weeks of stimulation (10,000 psi injection BHFP) ina strike-slip stress regime. Marker sizes are proportional to the log of the event magnitudes.194

Australian Geothermal Conference 2010Fig. 7. Log10 RMS effective permeability profile (magnitudes of diagonal terms) after 16 weeks (left) and 42 weeks (right) ofstimulation with 10,000 psi injection pressure, pre-production phase.195

Australian Geothermal Conference 2010Fig. 8. Log10 fluid RMS velocity profiles (magnitudes) after 10 years production for injector-producer pressure drops of 2000psi, stimulation time prior to running the production case was 16 weeks (left) and 42 weeks (right).196

Australian Geothermal Conference 2010Fig. 9. Rock temperature profile after 10 years production for a injector-producer pressure drops of 2000 psi. Blue indicatesrock at 80 o C and red indicates rock at 200 o C, stimulation time prior to running the production case was 16 weeks (left) and42 weeks (right).197

Australian Geothermal Conference 2010Fig. 10. Produced fluid temperatures vs. production time for GPK2 and GPK4. Stimulation time prior to running theproduction case was 42 weeks.Fig. 11a. GPK2 production rates for injector-producer pressure drops of 2000 psi and 1500 psi over 15 years, with areservoir stimulation time of 26 weeks.198

Australian Geothermal Conference 2010Fig. 11b. GPK2 production rates for injector-producer pressure drops of 2000 psi and 1500 psi over 15 years, with areservoir stimulation time of 42 weeks.Fig. 12a. Effective stresses in the x direction at 3 years (left) and 10 years (right) injection time. Stimulation time prior torunning the production case was 42 weeks.199

Australian Geothermal Conference 2010Fig. 12b. Effective stresses in the y direction at 3 years (left) and 10 years (right) injection time. Stimulation time prior torunning the production case was 42 weeks.200

Australian Geothermal Conference 2010Fig. 13. Average matrix temperature vs. production time for 1.2 km (North-South) by 1.0 km (East-West) block enclosingwells GPK2, GPK3 and GPK4. Stimulation time prior to running the production case was 42 weeks.Table 1. Characteristics of statistical fracture sets obtained from Gentier et al. (2010). Plane direction is measured positiveclockwise from North and dip positive downward from horizontal.201

Australian Geothermal Conference 2010Rock PropertiesYoung’s modulus (GPa) 40Poisson’s ratio 0.25Density (kg/m 3 ) 2700Fracture basic friction angle (deg) 40Shear dilation angle (deg) 2.890% closure stress (MPa) 20In situ mean permeability (m 2 ) 9.0 x 10 ‐17Fracture propertiesFractal Dimension, D 1.2Fracture density (m 2 /m 3 ) 0.12Smallest fracture radius (m) 15Largest fracture radius (m) 250Stress dataMaximum horizontal stress (MPa) 53.3Minimum horizontal stress (MPa) 78.9Fluid propertiesDensity (kg/m 3 ) 1000Viscosity (Pa s) 3 x 10 ‐4Hydrostatic fluid pressure (MPa) 34.5Injector pressure, stimulation (MPa) 68.9Injector pressure, production (MPa) 44.8Producer pressure, stimulation N/AProducer pressure, production 31.0Other reservoir dataWell radius (m) 0.1Number of injection wells 1Number of production wells 2Reservoir depth (m) 4430Table 2. Stress and reservoir data for strike-slip stress regime.202

Australian Geothermal Conference 2010Status of the Paralana 2 Hydraulic Stimulation ProgramP. W. Reid *, L. McAllister * and M. Messeiller **Petratherm Limited, 129 Greenhill Rd, Unley 5061, South AustraliaThe Paral ana Enginee red Geothe rmal Project i slocated 600 km north of t he city of Adelaid e inSouth Aust ralia (Fig ure 1). The p roject is te stingfor viable geothe rmal resou rces, within asedimentary basi n that lies immediatel y east ofknown high heat p roducing Me soproterozoicbasement rocks of the Mt Painter Regi on. In thisarea, 2D reflection sei smic su rvey data andpotential field geo physical (aeromagnetic,magneto-telluric and grav ity) data delineate amajor half graben info rmally termed the Poontan abasin. Based on the interpreted geophysical data,Petratherm postulates t hat the hi gh he atproducing basement rocks o bserved i n out crop,continue under the in sulating cover material, withthe maximum thickness of the sedim entary coverin section s of the Poo ntana Sub -basin beingmodeled at greater th an five kilom etres. Thi sfavourable arrangement of thick sedi mentsoverlying anomalously ra diogenic ba sementsuggests th at the Parala na a rea is an id eallocation to test the development of an EngineeredGeothermal System.Paralana 2 was de signed as an injector with theadvantage of sim plifying the well d esign an dallowing learning from this well to be incorporatedinto the well design and planning of the productionwell (Pa ralana 3) to maximize the ch ances ofachieving a commercial flow rate.During drilling of the l ower 8 ½” section severalfractures we re e ncountered b elow 34 00 m. H ightorques, drilling breaks, an increase in well-bor edeviation followe d by inflow of geoth ermal ove rpressurized brine s p rovide stro ng indi cations fo rthe presence of a natu ral and permeable fracturesystem between d epths of about 3 690 m a nd3864 m. Increa sed rates of penetrati on duri ngdrilling through the fract ured zones indicate achange in ro ck strength p ossibly relate d to openfractures. Shut in pre ssures indi cated a noverpressure of app roximately 3,300 p si an d th emud system required weighting up to 1 3.2 ppg tostop the inflow, and allo w drilling to continu esafely.Petratherm Limited in joint venture with a major oiland ga s (B each Energy) and po wer indu stryenergy utilities (T RUenergy) are initiall y see kingto build a 7.5 M We com mercial po werdevelopment to sup ply a local min e. In thesecond half of 2009 a deep geoth ermal wel l(Paralana 2) was drilled to 4012m. The well wa sdesigned as an inje ctor, the first of an initial twowell program to p rove circulation b etween wells.An innovative strategy for d evelopment of th eEGS re servoir i s pl anned, involving massiv ehydraulic stimulation of multiple ta rget zone swithin the sedimentary overburden. Multiple zonestimulation in creases th e cha nce of achieving acommercial flow rate which is the key commercialbarrier for EGS developments around the world.Keywords: EGS, Drilling, Fracture StimulationParalana 2 Well CompletionThe planned multizon e stimulation of Paralana 2involved the well being fully cased a nd cementedto allow better control on what intervals in the wellwould be stimulated in a cost effective manner. Inthe past, open hole hydraulic stimulations resultedin the main frac d eveloping at the zon e of leastresistance w hich is u sually at the base of th ecasing shoe (lowest pressure point) or a locationswhere large natural fractures already occur. Thisin effect leaves much of the open hole unaffectedby the stimulation pro gram. Being the first well,Figure 1: Regional Locality Map and extent of SAHFA(SAHFA modified from Neumann et. al . 2000)Due to wellbore stability problems thecharacterization of this zon e wa s limited tologging while drilling measurements down to 3740m depth. Th e final 7” casing stri ng was run to adepth of 3725 m as below this depth the well wastoo unstable to be abl e to set casing and cementin place (Figure 2).203

Australian Geothermal Conference 2010Paralana 2 GeologyThe litho- stratigraphic and structural environmentof the Paral ana Proj ect area i s co mplex andweakly con strained. T he interp retation of thestratigraphic succe ssion and litholo gy of theoverlying Phanerozoic seq uences and theircontact with the un derlying Adelai dean strata atParalana-2 i s based on the result s of drillingParalana-1B, drilled 1.5 km to west of P aralana-2.The Cambrian-Adelaidean un conformity isinterpreted at 1115m and the sequ ence b eginswith the A mberoona F m. The T apley HillFormation was e ntered a t 1612m in Paralana2and its co ntact with the unde rlying SturtianGlacials has been interpreted at about 1803m. At2855m, a 500mete rs thick haematicmetaquartzose with silice ous ce ment underliesthe tillite and grades to a litharenite with depth. Itsage is am biguous, but the metase diment is stillconsidered as Ad elaidean, within t he Sturtia nTillite or the Callanna Beds.At 3397m, a major ch ange of lithology wa sobserved during the drilling, associated with a keyunconformity. It is related to a s trong reflec torobserved on the sei smic. Below th e co ntact, thelithologies a re mai nly compo sed of dolomiti csiltstones and sand stones, inte rbedded withnumerous intervals of felsic tuffs. Several dole ritehorizons a re intercalated i n the pa ckage, from adepth of 3450m to the bottom of the hole. Zirconsextracted from the volc anic tuff returned a207 Pb/ 208 Pb age of 1585±11Ma using a LA-ICPMStechnique at the University of Adelaide. At 3910m,Paralana 2 e ntered a felsi c intru sive, confirmin ghigh h eat p roduction rate of the basement ofapproximately 10 to 12µWm-3 , and showi ng asimilar Mesoproterozoic age of 1580±10Ma.Numerous intervals of fractures were encounteredin the lower zone of the well based oninterpretation of the logs and expe rienced duringthe drilling operation. A more complex tectoni chistory an d the pre sence of dole rite in thesequence are indi cated by a chaoti c seismi csignature bel ow 3400m. The thi ck h omogenousquartzite ove rlying the ba sal sequ ence is highlycompetent with rock stre ngths rangi ng betwee n16,000 (110MPa) an d 2 9,000 p si (2 00MPa) ascalculated fro m the sonic l ogs. T his u nit may beconsidered to act a s a to p seal to the fractured,overpressured unit below.MEQ arrayThe Paralana Micro-seismic monitoring array hasbeen op erational sin ce A pril 200 8, recordin g thebackground seismi city at the Paralan aGeothermal Project site. The array ha s re centlybe up-graded to a real-time monitoring network toenable Petratherm and the joint ventu re partnersto actively re cord, analyse a nd lo catemicroseismic events duri ng the stimulation of thegeothermal reservoir. Th e gro wth of the fractu renetwork during fr acture stimul ation will bemonitored by sei smologists from the I nstitute ofEarth S cience an d En gineering, Au ckland, NewZealand. The array combines sensitive downholesondes with surface seismometers to enable th einterpretation of a wide spe ctrum of seismi cevents. All events will be analyse d, with autopickingsoftware, MIMO, develope d by theNorwegian S eismic Array (NORSA R), providi ngdata on the event location and magnitude.Figure 2: Paralana Well Completion and Geological log.Stimulation ProgramThe initial stimulation program is to be undertakenin two stages. The Stage 1 Injectivity test involvesperforation of the steel casing near the bottom ofthe Paral ana 2 well and injectio n of a sm allvolume of water to co nfirm fractu re ini tiation andpropagation. The test aim s to de rive informatio nof the insitu stre ss regime, re servoir prop ertiesand determine if the well i s already connected tofractures of the n atural overp ressured zoneencountered during drilling.The Stage 2 fracture stimulation involves injectionof large r vol ume of wate r at highe r rates. Th estimulation aims to create a fracture network andconnect to and e nhance the existi ng natu ralfracture network intersected lower in the well. Thestimulation also aims to generate significant microseismic events, mea sured by the MEQ array,greater th an 500 m etres f rom th e well bore. Thevolume and rate of the stimulation will bedependent o n the micro-sei smic response an dadjusted to meet the o bjectives. A se ries ofstepped inje ction rate te sts are pla nned du ring204

Australian Geothermal Conference 2010the stimulation to observe the development of thefracture network. Post the main water stimulation,tests will be performed to determine the injectivityflow rates. T his will be f ollowed by productiontesting if th e well allo ws, to un derstand lo ngerterm reservoir performance. Acidizing and u se ofproppant wit h gel in the fracture sti mulation iscontingent o n data obtai ned from th e injectivitytest. A se cond inte rval may be stimulateddependent on the results of the initial stimulation.ReferencesNeumann, N. , Sandiford, M., & Fode n, J. (2000)Regional geochemistry and continental heat flow:implications for the o rigin of the South Australi anHeat Flow anomaly. Earth and Planetary ScienceLetters, 183, 107-120.205

Australian Geothermal Conference 2010Three Dimensional Reservoir Simulations of Supercritical CO 2 EGSAlvin I. Remoroza 1 *, Behdad Moghtaderi 1 , Elham Doroodchi 11 Priority Research Centre for Energy, School of Engineering (Chemical Eng.), Faculty of Engineering & BuiltEnvironment, University of NewcastleUniversity Drive, Callaghan, NSW 2308, Australia*Corresponding author: remorozaair@yahoo.comFollowing the work of Pruess (2008) on theproduction behaviour of CO 2 as a working fluid inEGS, a three dimensional (3D) reservoirsensitivity analysis of CO 2 mass flow and heatextraction rates on injection temperature, rockpermeability, rock porosity and reservoirtemperature were performed. The 3D reservoirsimulations were performed using the TOUGH2modelling code with the modified ECO2N module.Keywords: CO 2 -EGS, reservoir simulation,TOUGH2, ECO2NBackground of the studyThe literature on the application of supercriticalCO 2 for Engineered Geothermal System (EGS) isrelatively scarece. Most of the available literatureare 1D and 2D simulations of the thermodynamicand transport properties as well as exergyanalysis (Atrens et al, 2008; Atrens et al, 2009;Atrens et al, 2009; Brown et al, 2000; Pruess etal, 2006; Reichman et al, 2008; Remoroza et al,2009).Pruess (2008) performed 2D and 3D reservoirsimulations of injection/production behaviour of anEGS operated with CO 2 as working fluid usingTOUGH2 with fluid property module “EOSM”which is not commercially available. Hissimulations examine production behaviour in 2Dareal model at different reservoir pressures andthen assessed 3D flow effects on energyrecovery. Table 1 lists the reservoir and CO 2injection parameters used by Pruess (2008) andin this study.The equivalent permeabilities calculated from theSoultz granite inferred from geophysical and flowlog analysis range from 5.2 x10 -17 m 2 to 9.6x10 -16m 2 (Sausse et al, 2006) while intact granite has1.6 to 3.8x10 -19 m 2 permeabilities (Selvadurai,2005). Soultz EGS average equivalentpermeability is 5x10 -16 m 2 .Porosities of granite range from 0.2 to 4%(http://www.granite-sandstone.com/granitephysical-properties.html).This study will expand the previous 3D reservoirsimulations of Pruess (2008) by determining theimpact of injection and reservoir parameters suchas permeability, porosity, reservoir temperatureand CO 2 injection temperature on the CO 2 massflow and heat extraction rates. Also, theapplicability of the modified fluid property moduleECO2N for EGS will be examined.Table 1. Reservoir and CO2 injection/production parameters.Pruess (2008) This studyFormationThickness, m 305 (5 layers) 305 (5 layers)Fracture spacing, m 50 50Permeable volume fraction 2% 2%Permeability in fracture 50 0.5, 5 and 50domain, x10x10 -15 m 2Nos. of MINC 5 5Porosity in fracture domain 50% 50%Permeability in rock 50 0.5, 5 and 50matrix, x10x10 -15 m 2Porosity in rock matrix 0.2%, 2%Rock grain density, kg/m 3 2650 2650Rock specific heat, kJ/kg 1000 1000Rock thermal conductivity, 2.1 2.1W/m- o CInitial conditionsReservoir fluid CO2, H2O CO2, H2OTemperature, o C 200 200, 225Pressure, bar 200 200Production/InjectionProduction area, km 2 1 1Fraction of the area 1/8 1/4modelledSpatial resolution, m 32.14 and 70.1 20.83, 45.45Injection Temperature, o C 20 20, 35Injection pressure, bar 210 gravityequilibratedfrom top layer210 gravityequilibratedfrom top layerProduction pressure, barMethodology190 gravityequilibratedfrom top layer190 gravityequilibratedfrom top layerBecause the fluid property module used in theonly published 3D reservoir modelling of CO 2flows in EGS is not publicly available; a modifiedECO2N fluid property module is used in thepresent study. ECO2N is a fluid property modulefor the TOUGH2 simulator (Version 2.0) that wasdesigned for applications to geologicsequestration of CO 2 in saline aquifers (Pruess,2005). The temperature limitation of this module is10 o C≤ T ≤ 100 o C. In the modified version ofECO2N the restriction on the upper temperatureis removed with the provision that only pure206

Australian Geothermal Conference 2010phases like CO 2 or H 2 O are present (i.e. nomixture).The 3D simulations in this study were conductedusing TOUGH2 in conjunction with the pre andpost-processing graphical interface PetraSim.Because of symmetry, only ¼ of the calculationdomain (1 km 2 five-spot well configuration) wassimulated in this study (Figure 1). Also, a modifiedCO2TAB file was used so that wider ranges ofpressure and temperature, which are moreappropriate for EGS application, can be studied.CO2TAB lists thermodynamic properties of CO 2 atdifferent temperature-pressure conditions whichare then used by TOUGH2.to all layers (Figure 4) and open only to topmost50 m layer (Figure 5).Figure 2: The CO2 mass and heat extraction rates from thisstudy match the previous study for a CO2 production wellopen in all six layers using the ¼ symmetric model having12x12 areal grids and rock matrix permeability of 1.9x10 -14m 2 and porosity of 0.2%.Figure 1: The 1/4 section of the five-spot well configurationshowing an injection-production segment.To validate the use of the modified ECO2N, theresult of the previous 3D reservoir simulationswere duplicated by finding the appropriate gridsize equivalent to the previous model used byPruess (2008). The previous study did not definerock wall specifications in the definition of fracturedomain, i.e. permeability and porosity of the rockmatrix. In the first attempt, different permeabilitieswere used for fracture domain and rock matrix(wall rock). A match was found for configurationwhere all layers of the production wells are openusing a 12x12 areal grid (41.67 m side length)and rock matrix permeability of 1.9x10 -14 m 2 andporosity of 0.2% (Figure 2). However, forconfiguration where only the top 50 m of theproduction well is open, the 12x12 areal grids ofthe ¼ symmetric model gives higher mass andheat extraction rates (Figure 3).Doubling the areal grid size to 24x24 (20.83 sidelength) and defining rock matrix permeabilityequal to fracture domain permeability (5x10 -14 m 2 )and rock matrix porosity to 0.2% gave an almostperfect match both for CO 2 production well openFigure 3: The CO2 mass and heat extraction rates from thisstudy are higher than the previous study for a CO2production well open only in topmost 50 m layer using a ¼symmetric model having 12x12 areal grids and rock matrixpermeability of 1.9x10 -14 m 2 and porosity of 0.2%.The reservoir simulation results from the ¼symmetric model with 24x24 areal grid size wasthen used as the reference for sensitivity analysis.The result from this section of the study alsoproved the applicability of modified ECO2N forreservoir simulation of pure phase CO 2 reservoirflows.Figure 4: CO2 mass and heat extraction rates from a ¼symmetric model having 24x24 areal grids and rock matrix207

Australian Geothermal Conference 2010permeability of 5x10 -14 m 2 and porosity of 0.2% match theprevious study from a CO2 production well open in all sixlayers.The 35 o C injection temperature is more or lessthe appropriate value for regions with arid climatelike Australia’s EGS locations.Figure 5: CO2 mass and heat extraction rates using a ¼symmetric model having 24x24 areal grids and rock matrixpermeability of 5x10-14 m2 and porosity of 0.2% match theprevious study from a CO2 production well open only intopmost 50 m layer.Results and DiscussionThe high CO 2 mass circulation at a reservoirtemperature of 200 o C for a production well opento all layers initially resulted in very high heatextraction rates and rapid decline of the reservoirthermal content (144 MW to 47 MW) due tothermal depletion of the reservoir. In contrast,H 2 O mass circulation was found to be low with arelatively slow decline rates and consequentlyslow decline in heat extraction rates (24 MW to 16MW, Figure 6). However, as the previous studyrecommended, for the case of CO 2 EGS whenonly the topmost 50 m layer configuration (Figure5) was used in the analysis stable massproduction and heat extraction rate of 64 MWafter 2 years were resulted.Figure 7: Effect of injection temperature on CO2 mass andheat extraction rates in a 200 o C EGS reservoir.Rock matrix permeability has dramatic effect onCO 2 mass production. In our studies the massproduction rates dropped from ~400 to 185 and29 kg/s when one and two orders of magnitudedecrease in permeability was implemented,respectively. Heat extraction rate, on the otherhand, declined to 67 and 11 MW, respectively(Figure 8).Figure 8: Effect of rock matrix permeabilities on CO2 massproduction and heat extraction rates.Rock matrix porosity has no significant effect onthe CO 2 mass production and heat extractionrates (Figure 9).Figure 6: CO2 and H2O pure phase mass and heat extractionrates at 200 o C reservoir.The effect of injection temperature on CO 2 massand heat extraction rates is shown in Figure 7.Increase in injection temperature above thecritical temperature (31.4 o C) resulted in highermass production but lower heat extraction rates.208

Australian Geothermal Conference 2010Figure 9: Effect of rock matrix porosity on CO2 massproduction and heat extraction rates.The average CO 2 mass flow rates did not varygreatly with reservoir temperature (411 kg/s at200 o C, 398 kg/s at 225 o C, and 395 kg/s at175 o C). Average heat extraction rates for the 200and 225 o C reservoir temperatures were found tobe similar at 117 and 113 MW, respectively.However, at low reservoir temperature, theaverage heat extraction rate decreased to about90 MW (Figure 10).Figure 10: Effect of reservoir temperature on CO2 massproduction and heat extraction rates.ReferencesSomerville, M., Wyborn, D., Chopra, P., Rahman,S., Estrella, D. and Van der Meulen, T., 1994, HotDry Rocks Feasibility Study, Energy Researchand Development Corporation, Report 243,133pp.Atrens, Al.D., Gurgenci, H., and Rudolph, V.,2008, Carbon Dioxide ThermosiphonOptimisation, Paper presented at the Proceedingsof the Sir Mark Oliphant International Frontiers ofScience and Technology Australian GeothermalEnergy Conference, Rydges Hotel, Melbourne.Atrens, Al.D., Gurgenci, H., and Rudolph, V.,2009a, CO 2 Thermosiphon for CompetitiveGeothermal Power Generation. Energy & Fuels,23(1), 553-557.Atrens, Al.D., Gurgenci, H., and Rudolph, V.,2009b, Exergy Analysis Of A Co 2 Thermosiphon,Paper presented at the PROCEEDINGS, Thirty-Fourth Workshop on Geothermal ReservoirEngineering, Stanford University, Stanford,California.Brown, D.W, 2000, A Hot Dry Rock GeothermalEnergy Concept Utilizing Supercritical CO 2Instead Of Water, Paper presented at thePROCEEDINGS, Twenty-Fifth Workshop onGeothermal Reservoir Engineering, StanfordUniversity, Stanford, California.Pruess, K., 2008, On production behavior ofenhanced geothermal systems with CO 2 asworking fluid, Energy Conversion & Management,49(6), 1446-1454.Pruess, K., and Azaroual, M., 2006, On TheFeasibility Of Using Supercritical CO 2 As HeatTransmission Fluid In An Engineered Hot DryRock Geothermal System, Paper presented at thePROCEEDINGS, Thirty-First Workshop onGeothermal Reservoir Engineering, StanfordUniversity, Stanford, California.Pruess, K., 2005. ECO2N: A TOUGH2 FluidProperty Module for Mixtures of Water, NaCl, andCO 2 . Report LBNL-57952. Earth SciencesDivision, Lawrence Berkeley National Laboratory:Berkeley, University of California.Reichman, J., Bresnehan, R., Evans, G., andSelin, C., 2008, Electricity Generation usingEnhanced Geothermal Systems with CO2 as HeatTransmission Fluid . 189. Paper presented at theProceeding of the Sir Mark Oliphant InternationalFrontiers of Science and Technology AustralianGeothermal Energy Conference, Rydges Hotel,Melbourne.Remoroza, A.I., Moghtaderi, B., and Doroodchi,E., 2009, Power Generation Potential of SC-CO 2Thermosiphon for Engineered GeothermalSystems, Paper presented at the AustralianGeothermal Energy Conference 2009, HiltonHotel Brisbane.Sausse, J., Fourar, M., and Genter, A. 2006Permeability and alteration within the Soultzgranite inferred from geophysical and flow loganalysis. Geothermics, 35(5-6), 544-560.Selvadurai, A. P. S., Boulon, M. J., and Nguyen,T. S., 2005, The Permeability of an Intact Granite.Pure and Applied Geophysics, 162(2), 373-407.209

Australian Geothermal Conference 2010Measuring the Success of EGS Projects: An Historical toPresent Day PerspectiveCatherine E. Stafford*Granite Power Ltd, Level 6, 9 Barrack Street, Sydney, NSW 2000.*Corresponding author: cstafford@granitepwr.comWorld wide, Enhanced Geothermal System (EGS)projects have been a round for over 35 years,commencing w ith the Fenton H ill research anddevelopment project in New Mexico back in 1972.Many years of knowledge have been accumulatedthrough various research a nd commercialprojects.As to be expected with an evolving industry, somesignificant development issues are still to be fullyor properly overcome, such as a ppropriate downwelltechnolo gies a nd ma nagement of induce dseismicity. H owever, sev eral f actors i ndicate t hatthis ‘ne w’ typ e of ge othermal technol ogy and it sassociated industry has moved beyond being justa re search and d evelopment concept. Su chfactors incl ude: the gro wth in the numbe r ofcommercial p rojects, with some of these now i nproduction; ceme nting o f the indu stry throug hassociations and governm ent incen tives; thedevelopment of geoth ermal re porting cod es forcommercial credibility (i.e. Australian andCanadian); and con siderable progre ss in theresolution of ongoing development issues.This paper provides a perspective on the successof EGS projects to date. Th is is very much a firstpassassessment as the te chnical an dcommercial data publicly available is currently toosparse an d project spe cific to ena ble a rigo rousquantitative study at thi s stage. However, it i sintended to offer a snapshot take on the successand evolution to d ate of the EGS se ctor of thegeothermal industry. It reveals that m any projectshave be en successful at what th ey set out t oachieve. It is also apparent that EGS developmentin Au stralia is like ly to b e mo re ‘su ccessful’ thanelsewhere because the continent’s stress regimeallows fav ourable sub-horizontal fractu redevelopment.Keywords: Enhan ced geo thermal system, EGS,hot dry rock, HDR, success, industry, Hot RockDefinition of EGSGeothermal power ha s been g enerated fromhydrothermal geothe rmal resou rces for man ydecades. However, such resources are limited toareas where accessible hydrothermal systems arefound, such as the world’s volca nic regions e.g.the Pacific Rim countries.Geothermal exploration in all area s requires abalance of a ccessible te mperature, water supplyand an a dequate flow rate in ord er to produ ceelectricity economi cally. Wate r may have to b eintroduced to the system or may be p resent. Ingeothermal plays away from conventionalgeothermal t erranes, re quired te mperatures willlikely b e f ound at g reater d epth whe repermeability is often decr eased (tight rocks), soreservoir enh ancement by physical or chemi calmeans is required to obt ain flow rate s that areconsidered economic.Over the years the concept of enha ncedgeothermal reservoirs h as be en d escribed withseveral acronyms e.g. Hot Dry Rocks (HDR), HotWet Rocks (H WR), Hot Fractured Roc ks (H FR)and HR (Hot Rocks). Such p rojects havecomprised the artificial creation of an undergroundheat exch anger by the drilli ng of a we ll into e.g.granite, th e stimulation o f that well to create areservoir (usually by hy draulic stimul ation a nd/orchemical sti mulation), and the drilling of aproducer well into the margi n of the create dreservoir.In the last few years EGS has be come the mostaccepted de scriptor in th e north ern hemisphere.Recent definitions include:- “EGS are a new type of geothermal powertechnologies that do not require naturalconvective hydrothermalresources.”(en.wikipedia.org/wiki/Enhanced_Geothermal_Systems)“EGS are engineered reservoirs created toproduce energy from geothermal resourcesthat are otherwise not economical due to lackof water and/or permeability.” Department OfEnergy, USA:http://www1.eere.energy.gov/geothermal/enhanced_geothermal_systems.html)“An Enhanced Geothermal System is anunderground reservoir that has been createdor improved artificially.” TP GEOELEC: thenewly formed GEOELEC-Platform, waslaunched on 2nd of December 2009, andcomprises more than 130 geothermal expertsfrom the industry and the research sector whovoted on the definition of EGS in March 2010.The secretariat of the panel is managed by theEuropean Geothermal Energy Council.(http://www.egec.org/ETP%20Geoelec/Conclusion%20EGS%20definition.pdf)This late st definition, would i nclude allconventional geoth ermal well s th at have b eenstimulated to improve reservoir performance.210

Australian Geothermal Conference 2010During the developm ent of the Geothe rmalIndustry Technology Roadmap (DRET, 2008) theAustralian community recognised that t erms suchas HFR, HDR and HWR were rather specific, andthat EGS coul d be a pplied to g eothermalresources with significant existing permeability.Therefore th e term Hot Rock was a dopted toencompass that end of the spectrum ofgeothermal resources th at requi red signifi cantpermeability enhan cement. The term HotSedimentary Aquifers is applied to that end of thespectrum where si gnificant permeability existsnaturally.Here we co nsider the success of enhancedgeothermal system s, with focus o n th eunconventional (non-volcanic related) systems.The Rapid Growth of the EGS SectorThis preliminary (and non-exhaustive) review hasfound that to date there a re in exist ence, or no wterminated, some fo rty-seven EGS projects (tomid Ju ne 2 010). Th ese proje cts a re listed inTable 1.The data presented i n this com pilation issomewhat incomplete, has been variably sourcedand as a result accuracy cannot be guaranteed. Inthe time-frame available for this preliminary study,project data was often dif ficult to acqui re o r wa snot acquired. It is re cognised that such omissionsimpact on the resul ts and he nce, an yinterpretation of those results. For example, it maybe more li kely that the re is non -publicity forunsuccessful proje cts or unsucce ssful part s ofprojects. A more rigo rous study is certainlyneeded but at thi s stage of the indu stry’sdevelopment may not b e po ssible due to thelimited and site specific nature of the data.The data sh ow that over the last four deca des,over 50% (27) of the EGS projects commenced inthe last 5 years (Figure 1 and Table 1 ). Over thewhole of the last decade 78% (37) o f the tota lprojects were comm enced (Figu re 1 ) with overhalf of these proj ects bei ng comme rcially funde das op posed to demon stration or R& D proj ects(Table 1).Conversely, durin g the 1 970’s, 80’ s and 90’sprojects were predomi nantly re search-driven(Table 1).The huge growth in the n umber of pro jects seenover the la st five years indicates that confidencein th is se ctor ha s g rown r apidly. T his can beattributed to the kno wledge a nd acquired skill sgained from the early proje cts, t echnologydevelopment e.g. drilling deeper being moreFigure 1: Bar chart showing the number of EGS projectscommenced from 1970 to 2010.easily attain ed and with less ri sk, governm entpolicy an d to the perceive d su ccess of previou sprojects.Indicators of SuccessIn a mature EGS industry, success would likely bemeasured by the amount of powe r pro duced i.e.in megawatts. However, relatively little d ata is yetavailable wit h only a modest num ber of project swell develop ed. Thus, measuri ng succe ss isdifficult at this stage. However, it is suggested thateven a p reliminary evalu ation is u seful as a to olfor all concerned.A broa d-brush a ssessment is presented he relooking at tech nical su ccess a nd commercia lsuccess of proje cts, with these bei ng definedbelow.In terms of th e development of the ind ustry, thereare many oth er fa ctors, which a re not discussedhere, that ca n be indi cative of the growth of thegeothermal industry as a whole, as well as for theEGS se ctor. The se incl ude: a cceptance of ri skand risk ma nagement; wide ap plication of th etechnology; governm ent suppo rt via grants;geothermal studies; legislation; se rvice indu strysupport via dedi cated group s and re search anddevelopment; numbe r of R&D/de monstrationprojects; number of commercial projects.211

Australian Geothermal Conference 2010Start Where Region Project WhoTable 1: EGS Projects Identified, as of Mid-June 2010SuccessPurposeStatusTechnicalCommercialReservoir rockDepthTemperatureProductiontemperaturePoweroutputMain source Commentkmo Co C MW1 1972 USA NM Fenton Hill R TE Y Y Granite 3.5 195 158 3 1, 7 Producing power for >20 years. Ended 19962 1976 Germany Bavaria Falkenberg R TE Y N Granite 0.3 85 1, 10 Successful fraccing, & short circ tests. Project finished 19853 1976 UK Cornwall Rosemanowes Camborne School of Mines R TE Y N Granite 2.7 79 1 Successful drill, frac & circulation for four yrs. Ended 19914 1977 France Le Mayet de Montagne R TE Y N Granite 0.8 33 1, 10 Granite from surface. Successful frac & circ tests.Project ended 19945 1982 Japan Ogachi Y N 1.3 240 1 Volcanic. Poor connectivity (~90% water loss)6 1983 Japan Higashi‐Hachimantai R TE Y N Granite 0.4 60 107 1985 Sweden Fjallbacka R TE Y N Granite 0.5 15 1 Successful 40 day circulation test. Ended 19898 1985 France Soultz‐en‐Foret ENGINE R OP Y Y Granite 4.2 200 155 1.5 1, 7, 8, 9 Prod. power for >15 years.9 1988 Japan Hijiori TE Y N Granodiorite 2.7 270 163 1, 4 A lot of short tests, with varied results; & water loss ~50%10 1999 Australia NSW Hunter Valley Pacific Power D TE Granite 1 Elcom. Discontinued due to funding issue. Site later acquire by Geodynamics11 2001 Australia NSW Jerry's Plain Geodynamics C Granite Inferred resource12 2001 Switzerland Basel D TE Y Granite 5 200 2 Fraccing caused 3.4R earthquake). Proj suspended.13 2002 Germany Bad Urach Y Mica‐syenite 3.3 170 2 Series of successful short frac & circ tests14 2002 Australia SA Cooper Basin Geodynamics C DR Y Y Granite 4.3 250 200 1, website Proof of concept achieved 2009 (flow btw 2 wells)15 2002 USA Coso R y 1 Stimulated same fracture via 2 wells. Second stimulation caused swarm of EQs16 2002 USA NV Desert Peak Ormat C 1 Adjacent to conventional geothermal area.17 2003 El Salvador Berlin C OP Y Y 2 Stimulation of tight injection well for existing geothermal field18 2003 Germany Hanover Horstberg (GeneSys) GEOZENTRUM R TE Y N Sandstone 3.8 2 Single well. Frac & circ was achieved, and inadequate circulation was proved.19 2004 Australia SA Paralana Petratherm C DR Metasediments 4.1 1 Have drilled injector well Paralana‐220 2005 Australia TAS Charlton‐Lemont KUTh C FU Granite website Inferred resource21 2006 Australia SA Olympic Dam Green Rock Energy Ltd C FU Granite website Inferred resource22 2006 Australia SA Parachilna Torrens Energy C FU Cryst Base/Sandst.155 website Power plant planned operational 201135 2009 Australia SA Roxby Southern Gold C FU Granite website Inferred resource36 2009 Australia WA Jurien‐Woodada New World Energy C FU website Inferred resource37 2009 Germany Unterhaching C OP Y Y 3.4 122 3.4 2, 5 Commercial. Operating. Commissioned 200938 2009 Switzerland St Gallen Geowatt C 4.1 150 (3‐5) 839 2009 UK Cornwall Eden Project EGS Energy C FU Granite 4 (3) website40 2009 UK Cornwall United Downs Geothermal Engineering Ltd C FU Granite 4.5 (10) website41 2009 USA AK Naknek Geo Project Naknek electric D 4.2 342 2009 USA NV New York Canyon TGP Development D 343 2009 USA OR Newberry Volcanic Bend Altarock/Davenport D PE Volcanics (15) 344 2009 Germany Hannover (Genesys) GEOZENTRUM R DR Sandstone 3.945 2010 Latvia Riga EGS (3‐4) 846 2010 Norway Oslo EGS 847 2010 Germany Rulzheim HotRock Verwaltungs C 3 websiteKEY:‐Purpose Status Technical success Commercial success References (see main text for full references)C Commercial DR Drilling Y indicates project was successful Y Indicates project achieved flow rates 1 MIT Report 2006D Demonstration OP Operational in stimulation and flow testing and temperatures sufficient to enable 2 M Back (re current Germany, pers con, 27/4/09) or M Haring (re Basel, pers con, June 09)R Research and development TE Terminated (if undertaken) a project to be commercially developed 3 Jennejohn 2010PE Permitting 4 Tezuka 2007FU Fundraising 5 Reif 2009DE Development 6 Baumgartner et al 2007Note: 7 Brown 2009Blank cells indicate data not yet collected or available 8 Holm et al 2010Parentheses indicate expected power output 9 Cornet 2009Projects in blue indicate assessment for technical success 10 Evans and Valley 2005212

Technical SuccessAt this stage of the industry’s development, and interms of ‘technical su ccess’, a p roject could bedeemed succe ssful if circulation, o r anothe rspecific technical goal was achieved e.g. creatinga re servoir. This i s particularly impo rtant whe nconsidering the ea rly R&D projects. For examplethe Rosema nowes proje ct, which di d not set outto commercially produce power but to prove rockmechanics concepts behind EGS an d achiev ecirculation. In that conte xt, the proje ct wa s asuccess, and as such has been valuable in drivingthe industry forward.There are several environmental factors that canbe said to affect the techni cal success of aproject. Of th ese arguably the m ost important arewater losses and indu ced seismicity. These risk shave impacted on the identified projects to varyingdegrees, ra nging fro m little to no imp act to thesuspension/cancellation of the project e.g. Basel.These co nstitute risks to the proj ect which ne edcareful ma nagement and mitigati on. Suchmanagement and mitig ation strategies have beendeveloped ov er the la st fe w years (e.g. Majer etal. 2008, M orelli a nd Malavazos 2008) an dcontinue to develop, wit h this havin g a hug eimpact on th e pu blic perception of the indu stry.However, fo r the pu rposes of thi s a ssessmentthey will not be considered to affect the techni calsuccess of a project.Commercial SuccessThe ‘comme rcial success’ of a proj ect ca n b emeasured wi th rega rd to the eco nomics of thatproject at th at parti cular site. It is dependent o ndrilling costs, temperatures, location relative topower infrast ructure, ma rkets, and fee d-in tariff sas well as other fact ors. Seve ral of these‘modifying’ fa ctors can change with ti me, as aresult of a chan ging or un certain regulato ryenvironment e.g. emission trading schemes a ndpower prices. This means that the economics of aproject ca n cha nge with time, all other facto rsbeing constant.Several of the EGS projects to date did not set outto commercially produce power, but to explore thedevelopment of EGS. Hence, these proje ctscannot be assessed in term s of being acommercial success.As a first-p ass and b road-brush ap proach t oassess commercial success, it is defined he re aseither production of p ower, or the project b eingadvanced e nough to h ave proven it s p otential toproduce power via long-term circulation tests.Environmental factors such as induced seismicityor unsustainable flo w rates can b e d eemed toaffect the co mmercial su ccess of a proje ct, asthey can lead to the closure of the project. Hence,they are considered here.Australian Geothermal Conference 2010Preliminary Assessment of theSuccess of EGS ProjectsTwenty p rojects all owed a first-p ass assessmentof technical success, as defined above. Of theseprojects o ne wa s deem ed not succe ssful due todrilling difficulties (serpentinite) leadi ng to asuspension of the proj ect (G eysers). Thi srepresents a 95% techni cal success rate for EGSprojects.Stimulations were noted to be successful within arange of lithologies (Table 1).With regard t o the c ommercial success of the seprojects: many of the earlier (1970-2000) projectsdid no t se t ou t to b e comme rcial, and o f th esenine projects (see Table 1, Rows 1 - 9) only twocan be considered as co mmercially succe ssfuland produced/is producing power (Fenton Hill andSoultz, Table 1, Rows 1 and 8).However, t he seven technically successfulprojects commenced in this last decade, for whichcommercial success could be a ssessed (Coo perBasin, Be rlin, Land au, Bruchsal, Gro ssSchonebeck, Insheim and Unterh aching), all areproducing power or have been proven capable ofproducing power an d a re planni ng developmentfor power production.Reservoir de velopment in Australia h as bee nshown (e.g. Geodynamics, Coo per Ba sin) to b eaided by the prevailing continental co mpressivestress regim e (Hillis an d Reynol ds 2000), Underthese con ditions, hydraulic re servoir stimulationslikely result i n sub-horizontal fracturing leading toenhanced well connectivity. Hence, in Australia, agreater level of success may be anticipated.ConclusionsIt is approp riate to be cau tious about the co nceptof success here, which has necessarily been keptrelatively s imple. In the f uture, the geothermalindustry coul d argu ably better define succe ss a sachieving a t ypical cost o f produ ction for po werdelivered into the retail market that is less than orequal to coal (on a pre-carbon tax basis) and thenprioritise it s R&D objectives acco rding to theprospective contri butions of various potentialtechnical a dvances to a chieving that benchma rkof success.At this moment in time, the early indicato rsgenerated from this industry worldwide show:-An expone ntial gro wth in the numb er ofEGS projects over the last 5 yearsA comme rcially dominat ed indu stry in2010 a s o pposed to R&D d ominatedactivity 10 years ago.Early indications that technical success isconsistently being achie ved with t his213

Australian Geothermal Conference 2010translating into commerc ial success(where this is an aim)The next decade will l ikely be a period ofconsolidation and furth er growth for th is pa rt ofthe geothermal sector, with several more projectslikely to b e producing power at the e nd of thi speriod. As part of that evolution, projects are likelyto increase in size.Regulatory frameworks a nd gove rnment polici esare needed which encourage this momentum andhelp consolidate the industry over this period withgovernment supp ort continuin g to be mad eavailable at approp riate l evels to fu rther growthand encourage investor interest.ReferencesBaumgärtner, J., Menzel, H., and Hauffe, P.,2007,“The Ge ox GmbH Proj ect in L andau - T he firstgeothermal p ower project in Pal atinate / UpperRhine Valley ”, in Proceedings First EuropeanGeothermal Review, Geothermal Energy forElectric Power Production, p. 33, Mainz,Rhineland Palatinate, Germany.Brown, D.W. , 2009, Hot dry rock g eothermalenergy: Important lessons from Fenton Hill. Proc.Thirty-Fourth Workshop on Geothermal ReservoirEngineeringStanford University, Stanford, California, February9–11.Cornet, F.H. , 2009, Fro m hot dry ro ck toenhanced g eothermal systems: the S oultz-sous-Forets project, AAPG Eur opean Region AnnualConference, Paris-Mal maison, Fra nce, 23 -24November 2009DRET (Department o f Re sources, En ergy andTourism) 2008, Canberra, 64pp.Evans, K.F. and Valley, B., 2005, An overview ofenhanced geo thermal systems,www.geothermal.ethz.ch/.../Geothermal%20publications/EvansEtAl_2005_EGS-Overview.pdfHillis, R. R. and Reyn olds, S.D., 2000. Th eAustralian St ress M ap. Journal of the GeologicalSociety, London, 157, 915-921.Holm, A., Blodgett, L., Jennejohn, D. A nd Gawall,K., 2010, Geothe rmal Energy: Internatio nalMarket Upda te, Geotherm al Energy Associ ationReport, May 2010.Jennejohn, D., 201 0, US G eothermal Po werProduction and Development Update, GeothermalEnergy Association Report, April 2010Majer, E., Baria, R. and Stark, M., 2008 , Protocolfor Induced Seismicity Associated with EnhancedGeothermal Systems. Report produced in Task DAnnex I (9 April 2008), Internation al Ene rgyAgency-Geothermal Imp lementing Agreement(incorporating comm ents by: C. Bromley, W.Cumming, A. Jelacic and L. Rybach).Morelli, C., and Malava zos, M. (2 008), Analysi sand Man agement of Seismic Risks Asso ciatedWith Engin eered G eothermal System Operation sin South Au stralia, in Gurgenci, H. a nd Budd, A.R. (editors), Proceedings of the Sir Ma rk OliphantInternational Frontiers of Science and TechnologyAustralian Geothe rmal Energy Confe rence,Geoscience Australia, RecordReif, T., 2009, Economi c aspe cts of geothe rmaldistrict he ating and powe r gene ration,Presentation, Tallinn Uni of Technol ogy, 17 April2009/GeotlS, Geothermische VereinigungTezuka, K., 2007, Japanese EGS Experience andmodelling efforts - a Review of the Hijiori HDRproject. ENG INE Worksho p 2, Volte rra, 1-4 Ap ril2007.214

Australian Geothermal Conference 2010Estimates of sustainable pumping in Hot Sedimentary Aquifers:Theoretical considerations, numerical simulations and theirapplication to resource mappingJ. F. Wellmann*, F. G. Horowitz, L. P. Ricard, K. Regenauer-LiebWestern Australian Geothermal Centre of Excellence, The University of Western Australia, 35Stirling Hwy, WA-6009 Crawley, Australia*Corresponding author: wellmann@cyllene.uwa.edu.auA method for a spatial analy sis of potentia lsustainability for the e arly stage of exploration inHot Sedime ntary Aquifers (HSA) is presentedhere. Our analyses are based on well establishedestimations f or the the rmal bre akthrough in adoublet well setting. We consider two significantlydifferent sce narios: the placement o f a welldoublet in a n aquife r wit hout si gnificant naturalflow, and the case whe re a natural ground waterflow exists.We integrate these two analytical estimations intoone workflow with ge ological mo delling a ndgeothermal simulation. A s a result, we obtainspatial an alyses of theoretical sustainablepumping rate s for a whole resource area. The semaps are sp ecifically suitable for the early sta geof exploration where a potential target area has tobe determined based on limited information.We p resent the appli cation of our m ethod to ageothermal reso urce are a in the North PerthBasin, from geological modelling, to the simulationof fluid and heat flow, and finally to map theanalysis of sustainable pumping rates fo r on eaquifer. Th e results con tain a high degree ofuncertainty, but indi cate t he di stribution of futureprospective areas. These maps can be combinedwith other spatial datase ts, e.g. infrastru cture.Also, as they are integrated into one workflow, anupdate of th e analyses i s dire ctly possible whennew data become available.Keywords: Hot Sedim entary Aquife r (HSA);Resource Analysis; Sustai nable Pumpi ng Rate s;Geothermal Simulation; Geological ModellingIntroductionThis paper presents a novel exploration method toidentify geot hermal p rospects ba sed on the rmaland hyd raulic p roperties of the su bsurface. Wecombine esti mates of su stainable pumping rate swith simulations of fluid a nd heat flow, and derivemaps of e stimations for sustai nable pumpin grates.Our work regards estimates of sustai nability forwell do ublet systems. After a certain time t B , thereinjected cold water f ront m ay reach th eextraction well and cool down th e extractedtemperature (Fig. 1, red curve). This will affect thegeothermal a pplication an d, at some stage, rul eout furth er effective u sage of the site. Anestimation of this breakthrough time t B is re quiredto evaluate the sustainability of a project.temperature [C]1009590Temperature development at production wellflow parallelno flowflow perpendicular0 10000 20000 30000 40000 50000production time [days]Figure 1: Comparison of temperature development at theextraction well for three different scenarios: i) If no advectiveflow is present, the cold reinjected water may reach theproduction/ extraction well (red curve) and the temperature ofthe pumped water will decrease; ii) For advectivegroundwater flow perpendicular to the wells, the temperaturedecrease is significantly slower (green curve); and iii) For thecase that the reinjection well is directly downstream of theproduction well, no thermal breakthrough occurs (bluecurve).Analytical e stimates of a su stainable long -termuse for ge othermal in stallations ha ve bee napplied for many years (e.g. Gri ngarten, 1978,Lippmann a nd Tsang, 1980). Mo st of theapproaches are ba sed o n many sim plificationsand assu mptions. They n onetheless d eliver animportant insight into th e distribution of promisingareas for sustainable flow in the sub surface,especially in the ea rly exp loration ph ase, as notonly available temperature and heat in pla ce areconsidered, but also hy draulic parameters likepermeability and porosity.Another standard tool in geothermal exploration isnumerical simulation of subsurfa ce fluid and h eatflow. (Se e e.g. O’Sullivan et. al. 2 001 fo r adetailed revision of appli cations.) A t horoughlyperformed study ca n deli ver detaile d i nsight intofluid and heat movement in the subsurface, withinthe usual limitations of data availability and modelaccuracy.One problem with both estimations, analytical andnumerical, is that they are usu ally only performedat one lo cation, i.e. at a previo usly identified215

Australian Geothermal Conference 2010target, to evaluate its lo ng term b ehaviour. Wepropose here that it is use ful to perform a ra steranalysis of sustainable pumping rate s. This canbe applied from the very first stages of geothermalexploration and sub sequently refined duringongoing exploration, wh en mo re data be comeavailable.We p resent here a m ethod to perf orm the sespatial analyses. Our approach is implemented ina com plete framework coveri ng geologicalmodelling and fluid and heat flow simulations. Theresults we obtain have to be analysed critically asthe many a ssumptions th at go into th e analysi sprohibit an a bsolute inte rpretation of t he results.For example, an estimated average pumping rateof 80 m 3 /s for a lifetime of 30 years may contain ahigh d egree of uncertaint y. But even if the totalvalues might vary, we con sider t he ge neraldistribution of the anal ysis to be a valuablerepresentation of p otential targ et a reas in aresource area.Theoretical considerations andsimulated examplesHere, we b riefly review some of the commonlyapplied theoretical estim ations of sustainabilitystudies. All the presented estimation s below aresuitable fo r appli cation in Hot S edimentaryAquifer/ po rous m edia sy stems. The situation i smuch m ore c omplex in f ractured sy stems (E GS)and special considerations are ne cessary. Fo rmore detail ed information , see the recent revie wof Banks (2009).Analytical estimationsThe longevity of a do ublet well can be defined asthe time it ta kes for the reinjec ted c ool water toreach the extraction well indicated at the poin twhen the extracted te mperature starts todecrease (e. g. Fig. 1). We are applying thi sdefinition he re, but it should be noted that thistime defines the lower en d of the usability. Afterthe therma l bre akthrough, the extracte dtemperature will decrease but possi bly theapplication will still be usable (e.g. Lipp mann andTsang, 1980, Banks, 2009).In this paper, we cons ider t wo cases f or t heestimation of a sustainable pumping rate, with andwithout advective background flow.Well doublet without advective background flowWe firstly con sider th e ca se of hydrauli cbreakthrough. This i s th e time t hyd the reinjectedwater tak es to reac h t he extrac tion well. Forsimple cases (flow al ong the sho rtest path,homogeneous aq uifer, no di spersion) th ehydraulic breakthrough ti me (e.g. Ho opes a ndHarleman, 1967) can be evaluated for a pumpingrate Q, an aquifer with poros ity , a thic kness hand a spa cing D bet ween extra ction an dreinjection wells as:2Dt hyd h3QThe hydraulic br eakthrough (i.e. th e time whenthe reinjected water reaches the extraction well) isnot eq ual to the thermal breakthrough (the tim ewhen the cold tempe rature fro nt re aches theextraction well). The temp erature front is delaye dby a retardation factor R th (Bodvarsson, 1972) thatdepends on the thermal properties (specific heat cand de nsity ) of the a quifer rock (a ) and th ewater (w):Rth1 ac cwawTherefore, th e time for the arrival of the therm alfront at the extraction well can be calculated as:tthe2 D h aca3 Q cwwInterestingly, the thermal breakthrough time in thiscase do es not depe nd on the hydrauli cconductivity / perme ability of the aquifer, but onlyon geometric and thermal properties.This estimation i s ba sed on ma ny a ssumptions;the most important are:1) Fluid prope rties are co nstant and do n otdepend on temperature.2) The flow itself is stea dy-state, inje ctionrate an d te mperature are co nstant andthere is n o mixing betwee n the reinje ctedfluid and the native water.3) The ge ometry of the aquifer i s verysimple: constant t hickness, co nstantporosity an d it is assumed to behorizontal.4) Cap rock an d bed rock of the aq uifer areimpermeable.(For a further detailed discu ssion see e.g.,Gringarten and Sauty, 1975.)Apart from t hese conditions, another co mmonassumption is that there i s no heat transport fromthe aq uifer into the surrounding ro cks byconduction. This assu mption is rea sonable inmany cases (see Gringarten and Sauty, 1975, foraccurate criteria) and we will adopt it here as well.Well doublet with advective background flowIf a native h ydraulic gra dient is pre sent in theaquifer, th e situation i s more co mplex. It is no wimportant to con sider th e well pl acement withrespect to the natural adv ective groundwater flow216

Australian Geothermal Conference 2010v 0 (Fig. 2). An analytical estimation for the thermalbreakthrough can be derived for the case that thereinjection we ll is p laced d ownstream o f theextraction well (Lippmann and Tsang, 1980):tthewhereA 1 acD / v ) c(0Q2hDv 0waw14Aarctan14A4A A1 4 This equation does not have a real solution whenthe natural groundwater velocity is above a criticalvaluev02QhDIn this case, no thermal breakthrough will occurand the sy stem i s, in prin ciple, compl etelysustainable and can be operate d without timelimitations.Validity of the analytical solutionThe a nalytical estim ations of hydraulic andthermal b reakthrough depen d on man yassumptions (se e ab ove). In a realist ic setting,some effects might reduce the b reakthrough time(hydraulic dispersion, heat conduction in the flui dphase) whil e othe rs might lea d to long erbreakthrough times (heat re supply fromsurrounding beds, stratification). A carefulexamination of these effects i s p ossible withnumerical simulations of pumping and reinjection.drawdown i n the reservoir. Concerning thetemperature breakthrough time, we want to havea large well spa cing D, but if we consider th epressure d rawdown, a smaller spa cing i s m orebeneficial. T he optim al value of D can not bedetermined analytically, but nume rical solution scan be applied (e.g. Kohl et al., 2003, Wellman net al., 2009).Numerical simulationsThe theo retical e stimations d escribed abovedeliver a ve ry useful e stimation ab out potentialgeothermal t argets. We t herefore consider themas ide al for the early ex ploration sta ge. But asthey depen d on ma ny assumpti ons andsimplifications, num erical simulations ofsubsurface fluid and heat flow have to be appliedto de rive a more realistic i nsight into th esustainability of the system.Consideration of pressure drawdownIn both case s presented above, with a nd withoutnatural grou ndwater flo w, we can consider amaximum pre ssure dr awdown s at th e extractionwell as another criterion for the determination of asustainable pumping rate. Grin garten (1 978)presented a relationship obtained fro m potentia ltheory:s Q2TlnDr wWe can see from this eq uation that the pressu redrawdown s depends on pumping rate Q, aquifertransmissivity T and well diameter r w , as can beexpected, but also o n the well spacing D, as thetwo wells int eract and a smaller spacing leads toless drawdown.Combined analysisFor a complete sustainability analysis fo r the welldoublet, we might co nsider the thermalbreakthrough time and a maximum pre ssureFigure 2: Fluid and heat flow representations for the threeexample scenarios described below, and presented in Fig.1.In the first example, without advective flow, we can see thatthe reinjected cold water (reinjection well: blue triangle)reaches the extraction well (red triangle) after a certain time,and the produced temperature decreases. Example 2: fornatural advective flow in the direction of the injection well, thecold temperature fan does not reach the extraction well, andthe system is completely sustainable. Example 3: for flow217

Australian Geothermal Conference 2010perpendicular to the wells, less cold water reaches theextraction well, and the temperature decrease is reduced.A variety of software codes exist s t o pe rformthese simulations (se e O ’Sullivan et al., 2001).We used SHEMAT (Simulator of HEat and MAssTransfer, Cl auser and Bartels, 2003) fo r th eresource scale simulations presented in this paper(e.g. Fig. 3). To test the validity of the analyticalestimation of breakthrough times, we si mulated awell d oublet (extra ction a nd reinje ction well) wit hSHEMAT an d, addition ally, with the petrole umreservoir engineering so ftware Tem pest/Morefrom Roxar.The plot s i n Fig. 2 show th e te mperaturedistribution (colo ur ma p) and fluid flow vecto rs(grey arrows) in the subsurface. The thre eexamples given relate to temperature decrease atthe extraction well for three different scenarios, asgiven in Fig. 1.In summary:1) Well d oublet in an a quifer withoutgroundwater flow, after thermalbreakthrough occurred. We can see thatthe flow field affect s a wide areaperpendicular to the direct con nectionbetween the wells.2) Natural g roundwater flo w, the reinjectionwell is in the downstream direction of theextraction well. The tem perature field isnow disturbed by the natural groundwaterflow. For the same pumping rate, wellspacing, and simulatio n time, the col dtemperature field does not reach theextraction well. No therma l brea kthroughoccurs.3) Natural groundwater flow perpendicular tothe con nection line of the well dou blet.The tempe rature field is again cle arlyaffected by the groundwater flow field andthe temperature de crease at theextraction well is slowed down.Example: North Perth BasinGeological modelWe a pplied the an alytical estimation s presentedabove to exploration-scale simulations of fluid andheat flo w. F or the first stage of exploration, weconsider th ese a nalytical a ssumptions as avaluable i ndication of pot ential ge othermal ta rgetareas.Figure 3: 3-D Geological model for a part of the North PerthBasin (In inset picture: model location=red square; blackcircle=Perth). Sedimentary formations are overlyingbasement which is offset by normal faults.As a n exam ple he re, we show the application ofthe method t o an area in the No rth Perth Basin.The geology is characterised by thick sedimentaryformations cut by normal faults in a graben setting(Fig. 3). The geological model wa s created withan impli cit p otential-field approach (Lajaunie, et.al. 1997 ), impleme nted in the G eoModellersoftware (Calcagno et al., 2008 ). The model is asimplified ve rsion of a more compl ex re gionalmodel.Geothermal simulationThe geological model is directly processed to aninput file for fluid and h eat flow simulation withSHEMAT (s ee C lauser and Bartel s, 20 03). Rockproperties (permeabilit y, porosity, thermalconductivity and h eat ca pacity) were assign edaccording to sample s i n this re gion wh ereavailable. A strong anisotropy (horizontal / vertical= 10) was applied to all perm eability values toachieve a more realistic model.Figure 4 is a rep resentation of the simu lated fluidand h eat flo w field fo r t he North P erth Basi nmodel. The effect of fluid flow o n the temperaturedistribution i s clearly visible. T he re sultingtemperature gradient s appear reasonable andqualitatively in accordan ce with measured value sin the area.For a q uantitative analysis of the re sults, themodel ha s t o be refin ed and adju sted further,especially at the b orders (bo undary conditions,see discussion). Resp ecting th ese cu rrentlimitations to model verification, we next apply ourresource an alysis m ethods to thi s mo del. As allsteps are integrated into one workflow, it is easilypossible to update the model an d a ll analyse slater, when more data become available.218

Australian Geothermal Conference 2010Figure 4: Visualisation of the temperature and fluid flow fieldsimulated for the 3-D geological model of the North PerthBasin. The orange isosurface shows the depth to 120 o C. Theblack lines indicate fluid flow pathways. General flowdirection is N-S.Novel resource analysis methodsNext we use the simulated fluid and heat flow fieldto estimate different a spects of geothe rmalresource sustainability. All the examples here areperformed fo r the old est sedimentary formatio n(see Fig. 3, light green unit). The minimum lifetimeto co nsider a geothermal p roject site a s“sustainable” is a ssumed to be 30 ye ars for t hefollowing analyses.Key novel aspects of all our analyses are:1) We p erform t he an alyses directly on th ebasis of the simulated fluid and heat flowfield for the resource area, linked to the 3-D geological model.2) All asp ects (maxim um su stainablepumping rate, heat i n place, pressuredrawdown) a re evalu ated on a spatialbasis, i.e. we derive 2 -D maps of theseproperties showing their distribution.3) All relevant steps a re integrated into oneworkflow, it is therefore readily possible toupdate the geological m odel and th egeothermal simulation wh en more databecome available.Maximum sustainable pumping rate, withoutconsideration of advective flowFollowing t he definitio n of the theoreti calbreakthrough time for the thermal front givenabove, we estimate a maximum pumping rate thatcould be exp ected for a d oublet system. Spatialanalysis is performed step-by-step at every pointin space (Fig. 5).Figure 5: Maximum pumping rates for a doublet system with800 m spacing and a lifetime of 30 years. All other propertiesrequired for the estimation of the pumping rate (seeequations above) are directly taken from the simulation (e.g.density) and the model (e.g. formation thickness).Consideration of natural groundwater flowIn Fig. 4 we can see from the di stribution of th estreamlines t hat gro undwater flow i s present atthe regional scale for this specific model. So, if wehypothetically intelligibly place a well doubl et inone of the flow areas, it is possible to determine apumping rate whe re a the rmal b reakthrough will,theoretically, never occur (see Exam ple 2 in Fig.2). If we then apply this analysi s agai n at everypoint in our model, we ca n de rive a spatialanalysis of t hese pum ping value s (Fi g. 6) fo rwhich thermal breakthrough will never occur.Figure 6: Estimation of a maximal pumping rate for thetheoretical case of no thermal breakthrough (Example 2 inFig. 2), in the presence of advection. This is of practicalsignificance as areas with a high value can also be expectedto allow a higher sustainable pumping rate (for a projectdurance less than infinity).Combination with other important factorsThese spati al analyses can n ow be u sed i ncombination with other relevant factors fo rgeothermal exploration, e.g. mean temperature atdepth for a target formation, or maps of local heatin place (Wellmann et al., 2009). Heat in pla ce isreferred to as one absolute number characterising219

Australian Geothermal Conference 2010the geothermal potential of a geothermal reservoirvolume. We define the lo cal heat in pl ace as theheat in place of a tiny subdivision of the reservoir.A local heat in place map gives crucial informationon the conn ectivity of the ge othermal re servoirand therefore is of high interests fo r re servoirengineering studies. As all results a re in mapview, they can ea sily be inclu ded in aGeoInformationSystem (GIS) and combined with,for example, infrastructure considerations.DiscussionWe h ave sh own that it i s po ssible to combi neanalytical co nsiderations of re sourcesustainability, with geoth ermal fluid and heat flowsimulations. Our approach enables a direct spatialanalysis of relevant factors fo r geothe rmalexploration i n Hot Sedi mentary Aq uifers. T hemajor advantage is that geothermal prospects canbe identified based on ph ysical reasoning (in thecontext of geolo gical modelling), geothermalsimulation, a nd id eally, a ll availabl e data. Wepropose that this method is a step forward for theidentification of geothe rmal target a reas from aregional analysis.The example presented above for the North PerthBasin is pe rformed for a reso urce-scale model,representative for an earl y stage of geothermalexploration. It contain s a hi gh deg ree ofuncertainty and the de termined nu mbers fo rsustainable pumping rat e are p robably notquantitatively correct. But as they are b ased on afull 3-D inte gration with geol ogical knowledge,physical simulation and all available data, we caninterpret the results spatially, i.e. identify areaswhich should be analysed more careful ly. This isa majo r a dvantage to stand ard re sourceestimation methods, e.g. heat in place, where onlyone val ue for the wh ole resource area i sdetermined.In a re alistic proj ect scenario, th e n ext stepswould be to refine the model and adjust boundaryconditions carefully to the loc al setting. But as wehave integrated all rel evant modelling, simulationand resource analysis steps into one workflow, anupdate at every stage is easily possible.We recognise that the results a re subject to alarge degree of uncertain ty. Two way s we wil laddress this in future work, will be to combine thisworkflow with an u ncertainty si mulation ofgeological modelling (Wellmann et al., 2010), andwith se nsitivity studies of the geothermalsimulation to derive a qu antitative evaluation o fthe sustainability map quality.AcknowledgementsThis work was en abled by an AustralianInternational Postgraduate Re search S cholarship(IPRS) a nd a Top -up S cholarship f rom GreenRock Ene rgy Ltd. The We stern AustralianGeothermal Centre of Ex cellence (WAGCoE) is ajoint centre of Curtin University, The University ofWestern Australia and the CSIRO and funded bythe State Go vernment of Western Au stralia. Weare ve ry thankful fo r t he comme nts of a nanonymous reviewer. We than k Roxar an dIntrepid Ge ophysics for providing acad emiclicenses of, res pectively, Tempes t/More andGeomodeller software.ReferencesBanks, David, 2009, Thermogeologicalassessment of open-loop well-doublet schemes: areview and synthesis of analytical approaches:Hydrogeology Journal, v. 17, no. 5, p. 1149–1155.Bodvarsson, G.,1972, Thermal problems in thesiting of reinjection wells: Geothermics, v. 1, no. 2,p. 63–66.Calcagno, Philippe; Chilès, Jean-Paul; Courrioux,Gabriel; Guillen, Antonio, 2008, Geologicalmodelling from field data and geologicalknowledge: Part I. Modelling method coupling 3Dpotential-field interpolation and geological rules.Recent Advances in ComputationalGeodynamics: Theory, Numerics andApplications: Physics of the Earth and PlanetaryInteriors, v. 171, no. 1-4, p. 147–157.Clauser, Christoph; Bartels, Jörn, 2003,Numerical simulation of reactive flow in hotaquifers. SHEMAT and processing SHEMAT:Berlin, Springer.Gringarten, A. C.; Sauty, J. P., 1975, A theoreticalstudy of heat extraction from aquifers with uniformregional flow: Journal of Geophysical Research-Solid Earth, v. 80, no. 35.Gringarten, Alain C., 1978, Reservoir lifetime andheat recovery factor in geothermal aquifers usedfor urban heating: Pure and Applied Geophysics,v. 117, no. 1, p. 297–308.Hoopes, J. A.; Harleman, D. R.F., 1967, Wastewater recharge and discharge in porous media: J.Hydraulics Div., Am, Soc. Civ. Eng, v. 93, p. 51-71.Kohl, Thomas; Andenmatten, Nathalie; Rybach,Ladislaus, 2003, Geothermal resource mappingexamplefrom northern Switzerland. SelectedPapers from the European GeothermalConference 2003: Geothermics, v. 32, no. 4-6, p.721–732.Lajaunie, Christian; Courrioux, Gabriel; Manuel,Laurent, 1997, Foliation fields and 3D cartographyin geology: Principles of a method based onpotential interpolation: Mathematical Geology, v.29, no. 4, p. 571–584.Lippmann, M. J.; Tsang, C. F. (1980): Groundwateruse for cooling: associated aquifer220

Australian Geothermal Conference 2010temperature changes: Ground Water, v. 18, no. 5,p. 452–458.O'Sullivan, M. J.; Pruess, K.; Lippmann, M. J.,2001, State of the art of geothermal reservoirsimulation: Geothermics, v. 30, no. 4, p. 395–429.Wellmann, J. F., Horowitz, F. G., Regenauer-Lieb,K., 2009, Co ncept of an i ntegrated workflow fo rgeothermal exploratio n in Hot SedimentaryAquifers; Australian Geoth ermal Ene rgyConference (AGEC), Brisbane, Abstracts.Wellmann, J. F.; Horowitz, F. G.; Schill, E.;Regenauer-Lieb, K., 2010, Towards incorporatinguncertainty of structural data in 3D geologicalinversion: Tectonophysics, v. 490, no. 3-4, p.141–151.221

Australian Geothermal Conference 2010Hydrothermal Spallation for the Treatment of Hydrothermal and EGSWells: A Cost-Effective Method for Substantially IncreasingReservoir Production Flow RatesThomas W. Wideman 1, Jared M. Potter 1* , Brett O’Leary 2 , Adrian Williams 2 ,1 Potter Drilling, Inc. 599 Seaport Blvd., Redwood City, CA 940632 Geodynamics Limited, Suite 1A Level 5 South Shore Centre, 85 The Esplanade SOUTH PERTH WA 6151*Corresponding author: jared@potterdrilling.comThere is a n eed in many geotherm al proje cts togreatly in crease th e p roductivity and/o r inje ctivityof wells, with a direct impact on proje cteconomics. Flow testing in wells and associatedmodelling of EGS reservoi rs has suggested thatlow p roductivity from EGS reservoi rs can resultfrom ne ar-wellbore imp edance, a restriction offluid flow wi thin the im mediate vici nity of theborehole walls. F urther modelling suggested thataltering the geometry of an existing wellbore bylongitudinally slotting the bore hole wall coul dsignificantly decrease ne ar-wellbore i mpedance.Hydrothermal spallation is an id eal technology tocreate sl ots in the hard ro ck fou nd in thesereservoirs. This pap er d escribes the results oftests of hydrothermal spallation well enhancementtechnology in the laborato ry and initial field trialsas well as its a pplication to incre ase productionfrom geothermal wells.Keywords: thermal sp allation, hydrothermalspallation, well e nhancement, Engine eredGeothermal Systems, EGSBackgroundEnhanced Geothermal Sy stems (E GS) have th epotential to generate cle an, renewable, base loadelectricity using h eated fluid prod uced fro mengineered rese rvoirs. These re servoirs a recreated by enha ncing n atural fractures in hotbasement rock usi ng hydrauli c fracturi ngtechnology. Once the stimulation proce ss iscompleted, i njection wells d eliver flui d from thesurface into the reserv oir where t he fluid i sheated, a nd prod uction well s drill ed into thefractured rock extract th e heated fluid to drive asurface-based power plant.This EGS te chnology is on the verge of reachingits tremend ous potential. Projects a round theworld a re a chieving thei r temperature objectives,but face the challenge of low circulation rates afterthe hydrauli c stimulatio n. Most model s sug gestthat EGS wells must pro duce 80-100 kg/se c offluid at 200 °C per well to be economically viable.However, such rate s ha ve proven difficult toachieve in EGS projects to date. T he potential tocreate additional fracture zones within a reservoirexists, but adds considerable cost an d ri sk to aproject a nd has not been a ccomplished at thi spoint. Meeting the prod uctivity targets in EGSwells is critical to realizing the full potential of thispromising renewable energy technology.Geodynamics Limited ("Geodynamics”), a leadingAustralia-based devel oper of EGS proje ctsworldwide, in colla boration with A ustralia’sCommonwealth Scientific and ResearchOrganisation (CSIRO), h as conducted extensivefield testing and a ssociated modelli ng of EGSreservoirs. These studi es sugge st that lowproductivity from some E GS reservoi rs prim arilyresults fro m near-wellb ore imp edance, therestriction of fluid flow within the immediat evicinity (sev eral in ches or a few fe et) of theborehole walls. This could be due to reduction infracture permeability from drilling muds or otherparticulates, but the most influential fact or is likelyfluid turbule nce due to rapidly increa sing fluidvelocities n ear the well bore. To enter t heproduction well, fluids mu st pass thro ugh po resand fra ctures in the walls of the bore hole. Fluidvelocities in t he vicinity of the bo rehole are muchhigher than in the rest of the reservoir because ofthe limited surface area of the borehole walls andare estimated to excee d 55 km per hour near theentry into the wellbore in some cases. This nearwellboreimp edance can l imit both the output ofproduction wells an d, to a lesser extent , the inputcapacity of injection wells.A simplerepresentation of this n ear wellbore effect i sshown in Figure 1.By altering t he g eometry of an existin g wellbo rethrough increasi ng the effective dia meter, it ispossible to redu ce the ne ar well bore effect. Thiscan be a ccomplished by slotting the boreh olewall, as illustrated in Figure 2.Spallation h as b een proposed a s a n effectiv emeans of expanding the diameter of a wellb ore ora way to create un derground cave rns. Fo rexample, in Pedernal, Ne w Mexico, Bob Potterand Ed Will iams, from Los Alamo s Nation alLaboratory, demonstrated the ability to create 60-cm diam eter holes in hard rock u sing o nly a 10-cm diam eter axially-orie nted flame jet. Anillustration of this process is shown in Figure 3.222

Australian Geothermal Conference 2010Figure 3: Flame spallation cavity makerFigure 1: Illustration of near wellbore impedanceFigure 2: Borehole geometry for reduced wellboreimpedanceFlame jet spallation p rocesses continue to b eused comme rcially. Fo r example, in Can ada,thermal fragmentation mining employs a variatio nof this ex cavation techniq ue to extra ct pre ciousmetals in thin, near surface ore depos its. In atypical operation, a small pilot hole is drilled int othe ore zone and an 8cm diameter flame jet drill isused to cut large r caverns up to about one m eterin diam eter. The spalls are the n p rocessed fortheir precious metals.However, this air-ba sed tech nology ca n only beused near the surfa ce and to moderat e depths i nregions where air drilling is feasibl e. At thedepths req uired for EGS, the wellbo re will verylikely need to be filled with fluid to preventexcessive water o r ga s flows and maintaincirculation and hol e st ability. For flame jetspallation, a fluid-filled wellbore poses seriouschallenges f or b oth the transport of chemicalsfrom the surface and for down-hole combustion.As a re sult, hydrothe rmal spall ation has b eeninvestigated as a vi able me ans for theenhancement of deep, water filled EGS wellbores.Experimental ApplicationA hydrothe rmal spallatio n test a pparatus wa sused for pro of-of-concept tests on the laborato ryscale. T he t est system shown in Figure 4 iscapable of indep endently applying controlle dhydrostatic, confining, a nd axial loa ds on 10 cmrock cores to simulate varying wellbore conditionsto 2400 m.223

Australian Geothermal Conference 2010Figure 4: Hydrothermal spallation deep well test rigThis system whi ch u ses su perheated stea m tospall the rock was used to create single axial slotsof the desired depth in holes and investigate theirinteractions with induced fractures.The process wa s then fu rther scal ed to cut theaxial slots 10 cm wide and more than 25 cm deepin open-hole sections of granite blo cks, as shownin Figure 5, usin g a full-scale well enh ancementtest rig.Figure 5: Axial slots cut into 10 cm open hole section ofSierra White Granite using a full scale toolFigure 6: Full scale well enhancement test rig224

Australian Geothermal Conference 2010A 8.8 cm ite ration of the slotting d rill hea d wasdesigned an d fabri cated f or field trial s. The drillhead interfaces with a dynamic seal to isolate theworking fluid from the cooling water, fol lowed byinstrumentation and controls, tension release, andconnector subassemblies, as shown in Figure 7.Geothermal Energy Conference, November 16-19in Adelaide.ReferencesCalaman, J. and Rolseth, H.C., 1961, TechnicalAdvances Expand Use of Jet-Piercing Process inTaconite Industry: International Symposium onMining Resources, University of Missouri.Dreesen, D. and Betz, R., Coiled-TubingDeployed Thermal Spallation Hard Rock CavityMaker, FWP-02FE15.O’Leary, B., 2010, Longitudinal Slotting forWellbore Enhancement Purposes: QuantifiedCost Benefit: Geodynamics Report.Potter, R. M., Potter, J. M., and Wideman, T.W.,2010, Laboratory Study and Field Demonstrationof Hydrothermal Spallation Drilling: GRC Meeting,Sacramento, California.Figure 7: Cutaway rendering of field prototype bottom holeassemblyA coiled tubing string was assembled for purposeby nesting a 9.5 mm OD stainless steel capillaryand 7-conductor wireline cable inside of a 50 m mOD, 3.4 0 wa ll, HS-90 ste el coiled tu bing string.The strin g wa s mounte d on a tracked coile dtubing u nit, shown in Fi gure 8. Rotating swivel sand slip rings ad ded to the reel al lowed fortransmission of the co oling wate r, ele ctrical, an dreactants through the hub.Williams, E., 1986, Thermal Spallation Excavationof Rock: U.S. Symposium on Rock Mechanics(USRMS), 27th annual meeting.An Evaluation of Enhanced Geothermal SystemsTechnology: Geothermal Technologies Program,U.S. Department of Energy, Energy Efficiency andRenewable Energy; IPGT Workshop, August 27-28, 2010. Reykjavik, Iceland.Flow Test Update: Geodynamics ASXAnnouncement, released April 22, 2008.The Future of Geothermal Energy: Impact ofenhanced geothermal systems (EGS) on theUnited States in the 21st century, MassachusettsInstitute of Technology and Department of EnergyReport, 2006: Idaho National Laboratory INL/EXT-06-11746.Figure 8: Coiled tubing unit for field testing of hydrothermalspallation well enhancement.A field locati on was chosen in the foot hills of theSierra Mountains i n no rthern California, whe recompetent granite co uld be found close to thesurface.At the time of publication, the results from the fielddrilling tests were not yet available. T he result swill be presented orally at the 2010 Australian225

Australian Geothermal Conference 2010Thermal effects on the wellbore stabilityBisheng Wu, Bailin Wu, Xi Zhang and Rob Jeffrey*CSIRO Earth Science and Resource Engineering, Clayton, VIC 3168, Australia*Tel: 03-95458354, E-mail: rob.jeffrey@csiro.auAbstractThermal-mechanical behaviour plays an importantrole in wellbore stability of enhanced geothermalsystems (EGS). Based on the thermo-poro-elasticmodel proposed by Coussy (1989) and by usingLaplace transformation methods, analyticalsolutions for the temperature, pore pressure andstresses around a vertical well are obtained in theLaplace space. Then the solutions in the timedomain are obtained via numerical Laplaceinverse transformation. The results show that thecooling of the wellbore has a considerable effectnot only on the pressure, in-plane radial and hoopstresses, but also on the vertical stress. It reducesthe compressive radial and hoop stresses alongthe maximum horizontal stress direction. Whenthe wellbore temperature is low enough, the hoopstress has a tendency to become tensile leadingto the wellbore breakdown and the lower verticalstress implies the generation of an in-planefracture. The possibility of wellbore shear failure isfound to exist at places with higher compressivehoop stress.Keywords: thermal effects, wellbore stability,Laplace transformation, numerical simulationIntroductionThe mechanical behaviour of a wellbore in a fluidsaturatedporous medium has attracted a greatdeal of attention during the past several decades(Rice and Cleary 1976; Detournay and Cheng1988; Rajapakse 1993; Ekbote et al. 2004). Mostof these studies focus on the pore pressure andstress distribution around the wellbore based onBiot’s formulation of linear poroelasticity (Biot1941). It is worthwhile to mention that Detournayand Cheng (1988) first provided the poroelastictransient and plane-strain analytical solutionsunder a non-hydrostatic stress field.Thermal effects should be considered ingeothermal reservoir stimulation and production.Cold fluid is generally injected into the wellboreand the surrounding rock is at a highertemperature. Due to different thermal coefficientsfor fluid and rock matrix, the pore space in therock will deform much less than the fluid in thepore when subject to a temperature variation.Obviously, this will lead to a higher (or lower) porepressure that affects fluid diffusion. On the otherhand, pervasive flow can carry low-temperaturewater into a pore and thus the local temperatureequilibrium is broken. To account for the thermaleffects, many theories have been developed(McTigue 1986; Kurashige 1989). However, ageneral thermo-poro-elastic theory proposed byCoussy (1989) appears more suitable for the lossof local temperature equilibrium, which may beimportant to consider for a geothermal system. Inthis theory, a latent heat associated with theincrease of fluid mass content is introduced with adifferent meaning from the conventionalterminology.For thermo-poro-elastic solids, Wang et al. (1994)and Li et al. (1998) obtained analytical solutionsfor stress distributions around a borehole. In theirmodels, the temperature is decoupled from porepressure. In this paper, the fully-coupled solutionsare obtained based on the linear thermo-poroelasticmodel proposed by Coussy (1989). As theborehole is much longer compared to its diameter,we consider the problem by assuming plane strainconditions for cross sections orthogonal to theborehole axis.Problem FormulationFigure 1 shows a vertical wellbore subjected to avertical tectonic stress σ v , a maximum and aminimum horizontal stresses, σ max and σ min . Theinitial pore pressure and temperature are P * R andT * R , respectively. When time t>0, the wellboretemperature and pressure are kept constants, i.e.maintained at T * w and P * W .Figure 1: Problem geometryGoverning equationsThe equations to be solved and initial andboundary conditions for this problem are listed asfollows:1. Linear thermo-poro-elastic constitutive law:226

K 2 G/3 2GBP3 K Tr B v r i B B K 2 G/3 2GBP3K T 2GrB v i Br K 2 G/3 BP3 K T .z B v i B B2. Compatibility equationsur ur 1 u11uruu r , , r= .r r r 2 r rr3. Equilibrium equations: r1r r r1 r 0, 2 0.r r r r r r4. Governing equations for pore pressure andtemperaturek 2 1 P L Tvw P Bi ,* M tt MTRt2 22 wLP* v wLTAT 3 BKBT R( C v ) ,*M t t MTRtwhere ijare effective stresses, latent heat forfluids is defined as L=T * (3α K -3α K )/ρ /B ,* R u u B B w iP=P * -P R and T=T * -T * R , λ A = λ w +(1-Φ) λ r and C v = Φρ w c w +(1-Φ) ρ r c r . We use the sign convention thatstresses are positive if in tension.The initial and boundary conditions for theproblem are given asPr ( ,0) 0, Tr ( ,0) 0,*rPw at r= rw,r R0 S0cos2 when r , R0 S0cos2 when r ,rS0sin 2 when r ,* *P P P at r= r , P 0 when r ,w R w*w RwT T T * at r= r , T 0 when r .Method for solutionIn order to make the problem tractable, the in-expressed in terms ofplane principal stresses areequivalent isotropic (or mean) and deviatoricstresses defined as followsR ( )/2, S ( )/2.0 max min 0 max minThe problem can then be decomposed into twomodes: axi-symmetric (mode I) and deviatoricloading (mode II) (Detournay and Cheng 1988).For the first mode, the well is subjected to anisotropic far-field mean stress R 0 , boreholepressure P * w and temperature T * w at the wellborewall. The second mode is associated with aborehole subject to a remote far-field pure shearstress. Moreover, the borehole boundaryconditions are all included in the first mode, sothat the boundary conditions for the second modeare homogeneous. Taking the stresses inducedby mode I as σ 1 , σ 1 , σ 1 , and the stress inducedr θ rθBAustralian Geothermal Conference 2010by mode II as σ r 2 , σ θ 2 , σ rθ 2 , the final stresses forthe complete thermo-poro-elastic problem arefound by superposition of the above solutions inLaplace space (Wu et al., 2009).Numerical resultsBy using the numerical Stehfest method(Detournay and Cheng 1988), the solution in thetime domain is obtained. The parameters used forthe example calculations are listed in Table 1, andare for the particular case of a granite-watersystem. In the following figures, the black, red andblue curves are for the pure elastic, poro-elastic(Detournay and Cheng 1988) and thermo-poroelasticsolutions (the present solution),respectively. And, their difference in time isdiscerned by line types, solid curves for t=0.9 s,medium dashed curves for t =1.5 mins, shortdashed curves for t=2.5 hrs, dotted curves fort=10.4 days, dash-dotted curves for t=2.85 years.Table 1 Parameters for calculationParameterValuesFluid, rock mass density ρw, ρr (Kg/m 3 ) 1000, 2700Fluid, rock specific heat cw, cr (J/Kg/K) 4200, 1000Fluid, rock thermal conductivit y λw, λr (W/m/K) 0.6, 2.4Rock permeability k (mD) 4.0×10 -5Fluid viscosity μ (Pa·s) 0.0004Drained thermal expansion coeff. αB (1/K) 5.0×10 -6Undrained thermal expansion coeff. αu (1/K) 1.0×10 -5Biot coefficient B i 0.47Max principal stress σ max (Pa)-1.7×10 8Min principal stress σmin (Pa) -1.3×10 8Vertical principal stress σ v (Pa)-1.1×108Initial pore pressure PR * (Pa) 7.0×10 7Wellbore pressure P w * (Pa)7.8×10 7Wellbore temperature T w * (K)323, 423Initial temperature T R * (K)523Shear modulus G (Pa) 1.5×10 10Drained and undrained poisson ratio v, vu 0.25, 0.34Porosity Φ 0.09Thermal effects on pore press ureFrom Figure 2(a) it is clear that the pore pressureobtained from the poro-elastic model for mode lIloading decreases monotonically with radialdistance and time to the reservoir pressure.However, for the thermo-poro-elastic solutions,the pore pressure in the region near the wellborecan decrease to below the reservoir pressure atsmall times if the far-field stresses are isotropic.When the cold fluid is injected into the wellbore,from Figure 3(a), we see that the temperaturegradient near the wellbore is very high, although itdecreases with time. This may lead to a reductionin pore pressure to below the reservoir pressuresince the contraction of cooling water results in a227

Australian Geothermal Conference 2010volume decrease of the pore fluid.The pore pressure and temperature induced bythe anti-symmetrical loading is displayed in Figure2(b) and Figure 3(b). We see that, even thoughthe thermal effect has been taken into account,the pore pressure curves from the thermo-poroelasticsolutions show little difference compared tothe poro-elastic solutions. This confirms, asexpected, that the shear deformation induced bythe anti-symmetric loading does not causesignificant temperature change, as shown inFigure 3(b). For the present case, the antisymmetricstress S 0 is 20 MPa, from Figure 3(b)the temperature change is in the order of 1ºC.Figure 2(c) is superposition of Figure 2(a) and (b),where the thermal effect on the pore pressure canbe strongly offset by the pressure change fromshear loading.Thermal effect on stressesIn mode II, the radial stress is not sensitive to thetemperature difference between the well and thefar field, as show Figure 4(b). But the radial stressdepends on the wellbore temperature, since thethermal responses and volumetric change areboth significant. At a certain distance to the wall ofthe well, the radial stress level is elevated by acooler injection fluid as indicated in Fig. 4(c).However, the radial stress on the well wall isimposed by the boundary condition.Figures 5(a)-(d) display the hoop stress evolutionfor two different well temperatures. Figure 5(b)shows that the hoop stresses induced by mode IIloading follow the poroelastic solutions closelyexcept at early time and near the wellbore.Figures 5(c) and 5(d) indicate that the cooler thewell, the higher the hoop stress in tension.Moreover, the hoop stress at the wall, although itdoes not change with time, can be increased byimposing a cooler well condition. The hoop stresscontours at different wellbore temperatures(T * w =373 K and 423 K) are shown in Figure 6.These plots clearly demonstrate the positionalong the maximum tectonic stress directionexperiences the lowest compressive hoop stress.In Figure 7, the vertical stress evolution isprovided. When the wellbore temperature is equalto the far-field rock temperature shown in Figure7(A), the vertical stress contours appearunchanged because of the very small temperaturegradient in the whole domain.However, when the wellbore is at a lowertemperature (Figure 7(B)), the vertical stresschange is also enhanced. The maximum tensilestress change in the vertical direction also existsat the position along the maximum tectonic stress.Moreover, if the vertical stress becomes less thanthe borehole pressure, a horizontal openingfracture may well be formed at this corner. Theplane strain assumption adopted here is expectedto produce the strongest estimate for change inthe vertical stress because of the built-inassumption of zero strain in the axial direction.P/|S0|P/|S0|P/|S 0 |4.14.03.93.83.70.0-0.2-0.4-0.6-0.8-1.0-1.24.24.03.83.63.43.23.02.82.6(a)2 4 6 8 10r/r w(b)2 4 6 8 10r/r w(c)2 4 6 8 10r/r wFigure 2: Pore pressure for (a) mode I, (b) mode II and (c)mode I+II for θ=0.Temperature T/TR*Temperature T/TR *1.000.950.900.850.800.750.0000-0.0001-0.0002-0.0003(a)t=0.9 st=1.5 minst=2.5 hrst=10.4 dayst=2.85 years2 4 6 8 10r/r w(b)t=0.9 st=1.5 minst=2.5 hrst=10.4 dayst=2.85 years2 4 6 8 10r/r wFigure 3: Temperature for (a) mode I and (b) mode II at θ=0228

Australian Geothermal Conference 2010-4-7-5-8(a)(a) r/|S0 |-6 /|S 0|-9-7-10-82 4 6 8 10r/r w-114.02 4 6 8 10r/r w0.43.5 r/|S0 |0.20.0-0.2-0.4(b) /|S 0|3.02.52.0(b)-0.61.5-0.8-1.0-42 4 6 8 10r/r w1.0-5.01.0 1.2 1.4 1.6 1.8 2.0r/r w-5.5-5(c)-6.0(c) r/|S0 |-6 |S 0|-6.5-7-7.0-7.5-8-42 4 6 8 10r/r w-5.02 4 6 8 10r/r w-5.5-5(d)-6.0(d) r/|S0 |-6 |S 0|-6.5-7-7.0-82 4 6 8 10r/r wFigure 4: Radial stresses at θ=0 for (a) mode I, (b) mode IIand (c) mode I+II at Tw * =373 K; (d) mode I+II at Tw * =423 K.Thermal effects on wellbore stabilityWellbore breakdownThe hoop stress on the wall of the well can becalculated theoretically. Based on rock strengththeory, the breakdown pressure is obtained asfollows (Wu et al. 2009)* *6 GK ( T T)*B B R WT2R0 2PR4S0KB 4 G/3pb2(1 )where σ T is the tensile strength. When S 0 0, θ is taken as /2 toobtain the minimum pressure.,-7.52 4 6 8 10r/r wFigure 5: Hoop stresses at θ=0 for (a) mode I, (b) mode IIand (c) mode I+II at Tw * =373 K; (d) mode I+II at Tw * =423 K.Wellbore shear failureThe Mohr Coulomb strength criterion is used toassess wellbore shear failure risk based on thesolutions for σ r , σ θ , σ rθ and σ z . In terms of principaleffective stresses, the criterion may be written as: (1sin ) / (1sin )''1 s3where σ s and are the unconfined compressivestrength and the internal friction angle’respectively. The stresses σ 1 ’ and σ 3 are themaximum and minimum principal effectivestresses based on Terzaghi’s definition. A safetyfactor for shear failure is defined based on theMohr-Coulomb strength criterion as' 'SF (1 ssin ) / (1 sin ) 3 / 1.229

Australian Geothermal Conference 2010SF>1 indicates that the effective stress state isbelow the shear strength envelope and thereforesafe from shear failure, whilst SF

Australian Geothermal Conference 2010can be reduced in magnitude if the well becomescooler. This may lead to the formation of openingfracture from the borehole.3. The breakdown pressure is reformulated toinclude the thermal effect.4. In addition to the shear failure in the minimumhorizontal stress direction often observed as widebreakout, shear failure can also occur in themaximum horizontal stress direction d ue to thewellbore cooling effect. However, the later isdependent on the plane strain assumption.(A)(B)Figure 8: Safety factor for different wellbore temperature.ConclusionsIn this paper, the semi-analytical solutions for thetemperature, pore pressure and stress around awellbore are obtained. By comparing the resultswith those from the poroelastic solutions, thefollowing interesting results are drawn:1. The thermal effect can affect the pore pressureat small times and in the near-well region, but thisinfluence decays with time.2. The thermal effects can elevate not only theradial stress but also the hoop stress. This effectbecomes more important with time and for acooler well. It is also found that the vertical stressReferencesRice, J.R. and Cleary, M.P., 1976, Some basicstress diffusion solutions for fluid saturated elasticporous media with compressible constituents,Rev. Geophys and Space Physics, v.14, no.2,p.227-241.Detournay, E. and Cheng, A. H-D., 1988,Poroelastic response of a borehole in a nonhydrostaticstress field, Int. J. Rock Mech. Min.Sci. & Geomech. Abstr., v.25, no.3, p.171-182.Ekbote, S., Abousleiman, Y., Cui, L. and Zaman,M., 2004, Analyses of inclined boreholes inporoelastic media, Int. J. Geomech., v.4, no.3,p.178-190.Rajapakse, R.D.. 1993, Stress analysis ofborehole in poroelastic medium, J. of Eng. Mech.v.119,no.6, p.1205–1227.Biot, M.A., 1941, General theory of three-M., 1989, A thermoelastic theory ofdimensional consolidation, J. Appl. Phys., v.12,no.2, p.154-164.McTigue, D.F., 1986, Thermoelastic reponse offluid-saturated porous rock, J. Geophys. Res.,v.91 (B9), p.9533-9542.Kurashige,fluid-filled porous materials, Int. J. Solids Struct.,v.25, no.9, p.1039-1052.Coussy, O. 1989, A general theory ofthermoporoelastoplasticity for saturated porousmaterials, Journal Transport in Porous Media, v.4,no.3, p.281-293.Wang, Y. L. and Papamichos, E., 1994,Conductive heat flow and thermally induced fluidflow around a wellbore in a poroelastic medium,Water Resour. Res., v.30, p.3375-3384.Li, X., Cui, L., and Roegiers, J.C., 1998,Thermoporoelastic analyses of inclined boreholes,In SPE/ISRM Rock Mechanics in PetroleumEngineering, Trondheim, Norway, p.443-451.Wu, B., Zhang, X. and Jeffrey, R. G., 2009,CSIRO internal report.231

Australian Geothermal Conference 20103D Geological Modelling for Geothermal Exploration in the TorrensHinge ZoneMatthew Zengerer 1 *, Chris Matthews 2, Jerome Randabel 3 .1 10 Holland St, Thebarton, SA, 5031.2 6 Hollywood Way, Glenalta, SA, 5052.3 Torrens Energy Limited. PO Box 506, Unley, SA, 5035.* Corresponding author: mzengerer@hotmail.comThis paper d escribes 2D/3 DGeophysical/Geological M odelling undertaken bythe autho r f or T orrens E nergy i n 20 09, for thepurpose of establi shing 3D Geol ogical Models o fthe uppe r crust within th eir Torren s Hinge Zonetenements in South Australia. Geolo gical Mo delsare a re quirement for a ny Geothe rmal Energyexplorer sea rching for Hot Dry Ro ck or HotSedimentary Aquifer targets. T hese Geol ogicalModels are assigne d thermal parametersdetermined by field and laborato ry measurementsbased on th e litholo gies included in t he mo del.Heat resources m ay then be computed for th egiven volume based on the geological model andthermal rock prope rties. Whil st the studie s we recarried out for the purp ose of cre atingtemperature model s an d geoth ermal resou rceestimates, this document will di scuss thegeological models, geophysical investigations andtectonic impl ications ge nerated by the stu dy.Examples of Geological M odels are sh own fromthe Torren s Energy Po rt Augusta, Y adlamalkaand Pa rachilna G eothermal Pro spects, ba sedover their GELs 230, 234, 235, 285 and 501.Keywords: South Aust ralia, 3 D GeologicalModelling, Geophy sical Modellin g, Heat Flow,Stratigraphy, Torren s Hinge Zo ne, Adelaid eGeosyncline, Magneti cs, Euler Deconvolution,Geothermal Energy.Introduction and Geological SettingThe principle behind 3D Geological Modelling is tointegrate existing geological and drillinginformation with ge ophysical data such a sseismic, gravity and magneti cs, in a 3 Denvironment which will simulate the 3D geology ofthe region of interest by creating a 3D geologicalmodel which ca n the n b e u sed a s a n inp ut fo rthermal modelling.The Torrens Hinge Zone is a lo ng but narrow (upto 40 km wide) geological transition zone betweenthe relativel y stable Eastern Ga wler Craton“Olympic Do main” to the west an d the folde dsedimentary ba sin known a s th e Adelaid eGeosyncline to the e ast (Drexel et al, 1993). TheTHZ is essentially a region of overlap between theGawler Craton and Adelaide Geosyncline (Figure1). In the study areas of the T HZ, the Adelaideanand Camb rian se dimentary seq uences areunderlain at depths between 2,000 and 7,000m byPaleoproterozoic to Meso proterozoic Ga wlerCraton Olym pic Domain rocks (Matth ews andGodsmark, 2009).Heat flo w va lues of over 90 m W/m 2 have bee nrecorded in several wells drilled in the THZ. Withseveral kilometres th ickness of mode rateconductivity sedim ents o verlying the crystallin ebasement in this re gion, predicted te mperaturesat 5000 m are u p to 3 00 o C in some are as(Matthews and Godsmark, 2009).Data and MethodologyAvailable ge ological data inclu ded hi storic an dTorrens En ergy d rillholes and South Australian(SA) geol ogical map ping. Geophy sical datacomprised SA magnetic and Bouguer Gravity dataalong with recently acquired Torrens Energy andGeoscience Australi a 2D de pth-convertedSeismic. Ro ck thermal p roperties were de rivedfrom lab oratory mea surements pe rfomed byTorrens En ergy an d top and downhol etemperatures measured from drilling.The Intrepid 3D Geo modeller suite wa s used tobuild the 3D Ge ological Mo del. The 3 DGeomodeller software interp olates bet weengeological b oundaries, orientatio n data an ddrillholes to g enerate a 3D geologi cal model thatis flexible an d rea dily ad aptable with addition algeological information. Geomodeller also supportsfaults and fo ld informatio n into its interpol ation.The final ou tput can be discretised into a 3D“voxel” mod el of desired resolution to use in 3 DGeothermal Modelling.232

Australian Geothermal Conference 2010Figure 1. Map of Central South Australia showing tectonic elements and study areas together with Torrens Energywells, seismic lines and prospects233

Australian Geothermal Conference 2010Basic Modelling requirementsThe follo wing method ology wa s a pplied toconstruct a 3 D Geological model. It relies he avilyon inte rpretation and inte gration of m any datatypes, as well as some cr eative th inking an dintuition. All 3D geol ogical mod els built byGeomodeller begin with 4 pieces of information.The first is a defined 3D volume of interest, set b yrectangular coordinates of the target area, and asuitable dept h dimen sion which will encom pass(preferably) the heat source, basin thickness andtopography. For both Parachil na and PortAugusta a d epth belo w Mean Sea L evel of 7kmand a maximum of 1.5km above MSL to clear therange topo graphy was deemed sufficient, givingan 8.5km thi ck mo del. T he are a cov ered in theParachilna model wa s 50x45 km, and 50x80kmover Port Augusta/Yadlamalka.A surface topography is re quired for the model tocompute th e interse ctions of the geologicalinterfaces at su rface, ie to have an initial 2 Dsection at which to i nput geolo gical i nformationand have it reproduced by the model computationfor validation. Topog raphy was imp orted into themodels from SRTM DEM grids at 90 m resolutionclipped to the Area Of Interest (AOI), Figure1.The third and fourth requirements are the definedgeology in the form of lith ological inte rfaces an dassociated stru ctural orie ntations, and astratigraphic pile whi ch defines th e re lationshipsbetween the lithologies. T he latter is particularlyimportant as contacts need to have rel ationshipsdefined a s ei ther Onl apping or Ero sional, as thisdetermines whether interface isolines intersect orremain ap art with repe ct to on e anothe r.Sequences of geolo gical units such a s fou nd i nsedimentary ba sins m ay also be grouped inonlapping se ries, whe reas intrusives are alwayserosional a nd ap pear where requi red in th estratigraphic pile to best sati sfy conta ctrelationships. In st rongly deform ed regi ons,structural requirements may imply that lithologie smust o bey o nlapping rath er tha n e rosional rules.A simplificati on of the stratig raphy wa s alsocreated fo r e ase of mode lling (Fig ure 2), whi chessentially has g rouped the Cam brian a ndNeoproterozoic sequences i nto t heir m ajortectono-stratigraphic units.Figure 2. Simplified Parachilna Stratigraphy.Geological i nformation a t surfa ce came fromimporting simplified or grouped shapefiles derivedfrom State 100k geol ogy polygons, and assigningthese to the sim plified stratigraphic f ormations.Likewise ori entation dat a, whe re known fro mmapping, have been supplied where possible andassigned to specific fo rmations. T orrens Ene rgyand historic diamond drillhole data have providedconstraint in the 3 rd dimensio n. Once again thegeological logs of these d rillholes were simplifiedto the stratigrap hy and the drillhole s importe dusing the import tools in Geomodeller (figure 3).Figure 3. Port Augusta 3D Interface Points and Drillholes inGeomodeller project.Supplementing Geology with GeophysicsUnless the re is a complete cove rage o f mappe dgeological information at surface, it is u nlikely thatreliable mode ls ma y b e d erived fr om s urfacemapping alo ne. To sup plement the existinggeological i nformation, interp retation of four 2 Ddepth-converted seismic lines, (two f rom th eParachilna area and two from the Port234

Australian Geothermal Conference 2010Figure 4. 2D Magnetics Model across Parachilna , showing main basin elements, possible magnetic intrusives andshallow fault-related magnetic bodiesFigure 5. Persepective view of 3D Bodies derived from 2.5D Magnetic modelling within Augusta/Yadlamalka acrossParachilna , including dykes, possible magnetic intrusives and sheet-like layers, probably thrusted Beda Volcanics.Figure 6. Depths to Base Sturtian, ie top Beda Volcanics, derived from Euler Deconvolution of Magnetics. Showntogether with drillholes from Augusta/Yadlamalka prospects.235

Australian Geothermal Conference 2010Augusta/Yadlamalka a rea), was performed (se eAppendix 1). These were con strained whereverpossible by drilling and outcrop i nformation.These seismic sections w ere im ported int oGeomodeller as im age backdrops alo ng verti calsections through the model and the interpretationtransformed into interfac es within these vertic alsections.The inte rpretation alon g these sections wasaugmented f urther in the Para chlina AOI by 2Dmodelling of state mag netic data al ong two line sadjacent to the 09TE-01 seismic line (figure 4). Inthe Port Augusta re gion, 2D mag netic modellingwas also performed in order to obtain estimates ofbasement de pth and mag netic bo dy shape s an dlocations (figure 5). In each case images or actual3D data poi nts from the result s we re importe dback into Geomodeller.Euler Deconvolution of magnetic d ata wa s alsoused in this area to pro vide spot estimates ofdepth to sh allow-dipping volca nic u nits, implyinglocal e stimates to the top of Willouran -age Bed aVolcanics (figure 6 ). Ad ditional scans u sing alarger window size revealed information about theshape and size of de eper magnetic bodies whichcorrelate well with 2D magnetic modelling results.Finally, structural interpretation of both gravity andmagnetic d ata wa s used t o define fault traces inGeomodeller (see Appendix 2).3D Geological ModelsComputation time of th e two 3D Geologicalmodels is dependent on the amou nt of datasupplied. Re sults which are strongly biased toorientation data were trea ted with ca ution. Someadditional inference of contacts and o rientationswas requi red to en sure geological co nsistency,especially near faults. Additional se ctions lin kingdrillholes p rovided strong er cro ss-correlation ofgeological i nterfaces, especially in the PortAugusta/Yadlamalka area where data was wid elyspaced. Although in some ca ses actu al outcro pgeology was further from the area of interest thanrequired, it was incl uded to provide geologicalreality and well-defined constraints on modelling.relatively flat-lying in the west to mod erately eastdippingat the Ediacara fault contact.Figure 7a. Projected section view of Parachilna 3D GeologyModel, view SE. Vertical. exaggeration= 2:1The Edia cara fault zone chan ges in characte rfrom south to north. Our m odelling shows it to bea ste eply west-dipping fe ature, u pthrown on th eeastern side, whi ch evolves i n the south from arelatively si mple antiform, to a complex horststructure inv olving seve ral faults an d possiblybasement o r diapiri c mat erial fro m the cent re ofthe Para chilna AOI to its northe rn edg e.Basement depths in the far west are predicted torange from 2km depth in t he north to 5 km in thesouth. In the vicinity of the Ediacara Fault footwallbasement depth reaches 6km, but this decreasesto 2.5-3km i n the fa r north. In the fa ulted h orstblocks of the Ediacara Fa ult region ba sement ordiapiric material may reach depths of 2km. East ofthe Ediaca ra Fault in the central and southe rnareas is a moderately folded di sturbed regi onleading eastward into a very d eep synclinal basinwithin the Cambri an – Adelaide an seq uencesadjacent to the ran ge fro nt fault. To the north ofthis are a the model pre dicts le ss inten se foldin gwith flatter basi n sequences, but still basinthicknesses of > 7km.As faults also require orientation information to becomputed in the modelling, their direction, dip andapparent mo vement was estimated e ither fromseismic ima ges or ge ological an d geophysicalimages.ParachilnaEmphasis on interpretatio n an d m odelling in th eParachilna area focu ssed on area s west of theEdiacara fault (Figure 7a, 7b, Appendix 2) wherethere is sim pler geology according to the seismicdata. In this area we se e from modelling that theEarly Camb rian to A delaidean ro cks transit f romFigure 7b. Interfaces and Faults, Parachilna 3D. View NE,VE=3:1236

Australian Geothermal Conference 2010Cainozoic erosional unit s ten d to range inthickness f rom 300 -400m and thi n o ut in thevicinity of the Ediacara Fa ult hangi ng wall. Earlyto Middle Cambri an unit s of the La ke Frome,Unnamed, and Ha wker Groups h ave beenconfirmed b y deep drilli ng, whil st Adelaid eanrocks of Mari noan to Sturt ian age s a re expecte dto lie b eneath the se based on their p resence inthe Edia cara Hill s to the north in th e fault zo neand outcrops of the se units to the we st on theStuart Shelf. Inference of Torrensian to Willouranrocks is ba sed on the apparent p resence ofdiapirs to the south an d east, although theserocks are expected to thi n out somewhere in thi sregion. Torre nsian ro cks in parti cular may notappear anywhere west of the Edi acara F aultsystem. Basem ent has b een largel yundifferentiated in the Pa rachilna 3 D model wit hthe exception of what are inferred to be two largemagnetic intrus ives in the c entral parts andnorthwest.Port Augusta/YadlamalkaA very large area has been modelled in this study,which naturally implie s a greater amount of b othinference and uncertainty. To complement this i smore complex geology and more drilling data, aswell a s mu ch more dive rse ma gnetic sign aturesand seismic imagery.The final 3D model is shown as sections andsurfaces in Figures 8a and 8b. In general thestratigraphy is identical to that used at Parachilna,with the addition of several intrusive basementunits of varying magnetic signature and a split ofthe Willouran-age rocks into BedaVolcanics/Backy Point Formation, generalWillouran – age rocks and several thrustedWillouran – age slices. The overall trendsmodelled show an eastward – thickening wedgeof sediments toward the range front, with possibledown-thrown rift basins evolving closer to therange. To the northeast thrust faults occur subparallelto the Gairdner Dyke Swarm before therange front and are possibly facilitated byCallanna group evaporitic units and/or diapirism.To the east of these significant thrusts, closer tothe range fault, sediments begin to dip morestrongly or are deformed. Some deformation isalso present immediately west of the thrusts onthe footwall side, but this peters out as sedimentsbecome more flat-lying to the west.In the south and south central areas in the vicinityof Port Augusta, the model predicts gently dippingAdelaidean rocks of 800 – 1200m thickness,overlying granitic or metamorphic basement,crossing a significant north-northeast trendingbasin-defining fault approximately 10km northwestof Port Augusta, coincident with the onset of asignificant gravity low and a step in the magneticsand seismic images. Here the basin thicknessincreases to at least 2000m, dropping awaygradually to 3000m approximately 9km west ofthe range front, before dipping away more steeplyto 4500m at the base of the range fault. Deeperreflectors in the Augusta seismic section areenigmatic but their pattern is suggestive of rifting,the timing of which should be investigated further,if proven (see Discussion). Modelling resultsbased on historical drilling near Wilkatana (centralsouthern Yadlamalka prospect) demonstratelocalised basins containing Cambrian sediments.The interpretation on the seismic lines 09TE-02(Augusta) and 08GA-01 (Arrowie) is still thesubject of some debate, until such time as somespecific drilling is performed along the lines andstratigraphic correlations are performed. Theinterpretation has been made through carefulcross-comparison of nearby drillholes usingGeomodeller as a guide, using logged depths tospecific units and comparing how they aredistributed across the area, and existinggeological mapping.Inferences have been made on the nature anddistribution of the Willouran and Torrensian agerocks in a fashion similar to the interpretation atParachilna, as well as deeper Proterozoicbasement units. These inferences will bediscussed in the next section.Figure 8a 3D volume model of Augusta/Yadlamalka. ViewNE, VE=3:1.The northern parts of the model, described asYadlamalka after the name of the stationsoutheast of Lake Torrens near a major thrustzone, has similar gently dipping Adelaideanstratigraphy ranging from 600m thickness in thefar west to 2300m before encountering asignificant thrust fault which displaces Willouranvolcanics close to surface, as evidenced by theArrowie seismic section, drilling and magnetics.Unlike the more obvious extensional faultingregime exhibited farther south, basin units appearless disturbed, but a very marked angularunconformity exists between thesedimentary/volcanic pile and the basement,which comprises a series of west-dippingreflectors. East of the Yadlamalka thrust the237

Australian Geothermal Conference 2010sedimentary pile is moderately folded and ofunclear composition and thickness. What is clearis that thrusting and compressional deformationhas disturbed the Adelaidean sequences in morethan one location. It is also speculated thatWillouran/Callanna Group salt formations mayagain be principally involved.Figure 8b 3D section view of Augusta/Yadlamalka. View NW,VE=3:1.In the northern part of the model magneticmodelling has suggested that a number ofmagnetic intrusives are present, along with slicesof Willouran age volcanics. These have beenemplaced into the model where most likely, butthe remainder of the basement has beenundifferentiated.Partial modelling of the range to the east hasbeen included to add both strratigraphic andstructural control, but has been performed at arepresentative level only. Cainozoic strata alsoappear relatively thin in this region but have beenincluded as a thin uppermost layer where evident.Overall, the Early Cambrian to Neoproterozoicbasin succession throughout the area remainsrelatively simple. Sequences dip gently eastwardand thicken before the onset of thrusting, folding,or extensional faulting. Roughly midway acrossthe model Torrensian lithologies are inferred toappear (see Discussion).DiscussionResults and some of the methodology used in thePort Augusta/Yadlamalka area require somediscussion due to inferences being made and thetectonic consequences of the modelledrelationshipsThese inferences particularly concern the natureof the rocks within the basin sequences west ofthe range front fault. Torrensian/Burra Grouprocks are not known to occur west of this structureor anywhere on the Stuart Shelf, however wehave inferred from the seismic that Torrensianrocks pinch out or fault out fairly consistently 20-30km west of the range front, mainly at depthsdeeper than what has been currently encounteredin drill holes. The implication from this is thatTorrensian sedimentation is controlled by asecond onset of strong rift-related extensioncloser to the Central Flinders Zone.Willouran age/Callanna Group rocks have longbeen enigmatic in terms of their relationship toother rocks of the Adelaide Fold Belt and theGawler Craton, but inferences on their distributionand geology have implications for understandingthe early evolution of the rift zone and laterdeformation. Basaltic rocks of the Beda Volcanicshave been encountered in many drillholes on theStuart Shelf, mostly at shallow depths, but alsomay have been confused with older GawlerRange Volcanics (GRV) in some holes. Theserocks and associated intertonguing sediments ofthe Backy Point Formation are thought to form thebasal units of the Adelaide Fold Belt rift sequence.Exactly how thick these units are, and whetheradditional sediments lie beneath the volcanics, isunknown, however drillhole information suggeststhat the Beda Volcanics are probably only a fewhundred metres thick, and their flat-lying naturesuggests that they should form a good reflector inseismic data.Both seismic lines in the area show good deepcontinuous reflectors just above or not far abovethe basement, but in places drillholes appear tohave intersected Beda Volcanics well above thesuspected top basement horizon, at anotherreflector. The Port Augusta seismic line evenappears to show what may be another rift grabenwell beneath what is interpreted as Willouran, butuntil this is confimed by further studies we haveassumed it to be basement, or at least GRV orsimilar age sediments. Elsewhere we haveinferred that the Beda Volcanics are reasonablythick or probably have much larger proportions ofBacky Point Formation beneath them for thepurpose of modelling. The unmodelled otherpossibilities are that beneath the Beda/BackyPoint rocks lie reasonable thicknesses ofMesoproterozoic rocks such as the PandurraFormation and Gawler Range Volcanics (GRV), orthat Torrensian and Sturtian rocks increasedramatically in thickness midway across themodel, or both.Further information was derived from themagnetics. Petrophysically, the Beda Volcanicsare known to be moderately magnetic, but theirflat-lying nature and probable magnetisationpattern means that unless they are tipped on theirsides, the net effect is only to add a regional dcshift to the overall magnetic signal. High pass andanalytic signal filtering of the state magnetics,however, has revealed where Beda Volcanicshave been tipped on their sides and thrust tosurface at the Yadlamalka thrust and Depot Creekon the range front, where they have been238

Australian Geothermal Conference 2010intersected at surface or in drillholes, with thethrust also visible in the Arrowie seismic section.Euler Deconvolution of magnetics is basedaround automated scanning of changes in themagnetic signal in windows across a grid, withdepth estimates to magnetic sources dependenton the size, intensity, and shape of bodies, andwindow size. Two different window sizes wereused, the first scanning to depths of 2km and theother to 8km. The results of the shallower windowtargeted mainly the tops of the Gairdner DykeSwarm, which are known to be Beda Volcanicequivalent ages. These are therefore byimplication minimum depths to the base ofSturtian Adelaidean rocks in the west, whereTorrensian rocks are not present, and base ofTorrensian rocks in the east where the methoddetected thrusted Beda Volcanics. These resultswere statistically culled for reasonableness andproximity to dykes, thrusted volcanics orcloseness to interception in drillholes.The remarkable consistency of Euler depth trendsin the west for the shallow scan gives credence tothe notion that Beda Volcanics overlie deepersediments and/or volcanics. The idea that GawlerRange Volcanics lie below the obvious basementunconformity seen in the Arrowie seismic is notsupported by this model, unless east of the majorvolcanic flows on the Gawler Craton the GRVwere deformed along with Hiltaba suite anderoded prior to either Mesoproterozoic clasticdeposition or Willouran rifting. Similar notionsapply in the Port Augusta area.The deeper window targeted magnetic intrusions,by implication a maximum depth to basementacross the area. Again these results werestatistically culled and sorted by proximity to theinferred “tops” of magnetic body depths. Theseresults, along with results of 2D magneticmodelling were interpreted to be from deepmagnetic intrusives, possibly even Iron OxideCopper Gold – bearing intrusives lying in thecentre of the region (though perhaps too deep fordrilling), rather than belonging to metasediments,GRV or younger volcanics. Further support forthis notion came from examination of filteredmagnetic grid images as described above.Values from the deeper Euler and 2D magneticmodelling were input back into Geomodeller asthe basis for intrusive 3D body shapes.Results of 3D modelling in the Parachilna areaare currently considered less contentious. Theprincipal features for discussion are the nature ofthe magnetic anomalies, whether they indeedcorrespond to deep intrusives, or are related todiapirism bringing basement magnetic material tocloser to surface, or a combination. The stronggravity anomaly running through the centre of theregion suggests it is more likely to be basementreactivated along the Ediacara Fault. Once againthe other major issue for consideration is thepresence and distribution of Torrensian andWillouran rocks.ConclusionTwo large 3D Geological models have been madeof se ctions o f the Torren s Hing e Zon e in SouthAustralia. Th e north ern model n ear Parachilnashows mo derately di pping Cambrian t oAdelaidean sediments of thickne sses between2000 and 4000m overlying unknown basement inthe we st. A modern a ctive fault system, theEdiacara Fault, has uplifte d basement locally an dforms the boundary of a d eep, moderatelydeformed n orth-south b asin a djacent to theFlinders Range front.The southe rn model bet ween La ke T orrens a ndPort Augu sta sho ws tra nsitions bet ween ge ntlydipping Adel aidean sequ ences of up to 2300 mthickness with thrus ted, deformed bas ins to thenorth a nd d eeper rifted basi ns to the so uth.Evidence fro m sei smic a nd ma gnetics suggeststhe presen ce of sub stantial sedim entary o rvolcanic sequences lying beneath Willouran BedaVolcanics and overlying deformed metamorphic orgranitic basement with sharp unconformity.ReferencesMatthews, C. G. & God smark, B., 2009.Geothermal Energy Prospectivity of the TorrensHinge Zone: Evidence from New Heat Flow Data.AGEG 2009 Conference Extended Abstracts.Drexel J. F. Preiss W. V. & Parker A. J. eds.1993. The Geology of South Australia, GeologicalSurvey of South Australia Bulletin 54239

Australian Geothermal Conference 2010Appendix 1: Seismic Surveys and Results (after Matthews and Godsmark, 2009)Parachilna Seismic Survey, completed January 2009Interpretation belowCambrianBase CainozoicNeoproterozoicEdiacara FaultMesoproterozoicAnd Older Basement240

Port Augusta Seismic Survey, completed May 2009Australian Geothermal Conference 2010Base CainozoicNeoproterozoicMesoproterozoicAnd Older BasementRangeFrontFault241

Appendix 2: Fault Interpretation from MagneticsAustralian Geothermal Conference 2010Fault Interpretation (blue and yellow lines) of SA Magnetics in the Port Augusta/Yadlamalka area.Image is a Tilt Angle filter draped over Analytical Signal Amplitude. Torrens Energy GELs in black.MGA Zone 53.242

Australian Geothermal Conference 2010Numerical simulation of geothermal reservoir systems withmultiphase fluidsZhang, J. 1 *, Xing, H.L. 11 Earth Systems Science Computational Centre, School of Earth Sciences, The University of Queensland, QLD 4072*Corresponding author: ji.zhang@uq.edu.auThis p aper descri bes a modula r framework forsimulating geothe rmal well and reservoi rperformance. The numerical model i s able toaccount for transient, three-dimensional, single- ortwo-phase flu id flow in normal hete rogeneous o rfractured-matrix formation s. Both cond uctive andconvective heat flow are accou nted for and fluidstates in the reservoi r ca n ran ge bet ween liquidphase, two -phase ste am-water mi xtures tosuperheated steam. Both short term well reactionand lo ng te rm reservoi r perfo rmance ca n bemonitored as dynamic ma ss/heat flow processes.Equations-of-state (E OS) for wate r an d CO 2 areintegrated as p art of the fully-cou pled nonlinearfinite element simulator. In this paper, the code hasbeen ap plied for use in geotechnical riskassessments for mine develop ments ingeothermal a reas. Com parison of EGS-CO 2 andEGS-water i s dem onstrated in the paper a ndresults show that the simulat orPANDAS/ThermoFluid ca n be u sed for bothconventional geothermal fields and enhance dgeothermal systems.Keywords: Nume rical simulatio n, Enhan cedgeothermal system, EG S, Geotech nical ri skassessment, Carbon dioxideIntroductionGeothermal energy is regarded a s a renewable,clean, co st effective ene rgy source, whi ch i sbecoming increasingly attractive, espe cially sincethe enhanced geothermal system (EGS) has beenproposed and applied as a new type of geothermalpower technology. Geothermal reservoir modelinghas been played an import ant role as a n integralpart of reservoir assessment and ma nagement inthe p ast few de cades. Although compute rmodeling is routinely applied in hydrothe rmalreservoir e ngineering, the re are severa l asp ectsthat continually need improvement:(1) Two-phase flow phenomenology. This includesthe impleme ntation an d calibration of relativepermeability and capillary pre ssure, t wo-phaseflow in fra ctures an d m ass tran sfer betwe enphases (evaporation and condensation);(2) Non-equilibrium models. To relax assumption sof local thermodynami c equilibri um, severalnon-equilibrium model s h ave bee n d eveloped,including double porosity models (Warren & Root,1963), explicit fracture treatment (fo r sp arselyfractured systems) an d st atistical models(“MINC”-TYPE, Pruess & Narasimhan, 1985) (fordensely fractured systems);(3) Natural-state modelin g. Further vali dation ofthe co mputed natu ral st ate against e arlyexploration well data is required;(4) Cons titutive repres entations. Res ervoir fluidswith dissol ved solid s (i.e. Na Cl) a ndnon-condensable ga ses are no w available inseveral si mulators. Mo deling of fluid un derextremely high tempe rature and pressu re(supercritical state) is already being conducted butnot in the geothermal modeling arena;(5) Coupled reservoir chemistry. There are a fewcalculations that have been reported incorporatingreactive che mistry, coupled with flui d flow bydissolution and precipitation of solids (creation anddestruction of porosity and permeability), however,this area still provides a challen ge to g eothermalmodeling;(6) Autom atic inv ersion te chniques. In versemodeling uses an ite rative inversion procedure todrive co nventional forward reservoi r model s(treated as sub routines). With the increa se inmodel complexity and de grees of fre edom, thecomputational cost is extremely high.In conclusion, further computational modeling andcode development is u rgently needed t o improveour understanding of geothermal reservoir and therelevant nature. It is also needed for the enhancedevolution such a s of enhanced g eothermalreservoir sy stem, and a chieves a m ore accurateand comp rehensive rep resentation of rese rvoirprocesses in more details. It also help s to redu cethe unce rtainties in model s, and to enhan ce thepractical utility and reliability of reservoir simulationas a basis for field development and management.This p aper will focus on our research ef fortstowards the simulation of enha nced g eothermalreservoir systems with multiphase fluids.PANDAS/ThermoFluidPANDAS (Parallel Adaptive Nonlinear DeformationAnalysis Sof tware) i s a m odular sy stem of finiteelement met hod ba sed modules. Cu rrently, itincludes the following four key components:• ESyS_Crustal for the interac ting fault sys temsimulation;• PANDAS/Fluid for simulati ng the fluid flow infractured porous media;• PANDAS/Thermo for the therm al analysis ofmetals and fractured porous media;243

Australian Geothermal Conference 2010• PANDAS/Pre and P ANDAS/Post forconceptual modelin g, mesh g eneration andvisualization.All of the abo ve modules can be u sed individuallyor togeth er to simulat e phe nomena su ch asinteracting fa ult system dynamics, hea t flow andfluid flow with or without coupling effects.PANDAS/ThermoFluid (the fully coupl ed modulesof PANDAS/Thermo and PANDAS/Fluid) is a finiteelement method based module for sim ulating thefluid and heat flow in a fra ctured porous media bysolving the conservation equations of macroscopicproperties numerically.The mass and energy conservation equations fortransient two-pha se water/steam couple dheat/fluid flow in porous medium used in PANDASare given in Equations (1) and (2):[ ( Sw w Ss s)] tKkrw(1)[w Pw wgZ]wKkrs[ s Ps sgZ]q ms[ ( Sw wuw Ss sus) (1 ) rcr] tKk(2)rw[ whw Pw wgZ]wKkrs[ shs Ps sgZ]( T)qesTo compl ete the governing eq uations, it isassumed th at Darcy’ s Law a pplies to themovement of each phase (momentum balance):Kkrwvw Pw wgZ(3)wKkrsvs Ps sgZ(4)swhere P is fluid pre ssure [Pa]; is the poro sity; Sis the phas e s aturation, with S w +S s =1; is thedensity [kg/m 3 ]; K is th e intrinsi c permeabilitytensor of the porous medium [m 2 ], k r is the relativepermeability of the phase; μ is the dynamicviscosity [kg/m·s]; g is gravity, Z is the depth; u andh are specific internal energy and specific enthalpyrespectively [kJ/ kg]; is the thermal condu ctivitytensor of th e po rous m edium [ W/m·K]; T istemperature [℃]; q m and q e are source/sink termsof the total mass and energy respectively [kg/m3·s,kJ/m3·s]; v is the fluid veloc ity vector in [m/s]; thesubscript w, s an d r denote wate r, ste am ph aseand rock matrix respectively.Thermodynamic p roperties of reservoir fluid s(such as water and CO 2 ) are of vital importance tounderstand the subsurface physical-chemical andgeological processes. The most acce ptedIAPWS-95 formulation (Wagner & Prub , 2002) fo rwater equation-of-state (EOS) is integ rated in oursimulator, which allows us to retrieve basicphysical parameters such as density and dynamicviscosity, as well as the saturated properties forphase transition modeling. SWEOS, an EOS forCO 2 which was o riginally developed by Span andWagner (1 996) ba sed on their algo rithm of anempirical representation of the fundam entalequation of Helmholtz e nergy, ha s also be enimplemented into our simulator to model EGS-CO 2processes.For more details of PANDAS, please refer to Xing& Makinouchi (2002), Xing et al. (2006a , b, 2007,2008, 2010 ). Model va lidation of PANDAS/ThermoFluid code is reported in Xing et al. (2008,2009).PANDAS/ThermoFluid application inGeotechnical risk assessments formine developments in geothermalareasTo demonstrate the feasibility and value ofmulti-phase fluid flow modeling as a pit design tooland for risk analysis of geothermal hazards duringmining in ho t ground, a numerical model wa sdeveloped f or a gold mine site usin g thePANDAS/ThermoFluid model code .PANDAS/ThermoFluid was used for sim ulation oftemperature, pressure an d steam dist ribution inthe se awall and fou ndation of the da m along anorth-northeast to south -southwest oriented crosssection.(a) Stage 1(b) Stage 2(c) Stage 3Steam Zone Mixed Zone LiquidWater ZoneFigure 1: PANDAS/ThermoFluid results of the cross sectionmodel of pit excavation for three development stages.For dem onstration pu rpose, three pit excavationstages were simulated. The first to a depth of 90m,the second reaching the t op of the b oiling zone at180m and the third down to 270m below surface.2

Australian Geothermal Conference 2010It was found that after the first stage of excavationto 90m stea m did not d evelop throu ghout themodel domain despite temperatures computed tobe p artly sig nificant a bove 100º C. Thi s is due tohigh hyd rostatic pre ssures preventin g water f romboiling at this stage. The simulation results for thesecond and third develo pment st ages predi ctedboiling to occur below the pit floor and in the lowerbenches of the seawall.Short-term/long-term modeling offracture dominated EGS-water andEGS-CO 2Operating e nhanced geothermal sy stems (E GS)with CO 2 instead of water as a heat tra nsmissionfluid is a new attra ctive concept wit h severalbenefits incl uding: The lowe r visco sity of CO 2would yield large r flow velocities for a givenpressure gra dient; less p ower con sumption forfluid circulati on system s due to buoyancy ef fectsand geologic storage of green house gases as anancillary benefit (Brown, 2000, Pruess, 2006). Thefollowing se ction demon strates simulations u singEGS-water and EGS-CO 2 under typica lgeothermal field situatio ns to compare th eefficiency and longevity of both systems as well asthe estimation amount of geologic storage of CO 2 .Pruess (2006) has compared the therm odynamicproperties of CO 2 and water under typical reservoirconditions. A five-spot well pattern geothermal fieldwas model ed by T OUGH2 (Prue ss, 20 06) an d aseries of co mparisons betwee n EGS-CO 2 andEGS-water h as dem onstrated the advant ages ofusing EGS -CO 2 su ch as the long-te rmperformance (up to 36 ye ars) heat exaction rate,mass flow rate and pressure/temperature field.In the deve lopment of enhanced g eothermalsystems, greater focus is shifted to the short-termdynamic response of well tests. A number of tests(with o r with out tracers) need to b e run betwee nthe injectio n and p roduction well s bef ore p owergeneration to assess the ability of prod uction andcirculation within an EGS. Here, a simplified500x200m fracture domi nated geoth ermal field(Figure 2) is modeled by PANDAS/ThermoFluid tocompare b oth long -term an d short-termperformance of EGS-CO 2 and EGS-water.Figure 3 sho ws the fluid ve locity evolution at anearly st age of the well test. With the sam epressure drop between injection well an dproduction well, CO 2 ha s highe r flow rates th anwater (2.5×10 -3 m/s for CO 2 to 8.7 ×10 -4 m/s forwater at its stable stage). It can also be seen thatCO 2 takes about 81 hours to reach the stable state,while water takes less than 50 hours. The reasonfor the longer stabalization time for CO 2 is becauseit has a la rger ratio of fluid den sity to viscosity(Pruess, 2006) and a larg er compressibility (about8 times of water) than water under reservoi rpressure/temperatures.Figure 3: Fluid velocity evolution at early stage of well test.Injection well pressure is 44MPa, constant pressure drop of10MPa is set between the injection well and production well.Figure 4 shows a long term temperaturedistribution along the f racture zone. For thi sparticular case, it t akes l ess than 10 years forEGS-CO 2 to redu ce temperature within the entireregion to und er 200℃, while EGS-water requiresmore tha n 2 0 years befo re temperature dro ps to200℃. This is due to CO 2 ’s higher heat extractionrate which was suggested by Pruess (2006).Figure 2: A fracture dominated geothermal field. L=500m,D=100m, H=0.1m. Permeability of the fracture zone is 5×10 -14 m 2 . Rock density is 2.65×10 3 kg/m 3 , specific heat is1kJ/(kgK), thermal conductivity is 2.1W/(mK). Initialtemperature of entire region is 250℃, the Injection fluid is 90℃.Figure 4: Comparison of fracture zone temperature atdifferent years. Constant pressure drop of 10MPa is setbetween the injection well and production well. With thefracture space of 100m, flow rate for water and CO2 are7.3kg/s and 10.8kg/s, respectively.ConclusionsComputer modeling of g eothermal systems ha sbeen widely used i n indu stry. Furthe rcomputational modeling and code development isurgently needed to impro ve our unde rstanding of

Australian Geothermal Conference 2010 22kkgeothermal reservoir and the relevant nature. It isalso needed for the enhanced evolution such as ofenhanced g eothermal rese rvoir system, andachieves a more a ccurate and com prehensiverepresentation of re servoir processe s in mo redetails. It also hel ps to reduce the uncertainties inmodels, and to enhance the pra ctical utility andreliability of reservoir simulation as a basis for fielddevelopment and man agement. This p aperintroduced a simul ator ‘PANDAS/ThermoFluid’ tomodel the transi ent, non-isothe rmal, multipha sefluid flow in hetero geneous o r fra ctured-matrixformations for simul ating geotherm al well andreservoir performance. It was applied for use in ageotechnical risk a ssessment of a n open pitmining site si tuated in hot ground. As well as tolook at sho rt-term/long-term modelin g of fractu redominated e nhanced geo thermal systems (EGS )with wate r and CO 2 as h eat transmi ssion fluid s,PANDAS/ThermoFluid ha s p roven t o be avaluable to ol to su pport the develo pment a ndimplementation of g eothermal risk managementstrategies to combat geothermal hazards. Resultsalso sh ow its ef ficiency and usefu lness insimulating both conventional geothermal fields andenhanced g eothermal sy stems with multiphasefluids.AcknowledgementsSupport is gratefully ackno wledged by theAustralian Rese arch Council an d Ge odynamicsLimited thro ugh the A RC Linkage proje ctLP0560932.ReferencesBrown, D.W ., 2000, A Hot Dry Rock g eothermalenergy concept utilizing su percritical CO 2 insteadof wate r: Proceedi ngs of the T wenty-FifthWorkshop on Geothermal Reservoir Engineering,p. 233–238.Pruess K., and Narasimhan T . N. , 1985, APractical Method for Modeling Fluid and Heat Flowin Fractured Porous Media: Society of PetroleumEngineers Journal, v. 25, no. 1, p. 14-26.Pruess K., 2 006, Enha nced ge othermal syste ms(EGS) us ing C O 2 a s worki ng fluid - A novelapproach for generatin g rene wable e nergy withsimultaneous sequ estration of carbon:Geothermics, v. 35, p. 351–367.Span, R., and W agner, W., 1996, A new equatio nof state for carbon dioxide covering the fluid regionfrom the triple-p oint temperature to 1 100K atpressures up to 800 MPa: Journal of Physical andChemical Reference Data, v . 25, no. 6, p.1509–1596.Wagner W., and Pru b, A., 2002, The IAPWS forthe therm odynamic prope rties of o rdinary watersubstance for general and scientific use: Journal ofPhysical and Chemical Reference Data, v. 31, no.2, p. 387–535.Warren, J. E., and Root, P. J., 1963, The behaviorof naturally frac tured reservoirs : SPE J ., v. 3, p.245-255.Xing, H.L., and Ma kinouchi, A., 2002, Thre edimensional finite element mo delling ofthermomechanical frictional contact between finitedeformation bodies u sing R-mi nimum strate gy:Computer M ethods in Applied Me chanics andEngineering, v. 191, p. 4193-4214.Xing, H.L., Mora, P., and Makinouchi, A., 2006a, Aunified frictio nal descri ption and it s ap plication tothe simulation of frictional instability using the finiteelement m ethod: Philo sophical Mag azine, v . 86,no. 21-22, p. 3453-3475.Xing, H.L., Mora, P ., 2006b, Co nstruction of anintraplate fa ult system mo del of South Australi a,and simulation tool for the iSER VO institute see dproject: Pure and Ap plied Geop hysics, v. 163, p.2297-2316. DOI 10.1007/s00024-006-0127-x.Xing, H.L., Makin ouchi, A. and Mora , P ., 2007,Finite element modeling of interacting fault system:Physics of the Earth and Planetary Interiors, v. 163,p. 106-121. doi:10.1016/j.pepi.2007.05.006.Xing, H.L. 2008, Progress Report: SupercomputerSimulation o f Hot Fra ctured Rock G eothermalReservoir S ystems: E SSCC/ACcESS T echnicalReport, The University of Queensland, p.1-48.Xing, H.L., Gao J., Zha ng J. and Li u Y., 2010,Towards An Integrated Si mulator F or EnhancedGeothermal Re servoirs: Proce edings WorldGeothermal Cong ress 2010. Bali, Indonesi a,25-29 April 2010.246

Australian Geothermal Conference 2010Development of Hydraulic Fractures and Conductivity near a WellboreXi Zhang and Robert G. Jeffrey*CSIRO Earth Science and Resource Engineering, Melbourne VIC, AustraliaABSTRACTHydraulic fra ctures play an imp ortant role i nestablishing a hi gh co nductive path wayconnecting the wellbore and reservoir. The n earwellborefracture develo pment is cru cial to theefficiency of a fractu re t reatment a s a smoothfracture path will redu ce the flow imp edance. Anumerical model is u sed to sim ulate thepropagation pathway s followed by hydrauli cfractures, in the presen ce of a p ressurised welland natural fractures. By including the coupling ofviscous fluid flow and rock defo rmation in theanalysis, we find that th e non-uniform pre ssuredistribution along the fractures and the removal ofnear-well fra cture closure result in growth of afracture with less curvature (a me asure o ftortuosity) i n its path. More over, th e fractu recurvature ca n be cha racterised b y tw odimensionless pa rameters. If multiple fractu rescoexist, fracture lin kage is po ssible due t omechanical i nteraction between fra ctures, p reexistingnatural f ractures a nd th e well bore.Natural fract ures can alte r the hydraulic fra cturepropagation direction, can offset the fra cture pathand can p rovide sites for new fracture growth. Inthe case of two hydraulic fractures initially growingin op posite directions from the well bore, theinteraction with a natu ral f racture on o ne si de ofthe well will result in the other fract ure wingpropagating faster. Th e complexity of the nearwellboref racture ge ometry can b e st udied withthe aid of our numerical model.Keywords: Near-well fracture tortuos ity, c ouplingmechanisms, naturally fractu red reservoi r,numerical modellingINTRODUCTIONMultiple sub parallel fra ctures can b e gene ratednear th e wel lbore because of fractu re initiationfrom pe rforations spaced alon g an d pha sedaround the wellbore o r from natural fra ctures andflaws that are usually not aligned with the far-fieldstress di rection. Fractu res initiate d from the sedefects will reo rient th emselves to a pl aneperpendicular to the minimum far-field stre ssdirection a s they grow a way from the wellbo re.The ove rall result of su ch a f racture initiationprocess produces ne ar-wellbore fractu retortuosity. As a consequ ence, the net pressu rerequired to extend the fractu res is i ncreasedconsiderably. Beca use of its imp ortance tooptimisation of fra cture treatments, th e stu dy onthis subject has continu ed for several de cades(Hamison a nd Fairhurst, 1970; Daneshy, 1973;Weng, 199 3; Soliman et al. 2008; Cherny et al .2009).Geomechanical models are of significant value i ntreating the se sort s of problems that rel ate tocompeting f racture growth, incl uding l ocal wi dthrestrictions and in creased viscou s frictionalpressure loss. Base d on such geo mechanicalanalyses, various operational remedial techniqueshave bee n p roposed to overco me or avoid thenear-wellbore effects, su ch a s usi ng highe rinjection rates, higher viscosity fluids and orientingslots o r p erforations for fractu re initiation(Manrique and Venkitaraman, 2001; Abass, et al.,2009). However, most studies were based on theuniform p ressurization assumptio n so that thefracture p roblem ca n be treated by conventionalfracture m echanics meth ods. In pa rticular, th efracture turn ing pro cesses, duri ng which thefracture reorients as it grows fro m a wellbore tobecome perpendicular to the minim um prin cipalstress di rection, involve s viscous fluid flow andfrictional slip. In addition, a moving bound aryexists associated with b oth the fluid flow and th efracture tip. The se aspects of th e proble mincrease its compl exity, b ut must be inclu ded i nany realistic model.In this pape r, we will e mploy a 2D numeri calmethod th at can model h ydraulic fra cture growthfrom a borehole. The 2D plane strai n assumptionis justified if the h eight of a slot or natural flaw islarge compared to the well radius. Moreover, themethod can dire ctly co pe with the fractureinteraction with the wellbore and natural fractures.Although the thermal effect is not yet considered,the re sults provide i nsights into fractu redevelopment around the wellbore in a hot dry rockreservoir.PROBLEM STATEMENTSThe pro perties used in the model for the rock,fluid and fra ctures a re a s follows. T he ela sticproperties for the intact rock are Young’s modulus(E), Pois son’s ratio ( ) and the Mode I fractu retoughness i s K IC. The injecte d fluid isincompressible and has a Ne wtonian visco sitygiven by . T he sum of a ll injection rates offracture b ranches g rowing from th e well isassumed to be constant and is denoted as Q in. Itmust be noted that the injected fluid rate into eachfracture varies in time because of the differentgrowth rate s of each fr acture. The total rate247

Australian Geothermal Conference 2010Q ininjected, , is sp ecified as co nstant andrepresents the rate fluid is injected into the well.In addition, the volume change (compressibility) ofthe bo rehole fluid an d ro ck system is n otaccounted for in determining the injection rate.LFigure 1 Fractures around a wellbore, in-situ stress andcoordinate system. The origin of the coordinate system is atthe wellbore centre.The two sides of a fra cture may come intocontact, espe cially ne ar th e well bore. T he mo delallows for contact stresses to develop and uses acohesionless Coulomb friction criteri on for slip onsurfaces in contact ( n p ) (1)f HP wywhere is the she ar stress along th e clo sedfractures, nis the n ormal stress a cting acro ssthem, pfis th e fluid pressure and is thecoefficient of friction.In addition to the tange ntial di splacementdiscontinuity (DD), v, associated with slip alon gthe clo sed fractu re segments, there exi stsopening displacement w along th e opene dhydraulic fra ctures, which contri butes to fractu reconductivity. All of these elastic di splacementsgive rise to changes in the rock stress or changesin the tractions acting along the fracture surfaces.The g overning eq uations for stresses, flui dpressure and displacements are given in Zhang etal. (200 7, 2 009), a nd a re given i n terms ofGreen’s fu nctions. We e mploy the di splacementdiscontinuity method (DDM) fo r the simulation ofrock deformation in a 2D h omogeneous a ndisotropic el astic mate rial, on which a uniformstress is applied at infinity.The governi ng equatio ns for fluid flow in thefracture channel are given by Zhang et al. (2009).There are t wo type of fluid migra tion alongfractures. The fluid not only can diffuse along theclosed fractu res without a ny mech anical openi ngof the fracture, but also can flow through an openchannel a s a result of openin g mode hydrauli cxFlaw hWellborefracturing processes. In the latter case, the fluidpressure is b alanced by the co mpressive stre sscaused by rock ela stic deformation , that is,n pf. We use Reynolds’ equation to describefluid movem ent insi de t he op ened hydra ulicfractures:(w ) ( w ) [ts3pf] (2)swhere w is the mechanical opening and vanishesfor th e closed s egments, is the pre-existinghydraulic aperture arising from surface roughnessand microstructures and . 12However, it must be me ntioned that effectivestress cha nges in the fluid-infiltra ted andpressurised fracture portion can produce changesof the hydra ulic ap erture without fully openin g afracture. The dilatation can s lightly affec t theinternal p ressure di stributions si nce th e re sultingfluid cond uctivity varies in location an d time. Asstated a bove, the hydr aulic aperture at thebeginning is assigned an initial value for eachclosed fra cture se gment. The evolution ofobeys a no nlinear spring model in response toany in crements of the internal pressure. Inparticular, an equation g overning the hydrauli caperture change associated with pressure changeis given as follows,d / dP f (3)where is a small constant with a value ofper Pa in the model. A pressure diffusion equationis applie d to calcul ate the fluid flow inside aclosed natural fracture (Zhang et al., 2009)pftw 0 p2 f c ( ) 0(4)ssw 010 8where c 1/(12). This cal culation applies toall clo sed p ortions of fract ures, wh ether they areundergoing or have undergone slip or not.To complete the problem formul ation, we n eedboundary and initial conditions fo r t he ab ovegoverning equatio ns. Initially, all fracturesegments are assumed to be evacu ated and therock mass is stationary. Rock failure cri terion willbe given bel ow, as well as the metho d use d forhandling fracture intersection and a ssociated fluxredistribution. For the f ractures connected to theborehole, the sum of the fluid flux entering them isequal to the injection rate, that is,N qi(0, t) Qin (5)i1where N is t he num ber o f fractures andq ithe248

Australian Geothermal Conference 2010fluid volumetric flux into fracture i.At the fracture tip, the ope ning and shearing DDsare zero, that is,w ( l) v(l) 0 (6)and the fracture opening profile near the crack tippossesses a square root shape with distance fromthe tip as p redicted by Linea r Elasti c Fra ctureMechanics (LEFM) theory.When the stress intensity factor at the f racture tipreaches the fracture toughn ess of the rock,fracture growth occurs. In general, the frac turepropagation direction may change as dictated bythe ne ar-tip stress fiel d ch aracterised by theStress Intensity Fac tors (SIFs). In part icular, themixed-mode fracture crite rion is used fo rdetermining f racture g rowth in the model. Thissame approach h as been use d by a number ofresearchers (e.g. Mogilevskaya et al., 2000).The co upled fluid flow (to obtain pre ssure) an drock d eformation (to ob tain fractu re openin g)problem is solved in an implicit m anner byincorporating the ela sticity equation into th eReynolds’ equation. Th e resp onses obt ained arehistory dependent because of the irreversible fluidflow and fri ctional sliding processes. Additionally,there are th ree m oving fronts i nherent in theproblem, namely the slip front, the fluid front an dthe fractu re opening (cra ck tip ) front. The re aderis refe rred t o Zhan g et al. (200 7, 2 009) fo r adetailed d escription of the num erical solutio nmethods applied in modelling these processes. Atthe borehole, the fracture entry pressures must beequal to the borehole pressure. T his condition i senforced through an it eration sch eme. Thefracture entry and bo rehole pressure s a reconstrained to be equal by enforcing Eq. (5). Theinjection rate into each fracture is found based onthe specified current borehole pressure, using themethod described in our previous work applied tofracture j unctions (Zhang et al., 2 007). In th ecalculation, the conve rgence of the solution i svery sen sitive to the time step and theconvergence error tol erance. T o rapi dly find aconverged wellbore pre ssure, a relaxa tion factoris used in updating the borehole pressure at eachiteration step. A tolera nce (< 10 4 ) is set for therelative erro r betwee n the wellb ore a nd fra ctureentry pressures. Therefore, in the program, a newiteration loo p is introd uced for o btaining aconverged b orehole a nd fracture e ntry pre ssuresolution.The nume rical method h as bee n verif ied, in ourprevious pa pers, by co mparison to publish edsolutions to a range of problems; see Zhang et al.(2009). However, to treat p roblems that include aborehole, th e prog ram was modifie d to allowcorrect treat ment of the interior and exterio rproblems g enerated when th e b orehole i sdiscretised usin g the DD (DisplacementDiscontinuity) method. Di scretising th e well boreusing conve ntional DD elements results in th eformation of a disc of ro ck insi de the wellbore,which can move rigidly. It must be mentioned thatthe resulting DDs along the wellbore boundary arefictitious alt hough they are ne cessary forcalculating p ressure a nd displa cement at theborehole wall and alon g the fractu res. The sefictitious DDs do n ot h ave a literal physi calinterpretation as DDs, bu t rathe r a re needed t oobtain the stress and di splacement everywhere inthe e xterior pa rt o f th e pr oblem. Mor eover, s incethe inte rior di sc is fre e to move a s a rigid b ody,the direct u se of a co nventional DDM is notpossible without sup pressing the rigid -bodymotion of this disc.A simple m ethod to suppress the se rigid -bodymotions is as follows. Two additional elements aredefined inside the borehole and assigned fictitiousDDs. Then we cal culate the displacement on thenegative sid e of an el ement ba sed on thefundamental elastic solutions. After th at the fou rdisplacements at these two additio nal element sare set to zero, in orde r to co nstruct a cl osed,well-posed problem s for all DDs. T hedisplacements on the negative si de of theelements are given as Eq. (5.6.1) in Crou ch andStarfield (19 83) fo r both normal and tange ntialdisplacement components.The ela sticity equation s for the n ormal andtangential DDs ( ukinclude s w an d v.) at themiddle of ea ch element a long th e bo rehole an dthe fractures and for the t wo additional elementsare, respectively,K u P k n,s (7)ijk k f ij ijM u 0 k n,s(8)ijkkwhere Kijkand Mijkare the coefficient matricesderived from the fundamental elastic solution as isdone i n conv entional boundary el ement method s(see detail s in Cro uch an d Starfield, 1983); an d ijare the resultant stresse s alon g the fracturesurface arising from far-field (in situ) stresses.The above first equation enforces equilibrium atDD elements along the fractures and the wellbore,and the second one enforces the con straint thateliminates rigid body motion of the interior disc byrequiring zero di splacements at two additio nalinterior ele ments. The above si multaneousequations (Eqns (7) and (8)) are solved for normaland shear DDs of all elements along the borehole,the fractu res and of the two addition al elements,once the flui d pressu re i s kno wn. T he DDs thatare obtained can then be used to calculate stressor displacement at any ot her point inside the rockmass.249

NUMERICAL RESULTSSingle hydraulic fractureWe first co mpare the o pening and path for ahydraulic fracture driven by a uniform p ressure orby injection o f a viscou s fluid. A uniform pre ssurecondition in a hydra ulic f racture i s eq uivalent tousing an i nviscid fluid (Detou rnay, 2 004). T hepressure ma gnitude is ad justed at e ach fra cturegrowth step to satis fy the fra cture extensi oncriterion at the fractu re tip. Fractu re reorientationis cle arly sh own in Fi g. 2 for both uniformlypressurised and viscous fluid flow cases. For zeroviscosity fluids as shown in Fig. 2(a), the portionadjacent to the well wall is in conta ct. In thisregion, fluid flow will be restricted and frictionalslip condition must be imp osed. For ge nerality ofresults, Mogil evskaya et al. (2000 ) sho wed tha tthe fractu re path is controlled by the followingdimensionless parameter,( Hh)R (9)Ky(m)y(m)0.40.30.20.10-0.1IC0.1 mm16.9 MPa4.22MPa-0.2-0.3 -0.2 -0.1 0 0.1 0.2 0.3x(m)0.40.30.20.10-0.10.1 mm(a)16.9 MPaTime=0.072 second4.22MPa-0.2-0.3 -0.2 -0.1 0 0.1 0.2 0.3x(m)(b)Australian Geothermal Conference 2010y(m)0.40.30.20.10-0.110 MPa16.9 MPaTime=0.072 second4.22MPa-0.2-0.3 -0.2 -0.1 0 0.1 0.2 0.3x(m)(c)Figure 2 Fracture trajectory, opening DD for uniformlypressurised fractures (a); opening DD (b) and internal fluidpressure (c) for fluid-driven fracture growth in the case of theborehole radius =0.1 m and crack starting angle of 170degree.If the fluid vi scosity is 0.01 Pa s and the injectionrate is specified as 0.0 004 m 2 /s, the frac turepathway an d asso ciated openi ng DD a ndpressure distributions are given in Figs. 2(b) and(c). The fra cture closure d oes n ot app ear in thiscase, but a high fluid pressure results as the fluidenters the n arrow fra cture an d op ens it again stthe high er normal stresses a cting alo ng its nearwell extent F ig. 2(c). Mo reover, a flui d lag zonedevelops wh en viscous dissipatio n cannot b eneglected an d is rep resented in Fig. 2(c) a s th eunpressurised region of the fra cture near the tip.This is o ne manifestation of the coupl ed flow anddeformation pro cess interacting with a movingboundary.To ch aracterize th e effect of fluid visco usdissipation on fractu re path s, we intro duceanother p arameter which provid es an appa rentfracture to ughness arising from viscou sdissipation, a s given in Jeffrey(1989), ba sed onan earlier somewhat different formulation given bySettari an d Price (1984), a nd as given byDetournay (2 004) i n the case of zero lag. Thisapparent toughness, when substituted for the rockfracture tou ghness in Eqn. 9 provide s adimensionless pa rameter, given by Eq n. 10, thatwe hypothesize will control the curvi ng of aviscosity do minated hyd raulic fra cture gro wingfrom a wellbore (Zhang et al. 2010),( Hh)RF3 1/4( QE)in(10)where E is th e plan e-strain modul us. I n thisformula, the product of injection rate and fluidQ in250

Australian Geothermal Conference 2010viscosity are parameters that can be controlledby a fracturing engineer.The above t wo parameters given by Eqs. (9 ) and(10) ca n fully cha racterise the fractu re path an drelated p ressure evolutio n for a singl e hydrauli cfracture g rowing from a wellbore. Z hang et al.(2010) provide numerical examples demonstratingthat these parameters control the curvature in thehydraulic fra cture path as it g rows from a nuncased wellbore.Two winged hydraulic fracturesfractures at angle 30° and 210°, do extend but toa shorter distance. These two fractures propagatetowards the long er f ractures afte r le aving th einfluence region around the borehole and a sharpkink is evident in Fig. 4. Th e tendency for fracturelinkage is a lso c lear. T he s lowly g rowing andpartially su ppressed fra ctures p ropagate towardsthe domin ant gro wing fractures and m ay link-upwith them a s previo usly f ound by We ng (199 3).These results sugg est that initiation of severalfractures from a wellbore subject to a larger stressratio favou rs intera ctions leadin g to link-up a sargued by Daneshy (1973).0.5 100 MPa0.450 MPaTime=0.165 second0.40.350 MPaTime=0.0846 second0.30.2100 MPa80 MPay(m)0.280 MPay(m)0.10.100-0.1-0.1-0.2-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4x(m)Figure 3 A pair of fractures growing from a wellbore: fracturetrajectory and internal fluid pressure for high-angle fractureinitiation. The fluid viscosity is 0.01 Pa s and the injectionrate is specified as 0.0004 m 2 /s.We n ext consi der t wo fractu res that gro wsimultaneously to form a single bi-win g fractu regeometry as shown in Fi g. 3. The pre ssure lo ssnear the wellbore wall is l ess evident and fracturerotation to align with the maximum principal stressoccurs in a very smooth way. This symmetri cgeometry is found in m any field obse rvations a sdescribed by Wen g (199 3) a nd Cherny et al.(2009).Multiple hydraulic fracturesThe m ore co mplicated init ial co ndition consistingof six starte r fractu res that are a rranged aroun dthe wellb ore is con sidered next. The starte rfractures h ave an un even ang ular di stribution tothe x-axis of 30°, 120°, 150°, 210°, 240° and 330°,as shown in Fig. 4. Th e fractu res at angle s of120° and 24 0° a re in a more compressive i nitialstress state, whi ch cau ses th eir gro wth to becompletely suppresse d co mpared with other fou rfractures. The remaining four starter fractures areall subjected to the same stress levels, except forthe shear stress directions. The fra cture at angl e150° is can extend more easily because it ha s aslightly long er initial fra cture l ength. As th efractures grow and turn to align with the horizontalx direction, the fractures at the a ngle of 150° and330° become more d ominant and form a bi -wingfracture configuration. Ho wever, the other two-0.2-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4x(m)Figure 4 Fracture trajectory and internal fluid pressure in thepresence of six initial flaws around the wellbore, where thefluid viscosity is 0.005 Pa s and the injection rate is specifiedas 0.0012 m 2 /s. The flaw alignment angles are 30°, 120°,150°, 210°, 240° and 330°.Effects of natural fracturesIf there i s a natu ral f racture in th e fra cturepropagation pathway, fluid will e nter it after ahydraulic fra cture interse cts it. Th e e xistence ofnatural fract ure will affect the growth of thehydraulic fra cture an d the inflation of the naturalfracture caused by the fluid pre ssure can indu ceslip alo ng t he natu ral fracture an d fractu rereinitiation from the end s of a finite-size natu ralfracture. The example fra cture geometry is give nin Fig. 5(a).y(m)0.30.20.10-0.180 MPa100 MPa50 MPaTime=0.16069 second-0.2 -0.1 0 0.1 0.2 0.3 0.4x(m)(a)251

Australian Geothermal Conference 2010y(m)0.30.20.10-0.180 MPa100 MPa50 MPaTime=0.14715 second-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3x(m)(b)Figure 5 Fracture trajectory and internal fluid pressure in thepresence of a natural fracture that has finite length and aninclination angle of 135 degree with respect to the x-axis.The coefficient of friction of the natural fracture is 0.3. Thefluid viscosity is 0.01 Pa s and the injection rate is specifiedas 0.0004 m 2 /s.The a rrest of one fra cture will faci litate thepropagation of a second win g of a t wo-wingedfracture, as shown in Fig. 5(b). The left-hand sidefracture p ropagates at a lower p ressure, takin gmost of injected fluid and growing much faster.ReferencesAbass, H. H., Soliman, M. Y., Tahi ni, A. M.,Surjaatmadja, J., Meadows, D. L., Sierra, L.,2009, Ori ented fractu ring: A new tech nology tohydraulically fractu re op enhole ho rizontal well ,SPE 124483, Proceedi ngs of the 2009 SPEAnnual Technical Conference and Exhibition, NewOrleans, Louisiana, October 4-7 2009Cherny, S., Chirkov, D., Lapin, V., Muran ov, A.,Bannkov, D., Miller, M., Willberg, D., Medvedev ,O. and Aleksee nko, O., 2009, Two-di mensionalmodelling of the near-well bore fracture tortuosityeffect, Int. J. Rock Mech. & Mining Sci. 46, p. 992-1000.Crouch, S. L. and Starfield, A. M., 1990, BoundaryElement Me thod in Sol id Me chanics, Un winHyman, BostonDaneshy, A. A., 1973, A stu dy of incli nedhydraulic fractures, SPE J 13: p. 61-68.Detournay, E., 2004, Prop agation regime of flui d-driven fractu res in imp ermeable ro cks, Int. J.Geomechanics 4, p. 35-45.Detournay, E. and Jeffrey, R., 19 86, Stre ssconditions for initiation o f seconda ry fractu resfrom a fract ured b orehole, proceedi ngs of theinternational symposium on ro ck stress and rockstress me asurements, St ockholm, S weden, 1 -3,September, 1986.Hamison, B. and Fai rhurst C., 1970 In-situ stressdetermination at gre at depth by means ofhydraulic fra cturing, in Ro ck M echanics—Theoryand Pra ctice, (ed.) W. H. Somerton, Am. Inst.Mining Engrg., 559–584, 1970Jeffrey, R. G., 1989 The combi ned effect of fluidlag and f racture tou ghness on hydraulic fra cturepropagation, SPE 18957, Pr oceedings of SPEJoint Rocky Mountain Regional/ Low PermeabilityReservoirs S ymposium a nd Exhibition , Denve r,Colorado, March 6-8, 1989.Manrique, J. F. and Ven kitaraman, A., 2001,Oriented fra cturing – A pra ctical te chnique forproduction optimization, Proceedings of the 2001SPE Annual Technical Conference and Exhibitionheld in New Orleans, Louisiana, 30 Se ptember–3October 2001Mogilevskaya, S. G., Rothe nburg, L. an dDusseault, M. B., 2000, Growth of pressu reinducedfra ctures in the vicinity of a wellbo re,International Journal of Fracture 104, p. L25-L30.Settari, A. and H. S. Pri ce., 19 84, Si mulation ofhydraulic fracturing in low-permeability reservoirs,SPEJ April 1984, p. 141-152.Soliman, M. Y., East, L. and Ad ams, D., 2008,Geomechanics aspects of multiple fracturi ng ofhorizontal and vertical wells, SPE 86889, SPEDrilling and Completion, Sept. 2008, 217-228.Weng, X., 1993, Fracture Initiation andPropagation from Deviated Well bores, SPE26597, pre sented at the 1993 AT CE, Hou ston,TX, 3-6 October, 1993.Zhang, X. Jeffrey, R. G. an d Thiercelin, M., 2007,Deflection of fluid-driven fractures a t beddin ginterfaces: A numerical investigation, J. StructuralGeology 29, p. 396-410.Zhang, X. Jeffrey, R. G. an d Thiercelin, M., 2009,Mechanics of fluid-drive n fra cture growth i nnaturally fractured reservoirs with si mple networkgeometries, J. Ge ophysical Research -S olidEarth, 114, B12406.Zhang, X. Jeffrey, R.G ., Bunger, A.P. andThiercelin, M., 2010 Initia tion and growth of ahydraulic fracture from a boreho le unde rtoughness- or visco sity- domin ated conditions,44 th US Ro ck Me chanics Symposiu m and 5 thU.S.-Canada R ock Me chanics Symp osium, Sa ltLake City, UT June 27–30, 2010.252

Australian Geothermal Conference 2010Economics of geothermal feedwater heating for steam Rankine cyclesDavid L. Battye * , Peter J. Ashman, Gus J. NathanCentre for Energy TechnologyUniversity of Adelaide, SA 5005.* Corresponding author: david.battye@adelaide.edu.auWe have investigated the economics ofgeothermal feedwater heating (GFWH) in a steamRankine cycle with base capacity of 583 MW(gross). The capital cost per kW geothermalpower (before geofluid pumping) is compared withthose of stand-alone geothermal plants of equalpower. The economic benefits of GFWH overstand-alone plants for enhanced geothermalsystems are, in order of importance:• 67% less operating and maintenance cost• 28% less capital cost• 30-35% less geofluid pumping loadsThe reduction in levelised geothermal energy costis estimated to be 48%.Keywords: geothermal power, cost, economics,hybrid.IntroductionIn steam Rankine cycles, boiler feedwater mustbe heated from a low temperature at thecondenser outlet (typically ~40 C) totemperatures of typically 250 C before enteringthe boiler. In “regenerative” steam cycles, this isaccomplished in a series of feedwater heaters,which use steam extracted from various stages inthe turbine. One feedwater heater is a deaerator,in which extracted steam and feedwater are indirect contact, liberating oxygen and other gasesin the feedwater for venting to atmosphere. Theother feedwater heaters are shell-and-tube heatexchangers where feedwater flows in the tubesand steam condenses on the outside.If feedwater heating can be accomplished usinggeothermal heat, the steam normally extracted forfeedwater heating can instead generate power inthe turbine. The extra power generated bygeothermal feedwater heating (GFWH) offers thefollowing potential benefits:• Existing plants may be retrofitted toaccommodate a proportion of geothermalpower• Geothermal power may be generated morecheaply than in a stand-alone plant• Steam cycles may be adapted for GFWH withno novel technology required.GFWH is suitable for coal, gas, solar-thermal andnuclear-fuelled boiler plants, but not combinedcyclegas turbine plants. The latter plants use hotexhaust gas for feedwater heating in the Rankinecycle section.Interest in geothermal feedwater heating hasexisted since the 1970s, when it was investigatedfor application at sites in California and Utah(Parsons, 1978). It has been recently studied byBruhn (2002), Buchta (2009) and Borsukiewicz-Gozdur (2010). All studies agree that geothermalpower generation by GFWH is significantly moreefficient thermodynamically than in conventionalstand-alone plants. However, to the authors'knowledge, no previous comparison has beenreported of the economic benefits of GFWH overstand alone geothermal power generation. Weseek to provide such a comparison, specificallywith respect to power generation from enhancedgeothermal systems.Condensate tocondenserFWH 1FWH 2FWH 3FWH 4FWH 5(deaerator)FWH 6FWH 7Feedwater from condenser49 C, 386 kg/sSteam72 C, 12 kg/s67 C88 CSteam132 C, 23 kg/s125 CSteam147 C, 10 kg/s140 CSteam186 C, 29 kg/s188 CSteam222 C, 22 kg/s218 CFeedwater to boiler251 C, 470 kg/sSteam94 C, 13 kg/sSteam255 C, 33 kg/sFigure 1: Feedwater heating section in a conventional 583MW (gross) steam Rankine cycle. Steam flows extractedfrom turbine stages are shown at right. These steam flowsmay be replaced by a single flow of hot water from ageothermal reservoir.253

Australian Geothermal Conference 2010MethodologyFor our economic analysis we consider the casethat GFWH is being used to provide extracapacity for a new Rankine steam cycle plant.This scenario is in contrast to the retrofit of anexisting plant, where the capacity is typically fixed.The basis for analysis is a 583 MW (gross)subcritical steam cycle with single reheat, forwhich design specifications and cost estimateshave been reported (NETL, 2007). A processflow diagram of the feedwater heating section isshown in Figure 1. In the proposed GFWHscenario, feedwater is heated by geofluid insteadof steam extracted from turbine stages. Theotherwise extracted steam flows are thenavailable for power generation in the turbine.The temperature to which the feedwater is heatedby the geofluid, T GFWH, is a variable in theanalysis. Above this temperature, we assumethat feedwater heating is achieved byconventional heaters as per normal. However,the deaerator must be retained in the cycle, andmay have to operate at higher temperatures thannormal. We assume that the upper limit of GFWHinput is one where feedwater is heated to 220 Cby geofluid, and a deaerator heats feedwater from220 C to 251 C.To estimate the increased costs of equipment inGFWH cases, the reported costs of componentsare increased in proportion to turbine exhauststeam flow (turbine, condenser, steam piping, andcooling system) or gross power (generator,accessory electrical plant, instrumentation andcontrol).The cost of feedwater heaters (apart from thedeaerator) is proportional to the heat transfersurface area, which varies with the overall heattransfer coefficient, U, and mean temperaturedifference between geofluid and feedwaterstreams, MTD. We estimate that U is decreasedby ~30% in GFWH, since the high film coefficientsassociated with condensing steam are absent.MTD is about 15 K on average in conventionalfeedwater heaters. In GFWH cases MTD is avariable. The resulting multiplier applied to theconventional feedwater heater train cost is: TGFWH 38 21 f 1 1 213 MTD Geofluid costs are estimated assuming:• Flow of 80 kg/s per well drilled• Drilling costs vary exponentially with depthaccording to:C drill 0.5e0.0001491d$M, 2000 US• Other costs (stimulation, geofluid pumps andpiping) are 20 % of drilling costs.Geofluid costs (2009 US dollars) versus geofluidtemperature are shown in Figure 2 for thermalgradients of 40 and 50 K/km.Geofluid cost, $K per kg/s2502001501005050 K/km40 K/km050 100 150 200 250 300Geofluid temperature, CFigure 2: Geofluid cost versus geofluid temperature forthermal gradients of 40 and 50 K/km.Estimates of the cost and performance of standalonegeothermal power plants are based on datain the Next Generation Geothermal Power Plantsreport (Brugman et al., 1995 and CE Holt, 2006)and documentation for the Geothermal ElectricTechnologies Evaluation Model (GETEM)(Entingh, 2006).All plants in the analysis are water-cooled and usea cooling tower.The costs determined from the literature havebeen updated to 2009 US dollars using theChemical Engineering Plant Cost Index.Results and discussionFigure 3 shows percentage extra gross power andexhaust steam flow versus T GFWH. The effectivegeothermal power output (before geofluidpumping) ranges from 9.6 MW at 100 C to 61MW at 220 C.Figure 4 shows the consumption of geofluid perunit geothermal energy generated (net beforegeofluid pumping) versus geofluid temperature.The GFWH plants require significantly lessgeofluid than the stand-alone plants, typically 30-35% less above 200 C. The explanation for thislies in the fact that the GFWH plants minimisethermodynamic inefficiency associated with heattransfer from the geofluid. The lower the MTD,the greater the thermodynamic benefit, but at theexpense of higher heat exchanger costs.(Entingh, 2006), where d is depth in feet.254

Australian Geothermal Conference 2010% extra power or steam flow35302520151050Exhaust steam flowGross power50 100 150 200 250T GFWH , CPlant cost, $/kW4000300020001000ORCDouble flashGFWH, MTD = 30 KGFWH, MTD = 20 K0100 150 200 250Geofluid temperature, CFigure 3: Percentage extra gross power and turbine exhauststeam flow versus geothermal feedwater heatingtemperature.0.1Figure 5: Plant cost versus geofluid temperature for equalpower outputs. GFWH plant costs correspond to theincrement on base plant cost.6000Geofluid consumption, kg/kJ0.080.060.040.020ORCDouble flashGFWH, MTD = 30 KGFWH, MTD = 20 KCapital cost, $/kW500040003000200010000ORCDouble flashGFWH100 150 200 250Geofluid temperature, C100 150 200 250Geofluid temperature, CFigure 4: Geofluid consumption per unit geothermal energygenerated (net before geofluid pumping) versus geofluidtemperature.Figure 5 compares plant costs for GFWH andstand-alone plants, ie. Organic Rankine Cycles(ORC) and double flash steam plants. Increasinggeofluid temperature reduces the cost of all plantsdue to improving thermal efficiency. However, thestand-alone plant cost also decreases due toeconomies of scale. GFWH plant costscorrespond to the increment on base plant cost.Since the base plant capacity is already manytimes greater than for the stand-alone plants,GFWH plant costs are much lower on averageover the geofluid temperature range. At the uppergeofluid temperatures in Figure 5, the GFWHplant costs are 15-25 % lower than for a doubleflash plant.Figure 6: Capital cost versus geofluid temperature forthermal gradient of 50 K/km.Capital cost, $/kW60005000400030002000ORC1000 Double flashGFWH0100 150 200 250Geofluid temperature, CFigure 7: Capital cost versus geofluid temperature forthermal gradient of 40 K/km.Total capital costs are shown in Figures 6 and 7for thermal gradients of 50 K/m and 40 K/mrespectively. Note that the $/kW value is forgeothermal power before geofluid pumping. The255

value of MTD is optimised in each case: 30 K for50 K/km and 20 K for 40 K/km. The following isobserved:• Optimal geofluid temperatures are 220 C orgreater.• Optimal GFWH plants reduce total cost by28% relative to optimal stand-alone plants inboth cases.• GFWH with a 40 K/km thermal gradient cangenerate power at the same cost as a standaloneplant with a 50 K/km gradient.• GFWH can generate power at the same costas optimal stand-alone plants, but at geofluidtemperatures of only 120-150 C as opposedto 220 C.Relative to stand-alone plants, GFWH has theadditional benefit of incurring less geofluidpumping power loads, which are not accountedfor in the preceding analysis. Since GFWHconsumes 30-35 % less geofluid above 200 C,pumping loads are reduced by the same amountrelative to stand-alone plants.Another economic benefit is that geothermaloperating and maintenance costs are drasticallyreduced by GFWH. For example, O&M costs in a61 MW stand-alone geothermal plant areestimated at 7.3 $M/year (Sanyal, 2004) whereas61 MW generated by GFWH only incurs 10 % ofthe annual O&M costs of the whole plant, ie. 2.4$M/year (NETL, 2007), a saving of 67%.A simple estimate of the reduction in the levelisedcost of geothermal energy can be made usingfigures reported by Sanyal (2007) for EGS power.Table 1 illustrates how the estimate is made andsuggests a reduction in total levelised costapproximating 48%.Table 1: Estimate of reduction in levelised cost of geothermalenergy by GFHW, based on estimates of Sanyal (2007) andpreceding analysis.Levelised costfor standaloneplant,¢/kWh(US, 2007)Reductionby GFWHGFWHlevelisedcost,¢/kWh(US, 2007)O&M 2.75 1.84 (67%) 0.91Capital andcost ofmoney2.68 0.75 (28%) 1.93Total 5.43 2.59 (48%) 2.84Australian Geothermal Conference 2010Concluding remarksThe economic analysis has shown geothermalfeedwater heating in a purpose-built steamRankine cycle to have economic merit. Giventhat good solar and EGS resources in Australiaare generally co-located, solar-geothermal hybridsusing GFWH are possible. Further work willinvestigate the economics of retrofitting existingsteam Rankine cycle plants for GFWH.ReferencesBorsukiewicz-Gozdur, A., 2010, Dual-fluid-hybridpower plant co-powered by low-temperaturegeothermal water, Geothermics, 39, 170-176.Brugman, J., Hattar, M., Nichols, K. and Esaki, Y.,1995, Next Generation Geothermal Power Plants,Electric Power Research Institute (EPRI) ReportNo. EPRI/RP 3657-01.Bruhn, M., 2002, Hybrid geothermal–fossilelectricity generation from low enthalpygeothermal resources: geothermal feedwaterpreheating in conventional power plants, Energy,27, 329–346Buchta, J., 2009, Green power from conventionalsteam power plant combined with geothermalwell, Proceedings of the IEEE InternationalConference on Industrial Technology , Article No.4939553.CE Holt Consulting Firm, 2006, NGGPP processdata for binary cycle plants, Document for U.S.Dept. of Energy.Entingh, D.J., 2006, US Dept. of Energy (DOE)Geothermal Electricity Technology EvaluationModel (GETEM): Volume I – Technical ReferenceManual.National Energy Technology Laboratory (NETL),2007, Cost and Performance Baseline for FossilEnergy Plants, Volume 1: Bituminous Coal andNatural Gas to Electricity Final Report, ReportNo. DOE/NETL-2007/1281.Parsons, R.M., 1978, System design verificationof a hybrid geothermal/coal fired power plant, Cityof Burbank and Ralph M. Parsons Company; USDept. of Energy (DOE), Report No. SAN-1572-T2.Sanyal, S.K., 2004, Cost of geothermal power andfactors that affect it, Proceedings of the Twenty-Ninth Workshop on Geothermal ReservoirEngineering, Stanford University, Stanford,California, Jan. 26-28, SGP-TR-175.Sanyal, S.K., 2007, Cost of electricity fromenhanced geothermal systems, Proceedings ofthe Thrity-Second Workshop on GeothermalReservoir Engineering, Stanford University,Stanford, California, Jan. 22-24, SGP-TR-175.256

Australian Geothermal Conference 2010A Protocol for Estimating and Mapping Global EGS PotentialGraeme Beardsmore 1* , Ladislaus Rybach 2 , David Blackwell 3 and Charles Baron 41. Hot Dry Rocks PL, PO Box 251, South Yarra VIC 3141, Australia2. Geowatt AG, Dohlenweg 28, CH-8050 Zürich, Switzerland3. Southern Methodist University, Dallas, TX 75275, USA4. Google.org, 1600 Amphitheatre Parkway, Mountain View, CA 94043, USA* Corresponding author: graeme.beardsmore@hotdryrocks.comAbstractWe present a Protocol to estimate and map theTheoretical and Technical potential forEngineered Geothermal Systems (EGS) in aglobally self-consistent manner compatible withcurrent public geothermal Reporting Codes. Thegoal of the Protocol is to standardise theproduction of regional estimates and maps ofEGS potential so that they are directlycomparable to one another globally.The Protocol is divided into five stages:1. Model the temperature, heat flow and availableheat of the Earth’s crust to a depth of 10,000 m2. Estimate the Theoretical Potential for EGSpower in the crust to a depth of 10,000 m3. Estimate the Technical Potential that can berealized with current technology, and consideringgeographic, ecologic, legal and regulatoryrestrictions4. Define a level of confidence in the estimatedTechnical Potential at each location, consistentwith public Reporting Codes5. Present results using KML visualization anddata architectureThe maps, estimates and source dataunderpinning the estimates and maps will bemade freely available for public use andpresented in the Keyhole Markup Language(KML) for Google Earth.Keywords:Engineered Geothermal Systems; EGS resourceestimation; Global geothermal resource inventory;Google Earth; Keyhole Markup Language; KMLIntroductionEngineered Geothermal Systems‘Engineered Geothermal Systems (EGS)’ is ageneric term for the process whereby heat isextracted from the Earth’s crust by circulatingwater through an artificially engineered set ofpermeable fractures in hot rocks (Figure 1).Although significant engineering and financialhurdles remain, EGS plants hold the promise ofnearly ubiquitous, low to zero CO 2 emission,secure, base-load power for millennia to come. Intheory, EGS plants may be constructed anywherethat the mechanical limits of drilling and fractureengineering allow. Furthermore, geothermalsystems have the second lowest land footprint ofall electrical generating technologies (McDonaldet al., 2009). These attributes make EGS anattractive potential major contributor to worldenergy supplies.Figure 1. Conceptual EGS power plant design (from MIT,2006)For EGS to play a material role in the globalenergy mix, improving public awareness anddispersing knowledge of the global potential andits regional distribution is a vital precursor toinformed R&D, energy policy making, and broadscalecommercial deployment.A Protocol for estimating and mapping globalEGS potentialThis paper summarises a Protocol to estimateand map the Theoretical Potential and TechnicalPotential (as defined by Rybach, 2010) for EGS ina globally self-consistent manner. Any estimate ormap of EGS potential in a region involves anumber of inputs about geology, thermalproperties, recovery factors, power conversionefficiencies, ambient temperatures and so on. Itfollows that an inventory of the global EGSpotential requires a globally consistentmethodology and a globally consistent set ofassumptions to fall back on when real data arenot available.The Protocol does not seek to provide a uniquepicture of the magnitude and distribution of theworld’s EGS potential. Alternative approaches toestimating EGS potential may be more relevant inparticular locations and more robust analyses will257

Australian Geothermal Conference 2010certainly be required to assess the commercialviability of EGS at specific sites. The Protocol will,however, provide consistent methodologies andassumptions that will ultimately allow a selfconsistentinventory and map of EGS potentialaround the world.The Protocol will provide utility for academia,policy makers and commercial entities bystandardizing technical language, improvingunderstanding of EGS generation potential,providing a consistent visualization platform, andfacilitating international commercialization efforts.The basis for the ProtocolThe Protocol closely follows the methodsunderpinning a report by the MassachusettsInstitute of Technology, which concluded in 2006that EGS could provide 100,000 MW of electricalgenerating capacity to the United States by 2050(MIT, 2006). An integral component of that studywas a review of the heat resource within the top10,000 m of the crust by Professor DavidBlackwell and his team at Southern MethodistUniversity (SMU; Blackwell et al., 2007). SMUassumed that conduction is the primary heattransfer mechanism in the crust, and that theupper crust can be broadly divided into sections of‘sediment’ and ‘basement’, each with its ownphysical properties of thermal conductivity andinternal heat generation.In 2008, the SMU team and Google.org convertedthe MIT findings into KML format for visualizationon the Google Earth platform (Figure 2). Thelayers are available for free download and viewingfrom www.google.org/egs/.Figure 2. Screen capture of a Google Earth layer depictingthe predicted temperature at 5.5 km and the EGS Potentialbase for Nevada, USA.Theoretical and Technical PotentialThe Protocol calls for an initial estimate of the‘Theoretical Potential’ for EGS across a region,following the terminology of Rybach (2010).Theoretical Potential is “defined solely by thephysical limits of use and thus marks the upperlimit of the theoretically realizable energy supplycontribution.” From the Theoretical Potential, theProtocol provides guidelines to estimate the‘Technical Potential’, or “the fraction of thetheoretical potential that can be used under theexisting technical restrictions…structural andecologic restrictions as well as legal andregulatory allowances” (Rybach, 2010). Theserestrictions vary greatly with geology, location andtime, providing some flexibility to modify TechnicalPotential based on current local conditions.Geothermal Resource reporting codesCodes for the reporting of Geothermal Resourceestimates exist for Australia and Canada. TheProtocol aims to conform to those reportingCodes in so far as respecting their underlyingprinciples of ‘transparency’, ‘materiality’ and‘competence’. These principles will be honouredthrough the inclusion of all relevant information(generally as metadata) with each set of mapsand tables produced, and by including thepersonal endorsement of one or more‘Competent’ or ‘Qualified’ Persons. The Protocolproposes the following minimum level ofinformation to comply with these principles:1. A statement that the data should not be reliedon to inform commercial investment decisions2. Sources of all data utilized for the estimates ofEGS potential3. A brief description of the modelling technique4. Assumed ambient temperatures, recoveryfactors, and conversion efficiencies5. Assumed lifespan of power generation6. Statement of relative accuracy / confidence7. The name(s) of the Competent or QualifiedPerson(s) who accept(s) responsibility for theResource estimate.MethodologyThe Protocol assumes that pure verticalconduction dominates heat transport through thecrust, and that a simple two-layer geologicalmodel (‘sediment’ on ‘basement’) approximatesthe top 10,000 m of crust in all continental areas.A region is divided into a grid-work of ‘cells’, andthe simple two-layer model is used to estimate thelocal thermal structure in each cell. The EGSTheoretical Potential (relative to a defined ‘basetemperature’) is then tallied over differentdepth/volume intervals by assuming density andspecific heat values for the rocks in question, andassuming a uniform heat–electricity conversionpathway. The discrete estimates of each cell maybe summed to estimate the total EGS potentialover the region or depth interval.In practice, the process is divided into five stages:1. Model temperature, heat flow and heat in theEarth’s crust down to a depth of 10,000 m258

Australian Geothermal Conference 20102. Estimate the Theoretical Potential of EGSpower in the crust down to a depth of 10,000 m3. Estimate the Technical Potential given currenttechnology, geographic, ecologic, legal andregulatory restrictions4. Define a level of confidence in the estimatedTechnical Potential at each location, consistentwith public Reporting Codes5. Present results using common visualization anddata architectureModel temperature, heat flow and heat in theEarth’s crust down to a depth of 10,000 mThe temperature in the crust can be estimatedusing a ‘top down’ approach, where surface heatflow (Q 0 ) is assumed to extend downwards,gradually decreasing with increasing depth due tothe distribution of heat generation in the rocks.Average thermal gradient can be estimated overany depth interval from the heat flow and thermalproperties of the rocks. The Protocol recommendsestimating the temperature profile through the top10,000 m of crust using a ‘top down’ approach.Figure 3 provides a flow chart for a process thatcan be applied for any location.The next step is to chart the average thickness of‘sediment’ overlying ‘basement’ in each 5’ x 5’cell—in effect, develop a ‘depth to basement’ mapfor the region of interest. For Australia, theSEEBASE TM database provides a first passestimate of the thickness of Phanerozoic basinsacross the continent (Figure 4), and can be freelydownloaded over the Internet. SEEBASE is aregistered trademark of FrOG Tech Pty Limited inAustralia, and stands for Structurally Enhancedview of Economic BASEment.Figure 4. A visualization of the SEEBASE TM database.Figure 3. General process for estimating the temperatureprofile of the crust to 10,000 m depth. See Glossary forsymbol definitions.The region under investigation is first divided intoa regular grid. 5’ x 5’ graticules are recommendedas the basic ‘cell’ size. A region such as Australia(7.6 million square kilometres at an averagelatitude of around 25°S) requires ~100,000 cells.Temperature can be predicted at any arbitrarydepth for a given surface heat flow (Q 0 ), thermalconductivity and heat generation (A) structure.The temperature prediction process requires thatthe sediment and basement sections of each cellbe individually characterised with values ofthermal conductivity (K S and K B for sediment andbasement, respectively) and heat generation (A Sand A B , respectively). To maintain consistencywith Blackwell et al. (2007), the Protocol assumesthat the thermal conductivity of sediment deeperthan 4,000 m is the same as the basement (K B ).Mean surface temperature (T 0 ) is an importantboundary condition for models of undergroundtemperature and for estimates of EGS potential.The Protocol assumes that mean surface rocktemperature is approximately equal to meansurface air temperature.Temperatures at depth are estimated in twosteps. The first step is to estimate the temperatureat the sediment–basement interface (T S ). T S isderived using one or both of the followingformulae, depending on whether sedimentthickness (S) is greater than or less than 4,000 m.If S < 4,000 m:T S = T 0 + [(Q 0 x S) / K S ] - A S x [S 2 / (2 x K S )]Eq 1259

Australian Geothermal Conference 2010If S > 4,000 m, the conductivity of that portion ofsediment deeper than 4,000 m is K B . In this case,first calculate T 4km using S = 4000 in Eq 1, then:T S = T 4km + [(Q 0 - 4000.A S ) x (S - 4000) / K B ] - A S x[(S - 4000) 2 / (2 x K B )] Eq 2Heat flow at the sediment–basement interface(Q S ) becomes the ‘surface heat flow’ forestimation of temperature at deeper levels. Q S isderived by subtracting the total contribution ofsedimentary heat from Q 0 :Q S = Q 0 - S x A S Eq 3The second step is to estimate temperature atdepth (X) in the basement (T X ):T X = T S + [(Q S x (X-S)) / K B ] - A B x [(X-S) 2 / (2xK B )]Eq 4At the completion of this step, a mean predictedtemperature profile to 10,000 m depth should beavailable for each 5’ x 5’ cell.Estimate the Theoretical Potential of EGS powerin the crust down to a depth of 10,000 mThe heat stored within a volume of rock isproportional to the temperature, heat capacity,density and volume of the rock. In addition, it canonly be estimated relative to a ‘base temperature’.Estimates of EGS potential, therefore, requirevalues for each of these parameters. Figure 5provides a flow chart of the recommended fivestepprocess for estimating the TheoreticalPotential for EGS in the top 10,000 m of crust inany location.Figure 5. General process for estimating stored heat energyand theoretical power generation potential. See Glossary forsymbol definitions.Crustal temperature is the key determinant ofTheoretical Potential for EGS at any specificlocation. Equation 4 provides this value for anyspecific depth. To relate temperature to heatcontent in a particular volume of crust, we need toassign a density (ρ) and specific heat (C p ) valuefor the volume of interest.Each depth interval beneath a surface cell willcontain a different amount of thermal energy. Thetotal available heat in exajoules (H) in a volume ofcrust (V c ) is a function of the temperature, density,specific heat, and a ‘base temperature’ (T r ):H = ρ x C p x V c x (T X - T r ) x 10 -18 Eq 5The base temperature is the temperature to whichthe crust can theoretically be reduced throughutilization of geothermal heat. The Protocolproposes following the lead of the USGS in arecent assessment of geothermal potential in theUSA (Williams et al., 2008), in which the USGSassumed approximately T r = T 0 + 80°C.Again following the lead of the USGS (Williams etal., 2008), Theoretical Potential power generationis derived using the following assumptions:1. All heat (H) above the base temperature istheoretically recoverable in all locations2. 30 years life span of power generation3. Cycle thermal efficiency, η th , is a function ofresource temperature as per MIT (2006):η th = 0.00052 x T + 0.032 Eq 6Note that the temperature appropriate for Eq 6 isthe average of the initial rock temperature and thebase temperature:T = (T X + T r ) / 2 Eq 7The potential power generation, P (MW e ), from avolume of rock with available heat, H, is:P = H x 10 12 x η th / 9.46x10 8 Eq 8Theoretical Potential power generation can becollated and tabulated for specific depth andtemperature intervals.Estimate the Technical Potential given currenttechnology, geographic, ecologic, legal andregulatory restrictionsIt is impossible to realize the entire TheoreticalPotential for EGS power in any location. Followingthe terminology of Rybach (2010), the ‘TechnicalPotential’ is that part of the Theoretical Potentialthat can be extracted after consideration ofcurrently ‘insurmountable’ technical limitations.‘Technical’ is defined in its broadest sense,including factors such as land access, rock type,drilling technology, fracture density, stressorientation, regulatory framework, powerconversion technology and availability of water.Rybach (2010) argues that “the EGS potentialcannot yet be termed ‘technical’”, but the Protocolproposes a set of assumptions for deriving anestimate of Technical Potential. The steps areillustrated in Figure 6.National parks, conservation areas, denselypopulated areas, mountains, large lakes andswamps, militarized zones, deserts with noavailable water resources, and other areas maybe excluded from EGS development. Theproportion of each cell that is accessible andavailable for EGS (R av ) is a value between 0–1.260

Australian Geothermal Conference 2010Figure 6. General process for estimating Technical Potentialof EGS from the Theoretical Potential. See Glossary forsymbol definitions.The Protocol recommends limiting estimates ofTechnical Potential for EGS to the top 6,500 m.This may change if there are significant advancesin hard-rock drilling technology.Recoverability factor (R) is the proportion of heatthat can ultimately be recovered from a volume ofrock. The Protocol suggests estimates of potentialbe based on a range of R values representing theexpected minimum, maximum and mean values.This Protocol proposes 0.02 as the minimum R,following the precedent of MIT (2006), and 0.14and 0.20 as the mean and maximum R,respectively, following the findings of Williams etal. (2008). These values are based on the resultsof numerical modelling. While they fulfil the aim ofthe Protocol to provide a globally consistent set ofassumptions, estimates of Technical Potentialusing this Protocol should only be viewed aspreliminary until such time as practical experienceprovides real data on recoverability.There is a practical limit to the temperaturedrawdown a power plant can withstand before itwill no longer operate effectively. This Protocolrecommends following the methodology of MIT(2006) by assuming a maximum allowabletemperature drawdown of 10°C. This effectivelyintroduces a ‘temperature drawdown’recoverability factor (R TD ) defined by:R TD = 10 / (T X - T r ) Eq 9The Technical Potential (P T ) is that part of theTheoretical Potential accessible from the surface,shallower than 6,500 m, accessible via fracturenetworks, and available with

Australian Geothermal Conference 2010Thirdly, the Protocol extends the methodologydescribed by MIT (2006) and Blackwell et al.(2007) to apply in areas where real data arescarce or non-existent.Fourthly, the Protocol recommends assessingEGS potential relative to a base temperature of T 0+ 80°C, rather than relative to T 0 .ConclusionsEstimates of EGS potential derived using theProtocol will not be ‘final’. They will continue to berefined as more relevant data become available.Theoretical Potential will be refined as newgeological and geophysical data improve ourunderstanding of the thermal structure of thecrust. Refinements here are expected to begradual. Technical Potential will be refined astechnological advancements in drilling, powerconversion and legal regimes allow greateramounts of the Theoretical Potential to berealized. Changes here are expected to besudden and dramatic.Application of the Protocol will undoubtedly revealgaps and uncertainties that will require theProtocol itself to be refined through time. TheProtocol will, therefore, be a ‘living document’.The authors hope that the EGS potential of mostof the world’s continental surface will eventuallybe assessed and charted following the guidelinesof the Protocol, allowing for the first time acoherent view of the size and distribution of the‘hidden’ energy stored in the rocks of the top10,000 m of the Earth’s crust.AcknowledgementsThe development of the Protocol was madepossible by the financial support of Google.orgthrough its ‘Renewable Energy Cheaper thanCoal’ initiative (RE

Australian Geothermal Conference 2010Changes to Geothermal Reporting Code andGuidelines on Company ReportingBeardsmore, G. 1 , Budd, A. 2 , Goldstein, B. 3 , Holgate, F. 4 , Jeffress, G. 5 , Larking, A. 6 , Lawless, J. 7 , Libby J. 8 , Middleton,M. 9 , Reid, P. 10*, Stafford, C. 11 , Ward, M. 12 , Williams, A. 131Hot Dry Rocks Pty Ltd, PO Box 251, South Yarra, VIC 3141; 2 Geoscience Australia, GPO Box 378, CanberraACT 2601; 3 Primary Industries and Resources, PO Box 277, Adelaide SA; 4 KUTh Energy Ltd, 35 Smith Street,North Hobart, TAS 7000; 5 CSA Global, PO Box 139, Burswood, WA 6100; 6 Green Rock Energy, PO Box1177,WestPerth, WA 6872; 7 Sinclair Knight Merz, PO Box 9806, Newmarket , Auckland NZ; 8 New World Energy, Level 3, 46Ord St, West Perth, WA 6005; 9 WA Department of Mines and Petroleum, 100 Plain Street, East Perth, WA 6004; 10Petratherm Limited, 129 Greenhill Rd, Unley 5061, South Australia; 11 Granite Power Limited, Level 6, 9 Barrack St,Sydney NSW 2000; 12 Geothermal Advisory Pty Ltd, PO Box 150, Richmond Tas 7025; 13 Geodynamics Limited, POBox 2046, Milton, Qld 4064* Corresponding Author: preid@petratherm.com.auWith the launch of the First Edition of theGeothermal Reporting Code in August 2008,Members of the Australian Geothermal EnergyAssociation (AGEA) agreed to report theirExploration Results, Geothermal Resources andGeothermal Reserves in accordance with theGeothermal Reporting Code. In its first year ofoperation, a number of issues became apparentin the use and interpretation of the GeothermalReporting Code. Most significant amongst thesewere: That the definition of an InferredGeothermal Resource was allowing largeestimates of stored heat to be reported.Whilst derived by legitimatemethodologies, it was perceived thatthese large figures may not be properlyunderstood by non-technical people andmay result in a diminution of credibility ofthe industry;Person sign-off on public reports andother company announcements.The Geothermal Reporting Code Committeeconsulted with industry and received feedbackfrom companies and practitioners both in Australiaand overseas. The Second Edition of theGeothermal Reporting Code addresses the issuesidentified above. In some cases definitions andmeanings have been changed (for example, there-defining of Geothermal Resources asrecoverable energy, rather than energy in place)whilst in others, aspects of the GeothermalReporting Code that were already in place havebeen made more explicit (for example, theobligations in respect of Competent Person signoffs).This presentation outlines the key changesto the Second Edition of the AustralianGeothermal Reporting Code, and providesclarification on the public reporting requirementsof Companies.The units of energy described in the FirstEdition were not always consistent orpractical; andCompanies were misunderstanding thecircumstances requiring a CompetentKeywords: Geothermal Energy, ResourceEstimation, AustraliaReferencesThe Australian Geothermal Code Committee,2008. “Australian Code for Reporting ofExploration Results, Geothermal Resources andGeothermal Reserves; The Geothermal Reporting263

Australian Geothermal Conference 2010Code 2008 Edition” Australian Geothermal EnergyGroup and the Australian Geothermal EnergyAssociation.The Australian Geothermal Code Committee,2010. “Australian Code for Reporting ofExploration Results, Geothermal Resources andGeothermal Reserves; The Geothermal ReportingCode 2010 Edition” Australian Geothermal EnergyGroup and the Australian Geothermal EnergyAssociation.264

Australian Geothermal Conference 20103D analysis of induced micro-seismic event records compared withgeology and structure: Preliminary results of a Cooper Basin projectChloé Bonet 1,2 , Stewart Hore 1 , Desmond FitzGerald 1 , Michel Rosener 3 ,Helen J. Gibson 1 and David Jepsen 41 Intrepid Geophysics, Unit 2, 1 Male Street, Brighton, Victoria, 3186, Australia.2Institut de Physique du Globe de Paris of Earth's Geophysics, Paris, France.3Geodynamics Limited, Level 2, 23 Graham Street, Milton, Queensland, 4064, Australia.4Geoscience Australia, Cnr Jerrabomberra Avenue & Hindmarsh Drive, Symonston, ACT, 2609, Australia.* Corresponding author: chloe@intrepid-geophysics.comAbstractThe concept of an EGS geothermal prospect isbased on fracture network permeability enhancedby hydraulic stimulation. Characterisation offracture/fault mechanisms and geometries aretherefore an important part of prospectdevelopment.In this study, building a prospect-scale 3Dgeology and structure model of the Cooper Basingeothermal field has assisted prospect explorationand evaluation, but even greater advantage willcome from the ability to integrate (in the sameworkspace) information from induced seismicevents records such as location, magnitude,timing, focal mechanism, and shear planeorientation.Keywords: Induced seismicity, fracture/fault networks,clustering, focal mechanisms, 3D geology models.IntroductionThe Cooper Basin geothermal field is located nearthe common borders of Queensland and SouthAustralia. Since geothermal exploration began in2002, 4 wells have been drilled into the underlyinggranite, with final depths around 4300 metres. Ofthese, the designated injection well, Habanero-1was hydraulically stimulated in 2003 and again in2005. Both stimulations induced detectable microseismicitycentred on TD of Habanero-1 (4421 m).Detailed studies by Geodynamics, GeoscienceAustralia, and Q-Con (Baisch et al, 2006)concluded that the hydraulic stimulationssuccessfully enhanced hydraulic permeabilitybetween the injection well (Habanero-1) and oneof the production wells (Habanero-2). Comparedwith the earlier stimulations, those in 2005extended the previous stimulated reservoir, aswell as further enhancing permeability.Beyond the goal of increased permeability,hydraulic stimulations also provide an opportunityto gather induced seismicity event records whichenable studies on: Fracture / fault mechanism and geometrycharacterization, andlikelihood of seismic risk to infrastructure fromgeothermal reservoir activities.Joint R&D project outlineSince March 1 st 2010, Intrepid Geophysics andGeodynamics have been engaged in acollaborative project with the following goals:1. Develop a 3D micro-seismic time recordsviewer so that new knowledge about faultgeometries and fault mechanisms can evolve andbe integrated in the context of all availablegeology and geoscience observations.2. Investigate clustering or condensing of theevent records in a manner that is coherent withgeological and geomechanical principals, with theaid of filtering capabilities.3. Characterize focal mechanisms in a suitableviewing format, e.g., triangular state diagram.4. Synthesize and integrate other disparateobservations of the geothermal field in a 3Dworkspace, such as: tectonic stress, well fracturerecords, and rock velocity data. The aim here is tofacilitate creation of workable 3D velocity models,and hence to improve the accuracy of locating thehypocenters of micro-seismic events.Induced Seismicity DatabaseThe primary database underpinning this project isthe induced seismicity events records from themid-September 2005 hydraulic re-stimulation ofHabanero-1. Over a period of 13 days 22,500m 3of water was injected into 4421 m. (For details ofthe injection flow rate and wellhead pressure, seeBaisch et al., 2009.)In April 2005 a continuous seismic monitoringsystem was installed at the Cooper Basingeothermal field, and this captured the events ofthe September 2005 re-stimulation. The seismicstation network includes instruments at depthsvarying from surface, to shallow depths (80-370m)265

Australian Geothermal Conference 2010and one deep station at 1780m. (See Baisch 2009for details.)During the mid-September 2005 injection ofHabanero-1, approximately 16,000 detectableseismic events were recorded. From this total, Q-Con determined absolute hypocentre locations for8886 events, that is, events for which at least fiveP-phase and three S-phase onsets could beidentified (Baisch et al. 2009).The data processed by Q-con (containinghypocenter locations, magnitudes and focalmechanism data for 8886 micro-seismic events)was selected for use in the joint R&D project,ahead of an alternative dataset available from aJapanese processing team.3D model of the Copper Basingeothermal fieldA preliminary 3D geology model of the CooperBasin geothermal field site has been constructedin 1 3D GeoModeller software (e.g., see Calcagnoet al, 2008) using formation tops data from thethree Habanero wells, and McLeod-1 drilled in1983 on a site approximately 440 m east-northeastof Habanero-1 (Table 1).WellsnamesLatitude Longitude Depths(m)DipDateFigure 1. Stratigraphic pile for the Cooper and EromangaBasins, within the geothermal field (excluding Cainozoic EyreFm), as used in the geology model shown in Figure 2, below.Habanero1Habanero227°48’57.0’’ 140°45’15.9’’ 4420.82 90° 14/10/0327°49’9.7’’ 140°45’4.9’’ 4357.73 90° 31/03/05Habanero327°48’43.3’’ 140°45’28.9’’ 4221.48 90° 05/02/08McLeod 1 27°48’53.2’’ 140°45’31.3’’ 3806.34 - 08/10/83Table 1. Location and description of the Habanero-1, -2, -3,and McLeod-1 wells.The geological model was also constrained byformation tops data derived from interpretedconventional seismic lines.The preliminary geological model has extents of10 x 10 x 6.5 km, and incorporates thestratigraphic successions of the Cooper andoverlying Eromanga Basin (Fig 1). The mainlyTriassic-aged basin fill, has an average beddingangle around horizontal, and top of the graniticbedrock lies between 3500 and 4200 m (Fig 2).Figure 2.Preliminary geologicalmodel for the CooperBasin geothermal field(view to the northeast).Exaggerated drill holeintersects arehighlighted forHabanero-1, -2 and -3,and for McLeod-1 (atthe rear). Hypocentrelocations estimated byQ-Con for 8886 microseismiceventsinduced during theSeptember 2005hydraulic simulation ofHabanero-1 are shownby the cloud of whitesymbols.266

Australian Geothermal Conference 2010Interpreted basement fault characterizationAs viewed in Figure 2, the hypocentre locationsestimated by Q-Con for 8886 micro-seismicevents induced during the September 2005hydraulic simulation of Habanero-1 are localisedin a zone approximately 3200 m x 1800 mhorizontally (long-axis oriented NNE) and within600 m vertically, centred on TD of Habanero-1(4421 m).Analysis of these event locations by Q-Consuggests induced seismicity aligns along a single,reactivated sub-horizontal fracture system (Baischet al, 2009; Rothert & Baisch, 2010). Indeed, abest-fit plane to the seismic cloud supports thisinterpretation.Pre-existence of this sub-horizontal fracture andit’s primary tectonic origin, is evidenced by loggingdata acquired before any reservoir tests began(D. Wyborn unpublished data, in: Baisch et al,2009).Viewing focal mechanism informationGoal number 3 of the joint R&D project beingundertaken by Geodynamics and IntrepidGeophysics is the ability to visualize focalmechanism information about faults at thehypocentres of induced seismicity. Plots createdon-the-fly from a whole or filtered database arerequired in order to differentiate discrete faults /fractures activated during hydraulic stimulation,and to determine their stress state, orientation,and slip direction, etc. This requires a preprocesseddataset to be loaded, and demandssophisticated plotting and visualization options.Plotting of triangular state diagrams is likely to beimplemented. These will allow groups of seismicevents representing similar focal mechanisms, tobe identified, and conclusions about the presenceof single or multiple structures to be made. Foreach diagram, the 3 poles will represent an endmembersense of movement on a fault: normal,thrust, or strike-slip. Two complete triangular statediagrams are required: one each containingdextral strike-slip and sinistral strike-slip sense.Fracture orientations in the Cooper Basingeothermal fieldAs above, focal mechanism information for 8886micro-seismic events recorded during theSeptember 2005 stimulation of Habanero-1 wasprocessed by Q-con, and provided for this R&Dproject by Geodynamics. From this information,we imported the fracture orientations of all seismicevents, noting (on a first-pass) the existence oftwo distinct populations of fracture orientations:a) 96% of fractures orientations are similarlyoriented, with a mean dip of 9° towards247° (pale blue discs in Fig 3)b) 3% of fractures orientations are similarlyoriented with a mean dip of 20° towards164° (red discs in Fig 3)c) 1% of fracture orientations are not definedas belonging to any population.Figure 3. Visualisation in 3D GeoModeller of fractureorientations of the seismic cloud captured during the 2005hydraulic stimulation of Habanero-1. One dominant and oneminor population constitute the dataset. (See text.)Figure 4 Enlargement of part of the seismic cloud from Fig 3,to highlight the variations in orientations.Compared with the simple analysis visualised inFigs 3 and 4, the ability to further filter and sortthe fracture orientation dataset is required, so thatindividual faults and fault-families can be tagged,assisting the interpreter to trial scenarios ofpossible fault network interpretations – whileworking with superimposed representations ofprior knowledge of geology and structure, andother independent data.267

Australian Geothermal Conference 2010Micro-seismic event magnitudeAdditionally, for the Cooper Basin geothermal fieldwe have facilitated visualization of variablemagnitudes of micro-seismic event records fromthe 2005 stimulation of Habanero-1 (Fig 5). Againthese data were made available through theprocessed dataset of Q-con which includes focalmechanism information.geological and geoscientific knowledge. In turn,improved knowledge on fault / fracture locations,orientations and connectivity will lead to a betterunderstanding of the overall behaviour of areservoir in terms of mechanics, and alsohydraulics.Acknowledgements1 3D GeoModeller is a commercial softwaredeveloped by BRGM and Intrepid Geophysics.ReferencesBaisch S, Vörös R, Weidler R And Wyborn D(2009). Investigation of fault mechanisms duringgeothermal reservoir stimulation experiments inthe cooper basin, Australia, Bulletin of theSeismological Society of America, 99, 148-158Baisch, S.; Weidler, R; Voros, R; Wyborn, D. Andde Graaf, L. (2006). Induced seismicity during thestimulation of a geothermal HFR reservoir in theCooper Basin, Australia. Bulletin of theSeismological Society of America, 96, 2242-2256Calcagno, P., Chiles J., Courrioux, G. and Guillen,A. (2008) Geological modelling from field data andgeological knowledge. Phys. Earth Planet.Interiors, doi: 10.1016 / j.pepi.2008.06.013Rothert E And Baisch S (2010). Passive seismicmonitoring: Mapping enhanced permeability,Proceeding world geothermal congress 2010.Figure 5. Oblique view to the northeast, through hypocentrelocations for micro-seismic events of the 2005 stimulation ofHabanero-1, including enlarged well traces for Habanero-1(lower), Habanero-3 (middle) and McLeod-1 (upper). Thevariations of event magnitude range from +2.9 to -1.0(Moment Magnitude Scale), with highest magnitudes in red,mid-scale in yellow, and lowest in blue.Concluding NoteBeyond customised objectives and early accessto the project results for Geodynamics Ltd, a keyobjective of this joint R&D project is to develop amicro-seismic time records viewer capability in1 3D GeoModeller. As described above, the viewerwill also have the ability to filter and sort recordsbased on attributes such as timing, magnitudeand focal mechanism.These improvements will give reservoir engineersa tool to directly interpret the micro-seismic eventrecord against the background of the prior268

Australian Geothermal Conference 2010Geothermal Energy's Transition from Resource to Power: LegalChallengesAnna Collyer*, Nadia Harrison and Natasha McNamaraAllens Arthur Robinson.Level 27, 530 Collins Street, Melbourne, 3000.* Corresponding author: Anna.Collyer@aar.com.auGrowth and development in the geothermalenergy sector will require the industry tounderstand and comply with several forms ofregulation. The geothermal energy sector hasmuch in common with mining of petroleum andminerals, and is regulated as such. However,because geothermal will ultimately serve as asource of electricity, it is equally important tounderstand the legal regime that governs thenational electricity market and the sector's entryinto the electricity market. Differences in Stateand Territory regulation of geothermal projectsmean that the sector will have different obligationswith which to comply in different jurisdictions.Different approaches to allocating priority amongholders of overlapping tenements are taken underthe various State and Territory regimes. Thegeothermal energy sector's engagement with theScale Efficient Network Extension (SENE)framework and the National Electricity Marketestablished by the National Electricity Law will becrucial if it is to have a significant role in electricityproduction.Keywords: geothermal; regulation; overlappingtenements; National Electricity Market; ScaleEfficient Network ExtensionsRegulation of geothermal resourcesState and Territory Regulatory RegimesThe geothermal sector must come to terms withthe different systems regulating geothermalprojects in each State and the Northern Territory.Each legislative scheme imposes differentobligations, and the differing classifications ofgeothermal resources have implications that willaffect project development. In South Australiaand Western Australia, petroleum legislation wasamended so that it now regulates both petroleumand geothermal energy. In Queensland, Victoriaand the Northern Territory, geothermal energy isregulated by a dedicated Act. In New SouthWales and Tasmania, a geothermal resource isdefined as a mineral and is regulated undermining legislation.In the Australian Geothermal IndustryDevelopment Framework, which was published bythe Federal Department of Resources, Energyand Tourism in 2008, concern was expressed thatthat differences in legislative schemes have thepotential to affect investment decision-making.Investment decisions may be influenced by thenumber of hurdles to exploration and exploitationa particular legislative scheme imposes and thedegree of certainty it provides. Thus, decisionsthat should be based on the potential of thegeothermal resource may be influenced bydifferences in regulation of the industry. One ofthe objectives outlined in the framework is to'harmonise the framework governing the industryacross Australia.' Given the existence of sevendifferent schemes for the regulation of geothermalenergy, this objective will be difficult to achieve.The practical implications of the differinglegislative regimes must be understood. Twoexamples serve to illustrate this. First, in SouthAustralia activity may only be conducted if theregion or land concerned is covered by aStatement of Environmental Objectives that hasbeen approved by the Minister. Geothermal andpetroleum activities are governed by the same Actin South Australia, so Statements ofEnvironmental Objectives developed forpetroleum activities can be adopted forgeothermal operations.Second, native title negotiations must beapproached based on the different regulatorytreatment of geothermal energy in eachjurisdiction. In South Australia, the Governmenthas made a determination that geothermalactivities are not mining activities. This meansthat holders of South Australian geothermallicenses are not subject to the additional criteriaunder the right to negotiate process associatedwith holding mining leases. In Queensland andWestern Australia, holders of geothermaltenements are subject to the more stringent rightto negotiate procedure. In Victoria and theNorthern Territory, native title is dealt withexplicitly under geothermal legislation. In Victoria,native title holders must be compensated for anyloss or damage sustained in relation to theirinterests as a result of geothermal activity. In theNorthern Territory, the Minister must be satisfiedthe native title holder's consent has been obtainedand the federal Native Title Act proceduresfollowed before granting the geothermal authorityapplied for.Overlapping tenementsTenements granted for the purpose ofdevelopment of geothermal resources mayoverlap with other mineral or petroleum269

Australian Geothermal Conference 2010tenements. For example, the Cooper Basin is thesite of a geothermal resource and significant gasand oil reserves. The legislative regimes resolvethe question of prioritisation of interests indifferent ways. The approach to overlappingtenements taken in each jurisdiction is linked toclassification of geothermal energy and the modeof regulation of geothermal energy adopted ineach state.In South Australia and Western Australia,geothermal tenements and petroleum tenementsmay subsist in the same area. Governed underthe same legislation in these states, geothermaland petroleum resources are treated equally.However, prior to the grant of the secondtenement (geothermal or petroleum), the Ministermust consult the existing tenement holder. Thus,in both of these jurisdictions, the Minister has aresponsibility to hear and take into account theconcerns of the existing tenement holder.In Queensland, a more prescriptive approach istaken. Geothermal exploration cannot beundertaken to the extent that it adversely affectsactivity under a mineral or petroleum tenementthat has already started. Conversely, activity ontenements granted under the mineral andpetroleum legislation cannot be undertaken to theextent that it adversely affects activity under ageothermal tenement that is already underdevelopment. This approach clearly gives priorityto the holder of the tenement that was obtainedfirst, whereas in Western Australia and SouthAustralia the existing tenement holder need onlybe consulted before the Minister grants thesecond interest. Although one of the purposes ofthe Queensland Act is to facilitate consultationwith and compensation for persons adverselyaffected by geothermal exploration, the Actprovides no guidance as to how to determinewhether a project is 'adversely affected' by ageothermal project. The Queensland legislationdictates that a project that is first in time andadversely affected by a second proposed activityon the same land has priority. However, discretionin relation to what constitutes an 'adverse effect'remains. The question of whether the first activitywill be 'adversely affected' by the second is amatter that will require resolution in each case ofoverlapping tenements.In New South Wales and Tasmania, geothermalresources have the same status as otherminerals. In New South Wales, an overlappingtenement may not be granted without the consentof the existing tenement holder. As in SouthAustralia and Western Australia, the Minister hasa role to play: disputes between the overlappingtenement holders are referred to the Minister fordetermination. In Tasmania, an explorationlicense or mining lease cannot be granted overland that is already the subject of a tenement.Because geothermal resources are notdistinguished from other minerals in these states,the approach taken to overlapping tenements ismore rigid.The Victorian Act is silent on the question ofoverlapping tenements. However, the OperationPlan that is submitted to the Minister prior to thecarrying out of the geothermal operation mustdetail the risk of injury to land users in the vicinityof the operation.As well as describing these differing regulatoryregimes in the abstract, analysis of thepracticalities associated with attempts to complywith them will also be provided.In the Northern Territory, the holder of ageothermal tenement must consult the holder ofan existing mining right or petroleum interestabout the proposed geothermal activities. Unlikethe Queensland legislation, the NorthernTerritory's Geothermal Energy Act does notprovide that holders of mining rights andpetroleum interests must consult with holders ofexisting geothermal tenements. Unlike the SouthAustralian and Western Australian requirementthat existing tenement holders be consulted priorto the grant of an overlapping tenement,geothermal tenement holders need only consultthe holder of an existing right or petroleuminterest before conducting geothermal activities.The resolution of issues associated withoverlapping tenements often rely on the exerciseof Ministerial jurisdiction, either at the time whenthe tenement is granted, or in the event of adispute. The exercise of Ministerial discretionmay be influenced by political considerations.Holders of overlapping tenements may benefitfrom using contract to overcome uncertainty as tohow their interests will be prioritised. Anothersituation in which contractual negotiation andagreement may also benefit the development ofthe project is where tenements over a singlegeothermal resource are granted to differentparties. In some jurisdictions, cooperation can bemandated. In Western Australia, license holderscan be required to enter into an agreement if theMinister is of the view that unit development of thegeothermal resources area would secure moreeffective recovery of geothermal energy. Similarlyin Victoria, the Minister may require holders ofextraction licences that cover adjacent areas toenter into cooperative agreements for theextraction of geothermal energy. Further insightswill be provided into how contracts are likely tooperate as a practical tool. Our analysis will drawfrom contexts which share importantcharacteristics with the geothermal energy sector,including experiences of the wind power sector.Regulation of geothermal energyWhile geothermal remains a young andunderdeveloped industry in Australia, there will be270

Australian Geothermal Conference 2010a tendency for industry participants to focus onregulatory systems governing preliminary issuessuch as exploration, prospecting and valuation.As geothermal resources are developed tobecome sources of power, parties with an interestin geothermal projects will need to becomefamiliar with the requirements and peculiarities ofthe National Electricity Market (NEM) and therules which apply to it.The NEM: an overviewAn outline of how the NEM operates and some ofthe key principles underpinning the NationalElectricity Law and the NEM will be provided.This will be distinguished from the system whichoperates in the Northern Territory and WesternAustralia. A brief description of the rules andstandards likely to be particularly important forgeothermal projects will also be provided.Entry into the NEM and ContractingElectricity prices in respect of base load powerare determined according to bids placed bygenerators at five minute intervals. In thiscompetitive environment, the capacity to sellelectricity at a specified price cannot be taken forgranted.ContractingThe production of geothermal energy requiressignificant capital investment when compared withother electricity generation methods. TheMassachusetts Institute of Technology citesCalifornian research indicating that capitalreimbursement and interest charges account for65% of the total cost of geothermal power. Thiscan be compared to combined-cycle natural gasplants for example, for which equivalent costsaccount for approximately 22%. In each case theremaining costs cover items including fuel (egwater), labour and access charges.There will therefore be an understandable desireamong industry participants to obtain somesecurity regarding electricity sales and prices.Generators may seek to negotiate longer termunderlying contracts to give themselves, theirfinanciers and their investors certainty. Suchcontracts are commonly relied upon within thecurrent market. The NEM structure means thatderivatives are often used for this purpose,although a PPA may also be appropriate incertain circumstances.As RECs will support the economic viability ofgeothermal projects under the Renewable Energy(Electricity) Act 2000 (Cth), it is likely thatproponents will sell a bundled product comprisingelectricity ('black') and RECs ('green').An outline of the operation of such contractswithin the electricity market currently, includingsome practical tips on contract drafting,negotiation and key risks will be provided.Scale Efficient Network Extensions(SENE)In order to promote development of theinfrastructure needed to transfer electricity from(often remote) energy generation sites to theelectricity grid, the Australian Energy MarketCommission has drafted proposed amendmentsto the National Electricity Rules which wouldfacilitate the construction of SENEs between newelectricity generators and existing infrastructure.While the extensions will be available to varioustypes of energy producers, they are likely to beparticularly important for the renewable energysector.According to the current proposal, SENEs wouldbe developed by Network Service Providers withGenerators paying for the right to transmit energythrough an SENE and the cost of any shortfallbeing passed through to customers. Customershere are those registered with AEMO ascustomers and therefore would include retailersrather than end use customers. Customers arehowever expected to benefit from the preventionof costly duplication which could result ifcompanies were left to each construct their owninfrastructure in order to connect to the network.Outlined below is a sample of some of the issuesraised by the draft amendment which are likely topresent challenges.CapacityThe fact that geothermal sites in any particularregion are likely to develop in a staggered mannermakes the determination of suitable capacity forSENEs particularly difficult. Clearly, there shouldbe enough to enable geothermal and othergenerators to transmit electricity to the market. Atthe same time, a potential over-sizing or 'assetstranding' would be paid for by electricitycustomers and would likely be passed through totheir end-use consumers in the form of higherelectricity costs.As the Network Service Providers responsible fordeveloping SENEs will be able to recover theircosts entirely from generators and customers, theAEMC notes that the incentives may favour oversizing(and excessive capacity) rather than undersizing(and inadequate capacity).Notwithstanding, geothermal industry participantsconsidering investing in reliance on transmissionservices offered through an SENE should notconsider this to constitute a capacity guarantee.The draft rules prescribe that the AER maydisallow proposed SENE projects. Moreover, aNetwork Service Provider will only be required tocommence the SENE planning procedure if thereis a reasonable likelihood of other generators271

Australian Geothermal Conference 2010connecting to the proposed SENE and itsdevelopment is likely to offer material scaleefficiencies. In the event that there is inadequatecapacity, the draft rule provides that the generatorwhose requirements cannot be met is required tofund an augmentation to the SENE. If thegenerator does not agree to fund such anaugmentation and its generation (in excess of itsagreed power transfer capability) has 'constrainedoff' another generator connected to the SENE, thefist generator will be required to pay the othercompensation. Importantly, compensation iscalculated with reference to a formula which willnot necessarily result in recovery of the 'trueeconomic cost of connection'.Relationship with shared networkDepending on the way in which the SENEinfrastructure develops, its interaction with theshared network could present challenges. Wherethe SENE generates benefits and/or costs forusers of the shared network, there is limitedallowance in the rules for their allocation. Theremay also be practical difficulties associated withthe fact that the shared network operates throughan open access system whereas after theintroduction of the SENE, generators will beallocated a specific agreed power transfercapability.Infrastructure development efficiencyThe Clean Energy Council has noted that theNetwork System Operators responsible fordeveloping SENEs have few incentives to avoidcost blow outs.ReferencesApproval Process for EGS Projects (includingInduced Seismicity), Workshop on GeothermalReservoir Engineering, 35 th , Stanford University,Abstract.Gove-White, G., and Frederiks, P., June 2010,Geodynamics - The Opportunity.Hedge, J., 2006, Recent Developments:Geothermal Energy Heating Up, AMPLA Limited13 th Annual Conference, Melbourne.Lund, J. W., 2007, Characteristics, Developmentand Utilization of Geothermal Resources, p. 1-3,7-9.Massachusetts Institute of Technology, 2006,The Future of Geothermal Energy: Impact ofEnhanced Geothermal Systems (EGS) on theUnited States in the 21st Century, Ch. 8-9.Steol Rives LLP Attorneys at Law,McKinsey, J.A., 2009, The Law of Lava, p. 10-17.Mineral Resources Development Act 1995 (Tas).Mining Act 1992 (NSW).Mining Regulation 2003 (NSW).Native Title Act 1993 (Cth).Petroleum and Geothermal Energy Act 2000 (SA).Petroleum and Geothermal Energy Resources Act1967 (WA).Geothermal Exploration Act 2004 (Qld).The authors have also relied on various internalresearch resources owned by Allens ArthurRobinson.Australian Energy Market Commission, 2010,Consultation Paper National ElectricityAmendment (Scale Efficient Network Extensions)Rule 2010.Australian Energy Market Operator, 2010,AEMO's Submission to the AEMC's ConsultationPaper on Scale Efficient Network Extensions, Ch.1, 6, 7, 9, 10.Bartlett, R. H., Native Title in Australia, 2004.Blodgett, L., Holm, A., Jennejohn, D. and Gawell,K., May 2010, Geothermal Energy: InternationalMarket Update, p. 4-10, 51-52.Department of Resources, Energy and Tourism,2008, Australian Geothermal IndustryDevelopment Framework.Geothermal Energy Resources Act 2005 (Vic).Geothermal Exploration Act 2004 (NT).Goldstein, B., Malavazos, M., Long, A., Bendall,B., Hill, T., and Alexander, E., 2010, TheRegulator's Perspective – Best Practice Activity272

Australian Geothermal Conference 2010Trusted Regulation for Geothermal Resource Development (IncludingInduced Seismicity Associated with EGS Operations)Barry Goldstein, Betina Bendall, Alexandra Long, Michael Malavazos, David Cockshell and David LovePrimary Industries & Resources South Australia (PIRSA), GPO Box 1671, Adelaide, South Australia, 5001AbstractOperations need be managed to reduce materialrisks to acceptable levels and as low asreasonably practical to meet communityexpectations. Public trust in both industry andregulators to deliver safe outcomes fostersinvestment and enables efficient land access andactivity approvals. These principles and practicesare especially important for EngineeredGeothermal System development which maypose uncertain risks from induced seismicityThe objective based, one-stop-shop, regulatoryframework in South Australia (Petroleum andGeothermal Energy Act 2000) and the behaviourof the regulator, Primary Industries andResources – South Australia (PIRSA) arerecognized as “a relatively straightforwardregulatory system, which could be considered abenchmark for other jurisdictions” (AustralianProductivity Commission, 2009).The regulatory instruments that deliver trust andefficiency in South Australia are: non-prescriptive;allow for innovation while ensuring that operatorsdemonstrate their ability to manage all possiblerisks to an acceptable level; and entail extensivestakeholder consultation to set standards whichare aligned with community expectations.Operations in South Australia includeinternationally significant Enhanced GeothermalSystems (EGS) developments. Recognizing EGSis (at least to the public) a new technology withuncertain risks, PIRSA has taken account ofinternational developments, commissionedresearch and is supporting national cooperation toincrease certainty in relation to risk managementfor EGS operations.This paper describes practices and technologiesthat can help to inform activity approvals for EGSoperations anywhere.Efficient Co-regulation through Statementsof Environmental Objectives (SEOs) andEnvironmental Impact Reports (EIRs)The South Australian geothermal licenses shownin figure 1 are governed by the Petroleum andGeothermal Energy Act 2000 (P&GE Act).In the context of the P&GE Act, the legalstandards set for the protection of natural, social,heritage and economic environments are agreedthrough a robust, open and transparent researchand consultation process that culminates in P&GEAct license operators owning and abiding bySEOs and associated EIRs (Laws, et al, 2002).EIRs detail potential impacts of proposedoperations and specify strategies to mitigate risksto as low as reasonably practical (ALARP).SEOs set standards for area and operationspecific compliance with co-regulatory objectivesfor the sustainability of natural, social, heritageand economic environments. Hence, SEOsenable PIRSA to act as a first-line, one-stop-shopfor co-regulation for the full-cycle of geothermal,upstream petroleum, high pressure pipelines andgas storage operations in the State of SouthAustralia.Figure 1: Figure 1.Geothermal licenses in South Australia and the SouthAustralian Heat Flow Anomaly (adapted from Neumann, et al 2000)Trust Underpins Efficient Co-regulationEmbedding relevant local, State and Federalobjectives and standards into SEOs make abreach of co-regulatory standards a breach of theP&GE Act.Stakeholders are engaged during thedevelopment of EIRs and draft SEOs and usuallyin a staged process that entails face-to-facemeetings. The process for SEO consultation withstakeholders can take 3 months or moredepending on the level of impact of the activity,and stakeholder consultation requirements. Givenpotential for relatively low environmental impactsand sufficient prior publicly developed anddisclosed criteria, the Minister (who is an electedParliamentarian) may agree public consultationmay be restricted to a time after a relevant SEO isfully developed for consideration. Publicconsultation is undertaken for the development ofother EIRs and draft SEOs.273

Australian Geothermal Conference 2010Final SEOs, final EIRs and annual statements oflicensees’ performance against SEOs and theMinister's determinations of the level ofenvironmental impact of proposals are freelyavailable to the public on PIRSA’s website. PIRSAalso proactively participates in public forums toenable well-informed consideration of riskmanagement strategies put in place to protectnatural, social and economic environments.This openness and transparency underpins trustin PIRSA’s roles as first-line regulator for: theintegrity of plant and equipment; the protection of:water; air; flora; fauna; landscape; heritage; native(aboriginal) title; and as an interlocutor fordisputes between P&GE Act licensees andstakeholders in multiple land use. In addition tothe first-line roles listed above, PIRSA also worksclosely with South Australia’s lead agency for theregulation of occupational health, safety andwelfare (SafeWorkSA) matters.Formal agreements and policies explicate mutualexpectations and underpin both the efficiency andeffectiveness of co-regulation. Hence, licenseeshave a one-stop-shop for regulation.Australia’s Productivity Commission (2009) reviewconcludes that PIRSA “has a clear mandate, clearregulatory responsibilities, good processes toengage with other agencies, and checks andbalances that apply in high risk situations” and “iswidely seen as a model for other jurisdictions toemulate”.Geothermal SEOs and EIRsThe P&GE Act requires that any activity can onlybe conducted if it is covered by an approved SEOfor the region and/or land system within which itwill be carried out. Hence, geothermal operatorsmust develop appropriate EIRs and SEOs orconclude a bridging assessment to make plain thepracticality of adopting and then abiding by preexistingSEOs for analogous operations inanalogous areas. When adopting pre-existingSEOs, the bridging assessment must satisfyrelevant co-regulators that location-specific risksare adequately covered in adopted SEOs. Iflocation-specific risks are not already adequatelycovered, then a new relevant SEO is drafted bythe operator for stakeholder consultation. In someinstances, SEOs developed for petroleum licenseactivities are adopted for analogous geothermaloperations.EIRs underpin the relevance and contents ofSEOs. All licensed field activities must be coveredby an SEO approved by relevant Minister(s).When considering EIRs, PIRSA makes adetermination whether proposed activities shouldbe characterized as low-, medium or high-levelenvironmental impact. For activities characterizedas:• Low-impact – PIRSA consults with relevant coregulatorygovernment agencies.• Medium impact – PIRSA undertakes publicconsultation with support from the operator;and• High impact – an Environmental ImpactStatement process is instigated under SouthAustralia’s Development Act 1993.In this way – final SEOs cover all materialconcerns raised by stakeholders.Licensees are required to report annually and byexception on their performance against SEOs.Five-yearly reviews consider the efficacy of SEOsand, following the principle of transparency, thesereports are available to the public from PIRSA’swebsite.Table 1 provides examples of EIRs and SEOsthat illuminate potential risks, strategies tomitigate risks to as low as reasonably practicaland standards for outcomes for geophysicalsurvey and well operations (including drilling)used to date for deep geothermal well operations.Table 1. Examples of EnvironmentalImpact Reports and Statements ofEnvironmental Objectives (SEOs)Geophysical Surveys State-wide EIR for non-seismic geophysical surveyswww.pir.sa.gov.au/__data/assets/pdf_file/0003/50844/EIR_GeoOps_NonSeis.pdf State-wide SEO for non-seismic geophysical surveyswww.pir.sa.gov.au/__data/assets/pdf_file/0004/50845/SEO_GeoOps_NonSeis.pdf Cooper Basin EIR for geophysical operationswww.pir.sa.gov.au/__data/assets/pdf_file/0011/27398/cooper_basin_geophysical_operations_eir.pdf Cooper Basin SEO for geophysical operationswww.pir.sa.gov.au/__data/assets/pdf_file/0010/27397/cooper_basin_geophysical_operations_seo.pdfDrilling and Well OperationsCooper Basin EIR for drilling and well operationswww.pir.sa.gov.au/__data/assets/pdf_file/0004/27409/drilling_and_well_operations_eir_february_2003.pdf5-Year Review of Operations, Addendum to CooperBasin EIR for drilling and well operationswww.pir.sa.gov.au/__data/assets/pdf_file/0009/123030/Santos_-_Drilling__and__Well_Ops_EIR_Addendum_-_November_2009_Final.pdfCooper Basin SEO for drilling and well operationswww.pir.sa.gov.au/__data/assets/pdf_file/0010/123031/Santos_-_Drilling_and_Well_Operations_SEO_-_November_2009_Final.pdfHabanero Well Operations EIRwww.pir.sa.gov.au/__data/assets/pdf_file/0018/27441/habanero1_eir_sept2002.pdfHabanero EIR and SEO for circulationwww.pir.sa.gov.au/__data/assets/pdf_file/0007/27439/habanero_circulation_eir_seo_oct2004.pdfJacaranda Ridge 2 EIRwww.pir.sa.gov.au/__data/assets/pdf_file/0010/52588/Adelaide_Energy_PEL_255_Region_EIR.pdfJacaranda Ridge 2 SEOwww.pir.sa.gov.au/__data/assets/pdf_file/0010/52597/Final_Adelaide_Energy_PEL_255_SEO.pdfGeodynamics demonstrated the existing CooperBasin SEO for drilling and well operations wasrelevant, and adopted that SEO for its operationsin the Habanero, Jolokia and Savina wells.Petratherm also adopted the Cooper Basin SEOfor its Paralana 2 drilling and well operations.Panax Geothermal has adopted the Otway Basin(Jacaranda Ridge) SEO for its Salamander 1drilling and well operations.Responsible Operator’s Needto Know Best (Practice)274

Australian Geothermal Conference 2010P&GE Act’s non-prescriptive requirements for riskmanagement enable experienced geothermaloperators to easily deploy their existing corporategovernance standards for risk management foruse in South Australia. New entrants togeothermal operations do sometimes express asense of uncertainty while documenting their ownstandard operating (including risk management)procedures. This uncertainty leads some relativelynew entrants to ask to be told how to operate, asa perceived easier path to attaining competencein managing operational risks. In such cases,PIRSA willingly (and often does) ‘hold-the-hand’of new entrant operators ascending learningcurves that result in licence-holders adopting ordeveloping their own activity- and area-specificSEOs and EIRs. Operator ownership of activityoutcomes (and potential liabilities) also satisfiesthe government’s policy objective that the public isprotected from insufficiently managed risks.Additionally, South Australia’s non-prescriptiveapproach allows operators to innovate andmanage risks so efficiently as practical to attainstandards set for outcomes. Sustainableoperations and benign outcomes elicit communitytrust in both the regulator and industry. This is avirtuous cycle.Activity ApprovalsLocal issues for particular field operations areaddressed case-by-case during activity approvalprocesses. In the activity approval process,PIRSA reviews: operator capabilities; fitness-forpurposeof plant and equipment; riskassessments concluded by licensees; and sitespecific environmental impacts. License operatorswho have demonstrated capabilities thatconsistently achieve regulatory compliancerequire low-level surveillance and only need tonotify the regulator of activities, rather thanseeking case-by-case activity approval.Advance Notice of EntryLicense operators must provide 21 days notice inwriting to users of the land that may be affectedby specific regulated activities to relevantstakeholders, including PIRSA. Land users have14 days to raise access-related concerns with thelicense operator and have the option of raising theconcern directly with the regulator, and the finaldispute resolution is a Warden’s Courtproceeding.Induced Micro-seismicityThe most advanced engineered geothermalsystem (EGS) projects in Australia are remotefrom population centers, so experience inAustralia will be gained while potential risks ofinduced micro-seismicity are effectively managed.This experience will be of great value in showingthe extent and magnitude of induced microseismicity,the reliability of pre-stimulationforecasts, and providing a logical basis forpredicting safe distances from fracture stimulationoperations in built-up areas.Regulatory Research for EGS OperationsMany of the geothermal resources in SouthAustralia are expected to be hydraulically fracturestimulatedto achieve optimum (high) rates of heatflow from well-bores. Fracture stimulation ofreservoirs inevitably induces seismic events ofsome measurable magnitude. Proper planningand management of EGS operations can ensurethat risks to people, buildings and infrastructureare reduced to as low as reasonably practical andacceptable levels.To inform regulation, mitigate potential risks andaddress concerns raised by stakeholders, in2005, PIRSA contracted University of Adelaideresearchers to address a critical uncertaintyshared by all geothermal licensees planning todemonstrate EGS in South Australia. Thatresearch (Hunt and Morelli, 2006) assessedinduced seismicity within the context of localgeologic conditions in the Innamincka area of theCooper Basin, and concluded:• Granite basement in the Cooper Basin inSouth Australia is ideally suited to EGSactivities in terms of its compressive stressregime (prone to sub-horizontal fracturepropagation), low levels of natural backgroundseismicity and the availability of extensive highquality reflection seismic to illuminate faultsand fracture trends;• Reactivation of faults in the vicinity of theHabanero site is unlikely. This is due to thenearby faults being beyond the reach of theinduced seismicity associated with EGSactivity.• Induced seismic events at the Habanero wellsite in the Cooper Basin could reasonably beexpected to fall below a ground acceleration of0.05 g, which is a safe level for the Habanerolocation and its surrounds.These findings informed the regulator andstakeholders that the fracture stimulation ofgeothermal wells in the Innamincka area of theCooper Basin could be safely managed so thatmicro-seismic events induced during the fracturestimulation:• would be well below potentially damaginglevels;• were unlikely to induce slip and consequent,larger seismic events on larger geologicalfaults; and• were unlikely to create hydrauliccommunication between the stimulated granite(basement) zones below 4,000 metres and theoverlying sedimentary Cooper Basin above3,700 metres.This last finding is based on:• the prevailing, natural, highly compressivestress regime acting to constrain fracturepropagation to sub-horizontal intervals;• acceleration attenuation is at least one order ofmagnitude greater in soft rocks and soils thanin crystalline rocks; and• high frequency motion is attenuated morequickly with distance than low frequency275

Australian Geothermal Conference 2010motion and the seismic events induced duringfracture stimulation and pumping into HotFractured Rocks (HFR) are characteristicallyhigh frequency (100–500 Hz).Indeed, fracture stimulation and injectionprograms in Geodynamics’ Cooper BasinHabanero wells were both conducted safely andwere successful in the enhancement and flowtesting of EGS reservoirs.Risk Management for Induced SeismicityGiven the results in the Cooper Basin, and lookingforward to many additional EGS projects inAustralia (and South Australia, in particular),PIRSA commissioned the development of riskmanagement protocols for induced seismicityassociated with EGS reservoir development in2007. The findings (Morelli and Malavazos 2008)are fully consistent with the findings of Majer,Baria and Stark (2008) and are summarized inTable 2.Running Ahead of the Frac CrewAn informed risk assessment for EGS operationsstarts with an analysis of:• historical (monitored) earth movementsmagnitude and location; and• geophysical survey data to relate earthmovements to faults and fracture trends.The adequacy of seismic monitoring arrays has abearing of the certainty of seismic eventmagnitudes, locations and sense of motion, andhence the usefulness of recorded (historical)base-line information. Equipment capable ofsensing seismicity (detectability) may only providecomplete records of all movements at a higherlevel (completeness), because instrumentationmaybe insufficient to detect many small events.The seismic monitoring stations in South Australiaare depicted in Figure 2. Additional seismicmonitoring stations are located in adjacentjurisdictions. The locations of four additionalstations in South Australia have been agreedbetween State and Federal Government agencies(as shown in Figure 2) and the equipment to bedeployed will enhance both the completeness andaccuracy of measuring seismic events.The detection limits of the existing seismographnetwork in and around South Australia is variable(as shown in Figure 2, based on Dent, 2009), andthis array is reliably locating all events aboveRichter magnitudes of 3.5 (e.g. recording iscomplete for events >3.5) and is more resolute(complete for events above Richter magnitudes of2.0) for settled areas. This is considered sufficientto manage public safety under currentcircumstances. If required, existing networks canbe augmented to provide higher resolution of thelocation and depth of hypocentres.Table 2 Information that can most help to inform activityapprovals (or otherwise) for the fracture stimulation ofgeothermal reservoirs includes: Characterization of the local environment,infrastructure and population for vulnerability toground movements and loss modeling (taking accountof design standards) High-quality records of seismicity waveforms,magnitude and location; Thickness and shear velocity of soil and weatheredcover over bedrock. Measuring shear velocities to 30metres depth is a generalized suggestion; Reservoir data for characterization – including:- Orientation and magnitude of stress fields;- Location, extent of faults and fracture trends;- Mechanical, thermal and chemical rock properties, and- Hydrologic parameters (extent, pressure, chemistry andnature of confining aquitards) Conclude loss modeling (taking account of designstandards and infrastructure that pre-dates designstandards)Non-exhaustive protocols for credible risk management forgeothermal operations that may induce seismicity Apply national or international standards for riskmanagement Proponent to demonstrate adequate assessment ofpotential consequence of induced seismicity for sitesselected for hydraulic stimulation or large scaleinjection. Stakeholder engagement to start as soon as is practical If required, augment the existing seismic monitoringnetwork to detect and gather seismic events ofmagnitudes (Richter scale) less than 3. It will beadvantageous to deploy seismic monitoring stations to:- Continuous digital high sample frequency (≥ 100 htz)recording;- Attain adequate network to accurately locate seismicevents and measure attenuation; and- Geophysical surveys to calibrate regolith responsemodels at EGS locations. Maintain the seismic monitoring network for the life ofthe project. As practical, deploy at least one sub-surface seismicmonitoring station (below regolith if possible) prior tohydraulic stimulation or large scale injection. Deploy down-hole and near surface monitoring stationsto determine attenuation and regolith amplification. Sustain an evergreen watching brief so new informationis assessed and considered for induced seismicity riskmanagement.The location and magnitude of historical, recordedearth movements in South Australia are depictedin Figure 3. This map does not express theuncertainty of epicenter locations, but thisuncertainty is a factor considered when assessingpotential risks posed by EGS operations.The most advanced EGS projects in Australia arethose of Geodynamics (Habanero, Jolokia andSavina wells – see figure 1) and Petratherm(Paralana 2 – see figure 1). Only Habanero wellshave been fracture stimulated by year-end 2009.High resolution seismic monitoring arrays havebeen installed at Habanero and Paralana to bettermeasure both background seismicity andseismicity induced during stimulation, productionand circulation operations. The array positioned atHabanero can detect and locate events as low as-2.5 (Richter scale) at a depth of 5 km, with a 3D276

Australian Geothermal Conference 2010locational accuracy of about 30metres, but thecompleteness of accurately computing all eventsis probably limited to -1 (Richter scale). Networkscan be augmented as required to provide greaterresolution and completeness of the epicenterlocations.Hot Sedimentary Aquifer projects that do notrequire fracture stimulation and entail welloperations largely analogous to petroleum welloperations do not necessarily need to deployseismic monitoring arrays.Reflection seismic is useful to optimize drillinglocations for EGS targets. Figure 4 depicts vastareas in the South Australian Heat Flow Anomalythat are remote from population centers, andcovered with at least some modern reflectionseismic information.Figure 2. Locations and approximate capability of seismographnetwork in South Australia (including use of seismographs inadjacent States and the Northern Territory. Adapted fromDent, 2009. The South Australian Heat Flow Anomaly is alsoshown (adapted from Neumann, et al 2000).• the Australian Building Code AS 1170.4 -(1993) characterizes this region as having a10% chance in 50 years of experiencing apeak ground acceleration of 0.05g, and• the calculated maximum peak groundacceleration from EGS operations in theHabanero wells is 0.041g, which for thislocation.Figure 3 Location and magnitude of historical earth movementepicenters in South Australia. The South Australian Heat FlowAnomaly is also shown (adapted from Neumann, et al 2000).Figure 4. Reflection seismic lines (2D and 3D surveys)onshore South Australia. The South Australian Heat FlowAnomaly is also shown (adapted from Neumann, et al 2000)Tools of Trade in AssessingAttenuationStandards for attenuation for induced seismicityneed be locally appropriate. The largest recordedmagnitude seismic event associated with EGSoperations at Habanero in the Cooper Basindetermined to be of magnitude 3.7 on the Richterscale e.g. an event felt but one that would rarelycause damage, and which had no significantnegative impacts on local social, natural oreconomic environments.Analysis of natural and induced seismicity(associated with EGS operations in the term2003-5) in the Innamincka region in the CooperBasin, Hunt and Morelli (2006) found:• natural seismicity can be expected to rangebetween Richter magnitudes of 3.5 and 4.0once every 50 to 167 years;• the largest event in the area prior to 2003 wasin 1979 with magnitude 2.9;Using Toro’s (1997) relationships, Hunt andMorelli (2006) mapped the attenuation distance ofpeak ground acceleration in units of gravity (g)and distance (in kms) from the Habanero site asredisplayed here in figure 5.Figure 5. Toro (1997) relationship calculation of attenuationdistance of peak ground acceleration in units of gravity (g) and277

Australian Geothermal Conference 2010distance (in kms) generated at the Habanero well locationsnear Innamincka in the Cooper Basin. From fig.8 in Hunt andMorelli, 2006)Figure 6. Estimated attenuation of earth movement intensity ofa Richter scale 3.7 magnitude seismic event with distancefrom an event hypocenter in northeast South Australia.Modified Mercalli Intensity ScaleAttenuation Distance from Hypocenters basedon Greenhalgh, et al (1990) function for Australia876543210 100Hypocentral Distance (Km )Another approach is illustrated with Figure 6,which is output from a spreadsheet tooldeveloped by PIRSA (Love, 2009) that usesestimates of earth movement (seismic wave)velocities to forecast modified Mercalli scaleintensity attenuation with distance fromhypocentres. This spreadsheet tool estimatesintensity attenuation based on five functionspublished in: Bierbaum, et al, 1994; Gaull, et al1990; Greenhalgh, et al, 1990; and Greenhalgh,et al 1994. The Greenhalgh, et al (1994) functionfor Australia is considered the best approximationof these five correlations to characteriseseismicity induced during EGS operations in theCooper Basin in South Australia. More detailedanalyses (akin to the analysis illustrated in figure5) are expected to be concluded by EGSoperators to optimise fracture stimulationprograms, assess potential hazards, and underpinconsultation with stakeholders. Based on thiscorrelation (figure 6), the maximum (3.7 on theRichter scale) recorded event at Habanero ischaracterised as having attenuated to a modifiedMercalli scale intensity of 4.8 at 10 km distancefrom the hypocenter and diminished to a slightintensity at a distance 20 km from the hypocentere.g.

Australian Geothermal Conference 2010Figure 7. Intensity map of the Beachport (South Australia)earthquake of 10 May 1897 based on print media reports fromhttp://www.pir.sa.gov.au/minerals/earthquakes/major_earthquakes_in_south_australia.Better Baseline DataA unique opportunity arises for cooperation toefficiently meet multiple objectives for publicsafety, and exploration for EGS, unconventionalgas reservoir sweet-spots and geosequestration.This would entail cooperation of: governmentagencies responsible for assessing geo-hazardsassociated with earth movements; proponents ofdeveloping fractured reservoirs for the productionof heat energy (e.g. EGS); proponents ofdeveloping coal bed methane, shale gas and tightgas reservoirs; and proponents of subsurfacegreenhouse gas storageIn particular – it will be advantageous forcompanies exploring for reservoir sweet-spotsrelated to tensile rock fabrics and seismicallyquiescent storage reservoirs and relevantgovernment agencies to coordinate plans in thecontext of:• publically managed seismic monitoringnetworks, so those networks are augmentedwith multiple objectives in mind;• privately installed monitoring stations becomepublic assets, post-decommissioning ofindustry’s projects; and• national and international forums to fosterinteroperability of databases and softwareapplications (input and output).Cooperation will advance both knowledge ofinduced seismicity risks and reservoirdevelopment opportunities.Conclusions1. Co-regulatory efficiency and effectiveness forgeothermal operations can be delivered withan objective-based and transparent one-stopshopapproach as applied in South Australia.2. PIRSA’s research into potential risks posed byEGS operations has informed regulatoryapprovals for fracture stimulation operations ingeothermal wells in areas that are remote frompopulation centers.3. The magnitude and extent of micro-movementinduced by fracture stimulation and injection atHabanero in the Cooper Basin were largely aspredicted e.g. EGS reservoirs were createdand circulated without adverse impacts.4. Calibration of predictive models for the riskmanagement of induced seismicity in remotelocations will provide benchmarks for theregulation of EGS projects nearer to populatedlocations.5. Australia is becoming a globally importantlaboratory for EGS operations.6. Given enough experience – risk managementstrategies for fracture stimulating and injectinginto geothermal reservoirs are expected toevolve and EGS operations are expected tobecome predictably profitable and reliablysafe. The outcome will be wide-spreadcommunity and investor trust in EGSdevelopment.ReferencesAustralian Building Code AS 1170.4 (1993),Minimum design loads on structures. Part 4Earthquake Loads. The Australian Standard wasprepared by Committee BD/6, Loading onStructures. It was approved on behalf of theCouncil of Standards Australia on 21 April 1993and published on 16 August 1993. A 2007 updateof this reference is available for purchase from thesales agent for Standards – Australia from:http://infostore.saiglobal.com/store2/Details.aspx?ProductID=215710.Bierbaum, S. J, 1994. Earthquake hazard andmicro-tremor analysis, South Australia” HonoursThesis, Flinders University of South Australia,Australia. Quotes formula used in Love (2009)that was developed by K. Malpas.Dent, V. F., 2009. Seismic network capability andmagnitude completeness maps, 1960 – 2005 forWest Australia, South Australia and the NorthernTerritory, proceedings from the AustralianEarthquake Engineering, Society Conference,Newcastle, NSW, Australia, 11-13 December2009 (see:http://www.aees.org.au/Proceedings/Proceedings.html).Gaull, B. A., Michael-Leiba, M. O., and Ryan, J.M. W., 1990; Probabilistic earthquake risk mapsof Australia, Australian Journal of Earth Sciencesv 37 pp 169-187.Greenhalgh S. A., and McDougall, R. M. 1990.Earthquake Risk in South Australia, AustralianCivil Engineering Transactions v 32 no 3.Greenhalgh S. A., Love, D., Malpas, K., andMcDougall, R. M. 1994. South Australianearthquake, 1980-92, Australian Journal of EarthSciences v 41 pp 483-495.Hunt, S.P. and Morelli, C., 2006. Cooper BasinHDR hazard evaluation: predictive modelling oflocal stress changes due to HFR geothermalenergy operations in South Australia. SouthAustralia Department of Primary, Industries andResources. Report Book 2006/16.279

Laws, R. A., Aust, T. and Malavazos, M., 2002.Environmental regulation of the upstreampetroleum industry in South Australia. The APPEAJournal, 42 (1), 683–96.Love, D., 2009. unpublished spreadsheet tocalculate intensity attenuation with distance fromhypocentres, Primary Industries & Resources –South Australia.Majer, E., Baria, R., Stark, M., Smith, B., Oates,S., Bommer, J., and Asanuma, H. (2006). InducedSeismicity Associated with Enhanced GeothermalSystems. Available at: http://www.ieagia.org/documents/ISWPf1MajerWebsecure20Sep06.docMajer, E., Baria, R. and Stark, M. (2008). Protocolfor induced seismicity associated with enhancedgeothermal systems. Report produced in Task DAnnex I (9 April 2008), International EnergyAgency-Geothermal Implementing Agreement(incorporating comments by: C. Bromley, W.Cumming, A. Jelacic and L. Rybach). Available at:http://www.iea-gia.org/publications.asp.Morelli, C., and Malavazos, M. (2008), “Analysisand Management of Seismic Risks AssociatedWith Engineered Geothermal System Operationsin South Australia, in Gurgenci, H. and Budd, A.R. (editors), Proceedings of the Sir Mark OliphantInternational Frontiers of Science and TechnologyAustralian Geothermal Energy Conference,Geoscience Australia, Record 2008/18, pp 113-116. Download fromhttp://www.ga.gov.au/image_cache/GA11825.pdfNeumann, N., M. Sandiford, J. Foden, 2000.Regional geochemistry and continental heat flow:implications for the origin of the South Australianheat flow anomaly Earth and Planetary ScienceLetters, 183 pp. 107-120.Productivity Commission, 2009, Review ofRegulatory Burden on the Upstream Petroleum(Oil and Gas) Sector, Research Report, April2009, Melbourne, Australia. Published by theCommonwealth of Australia. Download from:http://www.pc.gov.au/__data/assets/pdf_file/0011/87923/upstream-petroleum.pdfSouth Australian Government (2009), Petroleumand Geothermal Energy Act 2000, as amended inDecember 2009. Download from:http://www.legislation.sa.gov.au/LZ/C/A/PETROLEUM%20AND%20GEOTHERMAL%20ENERGY%20ACT%202000.aspx.Toro, G.R., N.A. Abrahamson and J.F. Schneider(1997). A Model of Strong GroundMotions from Earthquakes in Central and EasternNorth America: Best Estimates andUncertainties. Seismological Research Letters,v.68, no. 1, pp. 41-57.USGS, 2009. The Modified Mercalli IntensityScale. Abridged from The Severity of anEarthquake, USGS General Interest Publication1989-288-913. Webpage updated 27 October2009 at:www.earthquake.usgs.gov/learn/topics/mercalli.phpAustralian Geothermal Conference 2010AcknowledgmentsThe authors thank Ladsi Rybach (GeoWatt), RoyBaria (EGS-Energy) and PIRSA colleagues fortheir kind inputs, especially James Coda, BelindaHayter Carice Holland, Peter Hough, and ShaneFarrelly.Keywords: geothermal, EGS, induced seismicity,attenuation distance, one-stop-shop regulation,geohazards280

Australian Geothermal Conference 2010GREAT EXPECTATIONS FOR GEOTHERMAL ENERGY – FORECAST TO 2100Goldstein, B.A., Hiriart, G., Tester, J.,. B., Bertani, R., Bromley, Gutierrez-Negrin, L.,C.JHuenges, E., H, Ragnarsson, A., Mongillo, M.A. Muraoka, and V. I. Zui,INTRODUCTIONGeothermal energy systems have: a modest environmental footprint; compared to many renewableand conventional alternatives ; will not be impacted by climate change; and have potential to becomethe world’s lowest cost source of sustainable thermal fuel for zero-emission, base-load direct use andpower generation. Displacement of more emissive fossil energy supplies with geothermal energy canalso be expected to play a key role in climate change mitigation strategies.In this context, sharedchallenges on the road to a global portfolio of safe, secure, competitively priced energy supplies aredrivers for international cooperation in research, exploration, pilot demonstration and pre-competitivedevelopment of geothermal energy resources and technologies.The intellectual and financial inputs for international, pre-competitive initiatives are coming from publicand private investors with aspirations for low emissions, affordable, and globally deployable 24/7energy supplies. It is reasonable to conclude that the outcomes (improved technologies and methods)of these collective efforts over the next 20 years will underpin great expectations for widespread,profitable and environmentally sustainable use of geothermal energy for centuries to come. This paperprovides a synopsis of recent findings and current objectives of notable international fora enablingcooperation to reduce impediments to widespread use of geothermal energy.SHARING OF PRE-COMPETITIVE KNOWLEDGEThe co-authors of this paper have supported one or more of the international geothermal energy foralisted in Table 1.Table 1. Names and websites of 10 key international geothermal energy foraInternational Energy Agency’s Geothermal Implementing Agreement(IEA GIA)International Geothermal Association (IGA) and its World GeothermalCongress (WGC) (a)International Partnership for Geothermal Technologies (IPGT) (a)Geothermal Engineering Integrating Mitigation of Induced Seismicityin Reservoirs (GEISER)ENhanced Geothermal Innovative Network for Europe (ENGINE)European Energy Research Alliance Joint Programme on GeothermalEnergy (EERA JPGE)European Geothermal Energy Council (EGEC)Geothermal Resource Council (GRC) and its annual conference inparticularGeothermal Resource Association (GEA)Stanford University Geothermal WorkshopsInternational Panel for Climate Change (IPCC) Working Group III –Special Report on Renewable Energy (and in particular Chapter 4 –Geothermal)http://www.iea-gia.org/http://www.geothermal-energy.org/http://internationalgeothermal.org/http://www.geiser-fp7.eu/default.aspxhttp://engine.brgm.fr/http://www.eera-set.eu/index.php?index=36http://www.egec.org/http://www.geothermal.org/http://www.geo-energy.org/http://pangea.stanford.edu/ERE/research/geoth/conference/workshop.htmlhttp://www.ipcc-wg3.de/publications/specialreports/special-report-renewable-energy-sourcesSHARED CHALLENGES BEGET COMPLEMENTARY ACTIONImproved, evermore reliable, cost-effective methods to enhance the productivity of geothermalsystems will be essential to the competitiveness of geothermal resource in energy markets. Inparticular, the commercialisation of fracture and/or chemical stimulation methods to reliably createEngineered Geothermal Systems (EGS) will be one key milestone on the road to great expectationsfor widespread economic use of geothermal energy. A scan of the objectives of the internationalgeothermal energy fora define the high priorities summarized in Table 2.281

Australian Geothermal Conference 2010Table 2. Twenty key priorities for international geothermal energy fora. This is an update priorities presented inGoldstein et al., 2009.1. Openness to cooperation to engender complementary researchand the sharing of knowledge2. Informing industry experts, government policy makers andthe public at large through presentations, publications,websites, submissions to enquiries and the convening ofconferences, workshops and courses3. Creating effective standards for reporting geothermal4. For EGS, improved hard rock drill equipmentoperations, resources and reserves5. Predictive reservoir performance modelling 6. For EGS, improved multiple zone isolation7. Predictive stress field characterisation 8. For deep EGS, reliable submersible pumps9. For EGS, mitigate induced seismicity 10. Longevity of well cement and casing11. Condensers for high ambient-surface temperatures 12. For EGS: Optimum fracture stimulation methods13. Use of CO 2 as a circulating fluid 14. High temperature logging tools and sensors15. Improve power plant design 16. High temperature flow survey tools17. Technologies & methods to minimise water use 18. High temperature fluid flow tracers19. Predict heat flow and reservoirs ahead of the bit 20. Mitigation of formation damage, scale and corrosionUnited States3094 MWFrance16 MWPortugal29 MWIceland575 MWGermany6.6 MWAustria1.4 MWItaly843 MWTurkey82 MWEthiopia7.3 MWKenya167 MWChina24 MWRussia82 MWMexico958 MWJapan536 MWGuatemala52 MWPhilippines1904MWEl Salvador204 MWPapua‐N. G.56 MWNicaragua88 MWN. Zealand628 MWCosta Rica166 MWTotal : 10,715 MWeThailand0.3 MWIndonesia1197 MWAustralia1.1 MWFigure 1 Geothermal-electric installed capacity by country in 2009. Figure shows worldwide average heat flow in mW/m 2 and tectonic plates boundaries(illustration adapted from a figure in Hamza et al., 2008 with data from Bertani, 2010).This map of heat flow does not reconcile all geothermal information.The delineation of geothermal resources will be improved by integrating temperature gradient, heat flow and reservoir data.INFORMING POLICY MAKERS AND THE PUBLIC AT LARGEWORLD REPORT - GEOTHERMAL ENERGY USEGeothermal energy supplies are currently used to generate base-load electricity in 24 countries withan installed capacity of 11 gigawatts of electricity (GWe) and a global average capacity factor of 71%,with newer installations above 90%, providing 10% to 30% of their electricity demand in six countries(Bertani, 2010). Figure 1 is a map of estimated heat flow in milliwatts per square metre (Wx10 -6 /m -2 )annotated with geothermal electricity generation capacity by country. Geothermal energy supplies arealso used for direct use applications in 78 countries, accounting for 50 GW thermal including district(space) heating and cooling and ground-source (geothermal) heat pumps (GHPs), which haveachieved significant market penetration worldwide (Lund et al.,2010).282

Australian Geothermal Conference 2010BUILD (INNOVATE) AND MARKET BETTER MOUSE-TRAPSThe obvious generalised impediments to massive, global geothermal energy use are: currently insufficiently predictable reliability of geothermal reservoir performance (and in particular,the predicable reliability of engineered geothermal system reservoirs); and current costs of geothermal well deliverability (and in particular, fluid production levels fromstimulated engineered geothermal systems and the high costs of drilling deep wellsHence, the over-arching common and well justified objectives of global government initiatives are tostimulate technologic and learn-while-doing breakthroughs that will lead to a point where the cost ofgeothermal energy use is reliably cost-competitive and comparatively advantageous within markets.MARKET (COMMUNICATE!)Geothermal resources contain thermal energy that can be produced, stored and exchanged (flowed)in rock, gas (steam) and liquids (mostly water) in the subsurface of the earth.With proper management practice, geothermal resources are sustainable and renewable overreasonable time periods. As stored thermal energy is extracted from local regions in an activereservoir, it is continuously restored by natural conduction and convection from surrounding hotterregions, and the extracted geothermal fluids are replenished by natural recharge and by reinjection ofthe exhausted fluids. Additionally: geothermal plants have low- to emissions-free operations and relatively modest land footprints. Theaverage direct emissions yield of partially open cycle, hydrothermal flash and direct steam electricpower plants yield is about 120 g CO 2 /kWhe 1 . Current binary cycle plants with total reinjection yieldless than 1 g CO 2 /kWhe in direct emissions. Emissions from direct use applications are even lower(Fridleifsson et al., 2008). Over its full life-cycle (including the manufacture and transport ofmaterials and equipment), CO 2 -equivalent emissions range from 23-80 g/kWhe for binary plants(based on Frick et al. 2010 and Nill, 2004) and 14-202 g/kWht for district heating systems andGHPs (based on Kaltschmitt, 2000). This means geothermal resources are environmentallyadvantageous and the net energy supplied more than offsets the environmental impacts of human,energy and material inputs; geothermal electric power plants have characteristically high capacity factors; the average forpower generation in 2009 is71% (67,246 GWh electrical used from installed capacity of 10.714GW electrical based on Bertani, 2010), and modern geothermal power plants exhibit capacity factorsgreater than 90%. This makes geothermal energy well suited for base-load (24/7), dispatchableenergy use; the average estimated 27.5% capacity factor for direct use in 2009 (121.7 TWh thermal used frominstalled capacity of 50.6 GW thermal based on Lund, et al., 2010) can be improved with smart grids(as for domestic and industrial solar energy generation), by employing combined heat and powersystems, by using geothermal heat absorptive and vapour compression cooling technology, and byexpanding the distributed use of ground source heat pump energy generation; and properly managed geothermal reservoir systems are sustainable for very long term operation,comparable to or exceeding the foreseeable design-life of associated surface plant and equipment.CHARACTERISING GEOTHERMAL RESOURCES, RESERVES AND SUPPLIESThe theoretical global geothermal resource base corresponds to the thermal energy stored in theEarth’s crust. The technical (prospective) global geothermal resource is the fraction of the earth’sstored heat that is accessible and extractable for use with foreseeable technologies, without regard toeconomics. Technical resources can be subdivided into three categories in order of increasinggeological confidence: inferred, indicated and measured (AGEG-AGEA, 2009), with measuredgeothermal resources evidenced with subsurface information to demonstrate it is useable. Geothermalreserves are the portion of geothermal resources that can confidently be used for economic purposes.Geothermal reserves developed and connected to markets are energy supplies. Accessiblegeothermal resources are enormous as detailed in Figure 2. Resources size is clearly not a limitingfactor for global geothermal energy development.1 This is the weighted average of 85% of the world power plant capacity, according to Bertani and Thain, 2002,and Bloomfield et al., 2003283

Australian Geothermal Conference 2010Figure 2. Potential geothermal energy resources split into categories e.g. theoretical, technical, economic, developable and existing supplies for powergeneration and direct use. All categories for power generation assume a 71% capacity factor and 8.1% average efficiency for converting thermal intoelectrical energy, though both factors will likely improve (increase) in future. All direct use estimates assume an average 31% capacity factor, somewhathigher than the average (27.5%) in 2009. Adapted from a figure developed by Ladsi Rybach and published in the Proceedings of the Joint IEA-GIA~IGAWorkshop: Geothermal Energy Global Development Potential and Contribution to Mitigation of Climate Change, 5-6 May 2009, Madrid, Spain, edited byM.A. Mongillo, 20 March 2010FAQ: WHAT IS GEOTHERMAL ENERGY AND HOW DOES IT WORK? (COMMUNICATE!)Geothermal energy is the terrestrial heat stored in, or discharged from rocks and fluids (water, brines,gases) saturated pore space (including fractures), and is widely harnessed in two ways: for power(electricity) generation; and for direct use e.g. heating, cooling, aquaculture, horticulture, spas and avariety of industrial processes, including drying. The use of energy extracted from the constanttemperatures of the earth at shallow depth by means of ground source heat pumps (GSHP) is acommon form of geothermal energy use. The direct use of natural flows of geothermally heated watersto surface have been practised at least since the Middle Palaeolithic (Cataldi, 1999), and industrialutilisation began in Italy by exploiting boric acid from the geothermal zone of Larderello, where in 1904the first kilowatts of electric energy (kWe) were generated and in 1913 the first 250-kWe commercialgeothermal power plant was installed (Burgassi, 1999).Where very high temperature fluids (> 180 C) flow naturally to surface (e.g. where heat transfer byconduction dominates), geothermal resources are the manifestation of two factors: a geologic heat source to replenish thermal energy; and a hydrothermal reservoir that can be tapped to flow geothermal fluids for its direct use and/or forgenerating electricity.Elsewhere, a third geologic factor, the insulating capacity of rocks (acting thermal blankets) is anadditional necessary natural ingredient in the process of accumulating usable, stored heat energy ingeologic reservoirs that can be tapped to flow heat energy, and replenished by convective, conductiveand radiated heat flow from sources of geothermal energy.284

Australian Geothermal Conference 2010Usable geothermal systems occur in a variety of geological settings. These are frequently categorizedas follows: 1. High-temperature (>180°C) systems at depths above (approximately) 3.5 km aregenerally associated with recent volcanic activity and mantle hot spot anomalies.. Other hightemperature geothermal systems below (approximately) 3.5 km are associated with anomalously highheat producing crustal rocks, mostly granites. 2. Intermediate temperature systems (100-180°C) and3. Low temperature ( 180C) fluids to surface. Indeed, temperatures above 1000°C can occurat less than 10 km depth in areas of magma intrusion. Given the global average land area surfacetemperature of (about) 15C and an approximate global geothermal temperature gradient for landareas outside volcanic settings of (about) 30C, the same high temperature (> 180C) can be reached(on average) at a depth of about 5.5 km below ground level.The main types of geothermal power plants use direct steam (often called dry steam), flashed steamand binary cycles.Power plants that use dry and/or flashed steam to spin turbines are the most commonly deployed formof geothermal electricity generation. These plants use the heat energy contained in water and steamflowed from geothermal wells to spin turbines, converting thermal and kinetic energy to electricalenergy.Organic Rankine power plants employing secondary working fluids are increasingly being used forgeothermal power generation. These so-called binary closed-loop power plants do not flow producedgeothermal fluids directly into turbines. Thermal energy contained in water and/or steam producedfrom geothermal wells is transferred to a secondary working fluid using a heat exchanger (hence theterm binary closed-loop). Organic compounds with lower boiling points than water (such as propanethat boils at about 28 degrees C are often used as working fluids. The heat energy in the geothermalfluid boils the working fluid changing it from a liquid to a pressurized gas within the closed-loop, whichcan then be expanded in a turbine to spins a generator. The exhausted working fluid is cooled,condensed back into a liquid, pressurized and then recycled into the heat exchanger to complete thecycle.285

Australian Geothermal Conference 2010Table 1 Estimated global long term forecasts of installed capacity for geothermal power and direct uses (heat) and of electricand direct uses (heat) generation from Bertani, 2010 and Authors, 2010.Expected World UseExpected global useDirect(GWt)2020 2030 2050 2100Electric(GWe)Direct(GWt)Electric(GWe)Direct(GWt)Electric(GWe)Capacity 160.5 25.9 380.1 51.0 831.1 160.6Direct(GWt)1,367 to5,906Electric(GWe)264 to1,141TWh t /y TWh e /y TWh t /y TWh e /y TWh t /y TWh e /y TWh t /y TWh e /y421.9 181.8 998.8 380.0 2184.0 1266.43,592to15,5212,083to9,000EJ/y EJ/y EJ/y EJ/y EJ/y EJ/y EJ/y EJ/y12.9 to 7.5 to1.52 0.65 3.60 1.37 7.86 4.5655.9 32.4HOW BIG WILL GEOTHERMAL BE? (COMMUNICATE!)The extent or accessibility of geothermal resources will not be a limiting factor for deployment. Tables1 and 2 summarise the conclusions reached by the co-authors in 2010). These forecasts assumeimprovements in capacity factors power generation from the current average 71% to 90% by 2050, alevel already attained in efficient, existing geothermal power generation plants. Earlier estimates fordeployment beyond 2010 that were considered in developing forecasts in Table 1 include: IPCC,2007; IEA, 2008; and EREC, 2008.Table 2 World installed capacity, electricity production and capacity factor of geothermal power plants 1995-2010, forecastsfor 2015-2050 (adapted from data from Bertani, 2010 and Authors, 2010) and forecasts for 2100 based on 1% and 4%average annual growth for 50 years from 2050.YearInstalled Capacity (GWe) Electricity Production (GWh/y) Capacity FactorActual or mean forecast Actual or mean forecast(%)1995 6.8 38,035 642000 8.0 49,261 712005 8.9 56,786 732010 10.7 67,246 712015 18.5 121,600 752020 25.9 181,800 802030 51.0 380,000 852040 90.5 698,000 882050 160.6 1,266,400 902100 264 to 1,141 2,082,762 to 8,999,904 90+NEXT STEPS IN GLOBAL RESOURCE ASSESSMENTSA further global geothermal resource assessment is planned under an existing IEA GIA researchannex. This will include a probabilistic range of estimates, for example, assuming that a log normaldistribution adequately describes the range of recovery of stored heat from a minimum of 0.5% at a99% probability to a maximum of 40% of stored heat at a 1% probability. This implies: a low-siderecovery of 1.34% of stored (90% probability); a mid-range recovery of 4.47% of stored heat (50%probability); a Swanson’s mean 2 recovery of 6.68% of stored heat; and a high-side recovery of14.95% of stored heat (10% probability).CONCLUSIONS REGARDING DEPLOYMENT BY 2100Current global trends and regional research underpin credible expectations for great growth in theglobal use of geothermal energy over the next 90 years. Great expectations are:Power generation with binary plants and total re-injection will become common-place in countrieswithout high-temperature resources.2 Swanson’s mean is the weighted approximation for a log-normal distribution equal to the summation of 30% ofthe 90% probability value, 30% of the 10% probability value, and 40% of the 50% probability value e.g. (P90 x0.3) + (P10 x 0.3) + (P50 x 0.4) equals the Swanson’s mean value286

Australian Geothermal Conference 2010Geothermal energy utilization from conventional hydrothermal resources continues to accelerateand the advent of EGS is expected to rapidly increase growth after 10 to 15 years puttinggeothermal on the path to provide an expected generation global supply of 4.56 EJ per year(~160 GWe) by 2050, and between 7.5 EJ per year (with 1% growth per year) and 32.4 EJ peryear (with 4% growth per year) by 2100.In addition to the widespread deployment of EGS, the practicality of using supercriticaltemperatures and offshore resources is expected to be tested with experimental deployment ofone or both a possibility by 2100.Direct use of geothermal energy for heating and cooling, including geothermal heat pumps(GHPs) is expected to increase to 7.86 EJ /year (~815 GWt) by 2050 and between 12.9 EJ peryear (with 1% growth per year) and 55.9 EJ per year (with 4% growth per year)by 2100.Marketing and multiple internationally competitive supply chains will underpin this growth. Thisexpectation is supported information published by Rybach, 2005Geothermal energy is expected to meet between 2.5% and 4.1% of the total global demand forelectricity by 2050 and potentially more than 10% by 2100. It is also expected to provide about5% of the global demand for heating and cooling in by 2050 and potentially, more than 10% by2100. Geothermal energy will be a dominant source of base-load renewable energy in manycountries in the next century.With its natural thermal storage capacity, geothermal is especially suitable for supplying bothbase-load electric power generation and for fully dispatchable heating and cooling applications inbuildings, and thus is uniquely positioned to play a key role in climate change mitigationstrategies (Bromley, et al., 2010).ACKNOWLEDGEMENTSThe authors thank their international colleagues who have contributed so much of their professionallives and time to provide improved understanding of geothermal systems. We are especially gratefulto: John Lund, Ladsi Rybach, Mike Mongillo, Ken Williamson, David Newell, Trevor Demayo, ArthurLee, Subir Sanyal, Roland Horne, David Blackwell, and Doone Wyborn.REFERENCESAGEG-AGEA, 2009, Australian Code for Reporting of Exploration Results, Geothermal Resources andGeothermal Reserves, prepared by the Prepared by: The Australian Geothermal Code Committee - acommittee of the Australian Geothermal Energy Group (AGEG) and the Australian Geothermal EnergyAssociation (AGEA). Download fromhttp://www.pir.sa.gov.au/geothermal/ageg/geothermal_reporting_codeAuthors (of this report), 2010. Unpublished correspondence leading to tabulations of expectations forgeothermal energy deployment from resources below 3,500 metres in 2020, 2030 and 2050. Bertani,R. and Tester, J. were instrumental in developing these estimates.Bertani, R., 2010. World Update on Geothermal Electric Power Generation 2005-2009. ProceedingsWorld Geothermal Congress 2010, Bali, Indonesia, April 25-30, 2010.Bertani, R., and I. Thain, 2002. Geothermal power generating plant CO 2 emission survey. IGA News,49, pp. 1-3. (ISSN: 0160-7782)Bloomfield, K.K., J.N. Moore, and R.N. Neilson, 2003. Geothermal energy reduces greenhouse gases.Geothermal Resources Council Bulletin, Vol. 32, No. 2, pp. 77-79. (ISSN 01607782)Bromley, C.J., M.A. Mongillo, B. Goldstein, G. Hiriart, R. Bertani, E. Huenges, H. Muraoka, A.Ragnarsson, J. Tester, and V. Zui, 2010. IPCC Renewable Energy Report: the Potential Contributionof Geothermal Energy to Climate Change Mitigation. Proceedings World Geothermal Congress 2010,Bali, Indonesia, April 25-30, 2010.Burgassi, P.D., 1999. Historical Outline of Geothermal Technology in the Larderello Region to theMiddle of the 20th Century. In: Stories from a Heated Earth, R. Cataldi, S. Hodgson, J.W. Lund eds.,Geothermal Resources Council, Sacramento, CA. pp. 195-219. (ISBN: 0934412197)Cataldi, R., 1999. The Year Zero of Geothermics. In: Stories from a Heated Earth, R. Cataldi, S.Hodgson, J.W. Lund eds., Geothermal Resources Council, Sacramento, CA. pp. 7-17. (ISBN:0934412197)287

Australian Geothermal Conference 2010Dickson, M.H., and M. Fanelli, 2003. Geothermal energy: Utilization and technology. Publication ofUNESCO, New York, 205 pp. (ISBN: 9231039156)Electric Power Research Institute (EPRI), 1978. Geothermal energy prospects for the next 50 years.ER-611-SR, Special Report for the World Energy Conference 1978.Frick, S., G. Schröder, and M. Kaltschmitt, 2010. Life cycle analysis of geothermal binary power plantsusing enhanced low temperature reservoirs. Energy, Vol. 35, Issue 5, pp. 2281-2294. (ISSN: 0360-5442)Fridleifsson, I.B., R. Bertani, E. Huenges, J.W. Lund, A. Ragnarsson, and L. Rybach, 2008. ThePossible Role and Contribution of Geothermal Energy to the Mitigation of Climate Change. IPCCScoping Meeting on Renewable Energy Sources, Luebeck, Germany 21-25 January 2008. 36 p.Available at: http://www.ipcc.ch/pdf/supporting-material/proc-renewables-lubeck.pdfHamza, V.1; Cardoso, R.; and Ponte Neto, C., 2008. Spherical harmonic analysis of earth'sconductive heat flow, International Journal of Earth Sciences, Volume 97, Number 2, April 2008, pp.205-226(22), Publisher: SpringerIPCC 2007, Climate Change 2007, Fourth Assessment Report of the Intergovernmental Panel onClimate Change, Working Group III: Mitigation of Climate Change, Chapter 4 - Geothermal, Section4.3.3.4International Energy Agency (IEA), 2008, World Energy Outlook 2008, download fromhttp://www.iea.org/textbase/nppdf/free/2008/weo2008.pdfEuropean Renewable Energy Council (EREC) and Greenpeace International (GPI), 2008, Energy[R]evolution – A sustainable Global Energy Outlook (EREC-GPI-08). Download from:http://www.greenpeace.org/raw/content/international/press/reports/energyrevolutionreport.pdfGoldstein, B.A., Hill, A.J., Long, A., Budd, A. R., Holgate, F., and Malavazos, M., 2009. Hot RockGeothermal Energy Plays in Australia, Proceedings of the 34 th Workshop on Geothermal ReservoirEngineering, Stanford University, Stanford, California, February 9-11, 2009, SGP-TR-187Hiriart, G., R.M. Prol-Ledesma, S. Alcocer and G. Espíndola, 2010. Submarine Geothermics:Hydrothermal Vents and Electricity Generation. Proceedings World Geothermal Congress 2010, Bali,Indonesia, 25-29 April, 2010.Kaltschmitt, M., 2000. Environmental effects of heat provision from geothermal energy in comparisonto other resources of energy. Proceedings World Geothermal Congress 2000, Kyushu-Tohoku, Japan,May 28-June 10, 2000. (ISBN: 0473068117)Krewitt, W., K. Nienhaus, C. Klebmann, C. Capone, E. Stricker, W. Grauss, M. Hoggwijk, N.Supersberger, U. Von Winterfeld, and S. Samadi, 2009. Role and potential of renewable energy andenergy efficiency for global energy supply. Climate Change, 18, December 2009, 344 pp. (ISSN:1862-4359)Lund, J. W., Freeston, D.H. and Boyd, T.L 2010. Direct Utilization of Geothermal Energy 2010Worldwide Review. Proceedings World Geothermal Congress 2010, Bali, Indonesia, 25-30 April 2010.Mongillo, M.A., 2010, Proceedings of the Joint GIA-IGA Workshop - Geothermal Energy GlobalDevelopment Potential and Contribution to Mitigation of Climate Change, 5-6 May 2009, Madrid,Spain. Download from http://www.ieagia.org/documents/ProcGIA_IGAWorkshopMadridFinalprepress22Mar10_000.pdfNill, M., 2004. Die zukünftige Entwicklung von Stromerzeugungstechniken, Eine ökologische Analysevor dem Hintergrund technischer und ökonomischer Zusammenhänge, Fortschritt-Berichte VDI Nr.518. Düsseldorf, D: VDI-Verlag, 346 (in German). (ISSN: 0178-9414)Rybach, L., 2005. The advance of geothermal heat pumps world-wide. IEA Heat Pump CentreNewsletter, 23, pp. 13-18. (ISSN: 0724-7028)Sims, R.E.H., R.N. Schock, A. Adegbululgbe, J. Fenhann, I. Konstantinaviciute, W. Moomaw, H.B.Nimir, B. Schlamadinger, J. Torres-Martínez, C. Turner, Y. Uchiyama, S.J.V. Vuori, N. Wamukonya, X.Zhang, 2007. Energy supply. In Climate Change 2007: Mitigation. Contribution of Working Group III tothe Fourth Assessment Report of the Intergovernmental Panel on Climate Change [B. Metz, O.R.288

Australian Geothermal Conference 2010Davidson, P.R. Bosch, R. Dave, L.A. Meyer (eds)], Cambridge University Press, Cambridge, UnitedKingdom and New York, NY, USA. (ISBN-13: 9780521705981)Tester, J.W., E.M. Drake, M.W. Golay, M.J. Driscoll, and W.A. Peters, 2005. Sustainable Energy –Choosing Among Options, MIT Press, Cambridge, MA, 850 pp.Tester, J.W., B.J. Anderson, A.S. Batchelor, D.D. Blackwell, R. DiPippo, and E.M. Drake (eds.), 2006.The Future of Geothermal Energy Impact of Enhanced Geothermal Systems on the United States inthe 21st century. Prepared by the Massachusetts Institute of Technology, under Idaho NationalLaboratory Subcontract No. 63 00019 for the U.S. Department of Energy, Assistant Secretary forEnergy Efficiency and Renewable Energy, Office of Geothermal Technologies. 358 p. (ISBN-10:0486477711, ISBN-13: 978-0486477718)289

Australian Geothermal Conference 2010New Design Concepts for Natural Draft Dry Cooling TowersHal Gurgenci* and Zhiqiang GuanThe Queensland Geothermal Energy Centre of Excellence (QGECE), the University of Queensland, St Lucia,Brisbane, Qld 4072*Corresponding author: h.gurgenci@uq.edu.auDesign of efficient natural draft dry cooling systemis essential for Australian geothermal power plantapplications. In fact, dry cooling may be the onlyoption for most geothermal power plants plannedto be established in Australia since these areashave limited access to water resource.Natural draft cooling towers in coal-fired andnuclear power plants have been mainly ofconcrete construction. It has been reported thatthe highest concrete cooling tower in the world of200 meters high has been built at the RWE powerstation at Niederaussem (Busch et al, 2002). Dueto the lower thermal efficiencies of geothermalpower plants, the heat rejection per kWh(e) of netgeneration from these plants will be four or moretimes as great as from fossil fuelled plants (Kröger2004), which will require a much larger coolingtower for a geothermal power plant of comparedto a coal-fired power plant of similar capacity. Thedesign and construction of such a large concretecooling tower is extremely expensive and takes along time to build. The tower structure is heavyand requires substantial foundation.The Queensland Geothermal Energy Centre ofExcellence (QGECE) is to explore new conceptsfor natural draft dry cooling technologies forgeothermal power plants. The aims for the newdesign concepts are to increase the performanceof the cooling tower and reduce the overall cost.Two alternative designs are proposed: low costair-lift mobile cooling tower (Gurgenci and Guan,2009) and solar cooling tower.The design concept for air-lift mobile coolingtower is as follow: the tower is built as a flexibleshroud and is held in place in tension by thebuoyancy force provided by a lighter-than-air gas.One benefit for the air-lift cooling tower is that theindustry can construct very tall cooling towers atan acceptable cost. Other benefits includeconstructing such towers with a minimum of timeand expense, which are of relatively light weightand which are not subject to frequent repairs. Thetowers are mobile and can be disassembled andmoved to a new geothermal plant site with aminimum of time and effort.The design concept for solar cooling towerincludes a solar energy system which is used topreheat the air inside the cooling tower toincrease the buoyancy of the natural draft coolingtower. The enhanced buoyancy enable thecooling tower be constructed with either reducedsize or with increased tower performance, whichis especially beneficial during hottest time periodin a day.Keywords: Geothermal energy, Cooling tower andheat exchanger, natural draft coolingtechnologies.Natural Draft Dry Cooling TowersA natural draft dry cooling tower is a heat rejectiondevice that creates the flow of air through thebundles of finned tube heat exchangers by meansof buoyancy effects. Buoyancy occurs due to adifference in air density before and after heatexchanger resulting from temperature andmoisture differences. The greater the thermaldifference and the height of the tower structure,the greater the buoyancy force. Therefore, thevolume flow rate of air across the heat exchangerbundle is directly proportional to the height of thecooling tower.Natural draft dry cooling towers are particularlyattractive as a cost-saving solution for largerpower plants where water resource is limited andexpensive. The unique economic advantage ofnatural draft cooling towers lies in their very lowelectric energy requirement. Fan cost andsubstantial amount of fan power are saved in thissystem. The operating costs are minimal.Natural cooling towers have been designed invarious ways. However, reinforced concretestructures have become standard to handle thehigh cooling water flow rates in large powerplants.Fig.1 shows a configuration of natural draft drycooling tower used by the thermal power plants.Fig.2 is a photo taken from Yangchen powerplant of China.Fig.1 Natural draft dry cooling [GEA Aircooled Systems]290

Australian Geothermal Conference 2010concrete tower used in most coal-fired andnuclear power plants). The tension cables areconnected to the stiffening rings to add strengthand hold the shroud in place by tension. Thesealing skin can be made of light materials andcan be attached to the stiffening rings quickly andeconomically in site. The flexible shroud can havedurability for wind load and weather resistance.Fig.2 Yangchen Natural Draft Cooling TowerNew Design ConceptsThree design concepts are proposed in thispaper. The first concept relies on a conventionalsteel construction design. The second concept,the air-lifted tower, is a radical invention thatproposes the use of helium-filled toroidal rings toprovide the tower structure. The third concept is anovel combination of the solar chimney idea withnatural draft dry cooling towers.Steel Towers with Advanced Heat ExchangersOne of the reasons behind the use of reinforcedconcrete construction as a method of buildingnatural draft dry cooling towers in coal-fired powerplants is the concern about corrosion caused byexposure to some of the elements in the fluegases in combination with the ambient effects.These concerns do not apply to a binarygeothermal plant. Therefore, the use of steeltowers is allowable in geothermal plants and thischoice may offer new options that are notavailable to fossil fuel power plant designers. Acombination of steel construction methods withsome of the advanced heat exchanger optionsthat deliver high heat exchange rate/ pressuredrop ratios could then be a viable option. A steeltower design will be presented that will have thecapability to power air-cooled condensers for acommercial geothermal power plant subject toambient conditions typical of the Australianoutback.Air-lifted natural draft cooling towerIn this design, a flexible shroud is the main bodyof the tower structure and is held in place intension by the buoyancy force. This flexibleshroud includes stiffening rings, sealing skin andtensioning cables as shown in Figure 3. Theshroud can be divided into several sections andprefabricated in factory and assembled/installedquickly in site.The stiffening rings are made of differentdiameters and placed periodically to form thedesired shape required (such as hyperbolicFig.3 Proposed structure flexible shroudTo provide lifting force to the flexible shroud,stacked helium-filled toroidal vessels areconnected directly to stiffening rings separately asindicated in Figure 4 to provide adequate liftingforce. Each inflatable donut can pull either asingle stiffening ring or several rings together.Each inflatable donut can have its own air inletand can be pneumatically controlled. Redundancyis desirable in order to ensure that the towerintegrity can be maintained during themaintenance and to facilitate repair of leakingdonuts.Not all donuts will need helium or lighter than airgas. Normal air can be pumped into some donuts,probably those towards the bottom, withpredefined pressure to provide supporting forcefor the shroud. The donuts can be further dividedinto several sections vertically along the shroud(such as four or eight sections). Each verticalsection can be individual controlled pneumatically.291

Australian Geothermal Conference 2010Fig.4 Design of donut-shaped air-lifted cooling towerSolar enhanced natural draft cooling towerIn this design, based on the principal of solarchimney, a solar collector is placed at the base ofa natural draft cooling tower between the towerand the vertically arranged heat exchanger.Heating of the air behind the heat exchanger(inside the tower) will enhance convection, andhence more airflow through the heat exchanger.The cooling ability of this tower depends primarilyon two factors: the size of the solar collector areaand tower height. With a larger solar collectorarea, a greater volume of air is warmed to flow upthe tower. With a larger tower height, the pressuredifference increases the buoyancy effect.Figure 5 shows schematic drawing of proposeddesign of solar enhanced natural draft coolingtower. In this design, heat exchanger bundles areplaced circumferential at the base of the tower. Asolar collector is added between the heatexchanger bundles and the cooling tower to heatthe air below. Due to the upstream airtemperature of the heat exchanger has beenheated and much higher than the downstream airtemperature, the flow rate of the air through theheat exchanger increase, which results either thetower size can be reduced (for fixed heat sink) orthe performance of the tower increases (for fixedtower size).The cooling tower structure can be a standardconcrete tower used in power plant or it can be achimney shape as shown in figure 5. The air-liftcooling tower can also be used in this design.Fig.5 Schematic of solar natural draft cooling towerAir-Lifted Tower ConstructionTo erect the air-lift cooling tower shown in Figure4, the stiffening rings are placed first around thebases in order and the top stiffening ring isconnected to the first donut shape vessel and thetop end of the tension cables. Helium is thenpumped into the top donut vessel which lifts thefirst stiffening ring and part of the tension cables.After the first ring has been raised to a desiredposition above the ground, the second top ringcan be connected to the tension cables in thesame way. The ascending height of theassembled rings is controlled by the release oftension cables. The sealing skin can be formedand fixed to the already assembled stiffening ringswith various fixing mechanisms.With the same process repeated as above, thecontrolled upward motion of the lifting donutvessels result in a corresponding advance of thecooling tower shroud. The intermediate advanceof the installation of the cooling tower is shown inFigure 6.Once the bottom ring has been assembled andraised to desired position, the tension cables arethen secured to their respective anchors.292

Future WorkFig.6 Intermediate position of installationAn engineering feasibility study on the above newdesign concepts will be carried first based on theenvironmental conditions in the proposedgeothermal power plant sites.Should the feasibility study favour the air-liftcooling tower concepts, a detailed design of suchcooling tower for a planned 50MW geothermalpower plant will be conducted. A 3D model is tobe built to predict tower deformation and itsperformance under various wind conditions.Simulation and modelling of the solar natural draftcooling tower will be carried out by the PhDstudent Zeng Zhang who started his study inMarch 2010.Full economic study needs to be performed forboth conventional concrete cooling tower and thenew design concepts as proposed in this paper.Once the study favours air-lift option, a scaledphysical model will be built and tested in ourmobile plant. Different wind loads will besimulated and the best stabilizing controlledmechanism will be identified. All modelling andsimulating results will also be confirmed by thelaboratory testing results.Materials for the toroidal vessels, sealing skin andstiffening rings of the shroud play important rolesin cost reduction and durability of the tower. It istherefore required to search, identify and testdifferent materials suitable for the shroud andvessels.Optimisation of the area of solar collector and sizeof the tower for solar enhanced natural coolingtower will be carried in future by a PhD student.Stress analysis for tower structure for new designconcepts will be done to determine the strength ofthe structure.Australian Geothermal Conference 2010ConclusionsAir-lift cooling towers provide an alternative to thewidely used concrete cooling tower used in powerindustries. This alternative will achieve aconstruction with minimum of time and expense.Since materials of the tower can be made ofnonferrous and light materials, the resulting towerwill therefore be corrosion free and it will belightweight allowing quick installation with theadvantages of easy disassembly and mobilisationat a minimum of time and effort. The whole towercan be reused in a new site.The inflatable donuts (cells) can be divided intoportions which are so arranged that malfunction ofone or more of them will not interfere withoperation of the tower as a whole. Suchredundancy provides reliable maintenance andrepair.The shroud is made of light material and can befactory manufactured in segments or as a whole.This provides for a quick and easy fieldinstallation. Since there is less material (weight) tobe transported to the construction site, it willreduce the cost of transportation which may havesignificant impact when the site is located in aremote area.The most serious concern on air-lift cooling toweris its deformation and stabilization under strongwind which could affect the performance of thetower. Mechanism for improving design andcontrol will be identified.By adding solar collector to preheat the upstreamair temperature of the heat exchanger, it results amore air flow rate through the heat exchanger dueto a much higher buoyancy effect. This can eitherreduce the tower size (for fixed heat sinking) orincrease the performance of the tower (for fixedtower size).A comparative life cycle cost analysis for bothconventional and the new design concepts ofnatural draft dry cooling tower is necessary infuture work.ReferencesBusch D., Harte R., Kra¨tzig W.B. and Montag U.2002, New natural draft cooling tower of 200 m ofheight, Engineering Structures 24 (2002) 1509–1521.Kröger, D.G. 2004, Air-Cooled Heat Exchangersand Cooling Towers, Penwell Corp., Tulsa, USA.Gurgenci, H., and Guan, Z. 2009, Dirigible naturaldraft cooling tower in geothermal and solar powerplant applications, 14th IAHR Cooling Tower andAir-cooled Heat Exchanger Conference, 1 to 3December 2009, Stellenbosch, South Africa.293

Australian Geothermal Conference 2010QGECE Research on Heat Exchangers and Air-Cooled CondensersKamel HoomanQueensland Geothermal Energy Centre of Excellence,The University of Queensland, Brisbane, Australiak.hooman@uq.edu.auAbstractThe aim of this paper is to give an overview of theresearch conducted at Queensland GeothermalEnergy Centre of Excellence (QGECE) on aircooledcondensers for application in geothermalpower plants. The application of metal foam heatexchangers in Natural Draft Dry Cooling Towers(NDDCTs) has also been studied in detail.Samples of the numerical and experimentalresults are presented. These findings are thenbacked up by theoretical modelling that enablesus to run parametric studies which are sensible todifferent working/environmental conditions.Keywords: Air-cooled condensers, NDDCT, heatexchangers, theoretical, experimental, CFDIntroductionAustralia’s current policy targets an annualrenewable energy generation of 45,000 GWh by2020. Geothermal energy systems have thepotential to produce a base load generationcapacity capable of replacing existing coal-firedplants. However, there are some technicalchallenges to be overcome first. One of the majortechnical difficulties is the cooling system.Although wet cooling is more efficient than drycooling, water shortages and harsh environmentalconditions in areas such as the Australian deserthave forced designers to consider less efficientand more expensive air-cooled systems, or drycoolingas it is often termed.Air-cooled plants offer potential economicadvantages due to plant sitting flexibility. Bothnatural draft and mechanical draft dry-coolingtowers, equipped with air-cooled heat exchangers(almost always with extended airside surfacearea), are used. The fan-driven systems can bebuilt quickly and at relatively low cost but theiroperating costs are higher due to their highermaintenance requirements and the parasiticlosses associated with running the fans. Thecooling system is a significant cost item in thepower plant and affects the performance of theentire power cycle. If the cooling system does notprovide adequate cooling, the overall plantefficiency plunges with serious economicconsequences (e.g. decreased electricityproduction).It has been reported that approximately 0.3 GWh(per year) of electricity generation in the UnitedStates is lost because of cooling towers operatingbelow their design efficiency. This corresponds toan economic penalty of around 20 million USdollars per year [1]. It is therefore, very importantto design and analyse highly efficient dry coolingsystems for power plants.Though air-cooled condensers are not verypopular, their application to power industry is nota new practice. Interestingly, a hyperboloidcooling tower was patented in 1918. The reasonfor the renewed interest in air-cooled condensersis scarcity of water or high humidity [2]. In eithercase fan-cooled systems can lead to highparasitic losses. This makes the NDDCT an idealoption for such cases. The reason for QGECE’sinterest in NDDCT is obvious. Most of ourgeothermal resources are located where there isno water. Besides, the thermal efficiency of ourbinary power plants is already so low that wecannot afford parasitic losses (fans).MethodologyIn view of the above, QGECE has been lookinginto applying efficient heat exchangers to reducethe cost of towers. This would have beenimpossible unless a proper test bed is provided. Indoing so, a 2m height cooling tower is built and isoperational at our QGECE lab, see Figure 1.Figure 1: The QGECE lab scale model for a NDDCT.294

Australian Geothermal Conference 2010The tower has a square cross-section with1.4x1.4 m 2 base area and 0.9167x0.9167 m 2tower exit area. The heating elements are fourelectrically heated copper bars of 10 mmdiameter. At the same time, a numerical model isbuilt, as depicted by Figure 2, so that parametricstudy of the problem is possible [3]fluid flowing in the tubes undergoes phasechange. The same test section is used to collectdata for a finned-tube of 30 mm OD with (197 finsof 0.6 mm thickness per meter). This is a typicaldesign for an air-cooled condenser; similar to theASPEN design one analysed in [3].Figure 2: The computational QGECE lab scale model for aNDDCT.Parallel to these tests and simulations atheoretical model is developed to shed some lighton the scaling of such NDDCTs under no windconditions [4].A new design for tubes in air-cooled condensersis currently being examined at QGECE. The ideais to replace the fins by a layer of metal foams,see Figure 3, to increase the heat transfer areaand, hence, reduce the total tube numbers andheat exchanger size/cost [2]. This can significantlyreduce the cost as an efficient heat exchanger willlead to a less expensive (and shorter) coolingtower.Figure 4: The low speed wind tunnel test facility; the hotwater circulator and data acquisition system (sitting on thedesk).Results and discussionCooling tower resultsExperiments are conducted to evaluate the towerfrictional (shape) resistance, i.e. by excluding theheat exchanger resistances. A total of 25 datapointswere used to measure the air temperatureat the exit from the tower, as depicted by Figure 5.The results are then compared to numerical andtheoretical predictions. The result from thetemperature measurements involves theuncertainty of measurements which is estimatedby using a basic uncertainty analysis [6]. Themeasured air velocity at the exit is 0.543±0.064m/s which is very close to both theoretical (0.515m/s) and numerical (0.515 m/s) predictions.Figure 3: The metal foam-wrapped tube used in QGECE labsto conduct experiments on air-cooled condensers.Theoretical and numerical analysis of such foamwrappedtubes in air-cooled condensers havebeen reported in an earlier study [5]. Thus,experimental analysis followed initial high-leveltheoretical investigations. Figure 4 illustrates aphoto of our test section where air is forced (by asuction fan) to flow over a heated metal foamwrappedtube. The tube surface temperature iskept constant by using a circulation heater thatpumps water (at adjustable flow rates) through thetube to keep the tube surface temperature uniformand constant similar to a condenser where theFigure 5: The location of the thermocouples to measure theair temperature at the exit from the tower.295

Australian Geothermal Conference 2010The following formula [4] has been used to predictthe tower inlet velocity, U, for the empty towerheat transfer to pressure drop ratio) of the foamheat exchanger is higher than that of an ASPENdesigned finned- tube one; see Figure 7.(1)and for the case where heat exchangers introducea resistance to flow(2)with k, g, K, H, t, C F , and Δρ being the frictionalloss coefficient, gravitational acceleration, bundlepermeability, the tower height, the bundlethickness, form drag coefficient, the inlet-outlet airdensity difference, respectively.Figure 6 illustrates the results of numericalsimulations compared to theoretical predictionsfor taller cooling towers where the bundle velocityis plotted versus the tower height on a log-logscale. The 200m tower is designed to dump 283MW of heat for a power plant that generates50MWe [3]. The volumetric heat generation ratesfor shorter towers are then scaled down andimplemented as inputs to our CFD simulations. Asseen, the developed theory in [4] is capable ofpredicting the thermohydraulic performance of aNDDCT under no wind conditions. However, aswind can dramatically affect the performance of aNDDCT [7-12], QGECE is focusing on betterunderstanding of the effects of cross-wind onscaling of the cooling towers.Bundle Velocity (m/s)4.13.83.53.22.92.62.32Theoretical PredictionNumerical (CFD)1.720 50 80 110 140 170 200Tower Height (m)Figure 6: Numerical and theoretical prediction of bundlevelocity for towers of different heights.Metal foam resultsIn our last year report at Australian GeothermalConference, it was numerically shown that themetal foams can be used in air-cooledcondensers with a potential to reduce the size andcost of the condensers [13]. It has been arguedthat the area goodness factor (j/f; dimensionlessFigure 7: Normalized area goodness factor versus porouslayer thickness for different samples and free streamvelocities (samples 2-7 refer to commercially available ERGmetal foam samples).[13]Figure 8 illustrates a sample of our recentlycollected experimental data. This figure showsplot of heat transfer ratio versus the approachingair velocity.Foam/fin heat transfer ratio32.82.62.42.221.81.61.41.211 3 5 7 9 11Face Velocity (m/s)Figure 8: Heat transfer ratio (foam-wrapped tube divided byfinned-tube) versus face velocity. Finned-tube is designed byASPEN where tubes are 30 mm OD with 197 fins per meter.Fin diameter is 60 mm and the fin plate thickness is 0.6 mm;see [3] for more details.Interestingly, with the same pressure drop, theheat transfer from a foam-wrapped tube is always296

Australian Geothermal Conference 2010higher than that of a finned tube. The ratio canfigure out at 2.6 at high velocities which can leadto more compact tube bundles and thus results indecreasing the capital cost. This gain for fancooledsystems can lead to significantly reducedparasitic loses as with new design the number ofbundles is almost halved (so are the parasiticlosses). Furthermore, the pressure drop in the aircooled condenser is reduced as the number of thebundles is almost halved. Another point in favourof such compact systems is the lower tube-sidepressure drops because of the elimination of alarge number of the tubes in the condenser.Concluding remarksAn accurate model, compared to experimentaland CFD results, is developed to predict the heatand fluid flow through NDDCTs. Metal foam heatexchangers are tested and found to be superior tofins for air-cooled condensers. Further research isplanned at QGECE to examine different workingand environmental conditions.References[1] North American Electric ReliabilityCouncil, Generating Availability Data SystemReport, Princeton, NJ, 1986.[2] Hooman K, Gurgenci H. Different heatexchanger options for natural draft cooling towers.World Geothermal Congress Nusa Dua, Bali,Indonesia 2010: Paper # 2633.[3] Hooman K, Gurgenci H. Porous mediummodelling of air-cooled condensers. Transport inPorous Media 84 (2010): 257-273.[4] Hooman K. Dry cooling towers ascondensers for geothermal power plants.International Communications in Heat and MassTransfer 2010;in press.[5] Odabaee M, Hooman K, Gurgenci H.Metal foam heat exchangers for heat transferaugmentation from a cylinder in cross-flow.Transport in Porous Media 2010;in press.[6] Mee DJ. Uncertainty analysis ofConditions in the test section of the T4 shocktunnel, Department Research Report # 1993/4,Division of Mechanical Engineering, TheUniversity of Queensland, Brisbane, Australia,1993.[7] Kroger DG. Air-cooled heat exchangersand cooling towers. Pennwell Corp., 2004.[8] Al-Waked R, Behnia M. The Effect ofWindbreak Walls on the Thermal Performance ofNatural Draft Dry Cooling Towers. Heat TransferEngineering 26(2005): 50-62.[9] Zhai Z, Fu S. Improving cooling efficiencyof dry-cooling towers under cross-wind conditionsby using wind-break methods. Applied ThermalEngineering 26 (2006): 1008-1017.[10] Sabbagh-Yazdi SR, Goudarzi MA,Mastorakis N. Solution of Internal and ExternalWind Induced Pressure for a Single CoolingTower in Kazerun Power Plant (IRAN), in: S.H.Sohrab, H.J. Catrakis, N. Kobasko, S. Necasova,N. Markatos (Eds.), 6th IASME/WSEASInternational Conference on Fluid Mechanics andAerodynamics; 2008 Aug 20–22, World Scientificand Engineering Academy and Society, Rhodes,Greece.[11] Liu BM. Numerical Prediction for thePerformance of a Natural Draft Cooling Towerwith Comparison to the Field Test Data, in: G.F.Hewitt (Ed.), 10th International Heat TransferConference; 1994 Aug 14–18, Institute ofChemical Engineers, Brighton, England.[12] Iovino G, Mattoni H, Mazzocchi L, SenisR, Vanzan R. Optimal Sizing of Natural Draft, DryCooling Towers for ENEL Combined CyclePower-Plants, in: S.S. Stecco, M.J. Moran (Eds.),Florence World Energy Research Symp; 1990May 28–Jun 01, Pergamon Press Ltd, Florence,Italy.[13] Odabaee M, Hooman K, Gurgenci H.Comparing the tube fin heat exchangers to metalfoam heat exchangers. Australian GeothermalEnergy Conference, Brisbane, Australia,November 11-13, 2009.297

Australian Geothermal Conference 2010Equation-Free Summary of the 2010 Mathematics in Industry StudyGroup on Geothermal Data Analysis and OptimizationRobert McKibbin 1 ,*Franklin G. Horowitz 2,3 , Neville Fowkes 4 , Brendan Florio 41 Massey University Institute of Information & Mathematical Sciences2 Western Australian Geothermal Centre of Excellence,The University of Western Australia 3 School of Earth & Environment and 4 School of Mathematics & Statistics,AbstractThe Australia and New Zealand Industrial andApplied Mathematics (ANZIAM) division of theAustralian Mathematical Society (AustMS) held itsannual Mathematics and Statistics-in-IndustryStudy Group (MISG) both at RMIT in Melbourne,and via telepresence at UWA using the accessgrid from the 7 th through the 12 th of February2010. The Geothermal Working Group attracted atotal of 18 academic, professional, post-doctoral,and post-graduate applied mathematicians andothers to study problems of potential interest tothe geothermal projects in the Perth Basin HotSedimentary Aquifer plays. In the interests ofattracting the wider AGEC community’s attentionto the annual MISG efforts, reproduced below isthe “Equation Free Summary” of the week-longintensive effort. The detailed summary paper andseveral flow-on papers are currently beingprepared for publication. The author list of thiscurrent document are comprised of the so-called“Industry Representative”, the “ProjectModerators”, and “Student Moderator”. A full list ofthe participants in the geothermal study group arefound in Appendix 1.Equation-Free SummaryThe aim of this project was to assess theeconomic feasibility of extracting geothermalenergy from the deep sedimentary Perth Basin inWestern Australia. The WA Geothermal Centre ofExcellence (WAGCoE), based in UWA, has beencollecting data from (groundwater and oil)boreholes in the area as well as examiningremotely-sensed observational data of groundsurface temperatures. The question was how tointerpret the available data to gain information onthe modes of thermally-driven convective flows, ifany, in the groundwater aquifers, therebydetermining which parts of the system would bemost economical to use for energy extraction.The 2010 MISG "Geothermal" study groupconcentrated on problems of advective transportof heat in stratified geological units appropriate forthe conditions found in the Perth Basin of WesternAustralia. The main hypothesis examined – forseveral different sub-problems – is that advectivetransport of heat is an important contribution tothe overall heat transport. Measuredtemperatures in the system indicate anapproximate temperature increase through themain aquifers from about 40 ˚C to 80 ˚C. Clearly,conductive heat transport is present (aselsewhere over the Earth's surface), but wedesired to find out whether the system issusceptible to "unforced" or natural convection.There is a well-developed theory in the fluidmechanics/porous media literature about theconditions required for convective instability innon-homogeneous permeable systems that areheated from below. The criterion manifests itselfin a dimensionless combination of thermodynamicand rock matrix parameters known as theRayleigh number, denoted Ra. This was seen asrelevant for our study, so we concentrated onusing the available field data to determinewhether or not conditions in the Perth Basin wouldbe right or not.It is likely there is also an advective transportcomponent due to a hydraulic gradient from arecharge zone assumed to be at an elevatedheight in the Darling escarpment to the east,westwards to a discharge zone assumed to beoffshore. However, for the purposes of most ofour sub-problems, we neglected the hydraulicgradient effects.The geological backgroundRocks of the Perth Basin consist of multiple layersof sedimentary strata, where matrix (as opposedto fracture) hydraulic permeability is expected tobe dominant. Immediately below the groundsurface, there is an unconfined aquifer consistingof rocks geologically known as the "superficial"formations. These rocks are underlain by variousshallow confined aquifers consisting of rocks ofthe Kings Park Formation (locally), the LeedervilleFormation, and the South Perth Formation. Forthe study group's convenience, since these rocksare not in hydraulic communication with thedeeper and hotter aquifers, we lumped all of theseaquifers together into a group we labelled(perhaps confusingly to people familiar with thestratigraphy) as the "superficial" units andbasically ignored them (other than the boundaryconditions they provide) in our models.At a depth assumed for our purposes to be about600 m we find the top of a confined aquiferconsisting of rocks from the Gage Formation.Those rocks are assumed to be about 200 mthick, with uniform physical properties. Below theGage, but in hydraulic communication with it, wefind rocks of the Yarragadee Formation. Thecontact between the two formations is anerosional unconformity, so the thicknesses of thetwo units vary with map position. The rocks of theYarragadee are assumed to be about 1000 mthick, also with uniform physical properties.298

Australian Geothermal Conference 2010(These assumptions of uniform rock propertieswere relaxed in various ways in several of oursub-problems.) Below the Yarragadee lie therocks of the Cadda Formation, about 600 m thick,also in hydraulic communication with theYarragadee. Indeed, in the groundwatercommunity, rocks of all three formations (Gage,Yarragadee, and Cadda) are often called the"Yarragadee Aquifer".Below the Cadda Formation lie rocks of theCattamarra Coal Measures. These rocks aredominantly shales, with interbedded sandstonelayers. Hence, we treat the basal contact of theCadda with the Cattamarra as being a hydraulicseal.The result is our basic model of three stratifiedlayers, with physical properties internally uniform,all participating in a single hydraulically connectedsystem.Well Log constraintsThe closest deep borehole to the sites ofWAGCoE's proposed exploitation systems is theCockburn Number 1 oil well. Cockburn 1 wasdrilled in 1967, to a total depth of 10,020 feet(3054 m), with cores recovered from severaldiscrete sites within the bore. Results oflaboratory measurements of porosity, permeabilityand other physical parameters from these coresare found in the well completion report. Such welllogs have been used to try to determine a relationbetween porosity and permeability from the coremeasurements; it yielded two different estimatesof the vertical distribution of permeability.Statistics from these estimates were used in thesub-problem concerned with estimating Rayleighnumbers and critical values thereof for our PerthBasin problem.Mathematical modellingThe usual modelling paradigm for such problemsis to use a spatially averaged continuumapproximation for the rock matrix and fluidcomponents of the system. Standard fluid mass,fluid momentum (Darcy's law in this case) andthermal energy conservation laws enable a set ofdifferential equations to be formulated. Withappropriate geometric and thermal boundaryconditions, the mathematical problem may besolved; usually this has to be completed usingnumerical techniques since the equations and thethermodynamic parameters are non-linear.However, with simplified geometry and certainassumptions, the mathematical problem istractable using analytic methods.Some numbersTo get an idea of the times and fluid speedsinvolved, we estimated that a water particledescending into the aquifer in the Darling Rangewould flow underground to the west at a speed ofabout 3.5 m year –1 , and take about 8,500 years tocomplete the 30 km journey.If "waste" fluid from a geothermal energyinstallation was pumped into the aquifer at a rateof 100 l s –1 , after 1 year the injected fluid wouldhave penetrated to a radius of about 100 m, withthe fluid speed at the cylindrical front being about50 m year –1 at that time. This speed issignificantly greater than the estimated naturalflow above, so the injected fluid would be likely topenetrate the aquifers "upstream" against thenatural slow current.Convective instabilitySome effort was spent in attempting to determinewhether the system was prone to convectiveinstability; this would, according to theory andexperiment, result in large-scale convective "rolls"of fluid motion that sought to augment theconductive heat transport. A series of alignedrolls, each in counter-rotation to its neighbours,would induce hot spots or alignments near theground surface, as well as cooler areas near thedownflows. A consequent question was: Couldthe associated perturbed isotherm patterns andnon-uniform surface heat flux actually bemeasured? Even if there was convection, howwould we know the pattern?Figure 1 shows a simple idealized model of aporous slab that has a base maintained at auniform constant temperature that is higher thanthat at the top. The hotter fluid near the bottom isless dense than the cooler fluid near the top.Provided the net buoyancy force can overcomeviscous resistance everywhere, convective motionis induced. The most unstable mode is a set oflongitudinal "rolls".Figure 1: Idealized set-up for a permeable slab subjected toa vertical temperature gradient.Linear perturbation theory for thermal convectionin this idealized porous medium that is laterallyunconfined determines that, for a uniform matrix,the critical value of the Rayleigh number is Ra crit =4 2 , with the most unstable mode being horizontalrolls with square cross-sectional boundaries. Ifthe temperature gradient in the system is toosmall, so that Ra has a sub-critical value, then theheat transfer mechanism is conductive only.Above critical, heat transfer is by conduction andconvection. The simplified theory assumes a299

Australian Geothermal Conference 2010linearization of thermodynamic parametersaround a certain temperature.When the isothermal bounding surface above is atground level some distance above the porousregion, the above results are modified. Theisotherms near the top of the convection regionare then not horizontal. This is shownschematically in Figure 2, which also indicates thegeneral form of the convective rolls.Figure 2: Schematic of longitudinal rolls in the permeableaquifer beneath a confining stratum, with an exampleperturbed isotherm near the top of the convection region.Associated with supercritical conditions is anenhancement of the heat transfer. This increaseswith temperature gradient; the general trend of theso-called Nusselt number Nu, defined as the ratioof heat transferred by conduction plus convectionto that transferred by conduction alone, is shownin Figure 3.Figure 3: Trend of Nusselt number Nu with Rayleighnumber Ra; compilation of results from various theoreticaland experimental studies (from Nield & Bejan, 2006).First we considered a homogenized matrix systemwhere the rock parameters were assumeduniform with somewhat arbitrarily-averagedvalues from the borehole core values. Thecalculated values of Ra exceeded the criticalvalue only for high enough (and perhapsunrealistic) values of the datum. However, thelack of euphoria was briefly tempered by thethought that the layering of the system mayinfluence the calculation. It was furtherdetermined from the theory that anisotropy ofthermal conductivity and permeability of the matrixproduces an estimate of the critical Rayleighnumber that is different from that for ahomogeneous system.Anisotropy in the thermal conductivity theoreticallyincreases the critical Rayleigh number; however,the conductivity of rocks varies very little, so theanisotropy in that parameter is negligibly differentfrom unity (measured data gave a value of around1.06). Anisotropy in permeability is induced bymaterial layering. Geological strata necessarilyinduce a permeability that is averagely greater inthe bedding plane than perpendicular to it. Whatis more, such anisotropy generally reduces the(theoretical) critical Rayleigh number!Amidst some excitement, an assessment of thisanisotropy, from a rather complicated set of rockcore measurements and well logs, was made as astratum-thickness weighted average of intrinsicrock permeabilities, with the result that thehorizontal value is about 4.6 times that in thevertical. These estimates reduced the criticalRayleigh number to 21.6, about half thatcomputed from the homogenized parameters.Immediately, it became possible that conditionsfor convection may be met, even using thelinearised theory with the lowest datumtemperature (40 ˚C) for the linearised calculations.Convective rolls and heat transferIn optimistic anticipation of such a result, a smallgroup had been working on how enhanced heattransfer in the Yaragadee Aquifer would manifestitself in the temperature distribution in the nearsurfacesuperficial formations. This is a fairlystandard problem in heat transfer, and quickprogress was made. However, it was shown thatthere would be a very small signature at groundsurface level, and instrumentation would have tobe sensitive and widespread to obtain significantmeasurements. So, this encapsulated theproblem overall: Even if there was a convectiveprocess at work, how would measurements beable to be made to detect the upflow anddownflow regions? So, how could advantage betaken of the convection even if the Ra estimateswere correct?The UWA GroupiesOver in WA, the small "at home" group wasworking; with the assumption that convection wasoccurring, they trawled the literature forinformation about theoretical work on layeringeffects on cell-size, and the stability of suchconvection cells. The idea was that significant300

Australian Geothermal Conference 2010sub-layering caused by thick differentlypermeablestrata might produce vertically-alignedbut largely separated co-rotating cells that, intandem, would advect heat from the hot base tothe surface. At the end of the week, no firmconclusions had been able to be made.SimulationDriven by concern that the thermodynamicproperties of water were being approximated tooseverely in the linear stability theory, a numericalsimulation of 2-D rolls was made using acommercial computer package. This allowedmore accurate values for the water's temperaturedependentdensity, viscosity, etc. to be used.While the simulations were completed post-MISG,it turned out that there were only small qualitativedifferences in the flow patterns.Slope-induced thermal convectionA late idea resulted from current work by one ofthe group members; it considers the possibilitythat longitudinal convection may be induced in asloping aquifer by a near-vertical temperaturegradient. The warmer fluid near the bottom wouldmove up-slope (in this case, towards the east)while the returning cooler fluid near the top movesdown-slope. It was not clear how this would affectthe possible exploitation, because thetemperature profile would remain conductive, andonly be altered by convective rolls as werealready considered above.A small group of postgraduate students from thegroup seized upon some aspects of this, and theirefforts will be rewarded by co-authorship of apaper to be submitted later in the year. A moregeneral case will be considered, where inducedconvection in a layered sloping aquifer isinvestigated for optimal net convective heattransport by such a mechanism, as well as thestability of such motion to longitudinal rolls. Theenthusiasm to participate in this way reinforcesthe benefits of MISG to engagement of postgradsin mathematical enquiry into useful applications.SummaryThe problem was tractable, but depended onsuitable estimates of rock matrix properties fromdrill cores and well logs which were rathercomplex in structure. It also depended on goodestimates of aquifer temperatures; the deepervalues were elusive, and had to be deduced fromextrapolations of shallow well measurements.However, by the end of the study week, there wasa good increase of understanding of the problem,the issues to be resolved, and possiblemechanisms at work in the Perth Basin.Geothermal systems are complicated geologicalgeophysical-thermodynamicalentities. A multidisciplinaryapproach to understanding them isnecessary. However, the quantification of theirattributes is well-handled by mathematically-ablescientists and the MISG proved a suitable venueto tackle the WAGCoE problem.Keywords: Convection, Hot SedimentaryAquifers, AnisotropyReferenceNield, D. A., & Bejan, A. (2006). Convection inPorous Media . Springer, 3rd ed.AppendixParticipants in the 2010 MISG Geothermal StudyGroup (alphabetically by surname):Nurul Syaza Abdul Latif, Massey University NZManarah Eraki, (Affiliation Missing)Brendan Florio, University of Western AustraliaNeville Fowkes, University of Western AustraliaLuke Fullard, Massey University, NZTony Gibb, Consultant, South AustraliaNick Hale, Oxford University, UKDes Hill, University of Western AustraliaFrank Horowitz, University of Western AustraliaGrant Keady, University of Western AustraliaThiansiri (Pat) Luangwilai, ADFA, AustraliaFleur McDonald, Massey University, NZRobert McKibbin, Massey University, NZMelanie Roberts, University of Western AustraliaThomas Stemler, University of Western AustraliaRob Style, Oxford University, UKChristian Thomas, University of Western AustraliaNicole Walters, Victoria University of Wellington,NZ301

Australian Geothermal Conference 2010Status of Compliance With The Geothermal Reporting CodeJim Lawless*, Graeme Beardsmore, Malcolm Ward, Fiona Holgate and Graham Jeffress*SKM PO Box 9806 Newmarket, Auckland, NZJLawless@skm.co.nzAbstractThe Australian Geothermal Reporting Code hasnow been in operation since late 2008. The Codecurrently operates on a voluntary basis and as yethas no legal status or sanctions. However, in late2008 AGEA members agreed that its use inPublic Reports would be mandatory for them anda Compliance Subcommittee (currentlycomprising the above authors) was established.The Subcommittee has since reviewed all PublicReports that we have become aware of to checkfor compliance. This paper presents an analysisof instances of non-compliance broken intoseveral categories but without identification of thecompanies concerned. In each case thecompany concerned was informally notified aboutthe issue. In general there has been a high levelof willingness to comply with the Code andcompanies have responded positively to anyissues raised. Most instances of non-compliancehave been of a procedural nature such as omittingto include Competent Person statements in publicpresentations which include resource estimates.The experience gained has been valuable andhas been taken into account in drawing up theSecond Edition of the Code and its accompanyingLexicon. Other jurisdictions have been studyingthe Australian Code which has led to an almostidentical Code being established in Canada. It iscurrently understood that consideration is alsobeing given to equivalent Codes in the USA andEurope. The benefits of the Australian Code arebeing appreciated even in countries where it hasno formal standing and Code compliant reportshave recently been produced on resources inChile and Indonesia.Keywords: Geothermal, resources, reserves,compliance, Australiapaper presents an analysis of instances of noncompliance,broken into several categories, butwithout identification of the companies concerned.AnalysisMost instances of non-compliance have been of aprocedural nature such as omitting to includeCompetent Person (CP) statements in publicpresentations which include resource estimates.It is safe to say that there has been no case of anapparent intention to issue misleadinginformation, though a few statements have vergedon the over-enthusiastic at times.Figure 1 shows the number of reports reviewedand number of companies concerned againsttime. As is to be expected, the number hasincreased with time but has fluctuated dependingon the quarterly and annual reporting cycle aswell as the frequency of reviews. The latestreview has shown a drop in the number of reportsand several companies which have previouslyreported have not done so, which perhaps reflectsthe generally static state of the industry at thistime.IntroductionThe Australian Geothermal Reporting Code hasnow been in operation since late 2008. The Codecurrently operates on a voluntary basis and as yethas no legal status. However, in late 2008 AGEAmembers agreed that its use in Public Reportswould be mandatory for them and a ComplianceSubcommittee (currently comprising the aboveauthors) was established under the JointAustralian Geothermal Reporting CodeCommittee (JAGRCC). The Subcommittee hassince reviewed all Public Reports that we havebecome aware of to check for compliance. ThisFigure 1: Reports reviewed and number of companies withtimeFigure 2 show the percentage of reports whichhave been considered fully compliant. Althoughthere is still significant room for improvement, it ispleasing to see this increasing with time.Furthermore, a simple percentage does not reflectthat the initial instances of non-compliance weregenerally more significant and those recently havegenerally been of a less serious nature. Thecompanies have clearly embraced the reportingprocess and learned from it. It would also be fair302

Australian Geothermal Conference 2010to say that the Compliance Sub-committee hasalso refined their understanding and criteria overtime.investors, is a Public Report under the Code, sothe fundamental principle of requiring a CPstatement has been adhered to. In reaching thisconclusion the JAGRCC has adhered to astanding practice of consistency with the Joint OreReserve Committee (JORC) Code for mineralsresource statements on which the GeothermalCode was originally modelled.Wrong terminology. This has largely been amatter of companies coming to terms with theresource and reserve categories, and there arenow few problems with the latest reports.Figure 2: Percentage of fully compliant reports with timeInadequate detail. This has been a muchdebatedissue, as there is a tension betweenproviding so much information as to make reportstoo complex for the average investor while at thesame time fulfilling the requirements formateriality. The issue has been dealt with in partby allowing relatively brief Public Reports, but withmore detailed back up being available. A numberof example reports have also been produced toprovide guidance.No reference to Code. This is an issue that hasdecreased with time, although there has been asteady stream of new companies coming on tothe market who have had to be informed aboutthe Code. Responses have been positive.Figure 3: Break down of non-compliance by typeIn Figure 3, the instance of non- compliance arebroken down by category. Discussing each ofthese in turn:No (or incorrect) Competent Personstatement. This has generally speaking been themost common issue and one that has provensomewhat intractable with time in that instancesare still occurring. The usual situation is that acompliant resource estimate has been preparedwhich includes a CP statement, but then that issummarised down to a brief ASX announcement,conference or shareholder presentation in whichthe CP statement is omitted.It is noteworthy that in the Canadian geothermalreporting Code, which in most instances is basedon and identical to the Australian Code, strictcompliance in terms of including a CP statementin every Public Report is not required. However,this issue has been debated by the JAGRCC, andit is considered that every public announcementwhich refers to reserves or resources, and whichcould be deemed to provide information toMiscellaneous issues include items such asinappropriate aggregation of resources andconversion of heat-in-place to barrels of oilequivalent, delivered electricity or numbers ofhouseholds that could potentially be supplied,without stating the assumptions involved. Therehas sometimes been a degree of over-enthusiasmin making such claims. The Code is notprescriptive as to methodology in such instancesbut they remain matters of concern.ReactionIn each case on perceived non-compliance thecompany concerned has been informally notifiedabout the issue. In general there has been a highlevel of willingness to comply with the Code andcompanies have responded positively to anyissues raised.The Way ForwardThe experience gained in reviewing reports hasbeen valuable and has been taken into account indrawing up the Second Edition of the Code andLexicon which, at the time of writing, have beenfinalised and which should be released at thisconference. Other jurisdictions have beenstudying the Australian Code which has led to an303

Australian Geothermal Conference 2010almost identical Code being established inCanada, and it is understood that consideration isalso being given to Codes in the USA andEurope. The benefits of the Australian Code arebeing appreciated even in counties where it hasno formal standing and Code compliant reportshave recently been produced on resources inChile and Indonesia, as is described in acompanion paper.304

Australian Geothermal Conference 2010Application of the Australian Geothermal Reporting Code to“Conventional” Geothermal ProjectsJim Lawless*, Zim Aunzo and Jose Seastres, JrAbstractThe Geothermal Reserves and ResourcesReporting Code was developed in Australia withthe intention of it being principally applied toAustralian geothermal projects, though it wasframed in such a way as to apply to geothermalprojects of all types. Recently, SKM has appliedthe Code to a number of “conventional” hightemperature magmatic-related geothermalprojects outside Australia. This paper describeshow the Code principles and the underlyingresource estimation methodology have beenapplied in those cases. Some of the major issueswhich had to be addressed included:• What conversion efficiencies are applicablewhen the project is to include a flash steamcycle ?• How are resources and reserves to becategorised when estimates are based onnumerical simulation models ?• How is energy already extracted to be dealtwith in the case of project with an existingproduction history ?• How is energy remaining after the assumedproject life to be categorised ? Althoughthese projects have been based onmagmatic-related resources which do notoccur in Australia, some of the insightsgained in the process will have application asAustralian projects move towardscommercialisation, and as Australian listedcompanies develop this type of systemelsewhere in the world.*SKM PO Box 9806 Newmarket, Auckland, NZJLawless@skm.co.nz305

Australian Geothermal Conference 2010Small Australian seismic events and their effectsDavid N. Love 1* , Gary Gibson 2 , Russell Cuthbertson 3 and Wayne Peck 21 PIRSA, GPO Box 1671, Adelaide SA 50012 ES&S, 8 River St Richmond Vic 31213 ES&S, 81 Bishop St Kelvin Grove Qld 4059* Corresponding author: david.love@sa.gov.auWe review four small earthquakes in Australia andhighlight various seismological aspects that arerelevant when considering induced seismicity.The four events are in the magnitude range 2.5 to4.1, a range that is likely to be of interest for thoseinvestigating geothermally induced events.Keywords: Australia, earthquake, induced event,intensityMt Barker (South Australia) event16 April 2010 Magnitude 3.8This earthquake occurred at 11:27pm on Fridaynight, about 25 km from the Adelaide CBD. Itwas widely felt, with a massive response to policeand radio stations. There was no indication thatthe phone system failed, or that power outagesoccurred. The Primary Industries and ResourcesSA (PIRSA) earthquake network operatedsmoothly through the event, giving a very goodlocation and magnitude estimate in less than 3minutes. It took about 15 minutes to quicklyreview the event, and report it to the emergencyservices. The location was not entered on thePIRSA website until after the weekend. The firstepicentre entry on the US Geological Surveywebsite was about 100 km in error, but with alisted error estimate of 152 km. The firstGeoscience Australia (GA) magnitude estimatewas 3.2. While the internet remained operating,the GA and PIRSA websites were overloaded fora time. The PIRSA website received about13,000 hits, a record. Hundreds of theseemanated from overseas.ABC radio Adelaide is normally run from Sydneylate at night, but the Adelaide station wasreopened in response to overwhelming publicinterest. Twitter and Facebook registered a hugeamount of traffic. AdelaideNow website hadabout 1400 comments in short time; USGS DidYou Feel It (DYFI) website plotted over 300replies in less than one hour, and ES&S and GAhad many replies to their website questionnaire.Management, particularly those associated withemergency response, expressed frustration at notknowing where to go for the best information atthe eraly stages. In the following few weeks, alarge number of insurance claims were lodged,with an estimate of 1,000 by one loss adjuster.PIRSA circulated a questionnaire, but responseshave not yet been analysed. Of the responses,only the USGS Did You Feel It system wasavailable within a short time frame. As theearthquake was not in the US, the locationinformation was limited to 6 ‘city’ locations,instead of postcode areas, rendering it of limitedvalue.There is a clear need for a DYFI system to beoperated within Australia which can moreeffectively utilise location and intensityinformation, and disseminate it rapidly.The depth of this earthquake was 25 km, with anuncertainty of about 2 km. This is quite deep forAustralia. Normally depths of hypocentres arenot available due to poor network coverage,however the new Adelaide network is producingaccurate depths for a moderate number of events.The magnitudes for stations of the Adelaidenetwork ranged from 3.2 to 4.1, for stations of theGA network 3.5 to 4.2, and for stations of theES&S network 3.4 to 4.0. These magnitudeswere from peak amplitude measurements ofseismographs, not moment magnitudes. Verylittle has been published on magnitudes inAustralia in the last decade. Very few studieshave been done to investigate corrections forindividual sites on the basis of local geology, or tosee how well magnitudes compare betweennetworks.The Eugowra (New South Wales) event21 August 1994 Magnitude 4.1This was the largest of a swarm of hundreds ofevents that occurred near the town of Eugowra,east of Orange, in country NSW. The swarmwas very shallow, mostly less than 1 km deep.(Gibson et al, 1994)The number of events, initially increasing in sizeresulted in public meetings being called. Therewas some concern, but generally not great. Themain shock did stop the serious drinkers in thehotel for 5 minutes, but damage was limited tominor cracking and contents.Portable instrumentation was installed close to theactivity at an early stage in the swarm. From abeginning of one instrument, there was anincrease eventually to 8 instruments, atmoderately high sample rates of 200 to 400 sps.This resulted in probably the most accurately306

Australian Geothermal Conference 2010recorded sequence of events in Australianseismological history. It was possible todemonstrate the fault plane that accounted formost of the events in the sequence (figure 1).This, in conjunction with the local topography,gave a convincing geological story. It was alsopossible to roughly estimate the size of the mainshock at about 1 km square from the accuratelocation of the shocks in the following 9 days(figure 2). Aftershocks in the following 9 monthswere spread out still further. No other earthquakeunder magnitude 5.5 has been monitored wellenough to show the size, dip and strike of thefaulted area. No fault mechanism was producedfor any of the events, but the geologicalconclusion is that it was mainly thrust.Figure 2: Hypocentres of the Eugowra earthquake swarm.The 22-30 August hypocentres outline an area ofapproximately 1 square km, which was probably the rupturesurface of the largest event, magnitude 4.1.Figure 1: Hypocentres of the Eugowra earthquake swarmshowing a shallow fault rupture clearly dipping to the northwest.One accelerograph recorded a peak value of0.97g. This is the largest acceleration recordedto date from an earthquake in Australia. Thefrequency of this peak was above 50 Hz.Unfortunately the sampling rate on the instrumentwas insufficient to calculate a velocity. Thedamage was very minor, demonstrating thatacceleration is definitely not a good parameter touse when considering a management plan forinduced events. Velocity is a much betterpredictor of damage, although it does not solve allproblems. Detailed intensity information was notprocessed for this event, as a large effort was putinto monitoring.Acacia Ridge (Queensland) event18 July 1996 Magnitude 2.7This earthquake shows that even small events,which often may not be felt in other countries, arestill noticed in Australia. This earthquakeoccurred at 3:40pm in the suburbs of Brisbane,only 13km from the central business district. Itinitially produced only a small number of reports,however press reports and a phone surveycollected over 150 intensity reports (Lynam andCutherbertson 1997). The general radius ofperceptibility of the main felt area was about 14km with another felt area to the south-west and anisolated report as far as 70km away. Intensitiesassigned were low, with no reports reachingintensity 5 (figure 3).The higher intensities were concentrated aroundthe epicentre, and in a separate area to the southwest.The intervening area had lower intensitiesand even not felt replies. Variations in theintensities showed some correlation with surficialgeology; lower intensities were associated withareas of Mesozoic sandstones and coal measuresand Cainozoic sandstones and basalts, whilehigher intensities occurred on the Palaeozoicbasement rocks. This observation, somewhatcontrary to what is usually observed, is thought tobe due to the high frequency nature of thisrelatively small magnitude event. Attenuation ofthe high frequency energy in the Mesozoic andCainozoic rocks may have been more significant307

Australian Geothermal Conference 2010than the amplifying effects normally associatedwith surface sediments and lower frequencies oflarger events. Many smaller events are noticedmore by the noise, emanating from very highfrequencies, than the vibration.While there were nine seismographs within 90km, the distribution of these meant that there is anuncertainty of several kilometres in the epicentre(particularly in a NE-SW direction) and the depthis unknown. This is common for most events inAustralia. This lack of hypocentral depths meansthat there are limited data available to compareintensities with hypocentral distances at the closedistances that are of interest to the geothermalindustry.The good seismograph distribution meant that theepicentre and depth were known to within about 2km. This is fairly unusual for Australia. It hasthe benefit that any measured velocities andestimated intensities can be included in morespecific attenuation analyses for smallearthquakes at close range. Peak amplitudes forthe four closest seismographs were in the 10 to25 Hz range. Individual station magnitudesvaried from 2.0 to 3.4.Figure 3: Isoseismal map of the Acacia Ridge earthquake.Source: Geoscience AustraliaCaulfield (Melbourne) event22 October 2006 Magnitude 2.5An even smaller event was widely felt inMelbourne at 10:36pm while most people werequietly at home.Information for this isoseismal map (figure 4) wascollected by the intensity form on the ES&Swebsite. As a result, in contrast to the AcaciaRidge earthquake, it has only one not felt reply.There are only two MM5 intensity reports listed,again showing the expected low intensities.Despite the very low magnitude of the event, andthe low intensities reported to the ES&S website,there were many insurance claims made,primarily for cracked plaster, cracked and liftedtiles and other cosmetic damage.Figure 4: Isoseismal map of the Caulfield earthquakeConclusionsQuite small events are felt, heard and newsworthyin Australia, even though the real damageconsequences are minor or non-existent. Whilethere is a considerable amount of information ofinterest and value available to the geothermalindustry, some extra effort could improve the datacollection and interpretation to more suitablyaddress the needs of the industry.Information that is accurate, and quickly availablesaves the difficulty and complaints that comelater, from the inability of management and thepublic to find the information when desired.308

Australian Geothermal Conference 2010Given the large public response, even for smallevents, there is clearly a need for a goodcommunity consultation program which addressesperceived risk as well as the risk of damage.A ‘Did You Feel It’ intensity collection system, runin Australia would significantly improve thecollection of intensity data, particularly over theimportant populated areas, and make it rapidlyavailable in easily understood form. This will leadto improved attenuation formulae. The projectshould be easy and inexpensive. Improvedintensity data will give a much clearer indication ofthe level of induced seismicity that may betolerated, and the level of public response tovibrations.There are currently New Generation Attenuation(NGA) formulae in use in earthquake hazardassessment. (Power et al, 2008) Thedevelopment of modified NGA formulae to coverthe magnitude range 3 to 5 would also providevaluable tools for use in induced seismicity riskassessments. The current NGA formulae aredesigned for higher magnitudes and lowerfrequencies, and not suitable to be extrapolatedbeyond these ranges. There is a vast amount ofrecorded information for smaller earthquakes.Unfortunately these are nearly all from overseasevents. Events from stable continental areas,especially Australia would be needed, particularlyfor better handling of high frequencies.More detailed monitoring, particularly active andpopulated areas would lead to improved sourceinformation, and more local data for modified NGAformulae. Unfortunately this is a more expensiveexercise, and most populated areas are notmonitored to this extent. More portabledeployments of larger scale to measure swarmsor aftershocks at close range, and with highersample rates, would provide valuable informationon velocities, accelerations, rupture sizes andfocal mechanisms of use to the industry. This isnot done as often now as it was in the past.The collection of intensity data in the samelocalities as velocity data will further improve theestimation of effects and damage in populatedareas.There is a need to compare magnitude dataacross Australia, review magnitude formulaecurrently in use, begin usage of momentmagnitudes, and investigate seismograph sitemagnitude corrections. This is not difficult, buthas not been happening over the last decade,with limited interaction between seismologicalobservatories.ReferencesGibson, G., Wesson, V. and Jones, T., 1994, TheEugowra NSW Earthquake Swarm of 1994,Seminar Proceedings, Australian EarthquakeEngineering Society pp71 - 80Lynam, C. and Cuthberton, R., 1997, The AcaciaRidge (Brisbane) earthquake of 18 July 1996,Queensland University Advanced Centre forEarthquake Studies, Report no 2, pp71 - 81.Power, M., Chiou, B., Abrahamson, N.,Bozorgnia, Y., Shantz, T. and Roblee, C., 2008,An overview of the NGA Project 3, EarthquakeSpectra v1 n4, pp 3 - 22309

Australian Geothermal Conference 2010Community Consultation Best Practice ManualSteve Kennedy 2 , Chris Matthews 1 *, Damia Ettakadoumi 2 , Cynthia Crowe 21 Australian Geothermal Energy Association. PO Box 1048 Flagstaff Hill SA 51592 Department of Primary Industries GPO Box 4440 Melbourne VIC 3001The Australian Geothermal Energy GroupTechnical Interest Group 1 (AGEG TIG 1) and theAustralian Geothermal Energy Association(AGEA) have collaborated on a project to producea best practice manual (BPM) on communityconsultation for the development of geothermalprojects in Australia.There is a growing expectation from thecommunity, government regulators and leadingindustry players that stakeholders and the widercommunity will be fully informed and encouragedto provide input to decision making about theindustry activities that have the potential to affecttheir communities.AGEG and AGEA undertook the project to createthe BPM with the following objectives: Companies meet government regulationsand expectations Companies meet community expectationsand maintain the ‘social licence to operate’ All parties follow the Australian MinisterialCouncil for Minerals and PetroleumResources (MCMPR) Principles forengagement with communities andstakeholders Give clear advice to developers on theintricacies (such as timing) of reporting toboth the community and corporate regulatorssuch as the ASX.When scoping this project, AGEG and AGEArecognised that there is a significant body of workthat already exists on best practice communityengagement, both in the Australian resourcesindustry and in the international community.It is for this reason that AGEG and AGEA decidedto create tailored “best practice notes” that drawon the existing body of knowledge and provideguidance in the context of the emerginggeothermal energy sector in Australia.The “notes” format (rather than a single manual)means that each theme of communityengagement is a stand-alone document.Companies can pick and choose the topics thatare useful or relevant to them at the time. As thegeothermal industry matures and communityengagement principles change over time, thenotes can be individually modified without havingto update an entire manual.The best practice notes series is a precursor to amore complete BPM on community engagement,to be completed by AGEG and AGEA in the nearfuture.Keywords: Community and Stakeholderengagement, best practice manual, keymessages, communicating risks, mediamanagement, outrage management, geothermalenergy, Australia, South Australian ResourcesIndustry Code of Practice for Stakeholder andCommunity Engagement, Australian GeothermalEnergy Group, Australian Geothermal EnergyAssociation.Context: The need for communityconsultationThe emerging geothermal industry as a whole(including government regulators, developers andstakeholders) understands the need to establishcommunity understanding and support,sometimes known as the “social licence tooperate” (SLO). Variation in Australia’spopulation density and land use across thecountry means that the SLO will be more difficultto obtain in densely populated areas such asVictoria and New South Wales, compared withsparsely populated areas such as far northernSouth Australia. Failure to achieve the SLO maylead to delays in approving the project, result inexpensive or onerous conditions or, in somecases, may even lead to the rejection of aproposal by regulators.The regulatory process is likely to proceed moresmoothly and expeditiously if effective stakeholderand community engagement has, as far aspractical, addressed stakeholder concerns prior toapplying for the required approvals (SACOME2009).The best practice manual processSeveral organisations have worked over the lastfew years to establish a standard of practice thatmeets the requirements of best practicecommunity consultation in resource development.A number of benchmarking documents alreadyexist that outline standards for stakeholder andcommunity engagement that communities canexpect from geothermal developers. Theseinclude documents published by the MineralsCouncil of Australia (e.g. MCA 2005), the MCMPR(MCMPR 2005) and the International Associationfor Public Participation (e.g. IAP2 2007).400

A number of state government departments alsohave manuals on community engagement. Theseinclude the Department of Primary IndustriesVictoria (DPI Victoria 2010) and the WesternAustralian Government (WA Department ofIndigenous Affairs 2005).Drawing on and complying with these and otherdocuments, the South Australian Chamber ofMinerals and Energy (SACOME), in collaborationwith Community Engagement Group Australia(cega) and Primary Industry and Resources SA(PIRSA) produced a Code of Practice forStakeholder and Community Engagement in theSA resources industry (SACOME 2009). ASupporting Document, also published in 2009,supplemented the Code, which includedguidelines and a community engagement toolkitdesigned by cega.Best practice notesWith the knowledge that several benchmarkdocuments already exist that adequately meetnationally and internationally recognisedstandards, AGEG and AGEA decided to createtailored “best practice notes” that draw on theexisting body of knowledge and provide guidancein the context of the emerging geothermal energysector in Australia.Australian Geothermal Conference 2010ReferencesDepartment of Primary Industries, 2010,Geothermal Energy in Victoria LandholderInformation.International Association For Public Participation(IAP2), 2007, Core Values.Minerals Council of Australia, 2005, EnduringValue, the Australian Minerals IndustryFramework for Sustainable Development,Guidance for Implementation.Ministerial Council for Mineral and PetroleumResources, 2005, Principles for Engagement withCommunities and Stakeholders.SACOME, 2009, South Australian ResourceIndustry Code of Practice for Stakeholder andCommunity Engagement. Code and SupportingDocument.Western Australian Government Department ofIndigenous Affairs, 2005, Engaging withAboriginal Western Australians.Each note covers one of the following topics:Existing resources and tools for communityengagementIdentifying and engaging stakeholdersKey messages about geothermal energyHow to communicate messages about risksSkills required for community engagementMedia managementOutrage managementThe best practice notes series is a precursor to amore complete BPM on community engagement,to be completed by AGEG and AGEA in the nearfuture.AGEA Membership commitment tobest practiceThe final objective of the BPM process is thatmembers of AGEA, the peak geothermal industrybody, will sign up and agree to abide by bestpractice community engagement as a prerequisitefor being members.The undertaking of this commitment will give allstakeholders, whether they be governmentfunding bodies, communities, regulators or anyother stakeholder groups, comfort that world’sbest practice is being applied when Australiangeothermal projects are being undertaken.401

Australian Geothermal Conference 2010Optimising ORC use in Australian Geothermal ApplicationsCraig Morgan 1 , David Paul 21Pacific Heat and Power, Level 14, 500 Collins Street, Melbourne VIC 30002 Pratt & Whitney Power Systems, 400 Main Street, East Hartford, Connecticut, USA* Corresponding Author: craig.morgan@pacificheatandpower.com, +61 3 9013 4425IntroductionSince Pratt & Whitney Power Systems acquired amajority stake in Turboden in 2009, many hotsedimentary aquifer geothermal powerapplications across the world have beenassessed. A variety of power configurations havesubsequently been developed and costed thathave relevance for the future of the Australiangeothermal industry.This paper summarises the key considerationsmade in designing systems, as well as factors thataffect overall performance.Keywords: PureCycle®, Pratt & Whitney PowerSystems, Turboden, Organic Rankine Cycle,Pacific Heat and PowerInput conditionsInput conditions for ORC use in geothermalapplications across the world vary substantially: Geothermal brine input conditions varysuch as 74C in shallow and small scaleAlaskan wells, 98C in Birdsville, 130Cfrom a hot sedimentary aquifer in theUSA, 150C at the outlet of a hightemperature separator in New Zealand, oras high as 240C+ in enhancedgeothermal applications. Thesetemperature conditions, including flowrate may also change over time. Air temperatures vary from -5 C to 50C+during the day in desert regions ofAustralia. Water for cooling is sometimes available,but many times it is not. Energy prices vary between AUD 5c/kWhto 20 c/kWh, and is sometimes a marketrate or government-regulated. Incentives include Renewable EnergyCertificates equivalent to 2-4 c/kWh, ortax-incentive regimes. Well and steam field developments varyfrom free-flowing wells to 5km-deep wells. Noise restrictions vary from site to sitedepending on the location of the plant. Flammability and toxicity of ORC workingfluids is subject to a variety of planningregulations. Atmospheric conditions can include H2Swhich aggressively corrodes copper inplant and equipment. Some project developers wish to contractfor supply and commission of the ORConly, whilst others wish for a fullypackaged Engineer-Procure-Construct(EPC) solution.Amongst this, the ORC supplier must optimiseeach individual application according to a specificclient’s requirement, generally for highesteconomic return on their project and notnecessarily lowest specific cost of the power plant($/kW).ORC design considerationsApart from the input conditions, keyconsiderations in the ORC system design include: Corrosion → special and costly materialsfor the heat exchangers influence the unitcost significantly increase lead time. Scaling → there may be limits to theminimum temperature of the geothermalbrine due to deposition of silica or otherproducts. Fouling → removable covers and straightcleanable tubes assist in cleaning and ongoingefficiency. Cascade use and cogeneration (using thecondensing heat for other purposes) canimprove feasibility. Working fluid flammability can be an issuefor urban areas / insurance cost. Vapour plume from cooling towers areunsightly and there is a need for makeupwater in evaporative devices. There is a larger footprint required, andadditional noise emissions from the fansin an air condenser or fin/fan cooler.Water use is becoming a significant issue.Evaporative cooling towers feature: Smaller footprint. Lower noise emissions. Fresh water consumption. Chemical water treatment requiringoperation cost and environmentalimpacts.402

Australian Geothermal Conference 2010Air condensers feature: Larger footprint. Higher noise emissions. No water needed. Virtually no environmental impact and lowoperating costs.Other options include hybrid cooling systems thatoperate dry in winter and cooler ambientconditions, whilst using water in hotter conditions.Critical issues we observe are: Investment costs need to take intoconsideration initial capital requirements,ongoing costs, and future wateravailability. The generated yearly power output istightly linked to cooling conditions andparasitic loads, and is particularlysignificant for cooler resources.The best solution is dependent on the application,but in the main we see Australian clientspreferring dry or hybrid cooling systems of somesort.ORC performance & practical systemsAvailability of PWPS and Turboden ORCtechnologies in real world applications is at >98%. Maintenance costs are very low relative toother power generation technologies such asengines, gas turbines or steam turbines.The selection of working fluid is influenced by:• Cost.• Environmental impact.• Pressure levels selected.• Heat input curve.• Enthalpy drop & flow rate.• Flammability.• Cooling system.The optimal fluid, pressure levels, and turbine sizevaries from case to case.In hot sedimentary aquifer applications thegeothermal fluid normally stays in liquid phase.Currently the most proven and reliable ORCtechnology uses single-phase working fluids in‘normal’ pressure cycles. These systems can becascaded using multiple pressure levels toapproach the ideal Lorentz cycle, minimisingdisadvantages whilst maximising the benefits ofreliable systems.Technical innovationsThe product development team continue to makesystem improvements that increase efficiency,broaden the variety of suitable applications, andprovide the ability to operate in with zero water forcooling. Features available are:• Air condenser for use in areas with limitedaccess to water. The refrigerant iscondensed directly in the air-condenser tomaximise heat transfer.• Coatings and equipment modificationssuch as those necessary for a high H2Senvironment in some internationalgeothermal fields.• High pressure evaporators may also beselected on a case by case basis forsome applications.Price vs PerformanceA manufacturer is able to design for maximumefficiency, or minimum specific cost. It is criticalthat the client and equipment supplier worktogether to optimise the unit design for eachspecific application. The chart below illustratesthis.403

Australian Geothermal Conference 2010The table shows price/power ratios for differentpossible configurations of an air-cooled ORC unit(installed, BoP excluded) for the same Project.Generally speaking the price/power ratio isinfluenced by the boundary conditions as follows: a colder(hotter) geothermal water willraise(decrease) the price/power ratio. a more aggressive and impuregeothermal water will raise theprice/power ratio. a higher(lower) average annual airtemperature will raise(decrease) theprice/power ratio.But in the design phase the choice between abigger or a smaller heat exchanger will influenceperformance and price, that means also theprice/power ratioThe future and corporate updatePWPS and Turboden have recently beensuccessful in winning orders for a 5 MWgeothermal power plant in Germany (based onthe Turboden technology), and a 500kWgeothermal power plant in Taiwan (based onPureCycle®).ConclusionThere is a wide variety of input conditions anddesign considerations that influence the final costof an ORC power plant, and as such each project,whilst based on modular technology, will mostlikely be configured in a project-specific manner. Itis therefore critical that you work closely with yourequipment supplier to optimise the configurationto balance: Technical performance Efficiency Economic performance, Reliability and Operability404

Australian Geothermal Conference 2010Mineral scaling in geothermal fields: A reviewYung Ngothai 1 *, Norio Yanagisawa 3 , Allan Pring 2 , Peter Rose 4 , Brian O'Neill 1 , Joël Brugger 21School of Chemical Engineering, The University of Adelaide, SA 50052 South Australian Museum, North Terrace, Adelaide, SA, 50003 Institute for Geo-Resources and Environment, AIST, Tsukuba, 305-8567, JAPAN4 Energy and Geoscience Institute, The University of Utah, Utah 84109, USA.*Corresponding author: yung.ngothai@adelaide.edu.auAbstractGeothermal power is an established energysource in several countries, for example NewZealand and Iceland. However the proposedgeothermal operations in South Australia occursat a much greater depth (5 km) and the heatsource is radioactive decay rather than volcanism.A number of issues relating to the geochemistry ofgeothermal fluids are required to be consideredand explored to ensure safe, economic energyproduction from geothermal fields. Low pH andsaline waters, at temperatures much greater than200 o C, are highly corrosive, and it is vital toprevent the generation of scales as the brines aretransported to the surface. This paper provides areview on silica, calcite and metal sulphide scalingat various geothermal fields. The solubility of silicaand calcite as a function of temperature and/orpressure were discussed and how it affectsscaling at various locations in the geothermalplant.Keywords: silica scaling, calcite scaling, metalsulphide scaling, EGS, HDR.IntroductionIn many developed geothermal fields in USA,New Zealand, Indonesia, Japan etc., mineralscales tend to precipitate with change of fluidtemperature and/or pressure, leading to fluid flowproblems. To overcome the operational problemsassociated with this scaling, several methodshave been developed to eliminate and/or reducethe amount of mineral deposition. It is howeverimportant that we further investigate themechanisms involved and the relation betweenmineral scaling and the physical and geochemicalcharacteristics of the rocks and fluids in thegeothermal system under exploitation.For future geothermal systems, research anddevelopment of Enhanced Geothermal System(EGS) is required at a rapid pace, especially inAustralia.The outline of EGS (or Hot dry rock – HDR)system is described as follow; injected water isheated in artificial reservoir constructed usinghydrofracturing technology by flowing through ahigh temperature granite. The heated fluidproduced in reservoir, is used for powergeneration.The first EGS project was carried out at FentonHill, New Mexico (USA), which ran from 1970.After that other EGS projects were established ina number of countries e.g., Japan (at Hijiori),Soultz, France, etc. Metal corrosion and mineralscaling were observed at these EGS systems. Forexample, in Soultz, metal corrosion occurred dueto high Cl concentrations (Baticci et al, 2010)while in Hijiori, anhydrite and calcite scalingoccurred in the lower production wells andpipelines at ground level due to dissolution ofanhydrite in the reservoir by the injected lowtemperature water (Yanagisawa et al., 2008).These problems are common to the welldevelopedconventional geothermal power plantsand likely to feature in EGS systems.This paper presents a brief review of mineralscaling in geothermal fields and discussion forfuture EGS development.Mineral scalingThe fluid in a natural geothermal system has along residence time (thousands of years) at arelatively high temperature (> 150 o C – 300 o C)which therefore indicates a significant rock-waterinteraction. Since the system has had a longresidence time, it is often assumed to be in a stateof chemical equilibrium (Grigsby et al., 1989).Therefore, when fresh water is introduced into thesystem to extract heat, the system is no longer inequilibrium, which leads to precipitation anddissolution of different parts of the mineralassemblage. Precipitation, or scaling, is one ofthe major problems in geothermal energyextractions. Minerals may precipitate dependingon the compositions of the granite and theinjected brine.In geothermal fields, it has been reported thatsilica, calcium carbonate, anhydrite, metalsulphides and iron minerals scales precipitatedaround production and injection wells and in thepiping.Silica scalingSilica scaling is a well known problem inconventional geothermal systems, which occursdue to the presence of amorphous silica, duringheat extraction or partial flashing (Robinson,1982). Amorphous silica can contain severalmetals such as iron, magnesium, and calcium, etc.405

Australian Geothermal Conference 2010Silica scale precipitated in many geothermal fieldsdue to its high solubility at high temperature in thereservoir through water-rock interactions and thesolubility rapidly decreases with decreasingtemperature. The solubility of silica is alsostrongly dependant on pH. For example, ironbearedamorphous silica tends to precipitated atfluid mixing points. At this point, temperature,pressure and pH of fluid change rapidly andleading to significant changes in solubility andchemical equilibrium leading to silica precipitation.Furthermore, silica has several structural phases,for example, quartz and amorphous silica whilesilica exists in solution both as a polymer and amonomer form. This variety leads to a complexarray of mechanisms for silica precipitation,especially in lower temperature environment suchas at the injection well.The dissolution of solid silica phases (SiO 2 ) inwater has been extensively studied (Mackenzieand Gees, 1971; Owen, 1975; Robinson, 1982).The solubility data for various silica phases(Rimstidt, 1980 cited from Robinson, 1982) areshown in Figure 1 where the solubility of differentsilica phases is plotted versus temperature.Quartz is thermodynamically the most stablephase of silica thus its solubility at any giventemperature is lower relative to other silicaphases.dissolved silica in the brine may then becomesupersaturated. Depending on the degree ofsupersaturation the dissolved monomeric silicamay nucleate and deposit as amorphous silicascale on equipment, such as heat exchangersurfaces (Chan, 1989). The factors which affectthe solubility of amorphous silica in solutioninclude temperature, pH, pressure and saltcontent (Chan, 1989). Gunnarsson & Arnórsson(2005) state that there are two processesinvolving aqueous silica that occur in anamorphous silica super-saturated solution. Thefirst process is the direct precipitation ofamorphous silica directly onto the surface ofequipment. The second is the tendency for silicato polymerise and form colloids that remain insuspension for long periods of time. Polymericsilica has less tendency to precipitate fromsolution than monomeric silica.Kinetics of silica dissolution needs to beestablished in order to evaluate the likelihood ofsilica scaling. Factors that control the rate ofpolymerization of dissolved silica are pH, salinity,degree of supersaturation, presence of solidsubstances, and temperature (Angcoy, 2010).Bethke (1996) compiled the rate constants forquartz and amorphous silica dissolution whichwere originally determined by Rimstidt andBarnes (1980), are shown in Table 1.Table 1: Rate constants of quartz and amorphous silicadissolutionFigure 1: Equilibrium solubilities of silica phasesSilica concentrations in geothermal reservoirwaters at 200-350˚C are approximately 300-700mg/kg SiO 2 (Gunnarsson & Arnórsson, 2005).Fournier and Rowe (1966) state that the silicacontent in geothermal waters is controlled by thesolubility of quartz at depth and not the solubilityof amorphous silica at and near the surface of theground. When hot brine is brought to the surface itis either flashed or cooled while passing through apower plant system to release its availablethermal energy. Chan (1989) states that theCalcite scalingCalcite or aragonite (CaCO 3 ) scaling is also a wellknown problem in conventional geothermalsystems, Calcite scale occurs near the flashingpoint in the production wells due to decrease incalcite solubility.Calcium solubility varies with the pressure of CO 2(P CO2 ) and temperature of the fluid. Figure 2shows the calcium solubility curve as a function ofP CO2 and temperature by Fournier (1985). It isnoted from this curve that under the sametemperature condition, calcite solubility is higherwith respected to increase of P CO2 . This explainswhy it is common to observe scaling at theflashing point. During flashing, where vapourrelease occurs causes the P CO2 to decreases,406

Australian Geothermal Conference 2010calcite solubility therefore drops and calcite scaleprecipitates there. Furthermore, as calcitesolubility is lower at high temperature conditions,calcite precipitation tends to occur at deep pointsin production wells and could be a seriousproblem. Many geothermal fields with highcalcium or hydro-carbonate concentrationexperience calcite scale problems. In Japan,especially at the Mori and Oku-Aizu geothermalfield calcite scaling is a serious problem.Figure 2: Solubilities of calcite as a function of temperatureand pressure of CO2To prevent calcite scaling in wells, variouschemical inhibitors have been used in manygeothermal fields. For example, sodiumpolyacrylate, C 2 H 3 COONa, has been used forseveral Japanese geothermal fields and hot spas.Similar chemicals have been used around theworld. At high fluid temperature, sodiumpolyacrylate reacts with calcium ion to makecomplex. Calcium complex cannot react with thebicarbonate ion (HCO 3) to precipitate calciumcarbonate. Thus, this sodium polyacrylate inhibitoris effective to prevent calcite scaling. For aproduction well, a capillary tube is inserted intowell until it reaches the depth of the flashing pointand the chemical inhibitor is then injected directlyat this point and prevents scaling.In Japan, metal sulphide scaling is found in Oku-Aizu (Nitta et al., 1991) and Yamagawa (Akaku,1988). At Oku-Aizu geothermal field, the sulphurremoval system near power plant has been builtto prevent sulphide scaling.The sulphide mineral scaling indicate theconditions in the deep-seated reservoir,especially in granitic rocks. For example, at theKakkonda geothermal area, in north-easternJapan, there is a 80 MW geothermal power plantusing high temperature fluid from the reservoir atthe boundary between the Quaternary Kakkondagranite and the Pre-Tertiary formations at about3km depth. Metal sulphide minerals deposit in theproduction wellhead and pipelines (Yanagisawa etal., 2000).The metal sulphide scales are classified into twotypes, based on sulphide mineralogy. These arePb-Zn rich type and Cu rich type. Pb-Zn rich scaleis found in Well-19 located at the marginal part ofthe Kakkonda granite as shown in Figure 3. It ismainly composed of amorphous silica, galena(PbS), sphalerite (ZnS) and pyrite (FeS 2 ). Thebrine of WD-1a at 3.7km depth, in the QuaternaryKakkonda granite rock, underlied Well-19, is richin Pb and Zn and the scale in Well-19 is of similarcomposition.. Cu-rich scale is found in Well-13,located at the central part of the Kakkonda granite.It is mainly composed of amorphous silica,chalcocite (Cu 2 S), bornite (Cu 5 FeS4), loellingite(FeAs 2 ) and native antimony (Sb). It is also rich inAu, Ag, As, Cr, Ni and Mo compared to Pb-richscale.Similarly the solubility of anhydrite (CaSO 4 ) islower at higher temperature and tends toprecipitate deep in the production wells and at thepeak temperature point. In this case, with thetemperature of fluid from deeper parts increasesanhydrite precipitate due to its decreasingsolubility. It has been reported that in Japan, Moriand Sumikawa geothermal fields experienceanhydrite scaling.Metal sulphide scalingMetal sulphide scaling is often occurred involcanic or high Cl environment. For example, atSalton Sea geothermal field, the salinity (35%NaCl) and metal concentrations (around 20%copper) in the fluid are very high (Skinner et al.,1967), many sulphide minerals form scales, suchas bonite (Cu 5 FeS 4 ), chalcocite (Cu 2 S) have beenfound.Figure 3: Schematic geothermal cross-section of theKakkonda geothermal system including well name and metalin precipitated sulphide and brine in graniteScaling in EGS systemIn EGS system, several different types of scalemineral precipitated in the Hijiori system, located407

Australian Geothermal Conference 2010in Yamagata Prefecture of Japan (Yanagisawa etal., 2008). At the Hijiori field tests wereundertaken using a series of hydraulic stimulationexperiments. A heating reservoir was created at2000m depth, after which river water was injectedinto the fractured reservoir, heated in reservoirand returned to installations on the ground level..Scaling at surface installations and in deep wells(HDR-2 and HDR-3) are shown in Figure 4.Anhydrite deposited in the deeper parts of theproduction wells, while silica and calciumcarbonate precipitated at the surface downstreamof the wellhead. Amounts of precipitationdepended on the temperature and chemicalcomposition of the produced fluid. In HDR-2,which is closer to the injection well, most of thescale was calcium carbonate; in HDR-3, which isfurther away from the injection well, there wasslight precipitation of amorphous silica. As fluidcirculation progressed and temperaturedecreased, scaling in the flow line of well HDR-2changed from amorphous silica to calciumcarbonate.Figure 5 shows the depth temperature profile. Aswater was injected anhydrite dissolved around theinjection well, the water heated as it flowedthrough the reservoir, and anhydrite precipitateddue to lower solubility at higher temperatures.Injection60ton/hourWaterScales at pipeline of HDR-2 and 3Anhydrite dissolvedCaSO 4→ Ca 2+ + SO2-4Hot water lineSeparatorHDR2:Calcite/Aragonite HDR3:SilicaFlow in reservoirHeatingHDR2 : 130C, 24ton/hHDR3 : 170C, 8ton/hWellheadAnhydriteIncluding CalciteFigure 4: Scale sampling site and photo at Hijiori siteFigure 5: Temperature profiles in the Hijiori wells measuredin July 2002Reference:Akaku, K., 1988, Geochemistry of mineraldeposition from geothermal waters: DepositionProcesses of common minerals found in variousgeothermal fields and case study in the Fushimegeothermal field, Chinetsu, 25, p.154-171 (inJapanese).Angcoy, E.C. and Arnórsson, S., 2010, AnExperiment on Monomeric and Polymeric SilicaPrecipitation Rates from Supersaturated Solutions,Proceedings of World Geothermal Congress,2710, p.1-6.Azaroual, M. and Fouillac, C., (1997).Experimental Study and Modelling of Granite-Distilled Water Interactions at 180 o C and 14 bars.Applied Geochemistry, 12, pp. 55-73Baticci, F., Genter, A., Huttenloch, P. and Zorn, R.,2010, Corrosion and Scaling Detection in theSoultz EGS Power Plant, Upper Rhine Graben,France, Proceedings of World GeothermalCongress, 2721, p.1-11.Bethke, C. M., 1996, Geochemical ReactionModelling. Concepts and Applications, OxfordUniversity Press, New York.Chan, S., 1989, A Review on Solubility andPolymerisation of Silica’, Geothermics, vol. 18, pp.49-56.Fournier, R. O., 1985, Carbonate transport anddeposition in the epithermal environment; inBerger, B. R., and Bethke, P. M. (eds.), Geologyand geochemistry of epithermal systems: Reviewsin Economic Geology, v. 2, pp. 63–72.Gunnarsson, I. and Arnórsson, S., 2005, Impactof Silica Scaling on the Efficiency of HeatExtraction from High-Temperature GeothermalFluids, Geothermics, vol. 34, pp. 320-329.Mackenzie, F. T. and Gees, R., 1971, Quartz:Synthesis at Earth-Surface Conditions, Science,New Series, 173, pp. 533-535.408

Australian Geothermal Conference 2010Nitta, T., Adachi, M., Takahashi, M., Inoue, K. andAbe, Y., 1991, Heavy metal precipitation fromgeothermal fluid of 87N-15T production well in theOkuaizu geothermal field, Tohoku distinct, Japan.Resource Geology, 41, p.231-242 (in Japanese).Owen, L. B., 1975, Precipitation of AmorphousSilica from High-Temperature HypersalineGeothermal Brine, California University,Livermore (USA), Lawrence Livermore Lab.,Technical ReportRimstidt J. D. and Barnes, H. L., 1980, TheKinetics of Silica-Water Reactions, Geochimica etCosmochimica Acta, 44, pp. 1683-1700.Skinner, B. J., White, D.E., Rose, H. J., Jr. andMays R.E., 1967, Sulfides associated with theSalton Sea geothermal brine., Economic Geology,62, p.316-330.Yanagisawa, N., Fujimoto, K. and Hishi, Y.,2000,Sulfide scaling of deep-geothermal well atKakkonda geothermal field in Japan, Proceedingsof World Geothermal Congress, p.1969-1974.Yanagisawa, N., Matsunaga, I., Sugita, H., Sato,M. and Okabe, T., 2008, Temperature dependentscale precipitation in the Hijiori Hot Dry Rocksystem., Geothermics, 37, p.1-18.Robinson, B. A., 1982, Quartz Dissolution andSilica Deposition in Hot Dry Rock Geothermalsystems, Thesis.409

Australian Geothermal Conference 2010Direct GeoExchange Cooling for the Australian Square KilometerArray Pathfinder – Field TrialDonald J. B. Payne 1,2* , and Dezso Kiraly 31 School of Physics, University of Melbourne, Parkville, VIC, 3010.2Direct Energy Pty. Ltd., 6 Alma Rd, St Kilda, VIC, 3182.3 CSIRO Astronomy and Space Sciences, Corner Vimiera & Pembroke Rd, Marsfield, 2122.* Corresponding author: djbpayne@unimelb.edu.au; 0414 990 310Direct Geoexhange Heat Pumps (DGHPs)provide chilling via refrigerant-carrying copperloops buried in the ground which act as acondenser and achieve higher efficiency thanequivalent air source heat pumps because of theground’s constant heat capacity. DGHPs areparticularly suited to desert environments withmore extreme ambient temperatures. TheAustralian Square Kilometer Array Pathfinder(ASKAP) radio telescope will be an array of 36 x12-m diameter parabolic dish antennae situated inBoolardy, WA each requiring 5 – 9 kW th coolingfor computer equipment and pedestal. Followingsuccessful tests of a prototype installation atMarsfield, NSW (ground temperature 17 C) aDGHP is installed at antenna site #29 (the firsterected). The DGHP provides chilling to a variabletest load (3.6, 4.8, 8.4, 9.6, 13.2 kW) via a 200 Lwater buffer. 6 x 30m bores containing copperloops deliver up to 14 kW of chilling to maintain 13- 15 C water temperature in the buffer. 5 databores monitor ground temperature variation in thecentre of the 6 active bores and at 0.3, 0.6, 3.0,6.0 m (outside the bore field). System pressures &power are also monitored. A higher-than-expectedbaseline ground temperature of 27 C is measuredand verified by three methods. The nominal 14kW DGHP can maintain 15 C water temperatureat all loads including 13.2 kW. Note that 0.5 kWpump heating means the effective load is 13.7kW. We describe the results from this fieldinstallation in detail and present an abridged setof data.Keywords:Direct Geoexchange, Direct Use, GeothermalHeat PumpDirect GeoexchangeThe emerging geothermal industry in Australia isfocused on producing electricity (“indirect use”)and should begin delivering substantial resultsover the next decade. Direct Use geothermalenergy is more efficiently available as it entailsonly one energy conversion (absorption orradiation of heat), rather than the several thatoccur in electricity generation and usage, witheach step losing a percentage on conversion.Geothermal Heat Pumps (GHPs) are a direct useof geothermal energy which involve circulating afluid (water, brine or refrigerant) through earthloops (poly pipe or copper) a few tens of metresdeep (Fig. 1, Payne et al. 2008).Direct Geoexchange Heat Pumps (DGHPs) whichcirculate refrigerant through copper loops havegreater efficiency than water-loop GHPs because:(i) copper is more thermally conductive thaninsulating plastic;(ii) latent heat transfers directly with the groundon evaporation or condensation; and(iii) an intermediate water-to-refrigerant heatexchanger between the ground loops andcompressor is not required.DGHPs transfer heat via 30-metre deep, 75-mmdiameter bore holes compared with 100-metredeep, 150-mm diameter bores used for water-loopGHPs. A continuous, closed loop of copperpiping is inserted and sealed with a thermallyconductive grout (cement). Below 5 metres theEarth remains at a stable temperature all-yearround(27 o C at Boolardy, WA). The smallertemperature difference between the heatsource/sink and the building or water to beheated/cooled results in lower head pressure andenergy requirement of the compressor comparedto conventional heating and cooling systems. A“desuperheater” may be employed in chillingapplications to further optimise the performanceand produce hot water or air which can be usedusefully elsewhere.ASKAPThe Australian Square Kilometer Array Pathfinder(ASKAP) radio telescope is an array of 36 x 12-mdiameter parabolic dish antennae situated inBoolardy WA (Fig. 2) whose constructioncommenced in early 2010 and is a precursor tothe Square Kilometer Array (SKA) of severalthousand 12-m dishes. A prototype of theElectronics Systems (ES) and Phased Array Feed(PAF) package to be cooled has been installed atthe CSIRO Australian Telescope National Facility(ATNF) headquarters in Marsfield, NSW. Thesetwo components should not exceed 20 - 30 o Cunder all operating and weather conditions andrequire an estimated heat dissipation of 5 – 7kW th . Various cooling methods have beenconsidered and the extreme desert temperaturesand high price of diesel generated power stronglyfavour a DGHP solution. The results of testing ofa DGHP at Boolardy are summarised in thispaper.410

System DesignCooling is provided by a 14 kW th DGHP whichcirculates refrigerant (R407C) through 6 x 30-mcopper earth loops to a refrigerant-to-water, coilin-tankheat exchanger. Water in the 200 L buffertank is chilled to 7 – 15 o C. Water is circulatedfrom this buffer tank to the PAF & ES to deliverthe required cooling to the heating load (the“secondary” circuit). The design of the secondarycircuit is beyond the scope of this paper – herethe performance as a function of thermal load isexplored by a test load. The test load is a 50 Lwater tank with three elements (3.6, 4.8, 4.8 kW)that can be switched on to produce 3.6, 4.8, 8.4,9.6, 13.2 kW. A circulation pump continuallycirculates water through this test load andcontributes circa 0.5 kW of heating. A controllerachieves the specified water temperature.Earth Loop InstallationThe 14 kW th system requires 6 x 30-m copperearth loops (½-inch vapour and ¼-inch liquid) withPVC insulation on the upper 15-m of the liquid lineto minimise heat transfer between the two andstop flashing of the liquid refrigerant. 70-mmdiameter holes are drilled vertically 30-m deep ina hexagonal configuration around a point 10mfrom the pedestal foundation edge with 3-mspacing between them to allow sufficient thermaldiffusion through the ground. After insertion of theloops the holes are filled with a thermallyconductive (geothermal) grout (e.g. 111-mix,Therm-Ex, IDP-357, Barotherm) ensuring thatthere are no air pockets. The liquid & vapourlines from the earth loops are brazed to theirrespective manifolds (½-inch liquid & 7/8-inchvapour) using a 15% silver brazing alloy. Theliquid line is insulated with ½-inch non-corrosiveinsulation material (e.g. Armaflex, Insul-Lock) andrun towards the heat pump (compressor) with amaximum length of 40-m. 5 data bores were alsodrilled. The location of all bores was logged bytriangulation to GPS pegs on the foundation (asshown below).Australian Geothermal Conference 2010Buffer Tank, Heat Exchanger & Test LoadChilling is delivered through a submerged copperevaporator coil inside a 200 L buffer tank – this isin contrast to the brazed-plate heat exchangerused in the Marsfield installation. Water flows overthe coils via a central perforated copper pipe toensure effective heat transfer. A minimum flowrate of 0.6 L/s (36 L/min) is required to achievethis. A 50 L test tank equipped with threeswitchable heating coils (4.8, 4.8 & 3.6 kW)simulates the load. The test tank is connected tothe buffer tank via 32mm polyethylene pipe.Figure 4: Evaporator coil & load tankHeat Pump Design & RefrigerationThe DGHP installed at Boolardy differs from thatinstalled Marsfield in the following ways:- The desuperheater coil is built into theunit- The unit is cooling only having noreversing valve- It has a nominal capacity of 14 kWth (cf.10.5 kWth) to cater for pedestal coolingFigure 1: Loops, liquid & vapour manifold andtrenches.The components are described in turn.Compressor: A 3-phase Scroll compressor drivesthe heat pump with compression ratio of 4. Itsefficiency is a function of the evaporating andcondensing temperatures (pressures).411

Australian Geothermal Conference 2010Active Charge Control (ACC): This:• prevents liquid refrigerant from reaching thecompressor by acting as a reservoir;• evaporates refrigerant to keep the systemproperly charged and to eliminate superheat;• improves volumetric efficiency, reduces powerdraw, and lets compressor run cooler;• enables passage of oil entrained in refrigerant;• indicates refrigerant level via 3 sight glasses.Liquid Flow Control (LFC): This is an efficientThermal Expansion Valve (TXV) which:• Sets proper refrigerant flow rate based oncondenser (upstream) operating conditions;• Ensures zero sub-cooling so condenser is fullyactive;• Reduces compressor discharge pressure andlowers power requirement;• Prevents vapor from “blowing through”.Oil Separator: Oil lubricates the compressor andthe oil separator acts with the ACC to prevent oilfrom migrating down the earth loops.Dryer: This filters and dries the refrigerant beforeit enters the LFC.Refrigerant: R407C [HFC azeotrope: R32 (23%)+ R125 (25%) + R134a (52%)]. Future work willexplore other HFC and hydrocarbon refrigerants.Expected PerformanceThe Air-Conditioning, Heating and RefrigerationInstitute (AHRI) has a well-established standardfor DGHPs: ANSI/ARI Standard 870, 2001. Also,DGHPs are EnergyStar rated, endorsed by theEnvironmental Protection Authority (EPA) andElectrical Testing Laboratories (ETL) tested. TheDGHP manufacturer, EarthLinked, providesperformance tables which are derived from boththe Scroll compressor’s performance and fieldtrials. For an earth temperature of 27 o C andchilling water to 15 o C, the expected Coefficient ofPerformance (COP = thermal energy removed/electrical energy input) is 4.5 – 5.0 depending onload. The theoretical maximum (Carnot cycle)performance for cooling is given by T evap /(T cond –T evap ) where T cond is the condensing temperatureand T evap the evaporating temperature (Kelvin)and is 7.7 for T cond = 42 o C and T evap = 6 o C whichcorrespond to the above conditions. Select7software from Emerson Climate Technologies(manufacturer of Copeland Scroll compressors)suggests 14 kW.MethodologyFigure 5: Final setupTo find the most efficient means of cooling, theCOP is measured as a function of the controllableaspects of the system. The key controllablevariables are:Buffer Set Point (T b-set ): the temperature to whichthe buffer tank is controlled to be chilled – it ismeasured inside the buffer tank.Buffer Maximum (T b-max ): the buffer tanktemperature at which the controller switches onthe chilling DGHP.Cabinet Set Point (T cabinet ): the temperature towhich the cabinet is cooled.Secondary Load (P load ): the thermal power whichis simulated with a set of heaters.Tank Load (P tank ): there are no elements in thetank.Other input variables include:Ambient Temperature: this includes both the wet(T wet ) and dry (T dry ) bulb temperatures.Environmental Gain (P envt ): the tank & pipes areinsulated but there is still environmental gain.Tank Volume (V tank ): the buffer tank volume.Water Volume (V water ): volume of water in thesystem: tank, primary & secondary pipes.Output variables include:Compressor Electrical Power (P comp ): theelectrical power used by the compressor.Cycle Time (t cycle ): the time between compressorstart-ups.Compressor Time (t comp ): the time thecompressor is on during a cycle.Duty Cycle (D): the percentage of time thecompressor is on (t comp /t cycle ).Earth Loop Temperatures (T 5–30m ): Thetemperatures measured at 5, 10, 15, 20, 25 &30m.Refrigerant liquid & vapour temperature (T liq ,T vap ): the temperatures to and from the earthloopsSuction, Head & Return Pressures (P head , P suc,P ret ): The pressures in & out of the compressorand returning from the earth loops.412

Australian Geothermal Conference 2010ResultsCOP CalculationThe default input values are: T b-set = 13 o C, T b-max= 15 o C, T cabinet = 23 o C, P load = 8.4 kW th , P tank = 0,V tank = 200 L, V water = 273 L. The ambienttemperature varies between 10 and 35 o C duringthe test periods and future results will becalibrated against this. An estimate of theenvironmental gain was determined by cooling thebuffer to 15 o C, leaving the system off andmeasuring the time taken (t envt ) for thetemperature to increase a known amount (∆T test ).P envt = C pw .V water .ρ w . ∆T test /t envt where C pw = 4.2kJ/(kg.K) is the specific heat of water, ρ w = 1 kg/Lis the density of water. It was found that for ∆T test= 2 K, t envt = 135 min giving P envt = 208 W.Figure 5 shows the temperature and pressureoutputs for the above input conditions. Fromthese results, the COP can be derived. We findt cycle = 1160 sec, t comp = 780 sec, giving D =67.24%. There is a temperature overshoot oftypically 0.3 – 0.5 K below T b-set and above T b-maxgiving ∆T = 2.8 K. The power required to reducethe system water temperature is P water =C pw .V water .ρ w . ∆T/t comp = 4.12 kW th .The electric power consumed by the compressor+ pump + monitoring equipment is measuredcontinuously along with the electric powerrequired for the pump alone. This allows the totalelectrical energy consumed over a known numberof cycles to be computed. For this run, E = 3.081MJ is consumed over n cycle = 3 cycles (19.33 min)and the average power (directly monitored) P comp= E/(t comp .n cycle ) = 3.95 kW elec .COP = [P water + (P load + P envt )/D]/P comp = 4.46. Notethat D accounts for the fact that the load andenvironment are constantly delivering heat to thesystem which must be dissipated by the heatpump whilst on.Using this method (see appendix), the COP iscalculated as a function of load and DGHPconfiguration and plotted below. It is clear that thesystem should not be designed to run with a100% duty cycle. An initial configuration of thesystem had the heat pump unit 30 m away fromthe earth loops. After moving the DGHP to within4 m of the earth loops, the pressure drop acrossthe earth loops was 5 – 15 PSI lower. Also, themanifold pit was initially exposed for testing andthus about 15 m of the copper earth loops wasexposed to air and thus not efficiently dissipatingheat. Backfilling the manifold pit boosted theperformance.Mean Earth TemperatureThe undisturbed mean ground temperature wasmeasured every 5m to 30m down six bore holes(two active and four data) via temperaturesensors. One of the bore holes (B5) has sensorsat 1, 2 and 3m also. Initial temperature readingswere measured on 4 th May 2010 and they werecorroborated with data loggers (LASCAR EL-USB-2) on 2 nd June 2010 (late Spring in Australia,equivalent to November in the NorthernHemisphere). The temperature is plotted as afunction of depth below and corroborates withwhat is expected. Independent verification oftemperature data has been important as theexpected mean earth temperature was 23 C. A120m test hole for water geoexchange was drilledat the Murchison Radio Observatory central siteapproximately 2 km away from the DGHPinstallation at antenna #29. Bore water inhabitedthis hole at a depth of 31.3m. The data loggerswere hung down the hole to a depth of 30m on arope at depths 0.5,5,15,30 m and left for 20 hrwith resultant mean earth temperatures of 21,29.5, 27.5, 27 respectively (see figure). A sampleof the bore water (jar down rope) was taken andits temperature upon extraction was just under 27C measured with a probe thermometer. Thisprovides a third means of verifying the boretemperature and corroborates.413

Australian Geothermal Conference 2010DC19.00 DC1918Physical AI4-10 WT TANK 1 - OUTPUT4-11 AMBIENT - OUTPUT.AI.01.05.09 - OUTPUT.AI.01.05.10 - OUTPUT17161514131206/03 00:30 01:00 01:30 02:00 02:30 03:0000:31 Thu 06/03DC47.79 DC4846Physical AI3-1 LB1 TMP 5 - OUTPUT3-2 LB1 TMP 10 - OUTPUT3-3 LB1 TMP 15 - OUTPUT3-4 LB1 TMP 20 - OUTPUT3-5 LB1 TMP 25 - OUTPUT3-6 LB1 TMP 30 - OUTPUT444240383606/03 00:30 01:00 01:30 02:00 02:30 03:0002:42 Thu 06/03PSI393.93 PSI400350Physical AI.AI.01.02.01 - OUTPUT1-01 COMP LP - OUTPUT1-02 COMP HP - OUTPUT1-03 COMP L/L - OUTPUT300250200150100Ground pH & Bore water salinity5006/03 00:30 01:00 01:30 02:00 02:30 03:0001:46 Thu 06/03The pH and salinity of bore water used for thethermal grout was measured with testing strips.pH = 7.8 - 8.2 - acidity is thus not an issue forcorrosion of the copper loops. The chloride levelsare circa 850 mg/L. This is a borderline reading ofthe bore water. The aquifer was not penetrated bythe site #29 bores and the ground is probablybenign. Taking a conservative approach, aCathodic Protection System will be installed forthis system – a solar solution for its power isbeing exploring.Abridged ResultsFor the presentation of data, a period from 12:00– 3:00am on 3 rd June 2010 was chosen. Belowappear the water (buffer tank) temperatures,temperatures down LB1, pressures, 15m boretemperature for all active bores, refrigeranttemperatures throughout the system, and electricpower. Also initially available temperatures fromthe data bores are plotted: B5 0.3m data bore,15m bore temperature for all data bores, B4 6mdata bore.DC40393837363539.98 DCPhysical AI3406/03 00:30 01:00 01:30 02:00 02:30 03:00DC908070605040302085.77 DC00:16 Thu 06/031006/03 00:30 01:00 01:30 02:00 02:30 03:00KW544.97 KWPhysical AI01:41 Thu 06/03Physical AI3-13 LB3 TP 15 - OUTPUT3-3 LB1 TMP 15 - OUTPUT3-9 LB2 TMP 15 - OUTPUT5-8 B5 TMP 13M - OUTPUT3-14 LB4 TP 15 - OUTPUT3-15 LB5 TP 15 - OUTPUT3-16 LB6 TP 15 - OUTPUT4-2 HP TEMP - OUTPUT4-3 LIQUID TMP - OUTPUT4-6 WT IN 1A - OUTPUT4-8 WT OUT 1A - OUTPUT5-2 M/VAPOUR - OUTPUT5-3 M/ LIQUID - OUTPUTCOMP SUCT TEMP - OUTPUT4-13 KWH 1 - OUTPUT4-14 KWH 2 - OUTPUT321006/03 00:30 01:00 01:30 02:00 02:30 03:0001:22 Thu 06/03414

Australian Geothermal Conference 2010DC31.030.530.029.529.028.528.027.527.0BENCH TOP26.505/29 05/30 05/31 06/01 06/02 06/03DC29.229.028.828.628.428.228.027.827.627.49 DCBENCH TOP27.405/29 05/30 05/31 06/01 06/02 06/03DC30.029.529.028.528.027.36 DC03:17 Thu 06/03Physical AI5-8 B5 TMP 13M - OUTPUT5-4 B5 TMP 1M - OUTPUT5-5 B5 TMP 2M - OUTPUT5-6 B5 TEMP 3M - OUTPUT5-7 B5 TEMP 8M - OUTPUT5-4 B5 TMP 1M - OUTPUT5-5 B5 TMP 2M - OUTPUT5-6 B5 TEMP 3M - OUTPUT5-7 B5 TEMP 8M - OUTPUT5-8 B5 TMP 13M - OUTPUT5-8 B5 TMP 13M - OUTPUT5-8 B5 TMP 13M - OUTPUT2-12 B4 TMP 15 - OUTPUT2-06 B3 TMP 15 - OUTPUT1-6 B1 TMP 15M - OUTPUT1-12 B2 TMP 15 - OUTPUT2-12 B4 TMP 15 - OUTPUT2-06 B3 TMP 15 - OUTPUT1-6 B1 TMP 15M - OUTPUT1-12 B2 TMP 15 - OUTPUT2-15 B4 TMP 30 - OUTPUT2-14 B4 TMP 25 - OUTPUT2-13 B4 TMP 20 - OUTPUT2-12 B4 TMP 15 - OUTPUT2-11 B4 TMP 10 - OUTPUT2-10 B4 TMP 05 - OUTPUTThis corroborates with data obtained fromMarsfield – it was expected that the higher groundtemperature would slightly reduce output &performance. As in Marsfield, it has beendemonstrated that COP decreases as a functionof load in the secondary circuit (which determinesthe duty cycle) simulated by a test tank. Furtherresults are being obtained for this configurationand additional experiments include the use ofalternative refrigerants. The thermal conductivity& other properties of the ground will be calculatedby comparison with analytic and numericalmodels.AcknowledgementsThis scientific work uses data obtained from theMurchison Radio-astronomy Observatory. Weacknowledge the Wajarri Yamatji people as thetraditional owners of the Observatory site. Weacknowledge the CSIRO, CASS team includingGrant Hampson, Evan Davis, Antony Schinckel,Graham Allen, Steve Broadhurst along withRobert Rechtman & Mike Francis of DirectEnergy, David & Peter Manoni from GeothermalWA, Hugh Campbell & Nigel Foster of DongaraDrilling and Shannon Lovett, Nick Holt & MarkAlden of Emerson Climate Technologies.27.5References27.006/03 00:30 01:00 01:30 02:00 02:30 03:00DC5654525048464442403836343254.59 DC02:51 Thu 06/03Physical AI3006/02 12:00 16:00 20:00 06/0313:54 Wed 06/025-2 M/VAPOUR - OUTPUT5-3 M/ LIQUID - OUTPUT3-3 LB1 TMP 15 - OUTPUT3-2 LB1 TMP 10 - OUTPUT3-1 LB1 TMP 5 - OUTPUT3-4 LB1 TMP 20 - OUTPUT3-5 LB1 TMP 25 - OUTPUT3-6 LB1 TMP 30 - OUTPUTAnalytic derivation of the expected groundtemperatures will be presented in a future paper.As seen above, after the compressor starts, theground temperatures rise asymptotically towardssaturation at which thermal output matchesthermal dissipation in the ground. Maximum load(13.2 kW th ) was applied from 10:38am to 17:17and the water temperature was held down to 15C. After switching off the load, the groundrecovers. At 18:11 a 3.6 kW load is applied, at18:45, 4.8 kW and at 19:26:30, 8.4 kW.Conclusions & Further WorkDGHPs are an efficient solution for heating andcooling and are particularly suitable for cooling ina desert environment with no power infrastructure.It has been shown that a DGHP is able to deliver13-15 C chilled water despite high groundtemperatures of 27.5 C. It does with a COPbetween 4 and 5 for the expected load range.Payne, D. J. B., Hampson, G., Kiraly, D. andDavis, E. Direct GeoExchange Cooling for theAustralian Square Kilometer Array Pathfinder,Australian Geothermal Conference 2009Payne, D. J. B., Ward, M. And King, R. L.,Position Paper: Direct Use GeothermalApplications, Australian Geothermal EnergyAssociation (AGEA), Feb 2008.Yu MZ, Peng XF, Li XD, Fang ZH, A simplifiedmodel for measuring thermal properties of deepground soil, Experimental Heat Transfer, 17(2):119-130, 2004.Cui, Ping, H. Yang and Z. Fang, Heat transferanalysis of ground heat exchangers with inclinedboreholes, Applied Thermal Engineering, Vol.26,1169-1175Marcotte, Denis and Philippe Pasquier, The effectof borehole inclination of fluid and groundtemperature for GLHE systems, Geothermics,Volume 38, Issue 4, pp 392-398Bristow, K. L., R. D. White, and G. J. Kluitenberg,Comparison of single and dual probes formeasuring soil thermal properties with transientheating, Aust. J. Soil Res. 32, pp. 447–464415

Australian Geothermal Conference 2010A Research Well for the Pawsey Geothermal Supercomputer CoolingProjectKlaus Regenauer-Lieb* 1,2 and the WAGCoE team 1,2,3 , Peter Malin 4 and the IESE team, Steve Harvey 2and the CSIRO CESRE team, Edson Nakagawa 2 and CSIRO’S Geothermal & Petroleum Portfolio, HalGurgency 5 and the QGECE team1,2, 3 Western Australian Geothermal Centre of Excellence,1 The University of Western Australia, M004, 35 StirlingAv. 6009,Crawley2 CSIRO Earth Science & Resources Engineering, ARRC, 26 Dick Perry Avenue, Kensington, Western Australia 61513 Curtin University, Kent Street, Bentley, Perth Western Australia, 61024 Institute for Earth Science Engineering, University of Auckland, Auckland, NZ5 Queensland Geothermal Centre of Excellence, The University of QueenslandBrisbane, Queensland, 4072 AustraliaAbstract:The Australian Government has funded thePawsey supercomputer in Perth, providingcomputational infrastructure intended to supportthe future operations of the Australian SquareKilometre Array (SKA) and to boost nextgenerationcomputational geosciences inAustralia. Supplementary Federal EIF funds havebeen directed to the development of a geothermalexploration well to research the potential for directheat use applications at the Pawsey Centre site.Cooling the Pawsey supercomputer may beachieved by geothermal heat exchange ratherthan by conventional electrical power cooling,thus reducing the carbon footprint of the PawseyCentre and demonstrating an innovative greentechnology that is widely applicable in industryand urban centres across the world. Theexploration well is scheduled to be completed in2013, with drilling due to commence in the thirdquarter of 2011. One year is allocated to finalizingthe design of the exploration, monitoring andresearch well, and first concepts will be presentedat AGEC for discussion and participation.Success in the geothermal exploration andresearch program will result in an industrial-scalegeothermal cooling facility at the Pawsey Centre,and will provide a world-class student trainingenvironment in geothermal energy systems.Keywords: Direct Heat, Absorption Chiller, HotSedimentary Aquifer, SKA, High PerformanceComputingIntroductionJune 9th, 2010 marks the first step to one of thelargest direct heat demonstrators in Australia. Adirect heat geothermal demonstration project,using hot sedimentary aquifers to provide coolingand ventilation to the Pawsey High-PerformanceComputing Centre and the co-located CSIROfacility the Australian Resources Research Centresecured funding through the Sustainability Roundof the Education Investment Fund (EIF). ThePawsey Centre demonstrator is part of theinfrastructure of Australia and New Zealand’s bidto host the SKA (square kilometre array) radiotelescope. The Pawsey Centre (see Figure 1)supports the enormous data requirements fordeep space research and is also planned to be aprime enabling platform for Australia’scomputational geoscience and nanosciencecommunities. The supercomputer-cooling projectwill be the first of its kind and will demonstrate thefeasibility for widespread use of the clean andsustainable geothermal energy source to power,for example, hospitals, shopping centres andother industrial scale applications.The Pawsey Centre is to be situated on the deepsediments of the Perth Basin, where a verticalsequence of hot sedimentary aquifers suitable fordirect heat geothermal applications is now beingstudied.Construction on the exploration well project isscheduled to commence in November 2010,drilling of the exploration hole is scheduled forNovember 2011 and the project is to becompleted in August 2013 (see table 1), theprovisional timetable is coordinated with theinternational SKA decision-making process.Australia and New Zealand, together, are incompetition with an African SKA bid, involvingSouth Africa, Namibia, Botswana, Mozambique,Madagascar, Mauritius, Kenya and Ghana. Adecision is expected to be made in 2012. ThePawsey Centre geothermal demonstrator is partof a portfolio of solar and geothermaltechnologies, notably an additional ground sourceheat pump design for radio antenna cooling. Theoverall goal is to supply a reliable and clean,sustainable energy solution for the Australian/NZSKA bid.This contribution gives a brief overview of theentire concept, its location, timing, and the aboveand below ground technology concepts plannedto support the design of a production geothermallypowered cooling system. Collaborations betweenAustralian and NZ researchers on potentialdesigns for the exploration well are discussed.416

Australian Geothermal Conference 2010The Pawsey Supercomputer Coolingproject, a brief overviewOne of the strategic goals of the successful EIFSustainability Round proposal is to form throughdemonstration projects a joint research focus forscientists and students from all over Australia andNZ in renewable energy, as well as in astronomy,computer science, engineering, geology andenvironmental management.Figure 1. The Pawsey High-Performance ComputingCentre cooling project will be located on the CSIROTechnology Park site near the Perth CBD close toWestern Australian universities.The Pawsey Supercomputer Cooling (PSC)project is located centrally in Perth (Figure 1) andits collaborations with Australian and NZ partnerssets a sound basis for achieving this goal. Theadditional placement of a research well offers thenecessary infrastructure investment for researchcollaboration in the wider Australian and NZgeoscience community for testing, monitoring andexploitation of hot sedimentary aquifers.The PSC direct heat geothermal demonstratorutilises a hot sedimentary aquifer under Perth withan estimated temperature of 100 C at 3 kmdepth.Based on the available information andextrapolation from a 3 km deep petroleumexploration well (Cockburn 1, see Figure 3)located 21 km away from the Pawsey site onepossible scenario could be a 3 km deep extractionand injection well into the Cadda formation with ameasured permeability of 632-732 mDarcy and ameasured layer thickness of 350 m. Anotheroptional target layer is the Yarragadee aquiferdirectly above the Cadda formation featuringpermeabilities of up 1200 mDarcy according tothe Cockburn 1 records. The advantage of thistarget is its expected lower salinity than theCadda Formation. Shallower portions of theYarragadee aquifer have been successfully usedin ten different fully commercial low-temperatureprojects in Perth, mainly for swimming poolheating. The injection wells in such systems aremuch shallower than the extraction bores whichimplies minimal risk of thermal breakthrough. Theselection of injection depth is governed by localpermeability, net formation water balance andwater quality. Choice of the lower Yarragadeeover the Cadda will depend on the trade offbetween flow rates and temperaturesencountered in the exploration well.The first design of the above ground infrastructurecomprises two absorption chillers (Figure 4) at aFigure 2. Str atigraphy along a se ction through Cockburn 1 deep petroleum exploration well, passing ~20 km tothe south-west of the Pawsey Centre. Section from Crostella and Backhouse (2000).417

Australian Geothermal Conference 2010maximum capacity 5 MW th each, one unit isimplemented as a backup unit. The dimension foreach chiller is around 9.38m(L) x 2.2m (W) x 3.6m(H). The absorption chillers require heat rejectionwith an approximate maximum capacity of 25MW th . The preliminary design will be adjustedonce the actual cooling requirement of the PSC isknown.which could be abstracted for testing small scaleabove ground demonstrators. Key aspects of thesampling include water quality analyses thatprovide important indicators for heat exchangeplant and reinjection sustainability. Given, the longlead-up time for commencement of drilling (Table1) the design of the exploration well and theabove and below ground research componentsare fully open for discussion. In the spirit of theEIF funding we call here for active researchcollaborations that can complement the fundedinfrastructure with the best available expertise andinterpretation.Figure 3. The Pawsey High Performance Computercooling demonstrator design scheme consists ofabsorption chiller technology driven by heat from thehot sedimentary aquifer.Table1 Preliminary Timetable*Planned ProjectMilestonePlanned CompletionDateProject start October 2010Perth geothermal drillingprogram commencesPerth geothermal drillingcompletedAugust 2011November 2011Project completion December 2013* timings are tentative only and subject to revisionThe Dual Purpose of theExploration/Research WellThe exploration/research well is designed for adual purpose: (1) it breaks the ground and obtainsthe first direct samples of the chosen site prior toinstallation of the geothermal production wells;and (2) it offers the unique opportunity forimplementing purpose-designed downholesensors,long term monitoring and active testing.An additional attraction for research may be asmall flux of geothermal waters from the aquifer,Preliminary design scenarios of theExploration wellThere is no hard data indicating watertemperature (and chemistry) at depths of ~3 km inthe Perth Basin – the proposed depth of thegeothermal production wells. Data acquired fromthe drilling program associated with the UWAgeothermal demonstration site will provide abetter indication of water temperature at depth inthe Perth Basin, but will not obviate the need forthe ARRC geothermal exploration/research well.This depth of drilling is based on data sourcedfrom a ~800m Water Corporation bore adjacent tothe ARRC site. Data recently acquired from reloggingthis bore by the WA Geothermal Centre ofExcellence suggests water temperatures of~100 C at a depth of 3 km. Temperature, (fluidand sediment) chemistry and permeability datawill be acquired through the proposed geothermalexploration/research well. Data on the advectiveregimes operating at depth in the stratigraphicsequence will also be sought in order to facilitatethe construction of a geothermal reservoir modelfor the Pawsey Centre.The reservoir model will include hydrothermal flowprocesses necessary to describe circulationphenomena in moderate-permeability HSAgeothermal plays. The reservoir model will informthe design specifications of the subsequentproduction and injection well infrastructure.In preparation to the EIF bid IPS Australasia hasbeen tasked to prepare potential design scenariosfor the exploration well which are here laid openfor discussion. A number of scenarios have beenconsidered for the drilling, evaluation, testing andcompletion of the exploration well with proposedschematics shown in Figure 4.Scenario 1 (Base Case)Scenario 1 (Base Case) is designed with a three-final hole sectionstring casing program withcased with 178mm (7”) casing to 3000 m TrueVertical Depth from Rotary Table (TVD RT)(Figure 5). This design places the 244 mm (9-418

Australian Geothermal Conference 20105/8”) surface casing at 1000 m TVD RT in theYarragadee Formation and contains the aboveLeederville and South Perth Shale behind casing(Figure 2). This design offers a 216 mm (8-1/2”)open hole to conduct straddle testing of theproposed target intervals to 3000 m TVD RT.testing of the proposed target intervals to 3000 mTVD RT. The 152 mm (6”) hole section drilled to3500 m TVD RT will allow formation evaluation ofdeeper target intervals with the option of testingand casing with a 5” liner.Figure 4. Wellbore schematics of the exploration wellaccording to a Desk Study of IPS AustralasiaFigure 5. Wellbore schematics of Scenario 1exploration well suggested by IPS Australasia (2009).The casing configuration is summarized asfollows:The 340 mm (13-3/8”) conductor shoe isset at a depth between 50 and 100 mTVD RT.The 244 mm (9-5/8”) casing shoe is set ata depth of approximately 1000 m TVDRT.At well TD of 3000 m TVD RT followingopen hole testing of target intervals, a178 mm (7”) casing string will be set.Scenario 2Scenario 2 is of a similar design to Scenario 1with a three-string casing program to 3000 m TVDRT with the addition of a 152 mm (6”) hole sectiondrilled to 3500 m TVD RT. This design places the244 mm (9-5/8”) surface casing at 1000 m TVDRT in the Yarragadee Formation and contains theabove Leederville Formation and South PerthShale behind casing. This design offers a216 mm (8-1/2”) open hole to conduct straddleThe casing configuration is summarized asfollows:The 340 mm (13-3/8”) conductor shoe isset at a depth between 50 and 100 mTVD RT.The 244 mm (9-5/8”) casing shoe is set ata depth of approximately 1000 m TVDRT.Following open hole testing of targetintervals to 3000 m TVD RT, a 178 mm(7”) casing string will be set.Preliminary Design of the ResearchEquipmentThe preliminary list of research equipment fordeployment in the exploration well includes:Along-Casing Sensors: Optical fibretemperature monitoring system (Raman scatteringbased), Optical fibre pressure, temperature,acoustic, EM (Bragg grating/cladding based),Surface optical system, electromagnetic/magnetotelluric-receiver-station.419

Australian Geothermal Conference 2010Figure 6. Wellbore schematics of Scenario 2exploration well suggested by IPS Australasia (2009)Inside Hole: Water pressure tube, mechanicalpackers, banding standoffs (cable protectorsbetween pipes), pressure transducer,geochemical / environmental monitoring samplingsystemDownhole Sonde: Deep sonde (seismometer)thermistor, fluxmeters (desorption method whichneed to be deployed in packed zones to isolatefrom in-well gradients.Redeployable Wireline Exploration System:Wire / data cable and winch system flow meterdownhole formation imager, spectral gamma tool,electrical conductivity, caliper log.This equipment will allow us to monitor thepressure, temperature and acoustics of theaquifer along time and depth, collect fluid samplesat a range of depths for geochemistry analyses,seismic, flow rate and temperature data to monitorthe impact of the production plant.Above Ground Engineering ResearchWhile this project is demonstrating the direct useof the geothermal heat in space cooling, anotherobjective for the Australian geothermal energysector is the generation of electricity usinggeothermal heat. Additional targets are proof oftechnology for geothermal desalination ofseawater, saline aquifer water and industrialprocess water. Both Australian based inventionsare currently being built with support from differentCentres of Excellence in Australia.The Queensland Geothermal Energy Centre ofExcellence (QGECE) is developing newsupercritical cycle technologies to increaseelectricity production from a given geothermalresource by up to 25% for low temperaturegeothermal resources and by up to 50% for hightemperature geothermal resources. One of thecurrent research projects at the QGECE is aimingto develop a 100-kWe portable power plant todemonstrate a supercritical power generationcapability for a sub-150C geothermal resource.Using this plant as a test bed, optimum designand operating conditions can be ascertained forelectricity generation using the geothermal fluidproduced by a Perth Basin hot sedimentaryaquifer in local ambient conditions.The Western Australian Geothermal Centre ofExcellence (WAGCoE) in collaboration with theNational Centre of Excellence for Desalinationintends to use the extracted geothermal waterboth as a heat source and feedstock to develop aWAGCoE’s protected distillation baseddesalination technology by constructing acontainerized 4 m 3 /day desalination plant, asschematically suggested in Figure 7. Thistechnology is expected to boost the freshwateryield of conventional technology by 30%.Hot groundwaterMulti-effect distillation plantFigure 7. A geothermal distillation desalinationplantReferencesSalineground waterConcentrateDistilled waterCrostella and Backhouse (2000) Goelogy andPetroleum Exploration of the Central andSouthern Perth Basin, Western Australia,Geological Survey of Western Australia, DMPReport 57IPS Australasia (2009) A Desk Study: ProposedGeothermal Project at the Australian ResourcesResearch Centre.420

Australian Geothermal Energy Conference 2010Fracture Stimulation – legal risks and liabilitiesKyra Reznikov, Senior Associate, FinlaysonsAbstractFracture stimulation (also known as hydraulicstimulation) involves pumping water at highpressure into a well, which causes fracturing ofthe rock. This additional fracturing enhances thepermeability of the hot dry rock in order to createan efficient hydraulic loop for the purposes ofharnessing geothermal energy.seek to mitigate their potential liabilities wherefracture stimulation is proposed.The process of fracturing the rock can also giverise to seismicity events – that is, it can inducesmall earthquakes which have the potential todamage property and infrastructure not onlywithin the exploration licence area but also onadjacent land and further afield.The paper will examine the legal risks andpotential consequences of fracture stimulation,including:• damage caused by seismicity eventscould result in the commission of anoffence under environment protectionlegislation (which prohibits the causingof harm to the environment, propertyand the health, safety and amenity ofpeople) or under resources legislation(which for instance prohibitsinterference with mining operations);• landowners or operators ofinfrastructure that suffer damage totheir property may also seekcompensation for losses caused by theseismicity events;• if concerned about the potential impactson the environment and property,regulators could issue orderspreventing fracture stimulation, orallowing it only after precautionarymeasures are put in place.The paper will discuss the legal riskmanagement measures that geothermaloperators should include in project planning to421

Australian Geothermal Conference 2010QGECE Mobile Geothermal Test Plant & ORC Cycle ChallengesAndrew S. Rowlands and Jason CzaplaThe Queensland Geothermal Energy Centre of Excellence (QGECE),The University of Queensland, St Lucia, Brisbane, Qld 4072*Corresponding author: a.rowlands@uq.edu.auGeothermal energy reserves in Australia have thepotential to provide electrical energy for hundredsof years. However, due to the variety oftemperatures in geothermal resources a one-sizefits-allapproach to surface power infrastructure isnot appropriate. Furthermore, the use oftraditional steam as a working fluid is rarely afeasible option because resource temperatures liein the range of 100-200°C. To overcome this,organic fluids with lower boiling points than steammay be utilised as working fluids in organicrankine cycles (ORC) in binary power plants. Dueto differences in thermodynamic properties,certain fluids are able to extract more heat from agiven resource than others over certaintemperature and pressure ranges. Thisphenomenon enables the tailoring of power cycleinfrastructure to best match the geothermalresource through careful selection of the workingfluid and turbine design optimisation to yield theoptimum overall cycle performance. Here we lookat a selection of promising ORC cycles using arange of high-density working fluids operating atsub-, or trans-critical conditions that are beingstudied by QGECE as potential binary cyclesystems; discuss the challenges facing design ofORC cycles and how some of these challengesare being addressed in the QGECE MobileGeothermal Test Plant.Keywords: Geothermal energy, Organic RankineCycle, Working fluids.Modelling Binary Power CyclesIn this investigation, the thermodynamic cycle ofan ORC was simulated for conditionsrepresentative of a geothermal power station.The ORC was modelled as (1) pump work in, (2)heat addition in (i.e. pre-heater, evaporator,superheater), (3) turbine work out, (4) heatrejection out (i.e. pre-cooler, condenser) and (6)regeneration (when the temperatures allow).Figure 1 illustrates the basic layout of an ORCand denotes the state points for calculatingthermodynamic properties.The fluids database employed was the NIST’sREFPROP (NIST 2007). REFPROP offersequations of state for 84 pure fluids as well asuser defined mixtures. However, only pure fluidswere used in this analysis (Mixtures may be thefocus of future work).To allow for ranking results of the various cycles,the critical metric used for comparing fluids andcycles is Brine Consumption (β) (Franco, 2009)defined as follows:β = W net / mf geo (kJ/kg)Figure 1: General ORC LayoutWhere W net (kW) is the net work produced by thecycle and the mf geo (kg/s) is the mass flow rate ofthe geothermal brine. Greater β values indicate agreater yield of energy production per unit massof brine. This can be a good metric forgeothermal energy because it focuses onoptimising the energy produced for thegeothermal brine produced.The cycle analysis performed indicates that thepinch point in an evaporator has a significantimpact on β. This is in agreement with whatothers have stated previously in other publications(i.e. Saleh 2007). The pinch point is a product ofthe working fluid’s latent heat of evaporation(which is pressure dependent). Minimising pinchpoint effects will allow the cycle to make moreefficient use of the brine flow rate in theevaporator.In an ideal scenario the brine would cool to theinlet temperature of the working fluid and theworking fluid would heat up to the inlettemperature of the brine, transferring all the heatfrom the brine to the working fluid. However,there will always be a temperature differentialbetween the brine and working fluid becausesome temperature difference is required to drive422

heat through the walls of the heat exchanger.The aim then, is to minimise the temperaturedifferential and extract as much heat as possiblefrom the brine flow.Figure 2 shows an example of a T-s diagram andassociated pinch point plot. The cycle representsa subcritical cycle with superheating.T, o C120100806040200-20T-S Cycle Diagram (R152a)21ThdThc3 4Thb756T, o C150140130120110Brine Temp.Working Fluid Temp.Tca50 1001009080701000 1500 2000600Tcb TccEntropy, J/kg-KEvaporator Length ProportionFigure 2: T-s diagram and pinch point plot for R152a (Tgeo150°C and Tamb 40°C).Figure 3 shows an example of a T-s diagram andassociated pinch point plot. The cycle representsa supercritical cycle. It can be seen that atsupercritical conditions the effect of pinch point isgone.T, o C180160140120100806040Thb ThaT-S Cycle Diagram (C5F12)Thd21Tca200 200 400Entropy, J/kg-K7TcbThc34 56TccT, o C20018016014012010080Evaporator Pinch PointBrine Temp.Working Fluid Temp.600 50 100Evaporator Length ProportionFigure 3: T-s diagram and pinch point plot for C5F12 (Tgeo –200°C and Tamb 40°C).Supercritical cycles have the potential to yieldefficient cycles because they minimise pinch pointeffects in the evaporator and there tends to be alarge enthalpy differential available across theturbine in the cycle analysis. However, athorough turbine analysis must be undertaken inconjunction with the cycle analysis to validate thecalculated turbine work out from the cycleanalysis.Fluid PerformanceThree general categories of geothermaltemperature ranges were focused on to see howdifferent fluids performed at different operatingconditions. The ranges were labelled as low, midand high at temperatures of 100, 150 and 200°C,respectively. The ambient temperature was set at40 o C to represent high temperatures that can beseen in central Queensland where our research isAustralian Geothermal Conference 2010focused. For each temperature range the fluidswere ranked in ascending order of their calculatedβ. An abbreviated summary of the results isshown below in Tables 1-3.An obvious trend in the results tabulated above isthat haloalkanes perform well at temperatureconditions associated with geothermal resources.Table 1: Top 5 Fluids (Tgeo = 100°C)Fluid β (kJ/kg) Evap. CaseC4F10 2.07 SubcriticalC5F12 2.06 SubcriticalRC318 2.04 SubcriticalR227ea 2.04 SubcriticalR236FA 2.03 SubcriticalTable 2: Top 5 Fluids (Tgeo = 150°C)Fluid β (kJ/kg) Evap. CaseR152a 18.55 SuperheatedR134A 18.40 SuperheatedRC318 18.28 SuperheatedCOS 18.14 SuperheatedR236FA 17.78 SuperheatedTable 3: Top 5 Fluids (Tgeo = 200°C)Fluid β (kJ/kg) Evap. CaseC5F12 57.91 SupercriticalR236EA 55.67 SupercriticalISOBUTAN 54.61 SupercriticalR152a 52.06 SupercriticalDME 51.35 SupercriticalPerformance maps were also generated for eachfluid for a range of temperature conditions (T geoand T amb ). This allows for visualisation of howcycle performance can vary with changing inletbrine temperatures (i.e. a geothermal source maycool over its operating life) and ambient airtemperature (i.e. for air cooled systems thisrepresents shifts in temperature throughout theday/night and seasonal change). This can beuseful in ensuring that the fluid and cycle selectedwill operate effectively at all anticipated operatingconditions. Figure 5 shows an exampleperformance plot for R245fa. The top plot in thefigure shows T geo and T amb on the x and y-axeswhile β is on the z-axis. It is intuitive that whenT amb is at a minimum and T geo is at a maximum βwould be at its peak value. It is perhaps not sointuitive to see that evaporator pressureassociated with the optimum β for a given set ofT geo and T amb varies as it does. This can be an423

important factor in plants performance. Iftemperatures vary, then so should the pressure(along with other parameters that can beoptimised similarly) to achieve the bestperformance.Brine Consumption, kj/kgOptimum Evaporator Pressure, kpa10080604020320 0600050004000300020001000320 0310310300Tamb, K300Tamb, KOptimum Brine Consumption Performance Plot of R245fa290290280280360360380380400400420Tgeo, KOptimum Evaporator Pressure Performance Plot of R245fa420Tgeo, KFigure 4: Performance plots for R245fa, optimum brineconsumption (top) and associated evaporator pressure(bottom)Haloalkanes as Working FluidsMany haloalkanes, especially thehydrofluorocarbons, such as ,1,1,3,3-,Pentafluoropropane (HFC-245fa) and 1,1,1,2-Tetrafluoroethane (HFC-134a), which is inwidespread use as a refrigerant, are inflammable,non-toxic and relatively inert. These propertiesmake them highly attractive to use as workingfluids in binary power cycles.One of the drawbacks of these fluids is thermalstability, which, in most cases, is less than thenon-halogenated hydrocarbon equivalent andmay limit the maximum temperature for whichthey can be used in power cycles. This is an areaof significant uncertainty for ORC technologybased on current generation fluids.Some of the decomposition products can not onlycause system performance to degrade but arealso toxic and/or strong acids that may negativelyimpact plant. While some degradation of thecycle fluid to a static level may be tolerable incertain situations, gross scale degradation overtime of the working fluid is not acceptable.While there have been a number of studies(Angelino, 2003; Calderazzi, 1997) looking at thestability of these fluids, existing data does notnecessarily agree on exact degradationtemperatures and have, in general, looked only atthe stability when in contact with stainless steel,440440460460480480Australian Geothermal Conference 2010which has negligible catalytic interaction.However, in real-world applications the workingfluid may come into contact with a number ofdifferent materials, all of which may interactdifferently with the working fluid and may furtherreduce the reported degradation temperature.Further effects, including atmosphericcontamination (e.g. the presence of smallamounts of water, which may acceleratedecomposition) have not been thoroughlyinvestigated to the best of our knowledge.This is of concern when designing a plant as thepresence of materials that may act as catalysts fordegradation must be avoided in all parts of thesystem that come into contact with the workingfluid. Even common materials used in heatexchangers may be cause for concern.Optimising the Operating PressureNot only do certain fluids perform better thanothers at certain temperatures, but there exists anoptimum pressure for cycle performance.Net Power (kW)5004003002001000Power vs Pressure(150°C Resource)0 2000 4000 6000Pressure (kPa)HFC-134aHFC-245faFigure 5: Optimum cycle pressures at 150°Cfor HFC-134a and HFC-245fa.The general assumption that higher pressuredirectly translates to higher work output does notnecessarily hold true at pressures both below andabove the critical pressure, depending on the fluidin question. Figure 5 shows the calculated poweroutput over a range of pressures for HFC-134aand HFC-245fa at operating temperatures of150°C. As can be seen, at this temperature theoptimum system pressure for these two fluids isdifferent and produces significantly different netpower outputs.Optimising the TurbineWorking fluid properties, resource temperature,flow rate and condensing pressure are allvariables that dictate turbine design. In mostinstances, the goal is to generate the highestoutput power possible. Therefore designoptimisation centres around generating thehighest efficiency turbine. Generally, this resultsin turbine designs that need to operate at highrotational speeds that are not suitable for direct424

coupling to grid synchronous generators. QGECEis focussing on optimising turbines that haveoperational speeds of 24,000 RPM (400 Hz). Thisresults in high operating efficiencies and iscompatible with existing power electronics, suchas that found in aircraft. To date, current QGECEpreliminary turbine design has not indicateddramatic differences in turbine efficiencies forvarious ORC cycles operating with pressure ratiosof between 2 and 6.Optimising the CycleSo far we have touched on a number ofconsiderations of various key cycle componentsincluding the working fluid, system operatingpressure and turbine design. When combined, adifferent picture may begin to emerge as to thecomposition of elements that yields optimumpower output for a given resource temperature.For instance, depending on the design conditionsassumed in the thermodynamic cycle analysis,the pressure ratio calculated that gives anoptimum theoretical cycle performance may resultin a suboptimal turbine efficiency, which mayaffect the initial ranking of working fluid selection.Net Power (kW)5004003002001000Net Power vs Pump Efficiency0 0.2 0.4 0.6 0. 8Pump EfficiencyHFC-134aHFC-245faη pump = 0.65P = 1200 kPaFigure 6: Effect of ηpump on net power.Furthermore, variations in efficiencies of othercomponents in the system are often neglected intheoretical studies for ease of comparison, butcan have a large impact on overall systemperformance. For instance, the benefits of transandsupercritical cycles can be harnessed only ifhigh performance pumps or compressors exist forthe required conditions, otherwise a subcriticalcycle may become increasingly attractive due tolower pump work and reduced operatingpressures. Figure 6 shows the impact on netpower for a supercritical HFC-134a cycleoperating at 5 MPa compared to an optimisedsubcritical HFC-245fa cycle operating at 1.2 MPaat the same temperature.QGECE Mobile Test Plant (Terragen)The Terragen Project is a QGECE initiative todevelop a modular portable transcritical powerplant able to generate approx. 75 kW of powerfrom geothermal or waste heat sources. One ofAustralian Geothermal Conference 2010the primary goals is to demonstrate that usefulamounts of energy can be produced even fromlow-grade resources using transcritical fluids.Terragen consists of three containerised modules:a control module, power-plant module and an aircondenser module. The system is beingdesigned around two working fluids HFC-134aand HFC-245fa using oil-free technologies.Terragen will be a real-world demonstration andvalidation of cycles and technologies beingdeveloped at QGECE and will primarily be anexperimental rig based at The University ofQueensland’s rural Pinjarra campus. Theportable nature of Terragen means that it will alsobe available for companies to carry out welltestingand help prove the electricity-generationcapability of resources to assist with attractingpublic and private sector funding.Terragen is currently under development incollaboration with our project partner, Verdicorp,and is expected to be completed mid-late 2011.ConclusionsThe general assumption that higher pressuredirectly translates to higher work output does notnecessarily hold true at pressures both below andabove the critical pressure, depending on the fluidin question. Figure 5 above shows the calculatedpower output over a range of pressures for HFC-134a and HFC-245fa at operating temperatures of150°C.In developing an ORC that generates optimumcycle power output from a given resource thereare a number of critical considerations including,but not limited to, selection of best-suited fluid foroperating temperature; selecting optimisedoperating pressures for chosen fluid; optimisedturbine design and selection of auxiliary cyclecomponents that complement these major cycleelements.Future work will work towards verifying thepredicted performances experimentally both in thelab at small power scales and at larger scaleusing the QGECE Mobile Plant. Furthermore,future work will address the life cycle analysis ofimplementing various sub- and transcritical cyclesto verify that the cycles that give the bestperformance theoretically still out-perform theother cycles on an economic and feasibility basis.ReferencesAngelino G, Invernizzi C, 2003, Int J. Refrig. 26(1)Calderazzi L, di Paliano PC, 1997, Int J. Refrig. 20(1).Villani M. Franco, A., 2009, “Optimal design of binary cyclepower plants for water-dominated, medium-temperaturegeothermal fields”. Geothermics, 38(4): 379 – 391Lemmon, E.W., H. M. M. M. NIST Standard ReferenceDatabase 23: Reference FluidThermodynamic and TransportProperties-REFPROP 2007425

Saleh, B., Koglbauer, G., Wendland, M., Fischer, J., “WorkingFluids For Low Temperature Organic Rankine Cycle” Energy,2007, 32, 1210–1221Australian Geothermal Conference 2010426

Australian Geothermal Conference 2010Australian geothermal Industry working with Investors - Defining thekey parameters contributing to the long run marginal cost ofgeothermal and providing a comparison benchmark on the keyunknownsJonathan Teubner*,* Business Development Manager, Petratherm Ltd, 1/129 Greenhill Road Unley SA 5061,AbstractThe development of the Australian geothermalindustry is unique as, especially in the earlyexploratory phases, the industry is driven bycapital investment raised via equity markets. Thisequity is supported in some instances by grantsprovided by the Commonwealth Governmentthrough its Geothermal Drilling Program (GDP)and, in the case of the commercial demonstrationof the technology, via the Renewable EnergyDevelopment program (REDP).The costs involved in the deep exploration phaseare significant, as are those involved in thecommercial demonstration of the technology. It isunderstood by the industry that until it has beenable to prove the commercial reality of its projects,all funding is required to come from equitymarkets or government grants.Australian geothermal companies are therefore, inessence, competing for a limited pool of riskcapital from the equity markets and grant moniesfrom the Commonwealth government. This moneywill be provided to those projects and companieswhich are expected to deliver a commercialproject and this provide the best returns.A number of reviews by Equity brokerage firmshave been provided recently (Morgan Stanley,RBS Morgans and Goldman Sachs JB Were)reviewing the prospects of the industry andcommenting on each of the companies within thesector. It is not the purpose of this paper to reviewindividual companies’ prospects, however thesepapers have provided a summary of the keyparameters for company attractiveness: Experience of the board andmanagement; Presence of project partners, and theexperience and size of such; Development options of the company(effectively the breadth of the projectjteubner@petratherm.com.au4(x,t) x l* *u * portfolio to diversify risk);l Funding capability through access togovernment grants and large partners ateither the project or equity level; Project quality assessed in respect oflocation, quality and riskThe ability to attract the capital (human as well asfinancial) to develop a project is ultimately afunction of the project quality. The project qualityin turn is a function of a number of competingfactors which will be discussed in more detailbelow, but is ultimately determined by theexpected cost of production of geothermal energyarising from the project.Australian geothermal companies therefore arecompeting for limited capital allocations on thebasis of their project’s forecast cost of production.This cost of production is typically evidenced by acalculated Levelized Cost of Electricity (LCoE) orLong Run Marginal Cost (LRMC).Recognizing the importance of this figure in theallocation of capital within the industry, theAustralian Geothermal Energy Association(AGEA) formed an Economics Committee todevelop a framework for the reporting ofEconomic parameters to the public to ensureconsistency across releases.This paper represents that framework whichdefines the key parameters in the economicassessment of the project, and thus whichassumptions need to be released to support anyeconomic releases. Further, it also provides datasets where available of what has been achievablein respect of these assumptions elsewhere in theworld to give interested parties an independentbenchmark of what has been previously achievedand prompting explanations from companies as towhere their assumptions sit on this frequencydistribution curve and why.The paper highlights the close cooperation thatexists between the investment community and thegeothermal industry and is an example of how theinformational needs of investors can be met by acooperative industry approach. It is an excellentguide to how companies can work with investorsin other emerging technology markets.u u uu 0426

COMMUNITY TRUSTEngagement & CommunicationAustralian Geothermal Conference 2010O. Boushel, *T.Hazou, C. Scott & Roger WindersSINCLAIR KNIGHT MERZ 590 Orrong Road, Armadale, 3143, VIC*Corresponding author: thazou@skm.com.auEffective community engagement from the start ofany project is critical to gaining public support.The geothermal industry in Australia facesparticular challenges in consulting with thecommunity because its technology is relativelyunknown. The challenge includes:1. a process of public education;2. a process of managing public perceptionsof risk.Add to these the range of views about climatechange and confusion around the role ofrenewable energy in general and the communityengagement task can seem daunting.Local community concern about geothermalprojects in their vicinity should therefore come asno surprise.Planning for community concern and taking anapproach that prioritises long-term relationshipbuilding with the community is critical tosuccessful project management. Thispresentation will provide a practical introduction tocommunity relations and consultationmanagement drawing on lessons both from thegeothermal and the broader renewables sector.The presentation will provide a brief outline of thecontext of community consultation, a consultationframework, as well as some practical Do’s andDon’ts.Keywords:Stakeholder relations, community relations,consultation and engagement; outrage mitigation;action research; public relations, Strategiccommunications; risk; consultation tools andmethods; program evaluation.IntroductionStakeholder and community relations (SCR) is acomplex and demanding aspect of corporatemanagement and design and construct projectdelivery. SCR is now very much part of theindustry and business landscape and is seen asan integral aspect to Corporate SocialResponsibility (CSR) and Public Relations (PR).In the coming pages, an introduction to PR, CSRand SCR and their differences will be provided,followed by an overview of communityconsultation practice. This will include anillustration of the distinct aspects of SCR and whatdifferentiates it from PR and reputationmanagement. This will be followed by anexplanation of community consultation principles,frameworks, and tools. Finally, an illustration ofhow SCR is used in a practical context will beprovided with an explanation of some ‘rules ofengagement’ or practice Do’s and Don’ts. This willprovide an ability to identify the practices andskills required to perform in the field ofstakeholder and community relations.Stakeholders and communitiesStakeholder and community relations (SCR) is afield of practice that demands expertise, andprofessionalism. It has emerged from a range ofindustries and professions and covers a widevariety of work practices and organisationalfunctions. Understanding this diversity andintersections will assist in grasping the competingexpectations of the field.Public relations (PR) can be defined as “theethical and strategic management ofcommunication and relationships in order to buildand develop coalitions and policy, identify andmanage issues and create and direct messagesto achieve sound outcomes within a sociallyresponsible framework” (Johnston & Zawawi,2004). Through this definition the PR profession isoften seen as ‘the custodian of reputationmanagement.’ As such PR often overlaps manyaspects of organisational management such asmarketing, media and crisis management,investor and community relations, internalcommunications, ethical conduct and strategicplanning.If PR generally (in all its forms) is aboutcommunicating a message and “staying onmessage” then SCR is the ability to adapt andchange the message in order to prioritiserelationships. SCR is about ‘winning trust’ andunderstanding the community in a manner thatimpacts on actions and words for the long-term.SCR professionals are ideally the first to identifyissues and challenges and must then manage thediscussion on behalf of the organisation (Moore,1996).Corporate Social Responsibility (CSR) has alsoemerged as an important corporate function inrecent years. CSR is recognition by corporationsand organisations that they exist in a socialenvironment. It is a reflection of the ‘triple bottomline’ and the need for organisations to addresspublic concerns of ‘accountability andresponsibility’ (Regester, 2001). In fact, CSR has427

Australian Geothermal Conference 2010developed to the point where it is no longer “Dono harm” and it is “Do good beyond the narrowlimits of making profit” (Argenti and Forman,2002). With this has come a sophisticatedunderstanding of public communications on CSRmatters. Aside from issues of ethical conductcommunities have grown to understand that PR isthe outward sign of an organisations inwardcharacter (Moore, 1996), resulting in a growingparadigm shift towards social responsibility.CSR has growing recognition by management ofits importance and influence in creating positiveoutcomes for organisations and achieving longtermgoals. As a result of CSR expectationsgovernments, organisations, and companies arenow investing time and energy into SCR. TypicallySCR exists on two levels (Tymson and Lazar,2002). 1. The local level: to help an organisationcommunicate with local leaders, residents andorganisations to facilitate positive relations andgood outcomes. 2. The corporate level: toacknowledge an organisation as a citizen within awider social framework.As a result, SCR has developed as a distinct areafor four important reasons. First, it is concernedwith outcomes and impacts rather than image andmessage. That is it is concerned less with howsomething will look in the media and more withwhat will happen as a result (Fletcher, 1999).Second, SCR emphasises ‘outside-in’ thinkingand the ability to see an organisation from theoutside point of view inwards (Regester, 2001).Third, SCR is crucially distinguished by “two-waycommunication”. Communication is an interactiveprocess between parties rather than a process of‘delivering on message’; it requires a dynamicability to listen, reflect and respond to concernsand desires (Forrest and Mayes, 1997). Finally,SCR is about developing and maintaining positivelong-term relationships (Forrest and Mayes,1997).Changing socio-political environmentCommunities are increasingly active andinformed. We live in a media environment whereinformation is more accessible than it has everbeen, and an increasing number of technologiesassist communities to network and (if necessary)mobilise against projects. Both new technologyand changes in legislation are creating moreopportunities for communities to engage in andinfluence individual project proposals.Globalisation described as a process by whichregional economies, societies, and cultures havebecome integrated and shrinking through a globalnetwork of communication, transportation, andtrade (Appadurai, 1996) has had an acute impacton community, government and corporateexpectations.Social media, mobile phones and the internetenables social networks to activate with a speedthat was unthinkable until recently. The expenseof media access no longer impedes communityparticipation in the public sphere as new media iswidely accessible for minimal cost, and reachesbroad audiences. Evolution of Web 2.0 throughapplications that facilitate interactive sharing,inter-operability, user-centred design, andcollaboration on the World Wide Web is also key(Tim O'Reilly 2005. retrieved 2006-08-06).Traditional modes of PR communication usingstatic and mass communication channels arebecoming less relevant and proving less effectiveas people rely on alternative sources ofinformation (represented below). One-waychannels of connecting with the target are “atypical monologue model with little if any openexchange of ideas, thoughts, or information. Theone-way arrows represent one way informationflows, as opposed to dialogue.” (Mark Parker.retrieved 2010-09-04).Figure 1: http://smartselling.wordpress.com/tag/web-20/In this context, and the growth of the ‘informationage’ there is a higher demand on corporations “tomanage the establishment and maintenance ofcredibility and trust” (Tymson and Lazar, 2002).Johnston and Zawawi (2004) also point out that areflexive and socially conscious management ofcommunications within a socially aware andinformed environment is critical. People todaywant to and do know more about organisationsgenerally. “More people want to know more aboutthe organisation they are working for, theorganisation they are buying from, and theorganisations they are investing in” (Tymson andLazar, 2002).In recent years, community engagement hasbecome part of business as usual for capitalintensive projects and their proponents. Whilestakeholder engagement has always beenrequired to some extent, the geothermal industry(along with others) is now being asked toformalise and improve its approach.428

Community expectationsThe increase in community activism andawareness has been facilitated by changes inlegislation that reflect the public’s desire to beinvolved in helping guide how their communitiesgrow and adapt to projects. Often projects nowhave federal, state and council triggers inplanning processes that require publicconsultation to be included in any planningprocess. Importantly these triggers mean that thesupport of key stakeholders such as governmentfor a project, hinges on what benefit the projecthas for the wider public.At the most basic level these new technologiesmay be restricted to email, blogs or basicwebsites. However, improvements incommunications technologies have allowed eventhe smallest groups to advocate for or expresstheir concerns about projects in very polished,convincing campaigns using technologies such asinteractive websites such as, Facebook, YouTubeor Twitter.These technologies are not only being utilised bythe public, but increasingly by developers andproject proponents to effectively engage a widerrange of people, improve the effectiveness ofcommunication strategies and gather criticalinformation from the public that may affect thecommunity. For example, VicRoads isundertaking a consultation process for its HoddleStreet Study by using an online tool called “Bangthe Table” to gather feedback from the widercommunity about key considerations in theirplanning study such as existing operations, therole of public transport and bicycle access.The wider public is also becoming bettereducated, improving their ability and desire toengage in policy and planning processes inincreasingly sophisticated ways. Further, thereare increasing expectations for companies tomeasure up to community expectations of socialresponsibility and to earn their ‘social license tooperate’. These expectations are often shared bythe bodies responsible for approving projects.This can translate into unexpected and unforseendemands on companies to deliver in areas thatwere previously viewed as outside their area ofresponsibility. Relevant recent examples of thisinclude the $5 million Regional Benefits Programundertaken as part of the Sugarloaf PipelineProject (Melbourne Water. retrieved 2010-09-02)or the AGL Hallett Wind Farm Community Fundwith total annual total grants $22,000 available forcommunities and communities in the NorthernAreas Council (NAC. retrieved 2010-09-02).Government expectationsGovernments also expect an in-depth level ofcommunity consultation as part of corporate andindustry performance.Australian Geothermal Conference 2010Directions to proponents for the preparation ofdevelopment approvals often include specificelements relating either to community consultationand/or social impact assessment.Local government in particular is establishingconsultation requirements as part of approvalsprocesses above and beyond regular publicexhibitions and notices.Funding and tender requirements now emphasisecommunity consultation obligations forcingindustry to respond to contractual and approvalsrequirements.Industry expectationsThere is a growing awareness by industry thateffective community engagement rather thanbeing a regulatory burden, provides opportunitiesto improve project delivery through the earlyidentification of issues, contribute to socialwellbeing, and to deliver on CSR andsustainability agendas.Corporations (venture partners, banks andfinanciers) now have high expectations thatcompanies will manage community relations in aproactive and responsible manner.For these reasons, community expectations ofcompanies to deliver open and accountableconsultation processes are now, as great as theyhave ever been. Community engagementstrategies for geothermal projects mustacknowledge and navigate through thisenvironment.Effective engagementGeothermal energy, like many renewable energysources, is a relatively new industry in someregions. In these cases communities are unlikelyto have an understanding of what potentialdevelopment projects involve. As a result, anycommunity engagement will also need to play aneducative role with the public, shaping communityperceptions of the industry.Consultation methodologyA number of effective consultation methodologiesare available to guide public engagement. Thetwo major approaches utilised in Australia are theOrganisation for Economic Co-operation andDevelopment (OECD) model and InternationalAssociation of Public Participation (IAP2)spectrum of engagement.The OECD model provides three differentapproaches to consultation depending on thecontext in which it occurs:• Notification• Consultation• Participation429

Australian Geothermal Conference 2010Notification in itself is not consultation as it is aone way process in which the organisationinforms the public. This is often the first step inconsultation where people are notified about theproject;Consultation involves the two-way flow ofinformation between the project and the widercommunity. This process may be used to allowgroups to voice their concerns about the project,ask questions or provide feedback.Participation involves the public in the decisionmaking process.The IAP2 spectrum, which shares a lot incommon with the OECD model, is fast becomingaccepted as a baseline by a range ofstakeholders in Australia and internationally.The IAP2 spectrum has five engagementapproaches:1. Inform2. Consult3. Involve4. Collaborate5. EmpowerEach type of engagement is tailored to desiredengagement goals as shown in the table below.ApproachInformConsultInvolveCollaborateEmpowerGoalProvide the public with informationabout a project, plan or actionObtain feedback from the publicabout a project, plan or actionContinually obtain feedback andconsideration from the public atseveral stages throughout theprojectWork with the public on the project,involving them in the planning anddecision makingProviding the public with decisionmaking power on a project.accommodate differences in the public and thesocio-political environment.Importantly, where a community engagementprogram has worked well for a project, it is criticalto review and reassess before reapplying it.Different communities, environments andtechnologies will require a different approach.Planned and structured approaches to communityconsultation not only produce the best results butalso leave a lasting impression and communityconfidence in corporate processes.Action researchA good practical approach to evaluating SCRmethods and tools is to adopt an Action ResearchMethodology (ARM). Action research is simply aform of self-reflective enquiry undertaken byparticipants in social situations in order to improvethe rationality and basis of their own practices,their understanding of these practices and thesituations in which the practices are carried out.(Carr & Kemmis, 1986).Action Research does not set out to answer ahypothesis or find a single ‘objective truth’. Itfocuses on a development process that engagesindividuals and enables change and improvedperformance.Action research can be done through a range of‘typical’ survey and interview research methods. Acrucial aspect is that each method should inform acircular model of inquiry. This allows for thebuilding of improvements into practice. To do thisit involves participants in a collaborative approachto investigation in order to resolve specificproblems or create systematic (change) actions(Stringer, 1999).The action research cycle, involves four keysteps; planning, action, observation, reflection:Figure 2: IAP2 Approaches and Engagement GoalsThese approaches are not mutually exclusive.Often, several different approaches are used withdiffering aspects of a project. For example, aproponent may inform the public about a project,consult landholders about land access, involvelocal government in traffic management plansduring construction, and empower landholders inthe reinstatement process.In the context of a single project, the project teamneeds to continually assess the tools andapproaches used for their efficacy with the public.Where appropriate, tools should be adapted toFigure 3: http://cadres.pepperdine.edu/ccar/define.htmlAction research is particularly well suited to SCRas its intent is not only to ‘get the job done’ but tofacilitate improvements. It does this by taking an‘organisational change’ approach based onparticipation and involvement.430

Consultation toolsThere are a number of tools available for use toengage the community. As shown in the IAP2spectrum below, these can be grouped accordingto the engagement approach used. Typicalexamples include:• Information sheets / community bulletins• Web sites, blogs and email• Community Information sessions• Focus groups, surveys• Public meetings and public comment• Workshops and advisory committees• Community reference groups• Citizen juries and community ballotsAustralian Geothermal Conference 2010decided to not undertaken any engagement or touse an inform approach to let any adjoininglandholders know. As such, a letter to localhouseholds about the project may be all that isrequired.However, a greenfield development that will occurin close proximity to adjoining properties mayrequire a different approach. The possibility ofcommunity concerns about issues such asproperty acquisition, noise, vibration and amenityissues may mean that failure to consult the publicmay result in community outrage, confusion andmisinformation. To avoid this, a consultingapproach may be deemed most appropriate withadjoining landholders. This may involve in personmeetings, a public information session or othertools that are appropriate in this context.Advantages/DisadvantagesDifferent engagement approaches create differentlevels of expectations within the public. Theseengagement approaches carry implicitundertakings. These are described as follows:Figure 4: http://iap2.org.auThese tools need to be selected on the basis ofthe engagement goals of the project. Engagementgoals should be determined by a number offactors, most importantly the possible risks to theproject, the organisation and its reputation.While risks are best determined in terms ofspecific issues relative to a project, they are oftena product of a number of factors, as follows:• The number and type of stakeholders• The characteristics of the community• Potential issues associated with the project(perceived or real)• The potential benefits of the project for thecommunity (perceived or real)• Stage of the projectFor example, undertaking a number of upgradesto an existing plant that may have short termconstruction noise impacts in a sparselypopulated during the day may be determined tobe a low risk activity from a communityconsultation perspective. Therefore it may beGoalInformConsultInvolveCollaborateEmpowerFigure 5: Source: http://iap2.org.auUndertakingWe will keep you informed.We will keep you informed,listen to and acknowledgeconcerns and provide feedbackon how public input influencedthe decision.We will work with you to ensurethat your concerns andaspirations are directly reflectedin the alternatives developedand provide feedback on howpublic input influenced thedecision.We will look to you for directadvice and innovation informulating solutions andincorporate your advice andrecommendations into thedecisions to the maximumextent possible.We will implement what youdecide.The tools used often vary according to theengagement goals. Using the right tools allows fortargeted community engagement that is costeffective, timely and addresses risks posed bypotential public outrage.Each approach has a number of advantages anddisadvantages when they are used with differentcommunities or in a particular context.Consideration of these aspects when determiningwhich tools to use is a worthwhile exercise. This431

should not deter taking a certain approach, butrather allow strategic consideration and planningfor some of the challenges that may face aparticular project.The following table contrasts the advantages anddisadvantages of each public participation goaland will help in evaluating program planning anddecisions.Goal Advantage Disadvantage• Inexpensive• Limited• Allows control of information caninformationbe meaningfully• Widecommunicateddissemination of • Communityinformationconcerns can beignored• Allows for• Can be difficult tocommunitymanage publicfeedbackexpectations• Inexpensive• Superficial issues• can reach large can be raisednumbers of• Potential forpeoplesmall vocal• Improvedgroups todecision making dominate publicconsultationInformConsultInvolveCollaborateEmpower• Provides greaterinsight into publicissues andperspectives andhow these wereformed• Createscommunityownership of theproject• Createscommunityownership of theproject• Increase publictrust in theorganisation• Can fosterinnovative localsolutions• Well suited tosmaller issues• Createsownership in thecommunity overthe project• Can increasecommunitysupport of aproject• can increasepublic trust in theorganisation• Can createinnovative localsolutions• Well suited tosmall localisedissues• Can be resourceintensive• Smaller crosssection of thepublic engaged• Can lead to directconflict with thecommunity aboutproject goals• Can be resourceintensive• Smaller crosssection of thepublic engaged• Public decisionsnot alwaysfeasible, leadingto conflictbetweenproponents andthe community• Can be resourceintensive• Smaller crosssection of thepublic engaged• Public decisionsnot alwaysfeasible, leadingto conflictbetweenproponents andthe community• Not suited totechnicalsolutionsFigure 6: Advantages and Disadvantage of IAP2 GoalsAustralian Geothermal Conference 2010Do’s and Don’tsWhile each consultation program must be tailormadeto particular project circumstances, it ispossible to draw out a number of ‘consultationdo’s and don’ts’ from past experiences; including:Trust and transparency• Following through on what is promised. If youcan’t commit to a date, action or event, thendon’t promise you will.• It’s never enough to say ‘trust me’ to thepublic, be prepared with evidence to supportany claims made.• Inform the community where and to whatextent they can influence a decision andwhere they cannot.Timeliness and planning• Provide the opportunity for the public toparticipate as fully as possible within thetimeframe established• Ensure that Community engagement is anintegral part of the project planning processNew technology• Don’t be afraid to use new technologies,particularly where they can help you engage• Don’t assume that everyone has access to oris comfortable with the internetIntegrated design processes• Community engagement needs to beintegrated into the design process. Involvingit at this stage allows for the identification oflikely issues early in the process where theycan be addressed in a cost effective manner.Failure to do this can result in costly revisionsand risks of project delays during theconstruction stage.Evaluating Choices• Don’t be afraid to ask for help from within theindustry or outside of it. If you’re not sure youare better off getting some specialist advice.• Do take a systematic approach toconsidering tools, methods and consultationapproaches.Respect and Recognition• Do acknowledge that local communities havean interest in your project and will want a sayand input.• Don’t treat communities as ‘stupid’ or unableto grasp technology; if you do they may goelsewhere and get the ‘wrong’ information.• Do try to simplify things and use common‘spoken’ language to explain difficultconcepts or technology.• Do involve the community early and establisha ‘bank of goodwill’ that you can draw on ifsomething goes wrong or a future presentsitself.432

SummaryAs a fledgling industry in Australia, geothermalenergy developers have an opportunity to ‘standon the shoulders of others’, drawing on lessonslearnt by other industries and sharing experiencesfrom within the geothermal sector. Thispresentation aims to pool that knowledge and linkit to the engagement challenges faced by theindustry today.Stakeholder and community relations is clearlyabout communicating ‘early and often’establishing associations, mutual worth andinvestment in publics (Forrest and Mayes, 1997).Investing in a long-term relationship to understandcommunity views, predict issues, reflect concerns,communicate on target, and maintain credibilitycan only be done with pro-active andcomprehensive relations (Ledingham andBruning, 2000). If SCR is about establishingcredibility and trust then maintaining positiveenduring relationships becomes the cornerstoneof SCR practice.This puts the industry in a unique position.Geothermal power doesn’t have a legacy ofnegative community perceptions based on noise,amenity or other factors and has the opportunityto create an enduring positive image madepossible by clear communication and a ‘mature’,and sensible approach to communityengagement.ReferencesAppadurai, A. (1996) Modernity at Large: culturaldimensions of globalization. Minneapolis:University of Minnesota Press.Argenti, P and J Forman (2002) The Power ofCorporate Communication, crafting the voiceand image of your business, Boston:McGraw-Hill Pty. Ltd.Brackertz, N., Zwart, Meredyth, D. and LissRalston (2005) Community Consultation andhard to reach concepts and practice inVictoria local government. Hawthorn: Institutefor Social Research, Swinburne University ofTechnology.Riel, M. (2010) Center for Collaborative ActionResearchhttp://cadres.pepperdine.edu/ccar/define.html(CFCAR. retrieved 2010-08-04).Fletcher, M (1999) Managing Communication inLocal Government, London: Kogan Page Ltd.Forrest, C and R Mayes, (1997) The PracticalGuide to Environmental CommunityRelations, New York, John Wiley and SonsInc.Australian Geothermal Conference 2010Globalization. (2010)http://en.wikipedia.org/wiki/Globalisation(Wikipedia. retrieved 2010-09-17).Johnston, J and C Zawawi (2004) PublicRelations: Theory and Practice, 2nd Edition,NSW: Allen and Unwin.Ledingham and Bruning, (2000) Public Relationsas Relationship Management, Mahwah, NJ:Lawrence Erlbaum Associates Inc.Melbourne Water Sugarloaf Pipeline Alliance(2008)http://www.sugarloafpipeline.com.au/content/about_the_project/regional_benefits.asp(Melbourne Water. retrieved 2010-09-02)Moore, S (1996) An Invitation to Public Relations,London: Cassell Pty. Ltd.Northern Areas Council and AGL (2010)http://www.nacouncil.sa.gov.au/webdata/resources/files/AGL_HWF_Community_Fund_Application_NAC_10.pdf (NAC. retrieved 2010-09-02).Parker, M. (2010)http://smartselling.wordpress.com/tag/web-20/(Mark Parker. Retrieved 2010-09-04).Regester, M. Managing Corporate ReputationThrough Crisis in Jolly, A (Ed), (2001)Managing Corporate Reputations, London:Kogan Page Ltd.Sandman, P. (1993) Responding to communityoutrage: strategies for effective risk. AmericanIndustrial Hygiene Association, Fairfax.Stringer, T. (1999) Action Research 2nd Ed.Thousand Oaks: SAGE Publications Inc.Tymson, C, Lazar, P and R Lazar (2002) TheNew Australian and New Zealand PublicRelations Manual, Chatswood: TymsonCommunications.O'Reilly, T. (2005) "What Is Web 2.0". O'ReillyNetwork.http://www.oreillynet.com/pub/a/oreilly/tim/news/2005/09/30/what-is-web-20.html.Retrieved 2006-09-17433

Australian Geothermal Conference 2010Geothermal energy legislation developments in QueenslandMelissa Zillmann* and Myria MakrasEnergy Industry Policy Division, Department of Employment, Economic Development and Innovation, 61 Mary Street,Brisbane Qld 4000Corresponding author: melissa.zillmann@deedi.qld.gov.auThe Geothermal Energy Act 2010 (the Act) waspassed by the Queensland Parliament in August2010. The Act has been primarily established toprovide a regulatory framework for the productionof geothermal energy.On commencement, the Act will repeal the currentGeothermal Exploration Act 2004 (the ExplorationAct) which solely regulates geothermalexploration. The new Act largely duplicates thecurrent exploration legislation, howeverexperience and feedback gained from operationof the Exploration Act has led to some beneficialchanges to the exploration regime.Importantly the Act introduces a regulatory regimefor geothermal production which will enablegeothermal exploration permit holders to progressto a production tenure when a suitable geothermalresource has been identified.This paper outlines the key changes to theexploration regime and the framework forgeothermal production included in the Act.Keywords: Queensland, geothermal energylegislation, geothermal production, productionthreshold, geothermal exploration, GeothermalEnergy Act 2010, Geothermal Exploration Act2004.Geothermal Energy Act 2010The Queensland Government recognises theimportance of geothermal energy as a valuableenergy resource that has the potential tocontribute significantly towards reducing theState’s carbon footprint. The QueenslandRenewable Energy Plan launched in June 2009outlined a roadmap for the expansion of therenewable energy sector in Queensland. The Planforecasts 250 megawatts of generation capacitybased on geothermal energy in Queensland’sgeneration capacity mix by 2020. Development ofthis Act is a key milestone that will supportdevelopment of the geothermal industry inQueensland and enable this forecast to be met.The objectives of the Act are to encourage andfacilitate the safe exploration for, and productionof, geothermal energy. This will be achieved byproviding for the grant of geothermal explorationpermits to explore for sources of geothermalenergy across the State and for the grant ofgeothermal production leases for large-scalegeothermal production.Where possible, the Act aims to ensure maximumconsistency with Queensland’s existing resourcebasedlegislation and geothermal legislation ofother jurisdictions.Changes to geothermal energyexploration in QueenslandLand available for geothermal explorationUnder the current Exploration Act, an explorationpermit can only be obtained when the State hasfacilitated a land release for competitive tender.This means the timing of exploration, and the landmade available, is at the discretion of the State.Under the new Act, applications for explorationpermits will be available via two processes.Firstly, the tender process provided in theExploration Act has been modified and is referredto as a ‘released area’ process. This will providean ‘application period’ where applications may bemade. Any applications within this period will beconsidered competitively after the period ends.Secondly, a new process which enables eligiblepersons to apply for exploration permits over landof their choice and at a time appropriate to them isintroduced by the Act. This process is referred toas an ‘over the counter application’. Theintroduction of this new application process willprovide greater flexibility in gaining access to landfor geothermal exploration activities and supportincreased geothermal exploration in Queensland.In addition to making more land available forexploration, the exploration regime under the newAct increases the maximum area that can beapplied for under an exploration permit. The areawill be increased from 200 sub blocks (600 km 2 )to 50 blocks (3,750 km 2 ) which brings theexploration permit area into line with other Statesand increases the chances of identifying acommercially viable geothermal energy resource.The term of an exploration tenure has also beenincreased under the Act from 12 to 15 years.Exploration ExemptionTo encourage use of geothermal energy for thepurposes of geothermal heat pumps for heatingand cooling buildings, the Act does not regulateexploration for these purposes. Exploration of thisnature will generally depend on shalloweroperations within the relevant property boundary,which will not have the same impact on theenvironment and landholders as other explorationactivities.434

Australian Geothermal Conference 2010Retention StatusIt is recognised that there may be delays betweenthe discovery of a geothermal resource and itscommercialisation. These delays could be relatedto factors such as investment attraction,development of infrastructure or securing supplycontracts.To address these issues, the Act introduces aretention status by way of a ‘Potential GeothermalCommercial Area’. This is similar to the retentionstatus provided for under the Petroleum and Gas(Production and Safety) Act 2004. The ‘PotentialGeothermal Commercial Area’ can be granted forup to 5 years but it can not extend beyond themaximum term allowed for in the underlyingexploration permit.The grant of a ‘Potential Geothermal CommercialArea’ may provide an exemption fromrelinquishment provisions and may enablesuspension of the exploration permit’s workprogram where appropriate. However, anevaluation program will be required to ensure thatthe exploration permit continues to be activelydeveloped and brought to production.Production TestingThe Act provides clarity that restricted productiontesting can be carried out under an explorationpermit to determine whether a geothermalresource is suitable for commercial development.The commercial grade of the geothermal energyresource will need to be quantified before anapplication for a production lease can be made.Before testing can commence, a test plan must beprovided and any relevant authorities will need tobe obtained.Transitional provisions for ExplorationAct permit holdersThe Act provides existing exploration permitholders a period of 12 months to transition to thenew geothermal energy framework after thecommencement of the Act. Whilst the Actcontains extensive provisions that automaticallytransfer tenures, decisions and processes to theAct, current exploration permit holders will need tocomply with the requirement to obtain anenvironmental authority. A period of 12 monthsfrom commencement of the Act has been deemeda sufficient timeframe to comply with thisrequirement.Geothermal production regimeThe primary purpose of the Act is to introduce aframework for geothermal production. Theprincipal means for allocating geothermalresources for production are via a productionlease.Under the Act, a production lease can only beapplied for on the basis of an existing explorationpermit. A person who is not an exploration permitholder can only apply for a production lease if it isdone jointly with an exploration permit holder orwith the consent of the exploration permit holder.These provisions give priority production rights toexploration permit holders provided they complywith the requirements of the Act.Scope of the Act (for production)The proposed geothermal energy regime providesfor a sliding scale of regulation of geothermalenergy production under several QueenslandActs, according to the scale of production.Activities that are small or medium scaleproduction will not be regulated by the Act.Furthermore, the Act is not intended to regulateany associated use of the geothermal energyproduced, such as in power stations or industrialactivities. To minimise regulatory duplication, theuse of geothermal energy will be regulated whenit triggers existing legislation such as theElectricity Act 1994 and the Sustainable PlanningAct 2009. For example, a production lease holdermay be granted rights to extract geothermalenergy which may be used to produce electricity,but approvals to supply electricity must beobtained under the Electricity Act 1994 and thebuilding of any necessary infrastructure must beauthorised under the Sustainable Planning Act2009.Small to medium scale production thresholdThe use of heat pumps (considered as smallscale production) will be encouraged under thisframework since their use is not regulated by theAct and its requirements to obtain a productionlease and provide development plans.Under the Plumbing and Drainage Act 2002,installation of pipes for the purpose of conveyingwater within premises may only be undertaken bya licensed plumber. It is proposed to amend thePlumbing and Drainage Regulation 2003 to allowthe Queensland Plumbing and Wastewater Codeto provide a framework for the regulation of theinstallation of geothermal heat pumpsGeothermal energy production below the largescale production threshold that is not via heatpump technology is considered to be mediumscale production and it will not be regulated by theAct. Instead it will be regulated if the geothermalenergy use triggers requirements under existinglegislation such as the:ooSustainable Planning Act 2009 (materialchange of use provisions)Water Act 2000 (taking or interfering withwater provisions)o Environmental Protection Act 1994(environmentally relevant activitiesprovisions)435

Australian Geothermal Conference 2010Large scale production thresholdThe Act only regulates the large-scale productionof geothermal energy. The Act provides that thethreshold for large scale production will beprescribed in the supporting geothermalregulation.Currently a power threshold which is nottechnology-specific and is flexible enough to applyto a variety of geothermal production applicationsis being considered to define the large-scaleproduction threshold. A power threshold providesa direct measure of the rate of production of thegeothermal resource that does not depend onhow the resource is being utilised, thus avoidingthe need to consider conversion efficiency.A primary policy consideration for setting thethreshold is that it should be set at a level lowenough to enable commercial geothermalproduction activities of potential importance to bemonitored by the Government. However, thisneeds to be balanced so that the threshold is sethigh enough to ensure that the uptake ofsmaller-scale geothermal activities which mayoffset carbon emissions are not unnecessarilycaptured or stifled by a requirement to have aproduction lease. Further consultation will beundertaken during preparation of the supportingregulation about how the large-scale productionthreshold will be defined.Geothermal production leasesA geothermal production lease must be appliedfor if the proposed activities will trigger thethreshold for large sale production. Geothermalenergy production is defined in the Act as therecovery of geothermal energy from beneath oron the surface of the land in which it is contained.The area required for geothermal production isnot as large as the area required for geothermalexploration. Hence the Act allows for a maximumproduction lease area of 25 blocks (1, 875km 2 ).A production lease applicant must prepare adevelopment plan setting out how the geothermalenergy resource will be developed, and theproposed rate of energy production. The grant ofa production lease requires the holder tocommence commercial production within 2 yearsof its grant. To allow the ongoing monitoring ofresource development, the holder will be requiredto submit further development plans during theterm of the lease.A geothermal production lease has an initialmaximum term of 30 years. Subsequent renewalsof the lease are limited to a maximum of 20 years,with no limit on the number of renewals that canbe made.Sub-lease of geothermal production leaseA holder of a geothermal production lease maysublease an area of the lease to a third party. Thesublease is subject to approval by the Minister.Where a sublease has been approved andregistered, the holder of the geothermalproduction lease must continue to comply withany terms and conditions of the lease includingthe area of the sublease.RoyaltyThe Act provides that a royalty is payable on theuse of geothermal energy. The royalty rate will bespecified in the supporting regulation.The Queensland Government recognises that thegeothermal industry is in its infancy and facestechnological and financial barriers todevelopment. Key features of the royaltyprovisions include a royalty holiday and royaltyfree threshold that are aimed at reducing costs togeothermal production during the early start-upstages.A royalty holiday that provides a full exemptionfrom the payment of any royalties will be provideduntil 2020. Whilst the first half of the next decadeis expected to be characterised by increasedexploration activity, the later part of the decadeshould see geothermal projects coming intoproduction and benefiting from this initiative.After 2020, a royalty-free threshold will ensurethat in the early years of production when financialmargins may be low, a royalty is not payable if thewellhead value is below the royalty-free threshold.General authority provisionsA number of key provisions apply to bothexploration permits and production leases. Theseprovisions are outlined below.Environmental managementPersons proposing to carry out geothermalactivities are required to obtain the relevantenvironmental authority under the EnvironmentalProtection Act 1994 before an exploration permitor a production lease can be granted.Exploration activities conducted under anexploration permit will be classified as level 2environmentally relevant activities under theEnvironmental Protection Regulation 2008(environmental regulation), unless these activitiesoperate in an environmentally sensitive area orare associated with an environmentally relevantactivity stated in schedule 2 of the environmentalregulation.Production activities conducted under aproduction lease will be classified as level 1 orlevel 2 environmentally relevant activities,depending upon the scale of production andassociated environmental harm. This approach436

Australian Geothermal Conference 2010largely ensures consistency with existingQueensland resource-based legislation in relationto environmental management requirements.Where the geothermal energy produced is usedfor another activity that is an environmentallyrelevant activity under the environmentalregulation e.g. aquaculture, that activity willrequire the relevant authority under theEnvironmental Protection Act 2009 and theenvironmental regulation.Water ManagementThe Act does not create any rights to water. Anyrelevant authorisations under the Water Act 2000must be obtained before an exploration permit ora production lease can be granted. The waterauthority will ensure that the taking of water forgeothermal energy exploration or production ismanaged within a whole-of-catchment plan thatensures environmental flows are maintained.The grant of a geothermal tenure will not createany preference or priority for the grant of a waterentitlement. Geothermal tenure holders will betreated the same way as any other potential waterusers and may need to pay the market price for awater entitlement in order to consume water in theprocess of exploring for or producing geothermalenergy.The Act also mandates that an exploration permitor production lease applicant must assess anypotential structural or other impacts on aquifers ofthe carrying out of the proposed activities. TheMinister must consider this issue when decidingwhether to grant the exploration permit orproduction lease.SafetyGeothermal drilling and related activities aresimilar to those for petroleum and gas activities.Therefore, the safety and health aspects ofgeothermal energy exploration and production willprimarily be regulated under Queensland’sPetroleum and Gas (Production and Safety) Act2004 (P&G Act).Currently, the definitions of operating plant undersection 670 of the P&G Act only apply to certainauthorised activities carried out in a geothermaltenure area.In practice this would mean that other authorisedactivities on the tenure are regulated under theWorkplace Health and Safety Act 1995 (WH&SAct). The P&G Act and the WH&S Act have beenamended to clarify the different responsibilities ofthe Department of Employment, EconomicDevelopment and Innovation and WorkplaceHealth and Safety Queensland in relation tosafety management on a geothermal tenure.The construction of new operating plant willcontinue to be covered by the WH&S Act.Geothermal exploration activities that involve, forexample, seismic testing and drilling, will continueto be regulated under the P&G Act. Otheractivities associated with geothermal exploration(for example, construction activities) will continueto be regulated under the WH&S Act.Petroleum type drilling does not occur in theproduction of Hot Sedimentary Aquifer (HSA)sources of geothermal energy. As such, the Actwill not apply the safety provisions contained inthe P&G Act to HSA geothermal production,regardless of the scale of the operation. Theseoperations will continue to be covered by theWH&S Act.Overlapping authoritiesThe Act includes provisions about overlappingauthorities under different Acts. For example, aproponent may apply for a geothermal productionlease over an area where another resource tenureexists under the P&G Act, the Greenhouse GasStorage Act 2009 or the Mineral Resources Act1989.The purpose of allowing overlapping authorities isto ensure the optimum use and appropriatemanagement of the area’s resources and isenshrined in all Queensland resource-basedlegislation.The decision as to which resource authority getspriority rests with the Minister, who must considerspecific criteria, including submissions made byeach authority holder who would be affected bythe grant of a geothermal tenure over the sharedarea, as well as the public interest.The Act does not allow for an overlappingauthority to be provided for two geothermalenergy tenures. However, the Act allows ageothermal production lease holder to sub-lease apart of the land within the lease to any party thatis an eligible person, as defined by the Act.Land AccessThe Act contains landmark changes to the way anumber of Queensland resource tenure andauthority holders may access private land andhow compensation arrangements are arrived at.Under the changes, geothermal tenure applicantsand holders will have new responsibilities inrelation to land access.The Act introduces a requirement for allgeothermal authority holders to comply with aLand Access Code as a condition of the authority.The Land Access Code is structured in two parts:Part 1 which outlines non-statutory best practiceguidance for communication between authorityholders; and Part 2 which describes mandatoryconduct conditions that must be complied with byauthority holders.The Act introduces activity thresholds thatincrease the level of regulation as the intensity or437

Australian Geothermal Conference 2010impact of the proposed activity increases. This isa sensible measure that balances the level ofregulation with the level of impact.Early exploration activities may have minimalimpact on private or public landholders. The Actdefines these as ‘preliminary activities’ whichmight include taking rock, water or soil samplesand walking the land.For preliminary activities, a geothermal tenureholder will be required to give each owner oroccupier of land within the tenure area an entrynotice and copy of the Land Access Code at least10 days prior to entry. Access for these activitiesdoes not require a Conduct and CompensationAgreement.When the impacts of proposed activities on theprivate landholder or occupier are moresignificant, a higher level of accord is required.These activities are referred to as “advancedactivities”.To access land to conduct advanced activities,geothermal tenure holders must either be party toa conduct and compensation agreement or adeferral agreement with each ‘eligible claimant’ forthe land or have made an application to the LandCourt to determine compensation. As defined inthe Act, an ‘eligible claimant’ is each owner oroccupier of private land or public land that is in thearea of, or is access land for, the tenure. Whereagreement is not reached, the frameworkprovides for conferences or independentalternative dispute resolution (ADR) to be held.To assist in the negotiation process, standardconduct and compensation and deferralagreements have been developed in consultationwith resource and agriculture industry peakbodies, along with a negotiation tip sheet forlandholders.Conferences or ADR will become a legislativerequirement if agreement negotiations breakdown. Matters cannot be referred to the LandCourt for determination until a conference or ADRprocess has occurred. This requirement is aimedat resolving disputes early at the local level,without the expense of legal resolution.A detailed compliance and enforcement strategyis currently being developed to ensure that thesenew land access provisions achieve their fulleffect on the ground, particularly with respect tothe Land Access Code.Development of geothermal energysubordinate legislationThe Act will commence by proclamation. It isanticipated that the Act will commence in the laterpart of 2011. During the intervening period,supporting subordinate legislation, guidelines andtraining will be developed and systems upgradeswill be carried out in preparation for the Actcoming into force. There will be opportunities forindustry involvement in the development of thesubordinate legislation in late 2010/early 2011.438