Sarver and Jordan: Maintenance at Millstone

Nuclear Power Plant Maintenance Special SectionTHE NUCLEAR NEWS INTERVIEWSarver and Jordan:Maintenance at MillstoneIn combined service, Steve Sarverand Skip Jordan have worked forDominion (and its predecessor company,Virginia Power) for more than45 years. Sarver is director of operations and maintenanceat the Millstone nuclear power plant, and Jordan is theplant’s nuclear engineering director.Sarver has been with the company for 26 years and hasserved at Millstone since January 2003. He worked mostSteve Sarver (left) and Skip Jordan: On the turbine deck at theMillstone-2 nuclear power plant. Behind them is a new turbine that willreplace an existing component during an outage in October.The plant’s maintenance staff has become moreeffective at keeping emergent work from having animpact on the scheduled work.recently at Dominion’s Surry nuclear plant. Jordan, withthe company for 20 years, has been in his current positionsince last November. He has been around the circuitwith extensive experience in power generation, havingworked variously at the company’s fossil, hydro, and nuclearplants.Millstone, in Waterford, Conn., has two operatingunits. Millstone-2 is an 869-MWe (net) CombustionEngineering pressurized water reactor, and Millstone-3 is a 1136-MWe (net) Westinghouse PWR. A thirdunit on the site, Millstone-1, was a 660-MWe (net)General Electric boiling water reactor that was retiredin August 1998.Sarver and Jordan discussed how maintenance haschanged at Millstone since Dominion purchased theplant from Northeast Utilities in March 2001. The conversationalso delved into the TIP (Top Industry Practice)/FramatomeANP Vendor Award that Millstone receivedin May from the Nuclear Energy Institute. Theaward honored Millstone for its testing process thatsearches for possible leaks in control rod drive mechanisms.The process that the Millstone staff developed isan ultrasonic inspection technique that assesses the relevantsurface area of the reactor vessel head and penetrationnozzles. Using this method, plant technicians candetect leaks, verify the integrity of the carbon steel vesselhead, and perform a full volumetric testing of thenozzle base material.The interview was conducted by Rick Michal, NNsenior associate editor.40 N U C L E A R N E W S October 2003

CP-1 Special SectionHow have vessel-head examinations differedfor Millstone-2 and -3, considering that Unit2 was manufactured by Combustion Engineeringand Unit 3 by Westinghouse?Jordan: The inspections are not primarilydifferent because the units weremanufactured by different companies. Anydifference has more to do with susceptibilitydetermination and the significanceassociated with the primary water stresscorrosion cracking (PWSCC) issue. TheNuclear Regulatory Commission’s guidelinesplace various plants in different categoriesfor vessel head cracking susceptibility.At Millstone-2, we’ve done a fullvolumetric ultrasonic examination of eachnozzle from under the head, as well as aportion of a bare metal visual examinationon top of the vessel head. At Millstone-3,we’ve done only the bare metal visual inspectionat the top of the head. That is theonly difference between the two units atthis point.What susceptibility category is each unitin?Jordan: Unit 2, which started commercialoperation in December 1975, is in thehigh-susceptibility category for PWSCC.Unit 3, which started commercial operationin April 1986, is in the lower susceptibilitycategory. Operating age and the operatingtemperatures and pressures are the primaryconsiderations for putting the plants in thetwo separate categories. Both units had vesselhead inspections during their most previousrefueling outages. Unit 2’s was completedin February 2001, and Unit 3 in April2002.Has Dominion made a determination aboutreplacing the vessel heads at Units 2 and 3?Jordan: The vessel head replacement forUnit 2 is scheduled to take place in 2005.For Unit 3, we have no current plans to replacethe reactor vessel head.Could you talk about Millstone’s awardwinningtechnique for searching for controlrod drive mechanism (CRDM) leaks?Jordan: The process is known as theLeak Path Detection Technique. It assessesthe contact surface area of the interferencefit region between the penetration nozzleouter diameter surface and the carbonsteel vessel head penetration internal surfacearea. As we were preparing for Unit2’s outage, we looked at techniques thatwere being used at other utilities. All theother tests relied on bare-metal visual inspectionson top of the head, or surface examson the J-groove weld. Since neitherwere attractive options, we evaluated alternativenon-destructive examination methods.We were able to demonstrate from thedata that had been collected at the CRDMsthat if the contact surface of interference fitwas disturbed by erosion, corrosion, pitting,or the deposit of foreign material, it wouldbe detected by changes in the acoustic energy.In other words, the geometry of theacoustic energy would be significantly affectedby that disturbance. We would pickit up with the UT beam and be able to tellwhat kind of problem was in that area.Was this technique developed in-house?Jordan: Yes. We took a lot of data fromother utilities that were doing UT techniques,looked at that data, and decidedwhat would work for us. We put that intoplay and demonstrated with a vendor usingthe technique that the Leak Path DetectionTechnique was a viable option. The actualInterview: Sarver The and First Jordan PileUT was not developed in-house, but the useof the technique to prove viability for thisparticular application was done in-house.Has the technique spread to other PWRs inthe industry?Jordan: Yes. Other utilities are using it,hence our receipt of the TIP Award. Peoplerecognized that it’s a good thing and they’restarting to use it.Who at your plant was primarily responsiblefor the technique’s development?Jordan: It was developed primarily in theengineering ranks. The people I credit are theones who actually received the TIP Award.October 2003 N U C L E A R N E W S 41

Nuclear Power Plant Maintenance Special SectionOur nondestructive examination specialistMike Stark was involved in that, as were hissupervisor Harvey Beeman and the projectmanager for the reactor vessel head, Tim Petit.They were all looking at a way to get theinspection done for the least amount of dose,the least amount of time, and the leastamount of physical interference work to bedone on the head. The Leak Path DetectionTechnique is the process they came up with.Has maintenance improved in any waysince Millstone was acquired by Dominion?For example, do you share personnel withthe company’s North Anna and Surry“We are not yet to the pointof sharing significant numbersof people with North Annaand Surry, although [they]do a lot of sharing betweenthemselves.”plants, or purchase better equipment becauseof economies of scale?Sarver: We are not yet to the point ofsharing significant numbers of people withNorth Anna and Surry, although those twostations do a lot of sharing between themselves.Geographic proximity is the biggesthurdle there, in that we’re in Connecticut andthose units are in Virginia. So, the outagesof all of Dominion’s nuclear units are notnecessarily coordinated due to the differentgeographic regions in which we operate.I think our primary improvement in maintenancesince the acquisition by Dominionhas been the combining of the maintenancestaffs at Millstone. Before the acquisition, itwas recognized that we needed to bring theUnit 2 and Unit 3 maintenance staffs togetherso that we’d get economies of scaleacross the site, but it was never done. So,prior to the acquisition, there had been separatemaintenance staffs assigned to eachunit. It was only after Dominion purchasedthe site that the staffs were merged.In addition, we’ve adopted a team conceptrather than having the classic maintenanceorganizational structure of electrical, mechanical,and I&C. We also have specialtyteams at Millstone that focus on specific areasof the station or specific systems. Wehave a secondary team, an auxiliary team,and an NSSS (Nuclear Steam Supply System)team that focuses on specific groups ofequipment and thereby develop expertise inthe tooling, equipment, and equipment performancein those areas, such as our intakestructures, intake, and circulation waterequipment. It’s been a particular challengefor us. We have a specific team assigned thatpicks up nearly all of that maintenance. Thatamount of continuity, familiarity with theequipment, and how it’s worked and how it’sperforming is a significant improvement forthe site. We’re continuing to build on that.So, that is the difference between ourselvesand North Anna and Surry, in termsof how maintenance is organized. We’relooking to share lessons learned across ourthree sites for that.Are you saying that North Anna and Surry,which are two-unit sites, have maintenanceteams devoted to each unit at the sites?Sarver: No. Eachstation has one maintenancedepartmentthat services bothunits at the site. Theynever did have unitizationat Surry andNorth Anna. Whatthey don’t do that wedo at Millstone isthey don’t generallyhave a team conceptassigned to specificequipment areaswithin the plant. Ofcourse, they do it in limited fashion, but notnearly to the degree that we have organizedat Millstone.At Surry and North Anna, they have asingle maintenance organization focusedfor each two-unit site. They essentially havedual-unit sites—a common control room,and many of the facilities are common toboth of the units at North Anna and Surry.Millstone also is a two-unit site, but withtwo separate units and two separate controlrooms. Nearly all of the supporting facilitiesare separate for our units. That is a significantdifference there.Was it difficult to merge the two maintenancestaffs at Millstoneso they couldwork on both units?Sarver: I’m certainit wasn’t easy.The merger tookplace before I startedworking here. I thinkthe area where westill have a good bitof specialization associatedwith the units is in the I&C area.Dealing with the specific Combustion Engineeringversus Westinghouse processesand protection systems still requires a lot ofspecialization. It’s the most difficult area forbringing a common group up to par inknowledge and skills to span those differentprocess and control designs. With the electricaland mechanical work, it’s been far easierto bring that maintenance crew togetherand have people qualified to work on Unit 2Interview: Sarver and Jordanbreakers or Unit 3 breakers, or Unit 2 chargingpumps versus Unit 3 charging pumps.It’s been a far easier thing. So, we still dohave some specialization in our I&C ranks.Do you employ “FIN” teams—“Fix it now”teams—at Millstone?Sarver: Yes, we do. They’ve been in existencefor about five years now. We havetheir work scope very well defined in termsof what the FIN teams can and should workon. They are a very effective group at doingminor maintenance and high-prioritywork items that are more difficult to fit intoa normal crew’s schedule. We have both aday shift FIN team and one for overnightwork. Actually, our night shift maintenanceteam acts as a FIN team. We don’t schedulea large amount of routine work for our nightshift when we’re not in an outage, so ournight shift works as a FIN team to pick upemergent work from both units. They performthat maintenance if it’s within theirscope to do so.How many personnel are on the day shiftFIN team?Sarver: I estimate the day-shift FINteam consists of about six to seven people—asupervisor, two or three mechanics,two electricians, an I&C tech, and an operator.We always have an operations individualassigned to the FIN team to facilitateits getting to work in a quick fashion.Has a maintenance backlog ever been anissue at Millstone?Jordan: I think there are two areas wherethe FIN team has really helped. The first isworking off the backlog. The other is on protectingthe schedule and allowing the workthat we have scheduled to proceed. In otherwords, the emergent work that would normallycome in is being protected from havingan impact on the regular work crews. Inthe past, it was a challenge for us to handle“[T]he area where we stillhave a good bit ofspecialization associated withthe units is in the I&C area.”both of those. We found we were losingsome of the scheduled work and were havingto shift resources to emergent work. Nowwe’re much more effective in not letting theemergent work affect the scheduled work.What currently is your biggest challenge inmaintenance at Millstone?Sarver: I think our biggest challenge is theeffective coordination of maintenance resourcesin staffing and the scheduling of42 N U C L E A R N E W S October 2003

Nuclear Power Plant Maintenance Special Sectionwork. After Dominion acquired Millstone,and after the site went through a number ofother organizational changes, the maintenancestaff had a lot of new supervisors anda good bit of movement in maintenance staffresources. The biggest challenge was effectivecoordination of resources to focus on thework. We obviously have challenges inequipment obsolescence, in our ability tomaintain adequate supplies of spare materials,and in performing work on equipment thatgrows more obsolete by the day. And ourwork management system could stand someimprovement, too. But right now our biggestchallenge is the coordination of maintenanceresources to handle the work load.How has the role of the maintenance managerat your plant changed in the past fewyears?Sarver: Going back to bringing the twounits together and establishing a team conceptis the greatest change for the maintenancemanager. Part of that change is keepingup with adequate maintenance on sitewhile also addressing specific equipmentareas.Jordan: I don’t think that challenge isspecific just to the maintenance manager.Moving from the regulated to the deregulatedenvironment, the business-acumen focushas shifted, too. An individual in chargeof a department now realizes the impact thathis budget has on the overall station. Thatindividual has a keener awareness in termsof cost and overtime expenditures than waspresent in the regulated environment.Has maintenance changed in any way in thepost-9/11 world?Sarver: We see a direct effect in the wayplant access has changed in regard to thedelivery of materials to the site. It has becomea more cumbersome process simplybecause our site access has been modifiedto provide for additional security measures.While it has had an impact on the timelinessof materials to the site, I would not call it asignificant impact. We have figured outhow to get materials to the site in an effectivemanner so it has not degraded our abilityto perform maintenance. Beyond that,we have not identified anything significantin terms of additional maintenance activitiesdirectly related to 9/11.Are there any regulatory requirements fromthe past few years that have caused extraburden on maintenance?Sarver: We could point to a number ofindirect regulatory initiatives that are goingto create additional maintenance work. TheAlloy 600 inspection program is the largestone that is going to greatly expand our needto evaluate Alloy 600 components and componentlocations. There will be more detail inInterview: Sarver and Jordanthe way we evaluate the observance of boricacid on components in terms of its sourceand its impact on components. Beyond that,even if I point back to the NRC’s MaintenanceRule, EQ requirements, and a numberof those other regulatory initiatives, Iwouldn’t characterize them as creating anyadditional burden on maintenance. In fact, Ibelieve they’ve provided us with a frameworkto better prioritize our maintenance.In terms of maintenance burden, no, Idon’t think any recent regulatory requirementshave specifically implied an additionalburden on maintenance. Clearly, it’sgoing to change some of the priorities inhow we evaluate deficient component conditions,especially when it comes to boricacid and Alloy 600. Beyond that, I don’t believewe can point to any significant additionalburden from a regulatory standpoint.What are some of the Alloy 600 componentsthat you have to inspect?Jordan: Pressurizers, heater penetrations,primary hot leg and cold leg, andRTD thermal weld locations. We also havea number of safety injection system RTDand instrumentation small-bore penetrationsto the primary piping that have to beinspected. It is a significant expenditure interms of dose, dollars, and time. If the inspectionreveals some degradation and wemake the decision to change out whatever44 N U C L E A R N E W S October 2003

Nuclear Power Plant Maintenance Special Sectionpart it is, that new part would be of Alloy690 material, which is more corrosion-resistantthan Alloy 600.Has either Unit 2 or Unit 3 had to replacesome component for degradation reasons?Jordan: To date, we have not. We havewhat we call a “mensa” clamp, which is anapproved ASME Code temporary repair, ona couple of our pressurizer heater penetrations.We have not had to replace specificcomponents. For an upcoming refuelingoutage for Unit 2, we have set up contingencyplans for mensa clamps and somesmall-bore RTD penetrations that we mayneed, just in case.“We do . . . specializedtraining for the main steamisolation valves.That’s aunique setup in an areawhere we had problems withsolenoids that failed.”Are you trying to reduce your maintenancebudget? If so, how?Sarver: Yes, we certainly are. Twothings I would point out. First, we havemore effective planning than we used tohave. Increasing the productivity of ourmaintenance staff is the largest single impactwe can have on budget reduction andbudget effectiveness. Anytime we have ajob that doesn’t go exactly as plannedwhere we haven’t prestaged the appropriateparts, or have tagging issues—anythingthat gives us less than a perfect executionplan on a work order or job—theefficiency of a maintenance organizationis affected. That is where we are placingthe largest amount of emphasis. The bottomline is better planning for maintenancework.The second area where we need significanteffort—and where we’ll apply resourcesin the future—is in PM (preventivemaintenance) optimization. PM optimizationis something that can save us significantresources. Our PM program, while it’svery thorough, has some inefficiencies andhas not been optimized to ensure that we aredoing preventive maintenance when itneeds to be done, at the frequency it needsto be done, and to the extent it needs to beperformed on particular pieces of equipment.We can have a large impact on themaintenance budget by simply optimizingthat preventive maintenance and reducingthe amount that we’re doing on equipmentthat doesn’t need to be taken out of service.Have there been any significant changes inthe way your maintenance staff interfaceswith other organizations on site, such asoperations or engineering?Sarver: Yes, and I’ll point to two significantareas. First, operations now has asingle work control group assigned to it.That group covers both of Millstone’s units,and it provides the maintenance interfacewith operations for the entire site. Prior tothe group’s being established several yearsago, maintenancewas communicatingwith and interfacingwith different managersat the shiftmanager level.Establishment ofthe group has helpedfocus the maintenancestaff on themost important prioritiesacross bothunits. If we have 10jobs to do today, half on each unit, it maynot be appropriate that we work the first jobon each unit. So, the group has been effectiveat helping focus the maintenance resourceson day-to-day activities that affecteach unit.In addition, the engineering organizationhas sought to collectively bring a single focusto maintenance priorities on site. This ishelping maintenance when it comes to establishingthe correct priorities for whatneeds to be worked. We’ve recently formeda Station Equipment Reliability Team, calledSERT. It was formedat the initiative of theengineering group.SERT is establishingsingle prioritiesfor the site whenit comes to the consolidationof MaintenanceRule AlphaOne issues. When acomponent or a systemis placed intothe Alpha One category,it requires aspecific action planto restore it to Alpha Two status. Those issuesneed to be assigned priorities.A second change in how we interface withother organizations on site deals with operatorwork-arounds. Those work-aroundsidentify specific equipment deficiencies thatgive the operations group some additionalwork or complication. The deficiencies needto be properly prioritized. In that regard, weneeded to have a group that provided a singlefocus and a single set of priorities for themaintenance staff or we’d be working for toomany masters. The SERT, which is participatedin by engineering, operations, maintenance,planning, and several other key linefunctions on site, is now providing that forInterview: Sarver and Jordanus. We think SERT is not only unique, butalso a very useful team force at Millstone.Could you talk about any specialized trainingof Millstone’s maintenance personnel?Sarver: We do have some uniqueequipment here and we have unique trainingfor that equipment. What I would pointto is our significant focus on leadershiptraining for the maintenance supervisorystaff. This has been an area where we knew“Increasing the productivityof our maintenance staff isthe largest single impact wecan have on budget reductionand budget effectiveness.”that our maintenance leadership needed tohave a broader set of skills when it came toemployee relations, as well as the managementof our own particular areas within themaintenance group—business acumen, employeerelations, etc. We’re putting a largerfocus on not just getting to the maintenancemanager and his immediate directreport staff, but down to the first-line supervisor,and down even farther to lead craftpersonnel within groups that are being providedleadership training. The goal is tobroaden their perspective on how to bettercoordinate work activities and the importanceof their work as it relates to the overallplant.Could you give an example or two of someof your unique pieces of equipment forwhich workers need specialized training?Jordan: We do some specialized trainingfor the main steam isolation valves.That’s a unique setup in an area where wehad problems with solenoids that failed. Weput a design change in place to rectify thatcondition.Also, we have some flow-scan equipmenton which we provide specialized trainingand certification.Another area for us is acoustical monitoring.We have within the engineering organizationa condition-based maintenancegroup. We have some acoustical monitoringequipment we use, and we need specializedtraining to be able to use it. Theacoustical monitoring is used for welded-incheck valves. The monitoring gives us backan indication that is different from thenorm, and then we know something iswrong. For example, we have a baselinesignature that we’ve taken on some of theequipment, similar to what we do with vibrationmonitoring. That baseline signatureis the norm. The acoustical monitoring pro-46 N U C L E A R N E W S October 2003

Nuclear Power Plant Maintenance Special SectionInterview: Sarver and Jordanvides a difference from our baseline conditionto another condition, so it advises usthat something has changed in that component.We then can make a decision whetherto open up that piece of equipment and lookinside. In some cases, we’ll have to replacethat equipment.The vibration monitoring also helps usdetermine how rapidly we need to do work.It can give us a rate of degradation or thesignificance of the degradation. With informationgained from vibration monitoring,we’re able to schedule a piece ofequipment for work or put it in an emergentwork condition.Do you have any provision for solicitingsuggestions from plant and contractor personnelfor practical improvements in maintenancemethods?Sarver: We have a few things. One isour site-side reporting system, which accommodatesall kinds of condition reportingfor the site. To make that system work, wehave trained our people to identify deficienciesand improvements for the site. Andwe do get significant feedback for improvementsin maintenance methods. Itcould be related to a specific task where amechanic felt the job might go better ifsomething were changed for the next time.Or it might be a contractor who has seensomething we do that could be improvedwith an idea he’s brought from another sitehe’s worked at. Our site-wide reporting systemis a single repository that has the capabilityto allow us to sort, segregate, track,and trend feedback, whether it is improvementitems or factual information abouthow a job played out.Also, during outages where there are significantnumbers of contractor personnel onsite, we have lessons-learned tools that areavailable. We especially focus on our contractorstaff to give us feedback on evolutionsor activities performed during outagesso we can make improvements prior to thenext outage. All those suggestions are evaluatedby our outage integration teams tomake certain that we’re taking the benefitof those lessons learned.A final example is something we recentlystarted called our work week critique, althougha number of other plants have beendoing something similar. The work weekcritique takes a snapshot of all of the workthat was planned and executed during theprevious week. We review that work withthe maintenance, operations, engineering,and outage and planning staffs to see wherewe were successful or not successful, andwe evaluate where improvements can bemade for later work. This is a weeklyprocess that tries to capture fresh and detailedinformation about the successes orfailures of the previous week’s maintenancework. We factor it in so that we don’t makethe same mistakes twice.48 N U C L E A R N E W S October 2003