ML17256A403

Evaluation of Alternative Reactor Coolant Pump Trip Criteria for Re Ginna Nuclear Power Plant.

ML17256A403
Person / Time
Site:	Ginna
Issue date:	11/30/1982
From:	LEWIS R N WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML17256A400	List: ML17256A399 ML17256A401 ML17256A402 ML17256A403 ML17256A404 ML17256A405
References
3203Q:1-111582, NUDOCS 8211290439
Download: ML17256A403 (131)

Text

.E,ATTACHMENT BANALYSISOFPLANTRESPONSEDURINGJANUARY25,1982STEAMGENERATOR TUBEFAILUREATTHER.E.GINNANUCLEARPOWERPLANTNOVEMBER, 1982Preparedby:E.C.Volpenhein II'estinghouse ElectricCorporation NuclearEnergySystemsP.O.Box355Pittsburgh, Pennsylvania 15230PreparedforRochester GasandElectric89EastAvenueRochester, N.Y.1464982ii290405000244821122pDRPQOCKpDp32740:1/111782 TABLEOFCONTENTSSECTIONPAGEABSTRACTLISTOFTABLESLISTOFFIGURES~~~~~~~~.1~~'~~~~~11~~~~~~~oillI.INTRODUCTION 1II.ANALYSISOFPLANTRESPONSE~~~~~~II.lSystemsAnalysisCodeII.2PlantDataII.3InitialLeakRate'II.4Pre-tripSystemResponse~II.5Post-trip SystemResponse5.1PrimarySystemPressure5.2ReactorCoolantFlow5.3ReactorCoolantTemperature 5.4Pressurizer Level5.5BreakFlow5.6ReactorCoolantVoiding~5.7SteamGenerator OverfillII.6LongTermRecovery-I~~~~~~2~~245~~7162128335255586062III.SUMMARYANDCONCLUSIONS

...................65REFERENCES

............................67AppendixA:InitialLeakRateCalculation

~.-........

~~68AppendixB:BestEstimateBreakFlowModel............70:AppendixC:Calculation ofUpperHeadVoidSize-..743274/:1/111782 ABSTRACTPlantresponsetotheJanuary25,1982steamgenerator tubefailureattheR.E.GinnanuclearpowerplanthasbeenanalyzedusingthecomputercodeLOFTRANtoprovideadditional insightintoreactorcoolantvoiding,naturalcirculation loopflows,andprimary-to-secondary leakageduringtheevent.Resultsarecomparedtoavailable plantdata.3274(}:1/111782

/TABLEII.1-1LISTOFTABLESSEQUENCEOFMAJOREVENTSTABLEII.3-1PRE-TRIPPRESSURIZER LEVELTABLE11.5.1-1.SEQUENCEOFPORVOPERATION TABLEC-1UPPERHEADVOIDSIZE3274O:1/111782 LISTOFFIGURESFIGUREFIGUREII.1-1!I.4-1FIGUREII.4-2FIGUREII.4-3FIGURE,II.4-4FIGUREII.4-5FIGUREII.4-6FIGUREFIGUREFIGUREFIGUREFIGUREFIGUREFIGUREII.4-7II.5-1II.5.2II.5.3II.5-4II.5.1-1II.5.1-2FIGUREFIGUREFIGUREFIGUREII.5.1-3II.5.1-4II.5.2-1II.5.2.1-1 FIGUREFIGUREII.5.3.1-2II.5.3.2-1 FIGUREFIGUREII.5.3.2-2 II.5.3.2-3FIGUREFIGUREFIGUREFIGUREII.5.3.2-4 II.5.3.2-5 II.5.3.2-6II.5.3.3-1FIGUREII.5.2.1-2 FIGUREII.5.3-'1 FIGUREII.5.3.1-1GINNASAFETYINJECTION CAPACITYPRE-TRIPNORMALIZED COREPOWERPRE-TRIPSECONDARY SYSTEMPRESSUREPRE-TRIPPRESSURIZER PRESSUREPRE-TRIPPRESSURIZER LEVELPRE-TRIPAVERAGERCSCOOLANTTEMPERATURE PRE-TRIPPRESSURIZER PRESSURE:

CONSTANTCOOLANTTEMPERATURE PRE-TRIPPRESSURIZER LEVEL:CONSTANTCOOLANTTEMPERATURE NORMALIZED PRE-TRIPSTEAMFLOWNORMALIZED PRE-TRIPFfEDWATER FLOWINTACTSTEAMGENERATOR PRESSUREINTACTLOOPCOLDLEGTEMPERATURE REACTORCOOLANTSYSTEMPRESSURE'PRIMARY-TO-SECONDARY LEAKAGEANDTOTALSAFETYINJECTION FLOWPRESSURIZER WATERVOLUMEUPPERHEADFLUIDMASSVOLUMETRIC LOOPFLOWRATESCOMPARISON OF"MIXEDTEMPERATURE" REVERSEFLOWTHROUGHFAULTEDLOOPANDBREAKFLOWFROMTHESGOUTLETPLENUMFAULTEDLOOPCOLDLEGINLETANDOUTLETFLOWSPOST-TRIP REACTORCOOLANTTEMPERATURES STEAMDUMPVALVEOPERATION ANDAFWFLOWDURINGCOOLDOWNOFTHERCSGINNACOREEXITANDINTACTLOOPCOLDLEGTEMPERATURES COMPARISON OFINTACTANDFAULTEDLOOPCOLDLEGTEMPERATURES FOLLOWING REACTORTRIPFAULTEDLOOPCOLDLEGTEMPERATURES MIXINGVOLUMEFORVESSELDOWNCOMER TEMPERATURf CALCULATION MIXINGVOLUMELOOPFLOWANDSAFETYINJECTION FLOWMIXINGVOLUMEFLOWTEMPERATURfS BESTESTIMATEREACTORVESSELDOWNCOMER TfMPERATURE COREEXITFLUIDTEMPERATURE 3274(:1/111982 111

LISTOFFIGURES(Cont.)FIGUREFIGUREFIGUREFIGUREFIGUREFIGUREFIGUREFIGUREII.5.4-1II.5.5-1II.5.5-2II.5.7-1II.6-1II.6-28-1C-1FIGUREII.5.3.4-1FIGUREII.5.3.4-2FIGUREII.5.3.5-1FIGUREII.5.3.6-1FIGUREII.5.3.6-2BREAKFLOWFROMSGINLETANDOUTLETPLENUMSLOFTRANFAULTEDLOOPHOTLEGTEMPERATURE FAULTEDSGTUBEBUNDLEFLUIDTEMPERATURE POST-TRIP UPPERHEADFLUIDTEMPERATURE LOFTRANUPPERHEADFLUIDTEMPERATURE PRESSURIZER LEVELINDICATION LOFTRANANDBESTESTIMATEBREAKFLOWSFAULTEDSTEAMGENERATOR PRESSUREFAULTEDSTEAMGENERATOR WATERVOLUMERCSANDFAULTEDSTEAMGENERATOR PRESSURES LONGTERMPRESSURIZER LEVELRESPONSESGTUBERUPTUREFLOWMODELDIAGRAMUPPERHEADVOIDINGILLUSTRATION 32740:1/111982 1V I.INTRODUCTION AttherequestofRochester GasandElectric(RGE),Westinghouse hasanalyzedtheJanuary25,1982steamgenerator tuberuptureeventattheR.E.Ginnanuclearpowerplant.Theprinciple objective ofthiseffortistosupplement theexistingdatabase'oprovideamorethoroughunderstanding oftheplantr'esponse andactualsequenceofevents.Ofparticular interestarevoidingofthereactorcoolant,naturalcirculation loopflowbehavior, andprimary-to-secondary leakage.TheLOFTRANcomputercodewasusedfor(3)theseanalyses.

Anumberofauxiliary calculations arealsodescribed whichcomplement LOFTRANbyprqviding moredetailedmodelling oflocalized effects.Theplantresponsetothesteamgenerator tubefailureandsubsequent r'ecovery actionsispresented forthreedifferent, phasesoftheevent.Thepre-tripdatarecordprovidesinformation forestimating theinitialleakrateandextentoftubefailure.LOFTRAHanalysisofthisperiodextrapolates thisdatatodetermine theapproximate timeoftubefailureandhistoryoftube(leakage.Following reactortrip,severalautomatic protection systemswere~~~actuatedinrapidsuccession andasequenceofemergency recoveryactionswasinitiated tomitigatetheconsequences oftheaccident.

Thisemergency recov-eryperiodculminated intermination ofsafetyinjection.

Theplantresponsetotheautomatic protection systemsandrecoveryactionsduringthisphasewasalsoinvestigated usingLOFTRAN.Finally,thelongtermplantresponseandadditional leakageintothefaultedsteamgenerator aftertermination ofsafetyinjection isdiscussed.

Abriefdiscussion ofLOFTRANmodelling ispresented.

Severallimitations areidentified whicharesignificant whenappliedtotheGinnaeventandmust-beconsidered whenevaluating theanalysisresults.Theseresultsarecomparedtotheavailable plantdatainthefollowing sections.

32740:1/111882

II.ANALYSISOFPLANTRESPONSEII.1SystemsAnalysisCodeLOFTRANisafastrunning,digitalcomputercodedeveloped tosimulatetran-sientbehaviorinWestinghouse pressurized waterreactors.

Theprogrammodelsneutronkineticsaswellascontrolandprotection systemsontheprimaryandsecondary systems.Themostsignificant oftheprotection systems,theEmer-gencyCoreCoolingSystem(ECCS)andtheAuxiliary Feedwater (AFW)system,aredescribed below.TheECCSwasrepresented bythecombinedcapacityofthreehighheadsafetyinjection pumpsshowninFigureII.1-1.Safetyinjection initiated automatic-allyonlowpressurizer pressureof1740psiaandwasassumedequallydistrib-utedbetweenloops.Thesuctionofthesepumpswasinitially alignedtotwoboricacidtanks(BAT)containing approximately 4320gallonsofboratedwaterat140F.Onlowlevel,suctionwasautomatically re-aligned totheRefueling WaterStorageTank(RWST)whichcontained coolerwater.Intheanalysespre-sented,60FwaterfromtheRWSTwasassumedtobeinjectedthroughtheBATcontaining a140Fboricacidsolution.

Reactorcoolantmakeupfromthenor-malchargingpumpswasalsosimulated withsuctionfromthesepumpsalignedtotheRWST.Chargingflowisdiscussed onacasebycasebasis'inthefollowing sections.

TwomotordrivenAFWpumpsautomatically startedonasafetyinjection signal.Eachmotordrivenpumpprovidedapproximately 200GPMofwaterfromtheCondensate StorageTank(CST)anddelivered tooneofthetwosteamgener-ators.OnesteamdrivenAFWpumpstartedautomatically onlow-lowsteamgen-eratorlevel.Thesteamdrivenpumpsuppl.ied atotalof400GPMwhichwasavailable tobothsteamgenerators.

AFWpumpoperation wassimulated asdes-cribedinthesequenceofeventsinTableII.l-l.Apurgevolumeof200ft3containing normalfeedwater wasalsosimulated.

Thisrepresented adelayofapproximately 4minutesbeforecoldCSTwaterenteredthetuberegionof:theintactsteamgenerator.

3274(}:1/111882 22002000180016001400120010008004002000.020040060080010001200FIGUREII.1-1.GINiVASAFETYIHJECTIOiV CAPACITY.

3 PreviousLOFTRANanalyseshavesimulated thePrairieIslandtubefailureeventwell.However,LOFTRANissomewhatlimitedbythemodelling oftheupperheadregion,steamgenerator secondary side,andprimary-to-secondary leakage.Theupperheadmodelling assumeshomogeneous, thermodynamic equilibrium conditions duringflashingoftheupperheadfluid.Refilling oftheupperheadregionisartificially constrained tosimulatenon-equilibrium behavior.

Effectively, theupperheadregioncannotrefillduringnaturalcirculation flow.Furthermore, flowintotheupperheadregionviaguidetubesisnotrepresented.

Consequently, thecalculated upperheadfluidtemperature maybeunrealistic forplantswithsmallerupperhead"spray"nozzles,suchasGinna.LOFTRANisalsolimitedbythehomogeneous,

'saturated conditions withinthe'secondary whichpromotesanunrealisticlylethargic tubebundleregiontemperature response,to AFMflowandsecondary-to-primary heattransfer.

Inaddition, theseconditions resultinartifically reducedsteamgenerator pressures whennosteamflowoccurssincethesteamiseffectively assumedtobeincontactwiththesteamgenerator tubes.Thebreakflowcal-culations withinLOFTRANarebasedonconservative, i.e.maximumflow,criti-calflowcorrelations.

Theaccuracyofthesecorrelations inpredicting criticalflowtrendsoverawiderangeofsystemconditions isuncertain.

Furthermore, thebreakflowmodelling doesnotconsiderflowresistance throughthefailedtube,orfluidtemperature variations betweenthesteamgenerator inletandoutletplenums.Finally,LOFTRANdoesnotpermitreverseflowtooccurinthecoolantlooptowhichthepressurizer isconnected.

Fortheresultspresented, thepressurizer wasmodelledontheintactloopalthoughduringtheGinnaeventthepressurizer wasonthefaultedloop.Thismayresultinunrealistic loopflowsduringrefilling ofthepressurizer.

II.2PlantDataTheplantdatawhichformsthebasisoftheanalysesthatfollowwasobtainedfromvariouscomputerrecords,stripchartsofsystemparameters, andthesequenceofeventsasreconstructed byRGE.TheGinnaplantcomputerisarealtimecentralprocessing unitwhichstoresselectedplantparameters:

foruseduringnormaloperations.Severalperipherial devicesservicedbythiscentralunitprovidetheprincipal dataforpost-accident analyseswhichincludesprimaryandsecondary pressures, reactorcoolanttemperatures, and3274(}:1/111882 r~

pressurizer level.Thesedevicesincludeapre-tripeventrecorder, aTI-7000teletypeterminal, analarmtypewriter, andalogtypewriter.

Communications withRGEpersonnel supplemented thisdataandprovidedadditional insightintotheevent.Thesequenceofoperatoractionswasextracted fromthesequenceofeventsprovidedbyRGEandthechronology pfplantalarms.whenpossible.

Themajoreventsarepresented inTableII.l-l.Comprehensive plantdataandthecompletesequenceofeventsisavailable inreference l.II.3InitialLeakRateThepre-trippressurizer levelresponsetothelossofreactorcoolantwasanalyzedtoestimatetheinitialprimary-to-secondary leakrate.Rapidvaria-tionsinreactorcoolanttemperature duetoturbinerunbackandautomatic steamdumptothecondenser tendedtomasktheinventory loss.However,anaverageleakratepriortotripwas.estimated byconsidering theindicated pressurizer levelresponsebetweentimesofconstantaveragecoolanttempera-ture.Sincethecoolanttemperatures atthesetimeswereapproximately thesame,theeffectoftheturbinerunbackonthiscalculation wasminimized.

Thepre-trippressurizer level,adjustedforinstrumentation calibration, ispresented inTableII.3-1;Basedondiscussions withRGEpersonnel, twochargingpumpswereoperating priortothetubefailure,oneinmanualandtheotherinautomatic.

Eachpumpwasdelivering approximately 25GPMofflow.Totalletdown,including

'eactorcoolantpumpsealleakoff,was50GPM.Following tubefailure,onechargingpumpautomatically increased tothemaximumcapacityof60GPMaspressurizer leveldecreased.

Althoughathirdchargingpumpwasmanuallystartedapproximately 40secondsbeforereactortrip,9:27:30,itwouldhaveprovidedlittleflowbeforetrip.Consequently, thenormalchargingsystemwassupplying anexcessofapproximately 35GPMduringthisperiod.Theaveragepre-tripleakratewasestimated tobe573GPM(Appendix A).Aninitialleakrateof634GPMwascalculated byextrapolating theaverage-leakratetotheinitialsystemconditions basedonsubcooled criticalflow(5)throughthefailedtube.Aneffective breakareaof0.0033ft2wasdetermined byproportioning thecriticalflowrateascalculated byLOFTRANforthe.initialsystemconditions tomatchtheinitialleakrate.3274(}:1/111882

TABLEII.3-1PRE-TRIPPRESSURIZER LEVEL~~Time(A.M.)Indicated Level(Xspan)Adjusted*

Level('Xspan)Tavg(F)9:26:189:26:269:26:349:26:429:26'.589:27:069:27:149:27:229:27:309:27:389:27'469:27:549:28:029:28:10132.530.630.530.530.230.230.228.926.220.8'17.914.811.79.032.731.030.830.930.730.530.629.426.922.019.516.814.111.9571.2571.2571.9573.2575.2576.4576.8576.1575.3574.5573.4572.4571.4570.1*SeeAppendixA3274(}:1/1 11782

II.4Pre-TripSystemResponseThefirstindications ofabnormalconditions wererecordedatapproximately 9:25whenanumberofalarmssoundednearlysimultaneously.

Theseincludedlowpressurizer

pressure, lowpressurizer level,condenser airejectorradia-tion,andBsteamgenerator leveldeviation alarms.Thealarmrecorderindi-catesthatthelowpressurizer presurealarmsoundedfirst.Systemconditions werenormalat9:22withnoapparentsymptomsofprimary-to-secondary leakage.Thereactorcoolantsystempressureandtemperature priortoreactortripwereanalyzedusingnormalized corepower,FigureII.4-1,'nd secondary

pressure, FigureII.4-2,dataasforcingfunctions forthecalculations.

Alternative secondary sideboundaryconditions werealsoconsidered asforcingfunctions forthepre-tripcalculations, including normalized steamandfeedwater flows..Althoughtheseproducedreasonable results,theinstrument uncertain-tiesandresponsetimeswerenotasconducive assecondary pressuretopre-tripanalysis.

Thepressurizer pressureandlevelresponses calculated usingLOFTRANagreedverywellwithplantdataasillustrated inFiguresII.4-3andII.4-4,respectively.

Extrapolation ofthisdatawithaninitialleakrateof634GPMsuggeststhattubefailureoccurredat9:25:10(dmin).Thecalculated pressureattheactualtimeofreactortripwasapproximately 30PSIgreaterthanindicated.

Theaveragereactorcoolanttemperature iscomparedwithpre-tripdatainFigureII.4-5.Theincreaseintemperature duetoturbinerunbackmomentarily maskedthedecreaseinprimarycoolantinventory.

Simi-larly,whenthesteamdumpvalvesopened,theassociated cooldownenhancedreactorcoolantsystemdepressurization.

FiguresII.4-6andII.4-7illustrate thepredicted pressurizer pressureandlevelresponses whenreactorcoolanttemperature wasmaintained constant.

Asdemon'stated, thepressureandlevelresponses aresignificantly affectedbycoolanttemperature trends.3274(:1/111882

100GI<<HA(+)LOFTRAN(-)90~r80C)CL70C)LJJIVo..60C)509:24925926TINE(A.ii.)9:279:28929FIGUREII.4-1.PRE-TRIPthORllALIZED COREPOllER.

1100GINNA(+)LOFTRAN(-)10009008007006009249:25926TIVE(A.H.)9:279:28929FIGUREII.4-2.PRE-TRIPSECONDARY SYSTEMPRESSURE 2200GIN~iA(+)LOFTRAt~(-)20001800LU1600C'40012009:249:259:269:279:289:29TItiE(n.n.)FIGUREII.4-3.PRE-TRIPPRESSURI"ER PRESSURE.

10080GIr<NA(+)LOFTRAN(-)60402009:249259:26927928929TIflE(A.H.)FIGUREII.4-4.PRE-TRIPPRESSURIZER LEVEL.

58520GINNA(+)LOFTRAN(-)1612u5755659249:259:269:279:2809:29TII1E(A.H.)FIGUREII.4-5.PRE-TRIPAVERAGERCSCOOLANTTft1PERATURE

2200GINNA(+)LOFTRAN(-)2000180016001400'12009:249:259269:279:28TInE(A.W.)FIGUREII.4-6.PRE-TRIPPRESSURIZER PRESSURE:

CONSTANTCOOLANTTEf1PERATURE 100O60+O~r40UiCYUJ2009289249:259:27TIi'iE(A.t1.)FIGUREII.4-7.PRE-TRIPPRESSURIZER LEVEL:CONSTANTCOOLANTTEHPERATURE 9:29

TABLEII.l-l:SEQUENCEOFMMOREVENTSEventManual(0)Automatic (A)ActualTIME(Sec)Simulated TubeFailureTurbineRunbackAutomatic SteamDumpReactorTripSafetyInjection SignalFeedwater Isolation Auxiliary Feedwater StartReactorCoolantPumpTripBMotorDrivenAFWPumpOffManual.SteamDumpBLoopMSIVClosedAFWThrottled ASGAFWStoppedtoBSGChargingPumpsStartedPORVCycledSITerminated 078110182190192220230410(1)770(1)890("950(1)1250('2330(25404310(070118182198198239246410530(530(2)9501250233025404310Thesetimesareapproximate andtypically mayvarybyupto60seconds.SeesectionII.5foradiscussion onthesimulation oftheseevents.3274Q:1/111882 l5

II.5Post-Trip SystemResponseContinued leakageofprimarycoolantincombination withrapidcooldownofthereactorcoolantsystemfollowing actuation ofthesteamdumpsystem,causedanautomatic reactortriponlowpressurizer pressure.

Primarysystempressuredecreased rapidlyaspowergeneration wasabated,andseveralsupporting sys-tems,including theECCSandAFWsystem,startedinrelatively rapidsucces-sion.Aseriesofoperatoractionscommenced inaccordance withemergency responseprocedures torecovertheplanttoasafeshutdowncondition.

TheplantresponsetothesesystemsandoperatoractionswasanalyzedusingLOFTRANandtheresultsarepresented inthefollowing sections.

Theseanaly-seswerelimitedtothetimefrominitialfailureuntil10:40(75min),shortlyaftertermination ofsafetyinjection.

Beyondthistime,LOFTRANanalysiswasnotappropriate becauseofthehomogeneous, equilibrium secondary sidemodelling.

Additional leakageintothefaultedsteamgenerator wasestimated fromplantdatatodetermine themassdischarged fromthesteamgenerator afteroverfill.

Thesteamgenerator tubefailureoccurredintheBloopduringtheGinnaevent.Consequently, faultedloopandBlooparesynonymous inthefollowing sections.

Similarly, intactloopandAloopare*usedinterchangably.

However,AloopandBloopdesignations generally refertoplantdata,andintactandfaultedtoLOFTRANcalculations.

Fortheseanalyses, normalized steamflow,FigureII.5-1,andfeedwater flow,FigureII.5-2,fromplantdatawereusedpriortotripasforcingfunctions forthecalculations.

Thisprovidedadditional flexibility inmodelling sub-sequentoperatoractionswithL'OFTRAN.

TheLOFTRANanalysisresults,whichincludeanadjustment totheinputsteamflowtoaccountforincreased secon-daryp'ressure, arealsoshownforcomparison.

Afterreactortrip,theintactsteam'generator pressurewascontrolled toreproduce theindicated loopAcoldlegtemperatures.

Theintactsteamgenerator pressureinputtoLOFTRANiscomparedtoplantdatainFigureII.5-3.FigureII.5-4showsthecalculated andmeasuredintactloopfluidtemperatures atthecoldleginlet.NotethatloopAcoldlegfluidwassubcooled attheAloopsteamgenerator pressurebetweenapproximately 9:32(7min)and9:41(16min).Becauseofthehomo-geneousequilibrium secondary sidemodelling withLOFTRAN,suchsubcooling 32740:1/11188216 100GINNA(+)LOFTRAN(-)8060402009:249:259:26927TIi~>E(A.>>.)FIGUREII.5-1.i%OR/1ALIZED PRE-TRIPSTENlFLOW.9:28929

10080604020'9249'259:269:27TINE(A.ti.)FIGUREII.5-2.HORHALIZED PRE-TRIPFEEDHATER fLOW.9289:29

1200.01000.0GINNA(G)LOFTRAN(-)800.00CG600.00~400.00200.000.0CDCDCDTINKtMIN)FIGUREII.5-3.INTACTSTEANGENERATOR PRESSURE.

19

700.00GINHA(G)LOFTRAN(-)INLETOUTLETC)C)ClAJ7lHE(MlM)FIGUREII.5-4.INTACTLOOPCOLDLEGTEf1PERATURE.

20

couldnotbereproduced.

,However,thecooldownoftheAloopwassimulated byartifically steamingtheintactsteamgenerator.

Hence,thecalculated steamgenerator pressurewaslessthanmeasuredduringthisperiod.II.5.1PrimarySystemPressurePrimarypressurecontinued todecreasefollowing reactortripasbreakflowdepletedcoolantinventory andautomatic steamdumpcooledthereactorcoolantsystem.Safetyinjection wasactivated withinapproximately 16seconds,at9:28:28(3.2min),whenpressurizer pressurereached1740psia.Threehighheadsafetyinjection pumpsbegantoinjectshortlythereafter

.torestorecoolantinventory.

Pressurecontinued todecreasetoaminimumof1200psiabetween9:29(4min)and9:30(5min)asautomatic steamdumpestablished no-loadRCStemperature.

Asmallvoidmayhavedeveloped intheupperheadregion.duringthisinitialdepressurization althoughLOFTRANdidnotpredictflashing(seesectionII.5.6.2).

Thecalculated RCSpressurehistoryiscomparedtoplantdatainFigureII.5.1-1.

Whenthepost-trip cooldownsubsidedafterno-loadtemperature hadbeenestab-lished,safetyinjection flowinexcessofbreakflow,FigureII.5.1-2, repressurized thereactorcoolantsystemuntilapproximately 9:32(7min).Heat-upofthereactorcoolantduringthetransition fromforcedtonaturalcirculation duetoRCPtripcontributed tothisrepressurization.

LOFTRANanalysisdemonstrated amorerapidrepressurization duringthisperiodthanactuallyobservedpossiblybecauseofcollapseofanupperheadvoidduringtheactualevent.Reactorcoolantshrinkage, ascoldAFWenteredtheAloopsteamgenerator, incombination withbreakflowdecreased pressuretoamini-mumof1140psiabetween9:32(7min)and9:41(16min).Manualsteamreleasebeginning atapproximately 9:38(13min)contributed littletothisRCScool-downsincethesteamgenerator tubebundleregionwassubcooled.

AlthoughadecreaseinprimarysystempressureisevidentintheanalysisresultsshowninFigureII.5.1-1, theactualpressuredecreased significantly lower.Thisiscausedinpartbytheinitially highercalculated primarypressureat-9:32(7min).Inaddition, althoughLOFTRANindicates asmallamountofwaterremainedinthepressurizer duringthisperiod,FigureII.5;1-3, thepressur-izermayhaveactuallydrained.Thiswouldhaveenhanceddepressurization oftheprimarysystem.3274(:1/111882 21

2500.02250.0GINNA(G)LOFTRAN(-)1750.01500.0GGGGCGG1000.0CDCDCDCDAJTINE(MIN)CDCDCDC)CDFIGUREII.5.1-1.

REACTORCOOLANTSYSTE[1PRESSURE.

22

125.00LOFTRAN(-)100.00SIFLOll75.000lsJ~50.00025.000BREAKFLOW0.0C7C)EDC)C)AJC)C)CDC)TIVE(MlH)CDClC7CDtDC)C)CDC)EX7CDCDFIGUREII.5.1-2.

PRIORY-TO-SECONDARY LEAKAGEANDTOTALSAFETYINJECTION FLOW.23

800.00700.00LOFTRAN(-)600.00500.00F00.00300.00CD200.00100.000.0CDCDCDAJCDTlHK(MlH)C)CDCDEOCDCDCDCOFIGUREII.5.1-3.

PRESSURIZER WATERVOLUt1E.

WhenAFWflowwasterminated totheAloopsteamgenerator at9:41(16min),thecooldownofthereactorcoolantsystemsubsided.

Safetyinjection flowrepressurized theprimarysystemtowardanequilibrium pressureofapproxi-mately1320psiawherebreakflowandsafetyinjection werenearlyequal,asillustrated inFigureII.5.1-2.

Operation ofthesteamdumpvalvesoccasion-allyperturbed thisgeneraltrendandmaintained pressureslightlybelowequilibrium.

LOFTRANcalculations slightlyoverestimated thereactorcoolantsystempressureduringthisperiod.Twochargingpumpswithacombinedcapac-ityof120gpmwereassumedtoi'njectintothefaultedloopcoldlegbeginning at10:04(39min),asnotedinthesequenceofevents.Asaresult,thepre-dictedprimarypressureincreased towardanequilibrium valueof1410psiaby10:07(42min).Thisisconsistent withplantdatawhichindicates anequi-libriumpressureofapproximately 1390psia.Cyclingofapressurizer PowerOperatedReliefValve(PORV)wassimulated beginning at10:07(42min)andproceeded asindicated inTableII.5.1-1.

ThePORVwasmodelledtofullyopenorcloseinstantaneously.

Duringtheactualevent,thePORVfailedtocloseonthefourthcycleandwasmanuallyiso-lated.Althoughthisisolation wasassumedcompleted by10:10(45min),theactualtimemayhavebeenslightlylater.Thecalculated reactorcoolantsystempressureresponseduringthisperiodagreedwellwithavailable plantdata.Theminimumpressureduringthisperiodwascalculated tobe847psia.Following isolation ofthefailedPORV,thereactorcoolantsystempressureincreased rapidlytoapproximately 1400psiaassafetyinjection flowandreverseflowfromtherupturedsteamgenerator increased coolantinventory.

Theactualrepressurization wasslowerthancalculated byLOFTRAN.Asnotedpreviously, LOFTRANinhibitsrefilling oftheupperheadregionduringnaturalcirculation flow,asevidenced bytheconstantupperheadfluidmassbeyond10:10(45min)inFigureII.5.1-4.

Thisenhancedrefilling ofthepressurizer and,consequently, repressurization oftheprimarysystem.Hence,theslowerincreaseinpressureobservedintheactualeventisattributed toatleastpartialrefilling oftheupperheadregion.By10:17(52min),safetyinjection andchargingflowshadreestablished anequilibrium withbreakflowatapproximately 1400psia.Whensafetyinjection 32740:1/111782 25 3.00E+OiLOFTRAH(-)2.50Ei"2.00Ei,~t.50E&iSaturated Liquidl.OOEKli5000.0r>-Saturated Vapor0.0C)CDAJ8CDCDCD07TINE(MlM)FIGUREII.5.1-4.

UPPERHEADFLUID(NSS.26 TABLEII.5.1-1SEQUENCEOFPORVOPERATION CycleOpenedTime(A.M.)Closed10:07:30.5 10:07:35.5 10:07:49.3 10:07:57.3 10:08:44.0 10:08:52.7 10:09:10.1 Actualtimeofisolation mayhavebeenslightlylater32740:1/1 1178227

wasterminated at10:37(72min),primarysystempressuredecreased rapidlyfrom1370psiato945psia.LOFTRANanalysesdemonstrated asimilardepres-surization asthepressurizer steambubbleexpandedtoaccommodate residualbreakflowinexcessofreactorcoolantmakeup.Continued chargingflowandpressurizer heateroperation maintained primarypressuregreaterthan'thefaultedsteamgenerator pressureuntilprimary-to-secondary leakagewasterminated at12:30(185min).II.5.2ReactorCoolantFlowAtransition fromforcedtonaturalcirculation flowoccurredfollowing manualreactorcoolantpumptripat9:29:09(4min).Thisisevidenced bytheincreasing loopdelta-Tbetween9:29(4min)and9:31(6min).AFWflowpref-erentially cooledtheAsteamgenerator whichenhancednaturalcirculation flowintheAloopandretardedflowintheBloop.Thisresponsewasdemonstrated intheLOFTRANresultsshowninFigureII..5.2-1 from9:32(7min)to9:41(16min).WhenAFWflowwasthrottled at9:41(16'min),flowthroughtheAintactloopwascalculated todecreasetoapproximately 4Xofinitialconditions.

Flowthroughthefaultedloopmomentarily increased, asAFWflowfromtheturbinedrivenAFWpumpcontinued tocoolthefaultedsteamgenerator, until9:46(21min)whenAFWflowwasterminated; Asthecooldownoftheintactsteamgenerator continued, flowthroughthefaultedloopwas'alculated tostagnateat10:10(45min).Natural.circulation flowthroughtheintactloopwasmaintained between3Xand4gfullflowuntilthereactorcoolantpumpwasrestarted at11:19(114min).II.5.2.1LoopBColdLegFlowAlthoughLOFTRANdidnotsupportsignificant reverseflowthroughthefaultedloop,theeffect.ofbreakflowmodelling onthecalculated loopflowwasuncertain.

Hence,thepotential forprimary-to-secondary leakagegenerating sufficient reverseloopflowtoproducetheobservedBlooptemperature responsewasinvestigated.

FigureII.5.2.1-1 comparesthetotalreverseflow,i.e.safetyinjection flowandloopflowfromthevesseldowncomer, whichwouldhaveamixedtemperature identical totheindicated Blooptemperature, 28

10.000LOFTRAN(-)8.00006.00001.0000lINTACTLOOP4~2.0000FAULTEDLOOP0.0"l.0000ClClCDClT1HEtHlN)CDCDCDCDCDCDEKIFIGUREII~52lVOLUMETRIC LOOPFLOWRATES~

300.00250.00200.00ChargingPumpsStartedPORVClosedCalculated(+)"MixedTemperature" ReverseLoopFlow.150.00PORVOpened50.000OutletPlenumBreakFlow0.0-25.000CDCDCDCDCDCDCDTIME"(HIM)CDCDCDCDCDCDCDCDCQCDFIGURE'I!.5.2.1-1.

COiiPARISON OF"MIXEDTEYiPERATURE" REVERSEFLOWTHROUGHFAULTEDLOOPANDBREAKFLOWFROtlSGOUTLETPLENUi~i.

3O randthecalculated breakflowfromthesteamgenerator outletplenum.Asdemonstrated, primary-to-secondary leakagewasmuchlessthantherequiredmixedloopflow.Asanadditional assessment ofpotential reverseloopflow,theresponseofthetubebundlefluidtemperature inthefaultedsteamgenera-torwascalculated (seesectionII.5.3.5).

Theseresultssuggestthatifsufficient reverseloopflowdidoccurandproducedamixedtemperature responsesimilartotemperatures actuallyobserved, thefaultedsteamgenera-torwouldhavebeencolderthantheintactsteamgenerator.

Sincethiswouldpromoteforwardflowinthefaultedloop,itisunlikelythatsuchsustained reverseflowoccurred.

Ginnadatademonstrated propagation ofaportionofthesafetyinjection flowupstreamoftheinjection locationbeginning at9:39(14min).Inaddition, theobserved8loopcoldlegtemperature responsesuggestsacontinuous supplyofwarmfluidupstreamoftheinjection nozzle(seesectionII.5.3.3).

Thecalculated fluidflowsintoandfromthefaultedloopcoldlegareshowninFigureII.5.2.1-2.

Safetyinjection flowwascalculated tosplitwhenthefaultedloopflowstagnated at10:10(45min);asmallportionflowedtowardthesteamgenerator whilethemajorityflowedtowardthevessel.Acontinuous flowofwarmwaterwasnotobservedintheLOFTRANanalysisresults.Nosig-nificanttemperature increaseinthecalculated faultedloopcoldlegtempera-tureoccurred.

Theseresultssupporttheexistence ofacounter-current typeofflowregimeupstreamoftheinjection nozzle.The8looptemperature responserepresents mixingofaportionofthesafetyinjection flowupstreamoftheinjection nozzlewithastreamofwarmerwaterfromthesteamgenerator.

Suchmixingisnotsimulated intheonedimensional modelling ofLOFTRAN.Themagnitude offlowfromthefaultedsteamgenerator requiredtoproducethequasi-steady temperature responsewasestimated fromthecoldleginlettemperature andsafetyinjection flowinthefaultedloopcalculated withLOFTRAN.Experimental evidence'uggeststhatasignificant portionofsafetyinjection intoastagnantloopwouldpropagate upstreamoftheinjection nozzle.Basedonthisevidence, onethirdofthesafetyinjection flowwasassumedtomixupstreamoftheinjection location.

Theresultofthiscalculation indicates thataminimumloopflowof21ibm/sec(170gpm)existedafter10:07(42min).3274(}:1/111982 31

1000.0LOFTRAIN(-)800.00600.003100.00OUTLETu-200.00IINLET0.0"100.00CDCDCDCDAJCDCDTlMf(MlN)CDCDCDCQCDCDFIGUREII.5.2.1-2.

FAULTEDLOOPCOLDLEGINLETANDOUTLETFLOWS.32

II.5.3ReactorCoolantTemperatures Theearlyreactorcoolanttemperature responses weretypicalofreactortrip.Hotandcoldlegtemperatures decreased rapidlyastheautomatic steamdumpsystemandsecondary coolantabsorbedenergymorerapidlythandecayingcorepower.Thelargeflow/power mismatchreducedthecorecoolanttemperature risetoonlyafewdegreesandthesteamdumpsystemoperatedtomomentarily stabilize temperatures nearno-load.Themeasuredcoretemperature risedecreased toaminimumof2Fbeforereactorcoolantpumpsweretrippedandsteadilyincreased thereafter toapproximately 10'Fby9:31(6min).From9:31(6min)to9:38(13min)allsteamdumpvalveswereclosed.Duringthistime,safetyinjection andauxiliary feedwater flowsabsorbeddecayheatandstabilized temperatures asdemonstrated intheLOFTRANanalysisresultsshowninFigureII.5.3-1.

II.5.3.1ALoopColdLegTemperature ColdAFMflowrapidlycooledtheAloopcoldlegbeginning at9:32(7min).AlthoughtheAFWpumpswereautomatically startedshortlyafterreactortrip,thesteamgenerator feedlines anddowncomer volumedelayedinjection ofcoldwaterfromtheCondensate StorageTank(CST)intothetubebundleregion.Twosteamdumpvalveswere.manuallyopenedfromabout9:38(13min)until9:39(14min)todecreaseprimarycoolanttemperature asdirectedbytheemergency operating procedures.

Thisappearstohavehadlittleeffectonthecoldlegtemperature sincethesecondary coolantinthetubebundleregionwassub-cooled.TheAloopcoldlegtemperature decreased toaminimumof485Fat9:41(16min)whenAFMflowwasterminated totheAsteamgenerator.

Thehomogeneous equilibrium secondary sidemodelling withinLOFTRANtendedtounderestimate theprimarysystemcoolingduetoauxiliary feedwater.

Conse-quently,theintactsteamgenerator pressurewasusedasaforcingfunctiontotreproduce theAloopcoldlegtemperature

response, aspreviously noted.FiguresII.5-3andII.5-4comparetheintactsteamgenerator pressureandcalculated, coldlegtemperature, respectively, withplantdata.Withtheexception ofthecooldownduetoAFWflow,bothpressureandtemperature matchthedatareasonably well.Decreases inmeasuredcoldlegtemperature 3274(:1/111882 33 700.00650.00LOFTRAN(-)600.00HOTLEG-550.00~5¹00ICOLDLEGi5000400.00ClC7TIHf(HIN),FIGUREII.5.3-1.POST-TRIP REACTORCOOLAHTTEMPERATURES.

"

correlate withtheoperation ofsteamdumpvalvesandAFWflowasshowninFigureII.5.3.1-1.

Theinsurgeofsafetyinjection whenthepressurizer PORVwasopened,reducedthecoreexittemperature belowtheAloopcoldlegtem-perature, asshowninFigureII.5.3.1-2.

Thissuggeststhatsecondary-to-primaryheattransfermayhavemomentarily occurredintheAloop.TheLOFTRANanalysisresultsdemonstrated asimiliarresponse.

However,thedecreaseincoldlegtemperature between10:12(47min)and10:15(50min)wasunderestimated.

Thisappearstobeaconsequence ofthehomogeneous secondary side.II.5.3.2BLoopColdLegTemperaaure TheBloopcoldlegtemperature responsewasessentially thesameastheAloopuntilAFMflowwasreducedtothe8steamgenerator atapproximately 9:32(7min).From9:32(7min)until9:39(14min)theBlooptemperature decreased moreslowlyasillustrated inFigureII.5.3.2-1.

Atapproximately 9:39(14min),theBloopcoldlegtemperature decreased rapidly,indicative ofsafetyinjection flowupstreamoftheinjection nozzle.Beyondthistime,twodistincttrendsareevidentinthemeasuredtemperature response.

Therapiddecreaseintemperature beginning at9:39(14min)istypicalofa"mixing-cup" configuration whereaportionofthecoldsafetyinjection flowmixeswiththewarmerfluidwithinafixedvolume.Sincetheinventory ofwarmer'wateravailable formixingislimited,suchasystemischaracterized bya.continuous, exponential decreaseinfluidtemperature tothetemperature oftheincomingsafetyinjection flow.Equallyevidentinthe8loopcoldlegtemperature responseisaquasi-steady periodbeginning atapproximately 9:57(32min).Itisclearthat"mixing-cup" conditions donotdescribethisbehaviorsincesufficient mixingvolumeisnotavailable tosupportanexponential fittothetemperature timeresponse.

Furthermore, theincreaseincoldlegtemperature from10:11(46min)until10:18(53min)cannotbeexplained by"mixing-cup" behavior.

Thissuggeststhataflowofwarmerwatercontinued intotheBloopcoldleg.Thecalculated faultedloopcoldlegfluidtemperatures arecompareuwithBloopdatainFigureII.5.3.2-2.

Thecalculated coldlegoutlet,i.e.vesselinlet,temperature steadilydecreased, asflowthroughthefaultedsteamgenerator decayed,andapproached thetemperature ofthesafetyinjection 3274(}:1/111882 35 700.00650.00SteamDumpValvesOpen(0)SteamDumpValvesClosed(C)AFWFlowInitiated (I)AFWFlowTerminated (T)600.00455000OCLIz500.00i5000100.00CDC7CDTlMK(MIN)CDCPCDCDCDFIGUREII.5.3.1-1.

STEAN'DUHP VALVEOPERATION ANDAFWFLOWDURINGCOOLDOWNOFTHERCS.

500.00480.00COLDLEG(-)COREEXIT(+)460.00+++440.00l=:420.00400.00CDCDCDCDCD-CDCDCDCIlCDCDCDCDCDCDI/lCDCDCDV)T1MK(HlN)FIGUREII.5.3.1-2.

GIHNACOREEXITANDINTACTLOOPCOLDLEGTEMPERATURES.

37 100.00650.00LOFTRAN(-)600.00o550.00FAULTEDLOOPCL'IXx500.00IINTACTLOOPi50.00100.00CIC7TlME(HtN)C7C)EDClC7IAAJFIGUREII.5.3.2-1.

COl1PARISON OFINTACTANDFAULTEDLOOPCOLDLEGTEflPERATURES FOLLOhlING REACTORTRIP.

700.00600.00500.00GGCOLDLEGINLETGINNA(G)LOFTRAN(-)400.00~300.00IZi-200.00COLDLgGOUTLETd4'cC0G100.000.0CDCDTIME(MlN)CDCDCDCDKl)CD'DFIGUREII.5.3.2-2.FAULTFDLOOPCOLDLEGTEMPERATURES.

39

flow.AlthoughthisissimilartotheBlooptransition period,thecalcu-latedcoldlegoutlettemperature continued todecrease.

Thecalculated temp-eratureupstreamofsafetyinjection remainedrelatively hotuntilapproxi-mately10:10(45min),atwhichtimesafetyinjection flowwasfirstpredicted byLOFTRANtopropagate upstreamoftheinjection nozzle.Whenthepressurizer PORVwascycledbeginning at10:07(42min),asurgeofsafetyinjection flowdecreased theBloopfluidtemperature.

Thecalculated vesselinlettemperature demonstrated asimilarresponse.

Althoughthecal-culatedfaultedloopcoldleginlettemperature wasnotsignificantly

affected, thelocationofthepressurizer mayhaveartificially promotedflowtowardthevessel.Noincreaseinfluidtemperature isobservedintheLOFTRANanalysisresultsafterisolation ofthefailedPORV.Evaluation ofthepotential flowdistributions withinthefaultedloopcoldleg(seesectionII.5.2.1)suggeststhatmulti-dimensional behaviormayhavesignificantly affectedtheactualtemperature response.

Sucheffectsarebeyondthecapabilities ofLOFTRAN.However,theBlooptemperature responseindicates anadditional flowofwarmfluidintothecoldlegwhichisnotobservedintheLOFTRANanalysisresults.Consequently, thecalculated coldlegoutlettemperature showninFigureII.5.3.2-2 underestimates theexpectedminimumbulkfluidtemperature atthevesselinlet.Inordertomorerealisticly estimatetheminimumfluidtemperature inthereactorvessel,thevesseldowncomer, coldlegandcrossover legpiping,andreactorcoolantpumpweremodelledasasingle,mixingvolumeasshowninFigureII.5.3.2-3.

Thetemperature responseofthisconfiguration toflowfromthefaultedsteamgenerator andsafetyinjection flow,FigureII.5.3.2-4, wascalculated assumingperfectfluidmixing.Theseflowsandassociated temperatures, FigureII.5.3.2-5, wereobtainedfromtheLOFTRANanalysisresults;however,aminimumloopflowof21ibm/sec(170gpm)wasassumed,asdiscussed insectionII.5.2.1.

Metalheatadditionfromthereactorvessel,piping,andcoolantpumpwasdetermined fromaonedimensional conduction/

convection heattransfermodelbasedonthemeasuredfluidtemperature inthe~Bloopcoldleg.Thecalculated mixingvolumefluidtemperature iscomparedtotheLOFTRANanalysisvesselinlettemperature andBlooptemperature datainFigureII.5.3.2-6.

Themetalheatandadditional loopflowincreased the3274(}:1/111982 40

SAFETYIHJECTIONRCPCOLDLEGVESSELDOWNCOMER CROSSOVER LEGLOOPFLOWCOREMIDPLANEFIGUREII.5.3.2-3 MIXINGVOLUMEFORVESSELDOWNCOMER TEMPERATURE CALCULATION LOFTRAN(-)ESTIMATED

(------)LOOPFLOWSAFETYINJECTION FLOWTlwf(Nlm)FIGUREII.5.3.2-4'IXING VOLUMELOOPFLOWANDSAFETYINJECTION FLOW42 700.00600.00LOOPFLOW500.00i00.00ce300.00~200.00f0000SAFETYINJECTION FLOW0.0ClClQTlNflNil1FIGUREII.5.3.2-5MIXINGVOLUMEFLOWTEMPERATURES 43 700.00600.00l00.00GINNABLOOPDATA(G)ce300.00BESTESTIMATEVESSELDOWNCOMER t00.00LOFTRANVESSELINLET0.0C7C)C)T!NfllolN1FIGUREII.5.3.2-6 BESTESTIMATEREACTORVESSELDOWNCOMER TEMPERATURE 44

minimumcalculated downcomer fluidtemperature atthecoremidplaneelevation toapproximately 200'F.Inaddition, anincreaseinfluidtemperature ocurredaftersafetyinjection wasterminated at1037(42min)asobservedduringtheactualevent.Theseresultsrepresent amorerealistic estimateoftheminimumfluidtemperature inthevesseldowncomer.

II.5.3.3CoreExitFluidTemperature Thecalculated, coreexitfluidtemperature iscomparedtotheavailable datainFigureII.5.3.3-1.

Imperfect mixingatthecoreinletwassimulated intheresultspresented; consequently, thecoreexittemperature isslightly.

differ-entforthecoreregionsadjacenttothefaultedandintact,loops.Thecoreexitfluidtemperature trendedtheintactloopcoldlegandremainedsubcooled throughout thetransient.

IIncreased reactorcoolantmakeupfollowing startupofthechargingpumpsandcyclingofthepressurizer PORVdecreased thecoreexitfluidtemperature beginning at10:04(39min).Thistemporarily decreased thecoreexittemper-aturebelowtheAloopcoldlegtemperature asdiscussed insectionII.5.3.1.

II.5.3.4BLoopHotLegTemperature Primary-to-secondary leakagefromthesteamgenerator inletplenumprovidedamechanism forflowthroughthefaultedloophotlegevenforstagnantloopflowconditions.

Basedonestimates ofthisbreakflow,FigureII.5.3.4-1, andthehotlegvolume,thefaultedloophotlegtemperature wasestimated tolagthecoreexittemperature bylessthan10minutes.TheLOFTRANanalysisdemonstrated thistrend,asshowninFigureII.5.3.4-2.

Theseresultsindi-catethatthefaultedloophotlegfluidremainedsubcooled throughout theevent.II.5.3.5BSteamGenerator Temperature Thetubebundleregionfluidtemperature ofthefaultedsteamgenerator wascalculated bymodelling asingle,subcooled controlvolumeincommunication withtheprimarysystemviaprimary-to-secondary

.leakageandspecified loopflow.Perfectenergytransferwasassumedsothatbreakflowandloopflow3274(:1/111982 45

700.00GINNA(G)LOFTRAN(-)INTACTLOOPFAULTEDLOOPCICDCDCDAJCDTINK(MtN)CDCDCDFIGUREII.5.3.3-1.

COREEXITFLUIDTEIlPERATURE.

46

S0.00040.000INLETCalculated

(+)30.000OUTLETXCDCDCDCDCDCDAJCDCDCDTlHCtHIM)CDCDCDCQCDCDCDCDCDFIGUREII.5.3.4-1.

BREAKFLOWFRO[1S.G.INLETANDOUTLETPLENUi1S.

600.00HOTLEG(+)COREEXIT(-)CDCDCDAJCDCDCDCDCDCDCDCDCDCDTlME(MlN)FIGUREII.D.3.4-2.LOFTRAHFAULTEDLOOPHOTLEGTEHPERATURE.

48 acheivedthermalequilibrium withthecontrolvolumeinventory.

Twocasesofflowfromthefaultedsteamgenerator outletplenumwereconsidered:

1)reverseloopflowthroughthesteamgenerator equaltothatpresented inFigureII.5.2.1-1, and2)onlyprimary-to-secondary leakagefromtheoutletplenumintothefaultedsteamgenerator.

Thetemperature oftheseflowswasassumedequaltotheindicated Bloopcoldlegtemperature.

Forbothcases,primary-to-secondary leakagefromthesteamgenerator inletplenumwasalsoconsidered.

FigureII.5.3.5-1 comparestheresultsofthesecalculations withtheintactsteamgenerator temperature calculated withLOFTRAN.Case(1)suggeststhatifsufficient reverseloopflowdidoccurandproducedamixedtemperature responsesimilartotemperatures actuallyobserved, thefaultedsteamgenerator wouldhavebeencolderthantheintactsteamgenerator.

Sincethiswouldpromoteforwardflowinthefaultedloop,itisunlikelythatsuchsustained reverseflowoccurred.

Case(2)indicates thatprimary-to-secondary leakageeffectively cooledthetubebundleregionofthefaultedsteamgenerator.

II.5.3.6UpperHeadTemperature Duringnormaloperation, asmallfractionofthecoldlegflowisdivertedintotheupperheadregionofthereactorvesselandmixeswithflowfromtheupperplenumtomaintaintheupperheadfluidtemperature atGinnanear595F.Theseflowsremainnearlyconstantaslongasreactorcoolantpumpscon-tinuetorun.Afterreactortrip,thecoreexittemperature decreases rapidly.Withreactorcoolantpumpsrunning,theupperheadfluidtemperature willalsodecreaserapidlybutwilllagtheupperplenumandcoldlegtemper-atures.Theupperheadregiontemperature transient wasevaluated assumingconstantvolumetric upperheadflowsuntilreactorcoolantpumpsweretrippedat9:29:09(4min).Theaveragecoldlegandhotlegtemperatures asindicated bypre-tripdatawereassumedfortheflowsfromthecoldlegandupperplenum,respectively.

FigureII.5.3.6-1 presentsthecalculated upperheadregionfluidtemperature withandwithoutmetalheat.Forthecasewithmetalheat,themetaltemperature wasassumedtobeequaltothefluidtemperature.

Thefluidtemperature wascalculated tobeapproximately 553Fwhenreactor32740'1/111982 49 700600INTACTSG--------

FAULTEDSG500LaJI-400I-CASE2CASE1300200204060TI)1E(t1IN)80100..120FIGUREII.5.3.5-1.

FAULTEDSGTUBEBUNDLEFLUIDTEl1PERATURE.

50 600ep888p'p590580570I-560550540THOTpTCOLDTSATTUPPERHEAD(-)l8l]H/tlETALHEAT88IIIIIIIMII-CYEDVI-ll/0t1ETALItIlIIIIII'RR9269:27TItlE(A.f1.)9:289:29FIGUREII.5.3.6-1.

POST-TRIP UPPERHEADFLUIDTEtiPERATURE.

C

coolantpumpsweretripped.Thisisconsistent withtheupperheadthermo-Ccoupleindication of556Fat9:54(31min).Notethatvoidingmayhaveoccurredintheupperheatregionwhilereactorcoolantpumpswerestilloperating (seesectionII.5.6.2).Theupperheadregiontemperature calculated byLOFTRANisshowninFigureII.5.3.6-2.

Sinceflowfromtheupperplenumthroughtheguidetubeswasnotmodelled, thefluidtemperature atreactortripwaslessthanwouldbeexpected.

Hence,noupperheadvoidingocurredimmediately following reactortripintheLOFTRANresults.However,thefluidtemperature at10:07(42min),whenthepressurizer PORVwascycled,wasequaltothemeasuredupperheadthermocouple.

Hence,theupperheadvoidingcalculated duringthisperiodisexpectedtoberepresentative oftheGinnaevent.II.5.4Pressurizer LevelResponseThecalculated pressurizer waterlevelindication iscomparedwithplantdatainFigureII.5.4-1.

Theinitialdecreaseinlevelwaspredicted byLOFTRANverywell.Thepressurizer wascalculated todrainby9:29(4min)andbegintorefillsoonafterassafetyinjection flowrepressurized theprimarysys-tem,asillustrated inFigureII.5.1-3.

Thepressurizer mayhavedraineda'econdtimebetween9:32(7min)and9:38(13min)duringcooldownviaAFMflow.Ginnadataindicates thatpressurizer levelreturnedonspanapproxi-matelywhenthechargingpumpswerestarted.Anindicated leveldidnotreturnintheLOFTRANanalysisresultsuntilthepressurizer PORYwascycledbeginning at10:07(42min).Asprimarypressuredecreased whenthePORVwasopened,pressurizer levelincreased assafetyinjection flowinexcessofbreakflow,FigureII.5.1-2, replacedventedsteaminthepressurizer.

Soonafterwards, atapproximately 10:09:20(44,min),

theupperheadwaterbegantoflash.Materdisplaced fromtheupperheadregionrapidly'increased pressurizer inventory andtheindi-catedlevelincreased offscale.

TheLOFTRANanalysisdemonstrated similarlevelresponse; however,theindicated levelremainedonspan.Thisappearstobedueprimarily tothelowerinitiallevelpriortodepressurization oftheprimarysystem.Inaddition, noreverseflowfromthefaultedsteam3274(:1/111982 52

700.00GINNA(G)LOFTRAtl(-)CICIAjC)TIHj(HtN)C)C)CDFIGUREII.5.3.6-2.LOFTRANUPPERHEADFLUIDTEMPERATURE.

53

120.00100.00GINNA(G)LOFTRAN(-)80.000W>60.000~a0.00020.0000.0ClClClClAJClTIME(MIN)FIGUREII.5.4-1.

PRESSURIZER LEVELINDICATION.

generator occurredintheLOFTRANresults(seesectionII.5.5).Upperheadvoidingmayhavealsobeenslightlyunderestimated becauseofthehomogeneous modelling, assuggested byAppendixCcalcualtions.

Thecalculated pressurizer leveldecreased rapidlyaftersafetyinjection wasterminated at10:37(72min)asbreakflowdecreased coolantinventory.

Theseresultsindicatethat95ft3ofwaterwasdisplaced fromthepressurizer asprimarypressuredecreased to945psia.Sucha.decrease wouldnothavebeendetectedbythelevelinstrumentation ifthepressurizer hadbeennearlywatersolid.Beyondapproximately 10:40(75min),thecalculated decreasein'pres-surizerlevelwasunrealistic.

Primary-to-secondary leakagewasexaggerated afterthistimebecauseoftheunrealistic faultedsteamgenerator pressurecalculated byLOFTRAN.II.5.5BreakFlowPrimary-to-secondary leakagewascalculated inLOFTRANassuminganeffective breakflowareaandamodifiedZaloudekcriticalflowcorrelation.

Forunchokedflow,theorificeequationwasused.FigureII.5.5-1showstheprimary-to-secondary leakagecalculated byLOFTRANduringtheGinnaevent.Priortoreactortrip,breakflowdecreased asprimarypressurealsodecreased.

Immediately aftertrip,'herapidlydecreasing primarypressuredecreased breakflowuntilsafetyinjection flowbegantorepressurize thereactorcoolantsystem.Soonafterthefaultedloophotleghadcooledbelowthetemperature ofthefaultedsteamgenerator, 9:37(12min),flowthroughthefailedtubewascal-culatedtobecomeunchoked.

Beyondthistime,thecalculated breakflowwassensitive tothefaultedsteamgenerator pressure.

Asillustrated inFigureII.5.5-2, thesecondary sidemodelling withinLOFTRANunderpredicted thefaultedsteamgenerator pressureafter9:46(21min).Consequently, secondary-to-primary flowdidnotoccurintheLOFTRANanalysiswhenthepres-surizerPORVwasopened.Inordertoevaluatethelimitations ofLOFTRANbreakflowmodelling andassesstheeffectsontheanalysisresults,amoredetailedmodel(Appendix B)wasdeveloped tocalculate theflowfromeachsteamgenerator plenum.The3274Q:1/111982 55 75.000LOFTRAil(-)BestEstimate(+)50.000~25;000Ko.o+4+p+++++++ttNt~++~+++++-25.000++-50.000ClClCDC)CIC)IPJC)ClleTINK(HIH)C)CDC)OCPDw~~DOFIGUREII.5.5.-1.

LOFTRANAHOBESTESTIHATEBREAKFLOWS.56 1200.01000.0800.00600.00~ioo.oo200.00GI~iNA(G)LOFTRAN(-)0.0C)CITIME(MIN)CDEOC)C)CCIFIGUREII.5.5-2.

FAULTEDSTEAI1GENERATOR PRESSURE.

57

fluidtemperature inthesteamgenerator inletandoutletplenumsweretakenfromtheLOFTRANresults.ActualGipnadatawasusedfortheprimaryandfaultedsteamgenerator pressures.

Resultsofthesecalculations arealsoshowninFigureII.5.5-1.

Asdemonstrated, LOFTRANprovidedareasonable estimateofthebreakflow,withtheexception ofreverseflowthroughthefailedtube,untilshortlyaftersafetyinjection wasterminated.

Afterthattime,thelowerfaultedsteamgenerator pressureevidentintheLOFTRANanaly-sisresultedinoverestimated primary-to-secondary leakage.II.5.6ReactorCoolantVoidingDuringnaturalcirculation, portionsofthereactorcoolantsystemmaystag-nateandbecomeeffectively isolatedfromtheactivecoolantregions.Threesuchregionsmayexistontheprimarysideduringrecoveryfromasteamgener-atortuberuptureevent;thepr'essurizer, reactorvesselupperhead,andthefaultedsteamgenerator tubes.Assystempressureisreduced,hotfluidinthesestagnantregionsmayflashtosteam.TheextentofvoidingintheseregionsduringtheGinnaeventwasevaluated.

II.5.6.1Pressurizer SteamBubbleItisclearfrompressurizer leveldata,FigureII.5.4-1, thatasteambubbleexistedinthepressurizer untiltheprimarysystemwasmanuallydepressurized beginning at10:07(42min).Atthattime,theindicated pressurizer levelincreased rapidlyoff-scale.

LOFTRANresultssuggestthatthepressurizer didnotcompletely fill,FigureII.5.1-3; however,aspreviously noted,theincreaseinpressurizer levelmayhavebeenslightlyunderestimated.

Thepressurizer levelresponsefollowing termination ofsafetyinjection suggeststhatthepressurizer wasnearly.fullatthattime.Specificly, pressurizer leveldidnotreturnonspanduringtheGinnaeventwhenprimarypressuredecreased by440psi.II.5.6.2UpperHeadVoidingThecalcu1ated upperheadtemperature history,FigureII.5.3.6-1, indicates thatvoidingmayhaveoccurredintheupperheadregionpriortoreactorcool-antpumptrip.Themaximumvolumeofthisvoidwasestimated tobelessthan3274Q:1/111982

132ft(Appendix C).Anysteambubbleintheupperheadatthistimewouldhavebeenquicklycondensed sincereactorcoolantpumpscontinued tooperate.Itisunlikelythatsignificant additional voidingoccurredpriortomanualdepressurization oftheprimarysystemat10:07(42min)sincethecalculated upperheadfluidtemperature remainedsubcooled.

From10:07(42min)to10:10(45min),upperheadthermocouple andpressurizer levelresponses indicatethatvoidingalsooccurredwhenthepressurizer PORVwasopened.Theupperheadtemperature decreased fromapproximately 556'FwhenthePORVwasinitially openedtoaminimumofapproximately 525F,asshowninFigureII.5.3.6-2.

Theupperheadregionwascalculated tocom-pletelyvoidduringthisdepressurization (Appendix C).Flashinginthenon-activeregionoftheupperplenum,i.e.abovethetopofthehotlegnozzles,wouldnotbeexpectedbecauseofrelatively goodmixingcharacteristics.

Consequently, approximately 305ftofsteamvolumeexistedintheupperheadwhenthepressurizer PORVwasisolatedat10:10(45min).Noinstrumentation wasavailable abovetheupperheadflangeleveltotrackthesteambubblecollapse.

However,assafetyinjection repressurized thereactorcoolantsystem,theupperheadthermocouples increased approximately alongthesaturation linefrom525Fto540'F.Aspressurecontinued toincrease, temperature thendecreased toastabletemperature of525'F.Thissuggestsflowofcolderfluidfromtheupperplenumpasttheflangelevelthermocouples andisindicative ofpartialsteambubblecollapse.

Thisissupported bytheslowerrepressurization ofthe'primarysystemfollowing iso-lationofthefailedPORVascomparedwithLOFTRANanalysisresults(seesectionII.5.1).Thesizeofanysteambubblewhichexistedintheupperheadregionwhensafetyinjection wasterminated at10:37(72min)isuncertain.

Themeasuredtemperature andpressureresponses suggestthattheupperheadwasnotcom-pletelyvoidedandcontained significantly subcooled water.Theupperheadprobablyvoidedathirdtimeasprimarysystempressuredecreased to945.psiafollowing termination ofsafetyinjection.

Analysisoftheprimary-to-secondary leakage,chargingflow,andreactorcoolantexpansion suggestsamaximumof125ftofadditional voidingmayhaveoccurredduringthisperiod.Thisvoidmayhaveexisteduntilreactorcoolantpumpswere'restarted.

3274/'1/111982

II.5.6.3BSteamGenerator TubeYoidingThefluidtemperature inthefaultedsteamgenerator tubeswascalculated withLOFTRANtobe507Fat10:07(42min)whenthepressurizer PORYwasfirstopened.Thisisconsistent withplantdatawhichshowsthatpressureintheBsteamgenerator haddecreased to750psiaby9:46(21min)andisconservative

'ithrespecttothecalculations presented insectionII.5.3.5.

Sincethecalculated tubebundlefluidtemperature remainedsubcooled duringdepres-surization ofthereactorcoolantsystem,nosteamvoidwouldhavedeveloped inthisregion.11.5.7SteamGenerator OverfillPrimary-to-secondary leakageinexcessofsteamfloweventually filledtheBsteamgenerator withwaterandliftedthesecondary safetyvalve.TheLOFTRANanalysisindicates thatthefaultedsteamgenerator andmainsteamline wouldhavefilled'at 10:18(53min),asshowninFigureII.5.7-1.

However,thisisbelievedtobeearlierthanduringtheactualeventforseveralreasons.Inordertosimulatethecooldownoftheprimarysystemfrom9:32(7min)to9:41(16min)(seesectionII.5.3.1),

steamreleasefromthefaultedsteamgenera-tortothecondenser wasterminated 8minutesprematurely at9:32(7min).Thisunderestimated thesteamreleasedfromthefaultedsteamgenerator tothecondenser byamaximumof11000ibm.Inaddition, theconstraints onupperheadrefillmayhaveincreased carryover intothefaultedsteamgenerator byamaximumof300ft..Thetotalcarryover mayhavealsobeenslightlyover-estimated byLOFTRANsincereverseflowduringdepressurization oftheprimarysystemwasnotpredicted.

Thecombination oftheseeffectsmayhavedelayedoverfillbyanestimated 7minutes.Theinitialsafetyvalveliftsat10:19(54min)and10:27(62min)wouldhavealsodecreased steamgenerator inven-toryandfurtherdelayedoverfill.

Themassdischarged throughthefaultedsteamgenerator safetyvalvewasesti-matedfromtheprimary-to-secondary leakage.LOFTRANresultsindicatethat268,000ibmofprimarycoolantwastransferred intothefaultedsteamgener-atorpriortotermination ofsafetyinjection.

Approximately 104,000ibmofthisleakageoccurredafterthesteamgenerator wascalculated tofill.Eval-uationofthebreakflowfrom10:40(75min)until12.30(185min)asshownin3274/:1/111982

'j.OOE+OiiLOFTRAN(-)8000.06000.0S.G.andSteamline VolumeS.G.Volume~F000.0Z=I2000.00.0ClC)C7AJTIME(MIN)C)EDC)8FIGUREII.5.7-1.

FAULTEDSTEAYiGENERATOR WATERVOLUf1E.6l

FigureII.5.5-1suggeststhananadditional 132,000ibmwastransferred from~theprimarybeforeprimary-to-secondary leakagewasterminated.

II.6LONGTERMRECOVERYWhensafetyinjection wasterminated at10:37(72min),primarysystempres-suredecreased rapidlyfrom1370psiato945psia.TheLOFTRANanalysiswasterminated atthistimesincethehomogeneous equilibrium modelling onthesecondary sideoverestimates theprimary-to-secondary pressuredifferential and,consequently, leakagethroughthefailedtube.Continued chargingflowandoperation ofthepressurizer heatersmaintained primarypressureslightlygreaterthanthefaultedsteamgenerator

pressure, asshowninFigureII.6-1.TheBsteamgenerator pressureincreased asprimary-to-secondary leakagecon-tinued.Thesequenceofevents'indicates thatsafetyinjection wasreinitated at11:07(102min)inpreparation forreactorcoolantpumprestart.However,Jtheeffectoftheincreased coolantmakeupontheprimaryandfaultedsteamgenerator pressures isnotevidentatthattime.Althoughtheprimary-andfaultedsteamgenerator pressures increased slowly,thepressuredifferential decreased.

Atapproximately ll19(114min),arapiddecreaseinprimarysystempressureisevident,probablyduetothecollapseofanupperheadsteambubble(seesectionII.5.6.2) whenareactorcoolantpumpwasrestarted.

Althoughtheavailable dataislimited,itappearsfromtheBsteamgenerator pressureresponsethatasafetyvalvemayhaveliftedatapproximately thesametime.Safetyinjection flowrepressurized thereactorcoolantsystembeginning atll:26(121min)untilflowwasthrottled at11:35(130min).Thefaultedsteamgenerator pressurealsoincreased untilthesafetyvalveliftedforthefinaltimeatapproximately 11:37(132min).Afterthisfinallift,thefaultedsteamgenerator pressureremainedapproximately 150psiabelowtheprimarysuggesting continued leakageintothesteamgenerator.

Steamgener-atorblowdownlineradiation wasalsoincreasing duringthissameperiodsug-gestingflowthroughthisline.Pressurizer levelreturnedonspanatapprox-imately11:53(148min),asshowninFigureII.6-2,andcontinued todecreaseindicating alossofreactorcoolant.Asafetyinjection pumpwasoperatedintermittently from12:13(168min)until12:27(182min)tocontrollevel.By12:30(185min),thefaultedsteamgenerator pressurewasgreaterthahthereactorcoolantsystempressureandprimary-to-secondary leakagewasterminated.

3274(}:1/111982 62 1300.0GINNA(G)ccccccGGCCGCGCRCS(-G-)CCC6@QCCCCCCCCG CGccccc'ccCcCCCCCGGFAULTEDSG(G)CITIME(MlN)C7EDC7C7ClC7C)r~FIGUREII.6-1.RCSANDFAULTEDSTEAt1GENERATOR PRESSURES.

63

120.00100.0080.000GINNa(G)CGCCCGCCGCCGCGCGGCGGCCCGCCCGGCCGGCGCCGCCCCCCCGC CCGGccGGCCGGGCGo60.000~io.ooo)CGGGGGCCCCCG'CCc20.0000.0CDCDCDCDCDCDCDCDCDAJCDCDmCDCDCD.CDCDCDCDCDCDCDTIME(MtN)fIGUREII.6-2.LONG-TERtl PRESSURIZER LEVELRESPONSE.

III.SUMMARYANDCONCLUSIONS ThemaximumleakratethroughthefailedtubeduringtheGinnaeventwascal-culatedtobe634gpm.Adesignbasiseventwithconservative, FSARassump-tionsrepresents aninitialprimary-to-secondary leakrateof1147gpmforthesamesteamgenerator.

Hence,theinitialleakratewassignificantly 1'essthandesignbasis.Breakflowdepletedprimarycoolantinventory andresultedinautomatic reac-tortripandsafetyinjection withinapproximately 3minutesoftheinitiating event.Primarypressuredecreased rapidlyfollowing reactortripascoolant'temperature decreased andbreakflowfurtherreducedcoolantinventory.

Man-ualreactorcoolantpumptrip,whichoccurredwithin1minuteofreactortrip,wasfollowedbyasmoothtransition fromforcedtonaturalcirculation inbothloops.Naturalcirculation wasmaintained intheintactloopuntilareactorcoolantpumpwasrestarted.

Isolation ofthefaultedsteamgenerator incom-binationwiththecooldownoftheintactloopeventually stagnated flowinthefaultedloop.Analysisresultsandevaluation oftheBloopcoldlegtempera-turesuggestthatacounter-current, flowpatternmayhavedeveloped inthefaultedloopcoldlegupstreamoftheinjection nozzle.Withtheexception oftheupperheadregionandthepressurizer, thereactorcoolantsystemremainedsubcooled throughout theevent.Theupperheadmayhavevoidedthreetimes.Immediately following reactortrip,avoidmayhave'eveloped beforereactorcoolantpumpsweretripped.Noadditional voidingoccurreduntilthepressurizer PORVwasmanuallyopenedtodepressurize theprimarysystem.Theupperheadregioncompletely voidedfrom10:07(42min)to10:10(45min)resulting ina305ftsteambubble.Therelatively slowrepressurization ofthereactorcoolantsystemfrom10:10(45min)to10:17(52min)andtheupperheadthermocouple responsesuggestthatthissteambubblewasatleastpartially collapsed by10:37(72min).However,addi-tionalvoidingoftheupperheadprobablyoccurredathirdtimewhensafetyinjection wasterminated.

Thisvoidmayhaveexistedwhenthereactorcoolantpumpwasrestarted atll:19(114min).AlthoughLOFTRANresultsindicatethatthepressurizer didnotfillwithwater,pressurizer levelresponsefollowing termination ofsafetyinjection indicatethatthepressurizer wasnearlyfull.3274(}:1/111882 65 Thefaultedsteamgenerator wasestimated tohavefilledwithwaterbyapprox-imately10:25(60min).However,releasesduringtheearlysafetyvalveliftsmayhavereducedsteamgenerator inventory anddelayedoverfill.

Anestimated 400,000ibmofprimarycoolantweretransferred tothefaultedsteamgener-ator.Approximately 253,000ibmwascalculated tobedischarged fromthefaultedsteamgenerator untilprimary-to-secondary leakagewasterminated at12:30(185min).Anestimated 28000ibmwasreleasedassteamtothecon-denser.Theremaining 225000ibmrepresents anestimateofthetotalreleasefromthefaultedsteamgenerator safetyvalve.Consideration oftheuncer-taintyassociated withfeedwater flowtothefaultedsteamgenerator andrefilling oftheupperheadindicates thatthisestimatemaybeconservative byupto48000ibm.Inaddition, thecalculated leakageintothefaultedsteamgenerator from10:40(75min)until12:30(185min)reliesonmeasuredsystempressures whicharesubjecttoinstrument uncertainties.

3974n.1/111789 66

'

REFERENCES 1.LicenseeIncidentEvaluation ReportontheJanuary25,1982SteamGenerator TubeRuptureIncidentattheR.E.GinnaNuclearPowerPlant,DocketNo.50-144,April'(1982).2.NRCEvaluation oftheJanuary25,1982SteamGenerator TubeRuptureIncidentattheR.E.GinnaNuclearPowerPlant,NUREG-0909, April(1982).3.L.A.Campbell, et.al.,LOFTRANCODEDESCRIPTION, WCAP-7878, Rev.3,'anuary(1977).4.L.A.Campbell, et.al.,WESTINGHOUSE EYALUATION OFLICENSEEEVENT,No.SG79-11-030, Dec.(1979).5.F.R.Zaloudek, "Steam-Mater CriticalFlowFromHighPressureSystemsInterimReport".,

HanfordAtomicProductsOperation,

Richland, Washington, TID-4500, Jan.(1964).6.J.A.Block,FLUIDTHERMALMIXINGINAMODELCOLDLEGANDDOWNCOMER WITH~~~LOOPFLOW,CREAREInc.,Hanover,NewHampshire, EPRI-NP-2312, April(1982).7.S.LevyandJ.M.Lealzer,ANAPPROXIMATE PREDICTION OFHEATTRANSFERDURINGPRESSURIZED THERMALSHOCKWITHNOLOOPFLOWANDWITHMETALHEATADDITION, S.LevyInc.,Campbell, California, SLI-8220August(1982).3274(:1/111882 67 1

APPENDIXA:INITIALLEAKRATECALCULATION Theindicated pressurizer leveldecreased from32.5Xto11.7%over104seconds,asshowninTableII.3-1.Thislevelwasadjustedforpressurizer pressureasfollows:LPRZLINDX1+0/V1i0/V+100X'refVfVg10/V10/Vgrefwhere,PRZINDV~SubfSubgSubrefactualpressurizer levelindicated pressurizer levelfluidspecificvolumereferstosaturated liquidreferstosaturated vaporreferstonominalsystemconditions LPRZ9:26:18)=32.5x1.00.02698~~~=32.7LPRZ9:28:02)=11.7x1.0002698=14.10.1569+"'.1569'.17460.02617'.174610.15690.15690.19470.025430.1947Duringthistime,coolantinventory wasdepletedatanaveragerateof=1.202-=538GPM33BRK104sec5span'ecConsidering anexcessof'5gpmfromthechargingsystem,theaverageleakratewasapproximately 573gpm.Theinitialleakratewascalculated byextrapolating theaverageratetotheinitialsystemconditions of2250psiaand601F.Theaveragepressureandtemperature overthepre-tripperiodwereapproximately 2100psiaand601F,respectively.

Basedonsubcooled criticalflowthroughthebreak,theinitialflowratewasestimated tobe3274/:1/111682 68

qggy(0)=573x(2250-0.9x1555)x0.02336xx0.5=634GPN32740:1/']11682 69

APPENDIXB:BESTESTIMATEBREAKFLOWMODELFollowing asteamgenerator tubefailure,primarycoolantflowsthroughthebreakintothesecondary sideofthesteamgenerator.

Theprimary-to-secondary pressuredifferential providesthedrivingforceforthisflow.Thefailuresiteisconnected totwoprimaryfluidreserviors, i.e.steamgener-atorinletandoutletplenums,viathesegmented tube.Eachsegmentp'rovides asubstantial resistance tofluidflow.Forlargertubefailures, thisresis-tancerepresents alargefractionofthetotalresistance betweentheprimaryandsecondary systems.ThebreakflowmodelwithinLOFTRANdoesnotconsiderfrictional orformpressurelossesthrougheachtubesegment.Forcriticalflowconditions, thismaynotbeasignificant limitation sincethepressuredropisessentially localized atthebreaklocationorentrancetothefailedtube.However,temperature differences betweentheinletandoutletplenumswillaffectcriticalflowifentrancechokingoccurs.Thistemperature effectisalsonotsimulated withinLOFTRAN.Furthermore, theprimary-to-secondary pressuredifferential isnotaccurately predicted.

Consequently, amoredetailedmodelwasdeveloped whichusesprimaryandsecondary pressuredataincombination withfluidtemperature resultsfromLOFTRANtocalculate breakflow.Flowthroughthefailedtubewassimulated asshowninFigureB-l.Frictional pressurelossesthrougheachtubesegmentwererepresented byanappropriate singlephasefrictionfactor,length,anddiameter.

Entranceandexitlossesforeachtubesegmentandatthebreaklocationwerealsoincluded.

Thissystemleadstothefollowing setofsimultaneous equations whichdescribeflowthroughthefailedtube:xWBRK-SGBRKBRKBRK2xgcxABRK(B-1)RCSBRK2IPLIPLIPLENTEXTD22xcxATUBERCSBRK"LopLENTEXTIJ3274t}:1/111882 7O

FIGUREB-1.SGTUBERUPTUREFLOWMODELDIAGRAt1.

~RCSTzpLLxswj~GGGPIWLaoc.FRC5Topi71 BRKIPLOPL(B-4)where,PVENTEXTADgcSubBRKSubRCSSubIPLSubOPLSubTUBEpressureflowratefluidspecificvolumeTubeentrancelosscoefficient

=0.4,primary-to-secondary flow=1.0,secondary-to-primary flowTubeexitlosscoefficient ATUAE21.0-BRK=1.0,secondary-to-primary flowMoodyfrictionfactorflowareatubediametergravitational constantreferstobreaklocationreferstoprimarysidereferstosteamgenerator inletplenumreferstosteamgenerator outletplenumreferstosteamgenerator tubeCriti'cal flowthroughthefailedtubewascalculated usingamodifiedZaloudekcorrelation forsubcooled criticalflow.Chokedflowconditions foreachtubesegmentandatthebreaklocationwerecalculated fromWCIL=ATUBExClxWCOPL=ATUBExC2x2gfPRCS-C2xPsat(TIPL)]

IPLRCS2tOPL"OPL1/21/2(B-5)(B-6)WCBRK=ABRKxClx2gLPBRK-C2xPsat(TBRK)]

1/2BRK(B-7)32740:1/111882 72

where,TsatWCClC2'lui'dtemperature, Fsaturation

pressure, psiacriticalmassflowrate,ibm/secentranceeffectcoefficient (adjusted tomatchinitialleakrate)0.9Equations B-lthroughB-7weresolvedsimultaneously forbreakflowthroughthefailedtube.32740:1/111682 73

APPENDIXC:CALCULATION OFUPPER,HEADVOIDSIZETheupperheadregionofthereactorvesselwasmodelledasasingle,strati-fiednodewithonlyoutwardflowas.showninFigureC-l.Assystempressuredecreased belowsaturation oftheupperheadfluid,voidingwithintheupperheadregiondisplaced liquidintotheupperplenum.Theextentofvoidingwasestimated assumingthermodynamic equilibrium betweenphases.Metalintheupperheadregionwasconservatively assumedtobeatthefluidtemperature.

Amassandenergybalancebetweeninitialandfinalstateswithintheupperheadvolumeleadstothefollowing expression forthefractionoffinalsteamvolumeVUH,(hh)/V+(7hhf)/oVf(E-hf)(h-K)V+gVVfMC(TT)VUHosatwhere,=fluidspecificvolume'fluidenthalpy=VolumehT(MC)SuboSubfSubgSubUHaverageenthalpyofdisplace'd fluidfluidtemperature metalheatcapacityreferstoinitialcondi.tions referstosaturated liquidreferstosaturated vaporreferstoupperheadregionTableC-1liststheupperheadconditions forthefirsttwoincidents ofpotential upperheadvoidingduringtheGinnaevent(seesectionII.5.6.2).

Thecalculated upperheadregionvoidfractions fromequationC-1arealsopresented.

Fortheseresults,theenthalpyofdisplaced liquid,was assumedtobethelinear.averageoftheinitialandfinalstates.32740:1/111882 FIGUREC-1.UPPEP.HEADVOIDINGILLUSTRATION.

XII'TIAtVtLLH)hIII75 TABLE0-1UPPERHEADVOIDSIZETirade9:28:30Pp(Psia)1300Tp(Psia)577.5PF(Psia)1200Vg/VUH0.431V(Ft~)13210:0710985568451.030532740:1/11168276