ML18017A245: Difference between revisions
StriderTol (talk | contribs) (Created page by program invented by StriderTol) |
StriderTol (talk | contribs) (Created page by program invented by StriderTol) |
||
Line 40: | Line 40: | ||
PROPRIETARYEXHIBIT:"A".TABLE4-7TOTAL~U='.)CHERLOADSDURTi'IGSRVOPENING~1>7LoadHaximumValueDirectionTimeHisto~rinternaloverpressure27bars(377psiq)SeeFigure4-7ExternalloadhHaterdeflectionloadinsidethequencherTorque44kn<2>Simultaneouslyin(9891lb)thehorizontalandverticalquencherplanes620knVertical(139,376lb)40knmEnhorizontal.(29,501quencherplaneft-lb)SeeFigure4-8SeeFigure4-5SeeFigure4-9Externalloadduetobubbleoscillation(SeeSubsection4.1.2.4)<1>.Forthecaseofaslidingjointinthedischa"qelineclosetothequencher(Fiqure4-10),tnepressureinsidethepipeactsasanexternalforce.Thiscasisshownin,Fiqure4.11.,<2>Effectsofasymmetricholearrangementareincluded. | PROPRIETARYEXHIBIT:"A".TABLE4-7TOTAL~U='.)CHERLOADSDURTi'IGSRVOPENING~1>7LoadHaximumValueDirectionTimeHisto~rinternaloverpressure27bars(377psiq)SeeFigure4-7ExternalloadhHaterdeflectionloadinsidethequencherTorque44kn<2>Simultaneouslyin(9891lb)thehorizontalandverticalquencherplanes620knVertical(139,376lb)40knmEnhorizontal.(29,501quencherplaneft-lb)SeeFigure4-8SeeFigure4-5SeeFigure4-9Externalloadduetobubbleoscillation(SeeSubsection4.1.2.4)<1>.Forthecaseofaslidingjointinthedischa"qelineclosetothequencher(Fiqure4-10),tnepressureinsidethepipeactsasanexternalforce.Thiscasisshownin,Fiqure4.11.,<2>Effectsofasymmetricholearrangementareincluded. | ||
PROPRIETARYTABLE 8.EXHIBIT"A"~TOTALURNCHRRLOADSDURINGSRVCLOS1NGLoadMaximumValueDirectionTimeHistoryExternalLoadTorqu~4.5kn(1012lb)6knm(4425ft-lb)Simultaneouslyin-.hehorizontalandverti-calquencherplanesXnhorizontalquencherplaneSeePigure4-12SeeFigure.4-'12 | PROPRIETARYTABLE-4-8.EXHIBIT"A"~TOTALURNCHRRLOADSDURINGSRVCLOS1NGLoadMaximumValueDirectionTimeHistoryExternalLoadTorqu~4.5kn(1012lb)6knm(4425ft-lb)Simultaneouslyin-.hehorizontalandverti-calquencherplanesXnhorizontalquencherplaneSeePigure4-12SeeFigure.4-'12 | ||
PROPRIETARYTABLE4-9TOTALQUENCHERLOADSDURINGIRREGULARCONDENSATIONEXHIBIT."g".1LoadMaximumValueDirectionTimeHistoryExternalload317.5kn(3934lb)Simultaneouslyinthehoizontaland,verticalquencherplanesSeeFigure4-13Torque19knmInhorizontal(14,013quencheplane.Ct-lb)SeeFigure4-13 1 | PROPRIETARYTABLE4-9TOTALQUENCHERLOADSDURINGIRREGULARCONDENSATIONEXHIBIT."g".1LoadMaximumValueDirectionTimeHistoryExternalload317.5kn(3934lb)Simultaneouslyinthehoizontaland,verticalquencherplanesSeeFigure4-13Torque19knmInhorizontal(14,013quencheplane.Ct-lb)SeeFigure4-13 1 | ||
PROPRIETARYTABLE4-11~UENCHERARSLOAQSDURINGSRVCLOSINGEXHIBIT"A"'oad-'HaximumValueDirectionTimeHistorvinternaloverpressureExternalloadBendingmomentonmeldingseamatintersectionbetweenquencnera"mandquencherball22bars(304psiq)4.5Kn{1012lb)3Knm(2213tt-lb)SeeFigure4-18Simultaneouslyinthehorizontalandvertica1planesSeePigure4-19SimultaneouslySeeFigure4-19inthehorizonta1.andverticalplanesTherma1load-2190C(Internaltemper-=(426~-:)atu"-)SeeFigure4-18 f,11 PBOPREETABYTABLE4-12.EXHlHIT"g"QUENCHERARMLOADSDURIMGIRREGULARCOMDENSAIIOMLoadmaximumValueDirect.ionTimeHistoryEnternalpressureExternalload3.0bars(28.8psig)14.5KnSimultaneously(6638inthehorizontalft-lb)andverticaldirectionSeeFigure4-20SeeFigure4-21Bendingmoment,onweldingseamatintersectionbetweenquenche:armandquencherball9Knm(66380t-lb)Simultaneouslyinthehorizons.alandverticalplaneSeeFigure4-21Thermalload(Entlnaemuature)1330C(271.4~F)SeFigure4-20 | PROPRIETARYTABLE4-11~UENCHERARSLOAQSDURINGSRVCLOSINGEXHIBIT"A"'oad-'HaximumValueDirectionTimeHistorvinternaloverpressureExternalloadBendingmomentonmeldingseamatintersectionbetweenquencnera"mandquencherball22bars(304psiq)4.5Kn{1012lb)3Knm(2213tt-lb)SeeFigure4-18Simultaneouslyinthehorizontalandvertica1planesSeePigure4-19SimultaneouslySeeFigure4-19inthehorizonta1.andverticalplanesTherma1load-2190C(Internaltemper-=(426~-:)atu"-)SeeFigure4-18 f,11 PBOPREETABYTABLE4-12.EXHlHIT"g"QUENCHERARMLOADSDURIMGIRREGULARCOMDENSAIIOMLoadmaximumValueDirect.ionTimeHistoryEnternalpressureExternalload3.0bars(28.8psig)14.5KnSimultaneously(6638inthehorizontalft-lb)andverticaldirectionSeeFigure4-20SeeFigure4-21Bendingmoment,onweldingseamatintersectionbetweenquenche:armandquencherball9Knm(66380t-lb)Simultaneouslyinthehorizons.alandverticalplaneSeeFigure4-21Thermalload(Entlnaemuature)1330C(271.4~F)SeFigure4-20 | ||
PROPRIETARYCHAPTER8SSESQUENCHERVERIFICATIONTESTTABLEOFCONTENTS81INTRODUCTION8118.1.281.218.12.1.181.21.2812.281.2218.12.2.281.2.2.38.122-48.1-2.25PurposeofTestsTestConceptSingleCellApproachSingleCellTheoryApplicationofSingleCellApproachSimulationofSSESParameters'rimarySystemPressureSafetyBeliefValve(SBV)DischargeLineVacuumBreakersQuencherBEV1,3r79 82TESTFACILITYANDINSTRUMENTATION8.2.182.1.18.2.1118.2.1128.2.1138211.48.21.158.21168228.2.21822.282.238.2.231822.328.2.2.48.22.4.182.24282.2.5TestFacilityMechanicalSet-UpSteamBoilerSteamAccumulatorSteamLineandBufferTankSafety/RelicfValve(SRV)DischargeLineandQuencherTestTankInstrumentationGeneralDescriptionInstrumentationIdentificationOperatingInstrumentationDisplayonControlConsoleAcquisitionbyComputerTestInstrumentationMeasuringPointsSet-UpofMeasuringInstrumentsVisualRecordingREVli3/798-2 83TESTPARAiiETEHSANDMATRIX83I83.2VentClearingTestsCondensationTestsREV1,3/798-3 84'ZESTRESULTS84.1841.184128.413VentClearingTestResultsTestParametersBehavioroftheSRVandSystemPress'uresDynamicPressureLoadsonthePoolBoundaries841.48428.42.18.4.2.2842218.4.221.18.42.2.12842.2.2LoadsontheQuencherandBottomSupportSteamCondensationTestResultsTestParametersPresentationofTestResultsSurveyofObservedCondensationPhasesBlowdownatLowWaterTemperatureBlowdownatHighMaterTemperatureStatisticalEvaluationoftheDynamicPressureLoadsonthePoolBoundaries8422.21DependenceofDynamicBottomandWallPressuresonSystemPressureandWaterTemperature842222842223OccurrenceFrequencyDistributionsoftheDynamicBottomandMallPressuresStatisticalCharacteristicsoftheDynamicBottomandMallPressures84223TemperatureVariationsintheMaterRegionoftheTestTank8422.4843MaterLevelintheDischargeLineWhenOpeningandAfterClosingtheSRVCheckingandCal.ibrationoftheMeasuringInstrumentation844845AnalysisofMeasurementErrorsRepetitionTestsandReproducibilityoftheResultsREVl,3/798-4 85DATAANALYSISANDVERIFICATIONOFLOADSPECIFICATION8.5.18.5.11EvaluationofTestTankEffectsonBoundaryPressureMeasurementsEffectsofFreeWaterSurfaceandRigidWalls8.5128513851.4MethodofImagesTheTestStandasaSingleCellSpatialDistributionofPressureintheTestTank8.5.1.58515185152851.5.38515.4InvestigationoftheInfluenceofMovableWallsontheMeasurementResults(Fluid-StructureInteraction)GeneralRemarksExperimentalInvestigationoftheTank'sNaturalOscillationsExperimentalInvestigationoftheTank'sResponsetoVentclearingLoadsTheoreticalInvestigationsandModelCalculationsoftheInfluenceofFSI85.1.5.41851.5428.51543ComputationModelsModelParametersandInputforCalculationsWithoutFSI(RigidTank)ModelParametersandInputforCalculationsWithFSI85.1.544ResultsoftheFSICalculations8.5.285.21VerificationofSRVSystemLoadSpecificationDuetoSRVActuationPressuresDuringtheVentClearingProcess852118.5212VentClearingPressuresfortheLongLineVentClearingPressuresfortheShortlineREV.l.3/798-5 8521.3Transposition-oftheMeasurementValuestoSSESandComparisonwiththeDesignSpecification852.2PressuresDuringtheStationaryCondensationofSteam8522185222LongLineShortLine8.522.3TranspositionoftheMeasurementValuestoSSESandComparisonwiththeDesignSpecification8.52.3ExternalLoadsontheQuencherandBottomSupport8.5.2318.5.2.3.1.185231.2852312.185.231.2.2VerticalForceMeasurementoftheVerticalForceMeasuredVerticalForcesLongLineShortLine85231.3TranspositionoftheMeasurementValuestoSSES8523131852.31,.3.28.5.2.313-3852328523218.5.2.3-2.285232.2185232.22852323LongLineShortLineSummaryTorsionalMomentMeasurementoftheTorsionalMomentMeasuredTorsionalMomentslongLineShortLineTranspositionoftheMeasurementValuesto'SSES85-2338.5.2.3.31BendingMomentsattheQuencherArmsMeasurementoftheBendingMomentsREV1,3/798-6 8.5.2-33.28.5.2.3.3.3MeasuredBendingMomentsTranspositionoftheMeasurementResultsIntotheWeld8.5.233.4852.33.58.5-2.3.48523418.5234.28.5.2.3438.523448.5.2358-52-3.6SpecifiedStaticEquivalentLoadsEvaluati.onoftheMeasurementResultsBendingMomentsattheBottomSupportMeasurementoftheBendingMomentsMeasuredBendingMomentsSpecifiedStaticEquivalentLoadEvaluationoftheMeasurementResultsForcesontheQuencherInfluenceofanAdjacentQuencher~8-5.237LoadsontheQuencherDuringSteamCondensation852371ManifestionFormsofIntermittentCondensationintheKarlsteinTests8523728.5.237.3IllustrationoftheMeasurementValuesEvaluationoftheMeasurementResultsforthe.QuencherArm85237.4EvaluationoftheMeasurementResultsfortheBottomSupport8.5-2.375EvaluationoftheMeasuredTorsionalMoments852376EvaluationoftheMeasuredMaximumMomentsattheQuencherArmDuringIntermittentCondensation85.3VerificationofSuppressionPoolBoundaryLoadSpecificationDuetoSRVActuation8531EvaluationoftheLocalEffectsSeenatPressureTransducerP5.58532Veri.ficationoftheSpecifiedPressureAmplitudesandVerticalPressureProfilesafterVentClearingREVl~3/798-7 8532185321.185.3212OverpressuresVerticalPressureProfileVerticalPressureProfileIncludingLocalEffectsatP5.58532285322.185.3385.3.3.185331.18.5.33.11.1853311.28.5.3.3.1.1.385.331.14UnderpressuresVerticalPressureProfileVerificationofthePressureTimeHistoriesUsedfortheSSESContainmentAnalysisTranpositionMethodfortheOscillationFrequencyCalculationofMeasuredOscillationFrequenciesPPGLTestsatKarlsteinGKMModelQuencherTests'KBHotTestsConclusionfromtheFrequencyCalculations8.5.3.328.5.3.3.38533.3.18.53332MultipliersforConversionoftheBubbleFrequenciesFromtheTestStandtoSSESTranspositionMethodforthePressureAmplitudesPPGLQuencherTestsatKarlsteinKMUQuencherTestsintheModelTestStandinKarlstein8.53.3.338533.34853348533.41853.34.2AnalyticalCalculationsInfluenceofBackpressureonthePressureAmplitudesVerificationofDesignSpecificationFrequencyAnalysesofSelectedTestsShiftingofthePSD'sintheTranspositionFromtheTestStandtoSSESREV.li3/798-8 853.34.2l85.334.228533.4385.3.3.4485334585.33.4.6FrequencyShiftAmplitudeStretchingSymmetricalLoadCase(SimultaneousBlowdownofall16SRV's)UnsymmetricalLoadCase(BlowdownViaOneSRV)UnsymmetricalLoadCase(BlowdownViaThreeAdjacentSRV')AutomaticDepressurizationSystem(ADS)LoadCase853347853.3.58.533.518533.5285.33.53853.354SummaryEvaluationoftheMeasuredPressureOscillationsDuringCondensationTheQuencherisClearedContinuallyTheQuencherisNotClearedContinuallyCondensationintheBlowdownPipeandThrutheSlidingJointTransportationoftheMeasurementResultstoSSES854854.18.5.4285.438544PoolMixingDuringSRVActuationandThermalPerformanceoftheQuencherIntroduction'EquationofMotionoftheRotatingPoolDeterminationoftheFlowResistancesDeterminationoftheForceMovingthePool85.4.5WorkingEquationsfortheRotatingPoolofSSES854.68.5.4.7854.7185472EstimateoftheHeatingoftheSuppressionChamberWaterExperimentalProofsModelTankTestsKKBTestDuringtheNuclearCommissioningREVl,3/798-9 8.5.4.7-3GKMHalfScaleQuencherCondensationTest8.5488.5.5SummaryVerificationoftheSubmergedStructuresLoadSpecificationDuetoSRVActuation8.5.5.1855118551285513LoadsontheVentPipeMeasurementoftheLoadsMeasuredBendingMomentsExtrapolationoftheMeasurementResultsandComparisonwiththeSpecifiedValue8.55.2InfluenceofExpelledHaterDuringVentClearing8553SummaryREVl,3f'798-10 SECTION8.0FIGURESNumberTitle8-1MathematicalDesscriptionofaSingleCellConfigurationwithSolidWalls;SolidBottomandFreeWaterSurface8-28-38-48-58-68-7EguivalenceofaSingleCellConfigurationandaParallelBubbleFieldOscillatinginPhaseGeometricSingleCellPartitionoftheSuppressionPoolTestStandSchematicDiagramLongDischargeLineConfigurationShortDischargeLineConfigurationKarlsteinTestTankPlanVievTypicalVentClearingInstrumentation8-8KarlsteinTestTankC-DVievTypicalVentClearingInstrumentation8-9KarlsteinTestTankA-BVievTypicalVentClearingInstrumentation8-10KarlsteinTestTankPlanVievTypicalCondensationTestInstrumentation8-11KarlsteinTestTankC-DViewTypicalCondensationTestInstrumentation8-12KarlsteinTestTankA-BViewTypicalCondensationTestInstrumentation8-13T-QuencherShowingTypicalVentClearingInstrumentation8-14T-QuencherShovingTypicalCondensationTestInstrumentation8-158-168-178-188-19TestMatrigforVentClearingTestLocationofTestGroupNo.1intheOperationFieldLocationofTestGroupNo.2intheOperationFieldLocationofTestGroupNo.3intheOperationFieldLocation'fTestGroupNo.4intheOperationFieldREV1,3/798-11 8-208-218-228-23LocationofTestGroupNo.5intheOperationFieldLocationofTestGroupNo.6intheOperationFieldLocationofCondensationTestsintheOperationFieldValveOpeningTimeVersusAccumulatorPressureLongPipeVentClearingTests8-24ValveOpeningTimeVersusAccumulatorPressureShortPipeVentClearingTests8-25VentClearingPressureVersusSystemPressureLongLineVentClearingTests8-26VentClearingPressureVersusSystemPressureShortLineVentClearingTests8-27PeakPositiveWallandBottomPressuresVersusSystemPressure-Long.Line,CleanConditions,ColdPool8-28PeakPositiveMallandBottomPressuresVersusSystemPressure-ShortLineCleanConditions,ColdPool8-29PeakPositiveWallandBottomPressuesVersusSystemPressure-LongLineRealConditions,ColdPool8-30PeakPositiveWallandBottomPressuresVersusSystemPressure-ShortLine,RealConditions,ColdPool8-31PeakPositiveMallandBottomPresssuresVersusSystemPressure-LongLine,CleanConditions,Heatedpool8-32PeakPositiveWallandBottomPressuresVersusSystemPressure-ShortLine,CleanConditions,HeatedPool8-33PeakPositiveWallandBottomPressuresVersusSystemPressure-LongLine,RealConditions,HeatedPool8-34PeakPositiveMallandBottomPressuresVersusSystemPressure-ShortLine,RealConditions,HeatedPool8-35PeakPositiveMallandBottomPressuresVersusValveActuation-LongPipeTest148-36PeakPositiveWallandBottomPressuresVersusValveActuation-LongPipeTest58-37PeakPositiveMallandBottomPressuresVersusValveActuation-LongPipeTests4and4R8-38PeakPositiveWallandBottomPresuresVersusValveActuation-LongPipeTests15and15RREV1,3/798-12 8-3'98-408-418-428-438-448-458-468-478-488-498-508-518-528-538-548-558-568-578-588-598-608-618-628-638-648-65PeakPositiveMallandBottomPressureVersusValveActuation-ShortPipeTests19and19RPeakPositiveMallandBottomPressuresVersusValveActuation-ShortPipeTests20and20RVisicorderTraceP51-P5.10Test41.1.VisicorderTraceP5.1-P5.10Test4R.l.1VisicorderTraceP5.1-P5.10Test4.1.6VisicorderTraceP5.1-P5.10Test11.1VisicorderTraceP5.1-P5.10Test12.1VisicorderTraceP5.1-P5.10Test15.1.1VisicorderTraceP5.1-P5.10Test15.Rl.1VisicorderTraceP5.1-P5.10Test19.1.1VisicorderTraceP5.1-P5.10Test19.R2.1VisicorderTraceP5.1-P5.10Test19.R2.2VisicorderTraceP5.1-P5.10Test19.R2.3VisicorderTraceP5.1-P5.10Test19.R2.4UisicorderTraceP5.1-P5.10Test19.R2.5VisicorderTraceP5.1-P5.10Test19.R26VisicorderTraceP51-P510Test19.R2.7VisicorderTraceP5.1-P5.10Test19R2.8VisicorderTraceP5.1-P5.10Test19.R29VisicorderTraceP5.1-P5.10Test19.32.10VisicorderTraceP5.1-P5.10Test20.1.1VisicorderTraceP5.1-P5.10Test20.Rl.lVisicorderTraceP5.1-P5.10Test20.R1.10VisicorderTraceP5.1-P5.10Test21.1VisicorderTraceP5.1-P5.10Test21.2VisicorderTraceP5.1-P5.10Test25.1VisicorderTraceP5.1-P5.10Test25.R2REVlg3/798-13 8-66MaximumResultantBendingMomentatQuencherArm1-,LongPipeVentClearingTests8-67MaximumResultantBendingMomentatQuencherArm2-LongPipeVentClearingTests8-688-69MaximumResultantBendingMomentatQuencherArm1-ShortPipeVentClearingTestsMaximumResultantBendingMomentatQuencherArm2-ShortPipeVentClearingTests8-70MaximumResultant'BendingMomentattheQuencherSupport-LongPipeVentClearingTests8-718-728-73MaximumResultantBendingMomentattheQuencherSupport-ShortPipeVentClearingTestsObservedCondensationPhasesDuringTestsTypicalVisicorderTraceofStationaryOperationofQuencherTest33.2-10SecondsafterStart8-74TypicalVisicorderTraceofStationaryOperationofQuencherTest35.1-20-SecondsafterStart8-75VisicorderTraceShovingIntermittentOperationoftheQuencher-Test36.1SysemPressure-6.2-1.0barPoolWaterTemp-26~C-300C8-76VisicorderTraceShovingExcerptfromIntermittentOperationofQuencherTest36.1-280SecondsafterStart8-778-78VisicorderTraceShovingSingleEventOutofIntermittentCondensationTest36.1TypicalVisicorderTraceofStationaryOperationofQuencherTest37.2-13SecondsafterStart8-79TypicalVisicorderTraceofStationaryOperationofQuencherTest39.1-10SecondsafterStart8-80VisicorderTraceShovingIntermittentOperationofQuencher-Test40.1SystemPressure-2.5barPoolWaterTemp.-89~C-91~C8-81DynamicBottomPressuresduringtheBlowdownAlongtheUpperandLoverBoundaryoftheOperationField'-82DynamicWallPressuresDuringtheBlowdownAlongtheUpperandLoverBoundaryoftheOperationFieldREV.1,3/798-14 8-83OccurrenceFrequencyDistributionPositiveandNegativeDynamicAmplitudesfortheCondensationTestsPoolTemp.22~C-300C8-84OccurrenceFrequencyDistributionPositiveandNegativeDynamicAmplitudesfortheCondensationTestsPoolTemp.59~C-91>C8-85OccurrenceFrequencyDistributionPositiveandNegativePressureAmplitudeforCondensationTestsPoolTemp.22C-30C8-86OccurrenceFrequencyDistributionPositiveandNegativePressureAmplitudeforCondensationTestsPoolTemp59C-91C8-878-888-89NeanValuesoftheBottomDynamicPressuresDuringtheBlowdownsAlongtheUpperandLowerBoundaryoftheOperationFieldMeanValuesoftheWallDynamicPressuresDuringtheBlowdownsAlongtheUpperandLowerBoundaryoftheOperationFieldWaterTemperatureTimeHistoriesOnPoolWallCondensationTest33.28-90WaterTemperatureTimeHistoriesOnPoolWallCondensationTest35.18-91MaterTemperatureTimeHistoriesOnPoolMallCondensationTest37.28-92RaterTemperatureTimeCondensationTest39.1HistoriesOnPoolMall8-93WaterTemperatureTimeCondensationTest33.2HistoryOnQuencherArm18-94RaterTemperatureTimeHistoryonQuencherArm1CondensationTest35.18-95MaterTemperatureTimeHistoryonQuencherArm1CondensationTest3728-96MaterTemperatureTimeHistoryon.QuencherArm1CondensationTest-39.18-978-98CalibrationofSensorsandRegistrationInstrumentsIntervalsforCalibrationChecksandAdjustmentsofInstrumentation8-99CalibrationSystemREV.1>>3/79 8-100CalibrationResultsDeviationsfromNominalValue-P5.1-P5-108-1018-1028-1038-104MaterLevelinDischargeLineTest15.1MaterLevelinDischargeLineTest20.1WaterLevelinDischargeLineTest32EffectsofFreeSurfaceandRigidTankWallsonDynamicFluidPressure8-1058-1068-1078-1088-109MethodofImagesSSESSmallestUnitCellandtheKarlsteinTestTankPressureProfilesforDifferentBubbleLocationsPressureProfileforaOneandFourBubbleArrangementComparisonofMeasuredandCalculatedNormalizedPressureProfiles8-110ComparisonofPressureProfilesCalculatedfortheKarlsteinTestTankandtheSSESSuppressionPool8-111ComparisonofCalculatedandSpecifiedPressureProfiles8-1128-113TankArrangementShowingInstrumentationandExplosiveChargeLocationsforMeasuringTankReponseConfigurationofExplosiveContainerUsedtoGenerateUnderwaterPressureImpulse8-1148-1158-1168-1178-1188-1198-1208-1218-122TypicalTankResponseDuetoPressureImpulseFrequencyAnalysisofGageMA2FrequencyAnalysisofGageMA7FrequencyAnalysisofGageMA8FrequencyAnalysisofGageP5.10DisplacementCorrelationsfor13Hz-EigenmodeDisplacementforthe13HzEigenmodeTestTankArrangementforShakedownTestsTankDisplacementsand.PressureTraceDuringShakedownTest08.18-123FrequencyAnalysisofGageMA2ShakedownTest08.1REVli3/798-16 8-1248-1258-1268-127FrequencyAnalysisofGageWA7ShakedownTest08.1FrequencyAnalysisofGageWA8ShakedownTest081FrequencyAnalysisofP510ShakedownTest08.1AirNassPlowusedforKOVlBlComputerCodeCalculations8-128UnitWallDisplacementof13HzNodeUsedinKOVlB1ComputerCodeCalculations8-129BoundryPressureDistributionCalculatedforUnitDisplacementof13HzNode8-1308-1318-1328-133WallPresureCalculationwithKOVlB1ComputerCodeEffectsofFSZonBubblePrequencyTypicalPressureTraceinSRVDischargeLineTest4.1.4TypicalPressureTraceinSRVDischargeLineTest20.Rl78-134PressureinSteamLinebeforeSRVVersusPressureinBufferTankatValueOpening8-1358-136PressureinDischargeLineVersusReactorPressureatVentClearing-P4.1LongLineTestsPressureinDischargeLineVersusReactorPressureatVentClearing-P44LongLineTests8-137PressureinDischargeLineVersusReactorPressureatVentClearing-P4.1ShortLineTests8-138PressureinDischargeLineVersusReactorPressureatVentClearing-P4.4ShortLineTests8-1398-140VentClearingPressureVersusValveOpeningTimeSteadyStatePressureVersusReactorPressure-P4.1LongLineTests8-141SteadyStatePressureVersusReactorPressure-P4.4LongLineTests8-l42SteadyStatePressureVersusReactorPressure-P4.1ShortLineTests8-143SteadyStatePressureVersusReactorPressure-P4.4ShortLineTests8-144SteadyStatePressuresatDifferentLocationsAlongtheDischargeLineExtrapolatedto88BarReactorPressureREV1,3/798-17 8-1458-146.8-147TypicalTraceforVerticalLoadLongLineTestsVerticalLoadVersusClearingPressureLongLineTestsVerticalLoadVersusVentClearingPressureShortLine-Tests8-1488-149TypicalTraceforTorqueonBottomSupportLongLineTestBottomSupportTorqueVersusVentClearingPressureLongLineTests8-150BottomSupportTorqueVersusVentClearingPressureShortl.ineTests8-151TypicalTraceforBendingMomentsonQuencherArmsLongLineTests8-152ResultantQuencherArmBendingMomentVersusVentClearingPressureShortLineTests8-153FrequencyDistributionofMaximumResultant'BendingMomentonQuencherArmsandatWeldSeam8-154ResultantBottomSupportBendingMomentVersusVentClearingPressureShortLineTests8-155FrequencyDistributionofMaximumResultantBendingMomentonBottomSupportSG4.5-468-156FrequencyDistributionofMaximumResultantBendingMomentsOnQuencherArmsgtStrainGagesIntermittentCondensation8-157FrequencyDistributionofMaximumResultantBendingMomentatWeldSeamonQuencherArm-IntermittentCondensation8-158FrequencyDistributionofMaximumResultantBendingMomentsatBottomSupport-IntermittentCondensation-0.5mbelowQuencherCenter8-1598-160PowerSpectralDensitiesTestll.1-P5.5PowerSpectralDensitiesTestll1-P528-161PowerSpectralDensitie'sTest4.1.6-P5.58-1628-1638-164PowerSpectralDensitiesTest4.1.6-P5.2PowerSpectralDensitiesTest20.R1.10-P55PowerSpectralDensitiesTest20-R1.10-P5.2REV13/798-18 8-165MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValues-Overpressures8-166MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValuesConsidering.LocalEffects8-167MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValues-UnderPressures8-168KarlsteinTests-ComparisonofMeasuredandCalculatedBubbleFrequency-0$Humidity8-169KarlsteinTests-ComparisonofMeasuredandCalculatedBubbleFrequency-100%Humidity8-170GKMTests-ComparisonofMeasuredandCalculatedBubbleFrequency8-171GKMTests-ComparisonofMeasuredandCalculatedBubbleFrequency-Overpressure8-172KKBIn-PlantTests-ComparisonofMeasuredandCalculatedBubbleFrequency8-1738-174SSESCalculatedBubbleFrequenciesMultipliersforConversionofBubbleFrequenciestheKarlsteinTesttoSSES8-175OverpressureMultiplierforConversionofBubbleFrequencies8-176NormalizedAmplitudeSpectrumVersusBubbleFrequency-KarlsteinTests8-177KarlsteinModelTests-InfluenceofMaterSurfaceonPressureAmplitude8-178GKMTests-InfluenceofOverpressureonBubblePressure8-1798-1808-1818-1828-1838-1848-185PSDofKarlsteinTests-11.1and12.1-P5.10PSDofKarlsteinTests4.1.1and4.1.6-PS.10PSDofKarlsteinTests21.1and21.2-P5.10PSDofTest20.R110-P5.4PSD'sofTest11.1-P5.2,5.4and5.10PSDComparison-Test20Rl1andDesignSpecificationPDSComparisonTest4.1.1andDesignSpecificationREVlg3/798-19 8-1868-1878-1888-1898-1908-1918-1928-1938-194PSDComparisonTest20.R1.10andDesignSpecificationPSDComparisonTest211andDesignSpecificationPSDComparisonTests21.2and25.R2andDesignSpecificationPSDComparisonTest0.1.6andDesignSpecificationIBADrywellandWetvellPressureHistoryPSDComparisonTest11.1andDesignSpecificationTypicalCrossSectionofSSESSuppressionPoolRevisedQuencherArrangementVelocityofRotatingPoolforOneActuatingValveinOuterRow8-1958-1968-1978-1988-1998-2008-2018-202WaterMotionoftheAcceleratedPoolTestStandforMeasuringThrustMeasuredTemperatureDistributionintheKKBSuppressionPoolResultantBendingMomentonDummyVentVersusReactorPressureResultantBendingMomentonDummyVentVersusClearingPressure0~ResultantBendingMomentonDummyVentVersusPressureAmplitudeatP5.7'JSpecifiedPressureDistributiononDummyVentTypicalVisicorderTraceforBendingMomentonDummyVentREVli3/798-20 PROPRIETARYNumberSECTION8TABLESTitle818.28.3TypicalOperatingInstrumentationTypicalVentClearingTestInstrumentationTypicalCondensationTestInstrumentation8.4ParametersatTestStart-LongPipeVentClearingTestSeries85ParametersatTestStart-ShortPipeVentClearingTestSeries8.6ParametersatTestStart-CondensationTestSeries8.7BehavioroftheSRVandSystemPressures-LongPipeVentClearingTestSeries8.8BehavioroftheSRVandSystemPressures-ShortPipeVentClearingTestSeries8.9PeakDynamicPressuresonthePoolBoundaryDuringVentClearing-LongPipeVentClearingTests8.10PeakDynamicPressuresonthePoolBoundaryDuringVentClearing-ShortPipeVentClearingTests8.11Maximum,Strains,MomentsandVerticalLoadsontheQuencherArmsandSupportDuringVentClearing-LongPipeTests8.12MaximumStrains,MomentsandVerticalLoadontheQuencherArmsandSupportDuringVentClearing-ShortPipeTests8.13SystemPressuresandPoolWaterTemperaturesoftheCondensationTests810PeakDynamicPressuresAmplitudesDuringtheDifferentCondensationPhases8.15StatisticalCharacteristicsoftheBottomDynamicPressures(P5.2)8.16StatisticalCharacteristicsoftheWallDynamicPressures(P5.10)8.178.18REV.1,3/79RepetitionTests-ComparisonofRecordedValvesRepetitionTests-MeanValuesandDeviations8-21 PROPRIETARY80SSE~SUENCHERVERIFICATIONTEST81INTRODUCTION8.l.1Pu~roseoftheTestsTheoptimizedquencherdesignforSSESandtheloadspecificationonthewettedboundariesofthesuppressionpool,onthesubmer'gedstructuresandonthepressurereliefsystem,arebasedonparametricmodelteststudiesandfullscaleinplanttestresultsfromasimilarquencherdesign.TheloadspecificationsfortheSSESquencheraredescribedindetailinSection4.1.Inordertoverifytheseloadspecificationsandfurtherverifythequencher'ssteamcondensingcharacteristics,fullscalesinglecelltestswereconductedattheKraftwerkUnionlaboratoriesinKarlstein,WestGermany.8.1.2TestConceptTheconceptsusedtodesignandperformthetestswere:1)Useofaconservativelydefinedsinglecell2)Theclosesimulationofthemainsafetyreliefvalvesystemparameters81.2.1UnitCellApproach81.21.1SingleCellTheoryForagasbubbleoscillationinafreewaterspace,thewatermasscoupledtothebubbleisalternatelyacceleratedanddeceleratedDuringthisprocesstheoverpressureandunderpressureamplitudesdecreasewithincreasingdistancefromthebubble.Whenasolidwallisplacedneartheoscillatingbubble,thewateraccelerationisrestrictedinthedirectionofthewallandthedecreaseinpressureamplitudeinthedirectionofthewallisless.Thiseffectcanbeexpressedmathematicallybyreplacingthe.bubblebyapotentialsourceandaccountingforthewallbythemethodofimages.Theeffectsoftherealsourceandtheimagesourceareaddedforeachpointoftheflowfield.Forthecaseinwhichabubbleisenclosedinanarrowwaterspace,closelysurroundedbysolidwallsandasolidbottomwithafreewatersurfaceatthetop,thewaterspacebelowthebubbleisforallpracticalpurposesunmoved.Onlythewatervolumeabovethebubbleisfreetooscillate..Consequently,thepressuregradientinthelowerwaterspaceisnearlyzero,whilethepressureamplitudeabovethebubbledecreaseswithincreasingproximitytothewatersurface.Thepressureamplitudesarezeroatthewatersurfaceandthemethodofimagesapplies.REV.1,3/798-22 PROPRIETARYAnalytically,thecaseinwhichaplanarfieldofuniformstrengthbubblesarealloscillatinginphaseisthesameasthecaseinwhichsolidwallsexistbetweeneachoftheindividualbubbles.ThesinglecelltestconfigurationusedatKarlsteinsimulatesthisextremelyconservativecaseofparallelbubblesoscillatinginphasewiththesamesourcestrength.Adescriptionoftheequivalenceofthesinglecellconfigurations,usingthemethod-ofimages,iscontainedinFigures8.1and8.2.ForamoredetailedevaluationoftheKarlsteintesttanksinglecell,seeSection8.5.1.81.21.2A~licationofSingleCell~AroachThesubmergenceofthequencherinthetesttankisequaltothehighestvalueintheplant.Astothewatercross-sectionareathesinglecelltheorydescribedaboveisusedFigure8.3showsageometricalpartitionofwaterspace.Thewatercross-sectionareasrelatedtothedifferentquenchersarelistedbelow:QuencherAQuencherBQuencherCQuencherDQuencherEQuencherFQuencherGQuencherHQuencherJQuencherKQuencherLQuencherQuencherNQuencherPQuencherRQuencherSAverageWaterSurface3147mz(33862ftz)3147mz31.47mz3147mz31475231~47AI231.47mz31.47mz31.47mz3147mz3147mz3147mz31.47mz31.47mz31.47mz31.47mzRelatedWaterSurface21.4mz(230-26ftz)214mz31.3mz{336.79ftz)42mz(45192ftz)31~3mz31'mz42mz31m3mz31~3mz42mz31~3mz31a3mz214mz21.4mz31~3mz42mzREV1,3/798-23 PROPRIETARYThesmallestwatersurface(approximately21.4m~)issimulatedinthetests.Therefore,thedynamicpressureamplitudesatthewallsandthebottomaremeasuredunderconservativeboundaryconditions.8.1.22SimulationofSSESParametersThefo'llowingsectionprovidesadescriptionofthoseparametersthatweresimulatedintheKarlsteintestfacilityTheseparametersaretypicalofmostMKIIplants.FormoredetailonthetestfacilityseeSection8.28l.22.1Prima~rSgsternPressureThereactoroperatingpressureforSSESisapproximately1000psig(69bar)whilethehighestpressuresetpointforanySSESSafetyReliefvalveis1205psig(83bar),whichisclosetothehighestprimarypressurethatcanbesimualtedintheKarlsteintestfacility(82bar).Thisallowedthetestsimulationtoverycloselymatchtherangeofinitialprimarysystempressuresthat.canbeexpectedintheoperatingplant.8.1222Safet~ReliefVal'vegSRVJInordertomatchthecharacteristicsoftheSafetyReliefValve,anoriginalCrosbySRV,shippeddirectlyfromtheplantsite,wasinstalledintheteststandandusedinalltests.81223DischargeJ.ineInordertocovertherangeofdischarge'linelengthsandthereforeairvolumesthatexistinSSES,twoventclearingtestserieswererun;onewithadischargelinethatsimulatesthelongestSSESdischargeline{48m)andonethatsimulatestheshortestSSESdischargeline(35m).Inaddition,thenumberofbendsineachline,thei'nnerdiameterofthemainpartofthe'ine(303.9mm),andtheinnerdiameterofthelastverticalruntothequencher(2889mm)arecloselysimulatedtothatwhichexistsintheSSESplant.(schedule40pipeandschedule80pipe,respectively).Inadditiona24ft.submergence,correspondingtothehighestwaterlevelinthesuppressionpool,wasusedforalltests8.12.24VacuumBreakersInordertocloselysimulatetheeffectsofvacuumbreakeroperationonthetests,twosix-inchdiameterCrosbyvacuumbreakerswereshippedtoGermanyandinstalledintheteststandatthesamerelativelocationasplannedfortheSSFSplant.REVli3/798-24 PROPRIETARY81.22.5OuencherAfullsizeprototypeofthequencherinstalledintheSSESplantwasinstalledinthetestfacilityandusedforalltests.Figure8.13showsthequencherwithinstrumentationforventclearingtestswhilefigure8.14showsthequencherwithinstrumentationforthecondensationtests.8-2TESTFACILITYANDINSTRUHENTATION8.21TestFacility8.2.1.1MechanicalSet-upThetestconfigurationasconstructedistypicallyillustrateddiagrammaticallyinFigure8.4.Theteststandconfigurationcanbedividedinto:thesteamboiler,thesteamaccumulator,thesteamlinebeforetheSRVandthebuffertank,theSRV,thedischargelinebetweentheSRVandthewaterpoolwiththequencheraspipetermination,andthelargetankaswaterpool.8.2111SteamboilerThesteamboilerisanoil-fired,once-through,forced-flowboilerwithanoutputofapproximately20HWatamaximumsteampressureof170bar(2499psig)andamaximumsteamtemperatureof520~C(968~F).The.boilerisdesignedforaclosedoperatingmodeinnormaloperation.Afractionoftheboiler'soutputisrecoveredfromthecondensateviathehigh-pressurecooler.Whenthereisanopenloop{ie.,lostcondensate),theoutputisreduced.Thesteamflowavailableinthismodeisapproximately8to9kg/s(17.6to19.8ibm/s).Thelostcondensateresultsinatimelimitationoncontinuousoutput.Thefeedwatersupplyoftheboilerisabout20m3(705ft~).Oncethatamountisusedup,furthersteamsupplyascontinuousoutputispossibleonlyuptotheoutputofthefeedwaterconditioningsystem.Thatamountsto5m~/h(176ft3/h).Forlongertestperiodsitisnecessarytointerruptoperationfor4hoursinordertorefillthefeedwaterstoragetank.82.1.1.2SteamAccumulatorAsdescribedin8.2.1.1.1theamountsofsteamsuppliedcontinuouslybytheboileraretoosmalltotestanSRV.REV.1,3/798-25 PROPRIETARYToprovideawaytotestvalvesatflowratesofuptoapproximately22kg/s(484ibm/s),avalvetestfacilitywasbuiltusingtheboilerplantandapressurevesselconnectedtoit.Thisvesselischargedwithasteam/watermixturebytheboilerandisusedasasteamaccumulator.Fromthissteamaccumulator,highersteamflowratescanbedeliveredforashortperiodoftimeThedimensionsofthepressurevesselare1.5mdiameterand12mhigh,whichresultsinanaccumulatorvolumeofapproximately22m3.Adaptedtotherequiredsteamoutput,theaccumulatorisfilledwithsaturatedwaterandsaturatedsteamatthespecifiedratio.Thesteamisdrawndownwardthroughastandpipe.Thehighsteam,flowtobeextractedtransientlyfromtheaccumulatorresultsinarapiddecreaseofpressureandtemperature.Forstrengthreasons,thetemperaturedifferencebetweentheinsideandoutsideoftheaccumulatorvesselmustnotexceedacertainvalue.Thislimitsthemaximumpressuredropandthustheavailabletesttime.82.1.1.3SteamLineandBufferTankTheconnectionbetweenthesteamaccumulatorandthevalveteststandconsistsofanND250pipeline.ThislinecontainsisolatingdevicesforemergencyisolationandameasurementsectionconstructedasaVenturinozzle.Theexistingeguipmeatprovideforadirecthorizontalconnectionofthevalvebeingstudied.ThiscorrespondstothedesignoftheSRVsusedinGermanBMRplantsandtotheirarrangementattheendofataplinecomingfromthemainsteamline.ThesteamsupplylinewasrebuilttomatchthedesignfeaturesofSSES.ThepreviouslydescribedpipelinenowendsinaT-piece.XnordertosimulatetheSSESmainsteamlineandtokeepthesteamsupplyflowtothevalveasuniformaspossible,abuffertankhavingavolumeof5.2m3wasconnectedtothesecondhorizontaloutletoftheT-piece.Theverticaloutletoftheabove-describedT-pieceleadstothevalve.8.2.114Safet~ReliefValveQSRVQTheSRVusedinthetestsistheactualversionbeingusedforSSES.Thesevalvesarearrangedvertically,haveasteaminletfrombelowandanoutlettotheside.Asdescribedin8.211.3,thesteamsupplylinewasrebuiltinsuchawaythatthesamearrangementwaspossibleintheteststand.ThevalvewasmountedontheT-piece,usingthesameconnectiondimensionsasintheactualplants.REV1,3/798-26 PROPRIETARYOperationofthevalveduringthetestsrequirestheconnectionofpowersupplylines,controllinesandmeasurementlines.Theexistingequipmentatthevalvetestfacilitywasusedtosatisfymostofthoserequirements.Somemodificationsbecamenecessaryinordertoadapttotheconstructionofthevalve.TheSRVsinGermanBMRplantsareoperatedbyanelectricallyactuatedpilotvalvewithitsownoperatingmedium.Incontrast,theSS'>>Svalveusedinthetestwasopenedpneumatically.Accordingly,thecompressed-airconnectionwasrebuiltsothattheopeningconditionsintheactualplantcouldbesimulatedintheteststand.8.2~11.5DischargeLineandQuencherTheSRVdescribed.in8.2.1.1.4dischargesontheexhaust-steamsideintoapipewhichrepresentstheSRVdischargeline.ThelengthoftheSRVdischargelineandthenumberofbendsaredifferentforthe16SRV'sforSSES.Twolinelengthswereusedforthetests,correspondingtothelongestandshortestlengthsoftheSRVdischargelinesintheplant.IsometricdrawingsofthetwodischargelinesareshowninFigure8.5(longline)andFigure8.6(shortline).Pipesupportsandvibrationdampersweremountedattherequiredplaces.Theseplaceswerenotidenticaltothecorrespondingonesintheplant,becausethemountingsituationsandespeciallytheconcreteconstructionoftheplantcannotbe-simulateddirectlyinthetestfacility.Topreventthebuildupofalargeunderpressureinthepipe,twoactualvacuumbreakerswereinstalledinaverticalpartofthepipeline,asintheplant.ThequencherformstheterminationoftheSRVdischargeline(seeFigure8ll).Thesteamisconductedintothewaterthroughalargenumberofholeshavingadiameterof10mm.ThedesignofthequencherisdescribedindetailinSection4.1.AbottomsupportisprovidedtoholdthequencherinplaceinthetesttankItconnectsthequencherrigidlytothebottomofthetankandisconstructedinsuchawayastomakeitpossibletomeasuretheloadsexertedonthequencherduetoventclearingprocessesandsteamcondensation.Theslidingjointprovidedbetweenthequencherandthedischargelineintheplantissimulatedintheteststandhgdraulicallgbyacorrespondingannulargap8211.6TestTankForSSES,theexhauststeamfromthereliefvalvesisconductedintothesuppressionpoolandiscondensedthere.Xnthetestfacility,asectionofthatpoolissimulatedbyastiffenedREV-lg3/798-27 PROPRIETARYsteeltank(seeFigures8.7,88,89).Intheplant,thesuppressionpoolcanbesubdi'videdconceptuallyintosuhspaces,eachofwhichisassociatedwithasteamsupplyline(seeFigure8.3).Inordertoadapttheconditionsinthetesttanktothedimensionsofthesmallestgeometricalsinglesell,concreteshapedblockswereinsertedintothetesttank.TheconcreteshapedblocksareclearlyillustratedinFigure8.7.Theexposedcross-sectionalareaofthewaterspaceis7.2mx3.15m=22.7m~.Itcorrespondsconservativelytothesmallestindividualcellintheplant.Illuminatingdevicesandviewingportsmadepossiblethedirectobservationandalsophotographicrecordingoftheunderwaterprocesses.8.2.2InstrumentationInstrumentationisprovidedforcontrollingthetestprocedure,determiningtheprescribedmeasurementquantities,andrecordingthem.82.21GeneralDescriptionTheinstrumentationusedintheKarlsteintestfacilityconsistsofoperatinginstrumentationandtestinstrumentation.Operatinginstrumentationassuresthecontrolofthetestfacilityanditsenvironmentcorrelation.ThetestinstrumentationrecordstheloaddatawhichisusedtoverifytheconservatisminthedesignloadsasspecifiedfortheSSESinsection4.1ofthisDesignAssessmentReport.DetailsontheoperatinginstrumentationaregiveninSection8.2.2.3.AdetaileddescriptionofthetestinstrumentationcanbefoundinSection8.2.248.2.22InstrumentationIdentificationForidentification,themeasuringsensorsaredesignatedaccordingtoasystemoflettersandfigures.Thefirstoneortwocharactersareletterswhichidentifythetypeofinstrument:PTFLDGSGILPPressureTransducerTemperatureSensor(Thermocouple)FlowRateMeasurementsRaterLevelMeasurementsDisplacementGageStrainGageElectricalImpulseSignalLevelProbeREV.l,3/798-28 PROPRIETARYTheselettersarefollowedbyanumberwhichcharacterizesthelocationwithinthetestfacilitywheretheinstrumentissituated.Thefacilitywasdividedintosectionsasfollows:Section1containsthesteamsupply,includingtheaccumulator{onlytransducersoftheteststandinstrumentationsystemarecontainedinthissection).Section2containsthesteamlineuptothesafetyreliefvalveandincludesthebuffertank.Section3containsthesafetyreliefvalve.Section4containsthedischargelineandquencher.Section5containsthetesttank.Thesensordesignationiscompletedbyaddingadecimalpointandasequentialnumber.Forexample,"P5.6"means:thenumber6pressuretransducerinthetesttank.Additionalabbreviationsusedareasfollows:DPSCTCDCACFAHT-SGSRVPGRTDDataProcessingSystemCoatedThermocoupleDirectCurrentAmplifierCarrierFrequencyAmplifierHighTemperatureStrainGageSafetyReliefValvePressureGageResistorTemperatureDetector82.2.3Operating1nstrumentationTheoperatinginstrumentationisprovidedformeasurementofparametersinrelationtothesteamaccumulator,thesteamlinesandtheSRV'sAtotalof30sensorscanberecordedbyaprocesscomputerwhichispartoftheoperatinginstrumentationsystem.ThedataarestoredonamagneticdiskandcanbeprintedoutTherecordingfrequencyoftheprocesscomputerwasadaptedtoalignwiththeinstrumentationchanriels,coveringarangefrom0.5Hz,forthosesensorswhereonlysmalltransientsaretobeexpected,uptoabout200Hzforthesensorswherehigherfrequencysignalsareexpected(e.g.forpipevibrations)Theoperatinginstrumentationcomprisesthemeasuringdevicesusedtomonitorandcontrolthesystemandalsothedataacquisitiondevicesneededforthatpurpose.TypicalmeasuringlocationsforthetestsareillustratedinFigure8.4andlistedinTable8.1.BEV1,3/798-29 PROPRIETARYAccordingtothetypeofacquisitionanddisplay,themeasurementsensorscanbeclassifiedintotwogroups="DisplayonControlConsole"and"AcquisitionbyComputer".82.2.3.1Disp~la'nControlConsoleToenabletheoperatingpersonneltocontrolthetestequipment,anumberofquantitieswhichcharacterizetheoperatingconditionofthesystemaredisplayedcontinuously.Inparticular,theyare:Waterlevelin:Steamaccumulator,steamline,buffertank,dischargeline,testtankPressureinSteamaccumulator,buffertank,controlline,dischargelineTemperaturein:Steamaccumulator,buffertank,dischargeline,testtanka822.3.2A~cuisition~bComputerMostofthedatasensorscomprisingtheoperatinginstrumentationareinterrogatedbyacomputeratprescribedtimeintervalsbefore,duringandafterthete.t.Thevaluesarestoredonadisk.Thedataareprintedoutatprogrammedintervals.Ataninterrogaticnfrequencyof200Hz,thecapacityofthestoragedeviceissufficientforarecordingtimeof2minutes.Thefollowingmeasurementvaluesareinterrogated:WaterlevelSteamaccumulator,buffertankdischargeline,testtankPressureSteamaccumulator,buffertank,steamline,controlline,dischargelineTemperatureBeforeSRV,afterSRV,surfaceofSRV,dischargeline,testtankVibrationsValvetravelSwitchingtimeSteamlinebeforeSRV,dischargelineSRV,vacuumbreakersElectricalenergizationofSRVREV1g3/798-30 PROPRIETARY822.4Test,InstrumentationMesurementvaluesusedtoverifythetesttasksaredeterminedbythetestinstrumentation.Itisnecessarytoincludehereafewtypicalmeasuringpointsthatarealreadyusedformonitoringpurposesintheoperatinginstrumentationonthepipes.andSRV.Sincemostoftheseprocessesareofahigh-frequencynature,thedataisacquiredinanalogformbymeansofcarrier-frequencymeasuringamplifiersanddcamplifiersonanalogmagnetictape,andtoalargeextentalsoonvisicorders.Thevisicordertracesallowaninitialreviewandapre-evaluationofthetestdata.82.24.1MeasuringPointsMeasurementsaremadeofthepressureonthesteamlinebeforetheSRV;valveactuationandvalvetravel;pressurevariationinthedischargelineatfourpointsbetweentheSRVandquencher;temperatureinthedischargelineatthreepointsbetweentheSRVandquencher;waterlevelinthedischargelinebeforethequencherinletatfourpositionsforthelonglineandfivepositionsfortheshortline;bending,axialandtorsionalstrainsonthebottomsupport;bendingstrainsonthequencher;bendingstrainonadummyventpipe;temperaturedistributioninthetesttank;temperaturedistributionatthequencherforthecondensationtest;wallpressuresandbottompressuresinthetesttank.TypicalmeasurementpointsfortheventclearingtestsareillustratedinFigures8.7,8.8,8.9andlistedinTable8.2.TypicalmeasurementpointsforthecondensationtestsareillustratedinFigures8.10,8.11,8.12andlistedinTable8.3.8.2.24.2Set~uofMeasuringInstrumentsAllinstrumentationischannelledtoonecentralstationsituatedinthecontrolroomofthelaboratory.Eachinstrumentationchannelconsistsoftheindividualsensor,connectingcable,amplifier(carrierfrequencyamplifierordirectcurrentamplifier),attenuator;andarerecordedonmagnetictapesandvisicorders,mostchannelsbeinginparallelonbothsystems.Threemagnetictaperecordersandthreevisicorderswereusedinthecontrolroom.Eachunitallowstherecordingof12channelsand,inaddition,atimereferencesignalandaphysicalcorrelationtrace.REV1,3/798-31 PROPRIETARYThesensorsareconnectedbyshieldedcabletotheamplifiersvhicharelocatedneartherecordersinthecontrolroom.Forthestraingages,displacementgagesandpressuretransducers,carrierfrequencyamplifiersvereusedwhichallowafrequencyresolutionofupto1KHz.Fortemperaturemeasurements,directcurrentamplifiers(10Hz)vereusedtogethervitha10Hzlovpassfilter.82.25Visual-RecordingThreehigh-speedcamerasvereusedtofilmtheprocessesinthepoolduringtheblowdovnthroughthequencher.KMUusesa"HYCAM120m~'orthatpurpose.TvoLOCAMcameras(model51-0003)werebeingmadeavailablebytheStandfordResearchInstitute(SRI)Thepositioningofthecameraswasasfollovs:HYCAMcamerainfrontofonebull'seyeatquencherheight;LOCAMcamera1infrontofonebull'seyeatatankheightofapproximately4m;LOCAMcamera2ontheserviceplatformabovethetankataheightofapproximately9m.AcorrelationbetweenthemovingpicturesandthedatarecordingsontheVisicorderandmagnetictapevasaccomplishedbymeansofatimingmarkonthefi'lms.83TESTPARAMETERSANDMATRIX8.31VentClearingTestsThetestmatrixfortheventclearingtestsispresentedinFigure8.15.Thisfigureshowsthetestnumberandparameterconditionsusedforeachtest..Thenumberofbasictestswas25.These25testsweresplitinto5groupsoftestswherebyeachgroupcoveredasetoftestparameters.Testsnumbered26to32wereadditionaltestsvhichwerenotrequiredtoverifythequencherdesignbutwhichcouldproveusefulinevaluatingtheperformanceofthesafetyreliefsystem.Testsnumber27,28,30and31weretoinvestigateshorterthannormalSRVopeningtimes,but,asvalveopeningtimesverefoundtobequitefast,thesetestswerenotaddedtotherequiredtests.Testsnumber26and32,withonelockedvacuumbreaker,wereincludedintothetestmatrix.Theresultsshovedtheeffectofthelockedvacuumbreakertobeminimalsotestnumber29wasnotadded.REV1,3/798-32 PROPRIETARYTheallocationofeachtestgroupwithintheoperationrangeofthesafetyreliefsystemisshowninFigures8.16to8.21bytestpoints.BaseparametersinGroup1(Figure816)arelongdischargeline'length,normaldischargelineairtemperature,normalinitialwaterlevelinsidethedischargelineandnormalvalveopeningtime.EachofthefollowinggroupsvaryoneormoreoftheseGroup1baseparameters;Group2(Figure817)usesalowinitialwaterlevelinsidetheSRVpipe;Group3(Figure8.18)usesahighdischargelinetemperature;Group4(Figure8.19)usesashortdischargelinelengthandGroup5(Figure8.20)usesashortdischargelinelengthandahighdischargelinetemperature.Eachofthebasic25testswascomprisedoftwoormorevalveactuationswherebyonlythefirstactuationismadeatt,hespecifiedconditionsofthedischargeline(so-calledcleancondition).Anyotheractuationwasmadeattheprevailingdischargelinetemperatureandwaterlevel(so-calledRealCondition).Inthecaseofonlytwoactuationsatatestpointthetimeintervalbetweentheactuationswasapproximately10minutes.Inthecaseofmultipleactuationsatatestpointthetimeintervalsbetweenactuationswerevariedasfollows:Fortestpoints4,5,14,15thetimebetweensuccessiveactuationswasl.5/5/15/30/60/120seconds,accountingforsevenvalveactuations.Fortestpoints19and20thetimebetweensuccessiveactuationswas15/5/15/30/60/120/5/15/600seconds,accountingfortenvalveactuations.ForventclearingtestswithonlytwoSRVactuations,thehold-opentimefortheSRVwas2secondswhileforthemultiplevalueactuationteststhehold-opentimewas1.5seconds.'Fivetestpointswererepeated,theseweretestpoints4,15,19'0and25.RepeattestsatadesignatedtestpointareindicatedwithaletterRinthetestnumberi.e.Testnumber20.Rl.listhefirstvalueactuationoftherepeattestattestpoint20.11.AcompilationofactualparametersatthestartofeachtestistabulatedinTable8.4forthelongpipetestseriesandTable85fortheshortpipetestseries.8.32CondensationTestsInordertofurtherverifythesteamcondensationcapabilitiesofthequencherdeviceandprovidespecificinformationregardingitssteamcondensationcapabilitiesforthesafetyreliefsystemlREV.1,3/798-33 PROPHIETARYoperationrangeaseriesofeightextendedblowdowntestswereperformed.Thesetestsaredesignatedastestnumbers33to40.Eachtestwasperformedwiththeshortdischargelineconfigurationasdescribedinsection8.2.1.1.5andwithaninitialdischargelinetemperatureofapproximately90~C.ThelocationoftheinitialsystemconditionsforeachtestpointisplottedonthesafetyreliefsystemoperationrangeinFigure822InordertoinitiateeachtesttheSRVwasactuatedaswasdoneintheventclearingtests.Thevalvethenremainedopen-untilthesystempressurereachedthepredesignatedvalueforthattest.Atthistimethevalvewasclosedandthetestwascompleted.Thetotalallowablepressuredropintheaccumulatortankforeachinitialsystempressuredictatedthedurationofeachbiowdown.AcompilationofactualparametersatthestartofeachtestpointinthecondensationtestsmatrixistabulatedinTable8.6.84TESTRESULTSThissectionprovidesacompilationofthetestresultsfortheventclearingandsteamcondensationtestsconductedattheKraftwerkUnionlaboratoriesinKarlstein,WestGermanyinordertoverifytheloadspecificationandsteamcondensingcharacteristicsofthequencherdesignfortheSusquehannaSteamElectricStation.Includedinthissectionisinformationabouttheboundaryconditionsatthebeginningofeachtest,the'esultsofthebehavioroftheSRV,primarysystempressures,dynamicpressureloadsonthepoolboundariesandtheirprimaryfreguencyandtheloadsonthequencherandbottomsupportThisinformationisprovidedintheformoftables,figuresandactualvisicorderrecordings.841VentCleari~nTestResultsNineteentestswithatotalof67ventclearingprocesseswereperformedwiththelongdischargelineintheperiodfromMay8,1978toJune7,1978and13testswithatotalof58ventclearingprocesseswereperformedwiththeshortdischargelineintheperiodfromJune27,1978toJuly7,1978.841.1TestParametersThemostimportantoftheparametersbeinginvestigatedwasdescribedinSection8.3.AdetailedlistoftestparametersforeachvalveactuationisgivenforthelongdischargelinetestsinTable8.4andfortheshortdischargelinetestsinTable8.5.ThisincludesREV.1,3/798-34 PROPRIETARYtypeoftestlengthofdischargelineaccumulatorpressurewatertemperatureinthetesttankwaterlevelindischargelineairtemperatureindischargelineTheaccumulatorpressureP1.1AandthebuffertankpressureP2.6Aarethedeterminativevaluesforthesystempressureatthestartofeachtest.Thevalueswerereadbycomputerjustpriortothestartofthetest.Inadditionthesepressureswerestoredcontinuouslyonmagnetictape.IfalongperiodpassedbetweenthelastcomputerreadingandtheactualteststartthentheinitialvaluesfortheaccumulatorpressureweretakenfromthecorrespondingcomputerplotsTheinitialaccumulatorpressureswerealsoreadfromthoseplotsforthemultiplevalveactuationtests.Foraccumulatorpressuresbelow30bar(435psi),measuringpointP2.5wasusedtodeterminethesystempressure,sincemeasuringpointsPl.1AandP2.6Awereoutsidethemeasuringrange.Thewatertemperatureatthestartofthetestwastakeneitherfromthecomputerlistingsor,inthemultiplevalveactuationtests,fromthecomputerplotsDuetotheinertiaoftheBartoncell,themeasurementvalueforwaterlevelinthedischargeline(measuringpointL4.1)inthemultipleactuationtests,especiallyforthe2nd,3rdandifapplicable,the8thactuation,mustbedisregardedorconsideredonlyasanindicativevalue.Thetemperatureinthedischargelineatthestartofeachtestwastakenfromthecomputerlistingsorthecomputerplotsforthemultipleactuationtests841.2BehavioroftheSRVandSystemPressuresToevaluatethevalvebehavior,thevalveopeningtime,t,wasdeterminedfromtherecordedvalveliftvariationforalltests.0Thisinvolvesthetimefromthebeginningofvalveopeninguntilattainmentofthesteadystatelift(seesketchbelow).Theseopeningtimesarelisted,forthelongdischargelinetests,inTable8.7and,fortheshortdischargelinetests,inTable8.8.Theassociatedsteadystateliftsarealsoindicated.AplotofthemeasuredvalveopeningtimesasafunctionofaccumulatorpressureatthestartofeachtestisshowninFigure8.23forthelongdischargelinetestsandFigure8.24fortheshortdischargelinetests.Theso-calledventclearingtimestpzarealsogiveninTables8.7and88ThisisthetimefromthebeginningofvalveREV.1,3/798-.35 PROPRIETARYopeninguntiltheinstant'ofmaximumpressureatmeasuringpointP4.4inthedischargeline.(seesketchbelow)tsvalveliftventclearingpressurepressurebeforequencher1'wovaluesareindicatedinTables8.7and8.8forsystempressuresmeasuredin:buffertank-P2.6beforetheSRV-P2.5inthedischargeline-P4.1toP4.4Thesetwovaluesarethepressureattheventclearingtime(ventclearingpressure)andthepressureapproximately1.5secondsafterthestartoftest(steadypressure)Theinitialparametersofrelevancefortheclassificationof-testsareindicatedintherowheadings.Theventclearingpressureinthedischargelinebeforethequencherinlet(measuringpointP4.4)isplottedversussystempressure(measuringpointP2.6)underCleanConditionsinFigure8.25forthelongdischargelinetestsandinFigure8.26fortheshortdischargelinetests.SeeSection8.5.2.1foradiscussionoftheventclearingpressuresandtheirdependenceonreactorpressure.84.1.3DynamicPressureLoadsonthePoolBoundariesAsreadofftheVisicordertraces,thepeakpositiveandpeaknegativepressureamplitudesduringventclearingformeasuringpointsP5.1-P5.3(bottompressures)andP5.4-P5.10{wallpressures)arecompiledinTable8.9forthelongdischargelinetestsandinTable8.10fortheshortdischargelinetests.Ina'ddition,approximatevaluesforthepredominatefrequencyofthepressureoscillationsareindicated.Thesefrequencieswerereadfromthevisicordertraces.Figures8.27and8.28showthemeasuredpeakpositivepressureamplitudesatthetankbottomdirectlybeneaththequencher(P5.2)andontheconcretewallatthequencher'smid-heightREV1,3/798-36 PROPRIETARY(P5.10)asafunctionofsystempressureforthelongdischargelineandshortdischargelinetests.ThetestpointsplottedareallCleanConditiontestswithcoldwaterinthetesttank{approximately25~C)anddischargelinecold(approximately50~C)(Longdischargelinetests1.1,2.1,3.1,4.11,4.81.1and32.1andshortdischargelinetests16.1,171,J.8.1,19.1.1and19.R1.1)AsacomparisonFigures8.29and8.30representcorrespondingmeasuringpointsfortestsperformedunderRealCondition(Longdischargelinetestsl.2,2.2,3.2,.10.4and32.2andshortdischargelinetets16.2,17.2and18.2).AscanbeseenthepressureamplitudesareslightlyhigherfortheCleanConditiontestsandnosignificantchangewithsystempressureisobserved.Figures8.31and8.32showthemeasuredpeakpositivepressureamplitudesatmeasuringpointsP5.2andP5.10forCleanConditiontestswithheatedwater{45C-80~COinthetesttankforthelongdischargelinetestsandshortdischargelinetestsrespectively.(Longdischargelinetests5.1.1,6.1,71,8.1,9.1,151.1and15.R1.1andshortdischargelinetests20.1.1,20.Rl.l,22.1,23.1,24.1).Again,asacomparison,Figures8.33and8.34representcorrespondingmeasuringpcintsfortestsperformedunderRealConditions(Longdischargelinetests6.2,7.2,8.2,92,11.2and12.2andshortdischargelinetests20.R1.7,22.2,23.2and24.2)Incontrasttothetestswithcoldwaterinthetesttank,thepressureamplitudesareslightlyhigherfortheRealConditiontests,butaswiththecoldwatertests,nosignificantchangewithsystempressureisobserved.Figures8.35to8.40showthemeasuredpeakpositivepressureamplitudesatmeasuringpointsP5.2andP5.10foranumberofmultiplevalveactuationtestsplottedagainstthecorrespondingvalveactuation.Figures8.41to865showthefirstsecondofvisicorderpressurestraces(forthepoolboundarypressures,P5.1-P5.10)fromvarioustests.8~414LoadsOnThequencherandBottomSupportThebendingstrainsonthetwoarmsofthequencherandatthebottomsupportwereeachmeasuredintwomutuallyperpendiculardirections.Theresultantbendingstrainsandbendingmomentswerecalculatedfromtheseindividualvalues.Thestrain-versus-timevariationsstoredonmagnetictapewerereadforthemaximumresultantduringventclearing.Ahigh-passfilterhavingacutofffrequencyof2HzwasinsertedinordertoruleoutanyfalsificationoftheevaluationduetoslowdriftingofthezeropointTheupperfrequencylimitwasat400Hzduetothemechanica1conditions.REV1~3/798-37 PROPRIETARYThemaximumresultantbendingstrainsdeterminedinthismannerandthebendingmomentscalculatedfromthemarecompiledinTables8.11and8.12forthelongandshortdischargelinetestsrespectively.Toclarifythedirectiondistributionoftheresultingbendingmomentsonthequencherarms,thecomponentsofthemaximumresultantbendingmomentsaredepictedinpolarcoordinatesinFigures8.66and8.67forthelongdischargelinetestsandFigures8.68and8.69fortheshortdischargelinetests.AsshowntheresultantbendingmomentsonthequencherarmsoccurprincipallyintheverticaldirectionFigures870and8.71forthelongandshortdischargelinetestsshowacorrespondingdistributionofthemaximumresultantbendingmomentsatthebottomsupport.Tables8.11and8.12alsoindicatethemaximumtorsionalstrainsandtorsionalmomentsmeasuredatthebottomsupportandthemaximumverticalstrainsandverticalforcesmeasuredatthebottomsupportduringventclearing.Thisdataisbasedonasevaluationofthevisicordertraces.842SteamCondensationTestResultsEightcondensationtestswiththeshortdischargelinewereperformedintheperiodfromJuly18,1978toJuly.21,1978.8.4.21TestParametersThemostimportantoftheparametersbeing-investigatedwasdescribedinSection8.3.AdetailedlistoftestparametersisgiveninTable8.6.CompiledinthatTablearetheparametersatthebeginningofthetests,suchas:typeoftestlengthofdischargelineaccumulatorpressurewatertemperatureintesttankwaterlevelindischargelinewaterlevelintesttankairtemperatureindischargelineTheaccumulatorpressurePl.lAandbuffertankpressureP2.6Aarethedeterminativevaluesforthesystempressureatthestartofeachtest.Thevalueswerereadbycomputerjustpriortothestartofthetest.Inaddition,thesepressureswerestoredcontinuouslyontapebutonlyupto360secondsafterthestartoftests36.1and40.1.Thiswasdictatedbythelimitedstoragecapacityoftheoperatinginstrumentationcomputer'smagneticdisk.Thisdatawascontinuouslystoredonthevisicordertracesandthetestinstrumentationmagnetictapes.REV1i3/798-38 PROPRIETARYForaccumulatorpressuresbelow30bar(435psi),measuringpointP2.5was,usedtodeterminethesystempressure,sincemeasuringpointsPl.1AandP2.6Awereoutsidethemeasuringrange.ThewatertemperatureatthestartofatestwastakenfromthecomputerlistingsandattheendofatestfromthecomputerplotsThevaluesforthewaterlevelsandairtemperaturesinthedischargelineatthestartofatestweretakenfromthecomputerlistings.Table8.13showstherelationbetweentheteststep,testnumber,andrangesofpressureandwatertemperatureastheyactuallyoccurred.842.2PresentationofTestResultsFirstwewillpresentasurveyoftheobservedcondensationphases.Thatisfollowedbyapresentationofthedynamicpressureamplitudesinthewaterregionofthetesttank.Finallythetemperaturevariationsinthewaterregionaredescribed.8.4.221Surve~ofObservedCondensationPhasesIntheoperationfieldofthequencherasgivenbythetestmatrix,theobservedcondensationphasesareindicatedinFigure8.71forblowdownsalongtheupperandlowerboundarylinesoftheoperationfield.8422.1.1BlowdownatlowMaterTe~meratureFortheblowdownalongthelowerboundaryline,thefollowingcondensationphaseswereobservedforthetestedpressurerange:AbsolutesystemPressureinBarCondensationPhaseTests70-25Stationary33.2,34.1,35.1,andinitialsectionof36.125-2IntermittentMiddlesectionof35.12-1Inthepipe(1)Endsectionof36.1(1)Itshouldbenotedherethatatthebeginningofthisphaseaportionofthesteamflowhasemergedthroughtheannulargapabovethequencherinlet.AsnotedinSection8.2.1.1.5,REV1,3/798-39 PROPRIETARYthisannulargapsimulateshydraulicallytheslidingfitofthequencherinstalledatSSES.Figure8.73showsatypicalexampleofthemeasurementtracesobtainedwiththebottomandwallpressuresensorsforstationaryoperationofthequencherintheupperpressurerange(test33.2).Figure874showsatypicalexampleofthelowerpressure'ange(test35.1).High-frequencypressureoscillationsoccurwithverylowamplitude,andwithoutanyfixedfrequency.Toillustratetheintermittentoperation,thevariationofthebottomandwallpressuresandtwopipepressuresthroughouttheentiredurationoftest361isshowninanextremelytime-compressedforminFigure8.75.Theintermittentcondensationphaseisclearlyrecognizableinthemiddlesectionofthetest.Figure8.76showsamoretime-expandedexcerptfromthatphase.Supplementarily,Figure877showsatypicalpowerfulindividualeventinanextremelytime-expandedform.Thehigh-frequencypressurepeakssuperimposedonthelow-frequencysinusoidalpressurepulsationsareclearlydiscernibleinbothFigures875and8.76.Forthephaseofcondensationinthepipe,thetesttracesexhibitnegligiblylowamplitudes,whichareclosetotheresolutionlimitofthemeasuringchain.Therefore,noexampleofsuchatraceisshown.8.4.2212-BlowdownatHighWaterTe~meratureForblowdownalongtheupperboundaryline,thephasesdescribedin8.4.2.2.1.1wereobservedinpracticallythesamepressureranges.However,theappearanceofthepressureoscillationsdifferstosomeextentfromthatofthepressureoscillationsatlowwatertemperature.First,hereistheobservedrelationbetweenpressurerangeandcondensationphase:AbsolutesystemPressureinBarCondensationphaseTests>70-4.Stationary372e38ls39leandinitialsectionof4014-2IntermittentMiddlesectionof40-12-1Inthepipe<>>Endsectionof40.1REV1,3/798-40 PROPRIETARY(1)Itshouldbenotedherethatatthebeginningofthisphaseaportionofthesteamflowhasemergedthroughtheannulargapabovethequencherinlet.AsnotedinSection8.2.1.1.5,thisannulargapsimulateshydraulicallytheslidingfitofthequencherinstalledatSSES.Forstationaryoperationintheupperrangeofpressure,Figure8.78showsatypicalexamplefortest37.2.1helowerrangeofpressureforthisphaseisrepresentedbyanexamplefromtest391(Figure8.79).Therearealsohigher-frequencypressureoscillationswithlowandverylowamplitude,respectively,andwithoutanyfixedfrequency.Atypicalexampleofintermitten~toerationisshowninFigure8.80byanexcerptfromtest40.1.Comparedtothisphaseatlowwatertemperature(seeespeciallyFigure8.76),adistinctattenuationofthestrengthofthepressurepulsationsisobservableathighwatertemperature.Superimposedhigh-frequencypressurepeaksdonotoccur.Forthephaseofcondensationintheprie,the.testtracesexhibitnegligiblylowamplitudesevenatextremelyhighwatertemperatureofmorethan90oC.8.4.2.2.2StatisticalEvaluationofthe~DnamicPressureLoadsonthePoolBoundariesAsdescribedinSection8.4.22.1,thesteamcondensationdoesnothaveanyuniformformthroughouttheentirerangeofsystempressureandwatertemperature.Tonowbeabletoquantifythedistributionofdynamicpressureamplitudesduringablowdownfrom70bartoapproximately1bar,therecordingsfromarepresentativebottompressuresensorandwallpressuresensorforallthetestswerestatisticallyevaluated.Thisalsoallowedustoinvestigatetheinfluenceofsystempressureandwatertemperatureonthedynamicpressureamplitudes.B.a2.22.1Dependenceof~DnamicBottomandIlailPressuresonSystemPressureandMaterTemperatureThepressure-timehistoriesstoredonmagnetictapeforpressuresensorsP5.2(bottompressure)andP5.10(wallpressure)wereeachreadformaximumvalueatuniformtimeintervals.Ahigh-passfilterwithafrequencycutoffof2Hzandalow-passfilterwithafrequencycutoffof500Hzwereinsertedintothecircuit.In'hismanner,afalsificationoftheevaluationduetoslowREV1,3/798-41 PROPHIETARYdriftingofthezeropointorduetoelectricalinterferencewaslargelyexcluded.Portests33.2,34.1,351,37.2,38.1and39.1,.auniformintervalof1secondwaschosenbecauseoftherelativelyshorttestdurationofamaximumof64secondsintest39.1.Intests36.1and40.1withtestdurationsofover800seconds,theuniformintervalwas4seconds.Inthesetwotests,thephasesofstationaryandintermittentcondensationandcondensationinthepipewerecoveredseparatelyatthesametime.Noerrorwasintroducedintotheevaluationbythedifferentchoiceofintervals,sincethemaximumvalueswerecoveredineachcaseTheextremevaluesdeterminedforthepositiveandnegativedynamicpressureamplitudesatthebottomandonthewallareplottedversusthetransientvariationofthesystempressureinFigures881and8.82.Duetothelargenumberofextremevalues,aselectionwasmadewiththeaimofconsideringonlythehighervalues.ThetophalfoftheFigureshowsthemeasuredmaximumpressureamplitudesfortheblowdownathigherandhighwatertemperaturealongtheupperboundarylineoftheoperationfield.Thebottomhalfshowsthemfortheblowdownatlowwatertemperaturealongthelowerboundaryline.AsimilarillustrationforthemeasuredmaximumwallpressureamplitudesisgiveninFigure8.82.Thepeakbottom-pressureandwall-pressureloadsmeasuredduringtheindividualcondensationphasesareindicatedasafunctionofwatertemperatureinTable8.14.Promthesepeakvalues,wecanascertainaslightdecreaseofthepressurelevelwithahotpoolforthestationaryandintermittentcondensationphases.Porthephaseofcondensationinthepipe,ofcourse,therearepracticallynodiffe'rencesinthepressurelevelsforcoldandhotpool-84.22.2.2Occurrenc~efreuenceDistributionsofthe~DnanicBottomandMallPressuresInparallelwiththedeterminationofextremevaluesasdescribedinSection8.4.2.2.2.1allpositiveandnegativepeakvaluesbetweenthezeropassagesofthepressure-vs.-timevariationsweredetermined.ineachtimeintervalandclassifiedaccordingtomagnitude.Thiscountingmethod,knownas<<peakcountbetweenzeropassages"or"meancrossingpeakcountmethod>>,avoidstheinclusionandconsequentialoverassessmentofsmallintermediateoscillations.Onlytheabsolutemaximabetweentwozeropassages'reincludedinthecount.REVli3/798-42 PROPRIETARXThecountresultsuppliestheclassoccurrencefrequencydistributionatonce.Positiveandnegativepeakvaluesweretreatedseparately.Anyerrorinthecountresultsbythenoiselevelonthemagnetictapeswaslargelyeliminatedbymeansofa.prescribedamplitudesuppressionof10mV=0.015bar.Auniformclassintervalof0.025barwaschosenforthehistograms.Inthatway,thehistogramsoftheindividualtestswereabletobecombinedintoanoveralldistributionforblowdownswithcoldandhotpool.ThehistogramsofthepositiveandnegativeamplitudesofthedynamicbottompressuresatmeasuringpointP5.2areillustratedinFigures8.83and884forblowdownswithcoldandhotwater,respectively.AnalogoushistorgramsforthewallpressuresatmeasuringpointP5.10areshowninFigures8.85and886.8.4.2.2.2.3StatisticalCharacteristicsoftheDynamicBottomandMallPressuresInfluencesoftestparameterscanbereadofffromthestatisticallydeterminedmeanvalues,sincethosevaluesareobviouslymuchmoretypicalthanthemagnitudesofindividualandveryraremaximumvalues.Themeanvaluesweredeterminedbythegroupvaluemethodsusingthefollowingequation:PkZn.Pi~1KZn<illwherePG=meanvalue;Fifrequency.classmeanvalue;n=classThegroupvaluemethodwasindividualhistogramsofadistributions.Thosemeanto8.86alsousedforthecombiningoftheblowdowntogetthemutualfreqeuncyvaluesareindicatedinFigures8.83Ingeneral,thetrendsaresupportedbythemaximumvalues.Theunavoidablescatterofthemaximumvaluesisallowedforbyformingtheaveragevalueofthe10highestamplitudesineachtest.Duetothesmallnumber,theyweredeterminedbythesingle-valuemethod:wherePE=NZPi~1NREVli3/798-43 PROPRIETARYPE=meanvalue;P.=singleextremevalue;N=numberofextremevaluesTables815and8.16provideanoverviewoftheabovementionedmostimportant.statisticalcharacteristicsofthepressure-timehistoriesatthebottomandatthewall,respectivelyfortests33.2to40.1.Indicatedare:maximumvaluerelativetotheentiretest,meanvaluerelativetotheentiretest,-lowerlimitvalueofthe10highestvalues,meanvalueofthe10highestvalues.*Besidethedataconcerningthesystempressuresandwatertemperatures,thecondensationphasesarealsolisted.Intests36.1and40.1,thephasesofstationaryandintermittentcondensationandcondensationinthepipeweretreatedseparately.'igures8.87and8.88showplotsofthemeanvaluesrelativetotheentiretestortestsectionandthemeanvaluesofthe10highestvalues,asfunctionsofsystempressure.Themeanvaluesofthebottomandwallpressuresareslightlyhigherfortheblowdownwithacoldpool.Thistrend,alreadyalludedtoinSection8.4.2.2.2.1onthebasisoftheabsoluteextremevalues,isthereforeverifiedstatistically.,Thelevelofthemeanvaluesfromthe10highestvaluesishigherbya-factorofapproximately3-4thanthelevelofthemeanvaluesrelativetotheentiretestortestsection.8.4223Te~meratureVariationsintheRaterRegionoftheTestTankFourtestswereselectedtoillustratethetemperaturevariationsinthewaterregionofthetesttank:test33.2forhighsystempressureandcoldpool,test35.1forlowsystempressureandcoldpool,-test37.2forhighsystempressureandhotpool,test39.1forlowsystempressureandhotpool.Figures8.89to892showtheverticaltemperaturedistributionobtainedfromthemeasuringpointsT5.5,T5.2,T5.3andT5.4arrangedaboveoneanotherontheconcretewallIneachcase,themeasuredtemperaturesarescatteredaboutameancurve.ThescatterisgreatestformeasuringpointT52(approximatemax.a8oC).Thatmeasuringpointisattheheightofthequencherarmandisimpingedupondirectlybythesidewardsdirectedflowimpulse.ThescatterisleastformeasuringpointT5.4(approximatemax.a50C).Thescattercanbeexplainedbythehighdegreeofturbulenceinthe.pool.REV~1i3/798-44 PROPRIETARYFigures8.93to8.96showthetemperaturevariationsatquencherarm1forthesametests.AtmeasuringpointT5.8locatedinthemiddleoftheholearray(seefigure8.14)adistincttemperatureincreaseofapproximately15-200C,ontheaverage,wasrecordedrelativetothepooltemperature.Incontrast,thetemperaturesattheupperedgeoftheholearray(T5.9)andattheupperedgeoftheguencherarm(T5.10)aresomewhatlowerthanthepooltemperatureatT5.1duetoasufficient"coldwatersupply".Thisisanindicationofthegoodcirculationofwaterneartheguencher.This.confirmedtheexpectedcondensationbehaviorofthequencherasrelatedtothelayoutoftheholearray.(SeeSection4.1.1.1).84.2.24WaterLevelintheDischargeLineWhenOpeningandAfterClosi~ntheSRVInthetestswiththelongdischargeline,thewaterlevelinthepipewasmeasuredbythe"LevelProbes"LP4.1thruLP4.4atfourpositions,oneaboveanother.Inthetestswiththeshortdischargeline,thisinstrumentaitonwasextendedbythemeasuringpointLP4.5abovethemeasuringpointLP4.4;seeFigure8.8ThemeasurementsignalsfromtheseLevelProbeswererecordedonvisicordersandmagnetictape.ABartoncell,measuringpointL4.1inFigure8.4,wasusedtosetandmeasurethewaterlevelinthedischargelinebeforeteststart.ThereadingofthatmeasuringpointwasinterrogatedbythecomputerbeforeandduringthetestandwasstoredTheindicationsoftheLevelProbesandalsotheindicationsoftheBartoncellwereusedtodepictthetimevariationofthewater.levelinthedischargeline.ItmustbetakenintoconsiderationthattheresponsespeedoftheBartoncellistooslowfortherapidchangesofthewaterlevelduringventclearingandaftertheclosingoftheSRV.Themeasuringpointwasusedessentiallytodeterminethesteady-state.waterlevelsinthedischargeline.Figures8101and8.102showtwotypicalexamplesofthevariationofthewaterlevelinthepipefortheintervaltest15.1withthelongdischargelineand20.1withtheshortdischargeline.Itwasfoundthatintwoinstancesinintervaltest15.1(Figure8.101),thewatercolumnbrieflyexceededtheexternalwaterlevel,butfellbackimmediately.Thesetwotestpointsrepresentthemaximumwatercolumnrisemeasuredintheventclearingtests.Intheintervaltest20.1,thewatercolumndidnotreachtheleveloftheexternalwatersurfaceinanyinstanceafterclosingoftheSRV.Themaximumwaterlevelrisewasgenerallyfound,inalltests,tooccurafterthethirdvalveactuation.REV1,3/798-45 PROPRIETARYToevaluatetheeffectofvacuumbreakeroperationonthewatercolumnrefloodfollowingventclearing;Test32,withonelockedvacuumbreakerandatimeintervalof3secondsbetweentheclosingofthevalveafterthefirstactuationandthenextactuation,wasincluded.Figure8-105showsthevariationofthemovementofthewatercolumninTest32.Ascanbeseennoadverseeffectswererecorded.8.43CheckingandCalibrationoftheNeasuringInstrumentationThecalibrationandtheelectricalandphysicalcheckingofallsensorsbefore,duringandafterthetestswereperformedinaccordancewiththeTestandCalibrationSpecifications.Fig.897showsdiagrammaticallythephysicalcalibrationofthesensors,thesettingandcalibrationoftheamplifiersandrecordinginstruments,andthequalityinspectionofthesensors.Pig.8.98showsthetimeintervalsstiplatedforthechecksandcalibrationsintheTestandCalibrationSpecifications.Fig.8.99clarifiesthechainofthecalibrationsystemfromthenationalstandardsofthePhysikalisch-TechnischeBundesanstalt(PTB)tothemeasuringinstruments.ThepressuresensorsP5.1thruP5.10usedinthetestswerefullyoperableuntiltheendofthetests.Thelowestinsulationresistanceof1.2x10~0measuredatP5.1afterthetestscanbeclassifiedas"good".ThepipepressuresensorP4.1failedon31Nay1978Itwasreplacedbyanewsensorforthesubsequenttests.WiththisnewsensorP4.1~thelowestinsulationresistanceforthegroupofpipepressuresensorsafterthetestswas3x10~~,whichwasverygoodTherewerenofailuresforthestraingaugesSG41thruSG4.8,SG5.1andSG5.2Herealso,averygoodinsulationresistancelevelwasrecordedwithalowestvalueof3x10~uatSG4.6afterthetests.J.ikewise,noneofthetemperaturemeasuringpiontsT5.1thruT5.10failed.Thelowestinsulationresistanceof1.3x10~~wassufficientlyhigh.844AnalysisofNeasurementErrorsBasedoninformationfromthemanufacturersofthemeasuringinstruments,KWUsowninvestigations,andtakingintoconsiderationtheexperienceaccumulatedinsimilartestprojects,themaximummeasurementerrorsfortheindividualsensorscanbeindicatedasfollows:REV.1,3/798-46 PROPRIETARYPressuresensorsP5.1thruP5.10Linearityerrorofthesensor2.5%ofmeasuredvalueinrangeof0to2barError2.5'5Reproductionerrorofthesensor0.2%of5barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder0.01bar05%0.5%Max.totalerrorx[0.01bar+3.5%ofthemeasurementvalue]PressuresensorsP4.1thruP4.5ErrorLinearityerrorofthesensor0.5%ofmeasuredvalueinrangeof0to20bar0.5%Reproductionerrorofthesensor0.1%of35bar0.035barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder05%05%Max.totalerrora[0.035bar+1.5%ofthemeasurementvalue]PressuresensorsP2.3andP25ErrorLinearityerrorofthesensor1%ofmeasuredvalueinrangeof0to40barReproductionerrorofthesensor0.1%of140bar.0.14barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder0.5%05%Max.totalerrora[0.14bar+2%ofthemeasurementvalue]StraingaugesSG4.1thruSG48~SG5.1~andSG5.2ErrorToleranceoftheguagefactorInfluenceoftemperatureontheguagefactorREV1,3/798-47 PROPRIETARYErrorofthemeasuringamplifierErrorofthebalancingunitanrecorder05%0.5%Max.totalerrora5%ofthemeasurementvalueTemperaturemeasur~in~pintsT5.1thruT5.10ErrorofthesensorlocErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder05%0.5SMax.totalerrorx[l~C+1$ofthemeasurementvalue]AfterthefirsttestsonMay10,1978andafterconclusionofthetestsonAugust2,1978,additionalphysicalchecksofthepressuresensorsinthewaterregionwereperformedbyincrementalloweringofthewaterlevelinthetesttank.Themax.deviationsfromthenominalvaluewereapproximately+0.01and-0.02bar.Fig.8.100illustratesafrequencydistributionofthesedeviationscombinedfrombothchecksandforallpresuresensors.ItshowsatypicalGaussiandistribution.Inordertorecordthehigh-frequencyprocessescorrectlyinfrequencyandamplitude,thedatawasacquiredinanalogformonmagnetictape.Porasensoreigenfrequencyofapproximately30kHz,thedynamicrangewaslimitednotbythesensorsbutratherbythecarrier-frequencymeasuringamplifierslocatedfurtheroninthecircuit.Thefrequencycutoffofthemeasuringamplifierswasat1.5kHzandthatofthemagnetictaperecorderswasat2.5kHz.ThefrequencycutoffsofthevisicordersweredeterminedbytheutilizedgalvanometersThesefrequencycutoffsareapproximately1kHz.Thefrequencyresponseofeachindividualgalvanometerwascheckedpriortothetests.8.45RepetitionTestsand~ReroducibilityoftheResultsToverifythereproducibilityofthemeasurementresults,arepetitionof5testswasspecifiedintheTestMatrix.Basedonapreliminaryassessmentoftheresultsafterconclusionofthetestserieswiththelongandshortdischargelines,thefollowingtestswererepeated(asmentionedpreviously):Longline:4.1through4.Rl15.1through15.RlIntervaltestsIntervaltestsShortline:REV1,3/798-48 PROPRIETARY19.1through19.R2201through20.R125.1through25.R2IntervaltestsIntervaltestsSingleActuationtestsInadditiontotherelevantinitialconditions,Table8.17alsogivesthemeasuredventclearingpressure(measuringpointP4.4),max.dyn.bottompressures(measuringpointP5.2)imardyn.wallpressures{measuringpointP5.10)andfrequenciesofthepressureoscillationsforthefirstSRVactuationineachoftherepetitiontests(>>CleanConditionstests").Acomparisonoftheabove-citedvaluesfortherepetitiontestsassociatedwitheachotherdemonstratesthegoodreproducibilityunderCleanConditions.Themaximumdeviationsfromthemeanvalueforeachpairofrepetitiontestsare(seeTable8.18):fortheventclearingpressureforthebottomandwallpressures10.75barora6%a0.05barora7Xforthefrequencyofthepressureoscillations105Hzora7%Themeandeviationsfromthemeanvalueofrepetitiontests,averagedforall5pairsfortheventclearingpressureforthebottomandwallpressuresforthefrequencyofthepressoscillationseachpairofoftests,are:10.37baror13K1002barora6%RO2HzOrX5%Figures8.37and8.38illustratesthemax.dynamicpressuresinthepoolduringtheventclearingforthemultiplevalveactuationrepetitiontestswiththelongline.Figures8.39and840showsthesamethingforthemultipleactuationrepetitiontestswiththeshortlineIncomparisonwiththefirstSRVactuationsunderCleanConditions,somelargerdeviationsareexhibitedhereinthetestsunderRealConditions(2ndto7thand10thSRVactuations).Thereasonforthesedeviationsisthattheinitialconditionsdiffersignificantlyfromeachother.Thevisicordertracesforeach"cleancondition"actuationatarepetitiontestpointisprovided:Tests4.1.1and4.Rl.l-Figures8-41and8-42REV1i3/798-49 PROPRIETARYTests15.1.1and15Rl1-Figures8-46and8-47Tests19.1.1and19.R2.1-Figures8-48and8-49Tests20.1.1and20.Rl.l-Figures8-59and8-60Tests25.land25R2-Figures8-64and8-65Avisualcomparisonofthetracesfromeachrepititiontestalsoshowsgoodreproducibility.Accordingly,itcanbesaidthat:Iftheinitialconditionsofthetestsaresetinacontrolledmanner(CleanConditions),thenthetestresultsarereproducible.Iftheinitialconditionscorrespondtotherandomlyprevailingoperatingstates(RealConditions),thenthemeasurementvalueslieinalargerscatterrange.85DATAANALYSISANDVERIFICATIONOFLOADSPECIFICATION8.5.1EvaluationofTestTankEffectsonBoundaryPressureNeasurementsInthisSection,vepresenttheoreticalandexperimentalinvestigationswhichshowthattheKarlsteintesttankrepresentsagoodsimulationofthehydraulicconditionsoftheSSESsuppressionpool.Meareconcernedprimarilywiththeeffectsexertedontheprocessesinthevaterbytheexistingboundarysurfacessuchasthewatersurface,tankbottom,movableorimmovabletankwalls.TheresultsoftheinvestigationfacilitatetheevaluationandtranspositioncftheboundaryloadsmeasuredintheteststoSSES.85.1.1EffectsofFreeRaterSurfaceandR~iidWallsTheeffectsofthefreewatersurfaceandtherigidwallsofthetankonthefluidpressurewillbeexplainedfirstbymeansoftheexamplesillustratedinFigure8-104.ThetophalfoftheFigureshowsthevelocitypotentialandflowfieldofasphericalbubblesubjectedtooverpressureorunderpressureinaninfinitelyextended,incompressiblefluid.Thepotentialfieldisdescribedbyasimple1/rlaw(Reference35).If,forexample,thesamebubbleislocatedinacylindricalrigidtankwhichispartiallyfilledwithfluid,thenthepotentialfieldandflovfieldhaveavisiblydifferentappearance(Figure8-100,bottom).Thedifferences.inthenonstationaryfluidpressure,whichisproportionaltothevelocitypotentialforsufficientlylovflowvelocity(pressurefield=potentialfield;seeReference4forexample),areclearlyevidentinthepressureprofilesontherightsideoftheFigure8-104.Thefreewatersurfaceconstrainsthepressuretozero,vhilethecylindricalwallcausesanincreaseinglymorepoverfulpressurerisewithREV1,3/798-50 PROPRIETARYincreasingdepth.Thenarrowerthetank,thegreateristhepressurerise.ThecalculationsrelatingtoFigure8-104wereperformedbythefinite-elementsmethod{Reference34)foratankdiameterof3mandawaterdepthof6m.Thebubblewas2.8mdeepand0.8mindiameter.Besidesthepressurefield,thereisalsoaneffectonthewatermasswhichiseffectivelyentrainedbythebubbleduringpulsationmotions(pressureoscillations)andthusalsotheoscillationfrequency.InthecaseshowninFigure8-104,thebubbleinthetankhasalargercoupledmassthanintheinfinitelyextendedmedium.Thisismanifestedbythefactthatthepulsationfrequencyofthebubbleiscorrespondinglylower(seeSection8.5.3.2).8.512MethodofImagesThemethodofimagesisanimportantaidwhichmakesitpossibletoclearlyunderstandthehydraulicactionsofthewatersurfaceandrigidwallsandtocalculatethemquantitativelyinasimpleway(Reference35).Ztisbasedonthefactthattheinfluenceofaplanerigidwallontheflowfieldofahydrodynamicpointsourcecanberepresentedbyasuperpositionoftheflowfieldwithoutthewall(infinitelyextendedfluid)andtheflowfieldofanimagesourceofidenticalsignandidenticalstrengthlocatedbehindthewall(Fig.8-105).Thesameholdsforaplanefreewatersurface,exceptthattheimagesourcehastheoppositesign.Usingthismethodofimages,theflowfieldofapointsourceinarectangular,vesselisobtainedfinallybyrepeatedapplicationofsuitableimagingoperations(Figure8-105dandFigure8-2).Theimmediatesignificanceofthemethodofimagesliesinthefactthatapulsatingbubblecanbeconceivedofasahydrodynamicsource,thusprovidingasimplemethodtocalculatethepressurefield.Ofspecialimportancefortheperformanceoftestsistheconsequencederivedbyinversionofthemethodofimages:Aconfigurationofbubblesoscillatinginparallelcanbesimplifiedinatestbysurroundingonebubblewithrigidwalls.Thiswillbeclarifiedfurtherinthefollowing.851.3TheTestStandasaSincCleCellBasedontheabovediscussion,anoscillatingbubbleinarectangularvesselisequivalenttoaplanefieldofsimultaneouslyoscillatingbubbles{Figure8-2).FromFigure8-2itfollowsfurtherthatvesselswithseveralbubblesarealsoequivalent,sincebetweeneachpairofbubblestheimagingwallsectioncanalsobeomitted.REV1,3f'798-51 PROPR1ETARYApplicationofthemethodofimagestothetranspositionofasystemofvalvesblowingdownsimultaneouslyinaplanttoateststandwithaquencherleadstothecelldivisionillustratedinFigure8-3..Asdiscussedinsection8.1,thewaterspaceoftheteststandwasformedaccordingtotheinteriorsinglecellsC,F,KandN(Figures8-3and8-108),sincetheyarethenarrowestandwillthereforeexhibitthehighestwallandbottompressures.Thatcanbeseenbyobservingthat,accordingtotheimagingprinciple,theyconservatively'simulatemorequencherslyingclosertogetherthanisactuallythecaseintheSSESsuppressionpool8.51.4SpatialDistributionsofPressureintheTestTankTogetmeaningfultestresults,pressuresensorshavetobemountedatsuitablepointsinthetesttank.Aseriesoftheoreticalinvestigationswasperformedinordertobetterassesstheirarrangement.Theyconsistedofcalculatingthespatialdistributionofpressurealongthetankwallsforvariousbubbleconfigurationsunderwater.TheKRUcomputercodeVELPOTwasusedforthisinvestigation.Abubblewassimulatedbyapoint,sourcenormalizedtounitsourcestrength.TheresultsareillustratedinFigures8-107to8-109.Figure8-107showsthecalculatedwallpressuredistributionforabubbleinthreedifferentpositionsnearthequencher:Case1Sourceonthetankaxis,0.7mabovethequencheraxisCase2Sourceonthetankaxis,atquencherelevationCase3Sourceatcenterofthequencher(eccentric).Theresultsshowthat.,theeccentricarrangementofthequencherwhichbecamenecessarybecauseofspacelimitationsinthetank,includingthecorrespondingpositioningofthepressuresensors(blacksquaresinFigure8-107),results,theoretically,inslightlyhighermeasurementvaluesforthepressures.Thenextcalculation(case4,Figure8-108)servestoanswerthequestionastohowthebubble'sforminfluencesthepressuredistribution.Todothat,thesinglesourcefromcase3,figure8-107,wasreplacedbyfouridenticalsourceswiththesametotalsourcestrength.Figures8-108and8-109showthattherearenomajordifferences.Notealsothegoodagreementseenbetweenthemeasuredpressuresfromshakedowntest081andthecalculatedvaluesi'nFigure8.109.Themodelcases3and4(singlebubbleatcenterofquencherand4-bubblearrangement)arebestadaptedtotheteststandgeometry.Sincetheassociatedpressuredistributionshardlydifferatall{Figure8-109),itisdemonstratedthatanexactREV1,3/798-52 PROPRIETARYknowledgeoftheairdistributionunderwaterisnotnecessaryforacorrectarrangementofthepressuresensorsInordertodemonstratetheconservativenatureofthechosensinglecell,asalreadyexplainedinSection8.5.1.3,thepressuredistributionformodelcase4iscomparedtothedistributioncalculatedfortheSusquehannaplantinFigure8-110.ThepressuredistributionintheteststandenvelopsthepressuredistributionintheSSES.Furthermore,thepressuredistributionintheteststandisenvelopedbythespecifieddistribution{FigureB-ill).8.5.1.5InvestigationoftheInfluenceofSavableSalleontheMeasurementResults/Fluid-StructureInteracti~on8.5.151GeneralRemarksIntheprecedingdiscussion,itwasassumedthatthesinglecellhasrigidandimmovablewalls.TheconstructionoftheKarlstein.testtankissuchthatthetank,despiteaseriesofstiffeningribs(seeFigures8-10to8-12),stillhasaresidualcompliance.Thetime-varyingloadsactingduringtheblowdownofthequenchercanthereforeexcitethetankintooscillationduetoFluid-StructureInteraction(FSI).Usingexperimentalandtheoreticalinvestigations,itwillbeshownthatinfluencesoftankoscillationsonthemeasuredboundaryloadscanbeneglected.Theexperimentalinvestigationsconsisted,firstly,ofmeasuringthetank'sresponsetoashortpressureimpulsewhichwasproducedbyanexplosivechargedetonatednearthequencher(Section8.5.1.5.2).Measurementsmadeduringthestart-uptestsontheteststandthensuppliedthetank'sresponsetotheloadsoccurringduringventclearing(Section85.1.5.3).Takingintoconsiderationtheinpulseresponse,itturnsoutthateffectsoftankoscillationsattheeigenfrequenciesarenegligible.Thisstatementislaterconfirmedbycalculationsandalsoisextendedtoforcedoscillations.85.1.52Ex2erimentalInvestigationoftheTank'sNaturalOscillationsTheexperimentalinvestigationofthetank'snaturaloscillationswasperformedwithimpulsiveexcitationbyanexplosivechargeinthewaterandsimultaneousmeasurementofthedisplacementsofthewallandbottomsectionsandofthefluidpressure.ThearrangementofthechargeandsensorsinthetankisillustratedinFigure8-112.Thepositionofthechargewaschosensuchthatthespatialloadprofileinthetankmatchestheprofileoftheblowdownloadsaswellaspossible.ThechargeitselfwasastoichiometricmixtureofhydrogenandoxygenwhichREV1,3/798-53 PROPRIETARYwasignitedinaplasticallydeformableflatcontainer(Figure8-113).Eightdisplacementtransducers(WA1toWA8)wereavailableforthedisplacementmeasurements.Theywerepositionedwiththeaimofobtainingthemostusefulinformation.Thearrangementofthepressuremeasuringpointsinthewater(P5.1toP5.10,Figures8-10to8-12)wasthesameasinthelaterblowdowntests.AsfortheevaluationofthepressuretracesinSection8.5.3,transducerP5.10waschosenasreferencepressuretransducerThechargewaslocatedatdifferentpositionsnearthequencherasshowninfigure8-112,i.nordertoobtainenvelopingloadprofiles.AtypicalresultisillustratedinFigure8-114,whichshowstherecordingsfromdisplacementtransducersWAltoMA8andpressuretransducerP5.10fortestno.2(chargeinposition2).Thelowestoccurringfrequenciesarebelow1Hz,buthavenothingtodowiththetank'sresponse,butratherrepresentsashiftofthezeropointThelowesteigenfrequencyofthetankisatapproximately13HzandisseenclearlyintheresponsefromtransducersWA2andWA3oscillatinginphase.Bothgagesareseatedonthebox-shapedstiffeningringsasshowninfigure8-112.Atthewallsectionsbetweenthestiffeners(WA4andMA6)andatthebottom(WA8),thefrequenciesthatoccuraremainlybetween30and60Hz.Theoscillationsoftheflatlowerstiffenerrings(WA1andWA5)arelesspronounced.Thesmallestdisplacementsarefoundattheconcretesections(WA7),wheresomeoftheamplitudesaresmallerhyanorderofmagnitude.ThepressuresignalfromP5..10showsdistinctexcursionsonlyduringthefirst100ms.Tobeabletobetterevaluatethetank'sfrequencyresponse,themeasuredtimevariationswereZourieranalyzedandpowerspectrawereformed.Thespectraassociatedwiththedisplacementtransducersonthesteelwall(MA2),concretewall(MA7)andbottom(WA8)andthepressuretransducerPS10areshowninFigures8-115to8-118.Itturnsoutthatthepreviouslymentioned13Hzoscillationinthelow-frequencyrangeisofgreatestimportance.Theassociatedtankdeformation(eigenmode)canbederivedfromthepointcorrelationsshowninFigure8-119.There,thedisplacementsofthedisplacementtransducersWA2,WA3andMA7,filteredbyabandpassfilterat13Hz,areplottedagainsteachotheratthesametimes.Thefitlinethroughthesetofpointshasapositiveslopeinthetopgraphandanegativeslopeinthebottomgraph.Therefore,displacementtransducerWA3(steelwallaboveMA2;seeFigure8-106)-oscillatesinphasewithWA2,whiledisplacementtransducerWA7(concretewall)oscillatesoutofphase.Thismeansthatthe13Hzoscillationcorrespondstoanovalizingmotionofthewall(seeFigure8-120).REV1,3/798-54 PROPRIETARY8.5.1.53ExperimentalInves~tiationoftheTank'sRe~sonsetoVentClear~inLoadsTheinvestigationsofthetank'sresponsetoventclearingloadswereperformedduringtheteststandshakedowntests.Tomeasurethetank'sresponse,thechoicewasmadetouseonedi,splacementtransducereachonthesteelwall(WA2),ontheconcretewall(WA7)andonthebottom(WA8).TheinstrumentationisshowninFigure8-121.Test08.1representsatypicalexampleoftheshakedownteststhatwererun.ThemeasuredtimehistoriesofthewallandbottomdisplacementsandofthereferencepressureP5.10areshowninFigure8-122.Thezero-pointdriftmentionedabovewaseliminatedbyusinga2Hzhigh-passfilter.Itcanbeseenthatboththepressureandthedisplacementsoscillateatthesameprincipalfrequencyof5.1Hz.Thesteelwall(WA2)andbottom(QA8)moveinphase.Theverysmallmovementoftheconcretewall(WA7)isalmostoutofphasecomparedtothepressureP5.10.Inaddition,thedisplacementtransducerWA8recordsahigher-frequencyoscillationat30Hz.Ithasalreadybegunweaklyatteststart,thendevelopsstronglyataboutthetimeoftheventclearing',andthendecaysagainabout300mslaterThephysicalinterpretationofthe5Hzoscillationisobvious.Thepressureoscillationiscausedbythepulsationoftheairbubblewhichiscreatedduringventclearing.Atthesametime,thetankcarriesoutforcedoscillationsatthefrequencyoftheforcingforce(5Hzpulsationoftheairbubble).Thesometimesphase-opposednatureofthedisplacementsofthesteelwallandbottom,ontheonehand,andtheconcretewall,ontheotherhand,makesitevidentthattheabove-discussedovalizingeigenmodeplaysadominantrole.Theoriginoftherapidlydecaying30HzoscillationseenatWA-8attheteststartisattributedtolocalforcestransmittedthroughthedischargelineandthequenchersupportduringventclearing.'Figures8-123todisplacementtimeduringshakedownspectraldensityduringshakedownlittleinfluence30HzlocaleffecfromP510showseffects.8-125showthepowerspectraldensitiesofthehistoriesforgagesWA2,WA7andWA8measuredtest081.Figure8-126showsthepowerofthepressuretimehistoryforP5.10measuredtest08l.A,reviewofthesefiguresshowsveryfromthe13HztankeigenfrequencyorfromthetseenatWA8.Figure8-126showingthersultspracticallynoinfluencefromeitheroftheseREVli3/798-55 PROPRIETARYPromthisitcanbeconcludedthatforallpracticalpurposestheKarlsteintesttankisrigidandhasnoinfluenceonthepoolboundarypressuremeasurementsmadeduringthetests.8.5.1.54TheoreticalInvestigations.andModelCalculationsoftheInfluenceofFluid-StructureInteraction851.5.4.1ComputationModelsTheanalysisdescribedbelowtocomputetheFSIonthemeasuredpressuresintheKarlsteintesttankwasperformedbyusingtheKWUcomputercodeKOVIBlAwhichwasdevelopedoriginallyandusedsuccessfullyfortheanalysisoffluid'-structureinteractioninthewaterpoolofKWU's69ProductLineBWRPlant.Theunderlying-theoryfollowsfromauniformformulationofthemechanicalprocessesbasedonpotentialtheoryandclassicalLagrangeandynamics.ItunifiesthedynamicsofthebubbleandtheFSIbyusingtheresultsofmodalanalyses.Inparticular,thefeedbackeffectsbetweenbubbleandstructureviathefluidareinc1uded.8.5.1542ModelParametersand~InutforCalculationsWithoutThemodelparametersandinputquantitiesforcalculationsoftheairbubbleoscillationsintherigidtankare:airmassflowintothebubble,watertemperature(=airtemperatureinstationaryequilibrium),hydrostaticpressureatbubbleposition,hydrodynamicmassparameterofthebubble,spatialpressuredistribution,initialvalues,(hubbleradius,etc.).Thetotalairmass(integratedairmassflow),watertemperatureandstaticpressureatthebubblepositionareobtainedfromthetestdata.Thehydrodynamicmassconstantofthebubbleandthespatialpressuredistributionareobtainedfromthecorrespondingpotentialcalculations(Figure8-107,case1).Thetimevariationoftheairsupplyintothebubblewasadjustedheuristicallybymeansofsystematictrialanderror,inparallelwiththeinitialvalues,insuchawaythatthecalculatedandmeasuredtimevariationsofthepressureattransducerP5.10exhibitedoptimalagreementThestart-uptest08.1wasusedasreferencetestforthesecalculations.TheairmassflowdeterminedinthismannerisillustratedinFigure8-127REVli3/79~8-56 PROPRIETARY8.5.15.43NodelParametersandXnautforCalculationswithPdjJustasforthedeterminationoftheairsupplyintothewaterpool,asemiempiricalmethodisusedforthestructuraldynamicsdata.Theyaredeterminedonthebasisoftheeigenfrequencymeasurementsdescribedpreviously.Inputdataforthecalculationare:eigenfrequency,modalmass,modalweight,dynamicpressuredistribution.Basedontheimpulseresponseofthetank(Figures8-115to8-117),itisplausibletoselecttheoscillationmodelyingat13Hz.Thatfixesthefrequency.Themodalmasscannotbetakendirectlyfromtheexperiment,butrathercanbedeterminedindirectlyviathemeasuredunitdisplacementsofthewall.TheunitwalldisplacementisillustratedinFigure8-128.Itisobtainedfromdisplacementsatthedisplacementtransducersbybandpassfilteringat13Hzandplottingsimultaneousvaluesofdisplacementwhicharenormalizedto1atthewatersurface.Thedisplacementdirectionisdefinedaspositiveiftherelevantwallsectionmovesinward.Thehydrodynamiccomponentofthemodalmass,(coupledwatermass)isthencalculatedbymethodsofpotentialtheory.Themodalweight,whichisequaltotheintegralloadrelativetothemodalmassandaveragedovertheunitdisplacement,isbasedontheloaddistributioncalculatedforcase1(Figure8-107,centeredbubble).Thedynamicpressuredistribution(seeFigure8-129)isobtainedfromtheunitdisplacement.bymeansofpotentialcalculations.85.1.5.4.4ResultsoftheFSIcalculationsTheresultsofthecalculationsconcerningtheinfluenceofFSIareshowninFigures8-130and8-131.Figure8-130showsthecalculatedtimevariationofthepressureatpressuretransducerP5.10,firstintherigidtank(withoutFSI)andthenintheelastictankwiththe13Hzeigenfrequency.Thereisaveryslightreductioninthepressureamplitudes,butitiscertainlynegligibleincomparisontothescatterofthemeasurementvaluesthemselves.AsisevidentfromFigure8-131,thefrequencyinfluenceofFSIalsocanbeneglected.InthatFigure,theoscillationfrequencyofthebubbleisplottedagainstthebubblevolume.ThebubblehasaslightlylowerfrequencywithFSIeffectsincludedthanwithout.REV1,3/798-57 PROPRIETARYAphysicallyclearexplanationoftheveryslightFSIeffectsfoundintheKarlsteinTestTankcanbeobtained.bycomparingthevolumesoffluidwhicharemovedbytheoscillatingwallandbottomandbythepulsatingbubble.Forabubblevolume{longline)of2.2m~andpressurefluctuationsofx0.4bar{seeFigure8-126),thevolumechangeofthebubbleisapproximately1m~isentropically.Incontrasttothis,fordisplacementslikethosefoundinFigures8-124and8-125thewallsandbottomuseuponlyabout0.05m~,whichisonly5%ofthewatervolumecomingfromthebubble.Therefore,duetothecomplianceofthetank,95%ofthewaterflowsupwardinsteadof100%(rigidtank).Thus,theresultoftheexperimentalandtheoreticalFSIinvestigationsisthateffectsofthecomplianceoftheKarlsteintesttankwallsandbottomonthepressureloadsmeasuredontheboundariesofthetankduringthetestscanbeneglected.852VerificationofSHVSystemLoadSpecificationDuetoSRVActuationThepressuresinsidetheSRVdischargelineweremeasuredatfourmeasuringpoints:justbehindtheSRVatmeasuringpointP4.1,inthecenteroftheblowdownpipeatmeasuringpointP4.2(measuringpointP4.5fortheshortdischargeline),justabovethenormalwaterlevelatmeasuringpointP4.3,andjustbeforetheinletofthequencheratmeasuringpointP4.4(seeFigure84).ThelongandshortdischargelinesareillustratedinFigures8-5and8-6.ThemeasuredpressuresinthedischargelinearedocumentedinSection8.4.1.8.5.21PressuresDuri~ntheVentClearincCProcessTypicalmeasurementtracesofthepressuresinthedischargelineareshowninFigures8-132and8-133.TheventclearingpressureisreadoffatP4.4.AsdiscussedinSection8.41,theventclearingpresureisdefinedasthepressurewhichisreadoffatthefirstpressuremaximumatP4.4.Atypicalfeatureofthispressurevariationisthedynamicovershootofthepressureabovethestationaryvalue.This-phenomenondoesnotoccurinsuchapronouncedmannerattheotherpressuretransducersalongthedischargeline.Thisdynamiceffectindicatesthatthepressurerequiredtoexpelthe~atercolumnisgreaterthan.thepressurenecessarytobringthesteammassflowthroughthequencher.Theexpulsionofthewatercolumn,isalsoclearfromthedifferenttimevariationsatP4.3andP4.4.Thepressureat8EVli3/798-58 PROPRIETARYmeasuringpointP4.3(abovethevatercolumn)risesmuchmoresteeplythanthepressureatmeasuringpointP4.4(insidethevatercolumn)Thedifferencebetveenthetvopressurosisthepressurewhichisnecessaryfortheaccelerationofthevatercolumn.Atthetimeoftheventclearing,thetwopressureshaveapproximatelyequalvalues.Butaftertheventclearingtheydifferagain,thistimeduetothedifferentpressurelossescausedbyflowresistancesinthepipe.8.521.1VentClearingPressuresfortheL~onLineThesteammassflowthroughtheSRVisapracticallylinearfunctionofthestagnation'pressure(reactorpressure).Sincethesteammassflowisoneofthemainparametersforthepressurebuild-upintheairregionofthedischargelineandthusfortheaccelerationofthewatercolumn,wewillplotthepressuresinthedischargelineasafunctionofreactorpressure.Thepressureinthebuffertank(P2.6)andnotthepressureinthesteamlinebeforetheSRVisusedasthereactorpressureforthetestssincethepressureinthebuffertankmorecloselysimulatestherepresentativestagnationpressureinthereactor.(seeFigure8-134).Todescribethedependenceoftheventclearingpressureonthereactorpressure,onlythosetestsforwhichtheinitialconditionsweresetandthusknovnexactlywereused.Thosearethetestsvithso-called<>cleanconditions".FromFigures8-135and8-136,itcanbeseenthatthemeasurementresultshavegoodreproducibilityforthetestswithcleanconditions.Thepressuresinthepipeincreasepracticallylinearlywithreactorpressure.Thefollowingtrendscanbeobserved:1)Aloweredwaterlevelinthedischargelineresultsinloverpressuresduringtheventclearing.2)Ahotpiperesultsinhigherpressuresduringtheventclearing.Thisisduetothesmallerpercentageofcondensationonthepipewall.3)Thepressure(atthetimeofventclearing)behindtheSBVisalwayshigherthantheventclearingpressureclosetothequencher.Thedifferenceisattributabletotheflowlossalongtheline.REV1,3/798-59 PROPRXETARY4)*Thepressure(atthetimeofventclearing)behindtheSRVincreases-withincreasingreactorpressure{orincreasingsteamflowratethroughthereliefvalve).Besidestheclean-conditiontests,thereisalargenumberofreal-conditiontestsandintervaltests.SincetheinitialconditionsinthemwererandomandwerenotvariedinacontrolLedmanner,themeasurementvaluesarescatteredoveramuchwiderbandthanintheclean-conditiontests.Hence,thesetestsarenotusablefortrendanalyses,butmaybeusedforverificationofmaximumspecificationvalues.Themeasuredmaximumvaluesare:PressurebehindtheSRV(atventclearingtime):19baratareactorpressureof72barVentclearingpressurebeforethequencher:14.5baratareactorpressureof72bar852.12VentClearingPressuresfortheShortLineFigures8-137and8-138showthemeasuredpipepressuresplottedagainstreactorpressureforcleanconditiontestswiththeshortdischargeline.Thesametrendsasseenwiththelonglineareseenhere.'Sincetheshortlinehasasmallerairvolumethanthelongline,whilethewatercolumntobeclearedandotherparametersremainthesame,thepressuresintheshortlinearehigherthanthoseinthelongline.Themeasuredmaximumvaluesare:Pressurebehindthereliefvalve{atventclearingtime):22baratareactorpressureof73barVentclearingpressurebeforequencher:18baratareactorpressureof73bar.85.213Tran~sositionoftheMeasurementValuestoSSESandComparisonwiththeDesicCndecificationTheverificationtestsinKarlsteinwererunwiththeactualgeometryofthereliefsystem,theactualSRV,andthehighestwaterlevelinthedischarge-line(6.2mabovecenterofquencher)thatoccursforSSES.Themeasuredventclearingtimesforthatwaterlevelandahighreactorpressure(69-81bar)wasbetween250and400msREV.1,3/798-60 PROP3IETARYFortheseventclearingtimes,theopeningtimeoftheSRV(measuredopeningtimes:29-60ms)hasnonoticableeffectontheventclearingpressure(seeFigure8-139).Hence,inregardtotheventclearingpressure,theonlyvariablewhosemaximumvalueforSSESwasnotcompletelycoveredwasthereactorpressure.Thefollowingextrapolationappliesforthat:a)PressurebehindthevalveatventclearingtimeTheMeasuredmaximumvalueforthelonglineis19baratareactorpressureof72barASlopeof25%isseeninfigure8-135.Extrapolatingto88bar,theresultis:Pox=23barforthelonglineTheMeasuredmaximumvaluefortheshortlineis22baratareactorpresssureof73barSlopeof25%isseeninfigure8-137.Extrapolatingto88bar,theresultis:Pax=26barfortheshortlineThedesignvaluegiveninSection4.1.2.1is550psi=37.93bar.TheKarlsteintestsdemonstratethatthedesignvalueisveryconservativefortheventclearingcase.b)Ventcleari~npressureThemeasuredmaximumvalueforthelong.lineis14.5barforreactorpressureof72barASlopeof12.5%isseeninfigure8-136.Extrapolatingtoareactorpressureof88barresultsinPmax=16.5barforthelongline.Themeasuredmaximumvaluefortheshortlineis18baratareactorpressureof73bar.ASlopeof12.57isseeninfigure8-138.Extrapolatingtoareactorpressureof88barresultsinPmax=20barfortheshortline.ThespecificationvaluegiveninSection4.1.1.2isPmax27barTheKarlsteintestsdemonstratethatthespecificationvaluefortheventclearingpressureisveryconservative.85.2.2PressuresDuri~ntheStationaryCondensationofSteamAboutonesecondaftertheopeningoftheSRV,theventclearingprocessiscompletedandthephaseofsationarysteamcondensationbegins.Inthisphase,thepressuresinthedischargelinearedeterminedbythesteammassflowandtheflowresistance.SincethesteamREV1,3/798-61 PROPRIETARYmass.flowisproportionaltothereactorpressure,hereagainwewillinvestigatethedependenceofthepipepressurestothereactorpressure.85.221Lo~nLineFigures8-140and8-141showthedependenceofthesteadystatepressureonthereactorpressure.Meseethattherelationcanberepresentedverywellbyastraightline.Asaresultofpipefriction,thestationarypressurebehindtheSRVhashighervaluesthanthepressuresjustbeforethequencher.Italsoexhibitsafasterincreasewithreactorpressure.Themeasuredmaximumvaluesare:17.5baratreactorpressureof72barforthepressurebehindtheSRV(P41)>l0baratreactorpressureof70barforthepressurebeforetheinlettothequencher(P4.4)852.22ShortLineFigures8-142and8-143showthedependenceofthesteadystatepressureonthereactcrpressure.Thebehaviorofthepressurebeforethequencher(P4.4)ispracticallyidenticalfortheshortlineandlongline.Thisisnotsurprising,sincethispressuredependsonlyontheflowresistanceofthequencher.ThepressuresbehindtheSRVarelowerthanthoseforthelongline,butdisplaythesameincrease,withreactorpressure.Thedifferentflowresistancesofthetwodischargelinesaremanifestedhere.Toclarifythiseffect,thevariationofthestationarypressureatthemeasuringpointsalongthedischargelineareplottedinFigure8-144fortheshortandlonglines.Theaveragepressureswereused,i.e.,thepressureswerereadofffromtheinterpolation1'inesat88bar(seeFigures8-190to8-143).Themeasuredmaximumvaluesfortheshortlineare:PressurebehindtheSRV(P4.1)16baratareactorpressureof72bar,and15baratareactorpressureof63barREV1,3/798-62 PROPRIETARYPresurebeforeinlettothequencher(P4.4)9.5baratareactorpressureof71bar,and9.0baratareactorpressureof65bar.85.2.2.3TranspositionoftheMeasurementValuestoSSFSandcomparisonwiththeDesignSpecificationAswasthecasewiththeventclearingpressure,theonlyvariablewhosemaximumvalueintheSSESwasnotcompletelycoveredbytheteststandwasthereactorpressure.Anextrapolationofthemeasuredmaximumvaluestoareactorpressureof88baryieldsthefollowingresults:a)LongLineThemeasuredmaximumvaluebehindtheSRVis17.5baratareactorpressureof72bar.ASlopeof22%isseeninfigure8-140Extrapolatingto88bar,theresultis:Pm~~--21barThemeasuredmaximumvaluebeforequencherinletis10baratareactorpressureof70bar.ASlopeof16%isseeninfigure8-141.Extrapolatingto88bar,theresultis:Pmmx=13bar.b)ShortLineThemeasuredm'aximumvaluebehindtheSRVis16baratareactorpressureof72barand15baratareactorpressureof63bar.ASlopeof22%isseeninfigure8-142.Extrapolatedto88bartheresultis:Pmax=19.6barand20.5bar,respectively.Themeasuredmaximumvaluebeforequencherinletis9.5baratareactorpressureof71barand9baratareactorpressureof65bar.ASlopeof16%isseeninfigure8-143.Extrapolatedto88bar,theresultis:P=12.5barand13.0bar,respectively.maxItcanbestatedthatthedesignvalueof550psi=37.93barforthestationarypressurebehindthevalveisveryconservative.85.23ExternalLoadsonthe~uencherandBottomSup2ortInthisSectionweshalldiscussthemeasurementresultswhichprovideinformationabouttheexternalloadsonthequencherandREV1,3/798-63 PROPRIETARYbottomsupport.ThemeasuringpointsprovidedforthatpurposeareshowninFigure8-13,andareasfollows:SG41/42SG43/4SG45/46SG47SG48Bendingatquencherarm1Bendingatquencherarm2BendingatthebottomsupportLongitudinalstrainatthebottomsupport'orsionatthebottomsupportStrainsweremeasuredatallmeasuringpoints.Themeasuredstrainswereusedtocalculatetheloadswhichproducedthestrains.Theloadsthuscalculatedarestaticequivalentloadswhichcontainhydraulicandalsostructural-dynamicaleffects.85.2.31VerticalForce8.5.23.11MeasurementoftheVerticalForceTomeasuretheverticalforce,twostrainconnectedinsuchawaythattheymeasureverticalforces.Thefollowingrelationexistsbetweenthegauges,SG4.7,werestrainsresultingfromloadandstrain:F~A~E~eBF=33~ekNBwhereA~.016m252F~2.06x10N/mmIfweinserteinpm/m,wethengettheverticalforceinkN.Thisequationwasusedtoconvert'hemeasuredstrainsintoverticalforces.8.5.2.3.1.2MeasuredVerticalForcesFigure8-145showsatypicalmeasurementtracefortheverticalforceItincreasesrapidlyduring'theexpulsionofthewatercolumnand,afterreachingthemaximumvalue,returnsquicklytozero.8523.1.21Lo~nLineTheverticalforceexhibitsastrongrelationshipwithventclearingpressureasshowninFigure8-146Thisholdstrueforalltests,eventhosewithrandominitialconditionssuchastherealconditionsandmultipleactuationtest.AsdiscussedinSection8.5.21.3,theventclearingpressureisinturninfluencedbythereactorpressure,initialwatercolumninthe'dischargeline,dischargelinetemperature,etc.andwasREV13/798-64 PROPRIETARYextrapolatedouttoamaximumreactorpressureof88bar.Therefore,themaximumverticalloadwillbeextrapolatedtothemaximumventclearingpressurefromSection8.5.2.1.3.Themeasuredmaximumvaluefortheverticalforceis:149kNata128barventclearingpressure.8.5.2.3.1.22ShortLineFigure8-147illustratesthedependenceoftheverticalforceontheventclearingpressure.Inprinciple,thesamediscussionasinSection8.5.2.3.1.2.1forthelonglineappliesherealso.Themeasuredmaximumvaluefortheverticalforceis:192kNata168barvent-clearingpressure.Theverticalforcesrelativetotheventclearingpressurearepracticallythesame.85.2.3.1.3TranspositionoftheMeasurementValuestoSSESAswasdiscussedpreviouslyfortheextrapolationoftheventclearingpressures,themeasurementvaluesfortheverticalforcecanalsobetransposeddirectlytotheplant.Forverificationofextremeconditionsintheplant,themeasurementvaluesareextrapolatedtoareactorpressureof88bar.Theextrapolationcanbeperformeddirectlyviatheventclearingpressure.8.5.23.1.31LongLineThemeasuredmaximumvaluewas:149kNata12.8barvent-clearingpressureSlope=13kN/bar(Figure8-146)AccordingtoSection8.5213,theextrapolatedvent-clearingpressureforthelonglinewas16barExtrapolationoftheverticalforceto16baryields:Fymax=190kN85.23.1..3.2ShortlineThemeasuredmaximumvaluewas:192kNat168barvent-clearingpressureSlope=13kN/bar(Figure8-147)AccordingtoSection8.5.2.1.3theextrapolatedventclearingpressurefortheshortlinewas20bar.REV1,3/798-65 PROPRIETARYExtrapolationoftheverticalforceto20baryields:FvmxInadditionFigure8-147,showsameasuredvalueof149kNata12barvent-clearingpressure.Thisleadstoamaximumextrapolatedverticalforceof:F~~y=252kN8523.13.3SummaryTheextrapolationofthemeasurementresultsfortheverticalforceyieldsama'ximumvalueof:F~~~~=,252kN>>InFigure4-11,thespecifiedverticalforceisgivenas860kN.Dnthebasisofthemeasurementresults,thespecificationvaluecanbeviewedasextremelyconservative,bothinthemaximumvalueandalsointheload-versus-timefunction.8.5.232'ore'ionalMoment8~2~3.2lMeas~nementoftheTorsionalMomentTomeasurethetorsionalmoment,twostraingauges(SG4.8-Figure8-13)wereconnectedinsuchawaythattheymeasurestrainresultingfromtorsionalmomentonly.AccordingtoReference41,thereisaverysimplerelationbetweenthetorsionorshearstrainandthemeasuredstrain,whenthestraingaugesaremountedata45oanglerelativetotheprincipalshearstressdirection.Qehave:YmsshearstrainThereforeisincethestraingaugesSG4.8weremountedata45oinclinationtotheverticalaxis,vehave:Gz=shearstressGmsshearmodulusY=2.eandr~Da2REV1,3/798-66 PROPBIETAHYIpg=torsionalmomentI=polarmomentofinertiaPr~outsideradiusofthetwistedcylindricalbarYrG'PQethusobtaintherelationbetweentorsionalmomentandmeasuredstrax.nTheshearmodulusisdefinedasG2(1+p)MithE=2.06x10sN/mm~andDap=poisson'sratioMeget:'p=0~3G~7.9x10'/mmeThepolarmomentofinertiaisdefinedas7f~D(1-D/D)4p32Therefore:I4.64x10mInsertingthevariousnumericalvalues,weget:0.41',InsertingE.inMm/m,thisequationgivesusthetorsionalmomentinkN-mHEVlg3/798-67 PROPRIETARYThisequationwasusedtoconvertthemeasuredstrainsatSG4.8intotorsionalmoments.Thetorsionalmomentsobtainedinthismannerrepresentstaticequivalentloads.85.232.2MeasuredTorsionalMomentsFigure8-148showsatypicalmeasurementtraceforthetorsionalmoments.Aftertheendoftheventclearingprocess,(approximately1secondafterteststart)theamplitudesofthemeasuredtorsionalmomentsareverysmallcomparedtothemaximumamplitudeduringtheventclearingprocessThereisafactorof6-7differencebetweenthetwoofthem.Themaximumamplitudeofthetorsionalmomentoccursmuchlaterthantheexpulsionofthewatercolumn.8.5.2.32.2.1L~onLineThetorsionalmomentatthebottomsupporthas,itsoriginonlyinunsymmetricalprocessesatthequencherduringtheventclearingandduringthetransitiontostationarycondensation.iFigure8-149showsthedependenceofthetorsionalmomentontheventclearingpressure.Sincetheventclearingpressureisadirectinfluencingparameter(seeSection8.5.2.3.1.2.1)wewillcorrelatethetorsionalmomentwiththatvalue.Thesharplypronouncedscatterbandisanindicationthatarandomprocessissuperimposedonthatdependence.Thatisexpressedbythefactthatthetorsionalmomentisbroughtaboutbyrandomunsymmetry.Themeasuredmaximumvalueofthetorsionalmomentis:M=55.8kN-mata14barvent-clearingpressure.TmaX$.523.2.g2ShortLinePigure8-150againshowsthedependencesofthetorsionalmomentontheventclearingpressure.Inprinciple,thesituationisthesameasintheprecedingSectionfo-thelongline.Themeasuredmaximumvalueofthetorsionalmomen'tis:39.2kN-mata18barvent-clearingpressure.8.523.2.3Tran~sositionoftheMeasurementValuestoSSESShentransposingthemeasurementresultstoSSES,weshallconsiderinaconservativemannertheloadcarriedbythedischargeline,whichintheteststandisconnectedrigidly(butREVli3/798-68 PROPRIETARYnotinaleaktightmanner)tothequencherandbottomsupportbymeansofweldbrackets(seeFigure8-13and8-14)incontrasttothefreemovingslidingjointatSSES.Todothat,wemaketheassumptionthatthedischargelineisfixedinatorsionresistingmanneratthefirstbendabovethequencher.Thatresultsinthefollowingpicture:DischargeLineQuencherBottomsupport////ThetorsionalmomentN~actsatthequencher.ThetorsionalmomentN~~wasmeasured.atthebottomsupport.ThedischargelinecarriesthetorsionalmomentM><.Therefore:+M2Promtheequalityoftherotation,weget:Therefore:"TVYGIpGTl11G~Ipl="T2'2'2G~Ip2Tl=P1~22T2P211Mehavethefollowingdimensions:r=0.1775mlar1g0.125m0.45m1r=0.162m2a2g~0.3.445m~11.313m2REV1,3//798-69 PROPRIETARYTherefore:-444.64.10mZ4.0,.10"m4Therefore:-26.6Tl4.640.16211.313M240'17750'4526.6TTlŽT2Tl(1+1)26.6M1.0376M1Thus,theloadtransmittedtothedischargelineislessthan4gofthattransmittedtothebottomsupport.If,withouttakingintoconsiderationthedischargeline,wefirstusePigures8-149and8-150asthebasisforanextrapolationofthemeasuredmaximumvaluestomaximumvent-clearingpressureforthecorrespondingdischargeline,thenwegetthefollowingmaximumvalues:a)longlineMr,~=598kN-mb)short.lineMr)~ax=43.2kN-mIfwenowconsiderthetorsioncarriedbythedischargeline,thenthisvalueisincreasedtoamaximumof:"ri~x=62kN-mThetorsionalmomentspecifiedin4.1.2.6fortheguenchersupportwas40kN-mtobeappliedasastepfunctionAtorsionalmomentstepfunctionappliedtoanundampedonemassREVl,3/798-70 PROPRIETARYoscillator(quencheractingasinertialmassandbottomsupportasatorsionalspring)correspondstoamaximumresponseof:M<<=2(40)kN-m=80kN-mSincethemaximumtorsionalmomentderivedfromtheKarlsteintestsisM<=62kN-m,thespecificationisconservative./8-5.2.33Bean~inncnenteattheguenchecAten85.2.33.1MeasurementoftheBendingMomentsIntheKarlsteintests,thebendingmomentsveremeasuredinthehorizontalplane(paralleltothetank'sbottom)andalsointheverticalplane,atbothofthequencnerarms.Toaccomplishthat,twostraingaugeseachvereconnectedinsuch.awaythattheymeasuredunsymmetricalstrainsresultingfromnormalstresses(unsymmetricalcomponent).Thefollowingstraingaugesweremountedforthatpurpose(seeFigure8-13:SG4.1)MomentsinverticaldirectionSG43)SG4.2)MomentsinhorizontaldirectonSG44)Thestraingaugesweremountedapproximately150mmfromtheweldbetweenthequencherarmandthecentralball.Thesectionmodulusofonequencherarmis:3W+D(1--)aQehave:'D~0.4064maD=0.3744ma=cE=M/WM=cEWThisleadstotheequationbetveenquantities:M=0.38-c8-71 PROPRIETARYThisgivesthebendingmomentinkN-m,ifcisinsertedinpm/m.Withthisequation,allthemeasuredbendingstrainswereconvertedintobendingmoments.Thebendingmomentsthuscalculatedarestaticequivalentloads.8.5.23.3.2MeasuredBendingMomentsFigure8-151showsatypicalmeasurementtraceofthemeasuredbendingmomentsatthequencherarms.Weseeclearlythatthemaximumvaluesoccurmuchlaterthantheclearingofthequencher.Theevaluationoftheindividualbendingmomentsrelatestothetotalresultantbendingmoment,ie.,thebendingmomentwhichactuallyloadsthe.quencherarm.Theresultantbendingmomentisobtainedbyusingtherelationship:M~'gM+M2x'esyzThebendingmomentsMgarereadoffatSG4.2and4;4.ThebendingmomentsMzarereadoffatSG4.1and43Theresultantbendingmomentsexhibitnodeterministicdependenceontheventclearingpressure,asshovn.inFigure8-152.Therefore,theresultantbendingmomentsonthequencherarmsmustbeconsideredasstatisticalvalues.Themeasuredmaximumvalueofthereultantbendingmomentis63kN-m.e8.52.33.3TranspositionoftheMeasurementResultsintotheWeldInSection4l.2.5,thebendingmomentsintheweldwerespecified.IntheKarlsteinteststand,thestraingaugesweremountedabout150mmfromtheweldinordernottomeasurelocalizedstressesduetotheweldandtheintersectionbetweentheballcentralbodyandthequencherarm.Availableexperienceindicatesthatthisdistanceissufficienttomeasureastressprofilewhichisindependentofshapefactors.Fromthespecifiedforceandmoment(Table4-10),weobtainforthedistancebetweentheweldandtheforceproducingthebendingmoment:lp~~Oo6551929Bytreatingthequencherarmasacantileverbeam,weobtainforthemaximumstressandthusforthemaximumbendingmoment:g0.655=Mg(0.655-0.15).REV~1>>3/798-72 PROPRIETARYM=bendingmomentintheveldBmaxM-=measuredbendingmomentBmaxTherefore:=1.297MBmaxBmeasThus,basedonthemeasuredmaximumresultantbendingmomentof62KN-m(seeSection852.3.3.2),weobtainthefollowingmaximumbendingmomentintheweld:'aximumresultantbendingmoment:81kN-m852.33.4~SecifiedStaticEguivalentLoadsAsalreadynotedabove,themeasuredbendingmomentsaretobeconsideredasstaticeguivalentloadsInSection4.1.2.5Table4-10,two'ontributionswerespecifiedwithrespecttothebendingmomentintheweld:a)astepfunctionhavingastepheightof19kN-mb)amaximumdifferentialpressurevhich,accordingtoSection4.1.3.7,is08barfromKKBtraceNo.35witha0.5multiplier.Thisresultsinamaximumdifferentialpressureof0.4bar.Thecontributionofthedifferentialpressureistobeviewedstatically,since,accordingtoSection413.5,thefreguencyofthedifferentialpressureisapproximately6Hz.Thebendingeigenfreguencyoftheguencherarmisontheorderof100Hz.Thecontributionofthedifferentialpressuretothebendingmomentintheweldisthus:11.4kN-mThecontributionofthestepfuncionistobevieweddynamically.Therefore,thesameconsiderationsareapplicableasthosemadeforthetorsionalmomentsinSection8.5.2.3.2.3.Accordingly,wehavethefollowingstaticeguivalentloads:ComponentinoneDirectionContributionfromstepfunction=2X19=38KN-mContributionfromdifferentialpressure=11.4KN-mTotal=49.4KN-mREVli3/79'-73 PROPRIETARYResultantMomentContributionfromstepfunction=38x~2=53.7KN-mContributionfromdifferentialpressure=11.4KN-mTotal=65.1KN-m8.5.23.3.5FvaluationoftheMeasurementResultsAsalreadymentionedinSection8.5.23.3.2,thebendingmo'mentsonthequencherarmaretobetreatedasstatisticalvalues.Figure8-153showsthefrequencydistributionofthemeasuredmaximumbendingmomentsineachtestsandtheresultingfrequencydisrihutionofthevaluestransposedintotheweld.Thefrequencydistributionsarebasedonthepeakmaximumvalueofeachindividualtest,whichweremeasuredeitheratSG4.1/4.2oratSG4.3/4.4.Thespecifiedstaticequivalentloads(seeSection8.5.2.3.3.4areintroducedfor7000responsesofthereliefvalve.Therefore,theloadsaretobeevaluatedinafatigueanalysis.ItfollowsfromFigure8-153thatthemeanvalueofthemeasuredmaximumvaluestransposedintotheweldis35kN-m.Exceptforthreecases,thespecifiedresultantbendingmomentsalsocoverthemaximummeasuredvalues.Thequencherisbeingevaluatedforthesemeasuredmaximumvalues.Itshouldbenotedthatboththespecifiedstationaryinternalquencherpressureof22.0barandtheresultingthermalloadof219~Cwerefoundtobeveryconservativewhencomparedtothemaximumextrapolatedvaluesof13.0barandtheresultingsaturatedsteamtemperatureof195~Cmeasuredduringthetests.(Section8.5.2.23).8.5.23.4BendingMomentsattheBottomS~ugort852.3.4.1MeasurementoftheBending.MomentsTomeasurethebendingmomentsathebottomsupport,twostraingaugescapableofmeasuringthebendingstrainsweremounted.In,themeasurementarrangement,thebendingstrainscouldbemeasuredintwomutuallyperpendiculardirections(seeFigure8-13).Thestrainsformomentsaboutthex-axisweremeasuredwiththestraingaugesSG4.5.Thestrainsformomentsaboutthey-axisweremeasuredwiththestraingaugeSG46.REVlg3/798-74 PHOPRIETARYThesectionmodulusofthebottomsupportis:D4W=-D(1--)332a4aW~1.307x10m-33Wehavea~E~@~M/WThisleadstotheequation:M~0.27~cThisequationgivesthebendingmomentinkN-m,ifcisinsertedinpm/m.Thisequationwasusedtoconvertallmeasuredbendingstrainsofthebottomsupportintobendingmoments.Thebendingmomentsthuscalculatedarestaticequivalentloads.8.52.3.4.2MeasuredBendingMomentsInFigure8-151,thebendingmomentsatthebottomsupportcanbeseenunderthetracesofthebendingmomentsatthequencherarms.Themaximumvaluesoccuratalatertimethantheventclearing.Buttheyoccuratthesametimeasthemaximumvaluesofthebendingstrainsatthequencherarms.Themaximumstrainresultingfromtorsiondoesnotoccuratthetimeofthemaximumbendingstrain(seeFigure8-151,SG4.8).Theevaluationofthebendingmomentsrelatestotheresultantbendingmoment,i.e.,thebendingmomentwhichactuallyloadsthebottomsupport.Theresultantbendingmomentisobtainedbyinterconnectingtheactualload-versus-timefunctionsoftheindividualcomponentsthroughtherelation:ThebendingmomentsMzarereadoffatSG4.5andthebendingmomentsM>atSG4.6Themaximumresultantbendingmomentwas54.5kN-mTheresultantbendingmomentsdisplaynodependenceontheventclearingpresure,asshowninFigure8-154.Hence,thesameconclusionsthatweredrawnforthebendingmomentsatthequencherarmsareapplicablehere,also.BEV.1,3/798-75 PBOPRIETABY8.5.234.3SpecifiedStaticEquivalentLoad.Asalreadymentioned,themeasuredbendingmomentsaretobeviewedasstaticequivalentloads.Thebendingmomentsatthebottomsupportareintroducedthroughthequencher.Section4.12.4andTable4-7specifyatransverseforceof44kNonthequencherwasusedasstepfunction.Inaddition,amaximumdifferentialpressureof0.4baronthequencherwasspecified.Thecontributionresultingfromthedifferentialpressureistobeviewedasastaticallyactingload.Itamountsto48kN.Note:Thedischargelineandthebottomsupportwerenotconsideredhere.Thepresssuredifferencewasformulatedonlyovertheprojectedareaofthequencher.Thespecificationthenyieldsthefollowingtransverseforcesonthequencher:Contributionfromstepfun'ction=2x44=88kNContributionfromdifferentialpressure=48kNTotal=136kNStraingaugesSG4.5andSG4.6weremountedapproximately0.5mbelowthecenterofthequencher.Transposedtothislocation,thespecificationyields:68kN-m85.23.44EvaluationoftheMeasurementResultsFigure8-155showsthefrequencydistributionofthemeasuredmaximumbendingmomentsatthebottomsupport.Themeasuredmaximumvaluesarealsocoveredbythespecification.Thus,theKarlsteintestshavedemonstratedthatthespecifiedtransverseforcesonthequenchercanbeviewedasveryconservative.85235ForcesontheQuencherIntheKarlsteinQuencherTests,onlybendingmomentswereabletobedeterminedforthequencheritself.InSection4.1.2,forcesandmomentsonthequencherwerespecified.Thespecifiedmomentswerecalculatedfromtheforces.Themeasuredmomentsarewithinthespecification.Therefore,wecanconcludethattheforcesarealsoverified.REVli3/798-76 PROPRIETARY85.23.6InfluenceofanAdgacentQuencherDuringtheclearingofthequencher,strongturbulencesandeddiesoftheexpelledandambientwaterdeveloparoundthedischargingquencher.Inparticular,aftertheventclearingthequencherissurroundedbyalargenumberofairbubbleswhichrepresentalocallycompressiblevolumeinthewater.Thisstate,whichformsaroundthedischargingq<<niche>ipreventseffectsfromtheblowdownofanadjacentquencherfrompenetratingtothequencherunderconsideration.Itisthereforeunderstandablethat,intheKMUinplanttestswithintheBrunsbuttelandPhilippsburgnuclearpowerplants,noincreaseoftheloadonthequencherandbottomsupportwasfoundfortheresponseofseveralquenchersincomparisontotheresponseofonequencher(Reference6).Aneffectofaloadononequencherduetothefiringofanadjacentquencheristobeobservedonlywhentheadjacentquencherblowsdownalone.Inthatcase,adetailedevaluationwasmadefortheBrunsbuttelblowdowntests(Reference38).Theresultoftheinvestigationwasthatthemeasuredloadsareenvelopedbyapressuredifferenceof0.2barappliedovertheadjacentinternalstructuresinthepoolatthequencherlevel,i.e.,alsooverthequencher.Amaximumpressuredifferenceof0.4baroverthequencherarmswasspecifiedforSSES.TheventclearingpressuresanddynamicpressuresinthewaterpoolobtainedforSSESfromtheKarlsteintestsareofthesameorderofmagnitudeasthecorrespondingmeasurementresultsinBrunsbuttel.Therefore,thespecifieddifferentialpressureof0.4baroverthequencherarmscanbeviewedasconservativelyenveloping.8.5237LoadsontheQuencherDuringSteamCondensationThemaximummechanicalandthermalloadsonthequencherduringthecondensationphaseoccurduringthephaseofintermittentcondensation.InSection4.1.2.7,theloadsresultingfromintermittentcondensationweretakenasthebasisforthefatiguedesignofthequencher.TheevaluationoftheloadsonthequencherduringsteamcondensationintheKarlsteinteststhereforerelatesprimarilytothephaseofintermittentcondensation.REV1,3/798-77 PROPRIETARY8.5.2.371Manifestation/ormsofIntermittentCondensationintheKarlsteinTestsAsdiscussedinSection8.1.3,thecondensationtestsvereperformedalongthelowerandupperboundarylinesoftheoperationfieldforwatertemperatures<30~Candalsoforwatertemperatures>590C.Inbothregions,theintermittentcondensationphaseoccursforverylowreactorpressures(approximatelybetween2and4bar).InSection84.2itisshownthatthemaximumvaluesforthedynamicpressuresinthevaterregionoccur.duringintermittentcondensationincoldvaterThesameistruealsofortheloadsonthequencher.Foritheevaluationandcomparisonwiththespecification,weusethemeasurementvaluesofthebendingmomentsatthequencherduringtheintermittentcondensationinthecoldpool.ThemeasurementvaluesaredocumentedinSection8.4.2.85.23.7.2IllustrationoftheMeasurementValuesThetimedurationoftheintermittentcondensationinthecoldpoolwasabout100seconds.Thetotalnumberofcondensationeventsatthequencherwas52.ThemaximummeasurementvaluesoccurredintheverticaldirectionatSG4.3.Thefrequencydistributionoftheresultantbendingmoments(SG4.'3/4.4)atthequencherarmisshowinFigure8-156.Themeanvalueofthemaximummeasurementvaluesofeacheventis11.8kN-m.Themaximummeasuredvaluewas66.5kN-m.Thefrequencydistributionoftheresultantbendingmoments(SG4.5/4.6)atthebottomsupportisshowninfigure8-158.Themeanvalueofthemeasurementvaluesis89kN-m.Themaximumvaluewasapproximately30kN-mThemeasuredmaximumvalueofthetorsionalmomentduringtheintermittentcondensationis6.2kN-m.8.52373EvaluationoftheMeasurementResultsfortheOuencherArmFigure8-157showsthefrequencydistributionoftheresultantbendingmoments,whichweretransformedfromthemeasuringpointintotheweld(seeSection8.5.2.3.3.3.Themeanvalueofthesebendingmomentsis15.2kN-m.Themaximumvalueis86kN-m.Themeasuredbendingmomentsrepresentstaticequivalentloads.,InSection4.1.2.7andTable4-12,avalueof25.4kN-mwasspecified.fortheequivalentloadfortheresultantbendingmomentintheweldduringintermittentcondensation.Theloadsspecifiedareformulatedforanoccurrencefrequencyof106.REV.li3/798-78 PROPRIETARYInthefatigueanalysis,themechanicalloadsrepresentonlyoneloadcomponent.Anotherpartofthefatigueloadingisproducedbythealternatingthermalloading.Theassumptionmadeinthespecificationwas106temperaturestepsfrom35~Cto133~Candfrom133~Cto35oC.Thelow-frequencyoscillationsofthepipe'sinternalpressuremeasuredatP4.4areusedasabasisforthemeasuredtemperaturealternation.Thesaturated-steamtemperaturesarethencorrelatedwiththosepressures.Thepressureoscillationshaveanoscillationfrequencyofabout0.5Hzandamaximumamplitudeof05baroverpressure=approx.2barabsolutepressure.Thispressureliesbelowthespecifiedvalueof3bar.Themeasuredmaximumpressureof2barcorrespondstoasaturated-steamtemperatureof120~C.AssumingthattheinflowingwaterinSSESisatatemperaturecfatleast35~C,thenthetemperaturestepis85C.Atemperaturestepof98~Cisassumedinthespecification,sothatthereisareserveof13~C.Themeasurementvaluesformingthebasisfortheevaluationandcomparisonwiththespecificationwereobservedonlyduringthephaseofintermittentcondensationwithcoldwaterinthetesttank.Aswiththeboundarypressuresinthetesttank(Section8.4.2),theloadsonthequencherwereconsiderablylowerduringtheintermittentcondensationphasewithwarmwaterthanduringintermittentcondensationwithcoldwater.Themeasuredmaximumbendingmomentduringthiscondensationphasewas(1kN-mrelativetotheweldseam.Inaddition,KMUinplanttestsintheBrunsbuttelnuclearpowerplantshowedthat,forapoolwatertemperatureofapproximately35~andabove,intermittentcondensationloadsonaquencherweresmaller.Thisindicatesthattheregionwh'ereintermittentcondensationloadsofanyconsequencecanbeexpectedislimitedtothatofverylowpooltemperatures(approximately25~C)andverylowsteammassflowsandthatheatingofthepoolasmallamountresultsinareductioninloading8.52.3.7.4EvaluationoftheMeasurementResultsfortheBottomS~uportAnimpulsivelyactingtransverseforceof17.5kNwasspecifiedonthequencherforintermittentcondensation.REVl~3/798-79 PROPRIETARYThedistancefromthemiddleofthequenchertothemeasuringpointforthebendingmomentsatthebottomsupportis0.5m,sothatthespecifiedbendingmomentwithrespecttothebottomsupportis:(175kNx2)x0.5m=17.5KNm(staticequivalentload)Themaximumresultantbendingmomentfromthetestsisapproximately30KN-m.l85.2375EvaluationoftheMeasuredTorsionalMomentsAnimpulsivelyactingtorsionalmomentof19kN-mwasspecifiedfortheintermittentcondensation.Thisstepfunctionyieldsatorsionalmomentof:38kN-masthestaticequivalentloadThespecifiedtorsionalmomentsconservativelyenvelopthemeasuredmaximumvalueof6.2kN-m.852.3.76EvaluationoftheMeasuredMaximumMomentsattheQuencherArmduringIntermittentCondensationAmaximumresultantbendingmomentof665kN-matthequencherarmwasmeasuredintheintermittentcondensationphase,whichresultsinamomentof86kN-mintheweld.Themeasuredmaximumvaluesoftheresultantbendingmomentsatthequencherarmduringintermittentcondensationareontheorderofmagnitudeofthemeasuredmaximumvlauesduringtheventclearingphase(Section8.5.2.3.3.2).Fortheventclearing,atemperaturedifferenceof184~Cwasspecified.Fortheintermittentcondensation,atemperaturedifferenceof98~Cwasspecified.Thetotalstressesloadingthequencherarmarecomposedofmechanicalandthermalstresses.Thethermalstressesaredistinctlylargerthanthemechanicalstresses.Themaximumresultantbendingmomentatthequencherarmforintermittentcondensationexceedthevaluespecifiedfortheventclearingbyabout40%.,However,theassociatedtemperaturejumpisonlyabouthalfaslargeasfortheventclearing.REV-1,3/798-80 PROPRIETARY853VerificationofSuppressionPoolBoundaryLoadSpecificationDuetoSRVActuationInSection4l.3,threepressuretimehistoriesarespecifiedasthebasisforthecontainmentanalysisduetoSRVactuation.ThethreetracesveretakenfromalargenumberofbottompressuretimehistoriesfromvariousKKBinplanttests.TheevaluationofthepressureoscillationmeasurementsintheKarlsteinventclearingtestswillthereforeconcentrateondemonstratingthatthepressuretimehistoriesspecifiedareenveloping.Accordingly,analysisandassessment.oftheindividualmeasuredpressuretimehistoriesisrestrictedtoaminimum.8.5.3.1EvaluationoftheLocalEffectsSeenatPressureTransducerP5.5AsshowninFigures8-10to8-12,thepressuretransducerP5.5ismountedontheconcretewalloppositethemiddleoftheholearrayonthequencherarm.About0.25secondsafterexpulsionofthewatercolumn,P5.5,incomparisonwiththeotherpressuretransducers,exhibitshigh-frequencypositivepressurepeakswhicharenotobservedattheneighboringpressuretransducers.Thiseffectisfromthelocalturbu1ences.Thesehighfrequencypressurepeakshaveasmallenergycontentsothattheirrangeofactionislimitedtotheimmediatevicinityofthepressuretransducer.'hefollowingTableshouldmakethisclearInthisTable,theratioofthemeasuredpressureamplitudesoftheneighboringpressuretransducers{P5.10andP5.4)tothepressuremaximumatP5.5isindicatedforalltestsvhichexhibitedamaximumpressureamplitude>1baratpressuretransducerP5.5.REV1,3/798-81 PROPRIETARYp5.lOPS'+TestP54P5.10P5.5P5.4/55P5.10/P5.5(bar)(bar)(bar)4165.1.710R1.720Rl920Rl1025125R20,60,551,00,450,4li00,730,551,70,450,4.lr01,00,651,730,550,61,00,85081,550,60,450,430,450,580~550,550,550,40,320,40,380,60,52FromthisTablewecanseethatthemeasurementvaluehasdecayedbyhalfatabout1mfromthemeasuringpointP5.5.ThecomparisonmeasurementpointsP5.4andP5.10areintheregionoforiginationoftheairbubbleoscillation,sothatnoattenuationeffectduetodistanceeffectscouldoccuratthatmeasuringpointTherefore,thesharpdecreaseofthepressureamplitudewhichismeasuredneverthelessshowsclearlythatthepressuremeasuredatpressuretransducerP5.5islimitedtoitslocalvicinity.AsfurtherverificationthatthiseffectislimitedtotheareaaroundpressuretransducerP5.5,acomparisonismadebetweenthepowerspectraldensitiesfromP5.5andthebottompressuretransducerP5.2.REV.1,,3/798-82 PROPRIETARYThefollowingtestsvereselected:Test11.1ThistestexhibitedthehighestpowerspectrumatthedominantfrequencyTest4.1.6ThistestexhibitedthehighestpressureamplitudeatP5.5forthelongdischargelineTest20.R1.10ThistestexhibitedthehighestpressureamplitudeatP55fortheshortdischargeline.Thecomparisoncanbesummarizedasfollovs:Atthedominantfrequency,thepowerdensitiesarethesamemagnitudeforthepressureoscillationsatthebottompressuretransducerP5.2andat.pressuretransducerP5.5.Thedifferencesatthehigherfrequenciesissignificant.Fortests4.1.6and20.R1.10thefrequencyspectrumofP5.5exhibitssignificantlyhigherpowerdensitiesathigherfrequenciesthanthecorrespondingfrequencyspectrumatpressuretransducerP52.Thissignificantfactorisnotnotedforthefrequencyspectrumoftest11.1(seeFigures8-159and8-160).Inthattest,thedifferencebetweenthemaximumpressureamplitudesforpressuretransducersP5.5andP5.2was013bar.ThepressureratioisP55/P52=0'8/065=12.Intest4.1.6,thedifferenceinthepowerdensitiesatthehigherfrequenciesisalreadymorestronglyevident(seeFigures8-161and8-162).Inthattest,thedifferencebetweenthemaximumpressureamplitudesforP5.5andP5.2was0.5bar.ThepressureratioisP5.5/P52=1/0.5=2.Thedifferenceinthepowerdensitiesatthehigherfrequenciesisquitestronglypronouncedintests20.Rl.10(seeFigures8-163and8-164).ThedifferenceinthemaximumpressureamplitudesforP5.5andP5.2was1.1barinthattest.ThepressureratioisP5.5/P5.2=1.73/0.63=2.75.Thepressuredifferencesorpressureratiosarenotdiscernibleinthepowerspectraforthedominantfrequencies,butareatthehigherfrequenciesFromthatwecanconcludethatthepressureoscillationwhichwasmeasuredatpressuretransducerP5-5hasapproximatelythesameamplitudeatthedominantfrequencyasthepressureoscillationsvhichweremeasuredelsewhereinthevicinityofthequencher,eg.,atP5.2Inaddition,higherfrequencypressureoscillationcomponentshavingahighamplitudeareoccasionallysuperimposedonthefundamentaloscillationinthepressureoscillationsatP55.Thehigherfrequencycomponents,vhichoccuratpressureREV1,3/798-83 PROPRIETARYtransducerP5.5,decayrapidlyintimeandspace,sothattheeffectofthehighfrequencypressureoscillationsremainslimitedtotheimmediatevicinityofmeasuringlocationP5.5Therefore,asstatedbefore,themeasurementresultsforthedynamicpressuresatP5.5representlocaleventshavingnoglobaleffectonthecontainment.WewillthereforenotconsiderthepositivepressuremeasurementsatP5.5whenverifyingthedesignspecificationfortheoverallcontainmentanalysistheresultsfromthisgageareincludedfortheverificationoftheloadingsonthecolumns.8.5.3.2Verificationofthe~SecifiedPressureA~mlitudesandVerticalPressureProfilesafterVentClearingThemeasuredpeakpressureamplitudesforthe125ventclearingtestsaretabulatedinTables8.9and810.Section8.4.1alsopresentsanumberofFigures(8.27to8.34)whichshowthatthepressureamplitudesmeasuredinthetestshadnosignificantdependenceontheinitialreactorpressure.Therefore,nomodificationtothemeasuredpressureswillbemadetoaccountfordifferencesinthereactorpressurebetweenSSESandtheKarlsteinteststand.Inaddition,asexplainedintheprevioussection,thepositivepressuremeasurementsaP5.5willnotbeconsideredwhenverifyingthedesignspecificationfortheoverallcontainmentanalysis.8532.1OverpressuresThemaximumoverpressureamplitudemeasuredontheboundaryoftheKarlsteintesttankwas1.0barThatpressurewasmeasuredattheconcretewall(p5.4)intest20.R1.10.Amaximumpressureamplitudeofl.2barisspecifiedinsection4.1.3(KKBPressureTraceNo.35withthe1.5multiplier).Themaximumspecifiedoverpressureamplitudeof1.2bar.evelopsthemeasuredmaximumoverpresureamplitudeof10bar.8.5.32.1.1VerticalPressureProfileItcanbeassumedthatthemaximumdynamicpressurevilloccurinaspherewhichsurroundsthequencherandhasapproximatelytheradiusofaquencherarm,(5'-0").Atsomedistancefromit,themaximumvaluewillbeattenuatedinaccordancewithadistancelaw.Foraninfinitewaterspace,the1/Rlawisapplicableforthedecreaseofthepressurewithdistancefromthesource.Thatlawappliesinalldirections,i.e,intheverticaldirectionalso.Thevalidityofthe1/Rlawisbasedontheassumptionofastationary(i.e.,fixedposition)oscillatingbubbleintheinfinitewaterspace.Thatidealcasedoesnotholdfortheclearingofthereliefsystem.Alreadyshortlyaftertheexpulsionoftheair-steammixture,BEV1,3/798-84 PROPRIETARYsmallairparticlesmovetothesurfaceofthepoolbecauseofbuoyancy.Evenmoreimportant,however,isthefactthatthewatersurfaceandthetankboundarysurfacesinfluencethedistancelawandthatthepressureamplitudemustvanishatthewatersurfaceitself.Accordingly,specifiedin6e0(183m)<hatheight,surface.apressureprofileintheverticaldirectionisSection4.1.3.4providingforaconstant'pressureatabovethesuppressionpoolsbottomand,startingata,lineardecreaseofpressureuptothewaterFigure8-165showsthatthemaximumspecifiedpressuredistributionveryconservativelyenvelopsthemeasuredmaximumpressureamplitudes.Theconservativenessbecomesclearlyevidentif,basedonthemeasuredmaximumvalueofwallpressureamplitudeof1baratpressuretransducerP5.4,weassumealineardecreaseofpressurefromthatmeasuringpointtothewatersurface.Thatassumedlinearpressuredecrease(depictedinFigure8-165byadashedline)alsoenvelopsthemaximumpressureamplitudesmeasuredintheverticaldirection.Incomparisonwiththeassumedlinearpressuredecreaseandthespecifiedpressuredistribution,theconservativenessofthespecificationbecomesobvious.~' 3-~2..2VerticalPressureProfileIucluainulocalEffactsatP5.5Fortheevaluationoftheunpertubedpressuredistributionintheverticaldireciton,themeasuringpointP5.5wasomitted,eventhoughitliesinadirectlinewiththepressuretransducersP5.4,P5.6andP5.7.BecauseofthelocaleffectforP5.5,aseparateanalysisshallbeperformedhere.ThatanalysisstartswithanestimationoftheverticalzoneofinfluenceassociatedwiththepressurepeakmeasuredatP5.5.Thelateralholesinthequencherarmsextendoverananglerangeof72~oneachside.Theholesaredrilledradially,sothatinfirstapproximationwecanassumeasourceflowoftheemergingfluid.Thehigh-frequencypressurepeakatP5.5occursatamuchlatertimethantheventclearing.Itcanbesupposedthatatthattimethereisasteam-airmixtureflowingoutofthequencher.Thesteam-airjetsemergingfromtheholeshaveahighdegreeofturbulence.Thus,theedgesareverysoonmixedwiththesurroundingwater.Furthermore,theemergingsteamiscondensedimmediatelyandtheexpelledairiscooleddownquickly,sothattheexpelledcompactvolumeisreducedrapidly.Thereforetoestimatetherangeofaction,itisassumedthatthesourceflowactsoverameananglerangeof8=e/2=720/236o.ThetotalrangeofactionisthenREVli3/798-85 PROPRIETARYb=xtan36~x=1.575m{distancefromcenterlineofb=1.14mquencherarmtoconcretewall)Thisrangeofactionof'1.14misdividedintoequalpartsaboveandbelowthemeasuringpointP5.5,sothatweobtainarangeofactionof10.57mrelativetothemeasurementlocationBasedonthisrangeofactinthemeasuredverticalpressuredistributionconsideringthelocaleffectiscomparedwiththespecifiedpressuredistributioninFigure8-166.ThebasepointsofthepressureelevationatP5.5wereplacedonthestraightlineofthelinearpressuredropsymmetricallywithrespecttothequencher'scenterplane.FromFigure8-166itcanbeseenthatthemaximumspecifiedpressuredistributionresultsinalargerresultantforceonthecontainmentboundaryandcolumnsthandoesthemeasuredpressuredistributionincludingconsiderationofthelocaleffectThismeansthattheoverallspecifiedpressuredistributrionintheverticaldirectionalsoenvelopesthelocalpressureelevationatp5.5.8.5322Unde~rressuresThemaximumunderpressureamplitudemeasuredontheboundaryofKarlsteintesttankwas-0.68bar.Thatpressurewasmeasuredatheconcretewall.{P5.10)intest25.R2.Amaximumunderpressureamplitudeof-0.56barisspecifiedinSection4.1.3{KKBPressureTraceNo.76withthe1.5multiplier).The'nextlargestunderpressurerecordedduringtest25.R2was-050bar.Thenextlargestunderpressurerecordedanywhereduringtheventclearingtestswas-058baratP5.2intest25.1.Exceptforthetwomeasurementvaluescalledoutaboveallothermeasuredunderpressureswerehounded.bythemaximumspecifiedvalueof-0.56har.8.5.322.1VerticalPressureProfileFigure8.167showsaplotofthemaximumspecifiedunderpressuredistributionandthemaximummeasuredunderpressurevaluesfortheKarlsteintests.Itcanheseenthat,exceptfortheonevalueatP510fortest2S.R2,themaximumspecifiedpressuredistributionenvelopsthemaximummeasuredpressureamplitudes.REV.1,3/798-86 PROPRIETARYInaddition,forSSES,themostunfavorableboundaryconditioninthiscomparisonisthelowliquidlevelof22ft=6.70minthesuppressionpool.ThehydrostaticpressuredistributionwithrespecttothatliquidlevelisindicatedbyadashedlineinFigure8-167.Thecomparisonofthemeasuredworstunderpressuredistributionviththehydrostaticwaterloadresultingfromtheworstboundaryconditionforthiscomparison(lowestwaterlevelinthesuppressionpool)showsthatthecompressiveforcesfromthewaterloadandthetensileforcesfromtheunderpressuredistribution=maintaintheequilibrium."Thus,theKarlsteintestshave,inaddition,demonstratedthattheblowdownoftheSSESreliefsystemviththequencherdoesnotresultinanyresultanttensileforcesonthesteelliner,evenfortheworstpossiblesuperposition.85.33VerificationofthePressureTimeHistoriesUsedfortheSSESContainmentAnalysisXnordertoverifythatthepressuretimehistoriesusedfortheSSESdynamicanalysisduetoSRVactuationarebounding,thePowerSpectralDensities(PSDs)ofthespecifiedtimehistories(withtheappropriateamplitudeincreaseandfrequencyrangefromSection4.1.3)arecomparedwiththePSD'softheappropriatetimehistoriesrecordedintheKarlsteintesttankandtransposedtotheSSES"suppressionpool.Statementsconcerningtheclearingofparallelquenchersarebasedontheunrealisticandextremelyconservativeassumptionthattheexpelled,airbubblesareequallylargeandoscillateinphase.Aquantificationofthatconservativenessisnotgiven.Mevillfirstdiscussandverifythetheorytobeusedtotransposetheoscillationfrequenciesmeasuredinthetesttanktothesuppressionpool.Then,theappropriatemultipliersforthisfrequencytranspositionwillbeestablished.Adiscussionisalsoprovidedfortransposingthemeasuredpressureamplitudestothesuppressionpool.Finally,theactualverificationispresented.85.33.1Tr~ansositionmethodfortheOscillationFrequencyThetheoreticalbasisforthetranspositionofthepressuretimehistoriesmeasuredintheKarlsteinteststotheSSESsuppressionpoolisprovidedbytheKMUcomputercodesVELPOTandKOVIBlAByusingthetestresultsfromthePPGLquenchertestsinKarlstein,theGKMquenchertests,andthenon-nuclearhottestsintheBrunsbuttelnuclearpowerplant(KKBhottests),weshallfirstconfirmexperimentallythecorrectnessofthetranspositionBEV1,3/798-87 PROPRIETARYtheory.Thatisfollowedbyacalculationofthefrequenciesforthefollowingthreeblowdowncases:(1)Simultaneousblowdownofall16quenchers(2)Simultaneousblowdownofthe6quenchersrelatedtotheautomaticdepressurizationsystem(ADS)(3)BlowdownofoneouterquencherForeachcase,acomparisonofthetheoreticallycalculatedfrequencieswiththefrequenciesmeasuredintheteststand)providesanumber(frequencymultiplier)bywhichafrequencymeasuredintheteststandmustbemultipliedinordertogetthecorrespondingfrequencyintheSSESsuppressionpoolAfactorfortheinfluenceofthesuppressionpooloverpressureisalsodeterminedinthesameway.Thecorrespondingmeasuredpressuretimehistoryistransposedtotheplantbydividingbythisfactor853.3.1.1CalculationofMeasuredOscillation~r~cruencies85.33.1.11PPGLTestsatKarlsteinSinceitwasfoundthatFluid-StructureInteractionintheKarlsteintesttankhasnosignificantinfluenceonthemeasuredpressuretimehistories,itissufficienttocarryouttheanalysisforarigidtank.Thecomparisonofcalculatedandmeasuredoscillationfrequencieswillbebasedon'theassumptionofequalbubblevolumes.ThemeasuredoscillationfrequenciesaretakenfromTables8.9and8.10.Theassociatedbubblevolumeswerecalculatedfromthetestdata,usingtheformula:pp-piieeee'e)]~eo>[P-cP(7T[p-P(T'satpool)]TpipepipePpipePsatCTpoolpipefreepipevolume(ms)pressureinpipe(bar)hydrostaticpressureatthequencherlocation(bar)saturationsteampressure(bar)relativehumidity(s=1at1005)watertemperature(oC)meantemperatureinpipe(oC)TheaveragingofthetemperatureinthepipeisperformedbyusingtheformulaEi1NpipeiwherethepipewasdividedintoNequalsections.ThetemperatureTintheithsectionwasobtainedbyinterpolationbetweenthemeasuredtemperatures.REV.li3/798-88 PROPRIETARYThecomparisonbetweenthemeasuredandcalculatedbubblefrequencyisshowninFigures8-168and8-169inwhichthebubblepulsationfrequencyisplottedversustheequilibriumvolumeatstaticpressure.PorthemeasurementpointsinFigure8-168itwasassumedthatdryairwasinthepipepriortotheteststart,whilewetair(100%humidity)wasassumedinFigure8-169.Ingeneral,goodagreementisfoundbetweenthetheoryandmeasuredfrequency.However,wecannotoverlookthefactthatthemeasuredfrequenciesinfigure8-168)arehigherthanthecalculatedones,especiallyforsmallbubblevolumes.Thismayberelatedtothefactthattheactivevolumeofairunderwaterisactuallysmallerthanthevolumefoundfordryairfromthetestdata.Thisishintedatbythecalculationofthebubblevolumeundertheassumptionof100%humidityinthepipe.Therethemeasurementpointsareclosertothecalculatedcurve(Pigure8.169).Inordertokeeptheuncertaintiesassociatedwithsucheffectsassmallaspossible,onlytestsforwhichtheinitialpipetemperaturewasbelow700Cwerechosenforthecomparisonwiththetheoreticalcase.8.5.3.3.1.1.2GKMMode~luencherTestsAnothersorceusedtoverifythetheoryisofferedbytheGKNquenchertests(Ref1).Sincethepipetemperaturestherewereinthevicinityof300Corbelow,uncertaintiesinthebubble.volumeunderwateraredistinctlysmallerthanintheKarlsteintests.Inaddition,theGKMtestswerealsorunwithbackpressureinthesuppressionchamber,sothatinformationderivedfromthecomputercodesforblowdownofthequencherduringaloss-of-coolantaccidentcanalsobeverified.TheresultscanbefoundinFigures8-.170and8-171.Figure8-170showsthecalculatedandmeasureddependenceofthepulsationfrequencyonthebubblevolumeforvarioussubmergences(2m,4mand6m)withatmosphericpressureinthesuppressionchamber.ThetheoryandmeasuredfrequencyagreeevenbetterherethanintheKarlsteinquenchertests.Thisisprobablyduetothefactthatthebubblevolumesdeterminedfromthemeasurementvalueshaveamuchsmallerscatterduetothelowtemperaturesinthepipe.TheinfluenceofbackpressureonthepulsationfrequencyisshowninFigure8-171.Hereagain,thetheoryisverifiedbythetestdata.85.33113KKBHotTestsInordertodemonstratethecorrectnessofthetheoryforin-plantconditionsalso,calculationswereperformedfortheblowdowntestswithonevalveinthenon-nuclearhottestsintheBrunsbuttelBMRplant(Ref.3).Pigure8-172showstheresults.Theagreementbetweenthecalculatedandmeasured'requencyissimilartothatintheKarlsteintests.Thesameistrueforthescatterrangeofthemeasurementvalues.Sincethepipetemperatureherewasatabout90OC,alargerscatteractuallyREYlt3/798-89 PROPRIETARYwouldhavebeenexpected,butdidnotoccurbecausethepipewascarefullyflushedwithairpriortothebeginningofthesetests.8.5.3.3.1.1.4ConclusionfromtheFrequencyCalculationsThetestcalculationsdescribedaboveshowthatthetheory{VELPOTandKOVIB1Acomputerprograms)describesthemeasuredfrequenciesnotonlyinonespecialcase,butalsoforabroadrangeofgeometriesandbackpressure:(1)Thesizeofthewaterspacevariesfromapproximately7m>(GKN)toapproximately23m~(testtankatKarlstein)toapproximately400m~(suppressionchamberinBrunsbuttelnuclearpowerplant).(2)Thequenchersubmergencerangedfromapproximately2mto6m.(3)Thebubbleequilibrium.volumevariedbetweenapproximately015m~to37m~.(4)Thesuppressionchamberpressurevariedfrom1barto3bar.(5)Thewatertemperatureinthesuppressionpoolvariedbetweenapproximatley16~Cto800C.Thus,thetheorycanbeconsideredverifiedandcanbeusedtotransposethepulsationfrequenciesmeasuredintheKarlsteinteststandtotheSSESsuppressionpool.8.5.33.2Nuit~iliersforConversionoftheBubbleFrequenciesfromtheTestStandtoSSESUsingtheVELPOTandKOVIBlAcomputercodes,thefollowingthreeblowdowncasesareanalyzed:(1)'Simultaneousblowdownofall16quenchers(2)SimultaneousblowdownofthequenchersA,B,G,K,M,PwhichareincludedintheADS{3)Blowdownofonequencher(quencherB)TheresultsareillustratedinFigure8-173whichshowsthepulsationfrequencyasafunctionofbubblevolume(bubbleinhydrostaticequilibrium).Thebehaviorofthefrequencycurveforthe16-quenchercaseintheplantispracticallythesameasfortheteststand(Figure168),therebyconfirmingonceagainthesuitabilityoftheteststandgeometrythatwaschosen.Inthecaseofthe6quenchersintheADScase,thefrequenciesarehigherduetothelargersinglecellcorrespondingtothesmallerREVli3/79~8-90 PROPRIETARYhydrodynamicbubblemass.Theyareevenhigherinthecaseofonequencher.BasedontheresultsshowninFigures8-168and.8-173,asimpleformulacanbegivenforconvertingfromthemeasuredbubblefrequenciestothesefrequenciesfoundintheplantbyasking:Bywhatfactort"multiplier")mustabubblefrequencymeasuredintheteststandbemultipliedtogetacorrespondingfrequencyintheplant'?ThismultiplierisplottedinFigure8-174versusthe(measured)startingfrequency.Thus,wehave:v=f(v).vplantvtest'estinwhichthemuliiplierfforagiveninitialfreguencycanhereadofffromPigure8-173.ThegraphinFigure8-173isapplicableonlyforcaseswithapressureof1barinthesuppressionpoolairspace.However,theblowdownfortheADScaseduringaloss-of-coolantaccidentisassociatedwithasuppressonpooloverpressure.p~>lbarAnadditionalmultiplierfpKK(pKK)isnecessaryforsuchcases,sothatthefrequencyconversionmustbewritteninamoregeneralmanner:V=f(P).S(V).VplantPkk'testtestkkThemultiplierfpKK(pKK')canbetakenfromFigure8-175.Forasuppressionchamberpressureof7bar,ithastheavalueof1,asitmustbe.Themultipliersforthefrequencyalsofixthemultipliersfortheoscillationperiodwhentransposingthepressuretimehistoriesmeasuredintheteststandtotheplant:ttesttlPkkvtestkkI85.33.3TranspositionMethodforthePressureA~mlitudesAsalreadydescribedindetailinSection8.51,theteststandwassodesignedandthe.pressuretransducersweresoarrangedthatthemeasuredpressureamplitudescanbetransposedtotheplantwithoutchangeCorrespondingly,a1:1transpositionismade.Becauseofitsobviousconservativeness,sucha1:1REV1,3/798-91 PROPRIETARYaiplitudetranspositionofferstheadvantagethatmoreexactquantitativeproofsdonothavetobeprovided.Themostsignificantconservativefeaturesarethefolloving:(1)Inblowdowncase'swithseveralquenchers,itisassumedthatallbubblesareequallylargeandoscillateinphase.Deviationsfromthisassumption{suchasactuallyoccurintheplant)resultonlyinloverpressureamplitudes.(2)Blowdowncaseswithlessthan16quenchersareassignedthesamepessureamplitudeasthe16-quenchercase.Inreality,suchcaseshavealoweramplitudeduetothegeometry(largersinglecell).Theconservativenessdescribedin(1)hasnotyetbeenprovenexperimentallyinanyquenchertests,butitisalreadyobviousfromatheoreticalviewpoint,sinceatime-shiftedsuperpositionoftvotemporalmaximaalwaysyieldssmallervaluesthananadditionofthemaximumvalues.Concerningtheconservativenessof(2),thereareanumberqualitativeindicationsfromtheKarlsteinteststhemselves,fromcorrespondingmodelstudiesattheKarlsteinmodelteststand(Ref.1),andfromcalculationswiththeVELPOTandKOVIB1Aprograms.Theinformationobtainedfromallthreeoftheseinvestigationsshallbedescribedinthefollowingsections.Inaddition,wewillalsoexaminewhethertheconservativefeaturesareaffectedbyapossiblebackpressureinthesuppressionpoolairspace.85.33.31PPGL~uencherTestsatKarlsteinIndicationsoftheconservativenessdiscussedin(2)aboveareobtainedfromtheKarlsteintestsonthebasisofFigure8-176vhichillustratesthemeasuredrelationshipbetweenexcitation(relativeamplitude)andpressure-oscillationfrequencyfortheKarlsteintests.Thefrequencyanalysisforeachpressuretimehistoryhasatleasttwomaximaofthepowerdensity.Onepowerdensitymaximumliesatlowfrequenciesandtheotheratsomevhathigherfrequencies.Thereisafactorofapproximatelytwobetweenthetvofreqeuncies.Thefirstpeakofthepowerdensity(lowfrequency).isalwayslargerthanthesecondpeakofthepowerdensity(higherfrequency).Accordingly,thelovfrequencyisalvaysdesignatedasthedominantfrequencyForpressuretransducerP510,thepowerdensitiesofallanalyzedtestsareevaluatedinFigure8-176.DifferentanalysistimesvereselectedfortestshavingdifferentpressureoscillationfrequenciesThetimevassochosenthatREV.1,3/798-92 PROPRIETARYapproximatelythesameoscillationperiodscouldalvaysbeevaluated.Thefollowinganalysistimeswereselectedfortheevaluation:3HzTime:5HzTime:9HzTime:0-1.8seconds0-1.3seconds0-0.6secondsTheareabeneaththefrequencyspectrumvasdeterminedandthenthesquarerootofthatnumericalvaluewastaken.Thatresultsinvalueshavingthedimension>>har>>.Thosenumericalvalueswerenormalizedtothemaximumvalue.Theresultsarethen"relativepressures>>withrespecttothecalculatedmaximumpressurefromthefrequencyspectra.Sincenodominantfrequencieshigherthan6.5HzweremeasuredintheKarlsteintests,thesecondpeakswerealsousedtoevaluatethehigherfrequencies.Hence,thepowerdensitiesofboththedominantfrequencyandthenexthigherfrequencyareevaluatedinFigure8-176.Basedonanempiricalevaluation,it.followsfromFigure8-176thatthepressureoscillationsvithhigherfrequencieshavesmallerenergycontentthanthepressureoscillationswithloverfrequencies.Znaddition,asshowninFigure8-169,thehubblefrequencyincreaseswithdecreasinghubblevolume.Butdecreasingbubblevolumewithconstantsingle-cellsizemeans,accordingtothelawsofsimilarity,thesamethingasincreasingthecellsizewithconstantbubblevolumeTherefore,fromtheKarlsteintestdata,itcanbesaidthatthepressureamplitudesdecreasewithincreasingcellsize.8.5.33.32KM~UuencherTestsintheModelTestStandin~Ka1steinDuringthedevelopmentoftheKWUquencher,testsvereperformed.toexaminetheinfluenceofthesizeofthewaterspace(specifically:freewatersurface)inthemodelteststandinKarlstein(Ref.1).TheresultsareillustratedinFigure8-177,whichvastakenfromRefence1.Itshowsdirectlyhovthebottompressureamplitudesdecreasewithincreasingsizeofthewaterspace(singlecell).8.53.3.33AnalyticalCalculationsTheconservativenessdescribedin(2)aboveisalsoconfirmedfromresultsofcalculationswiththeVELPOTandKOVIB1AREV.1,3/798-93 PROPRIETARYprograms.Asforthefrequencyconversion,appropriatemultiplierscanbedeterminedalsofortheconversionofthepressureamplitudesfromtheteststandtotheplant.Theydependontheinfluenceofthewaterspaceonthestationaryvelocitypotential(spatialpressuredistributionnormalizedtounitsourcestrength)andonthehydrodynamicsourcestrengthassociatedwiththebubbledynamics.Thesourcestrengthitselfisdependentinturnonthepressureinthebubble,.whichisdeterminedbytheinterplay-of.bubblevolumeandairsupplyintothebubble.Sincetheairsupplyvariesaccordingtothedifferentoperatingconditionsduringtheblowdown,onlyaconservativeestimatecanbegivenwithintheframeworkofthepresentinvestigationsT'econversion.fromteststandtotheplantforonequenchermayserveasanexamplehere.Meobtainforthebottompressurebeneaththequencher:P(1quencher)<0.7Pplanttestasuppervalue.8.5.333.4InfluenceofBackgressureonthePressureAmplitudesAsforthebubbleoscillationfrequency,thequestionoftheeffectofbackpressureinthesuppressionpoolairspacemustbeinvestigated.Figure8-178showsthebottompressureamplitudesmeasuredintheGKNmodelquenchertestsforasuppressionpoolairspacepressuresof1and3barAscanbeseen,thepressureamplitudesdonotdependonthesuppressionpoolairspacepressure.8.533.4VerificationofDes~in~SecificationInthetranspositionofthepressureoscillationsmeasuredinKarlsteintotheSSES,theextremelyconservativeassumptionthatthesamepressuretimehistoriesareactingatallquencherssimultaneouslyisused.Differencesinthepressuretimehistoriesoriginatingfromthedifferentdischargelinesareneglected.Therefore,eachmeasuredpressureoscillationintheKarlsteinventclearingtestsisarepresentativecontainment,loadforallloadcases:symmetricalloadcase(simultaneousresponseofall16SRV'sunsymmetricalloadcase(responseofoneorthreeadjacentSRV'sautomaticdepressurizationinloss-of-coolantaccidentREV.1,3/798-94 PROPREETARYAtranspositionofthemeasurementresultstotheplantisperformedfortheseloadcases.TheKarlsteintesttankformsaconservativesinglecell.Therefore,conservativeenvelopingpressureamplitudesweremeasuredinthatteststand.Mhentransposingthepressureoscillationsfromthesinglecelltotheplant,thereisanincreaseofthepressureoscillationfrequenciesasdiscussedinSection8.5.3.3.2.Asstatedpreviously,theincreaseofthepressureoscillationfrequenciesisaccompaniedbyadecreaseoftheamplitudes.Thedecreaseoftheamplitudesisneglectedforthisevaluation,Theamplitudesofthemeasuredpressureoscillationsremainconstantforallfrequencies.Thatisanadditionalconservativefeature,asalreadydiscussedinSection8533385.33.4.1~foe~nencAnalysesofSelectedTestsThepressuretimehistoriesforselectedKarlsteintestsareillustratedinFigures8-41to8-65Thefreqeuncyanalyseswerecarriedout.withtheFourierAnalyzer5451madebyHewlettPackard.Thefrequencyanalysesweregeneratedaspowerspectraldensities.Thefrequenciesatwhichastructureisexcitedintooscillationcanbereadofffromthepowerspectraldensities.FreqeuncyanalyseswereperformedforpressuretransducersP5.2,P5.4,P5.5,andP5.10andforthefollowingtests:4el.l,4.16,12.1,llel,19R27~20Rl1,20eRle10,2lel~21.2,25.R2Pressureoscillationsatboththewallandthebottomareconsideredinthefreqeuncyanalyses.AlsoconsideredwasthefrequencyanalysisforpressuretransducerP5.5,whichshowsthefrlocaleffectThelimitationofthemeasuredfrequenciesofthepressureoscillationswasdeterminativeinselectingtheteststobeanalyzed.Thetestsselectedwerethosewhichexhibitedpressureamplitudes>0.3barbothatlowfrequencyandalsoathigherfrequencies.ThefrequencyspectraforseveralKarlsteintestsareillustratedinFigures8-179to8-182forpressuretransducersP5.10andp5.4.ThefrequencyspectrafortwotestswiththelongdischargelineandloweredwaterlevelareshowninFigure8-179.TheprincipalREV.1,3g798-95 PROPRIETARYfrequencyofthepressureoscillationsisat2-.3Hzforthesetests.TheyarethelowestpressureoscillationfrequenciesthatveremeasuredintheKarlsteintests.Figure8-180showsthedifferenceinthepressureoscillationfrequenciesfromclean-conditionteststoreal-conditionand/ormultiple-actuationtestsforthelonglineThepressureoscillationshaveaprincipalfreqeuncyof3.5Hzintest4.1.1(cleancondition)and5Hzintest4.1.6{realcondition)Fortheshortdischargeline,thefrequencyshiftsfromcleantorealconditionareillustratedinFigure8-181fortests21.1and21.2.Theresultfortheshortlineis:cleancondition:pressureoscillationfrequency5Hzrealcondition:pressureoscillationfrequency6.5HzThefollowingcanbesaidaboutthemeasuredgrin~cialfrequenciesfortheKarlsteintests:Thelowestpressureoscillationfrequencyvasmeasuredinthetestswiththelonglineandadischargelinewaterlevelloweredto2.5mabovethemiddleofthequencher.Itwas2.0-3Hz.2)Fortheclean-conditiontests,pressureoscillationfrequenciesof3.5-4Hzweremeasuredwiththelongdischargeline.3)Fortheclean-conditiontests,pressureoscillationfrequenciesof4.5-5Hzweremeasuredwiththeshortdischarge1ine.4)ThehighestfrequencyfortheKarlsteintestsvasmeasuredforthereal-conditionand/ormultiple-actuationtests.Themeasuredfrequenciesvere6-6.5Hz.Figure8-183shovsfrequencyanalysesfordifferentpressuretransducersforonetest.P5.2-sitsonthebottombeneaththemiddleofaquencherarm.P54-ismountedontheconcretewallattheintersectionofwallandbottom.P5.10-sitsontheconcretewalloppositethecenterpointof'heballofthequencher.Thefrequencyspectraofthepressuretransducersalldisplayapowermaximumatthesamefrequency(3Hz).Therefore,theREV13/798-96 PROPRIETARYlocationofthemeasurementandthestructureofthemountingpositioninthewaterregionoftheKarlsteinteststandhavenoinfluenceonthemeasuredfrequencyofthepressureoscillations.85.3.34e2ShiftingofthePSD'intheTra~nsositionfromtheTestStandtoSSESThecomparisonofthepressuretimehistoriesmeasuredintheKarlsteinquenchertestswiththepressuretimehistoriesspecifiedinSection4.13isaccomplishedbyusingthefrequencypowerspectra.ThefrequencyspectraoftheKKBtracesformingthebasisofthespecificationinSection4.1.3andareillustratedinFigures4-31to4-33Thespecifiedpressureoscillationshavetheirdominantfrequencyintherangeof6.5-8Hz.TocoverthepressureoscillationfrequenciesforSSES,thefollowingrulefortreatmentofthetraceswasgiven:Thethreetracesshouldbetime-expandedbyafactorintherangefrom0.9to1.8.Thepressureamplitudesshouldbemultipliedbyafactorof1.5.Tobeabletomakeacomparisonwiththemeasuredpressureoscillations,itisnecessarythatthefrequencyspectraofthethreetracesbeshiftedinfrequencyandstretchedinamplitude.InthisSection,weillustrateamethodbywhichthoseoperationsonthefrequencyspectracanbeperformed.8.5.3.3.4.2.1Pte5uencfshiftTheamplitudesarepreservedinthefrequencyshift.Toensurethat,theareaunderthepowerspectrummustbeheldconstant.Sincetheanalysistimerangeforthefrequencyanalysisisfinite,itmustbemadecertainthatthecomparisoninvolvesonlyspectrainwhichapproximatelythesamenumberofoscillationperiodswereanalyzedThetracesareexpandedorcompressedbythefactorf<,whilekeepingthezeropointfixedLetusdesignatetheexpandedorcompressedfrequencybyf'ndtheoriginalfrequencybyf.ApowerspectrumcanalwaysbesubdividedapproximatelyintotriangleswhosebaseisthefrequencyandwhosealtitudeisthepowerdensityIntheoriginalspectrum,theareabeneathatriangleis:f-fA21~h2REVli3/798-97 PROPRIETARYForthenewfrequency:fl=fxflfxfp2Therefore,wehaveforthenewarea:A''utsinceA'A,h=f~h''h~rThepowerdensityoftheshiftedspectrumisinverselyproportionaltothefrequencymultiplier.Inthisdefinition,thefrequencymultipliersaretobetakenfromSection4.1.3.Fromthefactor1.8wegetfV=1/1.8andfromthefactor0.9wegetfV=1/0.9.Ifthefrequencyisreducedtohalf,thepowerdensityisdoubled.85.3.34.2.2~AmlitudeStretchingThefollowingrelationprevailsbetweentheamplitudeofaload-vs.-timefunctionandthepowerdensity:a=k-bf'2k=correctionfactorForthestretchedamplitude,wehavea'fa.Therelationbetweenpowerdensityandamplitudeispreservedbythestretching,sothatthesamecorrectionfactorisalsovalidafterthestretching.Therefore:h''k-bf'andthus:hah)h'=f.hh2a8-98 PROPRIETARYThepowerdensityratiointheamplitudestretchingisproportionaltothesquareof'heamplitudemultiplier.8.53.3.43SymmetricalLoadCase~SimultaneousBlowdownofall16SRV's)AlltheKarlsteinclean-conditionandreal-conditiontestsareusedtoevaluatethisloadcase.Themultipleactuationtestsareconsideredasirrelevanttotheplantforthisloadcase.Theoneexceptionisthe10thblovdowntestofanentiremultipleactuationtestwiththeshortdischargeline.Thosetestsarestarted10,minutesaftercompletionofthe9thblowdowntest.Theyarethussubjecttothesameconditionsasthereal-conditiontests.Accordingly,the10thblowdovntestsofamultipleactuationtestwiththeshortdischargelinearetreatedasreal-conditiontests.ThetesttankinKarlsteinrepresentsthesmallestsinglecellwithrespecttothewaterspace.ThatmeansthatthemaximumpossiblepressureamplitudesforSSESweremeasured.AccordingtoSection8.5.3.2,themeasuredpressureamplitudesarecoveredbythespecificationForthisloadcase,themeasuredfrequenciesofthepressureoscillationscanalsobetransposeddirectlyfromtheKarlsteinteststandtoSSES(seeSection8.53.2).Thus,allthepressuretimehistorycanbetransposeddirectlyfromtheteststandtoSSES.Inordertoshowthatthemeasuredtimehistoriesarealsoenvelopedbythespecification,thefrequencyspectraofthemeasuredpressureoscillationsarecomparedwiththefrequencyspectraofthespecifiedtraces.Sincethemeasuredfrequenciesdifferfrom.thefrequenciesofthespecifiedtraces,thespectramustbetreatedbythemethodillustratedinSection8.5.3.3.42andbroughtintocoincidenceatthedominantfrequency.ThepressureoscillationsmeasuredatpressuretransducerP5.2areusedforthiscomparison,since,thepressuretransducerP5.2exhibitsthehighestpowerspectrumofallthepressuretransducersthatareuseablefortheoverallloadingofthecontainment(P5.5isnotconsidered-seeSection8.5.3.1).PressuretransducerP5.2ismountedonthebottomofthetesttank,directlybeneathaquencherarm.ThatpositionisalsopresentinSSES.Therefore,thispressuretransducermeasurespressureoscillationshavingthegreatestrelevancetoSSES.Purthermore,thespecifiedtracesarealsoresultsofameasurementmadewitha.bottompressuretransducerwhoselocationwassimilartothatofP5.2.REV1,3/798-99 PROPRIETARYThecomparisonofthefrequencypowerspectraisshowninFigures8-184to8-188WeseethatthefrequencyspectraoftheKKBtraces,whichwerefrequency-shiftedandamplitude-stretchedasdescribedinSection8.5.3.3.4.2envelopthefrequencyspectraofthemeasuredpressureoscillations.Therefore,itcanbestatedthat:a)theKarlsteinmeasurementresultsareconservativefortheloadcaseofsimultaneousclearingofall16quenchers{single-celleffect);'Ib)forthisloadcase,thepressureoscillationsareenvelopedbythespecificationwithrespecttotheiramplitude,theirfrequencypowerspectra,'andtheirspatialdistribution.8.5.3.3.44UnsymmetricalLoadCaseslowdownViaOneSRV)Forthisloadcase,alldeterminativeparameters,exceptforthe-,watersurfacearea,weresimulatedintheKarlsteinteststandaccordingtotheiractualvaluesforSSES.Fortheloadcaseofventclearingwithonequencher,alargerwatersurfaceareaisavailabletothequencherinSSESthaninthetestintheKarlsteinteststand.Accordingly,thepressureoscillationfrequenciesareraisedandthepressureamplitudesarelowered.Inthisverification,weconservativelymakenoallowancefortheamplitudedecreasewithincreasingwatersurfacearea.ThefrequenciescalculatedaccordingtoSection8.5.3.3.2fortheloadcaseofblowdownviaoneSRVarecompiledinthefollowingtable:Frequencyofthepressureoscillations(Hz)MeasuredFrequencymultiplierPlantSpecifiedfrequency.bandCLEANCONDITION3.5-41.54-1.485.4-5.9REALCONDITIONS5CLEANCONDITION1.421.427.17.13.75-8.9480RWREALCONDITIONS6.51.378.9REV.1,3/798-100 PROPRIETARYThefrequenciestransposedtotheplantareallenvelopedbythespecifiedfrequencyband.Fortheloadcaseofventclearingofonequencher,themultipleactuationtestsmustalsobeconsidered(theywereincludedunder"realconditions"intheTableabove)Fortheloadcaseofsimultaneousblowdownof16quenchers,itwasshownthatthemeasuredpowerspectraareenvelopedbythespecifiedpowerspectra.Thatstatementappliesforallfrequencyranges.Iftwopowerspectraarebroughtintocoincidenceatonefrequencyandifbothspectraaresubjectedtothesamefrequencyshift,thenthereisnochangeintherelationofthetwospectratoeachother.Therefore,thepowerspectraoftheclean-conditionandreal-conditiontestsarealsocoveredbythespecificationintheloadcaseofventclearingofonequencher,since,asstatedabove,thetransposedfrequenciesfromthetestareallenvelopedbythespecificationfrequencyrange.Forthemultipleactuationtests,test4.1.6isconsideredtobeenvelopingforthelongdischargeline,sinceitprovidedthehighestpressureamplitudes.Fortheshortdischargeline,test20.R1.10(whichformallycanbeclassifiedasamultipleactuationtest)isconsideredtobeenvelopingforthesamereason..Classifiedasareal-conditiontest,itwasshownintheprecedingSectionthatthespecifiedtracesenvelopthepressuretimehistoriesforthattest.InFigure8-189itisshownthatthepowerspectrumoftest4.16isalsoenvelopedbythespecifiedKKBtraces.EvenundertheveryconservativeassumptionthatthepressureamplitudesmeasuredinKarlsteincanbetransferredwithoutchangefortheloadcaseofvent,clearingofonequencher,thepressuretimehistoriesareenvelopedbythespecifiedtraces.85.3.34.5Un~smmetricalLoadCaseslowdownviaThreeA~d'acentSRV~sgThisloadcaseisboundedbytheloadcasesofsimultaneousventclearingof16quenchersandventclearingofonequencher.8.5.3.3.4-6AutomaticDeDzessuzizatio~asstem~A~DSLoadCaseInthissectionwediscusstheloadcasethatconsidersthefiringofthesixquenchersassociatedwiththeADSunderLOCAconditions.REVli3j'798-101 PROPRIETARYAsshowninFigure8-190,thefollowingconditionsprevailinthesuppressionchamberwhentheautomaticdepressurizationsystemisactuatedduringIBA:Absolutepressureinthewetwellairspace,approximatelyPressuredifference.betweendrywell'andsuppressionchamber2.55bar0.42barTheKarlsteintestswithloweredwaterlevelinthedischargelineareusedtoverifytheADScase.Thesetestsareusedastheycorrectlysimulatethedischargelineasitwouldbewithapositivepressuredifferentialofapproximately0.42barinthedrywell.Thispositivepressuredifferentialwouldresultintheloweringofthewaterlevelinthedischargelinetotheelevationofthebottomofthedowncomersaswassimulatedfortests10.3,ll1,12.1and13.1.Of'thosetests,thetest11.1(envelopinginamplitudeandpowerdensity)isusedasthebasisfortheverification.Theamplitude-reducinginfluenceofthelargerwatersurfaceareaassignedtotheindividualquencherintheADScaseisconservativelyneglected.Also,sinceearlierKMUtestsprovedthatthebackpressureinthesuppressionchamberhasnoinfluenceonthepressureamplitudes,themeasuredpressureamplitudesaretakenunalteredfromthecorrespondingKarlsteintests,inwhichthemeasurementsweremadeatatmosphericpressure.Thepredominantfrequencyintest11.1isat3Hz.AccordingtoSection8.5.3.3.2,Figures8-174and8-175,thefollowingfrequencymultipliersareobtainedfortheADScasefortranspositionofthepressureoscillationsfromtest11.1totheplant:InfluenceofthelargerwatersurfaceareaInfluenceofthe2.55barbackpressureTotalfrequencyfactorDomi.nantfrequency135141957HzNote:Themeasuredlowestdominantpressureoscillationfrequencywasmeasuredintests12.1and13.1,whichfallintothesamecategoryastestll1.Miththetotalmultiplier1.9,thefrequenciesareraisedto3.8Hzandthusliewithinthespecifiedfrequencyband(seeSection8.5.3.3.5).ThedominantfrequencyiswithinthespecifiedfrequencybandREVlg3/798-102 PROPRIETARYThecomparisonbetweenthepreparedtracefrompressuretransducerP5.2fortest11.1andthespecificationisshowninFigure8-191.Asfortheotherloadcases,thecomparisonismadeinthepowerspectra'ofthepressuretimehistories.Thespectrumoftest11.1wasshiftedfromthedominant.frequencyof3Hztothedominantfreqeuncyof5.7Hzwhilepreservingthearea(amplitude).TheKKBtraceoftest76wasshiftedfrom8Hzto5.7Hzwhilepreservingthearea,andthenstretchedbyafactorofl.5inamplitude.Figure8-191showsthatthetracefromthespecification,treatedinthismanner,envelopsthetraceofKarlsteintest11.1transformerdtotheADScasesincethetotalenergyrepresentedbytheareaunderthepowerspectrumcurvefromthespecificationisgreaterthanthatfromtheKarlsteintestll.185.33.47SummaryIthasbeendemonstratedthatthe,frequencypowerspectrumofthepressureoscillationsinthesuppressionchamberareenvelopedbythefrequencypowerspectrumspecifiedinSection4.1.3forallloadcases.Thus,thedesignspecificationprovidesenvelopingloadsalsoforthedynamicexcitationoftheSSFScontainmentbyventclearingofthereliefsystemwiththequencher.8.5.33.5EvaluationoftheNeasuredPressureOscillationsDuringCondensationAsdiscussedinSection8.4.2,threeregimescanbedistinguisedinthecondensationprocess:a)Thequencherisclearedcontinually.b)Thequencherisnotclearedcontinually.c)Onlytheslidingjointiscleared,andthesteamcondensesinthedischargeline.8.533.51TheguencherisClearedContinuallyThesteamiscondensedcontinuallyinthewaterpooloutsidethe-quencher.Calmcondensationprevailsforcoldwaterandalsoforhotwaterintheblowdowntank(seeFigures8-78and8-79)Themeasured.maximumpressureamplitudeiss0.13bar.Thiscondensationphasewasmeasuredforreactorpressuresuptoabout4bar.Thefrequenciesofthepressureoscillationsare70-120Hzforacoldpooland20-45Hzforahotpool.REV1,3/798-103 PROPRIETARY8.5.3.3.52TheQuencherisnotClearedContinu~allThiscondensatonphasebeginsvhenthecondensationrateoutsidethequencheris.greaterthanthesteammassflovthroughtheline.Thepressureinthequencherdropsbelovthehydrostaticpressureofthesurroundingvater.Thewaterpenetratesintothequencher.Thecondensationsurfaceareaistherebydecreasedandsoisthecondensationrate.Theresultisapressureriseinthedischargeline,sothatthewaterthathasflovedinisexpelledagain.Theinflowofvaterfromthesuppressionchamberintothequencherandthesubsequentbrakingandre-expulsionofthewaterisanonstationaryprocessvhichoccursperiodically.Forthatreason,thiscondensationphaseisalsocalledintermittentcondensation.Thephenomenonofintermittentcondensatonisdependentonthewatertemperature.Forcoldvaterthereisahigherrateofcondensationoutsidethequencher,resultinginalargergenerationofnegativepressureinsidethequencherandthereforeamorevigorousflowofwaterintothequencher.Foracoldwaterpool,theprofileofthedynamicpressuresissimilartotheprofilevhichisfamiliarfromthechuggingphaseofthecondensationattheventpipes;seeFigure8-76.Forheatedvaterinthesuppressionchamber,thecondensationrateoutsidethequencherissmaller,sothattheentireprocesstakesontheformofalow-frequencypressureoscillation(SeeFigure8-80)Thetestsin"Karlsteinyieldedasmaximummeasurementresultforthedynamicpressure:+0.28,-0.18bar,foracoldpool.Thetimebetweentwoeventsisabout1.0second.Foraheatedpool,themeasuredmaximumamplitudeis+0.12,-0.07,bar.8.533.5.3CondensationintheDischargeLineandThrutheSlidi~nJointIfthesteamflovdecreasesfurther,aconditionisfinallyreachedinwhichthequencherisnolongercleared,butratherremainscontinuallyfilledwithvater.Thenthereissteady-statecondensationofsteaminsidethedischargelineThiscondensationphaseproceedsverycalmlyandbeginsatreactorpressuresbelow2bar.Inthiscondensationphase,maximumdynamicpressuresof+0.08,-0.04barweremeasuredinthewaterpoolduringtheKarlsteintests.REVl,3/798-t04 PROPRIETARY853.3.54Tr~ansositionoftheMeasurementResultstoSSESInregardtosteamcondensation,theconditionsoftheKarlsteinteststandaredirecltytransposabletotheconditionsofSSES.Onthewhole,thepressureamplitudesduringcondensationaresmallcomparedtot'hoseduringventclearingandthereforearecoveredbythelatter.85.4PoolMixinqDurincnSRVActuationandThermalPerformanceofthe~uencher85.41IntroductionWhenanSRVresponds,steamiscondensedinthewaterofthesuppressionpoolviaaquencherAsthishappens,thewatermustabsorbtheheatofvaporizationofthesteam,andsoitisheated.Whenthereisalong-lastingdischargeofsteamviaaquencher,allthewaterinthesuppressionchambershouldparticpateintheheating,soastolimitthelocalheatinginthevicinityofthedischargingquencherInordertoobtaingoodmixingofthehotterandcolderwaterinthepool,allquenchersarepositionedatasmalldistancefromthebottom(3~6"=1.07m)(seeFigure8-192)).Thewaterheatednearaquencherisspecificallylighterthanthecolderwaterlyingaboveit.Therefore,thewarmerwaterwillriseandmixwiththecolderwater.Toobtainanadditionalmixingeffect,theholeoccupancyofthequenchersweremadeslightlyunsymmetrical(approximately8%).Mhereasthequencherarmshavethesameholeoccupanciesonthesides,onlyonearmofeachquencherhasholesontheendcap.Inthatway,aunilateralthrustcanbeexertedonthewaterinthesuppressionpool.Inthetopviewofthequencherarrangement(Figure8-193),weseethatthequenchersarearrangedintwograduatedcircles.Alongtheinnergraduatedcircle,thequencherarmsallpointinthecircumferentialdirection,andtheendcapwithholesallpointinthesamecircumferentialdirection.Ontheoutergraduatedcircle,thecolumnswouldpracticallypreventathrusteffectifthequencherswerearrangedinthesamemanner.Therefore,thequenchersweredirectedmoreradially,butturnedbyanangleofgf=300inthecircumferentialdirectionfromtheradii.Inthisway,50%ofthethrusttillactsinthecircumferentialdirection(equidirectionallywiththethrustofthequenchersontheinnergraduatedcircle).ItshouldbenotedthatthisnewarrangementsupersedestheoriginalarrangementshowninFigure1-4.Inthefollowing,weshallestimatetheaccelerationofthewaterpoolforthecaseinwhichonequencherontheoutergraduatedREVli3/798-105 PROPRIETARYcircleisoperatedforalongperiodoftimeatareactorpressureof70bar{valvefailureinopenposition).Thenweshallpresentsomemeasurementresultsfromatestwitha4-armquencherintheBrunsbuttelnuclearpowerplantandsomeinformationfromtheGEMNodelquenchertestsrelatedtosteamcondensationwithaquencher.85.42EquationofNotionoftheRotationPoolZtisassumedthatthewaterflowintherotatingpoolcanbeconsideredasastraight-linechannelflowduetothesmallcurvatureofthegraduatedcircleandthelowcircumferentialvelocity.Ifweplacetheoriginofthecoordinatesystematthecenterofthedischargingquencher,thentheequationofmotionoftherotatingpoolreads:5'.2mx+c2xFffmWcWeffThismassofwatertobeacceleratedinthesuppressionchambersumofallflowresistanceseffectivedrivingforcedifferentialequationhasthegeneralform:x+ax=bSubstitutingx=u,thedifferentialequationtakestheform:u+au~=bThisdifferentialequationisaspecialformoftheRiccatidifferentialequationThegeneralsolutionofthedifferentialequationsreadsR'ef.53:na.b+bTanha.b(t-K)u(t(n)=-.ga.b+a.q'Tanhga'b(t-c)Theinitialconditionfort=0reads:0n/ab+bTanh/ab-g)/ab=an.Tanh/ab(-g)REVl,3/798-106 PROPRIETARYThisconditionalequationissatisfiedonlyif6andn=0.Theinitialconditionthenleadstothesolution:b.Tanha.bu(t)=ga.bSinceu(t)=X(t),theequationforthevelocityoftherotatingpoolreads:x(t)=b)a.bTaab)a.btForthedistancecovered,wehave:x(t)=pJ'(v)dvThesolutionreads:X(t)=-ln[coshia.b.t(085.03'eterminationofthePlowResistancesThefollowingresistancesareconsidered:a)Wallresistanceofthechannelb)Resistanceforflowaroundthedischargelineswithquenchersandbottomsupportc)Resistanceforflowaroundtheventpipesd)ResistanceforflowaroundthecolumnsThechannelhasthefollowingdimensions:REV.li3/798-107 PHOPRIETARYThehydraulicdiameterofthechannelis:73(26~822-8~84Rect226.822-8.84(27.3)+,22.8cnFortheReynoldsnumber,wehave:Re!W.RRAccordingtoReference36,thekinematicviscosityforwaterat40oCisv=0.65lx10-~m~/s.Xfweassumeavelocityof10-2m/ssoas'tocoverthestart-upphasealso,weget:R+10x2.8.43x1024.651x10TheSSESsuppessionpoolislinedwithasteellinerwhichcannotbeconsideredhydraulicallysmooth.Forsuchlargesteelstructuresitmustbeassumedthattheindividualplatesarenotjoinedtogetherwiththeiredgesparallel,sothattheflowresistanceisincreasedbyprojectingedges.Wethereforeconservativelyassumeanabsoluteroughnessofk=2mm.Thenwehave:Kdh2.8x10-47.1x10Thiscorrespondstoafrictioncoefficientof>=0.022.Theresistancecoefficientisthen:~.1mWh26.844+8.84m'2~7fRR56mr-.022w'28)Cylindricalbodiesareimmersedintothewaterofthesuppressionchamber.Theyarethedischarge.lineswithguenchers,theventpipes,andthesteelcolumns.REV1,3/798-108 PROPRIETARYOutsidediametermSubmergencemQuantityDischargelinesVentpipesSteelcolumns03240.61106733357-3168712Fortheindividualstructuralcomponents,wethenhavethefollowingReynoldsnumber:v=0.01m/s(seeabove)WRe=-dmFortheroughness,weassumek=0.2mm.Then,accordingtoReference39:ReynoldsnumberSubmergencedDischargelinewithquenchervi.thbottomsupportVentpipe5x10~9.4x10~6.17x10-i6.28x1022.555073073Columnl63x10'9x10-+6.9073Theresistanceforceist,hen:p92Thesurfaceareaonwhichthewallresistanceactsis:24Furthermore:c6.16x.44+.73x50+.73x177.8+.73x93WAA16x0.324x9.650m.2c=238mA87x.61x3.35177.8.m2AS-12xl.06x7.3~93m2Sincethewaterregionofthesuppressionchamberalsocontainsafewstructuralcomponentswhichverenotconsideredhere,anadditionalallowanceshallbemade.Mechoose:2c300mWREV.13/798-109 PROPRIETARY8.5.4.4DeterminationoftheForceNovi~nthePoolForcesonthewatermassinthesuppressionpoolareproducedbythrustfromtheboreholesononeoftheendcapswhicharepresentoneachofthequenchers.Thesmallestthrustforceisproducedbythequenchersalongtheoutergraduatedcircle,sincetheydonothavetheirthrustboreholesarrangedinthecircumferentialdirection.Thequenchersalongtheoutergraduatedcircleareturnedbyanangle4=30orelativetotheradialdirection.g=50'DV=differencebetweenpressureinthequencherandambientpressureThethrustforceresultsfromtheimpulseoftheoutflowingsteam.F~APxA-+PxMDxA~UeffectiveoutletareaofquencherPD=densityoftheoutflowingsteamW=velocityoftheoutflowingsteamDAsaneffectiveoutletareaofaquencherendcap,thereisavailable:A~ga<xADUgeomeDC0.8(Section8.5.2.3)+0geom(~l2(4)6.9x10m-32ADOgeom5'~2x10m-32Aconstantreactorpressureof70barischosenfortheestimateof.theeffectivenessoftherotatingpool.AccordingtoReference37,themassflowthroughthereliefvalveatareactorpressureof70baris:REVlr3/798-110 PROPRIETARYm=illkg/sTheresultingstagnationpressureinthequencheris:p=llbarandthesteampressureinthequencherholesis:pD=64barTherefore,PD"=34kg/m>WD462m/sTheforceactinginthecircumferentialdirectionisthen:FeffFsinFeffTherefore:(AP+P+Wj))A"xsinQwithQ=30'DUFeff2~O'N+lo5KN3~5KN85.45WorkingEquationsfortheRotation'PoolofSSESTheequationofmotionfortherotatingpoolreads:5'.2mx+c2effwwThisdifferentialequationwassolvedingeneralforminSection85.4-2.Todetermineconsiderthehave:themassofwaterwhichistobemoved,wemustinternalstructureswhichreducethewatermass.MeI,4(26.822)-(8.84))x.73--x(.324)x7.3x164--x(61)x3.35x87--x(1.06)x12x7.3]43.5x10KgForthetotalresistancecoefficientwehaveaccordingtoSection8543:C<=300mandfortheeffectivelyactingforcewehaveaccordingtoSection8544:F=35KNeffREVl,3/798-111 PROPRIETARYTherefore,theeguationofmotionreads:3.5x10xX+1.5x10xX6-52or3.5x10Therefore,for:a=4.3x104b=9.9x102X+aX=bVab6.55x10tTheequationforthevelocityoftherotatingpoolreads: 3X(t)=,1.52x10Yanh6.55x10tTheequationforthedisplacementreads:X(t)=23.21nicosh6.55x10tiTheresultsareillustratedinFigures8-194and8-195.854.6EstimateoftheHeati~noftheSuppressionChamberMaterThelocalheatingofthesuppressionchambervaterresultsfromthebalanceoftheheatbroughtinbythecondensingsteamandtheheatdissipatedbytheflowingwater.Astimepasses,hovever,thepoolissetintomotionbytheimpulseof.theinflowingsteamandreachesavelocitysuchthatmostoftheheatbroughtinisdistributedoveralargervolumeofwaterthantheassumedlocalvolume.,Thedifferencebetweenthelocalandmeanwatertemperaturedecreases.85.47ExperimentalProofs854.71NodelTankTestsThrustmeasurementsonasteamjetveremadeintheKarlsteinmodeltankintheSpringof1973(Ref40).Thetestset-upisillustratedinFigure8-196Thesteampipeisconnectedby.aspringtothesidewallofthemodeltank.Theexcursionofthespringwiththesteampipeismeasuredbyadisplacementtransducer.Themeasurementsystemvascalibratedbydeterminingtheexcursionofthesteampipeforadefinedforce.Thesteamoutletopeninghadadiameterof10mm.Themassflowdensityvas600to630kg/m<s.Themeasuredreactionforceswere20-28N.REV1,3/798-112 PROPRIETARYAshortcalculationyields:OutletareaRestpressurebeforetheoutletopeningPressureaftertheoutletopeningSteamdensity(at2.6bar)=7.854x10-~m~4.5bar2.6bar.=1.44kg/m~Theresultingoutletvelocityis:W=gK-2.6x10~~1.135W-452.7m/sandthethrustforceis:F=(PW+hP)AffAff=08xAegeomF~(1.44x(452.72)+1.6x10)x0.8x7.854x10F~284NThemeasuredvaluesareloverthanthecalculatedvalues.Themeasurementshaveprovedclearlythattheimpulseoftheemergingsteamjetbecomesactiveasathrustandthat,vithrespecttothevelocitybuildupoftherotatingpool(andthusforthemaximumlocalheating),itisconservativelyboundedbythecalculatedvalues.85.47.2KKBTestDuringtheNuclearCommissioningThepressurereliefsystemwastestedduringthecommissioningphaseoftheBrunsbuttelnuclearpoverplant.Inonesuchtest,areliefvalvewasheldopenforatimeofabout270seconds.Thesuppressionchambercoolingsystemvasswitchedonduringthetest.Waterwasdrawnoffinthelowerpartofthepool,cooled,andsprayedfrompipesprovidedwithholesandlocatedunderthetopofthesuppressionchamber.12measuringpointsaremountedinthewaterregionofthesuppressionchamber.Theyarearrangedatthreedifferentelevations(14m,16.5m,18.2m)andatfourdifferentcircumferentialpositions(5o,75o,195o,245o).Thewaterlevelisataheightof18.89m.Figure8-197showsathreedimensionalspatialrepresentationofthemeasuredtemperaturefieldinthevaterjustbeforeteststart(curve1)andat228secondsafterteststart(curve2).InFigure8-197,theverticalpositionofthetransducerisrepresentedontheordinateandthecircumferentialpositionontheabscissaThetemperatureaxispointstotherear.Theheatingofthepoolisindicatedasthedifferenceofcurves2and1atthreeelevationpositions.Themeanwatertemperaturewasapproximately32.3oCbeforethetestandapproximately42.8oCREV.1,3/798-113 PROPRIETARYat228slater..Themaximummeasuredtemperaturewas500C,sothatthemaximumdeviationfromthemeanwas7.20C.Thedischargingquencherwaslocatedat285'tanelevationof14915mandacceleratedthewatertowardtheleftintheFigure.Correspondingly,the.watertemperatureishigheraboveandtotheleftofthequencher.Fromthatwecanseetheeffectivenessofthequencher'sarrangementnearthebottomandoftheunsymmetricalholearrangementwithre.,pecttouniformutilizationoftheheatsinkofthewaterpool.85473GKNHalfScaleQuencherCondensationTestAseriesofintermediatescale(1:2)condensationtestswereperformedintheGKMteststandtodemonstratethehightemperatureperformanceoftheguenchers(Ref.27).Condensationtestswererunonsevendifferentversionsofthequencherdevice.Thelastthreeversionshad10-mmdiameterholeson'hequencherarmsThespacingoftheholecenterlineswas1.5diameterscircumferentiallyand5.0diametersaxially.ThisholepatternisalsoadoptedintheactualSSESquencherdesign.Thesetestswererunatawatertemperaturerangingfrom13oCto100oC(56oF-2120F)andasteammassflux(withrespecttotheholearea)rangeof8to495kg/m~(1.6to101ibm/ft~s).Matertemperaturesashighas1070C(225~F)weremeasuredatcertainlocationsinthesetests.85.48SummaryTheKarlstein'uenchertestsandpreviousGKMhalfscalequenchertestsshowclearlythatsmoothsteamcondensationcanbeachievedatelevatedtemperatureswhichapproachthe.localsaturationlimit.XnadditionthecalculationsandKKB,inplanttestsprovideinformationwhichsuggestthatpoolmixingisenhancedbysteamdischargethroughtheholesintheendcapsofthequencher."85.5VerificationofSubmergedStructuresLoad~SecificationDueToSQVActuationSection4.1.3.7givesthedesignspecificationfortheloadsonsubmergedstructuresduetoSRVactuation.Thebasisforthespecificationisthethreepressuretimehistoriesusedforthecontainmentanalysisbutinsteadofaconstantamplitudemultiplierof1.5variousmultipliers,relatedtothecrossectionalareaoftheobject,areused.(seeTable4-15).TheloadingonthecolumnsincludingthelocalizedefectatP5.5hasbeendiscussedinSection8.5.3.2.1.2REV1,3/798-114 PROPREETARYInadditiontheeffectsofairbubbleoscillationloadsonthequenchershavebeendiscussedinSection8.5.23.6.ThefollowingsectionwilldiscusstheloadingsontheventpipesasmeasuredintheKarlsteintesttankandprovideadescriptionoftheinfluencefortheexpelledwaterduri'ngventclearing.8.5.51LoadsontheVentP~ie85.511MeasurementoftheLoadsXnordertodeterminetheloadingoftheventpipenearaquencher,aventpipehavingthesameoutsidediameterandwallthicknessasthatinSSESwasinstalledintheKarlsteinteststandandsupportedbytypicalbracing.(seeFigure8-10).Underneaththebracing,bendingstrainsweremeasuredintwomutuallyperpendicularplanesbymeansofstraingauges(SGS1andSGS2)(seeFigures8-11and8-12).Thestraingaugesweremountedabout100mmbelowthebracing.Theoutsidediameteroftheventpipeis:D=0609mandtheinsidediameteris:D;=0589mThus,thecross-sectionalareais:A~0.0188m2andthemomentofresistanceis:4Dl320Wehave:-332.77x10mxW~'M~GxExWTherefore:M~2.77x10'0'-3,11'ndhence;M~057cREV.1,3/798-115 PROPRIETARYIfweinsertcinmicrometerspermeterintothisequation,weobtainthebendingmomentinkN-mThebendingmomentscalculatedinthismannerarestaticequivalentloads.5 5.1.2MeasnnedBendi~nsaeentsFigures8-198to8-200showthedependenceofthemeasuredresultantbendingmomentsonthereactorpressure,ventclearingpressure,and.pressureoscillationamplitudethatweremeasuredneartheventpipeontheconcretewall.Onlythetestswithcleanconditionswereusedfortheplotofthemeasuredbendingmomentsversusreactorpressure,whereasalltestsinthereactorpressurerangeof60-81barwereusedfortheplotsofthebendingmomentversusventclearingpressureandpressureoscillationamplitude.Themeasurementsofthebendingstrainsattheventpipewereperformedonlyforthetestswiththelongdischargeline.Themeasuredmaximumbendingmomentwas14.6kN-mata74barreactorpressureanda13.8barventclearingpressure.85513Extr~aolationoftheMeasurementResultsand~ComarisonwiththeSpecifiedValueIfthemeasurementvaluesareextrapolatedtotheextremeconditionsintheplantonthebasisofFigures8-198and8-199,wegetthefollowingextrapolatedmaximumvalues:16.5kN-mwithrespecttoan88barreactorpressure,19.0kN-mwithrespecttotheventclearingpressureof16.5barforthelongdischargepipe,asextrapolatedinSection8.4fortheextremeboundaryconditionsintheplant.Inthespecification,amaximumpressuredifferenceof0.75x0.8=0.6barwasspecifiedfortheventpipewiththedistributionillustratedinFigure4-24.ThepressuredistributionfortheventpipeinstalledintheKarlsteinteststandisshowninFigure8-201Thefollowingrelationappliesforthepressureattheendofaventpipe:~dPdP7.3-1.837.3-3.65hPssOe4barREVli3/798-116 PROPRIETARYAttheclampingpointoftheventstrut,wehave:AP07.3-1.837.3-6.3AP=01barThepressuredistributionfromtheendoftheventpipetotheclampingpointofthevent-pipestrutistrapezoidal.L=0.1x-'-'-x2.652.65(0.4-0.1)2S'(2)3Theleverarmoftheactingforcewithrespecttotheclampingpointis:0.1+.0.42S1'59Forthebendingmomentattheclampingpointweget:M5=(~2~'2.65x0.6x1.59)10SP~SP63kNmRelativetothestraingauges,wehave:MBSP57kNmTheextrapolatedmaximummomentwas19kN-m.Itisthusdemonstratedthatthespecificationenvelopsthemeasurementvaluesandtheirextrapolation.Theproofthatthespecificationenvelopsthemeasurementvaluesandtheirextrapolationisbasedonapurelystaticanalysis.Suchananalysisispermissiblebecausetheexcitingpressureoscillationshaveafrequencyof4-6Hz.However,thestraingaugesindicateanaturaloscillationfrequencyof17-20HzfortheventpipewhichisveryclosetothenaturalfrequencyoftheventinSSES(19Hz)(seeFigure8-202).Hence,itcanbeassumedthatthedynamicloadfactorisclosetoone.8552InfluenceofExpelledWaterDuringVentClearinc[AreviewofthehighspeedfilmsandpressuretracesatP5.5fromtheKarlsteintestsshowsnegligableinfluenceoftheexpelledwateratthisgage.Inadditionthetotalpenetrationoftheexpelledvaterappearstobeapproximately3feetfora70barinitialsystempressure.Therefore,noadditionalloading,otherthanthatalreadyincludedinthepressuretracesvillheconsidered.REV1,3/798-117 PROPRIETARY(AtimecorrelationofahighspeedfilmtopressuretraceatP5.5willbesuppliedlater.}85.53SummaryTheloadsmeasuredonthedummyventpipearestaticequivalentloads,butloadswhichareasumofindividualcomponents.Inthespecification,thetransverseloadsoninternalstructuresoriginatingfromtheblowdownofthereliefsystemare.formulatedasdifferentialpressuresacrosstheinternalstructures.ThedifferentialpressureshavethesamepressuretimehistoryasthedynamicpressuresinthewaterregionofthesuppressionchamberThisformulationofthetransverseloadsontheventpipe(moregenerallyontheinternalstructuresinthewaterregionofthesuppressionpool)yieldstheenvelopingstaticeg'uivalentload.ThiswasalsoverifiedbytheKKBtestswiththeactualreliefsystem(Ref.38).Themaximumdifferentialpressurescalculatedfromthemeasurementresultsarep=0.16baratthequencherarm,andp=0.11barattheprotectivepipeonthedischargeline.TheyarebothconservativelyboundedbytheKKBspecifiedvalueofp=0.2bar.TheKKBtestresultsshowsthatthereisaclearseparationbetweenthespecifiedloadsandthemaximummeasuredloadsforboththelateralandverticalloadsoninternalsinthepoolofthesuppressionpool.BasedontheverificationofthetransverseloadsbytheKKBtestsandbasedonthecomparisonbetweenspecificationandmeasurementfortheKarlsteintests(seeSection8.5.5.1),itcanbestatedthatthevaluesformulatedinthespecificationforthetransverseloadsoninternalstructuresinthewaterregionyieldenvelopingstaticequivalentloads.REV1g3/798-118 KEY:1.Reliefvalve2.Compressedair3.HPsteamline4.Heating5.SRV6.CondensateeP,$$...'CIA,(o!>.d'1$$'4(C$!PACmli0a!0CUmmZCh0Zzzgzo$mLmZZr~Dmmnfh0Z | PROPRIETARYCHAPTER8SSESQUENCHERVERIFICATIONTESTTABLEOFCONTENTS81INTRODUCTION8118.1.281.218.12.1.181.21.2812.281.2218.12.2.281.2.2.38.122-48.1-2.25PurposeofTestsTestConceptSingleCellApproachSingleCellTheoryApplicationofSingleCellApproachSimulationofSSESParameters'rimarySystemPressureSafetyBeliefValve(SBV)DischargeLineVacuumBreakersQuencherBEV1,3r79 82TESTFACILITYANDINSTRUMENTATION8.2.182.1.18.2.1118.2.1128.2.1138211.48.21.158.21168228.2.21822.282.238.2.231822.328.2.2.48.22.4.182.24282.2.5TestFacilityMechanicalSet-UpSteamBoilerSteamAccumulatorSteamLineandBufferTankSafety/RelicfValve(SRV)DischargeLineandQuencherTestTankInstrumentationGeneralDescriptionInstrumentationIdentificationOperatingInstrumentationDisplayonControlConsoleAcquisitionbyComputerTestInstrumentationMeasuringPointsSet-UpofMeasuringInstrumentsVisualRecordingREVli3/798-2 83TESTPARAiiETEHSANDMATRIX83I83.2VentClearingTestsCondensationTestsREV1,3/798-3 84'ZESTRESULTS84.1841.184128.413VentClearingTestResultsTestParametersBehavioroftheSRVandSystemPress'uresDynamicPressureLoadsonthePoolBoundaries841.48428.42.18.4.2.2842218.4.221.18.42.2.12842.2.2LoadsontheQuencherandBottomSupportSteamCondensationTestResultsTestParametersPresentationofTestResultsSurveyofObservedCondensationPhasesBlowdownatLowWaterTemperatureBlowdownatHighMaterTemperatureStatisticalEvaluationoftheDynamicPressureLoadsonthePoolBoundaries8422.21DependenceofDynamicBottomandWallPressuresonSystemPressureandWaterTemperature842222842223OccurrenceFrequencyDistributionsoftheDynamicBottomandMallPressuresStatisticalCharacteristicsoftheDynamicBottomandMallPressures84223TemperatureVariationsintheMaterRegionoftheTestTank8422.4843MaterLevelintheDischargeLineWhenOpeningandAfterClosingtheSRVCheckingandCal.ibrationoftheMeasuringInstrumentation844845AnalysisofMeasurementErrorsRepetitionTestsandReproducibilityoftheResultsREVl,3/798-4 85DATAANALYSISANDVERIFICATIONOFLOADSPECIFICATION8.5.18.5.11EvaluationofTestTankEffectsonBoundaryPressureMeasurementsEffectsofFreeWaterSurfaceandRigidWalls8.5128513851.4MethodofImagesTheTestStandasaSingleCellSpatialDistributionofPressureintheTestTank8.5.1.58515185152851.5.38515.4InvestigationoftheInfluenceofMovableWallsontheMeasurementResults(Fluid-StructureInteraction)GeneralRemarksExperimentalInvestigationoftheTank'sNaturalOscillationsExperimentalInvestigationoftheTank'sResponsetoVentclearingLoadsTheoreticalInvestigationsandModelCalculationsoftheInfluenceofFSI85.1.5.41851.5428.51543ComputationModelsModelParametersandInputforCalculationsWithoutFSI(RigidTank)ModelParametersandInputforCalculationsWithFSI85.1.544ResultsoftheFSICalculations8.5.285.21VerificationofSRVSystemLoadSpecificationDuetoSRVActuationPressuresDuringtheVentClearingProcess852118.5212VentClearingPressuresfortheLongLineVentClearingPressuresfortheShortlineREV.l.3/798-5 8521.3Transposition-oftheMeasurementValuestoSSESandComparisonwiththeDesignSpecification852.2PressuresDuringtheStationaryCondensationofSteam8522185222LongLineShortLine8.522.3TranspositionoftheMeasurementValuestoSSESandComparisonwiththeDesignSpecification8.52.3ExternalLoadsontheQuencherandBottomSupport8.5.2318.5.2.3.1.185231.2852312.185.231.2.2VerticalForceMeasurementoftheVerticalForceMeasuredVerticalForcesLongLineShortLine85231.3TranspositionoftheMeasurementValuestoSSES8523131852.31,.3.28.5.2.313-3852328523218.5.2.3-2.285232.2185232.22852323LongLineShortLineSummaryTorsionalMomentMeasurementoftheTorsionalMomentMeasuredTorsionalMomentslongLineShortLineTranspositionoftheMeasurementValuesto'SSES85-2338.5.2.3.31BendingMomentsattheQuencherArmsMeasurementoftheBendingMomentsREV1,3/798-6 8.5.2-33.28.5.2.3.3.3MeasuredBendingMomentsTranspositionoftheMeasurementResultsIntotheWeld8.5.233.4852.33.58.5-2.3.48523418.5234.28.5.2.3438.523448.5.2358-52-3.6SpecifiedStaticEquivalentLoadsEvaluati.onoftheMeasurementResultsBendingMomentsattheBottomSupportMeasurementoftheBendingMomentsMeasuredBendingMomentsSpecifiedStaticEquivalentLoadEvaluationoftheMeasurementResultsForcesontheQuencherInfluenceofanAdjacentQuencher~8-5.237LoadsontheQuencherDuringSteamCondensation852371ManifestionFormsofIntermittentCondensationintheKarlsteinTests8523728.5.237.3IllustrationoftheMeasurementValuesEvaluationoftheMeasurementResultsforthe.QuencherArm85237.4EvaluationoftheMeasurementResultsfortheBottomSupport8.5-2.375EvaluationoftheMeasuredTorsionalMoments852376EvaluationoftheMeasuredMaximumMomentsattheQuencherArmDuringIntermittentCondensation85.3VerificationofSuppressionPoolBoundaryLoadSpecificationDuetoSRVActuation8531EvaluationoftheLocalEffectsSeenatPressureTransducerP5.58532Veri.ficationoftheSpecifiedPressureAmplitudesandVerticalPressureProfilesafterVentClearingREVl~3/798-7 8532185321.185.3212OverpressuresVerticalPressureProfileVerticalPressureProfileIncludingLocalEffectsatP5.58532285322.185.3385.3.3.185331.18.5.33.11.1853311.28.5.3.3.1.1.385.331.14UnderpressuresVerticalPressureProfileVerificationofthePressureTimeHistoriesUsedfortheSSESContainmentAnalysisTranpositionMethodfortheOscillationFrequencyCalculationofMeasuredOscillationFrequenciesPPGLTestsatKarlsteinGKMModelQuencherTests'KBHotTestsConclusionfromtheFrequencyCalculations8.5.3.328.5.3.3.38533.3.18.53332MultipliersforConversionoftheBubbleFrequenciesFromtheTestStandtoSSESTranspositionMethodforthePressureAmplitudesPPGLQuencherTestsatKarlsteinKMUQuencherTestsintheModelTestStandinKarlstein8.53.3.338533.34853348533.41853.34.2AnalyticalCalculationsInfluenceofBackpressureonthePressureAmplitudesVerificationofDesignSpecificationFrequencyAnalysesofSelectedTestsShiftingofthePSD'sintheTranspositionFromtheTestStandtoSSESREV.li3/798-8 853.34.2l85.334.228533.4385.3.3.4485334585.33.4.6FrequencyShiftAmplitudeStretchingSymmetricalLoadCase(SimultaneousBlowdownofall16SRV's)UnsymmetricalLoadCase(BlowdownViaOneSRV)UnsymmetricalLoadCase(BlowdownViaThreeAdjacentSRV')AutomaticDepressurizationSystem(ADS)LoadCase853347853.3.58.533.518533.5285.33.53853.354SummaryEvaluationoftheMeasuredPressureOscillationsDuringCondensationTheQuencherisClearedContinuallyTheQuencherisNotClearedContinuallyCondensationintheBlowdownPipeandThrutheSlidingJointTransportationoftheMeasurementResultstoSSES854854.18.5.4285.438544PoolMixingDuringSRVActuationandThermalPerformanceoftheQuencherIntroduction'EquationofMotionoftheRotatingPoolDeterminationoftheFlowResistancesDeterminationoftheForceMovingthePool85.4.5WorkingEquationsfortheRotatingPoolofSSES854.68.5.4.7854.7185472EstimateoftheHeatingoftheSuppressionChamberWaterExperimentalProofsModelTankTestsKKBTestDuringtheNuclearCommissioningREVl,3/798-9 8.5.4.7-3GKMHalfScaleQuencherCondensationTest8.5488.5.5SummaryVerificationoftheSubmergedStructuresLoadSpecificationDuetoSRVActuation8.5.5.1855118551285513LoadsontheVentPipeMeasurementoftheLoadsMeasuredBendingMomentsExtrapolationoftheMeasurementResultsandComparisonwiththeSpecifiedValue8.55.2InfluenceofExpelledHaterDuringVentClearing8553SummaryREVl,3f'798-10 SECTION8.0FIGURESNumberTitle8-1MathematicalDesscriptionofaSingleCellConfigurationwithSolidWalls;SolidBottomandFreeWaterSurface8-28-38-48-58-68-7EguivalenceofaSingleCellConfigurationandaParallelBubbleFieldOscillatinginPhaseGeometricSingleCellPartitionoftheSuppressionPoolTestStandSchematicDiagramLongDischargeLineConfigurationShortDischargeLineConfigurationKarlsteinTestTankPlanVievTypicalVentClearingInstrumentation8-8KarlsteinTestTankC-DVievTypicalVentClearingInstrumentation8-9KarlsteinTestTankA-BVievTypicalVentClearingInstrumentation8-10KarlsteinTestTankPlanVievTypicalCondensationTestInstrumentation8-11KarlsteinTestTankC-DViewTypicalCondensationTestInstrumentation8-12KarlsteinTestTankA-BViewTypicalCondensationTestInstrumentation8-13T-QuencherShowingTypicalVentClearingInstrumentation8-14T-QuencherShovingTypicalCondensationTestInstrumentation8-158-168-178-188-19TestMatrigforVentClearingTestLocationofTestGroupNo.1intheOperationFieldLocationofTestGroupNo.2intheOperationFieldLocationofTestGroupNo.3intheOperationFieldLocation'fTestGroupNo.4intheOperationFieldREV1,3/798-11 8-208-218-228-23LocationofTestGroupNo.5intheOperationFieldLocationofTestGroupNo.6intheOperationFieldLocationofCondensationTestsintheOperationFieldValveOpeningTimeVersusAccumulatorPressureLongPipeVentClearingTests8-24ValveOpeningTimeVersusAccumulatorPressureShortPipeVentClearingTests8-25VentClearingPressureVersusSystemPressureLongLineVentClearingTests8-26VentClearingPressureVersusSystemPressureShortLineVentClearingTests8-27PeakPositiveWallandBottomPressuresVersusSystemPressure-Long.Line,CleanConditions,ColdPool8-28PeakPositiveMallandBottomPressuresVersusSystemPressure-ShortLineCleanConditions,ColdPool8-29PeakPositiveWallandBottomPressuesVersusSystemPressure-LongLineRealConditions,ColdPool8-30PeakPositiveWallandBottomPressuresVersusSystemPressure-ShortLine,RealConditions,ColdPool8-31PeakPositiveMallandBottomPresssuresVersusSystemPressure-LongLine,CleanConditions,Heatedpool8-32PeakPositiveWallandBottomPressuresVersusSystemPressure-ShortLine,CleanConditions,HeatedPool8-33PeakPositiveWallandBottomPressuresVersusSystemPressure-LongLine,RealConditions,HeatedPool8-34PeakPositiveMallandBottomPressuresVersusSystemPressure-ShortLine,RealConditions,HeatedPool8-35PeakPositiveMallandBottomPressuresVersusValveActuation-LongPipeTest148-36PeakPositiveWallandBottomPressuresVersusValveActuation-LongPipeTest58-37PeakPositiveMallandBottomPressuresVersusValveActuation-LongPipeTests4and4R8-38PeakPositiveWallandBottomPresuresVersusValveActuation-LongPipeTests15and15RREV1,3/798-12 8-3'98-408-418-428-438-448-458-468-478-488-498-508-518-528-538-548-558-568-578-588-598-608-618-628-638-648-65PeakPositiveMallandBottomPressureVersusValveActuation-ShortPipeTests19and19RPeakPositiveMallandBottomPressuresVersusValveActuation-ShortPipeTests20and20RVisicorderTraceP51-P5.10Test41.1.VisicorderTraceP5.1-P5.10Test4R.l.1VisicorderTraceP5.1-P5.10Test4.1.6VisicorderTraceP5.1-P5.10Test11.1VisicorderTraceP5.1-P5.10Test12.1VisicorderTraceP5.1-P5.10Test15.1.1VisicorderTraceP5.1-P5.10Test15.Rl.1VisicorderTraceP5.1-P5.10Test19.1.1VisicorderTraceP5.1-P5.10Test19.R2.1VisicorderTraceP5.1-P5.10Test19.R2.2VisicorderTraceP5.1-P5.10Test19.R2.3VisicorderTraceP5.1-P5.10Test19.R2.4UisicorderTraceP5.1-P5.10Test19.R2.5VisicorderTraceP5.1-P5.10Test19.R26VisicorderTraceP51-P510Test19.R2.7VisicorderTraceP5.1-P5.10Test19R2.8VisicorderTraceP5.1-P5.10Test19.R29VisicorderTraceP5.1-P5.10Test19.32.10VisicorderTraceP5.1-P5.10Test20.1.1VisicorderTraceP5.1-P5.10Test20.Rl.lVisicorderTraceP5.1-P5.10Test20.R1.10VisicorderTraceP5.1-P5.10Test21.1VisicorderTraceP5.1-P5.10Test21.2VisicorderTraceP5.1-P5.10Test25.1VisicorderTraceP5.1-P5.10Test25.R2REVlg3/798-13 8-66MaximumResultantBendingMomentatQuencherArm1-,LongPipeVentClearingTests8-67MaximumResultantBendingMomentatQuencherArm2-LongPipeVentClearingTests8-688-69MaximumResultantBendingMomentatQuencherArm1-ShortPipeVentClearingTestsMaximumResultantBendingMomentatQuencherArm2-ShortPipeVentClearingTests8-70MaximumResultant'BendingMomentattheQuencherSupport-LongPipeVentClearingTests8-718-728-73MaximumResultantBendingMomentattheQuencherSupport-ShortPipeVentClearingTestsObservedCondensationPhasesDuringTestsTypicalVisicorderTraceofStationaryOperationofQuencherTest33.2-10SecondsafterStart8-74TypicalVisicorderTraceofStationaryOperationofQuencherTest35.1-20-SecondsafterStart8-75VisicorderTraceShovingIntermittentOperationoftheQuencher-Test36.1SysemPressure-6.2-1.0barPoolWaterTemp-26~C-300C8-76VisicorderTraceShovingExcerptfromIntermittentOperationofQuencherTest36.1-280SecondsafterStart8-778-78VisicorderTraceShovingSingleEventOutofIntermittentCondensationTest36.1TypicalVisicorderTraceofStationaryOperationofQuencherTest37.2-13SecondsafterStart8-79TypicalVisicorderTraceofStationaryOperationofQuencherTest39.1-10SecondsafterStart8-80VisicorderTraceShovingIntermittentOperationofQuencher-Test40.1SystemPressure-2.5barPoolWaterTemp.-89~C-91~C8-81DynamicBottomPressuresduringtheBlowdownAlongtheUpperandLoverBoundaryoftheOperationField'-82DynamicWallPressuresDuringtheBlowdownAlongtheUpperandLoverBoundaryoftheOperationFieldREV.1,3/798-14 8-83OccurrenceFrequencyDistributionPositiveandNegativeDynamicAmplitudesfortheCondensationTestsPoolTemp.22~C-300C8-84OccurrenceFrequencyDistributionPositiveandNegativeDynamicAmplitudesfortheCondensationTestsPoolTemp.59~C-91>C8-85OccurrenceFrequencyDistributionPositiveandNegativePressureAmplitudeforCondensationTestsPoolTemp.22C-30C8-86OccurrenceFrequencyDistributionPositiveandNegativePressureAmplitudeforCondensationTestsPoolTemp59C-91C8-878-888-89NeanValuesoftheBottomDynamicPressuresDuringtheBlowdownsAlongtheUpperandLowerBoundaryoftheOperationFieldMeanValuesoftheWallDynamicPressuresDuringtheBlowdownsAlongtheUpperandLowerBoundaryoftheOperationFieldWaterTemperatureTimeHistoriesOnPoolWallCondensationTest33.28-90WaterTemperatureTimeHistoriesOnPoolWallCondensationTest35.18-91MaterTemperatureTimeHistoriesOnPoolMallCondensationTest37.28-92RaterTemperatureTimeCondensationTest39.1HistoriesOnPoolMall8-93WaterTemperatureTimeCondensationTest33.2HistoryOnQuencherArm18-94RaterTemperatureTimeHistoryonQuencherArm1CondensationTest35.18-95MaterTemperatureTimeHistoryonQuencherArm1CondensationTest3728-96MaterTemperatureTimeHistoryon.QuencherArm1CondensationTest-39.18-978-98CalibrationofSensorsandRegistrationInstrumentsIntervalsforCalibrationChecksandAdjustmentsofInstrumentation8-99CalibrationSystemREV.1>>3/79 8-100CalibrationResultsDeviationsfromNominalValue-P5.1-P5-108-1018-1028-1038-104MaterLevelinDischargeLineTest15.1MaterLevelinDischargeLineTest20.1WaterLevelinDischargeLineTest32EffectsofFreeSurfaceandRigidTankWallsonDynamicFluidPressure8-1058-1068-1078-1088-109MethodofImagesSSESSmallestUnitCellandtheKarlsteinTestTankPressureProfilesforDifferentBubbleLocationsPressureProfileforaOneandFourBubbleArrangementComparisonofMeasuredandCalculatedNormalizedPressureProfiles8-110ComparisonofPressureProfilesCalculatedfortheKarlsteinTestTankandtheSSESSuppressionPool8-111ComparisonofCalculatedandSpecifiedPressureProfiles8-1128-113TankArrangementShowingInstrumentationandExplosiveChargeLocationsforMeasuringTankReponseConfigurationofExplosiveContainerUsedtoGenerateUnderwaterPressureImpulse8-1148-1158-1168-1178-1188-1198-1208-1218-122TypicalTankResponseDuetoPressureImpulseFrequencyAnalysisofGageMA2FrequencyAnalysisofGageMA7FrequencyAnalysisofGageMA8FrequencyAnalysisofGageP5.10DisplacementCorrelationsfor13Hz-EigenmodeDisplacementforthe13HzEigenmodeTestTankArrangementforShakedownTestsTankDisplacementsand.PressureTraceDuringShakedownTest08.18-123FrequencyAnalysisofGageMA2ShakedownTest08.1REVli3/798-16 8-1248-1258-1268-127FrequencyAnalysisofGageWA7ShakedownTest08.1FrequencyAnalysisofGageWA8ShakedownTest081FrequencyAnalysisofP510ShakedownTest08.1AirNassPlowusedforKOVlBlComputerCodeCalculations8-128UnitWallDisplacementof13HzNodeUsedinKOVlB1ComputerCodeCalculations8-129BoundryPressureDistributionCalculatedforUnitDisplacementof13HzNode8-1308-1318-1328-133WallPresureCalculationwithKOVlB1ComputerCodeEffectsofFSZonBubblePrequencyTypicalPressureTraceinSRVDischargeLineTest4.1.4TypicalPressureTraceinSRVDischargeLineTest20.Rl78-134PressureinSteamLinebeforeSRVVersusPressureinBufferTankatValueOpening8-1358-136PressureinDischargeLineVersusReactorPressureatVentClearing-P4.1LongLineTestsPressureinDischargeLineVersusReactorPressureatVentClearing-P44LongLineTests8-137PressureinDischargeLineVersusReactorPressureatVentClearing-P4.1ShortLineTests8-138PressureinDischargeLineVersusReactorPressureatVentClearing-P4.4ShortLineTests8-1398-140VentClearingPressureVersusValveOpeningTimeSteadyStatePressureVersusReactorPressure-P4.1LongLineTests8-141SteadyStatePressureVersusReactorPressure-P4.4LongLineTests8-l42SteadyStatePressureVersusReactorPressure-P4.1ShortLineTests8-143SteadyStatePressureVersusReactorPressure-P4.4ShortLineTests8-144SteadyStatePressuresatDifferentLocationsAlongtheDischargeLineExtrapolatedto88BarReactorPressureREV1,3/798-17 8-1458-146.8-147TypicalTraceforVerticalLoadLongLineTestsVerticalLoadVersusClearingPressureLongLineTestsVerticalLoadVersusVentClearingPressureShortLine-Tests8-1488-149TypicalTraceforTorqueonBottomSupportLongLineTestBottomSupportTorqueVersusVentClearingPressureLongLineTests8-150BottomSupportTorqueVersusVentClearingPressureShortl.ineTests8-151TypicalTraceforBendingMomentsonQuencherArmsLongLineTests8-152ResultantQuencherArmBendingMomentVersusVentClearingPressureShortLineTests8-153FrequencyDistributionofMaximumResultant'BendingMomentonQuencherArmsandatWeldSeam8-154ResultantBottomSupportBendingMomentVersusVentClearingPressureShortLineTests8-155FrequencyDistributionofMaximumResultantBendingMomentonBottomSupportSG4.5-468-156FrequencyDistributionofMaximumResultantBendingMomentsOnQuencherArmsgtStrainGagesIntermittentCondensation8-157FrequencyDistributionofMaximumResultantBendingMomentatWeldSeamonQuencherArm-IntermittentCondensation8-158FrequencyDistributionofMaximumResultantBendingMomentsatBottomSupport-IntermittentCondensation-0.5mbelowQuencherCenter8-1598-160PowerSpectralDensitiesTestll.1-P5.5PowerSpectralDensitiesTestll1-P528-161PowerSpectralDensitie'sTest4.1.6-P5.58-1628-1638-164PowerSpectralDensitiesTest4.1.6-P5.2PowerSpectralDensitiesTest20.R1.10-P55PowerSpectralDensitiesTest20-R1.10-P5.2REV13/798-18 8-165MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValues-Overpressures8-166MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValuesConsidering.LocalEffects8-167MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValues-UnderPressures8-168KarlsteinTests-ComparisonofMeasuredandCalculatedBubbleFrequency-0$Humidity8-169KarlsteinTests-ComparisonofMeasuredandCalculatedBubbleFrequency-100%Humidity8-170GKMTests-ComparisonofMeasuredandCalculatedBubbleFrequency8-171GKMTests-ComparisonofMeasuredandCalculatedBubbleFrequency-Overpressure8-172KKBIn-PlantTests-ComparisonofMeasuredandCalculatedBubbleFrequency8-1738-174SSESCalculatedBubbleFrequenciesMultipliersforConversionofBubbleFrequenciestheKarlsteinTesttoSSES8-175OverpressureMultiplierforConversionofBubbleFrequencies8-176NormalizedAmplitudeSpectrumVersusBubbleFrequency-KarlsteinTests8-177KarlsteinModelTests-InfluenceofMaterSurfaceonPressureAmplitude8-178GKMTests-InfluenceofOverpressureonBubblePressure8-1798-1808-1818-1828-1838-1848-185PSDofKarlsteinTests-11.1and12.1-P5.10PSDofKarlsteinTests4.1.1and4.1.6-PS.10PSDofKarlsteinTests21.1and21.2-P5.10PSDofTest20.R110-P5.4PSD'sofTest11.1-P5.2,5.4and5.10PSDComparison-Test20Rl1andDesignSpecificationPDSComparisonTest4.1.1andDesignSpecificationREVlg3/798-19 8-1868-1878-1888-1898-1908-1918-1928-1938-194PSDComparisonTest20.R1.10andDesignSpecificationPSDComparisonTest211andDesignSpecificationPSDComparisonTests21.2and25.R2andDesignSpecificationPSDComparisonTest0.1.6andDesignSpecificationIBADrywellandWetvellPressureHistoryPSDComparisonTest11.1andDesignSpecificationTypicalCrossSectionofSSESSuppressionPoolRevisedQuencherArrangementVelocityofRotatingPoolforOneActuatingValveinOuterRow8-1958-1968-1978-1988-1998-2008-2018-202WaterMotionoftheAcceleratedPoolTestStandforMeasuringThrustMeasuredTemperatureDistributionintheKKBSuppressionPoolResultantBendingMomentonDummyVentVersusReactorPressureResultantBendingMomentonDummyVentVersusClearingPressure0~ResultantBendingMomentonDummyVentVersusPressureAmplitudeatP5.7'JSpecifiedPressureDistributiononDummyVentTypicalVisicorderTraceforBendingMomentonDummyVentREVli3/798-20 PROPRIETARYNumberSECTION8TABLESTitle818.28.3TypicalOperatingInstrumentationTypicalVentClearingTestInstrumentationTypicalCondensationTestInstrumentation8.4ParametersatTestStart-LongPipeVentClearingTestSeries85ParametersatTestStart-ShortPipeVentClearingTestSeries8.6ParametersatTestStart-CondensationTestSeries8.7BehavioroftheSRVandSystemPressures-LongPipeVentClearingTestSeries8.8BehavioroftheSRVandSystemPressures-ShortPipeVentClearingTestSeries8.9PeakDynamicPressuresonthePoolBoundaryDuringVentClearing-LongPipeVentClearingTests8.10PeakDynamicPressuresonthePoolBoundaryDuringVentClearing-ShortPipeVentClearingTests8.11Maximum,Strains,MomentsandVerticalLoadsontheQuencherArmsandSupportDuringVentClearing-LongPipeTests8.12MaximumStrains,MomentsandVerticalLoadontheQuencherArmsandSupportDuringVentClearing-ShortPipeTests8.13SystemPressuresandPoolWaterTemperaturesoftheCondensationTests810PeakDynamicPressuresAmplitudesDuringtheDifferentCondensationPhases8.15StatisticalCharacteristicsoftheBottomDynamicPressures(P5.2)8.16StatisticalCharacteristicsoftheWallDynamicPressures(P5.10)8.178.18REV.1,3/79RepetitionTests-ComparisonofRecordedValvesRepetitionTests-MeanValuesandDeviations8-21 PROPRIETARY80SSE~SUENCHERVERIFICATIONTEST81INTRODUCTION8.l.1Pu~roseoftheTestsTheoptimizedquencherdesignforSSESandtheloadspecificationonthewettedboundariesofthesuppressionpool,onthesubmer'gedstructuresandonthepressurereliefsystem,arebasedonparametricmodelteststudiesandfullscaleinplanttestresultsfromasimilarquencherdesign.TheloadspecificationsfortheSSESquencheraredescribedindetailinSection4.1.Inordertoverifytheseloadspecificationsandfurtherverifythequencher'ssteamcondensingcharacteristics,fullscalesinglecelltestswereconductedattheKraftwerkUnionlaboratoriesinKarlstein,WestGermany.8.1.2TestConceptTheconceptsusedtodesignandperformthetestswere:1)Useofaconservativelydefinedsinglecell2)Theclosesimulationofthemainsafetyreliefvalvesystemparameters81.2.1UnitCellApproach81.21.1SingleCellTheoryForagasbubbleoscillationinafreewaterspace,thewatermasscoupledtothebubbleisalternatelyacceleratedanddeceleratedDuringthisprocesstheoverpressureandunderpressureamplitudesdecreasewithincreasingdistancefromthebubble.Whenasolidwallisplacedneartheoscillatingbubble,thewateraccelerationisrestrictedinthedirectionofthewallandthedecreaseinpressureamplitudeinthedirectionofthewallisless.Thiseffectcanbeexpressedmathematicallybyreplacingthe.bubblebyapotentialsourceandaccountingforthewallbythemethodofimages.Theeffectsoftherealsourceandtheimagesourceareaddedforeachpointoftheflowfield.Forthecaseinwhichabubbleisenclosedinanarrowwaterspace,closelysurroundedbysolidwallsandasolidbottomwithafreewatersurfaceatthetop,thewaterspacebelowthebubbleisforallpracticalpurposesunmoved.Onlythewatervolumeabovethebubbleisfreetooscillate..Consequently,thepressuregradientinthelowerwaterspaceisnearlyzero,whilethepressureamplitudeabovethebubbledecreaseswithincreasingproximitytothewatersurface.Thepressureamplitudesarezeroatthewatersurfaceandthemethodofimagesapplies.REV.1,3/798-22 PROPRIETARYAnalytically,thecaseinwhichaplanarfieldofuniformstrengthbubblesarealloscillatinginphaseisthesameasthecaseinwhichsolidwallsexistbetweeneachoftheindividualbubbles.ThesinglecelltestconfigurationusedatKarlsteinsimulatesthisextremelyconservativecaseofparallelbubblesoscillatinginphasewiththesamesourcestrength.Adescriptionoftheequivalenceofthesinglecellconfigurations,usingthemethod-ofimages,iscontainedinFigures8.1and8.2.ForamoredetailedevaluationoftheKarlsteintesttanksinglecell,seeSection8.5.1.81.21.2A~licationofSingleCell~AroachThesubmergenceofthequencherinthetesttankisequaltothehighestvalueintheplant.Astothewatercross-sectionareathesinglecelltheorydescribedaboveisusedFigure8.3showsageometricalpartitionofwaterspace.Thewatercross-sectionareasrelatedtothedifferentquenchersarelistedbelow:QuencherAQuencherBQuencherCQuencherDQuencherEQuencherFQuencherGQuencherHQuencherJQuencherKQuencherLQuencherQuencherNQuencherPQuencherRQuencherSAverageWaterSurface3147mz(33862ftz)3147mz31.47mz3147mz31475231~47AI231.47mz31.47mz31.47mz3147mz3147mz3147mz31.47mz31.47mz31.47mz31.47mzRelatedWaterSurface21.4mz(230-26ftz)214mz31.3mz{336.79ftz)42mz(45192ftz)31~3mz31'mz42mz31m3mz31~3mz42mz31~3mz31a3mz214mz21.4mz31~3mz42mzREV1,3/798-23 PROPRIETARYThesmallestwatersurface(approximately21.4m~)issimulatedinthetests.Therefore,thedynamicpressureamplitudesatthewallsandthebottomaremeasuredunderconservativeboundaryconditions.8.1.22SimulationofSSESParametersThefo'llowingsectionprovidesadescriptionofthoseparametersthatweresimulatedintheKarlsteintestfacilityTheseparametersaretypicalofmostMKIIplants.FormoredetailonthetestfacilityseeSection8.28l.22.1Prima~rSgsternPressureThereactoroperatingpressureforSSESisapproximately1000psig(69bar)whilethehighestpressuresetpointforanySSESSafetyReliefvalveis1205psig(83bar),whichisclosetothehighestprimarypressurethatcanbesimualtedintheKarlsteintestfacility(82bar).Thisallowedthetestsimulationtoverycloselymatchtherangeofinitialprimarysystempressuresthat.canbeexpectedintheoperatingplant.8.1222Safet~ReliefVal'vegSRVJInordertomatchthecharacteristicsoftheSafetyReliefValve,anoriginalCrosbySRV,shippeddirectlyfromtheplantsite,wasinstalledintheteststandandusedinalltests.81223DischargeJ.ineInordertocovertherangeofdischarge'linelengthsandthereforeairvolumesthatexistinSSES,twoventclearingtestserieswererun;onewithadischargelinethatsimulatesthelongestSSESdischargeline{48m)andonethatsimulatestheshortestSSESdischargeline(35m).Inaddition,thenumberofbendsineachline,thei'nnerdiameterofthemainpartofthe'ine(303.9mm),andtheinnerdiameterofthelastverticalruntothequencher(2889mm)arecloselysimulatedtothatwhichexistsintheSSESplant.(schedule40pipeandschedule80pipe,respectively).Inadditiona24ft.submergence,correspondingtothehighestwaterlevelinthesuppressionpool,wasusedforalltests8.12.24VacuumBreakersInordertocloselysimulatetheeffectsofvacuumbreakeroperationonthetests,twosix-inchdiameterCrosbyvacuumbreakerswereshippedtoGermanyandinstalledintheteststandatthesamerelativelocationasplannedfortheSSFSplant.REVli3/798-24 PROPRIETARY81.22.5OuencherAfullsizeprototypeofthequencherinstalledintheSSESplantwasinstalledinthetestfacilityandusedforalltests.Figure8.13showsthequencherwithinstrumentationforventclearingtestswhilefigure8.14showsthequencherwithinstrumentationforthecondensationtests.8-2TESTFACILITYANDINSTRUHENTATION8.21TestFacility8.2.1.1MechanicalSet-upThetestconfigurationasconstructedistypicallyillustrateddiagrammaticallyinFigure8.4.Theteststandconfigurationcanbedividedinto:thesteamboiler,thesteamaccumulator,thesteamlinebeforetheSRVandthebuffertank,theSRV,thedischargelinebetweentheSRVandthewaterpoolwiththequencheraspipetermination,andthelargetankaswaterpool.8.2111SteamboilerThesteamboilerisanoil-fired,once-through,forced-flowboilerwithanoutputofapproximately20HWatamaximumsteampressureof170bar(2499psig)andamaximumsteamtemperatureof520~C(968~F).The.boilerisdesignedforaclosedoperatingmodeinnormaloperation.Afractionoftheboiler'soutputisrecoveredfromthecondensateviathehigh-pressurecooler.Whenthereisanopenloop{ie.,lostcondensate),theoutputisreduced.Thesteamflowavailableinthismodeisapproximately8to9kg/s(17.6to19.8ibm/s).Thelostcondensateresultsinatimelimitationoncontinuousoutput.Thefeedwatersupplyoftheboilerisabout20m3(705ft~).Oncethatamountisusedup,furthersteamsupplyascontinuousoutputispossibleonlyuptotheoutputofthefeedwaterconditioningsystem.Thatamountsto5m~/h(176ft3/h).Forlongertestperiodsitisnecessarytointerruptoperationfor4hoursinordertorefillthefeedwaterstoragetank.82.1.1.2SteamAccumulatorAsdescribedin8.2.1.1.1theamountsofsteamsuppliedcontinuouslybytheboileraretoosmalltotestanSRV.REV.1,3/798-25 PROPRIETARYToprovideawaytotestvalvesatflowratesofuptoapproximately22kg/s(484ibm/s),avalvetestfacilitywasbuiltusingtheboilerplantandapressurevesselconnectedtoit.Thisvesselischargedwithasteam/watermixturebytheboilerandisusedasasteamaccumulator.Fromthissteamaccumulator,highersteamflowratescanbedeliveredforashortperiodoftimeThedimensionsofthepressurevesselare1.5mdiameterand12mhigh,whichresultsinanaccumulatorvolumeofapproximately22m3.Adaptedtotherequiredsteamoutput,theaccumulatorisfilledwithsaturatedwaterandsaturatedsteamatthespecifiedratio.Thesteamisdrawndownwardthroughastandpipe.Thehighsteam,flowtobeextractedtransientlyfromtheaccumulatorresultsinarapiddecreaseofpressureandtemperature.Forstrengthreasons,thetemperaturedifferencebetweentheinsideandoutsideoftheaccumulatorvesselmustnotexceedacertainvalue.Thislimitsthemaximumpressuredropandthustheavailabletesttime.82.1.1.3SteamLineandBufferTankTheconnectionbetweenthesteamaccumulatorandthevalveteststandconsistsofanND250pipeline.ThislinecontainsisolatingdevicesforemergencyisolationandameasurementsectionconstructedasaVenturinozzle.Theexistingeguipmeatprovideforadirecthorizontalconnectionofthevalvebeingstudied.ThiscorrespondstothedesignoftheSRVsusedinGermanBMRplantsandtotheirarrangementattheendofataplinecomingfromthemainsteamline.ThesteamsupplylinewasrebuilttomatchthedesignfeaturesofSSES.ThepreviouslydescribedpipelinenowendsinaT-piece.XnordertosimulatetheSSESmainsteamlineandtokeepthesteamsupplyflowtothevalveasuniformaspossible,abuffertankhavingavolumeof5.2m3wasconnectedtothesecondhorizontaloutletoftheT-piece.Theverticaloutletoftheabove-describedT-pieceleadstothevalve.8.2.114Safet~ReliefValveQSRVQTheSRVusedinthetestsistheactualversionbeingusedforSSES.Thesevalvesarearrangedvertically,haveasteaminletfrombelowandanoutlettotheside.Asdescribedin8.211.3,thesteamsupplylinewasrebuiltinsuchawaythatthesamearrangementwaspossibleintheteststand.ThevalvewasmountedontheT-piece,usingthesameconnectiondimensionsasintheactualplants.REV1,3/798-26 PROPRIETARYOperationofthevalveduringthetestsrequirestheconnectionofpowersupplylines,controllinesandmeasurementlines.Theexistingequipmentatthevalvetestfacilitywasusedtosatisfymostofthoserequirements.Somemodificationsbecamenecessaryinordertoadapttotheconstructionofthevalve.TheSRVsinGermanBMRplantsareoperatedbyanelectricallyactuatedpilotvalvewithitsownoperatingmedium.Incontrast,theSS'>>Svalveusedinthetestwasopenedpneumatically.Accordingly,thecompressed-airconnectionwasrebuiltsothattheopeningconditionsintheactualplantcouldbesimulatedintheteststand.8.2~11.5DischargeLineandQuencherTheSRVdescribed.in8.2.1.1.4dischargesontheexhaust-steamsideintoapipewhichrepresentstheSRVdischargeline.ThelengthoftheSRVdischargelineandthenumberofbendsaredifferentforthe16SRV'sforSSES.Twolinelengthswereusedforthetests,correspondingtothelongestandshortestlengthsoftheSRVdischargelinesintheplant.IsometricdrawingsofthetwodischargelinesareshowninFigure8.5(longline)andFigure8.6(shortline).Pipesupportsandvibrationdampersweremountedattherequiredplaces.Theseplaceswerenotidenticaltothecorrespondingonesintheplant,becausethemountingsituationsandespeciallytheconcreteconstructionoftheplantcannotbe-simulateddirectlyinthetestfacility.Topreventthebuildupofalargeunderpressureinthepipe,twoactualvacuumbreakerswereinstalledinaverticalpartofthepipeline,asintheplant.ThequencherformstheterminationoftheSRVdischargeline(seeFigure8ll).Thesteamisconductedintothewaterthroughalargenumberofholeshavingadiameterof10mm.ThedesignofthequencherisdescribedindetailinSection4.1.AbottomsupportisprovidedtoholdthequencherinplaceinthetesttankItconnectsthequencherrigidlytothebottomofthetankandisconstructedinsuchawayastomakeitpossibletomeasuretheloadsexertedonthequencherduetoventclearingprocessesandsteamcondensation.Theslidingjointprovidedbetweenthequencherandthedischargelineintheplantissimulatedintheteststandhgdraulicallgbyacorrespondingannulargap8211.6TestTankForSSES,theexhauststeamfromthereliefvalvesisconductedintothesuppressionpoolandiscondensedthere.Xnthetestfacility,asectionofthatpoolissimulatedbyastiffenedREV-lg3/798-27 PROPRIETARYsteeltank(seeFigures8.7,88,89).Intheplant,thesuppressionpoolcanbesubdi'videdconceptuallyintosuhspaces,eachofwhichisassociatedwithasteamsupplyline(seeFigure8.3).Inordertoadapttheconditionsinthetesttanktothedimensionsofthesmallestgeometricalsinglesell,concreteshapedblockswereinsertedintothetesttank.TheconcreteshapedblocksareclearlyillustratedinFigure8.7.Theexposedcross-sectionalareaofthewaterspaceis7.2mx3.15m=22.7m~.Itcorrespondsconservativelytothesmallestindividualcellintheplant.Illuminatingdevicesandviewingportsmadepossiblethedirectobservationandalsophotographicrecordingoftheunderwaterprocesses.8.2.2InstrumentationInstrumentationisprovidedforcontrollingthetestprocedure,determiningtheprescribedmeasurementquantities,andrecordingthem.82.21GeneralDescriptionTheinstrumentationusedintheKarlsteintestfacilityconsistsofoperatinginstrumentationandtestinstrumentation.Operatinginstrumentationassuresthecontrolofthetestfacilityanditsenvironmentcorrelation.ThetestinstrumentationrecordstheloaddatawhichisusedtoverifytheconservatisminthedesignloadsasspecifiedfortheSSESinsection4.1ofthisDesignAssessmentReport.DetailsontheoperatinginstrumentationaregiveninSection8.2.2.3.AdetaileddescriptionofthetestinstrumentationcanbefoundinSection8.2.248.2.22InstrumentationIdentificationForidentification,themeasuringsensorsaredesignatedaccordingtoasystemoflettersandfigures.Thefirstoneortwocharactersareletterswhichidentifythetypeofinstrument:PTFLDGSGILPPressureTransducerTemperatureSensor(Thermocouple)FlowRateMeasurementsRaterLevelMeasurementsDisplacementGageStrainGageElectricalImpulseSignalLevelProbeREV.l,3/798-28 PROPRIETARYTheselettersarefollowedbyanumberwhichcharacterizesthelocationwithinthetestfacilitywheretheinstrumentissituated.Thefacilitywasdividedintosectionsasfollows:Section1containsthesteamsupply,includingtheaccumulator{onlytransducersoftheteststandinstrumentationsystemarecontainedinthissection).Section2containsthesteamlineuptothesafetyreliefvalveandincludesthebuffertank.Section3containsthesafetyreliefvalve.Section4containsthedischargelineandquencher.Section5containsthetesttank.Thesensordesignationiscompletedbyaddingadecimalpointandasequentialnumber.Forexample,"P5.6"means:thenumber6pressuretransducerinthetesttank.Additionalabbreviationsusedareasfollows:DPSCTCDCACFAHT-SGSRVPGRTDDataProcessingSystemCoatedThermocoupleDirectCurrentAmplifierCarrierFrequencyAmplifierHighTemperatureStrainGageSafetyReliefValvePressureGageResistorTemperatureDetector82.2.3Operating1nstrumentationTheoperatinginstrumentationisprovidedformeasurementofparametersinrelationtothesteamaccumulator,thesteamlinesandtheSRV'sAtotalof30sensorscanberecordedbyaprocesscomputerwhichispartoftheoperatinginstrumentationsystem.ThedataarestoredonamagneticdiskandcanbeprintedoutTherecordingfrequencyoftheprocesscomputerwasadaptedtoalignwiththeinstrumentationchanriels,coveringarangefrom0.5Hz,forthosesensorswhereonlysmalltransientsaretobeexpected,uptoabout200Hzforthesensorswherehigherfrequencysignalsareexpected(e.g.forpipevibrations)Theoperatinginstrumentationcomprisesthemeasuringdevicesusedtomonitorandcontrolthesystemandalsothedataacquisitiondevicesneededforthatpurpose.TypicalmeasuringlocationsforthetestsareillustratedinFigure8.4andlistedinTable8.1.BEV1,3/798-29 PROPRIETARYAccordingtothetypeofacquisitionanddisplay,themeasurementsensorscanbeclassifiedintotwogroups="DisplayonControlConsole"and"AcquisitionbyComputer".82.2.3.1Disp~la'nControlConsoleToenabletheoperatingpersonneltocontrolthetestequipment,anumberofquantitieswhichcharacterizetheoperatingconditionofthesystemaredisplayedcontinuously.Inparticular,theyare:Waterlevelin:Steamaccumulator,steamline,buffertank,dischargeline,testtankPressureinSteamaccumulator,buffertank,controlline,dischargelineTemperaturein:Steamaccumulator,buffertank,dischargeline,testtanka822.3.2A~cuisition~bComputerMostofthedatasensorscomprisingtheoperatinginstrumentationareinterrogatedbyacomputeratprescribedtimeintervalsbefore,duringandafterthete.t.Thevaluesarestoredonadisk.Thedataareprintedoutatprogrammedintervals.Ataninterrogaticnfrequencyof200Hz,thecapacityofthestoragedeviceissufficientforarecordingtimeof2minutes.Thefollowingmeasurementvaluesareinterrogated:WaterlevelSteamaccumulator,buffertankdischargeline,testtankPressureSteamaccumulator,buffertank,steamline,controlline,dischargelineTemperatureBeforeSRV,afterSRV,surfaceofSRV,dischargeline,testtankVibrationsValvetravelSwitchingtimeSteamlinebeforeSRV,dischargelineSRV,vacuumbreakersElectricalenergizationofSRVREV1g3/798-30 PROPRIETARY822.4Test,InstrumentationMesurementvaluesusedtoverifythetesttasksaredeterminedbythetestinstrumentation.Itisnecessarytoincludehereafewtypicalmeasuringpointsthatarealreadyusedformonitoringpurposesintheoperatinginstrumentationonthepipes.andSRV.Sincemostoftheseprocessesareofahigh-frequencynature,thedataisacquiredinanalogformbymeansofcarrier-frequencymeasuringamplifiersanddcamplifiersonanalogmagnetictape,andtoalargeextentalsoonvisicorders.Thevisicordertracesallowaninitialreviewandapre-evaluationofthetestdata.82.24.1MeasuringPointsMeasurementsaremadeofthepressureonthesteamlinebeforetheSRV;valveactuationandvalvetravel;pressurevariationinthedischargelineatfourpointsbetweentheSRVandquencher;temperatureinthedischargelineatthreepointsbetweentheSRVandquencher;waterlevelinthedischargelinebeforethequencherinletatfourpositionsforthelonglineandfivepositionsfortheshortline;bending,axialandtorsionalstrainsonthebottomsupport;bendingstrainsonthequencher;bendingstrainonadummyventpipe;temperaturedistributioninthetesttank;temperaturedistributionatthequencherforthecondensationtest;wallpressuresandbottompressuresinthetesttank.TypicalmeasurementpointsfortheventclearingtestsareillustratedinFigures8.7,8.8,8.9andlistedinTable8.2.TypicalmeasurementpointsforthecondensationtestsareillustratedinFigures8.10,8.11,8.12andlistedinTable8.3.8.2.24.2Set~uofMeasuringInstrumentsAllinstrumentationischannelledtoonecentralstationsituatedinthecontrolroomofthelaboratory.Eachinstrumentationchannelconsistsoftheindividualsensor,connectingcable,amplifier(carrierfrequencyamplifierordirectcurrentamplifier),attenuator;andarerecordedonmagnetictapesandvisicorders,mostchannelsbeinginparallelonbothsystems.Threemagnetictaperecordersandthreevisicorderswereusedinthecontrolroom.Eachunitallowstherecordingof12channelsand,inaddition,atimereferencesignalandaphysicalcorrelationtrace.REV1,3/798-31 PROPRIETARYThesensorsareconnectedbyshieldedcabletotheamplifiersvhicharelocatedneartherecordersinthecontrolroom.Forthestraingages,displacementgagesandpressuretransducers,carrierfrequencyamplifiersvereusedwhichallowafrequencyresolutionofupto1KHz.Fortemperaturemeasurements,directcurrentamplifiers(10Hz)vereusedtogethervitha10Hzlovpassfilter.82.25Visual-RecordingThreehigh-speedcamerasvereusedtofilmtheprocessesinthepoolduringtheblowdovnthroughthequencher.KMUusesa"HYCAM120m~'orthatpurpose.TvoLOCAMcameras(model51-0003)werebeingmadeavailablebytheStandfordResearchInstitute(SRI)Thepositioningofthecameraswasasfollovs:HYCAMcamerainfrontofonebull'seyeatquencherheight;LOCAMcamera1infrontofonebull'seyeatatankheightofapproximately4m;LOCAMcamera2ontheserviceplatformabovethetankataheightofapproximately9m.AcorrelationbetweenthemovingpicturesandthedatarecordingsontheVisicorderandmagnetictapevasaccomplishedbymeansofatimingmarkonthefi'lms.83TESTPARAMETERSANDMATRIX8.31VentClearingTestsThetestmatrixfortheventclearingtestsispresentedinFigure8.15.Thisfigureshowsthetestnumberandparameterconditionsusedforeachtest..Thenumberofbasictestswas25.These25testsweresplitinto5groupsoftestswherebyeachgroupcoveredasetoftestparameters.Testsnumbered26to32wereadditionaltestsvhichwerenotrequiredtoverifythequencherdesignbutwhichcouldproveusefulinevaluatingtheperformanceofthesafetyreliefsystem.Testsnumber27,28,30and31weretoinvestigateshorterthannormalSRVopeningtimes,but,asvalveopeningtimesverefoundtobequitefast,thesetestswerenotaddedtotherequiredtests.Testsnumber26and32,withonelockedvacuumbreaker,wereincludedintothetestmatrix.Theresultsshovedtheeffectofthelockedvacuumbreakertobeminimalsotestnumber29wasnotadded.REV1,3/798-32 PROPRIETARYTheallocationofeachtestgroupwithintheoperationrangeofthesafetyreliefsystemisshowninFigures8.16to8.21bytestpoints.BaseparametersinGroup1(Figure816)arelongdischargeline'length,normaldischargelineairtemperature,normalinitialwaterlevelinsidethedischargelineandnormalvalveopeningtime.EachofthefollowinggroupsvaryoneormoreoftheseGroup1baseparameters;Group2(Figure817)usesalowinitialwaterlevelinsidetheSRVpipe;Group3(Figure8.18)usesahighdischargelinetemperature;Group4(Figure8.19)usesashortdischargelinelengthandGroup5(Figure8.20)usesashortdischargelinelengthandahighdischargelinetemperature.Eachofthebasic25testswascomprisedoftwoormorevalveactuationswherebyonlythefirstactuationismadeatt,hespecifiedconditionsofthedischargeline(so-calledcleancondition).Anyotheractuationwasmadeattheprevailingdischargelinetemperatureandwaterlevel(so-calledRealCondition).Inthecaseofonlytwoactuationsatatestpointthetimeintervalbetweentheactuationswasapproximately10minutes.Inthecaseofmultipleactuationsatatestpointthetimeintervalsbetweenactuationswerevariedasfollows:Fortestpoints4,5,14,15thetimebetweensuccessiveactuationswasl.5/5/15/30/60/120seconds,accountingforsevenvalveactuations.Fortestpoints19and20thetimebetweensuccessiveactuationswas15/5/15/30/60/120/5/15/600seconds,accountingfortenvalveactuations.ForventclearingtestswithonlytwoSRVactuations,thehold-opentimefortheSRVwas2secondswhileforthemultiplevalueactuationteststhehold-opentimewas1.5seconds.'Fivetestpointswererepeated,theseweretestpoints4,15,19'0and25.RepeattestsatadesignatedtestpointareindicatedwithaletterRinthetestnumberi.e.Testnumber20.Rl.listhefirstvalueactuationoftherepeattestattestpoint20.11.AcompilationofactualparametersatthestartofeachtestistabulatedinTable8.4forthelongpipetestseriesandTable85fortheshortpipetestseries.8.32CondensationTestsInordertofurtherverifythesteamcondensationcapabilitiesofthequencherdeviceandprovidespecificinformationregardingitssteamcondensationcapabilitiesforthesafetyreliefsystemlREV.1,3/798-33 PROPHIETARYoperationrangeaseriesofeightextendedblowdowntestswereperformed.Thesetestsaredesignatedastestnumbers33to40.Eachtestwasperformedwiththeshortdischargelineconfigurationasdescribedinsection8.2.1.1.5andwithaninitialdischargelinetemperatureofapproximately90~C.ThelocationoftheinitialsystemconditionsforeachtestpointisplottedonthesafetyreliefsystemoperationrangeinFigure822InordertoinitiateeachtesttheSRVwasactuatedaswasdoneintheventclearingtests.Thevalvethenremainedopen-untilthesystempressurereachedthepredesignatedvalueforthattest.Atthistimethevalvewasclosedandthetestwascompleted.Thetotalallowablepressuredropintheaccumulatortankforeachinitialsystempressuredictatedthedurationofeachbiowdown.AcompilationofactualparametersatthestartofeachtestpointinthecondensationtestsmatrixistabulatedinTable8.6.84TESTRESULTSThissectionprovidesacompilationofthetestresultsfortheventclearingandsteamcondensationtestsconductedattheKraftwerkUnionlaboratoriesinKarlstein,WestGermanyinordertoverifytheloadspecificationandsteamcondensingcharacteristicsofthequencherdesignfortheSusquehannaSteamElectricStation.Includedinthissectionisinformationabouttheboundaryconditionsatthebeginningofeachtest,the'esultsofthebehavioroftheSRV,primarysystempressures,dynamicpressureloadsonthepoolboundariesandtheirprimaryfreguencyandtheloadsonthequencherandbottomsupportThisinformationisprovidedintheformoftables,figuresandactualvisicorderrecordings.841VentCleari~nTestResultsNineteentestswithatotalof67ventclearingprocesseswereperformedwiththelongdischargelineintheperiodfromMay8,1978toJune7,1978and13testswithatotalof58ventclearingprocesseswereperformedwiththeshortdischargelineintheperiodfromJune27,1978toJuly7,1978.841.1TestParametersThemostimportantoftheparametersbeinginvestigatedwasdescribedinSection8.3.AdetailedlistoftestparametersforeachvalveactuationisgivenforthelongdischargelinetestsinTable8.4andfortheshortdischargelinetestsinTable8.5.ThisincludesREV.1,3/798-34 PROPRIETARYtypeoftestlengthofdischargelineaccumulatorpressurewatertemperatureinthetesttankwaterlevelindischargelineairtemperatureindischargelineTheaccumulatorpressureP1.1AandthebuffertankpressureP2.6Aarethedeterminativevaluesforthesystempressureatthestartofeachtest.Thevalueswerereadbycomputerjustpriortothestartofthetest.Inadditionthesepressureswerestoredcontinuouslyonmagnetictape.IfalongperiodpassedbetweenthelastcomputerreadingandtheactualteststartthentheinitialvaluesfortheaccumulatorpressureweretakenfromthecorrespondingcomputerplotsTheinitialaccumulatorpressureswerealsoreadfromthoseplotsforthemultiplevalveactuationtests.Foraccumulatorpressuresbelow30bar(435psi),measuringpointP2.5wasusedtodeterminethesystempressure,sincemeasuringpointsPl.1AandP2.6Awereoutsidethemeasuringrange.Thewatertemperatureatthestartofthetestwastakeneitherfromthecomputerlistingsor,inthemultiplevalveactuationtests,fromthecomputerplotsDuetotheinertiaoftheBartoncell,themeasurementvalueforwaterlevelinthedischargeline(measuringpointL4.1)inthemultipleactuationtests,especiallyforthe2nd,3rdandifapplicable,the8thactuation,mustbedisregardedorconsideredonlyasanindicativevalue.Thetemperatureinthedischargelineatthestartofeachtestwastakenfromthecomputerlistingsorthecomputerplotsforthemultipleactuationtests841.2BehavioroftheSRVandSystemPressuresToevaluatethevalvebehavior,thevalveopeningtime,t,wasdeterminedfromtherecordedvalveliftvariationforalltests.0Thisinvolvesthetimefromthebeginningofvalveopeninguntilattainmentofthesteadystatelift(seesketchbelow).Theseopeningtimesarelisted,forthelongdischargelinetests,inTable8.7and,fortheshortdischargelinetests,inTable8.8.Theassociatedsteadystateliftsarealsoindicated.AplotofthemeasuredvalveopeningtimesasafunctionofaccumulatorpressureatthestartofeachtestisshowninFigure8.23forthelongdischargelinetestsandFigure8.24fortheshortdischargelinetests.Theso-calledventclearingtimestpzarealsogiveninTables8.7and88ThisisthetimefromthebeginningofvalveREV.1,3/798-.35 PROPRIETARYopeninguntiltheinstant'ofmaximumpressureatmeasuringpointP4.4inthedischargeline.(seesketchbelow)tsvalveliftventclearingpressurepressurebeforequencher1'wovaluesareindicatedinTables8.7and8.8forsystempressuresmeasuredin:buffertank-P2.6beforetheSRV-P2.5inthedischargeline-P4.1toP4.4Thesetwovaluesarethepressureattheventclearingtime(ventclearingpressure)andthepressureapproximately1.5secondsafterthestartoftest(steadypressure)Theinitialparametersofrelevancefortheclassificationof-testsareindicatedintherowheadings.Theventclearingpressureinthedischargelinebeforethequencherinlet(measuringpointP4.4)isplottedversussystempressure(measuringpointP2.6)underCleanConditionsinFigure8.25forthelongdischargelinetestsandinFigure8.26fortheshortdischargelinetests.SeeSection8.5.2.1foradiscussionoftheventclearingpressuresandtheirdependenceonreactorpressure.84.1.3DynamicPressureLoadsonthePoolBoundariesAsreadofftheVisicordertraces,thepeakpositiveandpeaknegativepressureamplitudesduringventclearingformeasuringpointsP5.1-P5.3(bottompressures)andP5.4-P5.10{wallpressures)arecompiledinTable8.9forthelongdischargelinetestsandinTable8.10fortheshortdischargelinetests.Ina'ddition,approximatevaluesforthepredominatefrequencyofthepressureoscillationsareindicated.Thesefrequencieswerereadfromthevisicordertraces.Figures8.27and8.28showthemeasuredpeakpositivepressureamplitudesatthetankbottomdirectlybeneaththequencher(P5.2)andontheconcretewallatthequencher'smid-heightREV1,3/798-36 PROPRIETARY(P5.10)asafunctionofsystempressureforthelongdischargelineandshortdischargelinetests.ThetestpointsplottedareallCleanConditiontestswithcoldwaterinthetesttank{approximately25~C)anddischargelinecold(approximately50~C)(Longdischargelinetests1.1,2.1,3.1,4.11,4.81.1and32.1andshortdischargelinetests16.1,171,J.8.1,19.1.1and19.R1.1)AsacomparisonFigures8.29and8.30representcorrespondingmeasuringpointsfortestsperformedunderRealCondition(Longdischargelinetestsl.2,2.2,3.2,.10.4and32.2andshortdischargelinetets16.2,17.2and18.2).AscanbeseenthepressureamplitudesareslightlyhigherfortheCleanConditiontestsandnosignificantchangewithsystempressureisobserved.Figures8.31and8.32showthemeasuredpeakpositivepressureamplitudesatmeasuringpointsP5.2andP5.10forCleanConditiontestswithheatedwater{45C-80~COinthetesttankforthelongdischargelinetestsandshortdischargelinetestsrespectively.(Longdischargelinetests5.1.1,6.1,71,8.1,9.1,151.1and15.R1.1andshortdischargelinetests20.1.1,20.Rl.l,22.1,23.1,24.1).Again,asacomparison,Figures8.33and8.34representcorrespondingmeasuringpcintsfortestsperformedunderRealConditions(Longdischargelinetests6.2,7.2,8.2,92,11.2and12.2andshortdischargelinetests20.R1.7,22.2,23.2and24.2)Incontrasttothetestswithcoldwaterinthetesttank,thepressureamplitudesareslightlyhigherfortheRealConditiontests,butaswiththecoldwatertests,nosignificantchangewithsystempressureisobserved.Figures8.35to8.40showthemeasuredpeakpositivepressureamplitudesatmeasuringpointsP5.2andP5.10foranumberofmultiplevalveactuationtestsplottedagainstthecorrespondingvalveactuation.Figures8.41to865showthefirstsecondofvisicorderpressurestraces(forthepoolboundarypressures,P5.1-P5.10)fromvarioustests.8~414LoadsOnThequencherandBottomSupportThebendingstrainsonthetwoarmsofthequencherandatthebottomsupportwereeachmeasuredintwomutuallyperpendiculardirections.Theresultantbendingstrainsandbendingmomentswerecalculatedfromtheseindividualvalues.Thestrain-versus-timevariationsstoredonmagnetictapewerereadforthemaximumresultantduringventclearing.Ahigh-passfilterhavingacutofffrequencyof2HzwasinsertedinordertoruleoutanyfalsificationoftheevaluationduetoslowdriftingofthezeropointTheupperfrequencylimitwasat400Hzduetothemechanica1conditions.REV1~3/798-37 PROPRIETARYThemaximumresultantbendingstrainsdeterminedinthismannerandthebendingmomentscalculatedfromthemarecompiledinTables8.11and8.12forthelongandshortdischargelinetestsrespectively.Toclarifythedirectiondistributionoftheresultingbendingmomentsonthequencherarms,thecomponentsofthemaximumresultantbendingmomentsaredepictedinpolarcoordinatesinFigures8.66and8.67forthelongdischargelinetestsandFigures8.68and8.69fortheshortdischargelinetests.AsshowntheresultantbendingmomentsonthequencherarmsoccurprincipallyintheverticaldirectionFigures870and8.71forthelongandshortdischargelinetestsshowacorrespondingdistributionofthemaximumresultantbendingmomentsatthebottomsupport.Tables8.11and8.12alsoindicatethemaximumtorsionalstrainsandtorsionalmomentsmeasuredatthebottomsupportandthemaximumverticalstrainsandverticalforcesmeasuredatthebottomsupportduringventclearing.Thisdataisbasedonasevaluationofthevisicordertraces.842SteamCondensationTestResultsEightcondensationtestswiththeshortdischargelinewereperformedintheperiodfromJuly18,1978toJuly.21,1978.8.4.21TestParametersThemostimportantoftheparametersbeing-investigatedwasdescribedinSection8.3.AdetailedlistoftestparametersisgiveninTable8.6.CompiledinthatTablearetheparametersatthebeginningofthetests,suchas:typeoftestlengthofdischargelineaccumulatorpressurewatertemperatureintesttankwaterlevelindischargelinewaterlevelintesttankairtemperatureindischargelineTheaccumulatorpressurePl.lAandbuffertankpressureP2.6Aarethedeterminativevaluesforthesystempressureatthestartofeachtest.Thevalueswerereadbycomputerjustpriortothestartofthetest.Inaddition,thesepressureswerestoredcontinuouslyontapebutonlyupto360secondsafterthestartoftests36.1and40.1.Thiswasdictatedbythelimitedstoragecapacityoftheoperatinginstrumentationcomputer'smagneticdisk.Thisdatawascontinuouslystoredonthevisicordertracesandthetestinstrumentationmagnetictapes.REV1i3/798-38 PROPRIETARYForaccumulatorpressuresbelow30bar(435psi),measuringpointP2.5was,usedtodeterminethesystempressure,sincemeasuringpointsPl.1AandP2.6Awereoutsidethemeasuringrange.ThewatertemperatureatthestartofatestwastakenfromthecomputerlistingsandattheendofatestfromthecomputerplotsThevaluesforthewaterlevelsandairtemperaturesinthedischargelineatthestartofatestweretakenfromthecomputerlistings.Table8.13showstherelationbetweentheteststep,testnumber,andrangesofpressureandwatertemperatureastheyactuallyoccurred.842.2PresentationofTestResultsFirstwewillpresentasurveyoftheobservedcondensationphases.Thatisfollowedbyapresentationofthedynamicpressureamplitudesinthewaterregionofthetesttank.Finallythetemperaturevariationsinthewaterregionaredescribed.8.4.221Surve~ofObservedCondensationPhasesIntheoperationfieldofthequencherasgivenbythetestmatrix,theobservedcondensationphasesareindicatedinFigure8.71forblowdownsalongtheupperandlowerboundarylinesoftheoperationfield.8422.1.1BlowdownatlowMaterTe~meratureFortheblowdownalongthelowerboundaryline,thefollowingcondensationphaseswereobservedforthetestedpressurerange:AbsolutesystemPressureinBarCondensationPhaseTests70-25Stationary33.2,34.1,35.1,andinitialsectionof36.125-2IntermittentMiddlesectionof35.12-1Inthepipe(1)Endsectionof36.1(1)Itshouldbenotedherethatatthebeginningofthisphaseaportionofthesteamflowhasemergedthroughtheannulargapabovethequencherinlet.AsnotedinSection8.2.1.1.5,REV1,3/798-39 PROPRIETARYthisannulargapsimulateshydraulicallytheslidingfitofthequencherinstalledatSSES.Figure8.73showsatypicalexampleofthemeasurementtracesobtainedwiththebottomandwallpressuresensorsforstationaryoperationofthequencherintheupperpressurerange(test33.2).Figure874showsatypicalexampleofthelowerpressure'ange(test35.1).High-frequencypressureoscillationsoccurwithverylowamplitude,andwithoutanyfixedfrequency.Toillustratetheintermittentoperation,thevariationofthebottomandwallpressuresandtwopipepressuresthroughouttheentiredurationoftest361isshowninanextremelytime-compressedforminFigure8.75.Theintermittentcondensationphaseisclearlyrecognizableinthemiddlesectionofthetest.Figure8.76showsamoretime-expandedexcerptfromthatphase.Supplementarily,Figure877showsatypicalpowerfulindividualeventinanextremelytime-expandedform.Thehigh-frequencypressurepeakssuperimposedonthelow-frequencysinusoidalpressurepulsationsareclearlydiscernibleinbothFigures875and8.76.Forthephaseofcondensationinthepipe,thetesttracesexhibitnegligiblylowamplitudes,whichareclosetotheresolutionlimitofthemeasuringchain.Therefore,noexampleofsuchatraceisshown.8.4.2212-BlowdownatHighWaterTe~meratureForblowdownalongtheupperboundaryline,thephasesdescribedin8.4.2.2.1.1wereobservedinpracticallythesamepressureranges.However,theappearanceofthepressureoscillationsdifferstosomeextentfromthatofthepressureoscillationsatlowwatertemperature.First,hereistheobservedrelationbetweenpressurerangeandcondensationphase:AbsolutesystemPressureinBarCondensationphaseTests>70-4.Stationary372e38ls39leandinitialsectionof4014-2IntermittentMiddlesectionof40-12-1Inthepipe<>>Endsectionof40.1REV1,3/798-40 PROPRIETARY(1)Itshouldbenotedherethatatthebeginningofthisphaseaportionofthesteamflowhasemergedthroughtheannulargapabovethequencherinlet.AsnotedinSection8.2.1.1.5,thisannulargapsimulateshydraulicallytheslidingfitofthequencherinstalledatSSES.Forstationaryoperationintheupperrangeofpressure,Figure8.78showsatypicalexamplefortest37.2.1helowerrangeofpressureforthisphaseisrepresentedbyanexamplefromtest391(Figure8.79).Therearealsohigher-frequencypressureoscillationswithlowandverylowamplitude,respectively,andwithoutanyfixedfrequency.Atypicalexampleofintermitten~toerationisshowninFigure8.80byanexcerptfromtest40.1.Comparedtothisphaseatlowwatertemperature(seeespeciallyFigure8.76),adistinctattenuationofthestrengthofthepressurepulsationsisobservableathighwatertemperature.Superimposedhigh-frequencypressurepeaksdonotoccur.Forthephaseofcondensationintheprie,the.testtracesexhibitnegligiblylowamplitudesevenatextremelyhighwatertemperatureofmorethan90oC.8.4.2.2.2StatisticalEvaluationofthe~DnamicPressureLoadsonthePoolBoundariesAsdescribedinSection8.4.22.1,thesteamcondensationdoesnothaveanyuniformformthroughouttheentirerangeofsystempressureandwatertemperature.Tonowbeabletoquantifythedistributionofdynamicpressureamplitudesduringablowdownfrom70bartoapproximately1bar,therecordingsfromarepresentativebottompressuresensorandwallpressuresensorforallthetestswerestatisticallyevaluated.Thisalsoallowedustoinvestigatetheinfluenceofsystempressureandwatertemperatureonthedynamicpressureamplitudes.B.a2.22.1Dependenceof~DnamicBottomandIlailPressuresonSystemPressureandMaterTemperatureThepressure-timehistoriesstoredonmagnetictapeforpressuresensorsP5.2(bottompressure)andP5.10(wallpressure)wereeachreadformaximumvalueatuniformtimeintervals.Ahigh-passfilterwithafrequencycutoffof2Hzandalow-passfilterwithafrequencycutoffof500Hzwereinsertedintothecircuit.In'hismanner,afalsificationoftheevaluationduetoslowREV1,3/798-41 PROPHIETARYdriftingofthezeropointorduetoelectricalinterferencewaslargelyexcluded.Portests33.2,34.1,351,37.2,38.1and39.1,.auniformintervalof1secondwaschosenbecauseoftherelativelyshorttestdurationofamaximumof64secondsintest39.1.Intests36.1and40.1withtestdurationsofover800seconds,theuniformintervalwas4seconds.Inthesetwotests,thephasesofstationaryandintermittentcondensationandcondensationinthepipewerecoveredseparatelyatthesametime.Noerrorwasintroducedintotheevaluationbythedifferentchoiceofintervals,sincethemaximumvalueswerecoveredineachcaseTheextremevaluesdeterminedforthepositiveandnegativedynamicpressureamplitudesatthebottomandonthewallareplottedversusthetransientvariationofthesystempressureinFigures881and8.82.Duetothelargenumberofextremevalues,aselectionwasmadewiththeaimofconsideringonlythehighervalues.ThetophalfoftheFigureshowsthemeasuredmaximumpressureamplitudesfortheblowdownathigherandhighwatertemperaturealongtheupperboundarylineoftheoperationfield.Thebottomhalfshowsthemfortheblowdownatlowwatertemperaturealongthelowerboundaryline.AsimilarillustrationforthemeasuredmaximumwallpressureamplitudesisgiveninFigure8.82.Thepeakbottom-pressureandwall-pressureloadsmeasuredduringtheindividualcondensationphasesareindicatedasafunctionofwatertemperatureinTable8.14.Promthesepeakvalues,wecanascertainaslightdecreaseofthepressurelevelwithahotpoolforthestationaryandintermittentcondensationphases.Porthephaseofcondensationinthepipe,ofcourse,therearepracticallynodiffe'rencesinthepressurelevelsforcoldandhotpool-84.22.2.2Occurrenc~efreuenceDistributionsofthe~DnanicBottomandMallPressuresInparallelwiththedeterminationofextremevaluesasdescribedinSection8.4.2.2.2.1allpositiveandnegativepeakvaluesbetweenthezeropassagesofthepressure-vs.-timevariationsweredetermined.ineachtimeintervalandclassifiedaccordingtomagnitude.Thiscountingmethod,knownas<<peakcountbetweenzeropassages"or"meancrossingpeakcountmethod>>,avoidstheinclusionandconsequentialoverassessmentofsmallintermediateoscillations.Onlytheabsolutemaximabetweentwozeropassages'reincludedinthecount.REVli3/798-42 PROPRIETARXThecountresultsuppliestheclassoccurrencefrequencydistributionatonce.Positiveandnegativepeakvaluesweretreatedseparately.Anyerrorinthecountresultsbythenoiselevelonthemagnetictapeswaslargelyeliminatedbymeansofa.prescribedamplitudesuppressionof10mV=0.015bar.Auniformclassintervalof0.025barwaschosenforthehistograms.Inthatway,thehistogramsoftheindividualtestswereabletobecombinedintoanoveralldistributionforblowdownswithcoldandhotpool.ThehistogramsofthepositiveandnegativeamplitudesofthedynamicbottompressuresatmeasuringpointP5.2areillustratedinFigures8.83and884forblowdownswithcoldandhotwater,respectively.AnalogoushistorgramsforthewallpressuresatmeasuringpointP5.10areshowninFigures8.85and886.8.4.2.2.2.3StatisticalCharacteristicsoftheDynamicBottomandMallPressuresInfluencesoftestparameterscanbereadofffromthestatisticallydeterminedmeanvalues,sincethosevaluesareobviouslymuchmoretypicalthanthemagnitudesofindividualandveryraremaximumvalues.Themeanvaluesweredeterminedbythegroupvaluemethodsusingthefollowingequation:PkZn.Pi~1KZn<illwherePG=meanvalue;Fifrequency.classmeanvalue;n=classThegroupvaluemethodwasindividualhistogramsofadistributions.Thosemeanto8.86alsousedforthecombiningoftheblowdowntogetthemutualfreqeuncyvaluesareindicatedinFigures8.83Ingeneral,thetrendsaresupportedbythemaximumvalues.Theunavoidablescatterofthemaximumvaluesisallowedforbyformingtheaveragevalueofthe10highestamplitudesineachtest.Duetothesmallnumber,theyweredeterminedbythesingle-valuemethod:wherePE=NZPi~1NREVli3/798-43 PROPRIETARYPE=meanvalue;P.=singleextremevalue;N=numberofextremevaluesTables815and8.16provideanoverviewoftheabovementionedmostimportant.statisticalcharacteristicsofthepressure-timehistoriesatthebottomandatthewall,respectivelyfortests33.2to40.1.Indicatedare:maximumvaluerelativetotheentiretest,meanvaluerelativetotheentiretest,-lowerlimitvalueofthe10highestvalues,meanvalueofthe10highestvalues.*Besidethedataconcerningthesystempressuresandwatertemperatures,thecondensationphasesarealsolisted.Intests36.1and40.1,thephasesofstationaryandintermittentcondensationandcondensationinthepipeweretreatedseparately.'igures8.87and8.88showplotsofthemeanvaluesrelativetotheentiretestortestsectionandthemeanvaluesofthe10highestvalues,asfunctionsofsystempressure.Themeanvaluesofthebottomandwallpressuresareslightlyhigherfortheblowdownwithacoldpool.Thistrend,alreadyalludedtoinSection8.4.2.2.2.1onthebasisoftheabsoluteextremevalues,isthereforeverifiedstatistically.,Thelevelofthemeanvaluesfromthe10highestvaluesishigherbya-factorofapproximately3-4thanthelevelofthemeanvaluesrelativetotheentiretestortestsection.8.4223Te~meratureVariationsintheRaterRegionoftheTestTankFourtestswereselectedtoillustratethetemperaturevariationsinthewaterregionofthetesttank:test33.2forhighsystempressureandcoldpool,test35.1forlowsystempressureandcoldpool,-test37.2forhighsystempressureandhotpool,test39.1forlowsystempressureandhotpool.Figures8.89to892showtheverticaltemperaturedistributionobtainedfromthemeasuringpointsT5.5,T5.2,T5.3andT5.4arrangedaboveoneanotherontheconcretewallIneachcase,themeasuredtemperaturesarescatteredaboutameancurve.ThescatterisgreatestformeasuringpointT52(approximatemax.a8oC).Thatmeasuringpointisattheheightofthequencherarmandisimpingedupondirectlybythesidewardsdirectedflowimpulse.ThescatterisleastformeasuringpointT5.4(approximatemax.a50C).Thescattercanbeexplainedbythehighdegreeofturbulenceinthe.pool.REV~1i3/798-44 PROPRIETARYFigures8.93to8.96showthetemperaturevariationsatquencherarm1forthesametests.AtmeasuringpointT5.8locatedinthemiddleoftheholearray(seefigure8.14)adistincttemperatureincreaseofapproximately15-200C,ontheaverage,wasrecordedrelativetothepooltemperature.Incontrast,thetemperaturesattheupperedgeoftheholearray(T5.9)andattheupperedgeoftheguencherarm(T5.10)aresomewhatlowerthanthepooltemperatureatT5.1duetoasufficient"coldwatersupply".Thisisanindicationofthegoodcirculationofwaterneartheguencher.This.confirmedtheexpectedcondensationbehaviorofthequencherasrelatedtothelayoutoftheholearray.(SeeSection4.1.1.1).84.2.24WaterLevelintheDischargeLineWhenOpeningandAfterClosi~ntheSRVInthetestswiththelongdischargeline,thewaterlevelinthepipewasmeasuredbythe"LevelProbes"LP4.1thruLP4.4atfourpositions,oneaboveanother.Inthetestswiththeshortdischargeline,thisinstrumentaitonwasextendedbythemeasuringpointLP4.5abovethemeasuringpointLP4.4;seeFigure8.8ThemeasurementsignalsfromtheseLevelProbeswererecordedonvisicordersandmagnetictape.ABartoncell,measuringpointL4.1inFigure8.4,wasusedtosetandmeasurethewaterlevelinthedischargelinebeforeteststart.ThereadingofthatmeasuringpointwasinterrogatedbythecomputerbeforeandduringthetestandwasstoredTheindicationsoftheLevelProbesandalsotheindicationsoftheBartoncellwereusedtodepictthetimevariationofthewater.levelinthedischargeline.ItmustbetakenintoconsiderationthattheresponsespeedoftheBartoncellistooslowfortherapidchangesofthewaterlevelduringventclearingandaftertheclosingoftheSRV.Themeasuringpointwasusedessentiallytodeterminethesteady-state.waterlevelsinthedischargeline.Figures8101and8.102showtwotypicalexamplesofthevariationofthewaterlevelinthepipefortheintervaltest15.1withthelongdischargelineand20.1withtheshortdischargeline.Itwasfoundthatintwoinstancesinintervaltest15.1(Figure8.101),thewatercolumnbrieflyexceededtheexternalwaterlevel,butfellbackimmediately.Thesetwotestpointsrepresentthemaximumwatercolumnrisemeasuredintheventclearingtests.Intheintervaltest20.1,thewatercolumndidnotreachtheleveloftheexternalwatersurfaceinanyinstanceafterclosingoftheSRV.Themaximumwaterlevelrisewasgenerallyfound,inalltests,tooccurafterthethirdvalveactuation.REV1,3/798-45 PROPRIETARYToevaluatetheeffectofvacuumbreakeroperationonthewatercolumnrefloodfollowingventclearing;Test32,withonelockedvacuumbreakerandatimeintervalof3secondsbetweentheclosingofthevalveafterthefirstactuationandthenextactuation,wasincluded.Figure8-105showsthevariationofthemovementofthewatercolumninTest32.Ascanbeseennoadverseeffectswererecorded.8.43CheckingandCalibrationoftheNeasuringInstrumentationThecalibrationandtheelectricalandphysicalcheckingofallsensorsbefore,duringandafterthetestswereperformedinaccordancewiththeTestandCalibrationSpecifications.Fig.897showsdiagrammaticallythephysicalcalibrationofthesensors,thesettingandcalibrationoftheamplifiersandrecordinginstruments,andthequalityinspectionofthesensors.Pig.8.98showsthetimeintervalsstiplatedforthechecksandcalibrationsintheTestandCalibrationSpecifications.Fig.8.99clarifiesthechainofthecalibrationsystemfromthenationalstandardsofthePhysikalisch-TechnischeBundesanstalt(PTB)tothemeasuringinstruments.ThepressuresensorsP5.1thruP5.10usedinthetestswerefullyoperableuntiltheendofthetests.Thelowestinsulationresistanceof1.2x10~0measuredatP5.1afterthetestscanbeclassifiedas"good".ThepipepressuresensorP4.1failedon31Nay1978Itwasreplacedbyanewsensorforthesubsequenttests.WiththisnewsensorP4.1~thelowestinsulationresistanceforthegroupofpipepressuresensorsafterthetestswas3x10~~,whichwasverygoodTherewerenofailuresforthestraingaugesSG41thruSG4.8,SG5.1andSG5.2Herealso,averygoodinsulationresistancelevelwasrecordedwithalowestvalueof3x10~uatSG4.6afterthetests.J.ikewise,noneofthetemperaturemeasuringpiontsT5.1thruT5.10failed.Thelowestinsulationresistanceof1.3x10~~wassufficientlyhigh.844AnalysisofNeasurementErrorsBasedoninformationfromthemanufacturersofthemeasuringinstruments,KWUsowninvestigations,andtakingintoconsiderationtheexperienceaccumulatedinsimilartestprojects,themaximummeasurementerrorsfortheindividualsensorscanbeindicatedasfollows:REV.1,3/798-46 PROPRIETARYPressuresensorsP5.1thruP5.10Linearityerrorofthesensor2.5%ofmeasuredvalueinrangeof0to2barError2.5'5Reproductionerrorofthesensor0.2%of5barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder0.01bar05%0.5%Max.totalerrorx[0.01bar+3.5%ofthemeasurementvalue]PressuresensorsP4.1thruP4.5ErrorLinearityerrorofthesensor0.5%ofmeasuredvalueinrangeof0to20bar0.5%Reproductionerrorofthesensor0.1%of35bar0.035barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder05%05%Max.totalerrora[0.035bar+1.5%ofthemeasurementvalue]PressuresensorsP2.3andP25ErrorLinearityerrorofthesensor1%ofmeasuredvalueinrangeof0to40barReproductionerrorofthesensor0.1%of140bar.0.14barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder0.5%05%Max.totalerrora[0.14bar+2%ofthemeasurementvalue]StraingaugesSG4.1thruSG48~SG5.1~andSG5.2ErrorToleranceoftheguagefactorInfluenceoftemperatureontheguagefactorREV1,3/798-47 PROPRIETARYErrorofthemeasuringamplifierErrorofthebalancingunitanrecorder05%0.5%Max.totalerrora5%ofthemeasurementvalueTemperaturemeasur~in~pintsT5.1thruT5.10ErrorofthesensorlocErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder05%0.5SMax.totalerrorx[l~C+1$ofthemeasurementvalue]AfterthefirsttestsonMay10,1978andafterconclusionofthetestsonAugust2,1978,additionalphysicalchecksofthepressuresensorsinthewaterregionwereperformedbyincrementalloweringofthewaterlevelinthetesttank.Themax.deviationsfromthenominalvaluewereapproximately+0.01and-0.02bar.Fig.8.100illustratesafrequencydistributionofthesedeviationscombinedfrombothchecksandforallpresuresensors.ItshowsatypicalGaussiandistribution.Inordertorecordthehigh-frequencyprocessescorrectlyinfrequencyandamplitude,thedatawasacquiredinanalogformonmagnetictape.Porasensoreigenfrequencyofapproximately30kHz,thedynamicrangewaslimitednotbythesensorsbutratherbythecarrier-frequencymeasuringamplifierslocatedfurtheroninthecircuit.Thefrequencycutoffofthemeasuringamplifierswasat1.5kHzandthatofthemagnetictaperecorderswasat2.5kHz.ThefrequencycutoffsofthevisicordersweredeterminedbytheutilizedgalvanometersThesefrequencycutoffsareapproximately1kHz.Thefrequencyresponseofeachindividualgalvanometerwascheckedpriortothetests.8.45RepetitionTestsand~ReroducibilityoftheResultsToverifythereproducibilityofthemeasurementresults,arepetitionof5testswasspecifiedintheTestMatrix.Basedonapreliminaryassessmentoftheresultsafterconclusionofthetestserieswiththelongandshortdischargelines,thefollowingtestswererepeated(asmentionedpreviously):Longline:4.1through4.Rl15.1through15.RlIntervaltestsIntervaltestsShortline:REV1,3/798-48 PROPRIETARY19.1through19.R2201through20.R125.1through25.R2IntervaltestsIntervaltestsSingleActuationtestsInadditiontotherelevantinitialconditions,Table8.17alsogivesthemeasuredventclearingpressure(measuringpointP4.4),max.dyn.bottompressures(measuringpointP5.2)imardyn.wallpressures{measuringpointP5.10)andfrequenciesofthepressureoscillationsforthefirstSRVactuationineachoftherepetitiontests(>>CleanConditionstests").Acomparisonoftheabove-citedvaluesfortherepetitiontestsassociatedwitheachotherdemonstratesthegoodreproducibilityunderCleanConditions.Themaximumdeviationsfromthemeanvalueforeachpairofrepetitiontestsare(seeTable8.18):fortheventclearingpressureforthebottomandwallpressures10.75barora6%a0.05barora7Xforthefrequencyofthepressureoscillations105Hzora7%Themeandeviationsfromthemeanvalueofrepetitiontests,averagedforall5pairsfortheventclearingpressureforthebottomandwallpressuresforthefrequencyofthepressoscillationseachpairofoftests,are:10.37baror13K1002barora6%RO2HzOrX5%Figures8.37and8.38illustratesthemax.dynamicpressuresinthepoolduringtheventclearingforthemultiplevalveactuationrepetitiontestswiththelongline.Figures8.39and840showsthesamethingforthemultipleactuationrepetitiontestswiththeshortlineIncomparisonwiththefirstSRVactuationsunderCleanConditions,somelargerdeviationsareexhibitedhereinthetestsunderRealConditions(2ndto7thand10thSRVactuations).Thereasonforthesedeviationsisthattheinitialconditionsdiffersignificantlyfromeachother.Thevisicordertracesforeach"cleancondition"actuationatarepetitiontestpointisprovided:Tests4.1.1and4.Rl.l-Figures8-41and8-42REV1i3/798-49 PROPRIETARYTests15.1.1and15Rl1-Figures8-46and8-47Tests19.1.1and19.R2.1-Figures8-48and8-49Tests20.1.1and20.Rl.l-Figures8-59and8-60Tests25.land25R2-Figures8-64and8-65Avisualcomparisonofthetracesfromeachrepititiontestalsoshowsgoodreproducibility.Accordingly,itcanbesaidthat:Iftheinitialconditionsofthetestsaresetinacontrolledmanner(CleanConditions),thenthetestresultsarereproducible.Iftheinitialconditionscorrespondtotherandomlyprevailingoperatingstates(RealConditions),thenthemeasurementvalueslieinalargerscatterrange.85DATAANALYSISANDVERIFICATIONOFLOADSPECIFICATION8.5.1EvaluationofTestTankEffectsonBoundaryPressureNeasurementsInthisSection,vepresenttheoreticalandexperimentalinvestigationswhichshowthattheKarlsteintesttankrepresentsagoodsimulationofthehydraulicconditionsoftheSSESsuppressionpool.Meareconcernedprimarilywiththeeffectsexertedontheprocessesinthevaterbytheexistingboundarysurfacessuchasthewatersurface,tankbottom,movableorimmovabletankwalls.TheresultsoftheinvestigationfacilitatetheevaluationandtranspositioncftheboundaryloadsmeasuredintheteststoSSES.85.1.1EffectsofFreeRaterSurfaceandR~iidWallsTheeffectsofthefreewatersurfaceandtherigidwallsofthetankonthefluidpressurewillbeexplainedfirstbymeansoftheexamplesillustratedinFigure8-104.ThetophalfoftheFigureshowsthevelocitypotentialandflowfieldofasphericalbubblesubjectedtooverpressureorunderpressureinaninfinitelyextended,incompressiblefluid.Thepotentialfieldisdescribedbyasimple1/rlaw(Reference35).If,forexample,thesamebubbleislocatedinacylindricalrigidtankwhichispartiallyfilledwithfluid,thenthepotentialfieldandflovfieldhaveavisiblydifferentappearance(Figure8-100,bottom).Thedifferences.inthenonstationaryfluidpressure,whichisproportionaltothevelocitypotentialforsufficientlylovflowvelocity(pressurefield=potentialfield;seeReference4forexample),areclearlyevidentinthepressureprofilesontherightsideoftheFigure8-104.Thefreewatersurfaceconstrainsthepressuretozero,vhilethecylindricalwallcausesanincreaseinglymorepoverfulpressurerisewithREV1,3/798-50 PROPRIETARYincreasingdepth.Thenarrowerthetank,thegreateristhepressurerise.ThecalculationsrelatingtoFigure8-104wereperformedbythefinite-elementsmethod{Reference34)foratankdiameterof3mandawaterdepthof6m.Thebubblewas2.8mdeepand0.8mindiameter.Besidesthepressurefield,thereisalsoaneffectonthewatermasswhichiseffectivelyentrainedbythebubbleduringpulsationmotions(pressureoscillations)andthusalsotheoscillationfrequency.InthecaseshowninFigure8-104,thebubbleinthetankhasalargercoupledmassthanintheinfinitelyextendedmedium.Thisismanifestedbythefactthatthepulsationfrequencyofthebubbleiscorrespondinglylower(seeSection8.5.3.2).8.512MethodofImagesThemethodofimagesisanimportantaidwhichmakesitpossibletoclearlyunderstandthehydraulicactionsofthewatersurfaceandrigidwallsandtocalculatethemquantitativelyinasimpleway(Reference35).Ztisbasedonthefactthattheinfluenceofaplanerigidwallontheflowfieldofahydrodynamicpointsourcecanberepresentedbyasuperpositionoftheflowfieldwithoutthewall(infinitelyextendedfluid)andtheflowfieldofanimagesourceofidenticalsignandidenticalstrengthlocatedbehindthewall(Fig.8-105).Thesameholdsforaplanefreewatersurface,exceptthattheimagesourcehastheoppositesign.Usingthismethodofimages,theflowfieldofapointsourceinarectangular,vesselisobtainedfinallybyrepeatedapplicationofsuitableimagingoperations(Figure8-105dandFigure8-2).Theimmediatesignificanceofthemethodofimagesliesinthefactthatapulsatingbubblecanbeconceivedofasahydrodynamicsource,thusprovidingasimplemethodtocalculatethepressurefield.Ofspecialimportancefortheperformanceoftestsistheconsequencederivedbyinversionofthemethodofimages:Aconfigurationofbubblesoscillatinginparallelcanbesimplifiedinatestbysurroundingonebubblewithrigidwalls.Thiswillbeclarifiedfurtherinthefollowing.851.3TheTestStandasaSincCleCellBasedontheabovediscussion,anoscillatingbubbleinarectangularvesselisequivalenttoaplanefieldofsimultaneouslyoscillatingbubbles{Figure8-2).FromFigure8-2itfollowsfurtherthatvesselswithseveralbubblesarealsoequivalent,sincebetweeneachpairofbubblestheimagingwallsectioncanalsobeomitted.REV1,3f'798-51 PROPR1ETARYApplicationofthemethodofimagestothetranspositionofasystemofvalvesblowingdownsimultaneouslyinaplanttoateststandwithaquencherleadstothecelldivisionillustratedinFigure8-3..Asdiscussedinsection8.1,thewaterspaceoftheteststandwasformedaccordingtotheinteriorsinglecellsC,F,KandN(Figures8-3and8-108),sincetheyarethenarrowestandwillthereforeexhibitthehighestwallandbottompressures.Thatcanbeseenbyobservingthat,accordingtotheimagingprinciple,theyconservatively'simulatemorequencherslyingclosertogetherthanisactuallythecaseintheSSESsuppressionpool8.51.4SpatialDistributionsofPressureintheTestTankTogetmeaningfultestresults,pressuresensorshavetobemountedatsuitablepointsinthetesttank.Aseriesoftheoreticalinvestigationswasperformedinordertobetterassesstheirarrangement.Theyconsistedofcalculatingthespatialdistributionofpressurealongthetankwallsforvariousbubbleconfigurationsunderwater.TheKRUcomputercodeVELPOTwasusedforthisinvestigation.Abubblewassimulatedbyapoint,sourcenormalizedtounitsourcestrength.TheresultsareillustratedinFigures8-107to8-109.Figure8-107showsthecalculatedwallpressuredistributionforabubbleinthreedifferentpositionsnearthequencher:Case1Sourceonthetankaxis,0.7mabovethequencheraxisCase2Sourceonthetankaxis,atquencherelevationCase3Sourceatcenterofthequencher(eccentric).Theresultsshowthat.,theeccentricarrangementofthequencherwhichbecamenecessarybecauseofspacelimitationsinthetank,includingthecorrespondingpositioningofthepressuresensors(blacksquaresinFigure8-107),results,theoretically,inslightlyhighermeasurementvaluesforthepressures.Thenextcalculation(case4,Figure8-108)servestoanswerthequestionastohowthebubble'sforminfluencesthepressuredistribution.Todothat,thesinglesourcefromcase3,figure8-107,wasreplacedbyfouridenticalsourceswiththesametotalsourcestrength.Figures8-108and8-109showthattherearenomajordifferences.Notealsothegoodagreementseenbetweenthemeasuredpressuresfromshakedowntest081andthecalculatedvaluesi'nFigure8.109.Themodelcases3and4(singlebubbleatcenterofquencherand4-bubblearrangement)arebestadaptedtotheteststandgeometry.Sincetheassociatedpressuredistributionshardlydifferatall{Figure8-109),itisdemonstratedthatanexactREV1,3/798-52 PROPRIETARYknowledgeoftheairdistributionunderwaterisnotnecessaryforacorrectarrangementofthepressuresensorsInordertodemonstratetheconservativenatureofthechosensinglecell,asalreadyexplainedinSection8.5.1.3,thepressuredistributionformodelcase4iscomparedtothedistributioncalculatedfortheSusquehannaplantinFigure8-110.ThepressuredistributionintheteststandenvelopsthepressuredistributionintheSSES.Furthermore,thepressuredistributionintheteststandisenvelopedbythespecifieddistribution{FigureB-ill).8.5.1.5InvestigationoftheInfluenceofSavableSalleontheMeasurementResults/Fluid-StructureInteracti~on8.5.151GeneralRemarksIntheprecedingdiscussion,itwasassumedthatthesinglecellhasrigidandimmovablewalls.TheconstructionoftheKarlstein.testtankissuchthatthetank,despiteaseriesofstiffeningribs(seeFigures8-10to8-12),stillhasaresidualcompliance.Thetime-varyingloadsactingduringtheblowdownofthequenchercanthereforeexcitethetankintooscillationduetoFluid-StructureInteraction(FSI).Usingexperimentalandtheoreticalinvestigations,itwillbeshownthatinfluencesoftankoscillationsonthemeasuredboundaryloadscanbeneglected.Theexperimentalinvestigationsconsisted,firstly,ofmeasuringthetank'sresponsetoashortpressureimpulsewhichwasproducedbyanexplosivechargedetonatednearthequencher(Section8.5.1.5.2).Measurementsmadeduringthestart-uptestsontheteststandthensuppliedthetank'sresponsetotheloadsoccurringduringventclearing(Section85.1.5.3).Takingintoconsiderationtheinpulseresponse,itturnsoutthateffectsoftankoscillationsattheeigenfrequenciesarenegligible.Thisstatementislaterconfirmedbycalculationsandalsoisextendedtoforcedoscillations.85.1.52Ex2erimentalInvestigationoftheTank'sNaturalOscillationsTheexperimentalinvestigationofthetank'snaturaloscillationswasperformedwithimpulsiveexcitationbyanexplosivechargeinthewaterandsimultaneousmeasurementofthedisplacementsofthewallandbottomsectionsandofthefluidpressure.ThearrangementofthechargeandsensorsinthetankisillustratedinFigure8-112.Thepositionofthechargewaschosensuchthatthespatialloadprofileinthetankmatchestheprofileoftheblowdownloadsaswellaspossible.ThechargeitselfwasastoichiometricmixtureofhydrogenandoxygenwhichREV1,3/798-53 PROPRIETARYwasignitedinaplasticallydeformableflatcontainer(Figure8-113).Eightdisplacementtransducers(WA1toWA8)wereavailableforthedisplacementmeasurements.Theywerepositionedwiththeaimofobtainingthemostusefulinformation.Thearrangementofthepressuremeasuringpointsinthewater(P5.1toP5.10,Figures8-10to8-12)wasthesameasinthelaterblowdowntests.AsfortheevaluationofthepressuretracesinSection8.5.3,transducerP5.10waschosenasreferencepressuretransducerThechargewaslocatedatdifferentpositionsnearthequencherasshowninfigure8-112,i.nordertoobtainenvelopingloadprofiles.AtypicalresultisillustratedinFigure8-114,whichshowstherecordingsfromdisplacementtransducersWAltoMA8andpressuretransducerP5.10fortestno.2(chargeinposition2).Thelowestoccurringfrequenciesarebelow1Hz,buthavenothingtodowiththetank'sresponse,butratherrepresentsashiftofthezeropointThelowesteigenfrequencyofthetankisatapproximately13HzandisseenclearlyintheresponsefromtransducersWA2andWA3oscillatinginphase.Bothgagesareseatedonthebox-shapedstiffeningringsasshowninfigure8-112.Atthewallsectionsbetweenthestiffeners(WA4andMA6)andatthebottom(WA8),thefrequenciesthatoccuraremainlybetween30and60Hz.Theoscillationsoftheflatlowerstiffenerrings(WA1andWA5)arelesspronounced.Thesmallestdisplacementsarefoundattheconcretesections(WA7),wheresomeoftheamplitudesaresmallerhyanorderofmagnitude.ThepressuresignalfromP5..10showsdistinctexcursionsonlyduringthefirst100ms.Tobeabletobetterevaluatethetank'sfrequencyresponse,themeasuredtimevariationswereZourieranalyzedandpowerspectrawereformed.Thespectraassociatedwiththedisplacementtransducersonthesteelwall(MA2),concretewall(MA7)andbottom(WA8)andthepressuretransducerPS10areshowninFigures8-115to8-118.Itturnsoutthatthepreviouslymentioned13Hzoscillationinthelow-frequencyrangeisofgreatestimportance.Theassociatedtankdeformation(eigenmode)canbederivedfromthepointcorrelationsshowninFigure8-119.There,thedisplacementsofthedisplacementtransducersWA2,WA3andMA7,filteredbyabandpassfilterat13Hz,areplottedagainsteachotheratthesametimes.Thefitlinethroughthesetofpointshasapositiveslopeinthetopgraphandanegativeslopeinthebottomgraph.Therefore,displacementtransducerWA3(steelwallaboveMA2;seeFigure8-106)-oscillatesinphasewithWA2,whiledisplacementtransducerWA7(concretewall)oscillatesoutofphase.Thismeansthatthe13Hzoscillationcorrespondstoanovalizingmotionofthewall(seeFigure8-120).REV1,3/798-54 PROPRIETARY8.5.1.53ExperimentalInves~tiationoftheTank'sRe~sonsetoVentClear~inLoadsTheinvestigationsofthetank'sresponsetoventclearingloadswereperformedduringtheteststandshakedowntests.Tomeasurethetank'sresponse,thechoicewasmadetouseonedi,splacementtransducereachonthesteelwall(WA2),ontheconcretewall(WA7)andonthebottom(WA8).TheinstrumentationisshowninFigure8-121.Test08.1representsatypicalexampleoftheshakedownteststhatwererun.ThemeasuredtimehistoriesofthewallandbottomdisplacementsandofthereferencepressureP5.10areshowninFigure8-122.Thezero-pointdriftmentionedabovewaseliminatedbyusinga2Hzhigh-passfilter.Itcanbeseenthatboththepressureandthedisplacementsoscillateatthesameprincipalfrequencyof5.1Hz.Thesteelwall(WA2)andbottom(QA8)moveinphase.Theverysmallmovementoftheconcretewall(WA7)isalmostoutofphasecomparedtothepressureP5.10.Inaddition,thedisplacementtransducerWA8recordsahigher-frequencyoscillationat30Hz.Ithasalreadybegunweaklyatteststart,thendevelopsstronglyataboutthetimeoftheventclearing',andthendecaysagainabout300mslaterThephysicalinterpretationofthe5Hzoscillationisobvious.Thepressureoscillationiscausedbythepulsationoftheairbubblewhichiscreatedduringventclearing.Atthesametime,thetankcarriesoutforcedoscillationsatthefrequencyoftheforcingforce(5Hzpulsationoftheairbubble).Thesometimesphase-opposednatureofthedisplacementsofthesteelwallandbottom,ontheonehand,andtheconcretewall,ontheotherhand,makesitevidentthattheabove-discussedovalizingeigenmodeplaysadominantrole.Theoriginoftherapidlydecaying30HzoscillationseenatWA-8attheteststartisattributedtolocalforcestransmittedthroughthedischargelineandthequenchersupportduringventclearing.'Figures8-123todisplacementtimeduringshakedownspectraldensityduringshakedownlittleinfluence30HzlocaleffecfromP510showseffects.8-125showthepowerspectraldensitiesofthehistoriesforgagesWA2,WA7andWA8measuredtest081.Figure8-126showsthepowerofthepressuretimehistoryforP5.10measuredtest08l.A,reviewofthesefiguresshowsveryfromthe13HztankeigenfrequencyorfromthetseenatWA8.Figure8-126showingthersultspracticallynoinfluencefromeitheroftheseREVli3/798-55 PROPRIETARYPromthisitcanbeconcludedthatforallpracticalpurposestheKarlsteintesttankisrigidandhasnoinfluenceonthepoolboundarypressuremeasurementsmadeduringthetests.8.5.1.54TheoreticalInvestigations.andModelCalculationsoftheInfluenceofFluid-StructureInteraction851.5.4.1ComputationModelsTheanalysisdescribedbelowtocomputetheFSIonthemeasuredpressuresintheKarlsteintesttankwasperformedbyusingtheKWUcomputercodeKOVIBlAwhichwasdevelopedoriginallyandusedsuccessfullyfortheanalysisoffluid'-structureinteractioninthewaterpoolofKWU's69ProductLineBWRPlant.Theunderlying-theoryfollowsfromauniformformulationofthemechanicalprocessesbasedonpotentialtheoryandclassicalLagrangeandynamics.ItunifiesthedynamicsofthebubbleandtheFSIbyusingtheresultsofmodalanalyses.Inparticular,thefeedbackeffectsbetweenbubbleandstructureviathefluidareinc1uded.8.5.1542ModelParametersand~InutforCalculationsWithoutThemodelparametersandinputquantitiesforcalculationsoftheairbubbleoscillationsintherigidtankare:airmassflowintothebubble,watertemperature(=airtemperatureinstationaryequilibrium),hydrostaticpressureatbubbleposition,hydrodynamicmassparameterofthebubble,spatialpressuredistribution,initialvalues,(hubbleradius,etc.).Thetotalairmass(integratedairmassflow),watertemperatureandstaticpressureatthebubblepositionareobtainedfromthetestdata.Thehydrodynamicmassconstantofthebubbleandthespatialpressuredistributionareobtainedfromthecorrespondingpotentialcalculations(Figure8-107,case1).Thetimevariationoftheairsupplyintothebubblewasadjustedheuristicallybymeansofsystematictrialanderror,inparallelwiththeinitialvalues,insuchawaythatthecalculatedandmeasuredtimevariationsofthepressureattransducerP5.10exhibitedoptimalagreementThestart-uptest08.1wasusedasreferencetestforthesecalculations.TheairmassflowdeterminedinthismannerisillustratedinFigure8-127REVli3/79~8-56 PROPRIETARY8.5.15.43NodelParametersandXnautforCalculationswithPdjJustasforthedeterminationoftheairsupplyintothewaterpool,asemiempiricalmethodisusedforthestructuraldynamicsdata.Theyaredeterminedonthebasisoftheeigenfrequencymeasurementsdescribedpreviously.Inputdataforthecalculationare:eigenfrequency,modalmass,modalweight,dynamicpressuredistribution.Basedontheimpulseresponseofthetank(Figures8-115to8-117),itisplausibletoselecttheoscillationmodelyingat13Hz.Thatfixesthefrequency.Themodalmasscannotbetakendirectlyfromtheexperiment,butrathercanbedeterminedindirectlyviathemeasuredunitdisplacementsofthewall.TheunitwalldisplacementisillustratedinFigure8-128.Itisobtainedfromdisplacementsatthedisplacementtransducersbybandpassfilteringat13Hzandplottingsimultaneousvaluesofdisplacementwhicharenormalizedto1atthewatersurface.Thedisplacementdirectionisdefinedaspositiveiftherelevantwallsectionmovesinward.Thehydrodynamiccomponentofthemodalmass,(coupledwatermass)isthencalculatedbymethodsofpotentialtheory.Themodalweight,whichisequaltotheintegralloadrelativetothemodalmassandaveragedovertheunitdisplacement,isbasedontheloaddistributioncalculatedforcase1(Figure8-107,centeredbubble).Thedynamicpressuredistribution(seeFigure8-129)isobtainedfromtheunitdisplacement.bymeansofpotentialcalculations.85.1.5.4.4ResultsoftheFSIcalculationsTheresultsofthecalculationsconcerningtheinfluenceofFSIareshowninFigures8-130and8-131.Figure8-130showsthecalculatedtimevariationofthepressureatpressuretransducerP5.10,firstintherigidtank(withoutFSI)andthenintheelastictankwiththe13Hzeigenfrequency.Thereisaveryslightreductioninthepressureamplitudes,butitiscertainlynegligibleincomparisontothescatterofthemeasurementvaluesthemselves.AsisevidentfromFigure8-131,thefrequencyinfluenceofFSIalsocanbeneglected.InthatFigure,theoscillationfrequencyofthebubbleisplottedagainstthebubblevolume.ThebubblehasaslightlylowerfrequencywithFSIeffectsincludedthanwithout.REV1,3/798-57 PROPRIETARYAphysicallyclearexplanationoftheveryslightFSIeffectsfoundintheKarlsteinTestTankcanbeobtained.bycomparingthevolumesoffluidwhicharemovedbytheoscillatingwallandbottomandbythepulsatingbubble.Forabubblevolume{longline)of2.2m~andpressurefluctuationsofx0.4bar{seeFigure8-126),thevolumechangeofthebubbleisapproximately1m~isentropically.Incontrasttothis,fordisplacementslikethosefoundinFigures8-124and8-125thewallsandbottomuseuponlyabout0.05m~,whichisonly5%ofthewatervolumecomingfromthebubble.Therefore,duetothecomplianceofthetank,95%ofthewaterflowsupwardinsteadof100%(rigidtank).Thus,theresultoftheexperimentalandtheoreticalFSIinvestigationsisthateffectsofthecomplianceoftheKarlsteintesttankwallsandbottomonthepressureloadsmeasuredontheboundariesofthetankduringthetestscanbeneglected.852VerificationofSHVSystemLoadSpecificationDuetoSRVActuationThepressuresinsidetheSRVdischargelineweremeasuredatfourmeasuringpoints:justbehindtheSRVatmeasuringpointP4.1,inthecenteroftheblowdownpipeatmeasuringpointP4.2(measuringpointP4.5fortheshortdischargeline),justabovethenormalwaterlevelatmeasuringpointP4.3,andjustbeforetheinletofthequencheratmeasuringpointP4.4(seeFigure84).ThelongandshortdischargelinesareillustratedinFigures8-5and8-6.ThemeasuredpressuresinthedischargelinearedocumentedinSection8.4.1.8.5.21PressuresDuri~ntheVentClearincCProcessTypicalmeasurementtracesofthepressuresinthedischargelineareshowninFigures8-132and8-133.TheventclearingpressureisreadoffatP4.4.AsdiscussedinSection8.41,theventclearingpresureisdefinedasthepressurewhichisreadoffatthefirstpressuremaximumatP4.4.Atypicalfeatureofthispressurevariationisthedynamicovershootofthepressureabovethestationaryvalue.This-phenomenondoesnotoccurinsuchapronouncedmannerattheotherpressuretransducersalongthedischargeline.Thisdynamiceffectindicatesthatthepressurerequiredtoexpelthe~atercolumnisgreaterthan.thepressurenecessarytobringthesteammassflowthroughthequencher.Theexpulsionofthewatercolumn,isalsoclearfromthedifferenttimevariationsatP4.3andP4.4.Thepressureat8EVli3/798-58 PROPRIETARYmeasuringpointP4.3(abovethevatercolumn)risesmuchmoresteeplythanthepressureatmeasuringpointP4.4(insidethevatercolumn)Thedifferencebetveenthetvopressurosisthepressurewhichisnecessaryfortheaccelerationofthevatercolumn.Atthetimeoftheventclearing,thetwopressureshaveapproximatelyequalvalues.Butaftertheventclearingtheydifferagain,thistimeduetothedifferentpressurelossescausedbyflowresistancesinthepipe.8.521.1VentClearingPressuresfortheL~onLineThesteammassflowthroughtheSRVisapracticallylinearfunctionofthestagnation'pressure(reactorpressure).Sincethesteammassflowisoneofthemainparametersforthepressurebuild-upintheairregionofthedischargelineandthusfortheaccelerationofthewatercolumn,wewillplotthepressuresinthedischargelineasafunctionofreactorpressure.Thepressureinthebuffertank(P2.6)andnotthepressureinthesteamlinebeforetheSRVisusedasthereactorpressureforthetestssincethepressureinthebuffertankmorecloselysimulatestherepresentativestagnationpressureinthereactor.(seeFigure8-134).Todescribethedependenceoftheventclearingpressureonthereactorpressure,onlythosetestsforwhichtheinitialconditionsweresetandthusknovnexactlywereused.Thosearethetestsvithso-called<>cleanconditions".FromFigures8-135and8-136,itcanbeseenthatthemeasurementresultshavegoodreproducibilityforthetestswithcleanconditions.Thepressuresinthepipeincreasepracticallylinearlywithreactorpressure.Thefollowingtrendscanbeobserved:1)Aloweredwaterlevelinthedischargelineresultsinloverpressuresduringtheventclearing.2)Ahotpiperesultsinhigherpressuresduringtheventclearing.Thisisduetothesmallerpercentageofcondensationonthepipewall.3)Thepressure(atthetimeofventclearing)behindtheSBVisalwayshigherthantheventclearingpressureclosetothequencher.Thedifferenceisattributabletotheflowlossalongtheline.REV1,3/798-59 PROPRXETARY4)*Thepressure(atthetimeofventclearing)behindtheSRVincreases-withincreasingreactorpressure{orincreasingsteamflowratethroughthereliefvalve).Besidestheclean-conditiontests,thereisalargenumberofreal-conditiontestsandintervaltests.SincetheinitialconditionsinthemwererandomandwerenotvariedinacontrolLedmanner,themeasurementvaluesarescatteredoveramuchwiderbandthanintheclean-conditiontests.Hence,thesetestsarenotusablefortrendanalyses,butmaybeusedforverificationofmaximumspecificationvalues.Themeasuredmaximumvaluesare:PressurebehindtheSRV(atventclearingtime):19baratareactorpressureof72barVentclearingpressurebeforethequencher:14.5baratareactorpressureof72bar852.12VentClearingPressuresfortheShortLineFigures8-137and8-138showthemeasuredpipepressuresplottedagainstreactorpressureforcleanconditiontestswiththeshortdischargeline.Thesametrendsasseenwiththelonglineareseenhere.'Sincetheshortlinehasasmallerairvolumethanthelongline,whilethewatercolumntobeclearedandotherparametersremainthesame,thepressuresintheshortlinearehigherthanthoseinthelongline.Themeasuredmaximumvaluesare:Pressurebehindthereliefvalve{atventclearingtime):22baratareactorpressureof73barVentclearingpressurebeforequencher:18baratareactorpressureof73bar.85.213Tran~sositionoftheMeasurementValuestoSSESandComparisonwiththeDesicCndecificationTheverificationtestsinKarlsteinwererunwiththeactualgeometryofthereliefsystem,theactualSRV,andthehighestwaterlevelinthedischarge-line(6.2mabovecenterofquencher)thatoccursforSSES.Themeasuredventclearingtimesforthatwaterlevelandahighreactorpressure(69-81bar)wasbetween250and400msREV.1,3/798-60 PROP3IETARYFortheseventclearingtimes,theopeningtimeoftheSRV(measuredopeningtimes:29-60ms)hasnonoticableeffectontheventclearingpressure(seeFigure8-139).Hence,inregardtotheventclearingpressure,theonlyvariablewhosemaximumvalueforSSESwasnotcompletelycoveredwasthereactorpressure.Thefollowingextrapolationappliesforthat:a)PressurebehindthevalveatventclearingtimeTheMeasuredmaximumvalueforthelonglineis19baratareactorpressureof72barASlopeof25%isseeninfigure8-135.Extrapolatingto88bar,theresultis:Pox=23barforthelonglineTheMeasuredmaximumvaluefortheshortlineis22baratareactorpresssureof73barSlopeof25%isseeninfigure8-137.Extrapolatingto88bar,theresultis:Pax=26barfortheshortlineThedesignvaluegiveninSection4.1.2.1is550psi=37.93bar.TheKarlsteintestsdemonstratethatthedesignvalueisveryconservativefortheventclearingcase.b)Ventcleari~npressureThemeasuredmaximumvalueforthelong.lineis14.5barforreactorpressureof72barASlopeof12.5%isseeninfigure8-136.Extrapolatingtoareactorpressureof88barresultsinPmax=16.5barforthelongline.Themeasuredmaximumvaluefortheshortlineis18baratareactorpressureof73bar.ASlopeof12.57isseeninfigure8-138.Extrapolatingtoareactorpressureof88barresultsinPmax=20barfortheshortline.ThespecificationvaluegiveninSection4.1.1.2isPmax27barTheKarlsteintestsdemonstratethatthespecificationvaluefortheventclearingpressureisveryconservative.85.2.2PressuresDuri~ntheStationaryCondensationofSteamAboutonesecondaftertheopeningoftheSRV,theventclearingprocessiscompletedandthephaseofsationarysteamcondensationbegins.Inthisphase,thepressuresinthedischargelinearedeterminedbythesteammassflowandtheflowresistance.SincethesteamREV1,3/798-61 PROPRIETARYmass.flowisproportionaltothereactorpressure,hereagainwewillinvestigatethedependenceofthepipepressurestothereactorpressure.85.221Lo~nLineFigures8-140and8-141showthedependenceofthesteadystatepressureonthereactorpressure.Meseethattherelationcanberepresentedverywellbyastraightline.Asaresultofpipefriction,thestationarypressurebehindtheSRVhashighervaluesthanthepressuresjustbeforethequencher.Italsoexhibitsafasterincreasewithreactorpressure.Themeasuredmaximumvaluesare:17.5baratreactorpressureof72barforthepressurebehindtheSRV(P41)>l0baratreactorpressureof70barforthepressurebeforetheinlettothequencher(P4.4)852.22ShortLineFigures8-142and8-143showthedependenceofthesteadystatepressureonthereactcrpressure.Thebehaviorofthepressurebeforethequencher(P4.4)ispracticallyidenticalfortheshortlineandlongline.Thisisnotsurprising,sincethispressuredependsonlyontheflowresistanceofthequencher.ThepressuresbehindtheSRVarelowerthanthoseforthelongline,butdisplaythesameincrease,withreactorpressure.Thedifferentflowresistancesofthetwodischargelinesaremanifestedhere.Toclarifythiseffect,thevariationofthestationarypressureatthemeasuringpointsalongthedischargelineareplottedinFigure8-144fortheshortandlonglines.Theaveragepressureswereused,i.e.,thepressureswerereadofffromtheinterpolation1'inesat88bar(seeFigures8-190to8-143).Themeasuredmaximumvaluesfortheshortlineare:PressurebehindtheSRV(P4.1)16baratareactorpressureof72bar,and15baratareactorpressureof63barREV1,3/798-62 PROPRIETARYPresurebeforeinlettothequencher(P4.4)9.5baratareactorpressureof71bar,and9.0baratareactorpressureof65bar.85.2.2.3TranspositionoftheMeasurementValuestoSSFSandcomparisonwiththeDesignSpecificationAswasthecasewiththeventclearingpressure,theonlyvariablewhosemaximumvalueintheSSESwasnotcompletelycoveredbytheteststandwasthereactorpressure.Anextrapolationofthemeasuredmaximumvaluestoareactorpressureof88baryieldsthefollowingresults:a)LongLineThemeasuredmaximumvaluebehindtheSRVis17.5baratareactorpressureof72bar.ASlopeof22%isseeninfigure8-140Extrapolatingto88bar,theresultis:Pm~~--21barThemeasuredmaximumvaluebeforequencherinletis10baratareactorpressureof70bar.ASlopeof16%isseeninfigure8-141.Extrapolatingto88bar,theresultis:Pmmx=13bar.b)ShortLineThemeasuredm'aximumvaluebehindtheSRVis16baratareactorpressureof72barand15baratareactorpressureof63bar.ASlopeof22%isseeninfigure8-142.Extrapolatedto88bartheresultis:Pmax=19.6barand20.5bar,respectively.Themeasuredmaximumvaluebeforequencherinletis9.5baratareactorpressureof71barand9baratareactorpressureof65bar.ASlopeof16%isseeninfigure8-143.Extrapolatedto88bar,theresultis:P=12.5barand13.0bar,respectively.maxItcanbestatedthatthedesignvalueof550psi=37.93barforthestationarypressurebehindthevalveisveryconservative.85.23ExternalLoadsonthe~uencherandBottomSup2ortInthisSectionweshalldiscussthemeasurementresultswhichprovideinformationabouttheexternalloadsonthequencherandREV1,3/798-63 PROPRIETARYbottomsupport.ThemeasuringpointsprovidedforthatpurposeareshowninFigure8-13,andareasfollows:SG41/42SG43/4SG45/46SG47SG48Bendingatquencherarm1Bendingatquencherarm2BendingatthebottomsupportLongitudinalstrainatthebottomsupport'orsionatthebottomsupportStrainsweremeasuredatallmeasuringpoints.Themeasuredstrainswereusedtocalculatetheloadswhichproducedthestrains.Theloadsthuscalculatedarestaticequivalentloadswhichcontainhydraulicandalsostructural-dynamicaleffects.85.2.31VerticalForce8.5.23.11MeasurementoftheVerticalForceTomeasuretheverticalforce,twostrainconnectedinsuchawaythattheymeasureverticalforces.Thefollowingrelationexistsbetweenthegauges,SG4.7,werestrainsresultingfromloadandstrain:F~A~E~eBF=33~ekNBwhereA~.016m252F~2.06x10N/mmIfweinserteinpm/m,wethengettheverticalforceinkN.Thisequationwasusedtoconvert'hemeasuredstrainsintoverticalforces.8.5.2.3.1.2MeasuredVerticalForcesFigure8-145showsatypicalmeasurementtracefortheverticalforceItincreasesrapidlyduring'theexpulsionofthewatercolumnand,afterreachingthemaximumvalue,returnsquicklytozero.8523.1.21Lo~nLineTheverticalforceexhibitsastrongrelationshipwithventclearingpressureasshowninFigure8-146Thisholdstrueforalltests,eventhosewithrandominitialconditionssuchastherealconditionsandmultipleactuationtest.AsdiscussedinSection8.5.21.3,theventclearingpressureisinturninfluencedbythereactorpressure,initialwatercolumninthe'dischargeline,dischargelinetemperature,etc.andwasREV13/798-64 PROPRIETARYextrapolatedouttoamaximumreactorpressureof88bar.Therefore,themaximumverticalloadwillbeextrapolatedtothemaximumventclearingpressurefromSection8.5.2.1.3.Themeasuredmaximumvaluefortheverticalforceis:149kNata128barventclearingpressure.8.5.2.3.1.22ShortLineFigure8-147illustratesthedependenceoftheverticalforceontheventclearingpressure.Inprinciple,thesamediscussionasinSection8.5.2.3.1.2.1forthelonglineappliesherealso.Themeasuredmaximumvaluefortheverticalforceis:192kNata168barvent-clearingpressure.Theverticalforcesrelativetotheventclearingpressurearepracticallythesame.85.2.3.1.3TranspositionoftheMeasurementValuestoSSESAswasdiscussedpreviouslyfortheextrapolationoftheventclearingpressures,themeasurementvaluesfortheverticalforcecanalsobetransposeddirectlytotheplant.Forverificationofextremeconditionsintheplant,themeasurementvaluesareextrapolatedtoareactorpressureof88bar.Theextrapolationcanbeperformeddirectlyviatheventclearingpressure.8.5.23.1.31LongLineThemeasuredmaximumvaluewas:149kNata12.8barvent-clearingpressureSlope=13kN/bar(Figure8-146)AccordingtoSection8.5213,theextrapolatedvent-clearingpressureforthelonglinewas16barExtrapolationoftheverticalforceto16baryields:Fymax=190kN85.23.1..3.2ShortlineThemeasuredmaximumvaluewas:192kNat168barvent-clearingpressureSlope=13kN/bar(Figure8-147)AccordingtoSection8.5.2.1.3theextrapolatedventclearingpressurefortheshortlinewas20bar.REV1,3/798-65 PROPRIETARYExtrapolationoftheverticalforceto20baryields:FvmxInadditionFigure8-147,showsameasuredvalueof149kNata12barvent-clearingpressure.Thisleadstoamaximumextrapolatedverticalforceof:F~~y=252kN8523.13.3SummaryTheextrapolationofthemeasurementresultsfortheverticalforceyieldsama'ximumvalueof:F~~~~=,252kN>>InFigure4-11,thespecifiedverticalforceisgivenas860kN.Dnthebasisofthemeasurementresults,thespecificationvaluecanbeviewedasextremelyconservative,bothinthemaximumvalueandalsointheload-versus-timefunction.8.5.232'ore'ionalMoment8~2~3.2lMeas~nementoftheTorsionalMomentTomeasurethetorsionalmoment,twostraingauges(SG4.8-Figure8-13)wereconnectedinsuchawaythattheymeasurestrainresultingfromtorsionalmomentonly.AccordingtoReference41,thereisaverysimplerelationbetweenthetorsionorshearstrainandthemeasuredstrain,whenthestraingaugesaremountedata45oanglerelativetotheprincipalshearstressdirection.Qehave:YmsshearstrainThereforeisincethestraingaugesSG4.8weremountedata45oinclinationtotheverticalaxis,vehave:Gz=shearstressGmsshearmodulusY=2.eandr~Da2REV1,3/798-66 PROPBIETAHYIpg=torsionalmomentI=polarmomentofinertiaPr~outsideradiusofthetwistedcylindricalbarYrG'PQethusobtaintherelationbetweentorsionalmomentandmeasuredstrax.nTheshearmodulusisdefinedasG2(1+p)MithE=2.06x10sN/mm~andDap=poisson'sratioMeget:'p=0~3G~7.9x10'/mmeThepolarmomentofinertiaisdefinedas7f~D(1-D/D)4p32Therefore:I4.64x10mInsertingthevariousnumericalvalues,weget:0.41',InsertingE.inMm/m,thisequationgivesusthetorsionalmomentinkN-mHEVlg3/798-67 PROPRIETARYThisequationwasusedtoconvertthemeasuredstrainsatSG4.8intotorsionalmoments.Thetorsionalmomentsobtainedinthismannerrepresentstaticequivalentloads.85.232.2MeasuredTorsionalMomentsFigure8-148showsatypicalmeasurementtraceforthetorsionalmoments.Aftertheendoftheventclearingprocess,(approximately1secondafterteststart)theamplitudesofthemeasuredtorsionalmomentsareverysmallcomparedtothemaximumamplitudeduringtheventclearingprocessThereisafactorof6-7differencebetweenthetwoofthem.Themaximumamplitudeofthetorsionalmomentoccursmuchlaterthantheexpulsionofthewatercolumn.8.5.2.32.2.1L~onLineThetorsionalmomentatthebottomsupporthas,itsoriginonlyinunsymmetricalprocessesatthequencherduringtheventclearingandduringthetransitiontostationarycondensation.iFigure8-149showsthedependenceofthetorsionalmomentontheventclearingpressure.Sincetheventclearingpressureisadirectinfluencingparameter(seeSection8.5.2.3.1.2.1)wewillcorrelatethetorsionalmomentwiththatvalue.Thesharplypronouncedscatterbandisanindicationthatarandomprocessissuperimposedonthatdependence.Thatisexpressedbythefactthatthetorsionalmomentisbroughtaboutbyrandomunsymmetry.Themeasuredmaximumvalueofthetorsionalmomentis:M=55.8kN-mata14barvent-clearingpressure.TmaX$.523.2.g2ShortLinePigure8-150againshowsthedependencesofthetorsionalmomentontheventclearingpressure.Inprinciple,thesituationisthesameasintheprecedingSectionfo-thelongline.Themeasuredmaximumvalueofthetorsionalmomen'tis:39.2kN-mata18barvent-clearingpressure.8.523.2.3Tran~sositionoftheMeasurementValuestoSSESShentransposingthemeasurementresultstoSSES,weshallconsiderinaconservativemannertheloadcarriedbythedischargeline,whichintheteststandisconnectedrigidly(butREVli3/798-68 PROPRIETARYnotinaleaktightmanner)tothequencherandbottomsupportbymeansofweldbrackets(seeFigure8-13and8-14)incontrasttothefreemovingslidingjointatSSES.Todothat,wemaketheassumptionthatthedischargelineisfixedinatorsionresistingmanneratthefirstbendabovethequencher.Thatresultsinthefollowingpicture:DischargeLineQuencherBottomsupport////ThetorsionalmomentN~actsatthequencher.ThetorsionalmomentN~~wasmeasured.atthebottomsupport.ThedischargelinecarriesthetorsionalmomentM><.Therefore:+M2Promtheequalityoftherotation,weget:Therefore:"TVYGIpGTl11G~Ipl="T2'2'2G~Ip2Tl=P1~22T2P211Mehavethefollowingdimensions:r=0.1775mlar1g0.125m0.45m1r=0.162m2a2g~0.3.445m~11.313m2REV1,3//798-69 PROPRIETARYTherefore:-444.64.10mZ4.0,.10"m4Therefore:-26.6Tl4.640.16211.313M240'17750'4526.6TTlŽT2Tl(1+1)26.6M1.0376M1Thus,theloadtransmittedtothedischargelineislessthan4gofthattransmittedtothebottomsupport.If,withouttakingintoconsiderationthedischargeline,wefirstusePigures8-149and8-150asthebasisforanextrapolationofthemeasuredmaximumvaluestomaximumvent-clearingpressureforthecorrespondingdischargeline,thenwegetthefollowingmaximumvalues:a)longlineMr,~=598kN-mb)short.lineMr)~ax=43.2kN-mIfwenowconsiderthetorsioncarriedbythedischargeline,thenthisvalueisincreasedtoamaximumof:"ri~x=62kN-mThetorsionalmomentspecifiedin4.1.2.6fortheguenchersupportwas40kN-mtobeappliedasastepfunctionAtorsionalmomentstepfunctionappliedtoanundampedonemassREVl,3/798-70 PROPRIETARYoscillator(quencheractingasinertialmassandbottomsupportasatorsionalspring)correspondstoamaximumresponseof:M<<=2(40)kN-m=80kN-mSincethemaximumtorsionalmomentderivedfromtheKarlsteintestsisM<=62kN-m,thespecificationisconservative./8-5.2.33Bean~inncnenteattheguenchecAten85.2.33.1MeasurementoftheBendingMomentsIntheKarlsteintests,thebendingmomentsveremeasuredinthehorizontalplane(paralleltothetank'sbottom)andalsointheverticalplane,atbothofthequencnerarms.Toaccomplishthat,twostraingaugeseachvereconnectedinsuch.awaythattheymeasuredunsymmetricalstrainsresultingfromnormalstresses(unsymmetricalcomponent).Thefollowingstraingaugesweremountedforthatpurpose(seeFigure8-13:SG4.1)MomentsinverticaldirectionSG43)SG4.2)MomentsinhorizontaldirectonSG44)Thestraingaugesweremountedapproximately150mmfromtheweldbetweenthequencherarmandthecentralball.Thesectionmodulusofonequencherarmis:3W+D(1--)aQehave:'D~0.4064maD=0.3744ma=cE=M/WM=cEWThisleadstotheequationbetveenquantities:M=0.38-c8-71 PROPRIETARYThisgivesthebendingmomentinkN-m,ifcisinsertedinpm/m.Withthisequation,allthemeasuredbendingstrainswereconvertedintobendingmoments.Thebendingmomentsthuscalculatedarestaticequivalentloads.8.5.23.3.2MeasuredBendingMomentsFigure8-151showsatypicalmeasurementtraceofthemeasuredbendingmomentsatthequencherarms.Weseeclearlythatthemaximumvaluesoccurmuchlaterthantheclearingofthequencher.Theevaluationoftheindividualbendingmomentsrelatestothetotalresultantbendingmoment,ie.,thebendingmomentwhichactuallyloadsthe.quencherarm.Theresultantbendingmomentisobtainedbyusingtherelationship:M~'gM+M2x'esyzThebendingmomentsMgarereadoffatSG4.2and4;4.ThebendingmomentsMzarereadoffatSG4.1and43Theresultantbendingmomentsexhibitnodeterministicdependenceontheventclearingpressure,asshovn.inFigure8-152.Therefore,theresultantbendingmomentsonthequencherarmsmustbeconsideredasstatisticalvalues.Themeasuredmaximumvalueofthereultantbendingmomentis63kN-m.e8.52.33.3TranspositionoftheMeasurementResultsintotheWeldInSection4l.2.5,thebendingmomentsintheweldwerespecified.IntheKarlsteinteststand,thestraingaugesweremountedabout150mmfromtheweldinordernottomeasurelocalizedstressesduetotheweldandtheintersectionbetweentheballcentralbodyandthequencherarm.Availableexperienceindicatesthatthisdistanceissufficienttomeasureastressprofilewhichisindependentofshapefactors.Fromthespecifiedforceandmoment(Table4-10),weobtainforthedistancebetweentheweldandtheforceproducingthebendingmoment:lp~~Oo6551929Bytreatingthequencherarmasacantileverbeam,weobtainforthemaximumstressandthusforthemaximumbendingmoment:g0.655=Mg(0.655-0.15).REV~1>>3/798-72 PROPRIETARYM=bendingmomentintheveldBmaxM-=measuredbendingmomentBmaxTherefore:=1.297MBmaxBmeasThus,basedonthemeasuredmaximumresultantbendingmomentof62KN-m(seeSection852.3.3.2),weobtainthefollowingmaximumbendingmomentintheweld:'aximumresultantbendingmoment:81kN-m852.33.4~SecifiedStaticEguivalentLoadsAsalreadynotedabove,themeasuredbendingmomentsaretobeconsideredasstaticeguivalentloadsInSection4.1.2.5Table4-10,two'ontributionswerespecifiedwithrespecttothebendingmomentintheweld:a)astepfunctionhavingastepheightof19kN-mb)amaximumdifferentialpressurevhich,accordingtoSection4.1.3.7,is08barfromKKBtraceNo.35witha0.5multiplier.Thisresultsinamaximumdifferentialpressureof0.4bar.Thecontributionofthedifferentialpressureistobeviewedstatically,since,accordingtoSection413.5,thefreguencyofthedifferentialpressureisapproximately6Hz.Thebendingeigenfreguencyoftheguencherarmisontheorderof100Hz.Thecontributionofthedifferentialpressuretothebendingmomentintheweldisthus:11.4kN-mThecontributionofthestepfuncionistobevieweddynamically.Therefore,thesameconsiderationsareapplicableasthosemadeforthetorsionalmomentsinSection8.5.2.3.2.3.Accordingly,wehavethefollowingstaticeguivalentloads:ComponentinoneDirectionContributionfromstepfunction=2X19=38KN-mContributionfromdifferentialpressure=11.4KN-mTotal=49.4KN-mREVli3/79'-73 PROPRIETARYResultantMomentContributionfromstepfunction=38x~2=53.7KN-mContributionfromdifferentialpressure=11.4KN-mTotal=65.1KN-m8.5.23.3.5FvaluationoftheMeasurementResultsAsalreadymentionedinSection8.5.23.3.2,thebendingmo'mentsonthequencherarmaretobetreatedasstatisticalvalues.Figure8-153showsthefrequencydistributionofthemeasuredmaximumbendingmomentsineachtestsandtheresultingfrequencydisrihutionofthevaluestransposedintotheweld.Thefrequencydistributionsarebasedonthepeakmaximumvalueofeachindividualtest,whichweremeasuredeitheratSG4.1/4.2oratSG4.3/4.4.Thespecifiedstaticequivalentloads(seeSection8.5.2.3.3.4areintroducedfor7000responsesofthereliefvalve.Therefore,theloadsaretobeevaluatedinafatigueanalysis.ItfollowsfromFigure8-153thatthemeanvalueofthemeasuredmaximumvaluestransposedintotheweldis35kN-m.Exceptforthreecases,thespecifiedresultantbendingmomentsalsocoverthemaximummeasuredvalues.Thequencherisbeingevaluatedforthesemeasuredmaximumvalues.Itshouldbenotedthatboththespecifiedstationaryinternalquencherpressureof22.0barandtheresultingthermalloadof219~Cwerefoundtobeveryconservativewhencomparedtothemaximumextrapolatedvaluesof13.0barandtheresultingsaturatedsteamtemperatureof195~Cmeasuredduringthetests.(Section8.5.2.23).8.5.23.4BendingMomentsattheBottomS~ugort852.3.4.1MeasurementoftheBending.MomentsTomeasurethebendingmomentsathebottomsupport,twostraingaugescapableofmeasuringthebendingstrainsweremounted.In,themeasurementarrangement,thebendingstrainscouldbemeasuredintwomutuallyperpendiculardirections(seeFigure8-13).Thestrainsformomentsaboutthex-axisweremeasuredwiththestraingaugesSG4.5.Thestrainsformomentsaboutthey-axisweremeasuredwiththestraingaugeSG46.REVlg3/798-74 PHOPRIETARYThesectionmodulusofthebottomsupportis:D4W=-D(1--)332a4aW~1.307x10m-33Wehavea~E~@~M/WThisleadstotheequation:M~0.27~cThisequationgivesthebendingmomentinkN-m,ifcisinsertedinpm/m.Thisequationwasusedtoconvertallmeasuredbendingstrainsofthebottomsupportintobendingmoments.Thebendingmomentsthuscalculatedarestaticequivalentloads.8.52.3.4.2MeasuredBendingMomentsInFigure8-151,thebendingmomentsatthebottomsupportcanbeseenunderthetracesofthebendingmomentsatthequencherarms.Themaximumvaluesoccuratalatertimethantheventclearing.Buttheyoccuratthesametimeasthemaximumvaluesofthebendingstrainsatthequencherarms.Themaximumstrainresultingfromtorsiondoesnotoccuratthetimeofthemaximumbendingstrain(seeFigure8-151,SG4.8).Theevaluationofthebendingmomentsrelatestotheresultantbendingmoment,i.e.,thebendingmomentwhichactuallyloadsthebottomsupport.Theresultantbendingmomentisobtainedbyinterconnectingtheactualload-versus-timefunctionsoftheindividualcomponentsthroughtherelation:ThebendingmomentsMzarereadoffatSG4.5andthebendingmomentsM>atSG4.6Themaximumresultantbendingmomentwas54.5kN-mTheresultantbendingmomentsdisplaynodependenceontheventclearingpresure,asshowninFigure8-154.Hence,thesameconclusionsthatweredrawnforthebendingmomentsatthequencherarmsareapplicablehere,also.BEV.1,3/798-75 PBOPRIETABY8.5.234.3SpecifiedStaticEquivalentLoad.Asalreadymentioned,themeasuredbendingmomentsaretobeviewedasstaticequivalentloads.Thebendingmomentsatthebottomsupportareintroducedthroughthequencher.Section4.12.4andTable4-7specifyatransverseforceof44kNonthequencherwasusedasstepfunction.Inaddition,amaximumdifferentialpressureof0.4baronthequencherwasspecified.Thecontributionresultingfromthedifferentialpressureistobeviewedasastaticallyactingload.Itamountsto48kN.Note:Thedischargelineandthebottomsupportwerenotconsideredhere.Thepresssuredifferencewasformulatedonlyovertheprojectedareaofthequencher.Thespecificationthenyieldsthefollowingtransverseforcesonthequencher:Contributionfromstepfun'ction=2x44=88kNContributionfromdifferentialpressure=48kNTotal=136kNStraingaugesSG4.5andSG4.6weremountedapproximately0.5mbelowthecenterofthequencher.Transposedtothislocation,thespecificationyields:68kN-m85.23.44EvaluationoftheMeasurementResultsFigure8-155showsthefrequencydistributionofthemeasuredmaximumbendingmomentsatthebottomsupport.Themeasuredmaximumvaluesarealsocoveredbythespecification.Thus,theKarlsteintestshavedemonstratedthatthespecifiedtransverseforcesonthequenchercanbeviewedasveryconservative.85235ForcesontheQuencherIntheKarlsteinQuencherTests,onlybendingmomentswereabletobedeterminedforthequencheritself.InSection4.1.2,forcesandmomentsonthequencherwerespecified.Thespecifiedmomentswerecalculatedfromtheforces.Themeasuredmomentsarewithinthespecification.Therefore,wecanconcludethattheforcesarealsoverified.REVli3/798-76 PROPRIETARY85.23.6InfluenceofanAdgacentQuencherDuringtheclearingofthequencher,strongturbulencesandeddiesoftheexpelledandambientwaterdeveloparoundthedischargingquencher.Inparticular,aftertheventclearingthequencherissurroundedbyalargenumberofairbubbleswhichrepresentalocallycompressiblevolumeinthewater.Thisstate,whichformsaroundthedischargingq<<niche>ipreventseffectsfromtheblowdownofanadjacentquencherfrompenetratingtothequencherunderconsideration.Itisthereforeunderstandablethat,intheKMUinplanttestswithintheBrunsbuttelandPhilippsburgnuclearpowerplants,noincreaseoftheloadonthequencherandbottomsupportwasfoundfortheresponseofseveralquenchersincomparisontotheresponseofonequencher(Reference6).Aneffectofaloadononequencherduetothefiringofanadjacentquencheristobeobservedonlywhentheadjacentquencherblowsdownalone.Inthatcase,adetailedevaluationwasmadefortheBrunsbuttelblowdowntests(Reference38).Theresultoftheinvestigationwasthatthemeasuredloadsareenvelopedbyapressuredifferenceof0.2barappliedovertheadjacentinternalstructuresinthepoolatthequencherlevel,i.e.,alsooverthequencher.Amaximumpressuredifferenceof0.4baroverthequencherarmswasspecifiedforSSES.TheventclearingpressuresanddynamicpressuresinthewaterpoolobtainedforSSESfromtheKarlsteintestsareofthesameorderofmagnitudeasthecorrespondingmeasurementresultsinBrunsbuttel.Therefore,thespecifieddifferentialpressureof0.4baroverthequencherarmscanbeviewedasconservativelyenveloping.8.5237LoadsontheQuencherDuringSteamCondensationThemaximummechanicalandthermalloadsonthequencherduringthecondensationphaseoccurduringthephaseofintermittentcondensation.InSection4.1.2.7,theloadsresultingfromintermittentcondensationweretakenasthebasisforthefatiguedesignofthequencher.TheevaluationoftheloadsonthequencherduringsteamcondensationintheKarlsteinteststhereforerelatesprimarilytothephaseofintermittentcondensation.REV1,3/798-77 PROPRIETARY8.5.2.371Manifestation/ormsofIntermittentCondensationintheKarlsteinTestsAsdiscussedinSection8.1.3,thecondensationtestsvereperformedalongthelowerandupperboundarylinesoftheoperationfieldforwatertemperatures<30~Candalsoforwatertemperatures>590C.Inbothregions,theintermittentcondensationphaseoccursforverylowreactorpressures(approximatelybetween2and4bar).InSection84.2itisshownthatthemaximumvaluesforthedynamicpressuresinthevaterregionoccur.duringintermittentcondensationincoldvaterThesameistruealsofortheloadsonthequencher.Foritheevaluationandcomparisonwiththespecification,weusethemeasurementvaluesofthebendingmomentsatthequencherduringtheintermittentcondensationinthecoldpool.ThemeasurementvaluesaredocumentedinSection8.4.2.85.23.7.2IllustrationoftheMeasurementValuesThetimedurationoftheintermittentcondensationinthecoldpoolwasabout100seconds.Thetotalnumberofcondensationeventsatthequencherwas52.ThemaximummeasurementvaluesoccurredintheverticaldirectionatSG4.3.Thefrequencydistributionoftheresultantbendingmoments(SG4.'3/4.4)atthequencherarmisshowinFigure8-156.Themeanvalueofthemaximummeasurementvaluesofeacheventis11.8kN-m.Themaximummeasuredvaluewas66.5kN-m.Thefrequencydistributionoftheresultantbendingmoments(SG4.5/4.6)atthebottomsupportisshowninfigure8-158.Themeanvalueofthemeasurementvaluesis89kN-m.Themaximumvaluewasapproximately30kN-mThemeasuredmaximumvalueofthetorsionalmomentduringtheintermittentcondensationis6.2kN-m.8.52373EvaluationoftheMeasurementResultsfortheOuencherArmFigure8-157showsthefrequencydistributionoftheresultantbendingmoments,whichweretransformedfromthemeasuringpointintotheweld(seeSection8.5.2.3.3.3.Themeanvalueofthesebendingmomentsis15.2kN-m.Themaximumvalueis86kN-m.Themeasuredbendingmomentsrepresentstaticequivalentloads.,InSection4.1.2.7andTable4-12,avalueof25.4kN-mwasspecified.fortheequivalentloadfortheresultantbendingmomentintheweldduringintermittentcondensation.Theloadsspecifiedareformulatedforanoccurrencefrequencyof106.REV.li3/798-78 PROPRIETARYInthefatigueanalysis,themechanicalloadsrepresentonlyoneloadcomponent.Anotherpartofthefatigueloadingisproducedbythealternatingthermalloading.Theassumptionmadeinthespecificationwas106temperaturestepsfrom35~Cto133~Candfrom133~Cto35oC.Thelow-frequencyoscillationsofthepipe'sinternalpressuremeasuredatP4.4areusedasabasisforthemeasuredtemperaturealternation.Thesaturated-steamtemperaturesarethencorrelatedwiththosepressures.Thepressureoscillationshaveanoscillationfrequencyofabout0.5Hzandamaximumamplitudeof05baroverpressure=approx.2barabsolutepressure.Thispressureliesbelowthespecifiedvalueof3bar.Themeasuredmaximumpressureof2barcorrespondstoasaturated-steamtemperatureof120~C.AssumingthattheinflowingwaterinSSESisatatemperaturecfatleast35~C,thenthetemperaturestepis85C.Atemperaturestepof98~Cisassumedinthespecification,sothatthereisareserveof13~C.Themeasurementvaluesformingthebasisfortheevaluationandcomparisonwiththespecificationwereobservedonlyduringthephaseofintermittentcondensationwithcoldwaterinthetesttank.Aswiththeboundarypressuresinthetesttank(Section8.4.2),theloadsonthequencherwereconsiderablylowerduringtheintermittentcondensationphasewithwarmwaterthanduringintermittentcondensationwithcoldwater.Themeasuredmaximumbendingmomentduringthiscondensationphasewas(1kN-mrelativetotheweldseam.Inaddition,KMUinplanttestsintheBrunsbuttelnuclearpowerplantshowedthat,forapoolwatertemperatureofapproximately35~andabove,intermittentcondensationloadsonaquencherweresmaller.Thisindicatesthattheregionwh'ereintermittentcondensationloadsofanyconsequencecanbeexpectedislimitedtothatofverylowpooltemperatures(approximately25~C)andverylowsteammassflowsandthatheatingofthepoolasmallamountresultsinareductioninloading8.52.3.7.4EvaluationoftheMeasurementResultsfortheBottomS~uportAnimpulsivelyactingtransverseforceof17.5kNwasspecifiedonthequencherforintermittentcondensation.REVl~3/798-79 PROPRIETARYThedistancefromthemiddleofthequenchertothemeasuringpointforthebendingmomentsatthebottomsupportis0.5m,sothatthespecifiedbendingmomentwithrespecttothebottomsupportis:(175kNx2)x0.5m=17.5KNm(staticequivalentload)Themaximumresultantbendingmomentfromthetestsisapproximately30KN-m.l85.2375EvaluationoftheMeasuredTorsionalMomentsAnimpulsivelyactingtorsionalmomentof19kN-mwasspecifiedfortheintermittentcondensation.Thisstepfunctionyieldsatorsionalmomentof:38kN-masthestaticequivalentloadThespecifiedtorsionalmomentsconservativelyenvelopthemeasuredmaximumvalueof6.2kN-m.852.3.76EvaluationoftheMeasuredMaximumMomentsattheQuencherArmduringIntermittentCondensationAmaximumresultantbendingmomentof665kN-matthequencherarmwasmeasuredintheintermittentcondensationphase,whichresultsinamomentof86kN-mintheweld.Themeasuredmaximumvaluesoftheresultantbendingmomentsatthequencherarmduringintermittentcondensationareontheorderofmagnitudeofthemeasuredmaximumvlauesduringtheventclearingphase(Section8.5.2.3.3.2).Fortheventclearing,atemperaturedifferenceof184~Cwasspecified.Fortheintermittentcondensation,atemperaturedifferenceof98~Cwasspecified.Thetotalstressesloadingthequencherarmarecomposedofmechanicalandthermalstresses.Thethermalstressesaredistinctlylargerthanthemechanicalstresses.Themaximumresultantbendingmomentatthequencherarmforintermittentcondensationexceedthevaluespecifiedfortheventclearingbyabout40%.,However,theassociatedtemperaturejumpisonlyabouthalfaslargeasfortheventclearing.REV-1,3/798-80 PROPRIETARY853VerificationofSuppressionPoolBoundaryLoadSpecificationDuetoSRVActuationInSection4l.3,threepressuretimehistoriesarespecifiedasthebasisforthecontainmentanalysisduetoSRVactuation.ThethreetracesveretakenfromalargenumberofbottompressuretimehistoriesfromvariousKKBinplanttests.TheevaluationofthepressureoscillationmeasurementsintheKarlsteinventclearingtestswillthereforeconcentrateondemonstratingthatthepressuretimehistoriesspecifiedareenveloping.Accordingly,analysisandassessment.oftheindividualmeasuredpressuretimehistoriesisrestrictedtoaminimum.8.5.3.1EvaluationoftheLocalEffectsSeenatPressureTransducerP5.5AsshowninFigures8-10to8-12,thepressuretransducerP5.5ismountedontheconcretewalloppositethemiddleoftheholearrayonthequencherarm.About0.25secondsafterexpulsionofthewatercolumn,P5.5,incomparisonwiththeotherpressuretransducers,exhibitshigh-frequencypositivepressurepeakswhicharenotobservedattheneighboringpressuretransducers.Thiseffectisfromthelocalturbu1ences.Thesehighfrequencypressurepeakshaveasmallenergycontentsothattheirrangeofactionislimitedtotheimmediatevicinityofthepressuretransducer.'hefollowingTableshouldmakethisclearInthisTable,theratioofthemeasuredpressureamplitudesoftheneighboringpressuretransducers{P5.10andP5.4)tothepressuremaximumatP5.5isindicatedforalltestsvhichexhibitedamaximumpressureamplitude>1baratpressuretransducerP5.5.REV1,3/798-81 PROPRIETARYp5.lOPS'+TestP54P5.10P5.5P5.4/55P5.10/P5.5(bar)(bar)(bar)4165.1.710R1.720Rl920Rl1025125R20,60,551,00,450,4li00,730,551,70,450,4.lr01,00,651,730,550,61,00,85081,550,60,450,430,450,580~550,550,550,40,320,40,380,60,52FromthisTablewecanseethatthemeasurementvaluehasdecayedbyhalfatabout1mfromthemeasuringpointP5.5.ThecomparisonmeasurementpointsP5.4andP5.10areintheregionoforiginationoftheairbubbleoscillation,sothatnoattenuationeffectduetodistanceeffectscouldoccuratthatmeasuringpointTherefore,thesharpdecreaseofthepressureamplitudewhichismeasuredneverthelessshowsclearlythatthepressuremeasuredatpressuretransducerP5.5islimitedtoitslocalvicinity.AsfurtherverificationthatthiseffectislimitedtotheareaaroundpressuretransducerP5.5,acomparisonismadebetweenthepowerspectraldensitiesfromP5.5andthebottompressuretransducerP5.2.REV.1,,3/798-82 PROPRIETARYThefollowingtestsvereselected:Test11.1ThistestexhibitedthehighestpowerspectrumatthedominantfrequencyTest4.1.6ThistestexhibitedthehighestpressureamplitudeatP5.5forthelongdischargelineTest20.R1.10ThistestexhibitedthehighestpressureamplitudeatP55fortheshortdischargeline.Thecomparisoncanbesummarizedasfollovs:Atthedominantfrequency,thepowerdensitiesarethesamemagnitudeforthepressureoscillationsatthebottompressuretransducerP5.2andat.pressuretransducerP5.5.Thedifferencesatthehigherfrequenciesissignificant.Fortests4.1.6and20.R1.10thefrequencyspectrumofP5.5exhibitssignificantlyhigherpowerdensitiesathigherfrequenciesthanthecorrespondingfrequencyspectrumatpressuretransducerP52.Thissignificantfactorisnotnotedforthefrequencyspectrumoftest11.1(seeFigures8-159and8-160).Inthattest,thedifferencebetweenthemaximumpressureamplitudesforpressuretransducersP5.5andP5.2was013bar.ThepressureratioisP55/P52=0'8/065=12.Intest4.1.6,thedifferenceinthepowerdensitiesatthehigherfrequenciesisalreadymorestronglyevident(seeFigures8-161and8-162).Inthattest,thedifferencebetweenthemaximumpressureamplitudesforP5.5andP5.2was0.5bar.ThepressureratioisP5.5/P52=1/0.5=2.Thedifferenceinthepowerdensitiesatthehigherfrequenciesisquitestronglypronouncedintests20.Rl.10(seeFigures8-163and8-164).ThedifferenceinthemaximumpressureamplitudesforP5.5andP5.2was1.1barinthattest.ThepressureratioisP5.5/P5.2=1.73/0.63=2.75.Thepressuredifferencesorpressureratiosarenotdiscernibleinthepowerspectraforthedominantfrequencies,butareatthehigherfrequenciesFromthatwecanconcludethatthepressureoscillationwhichwasmeasuredatpressuretransducerP5-5hasapproximatelythesameamplitudeatthedominantfrequencyasthepressureoscillationsvhichweremeasuredelsewhereinthevicinityofthequencher,eg.,atP5.2Inaddition,higherfrequencypressureoscillationcomponentshavingahighamplitudeareoccasionallysuperimposedonthefundamentaloscillationinthepressureoscillationsatP55.Thehigherfrequencycomponents,vhichoccuratpressureREV1,3/798-83 PROPRIETARYtransducerP5.5,decayrapidlyintimeandspace,sothattheeffectofthehighfrequencypressureoscillationsremainslimitedtotheimmediatevicinityofmeasuringlocationP5.5Therefore,asstatedbefore,themeasurementresultsforthedynamicpressuresatP5.5representlocaleventshavingnoglobaleffectonthecontainment.WewillthereforenotconsiderthepositivepressuremeasurementsatP5.5whenverifyingthedesignspecificationfortheoverallcontainmentanalysistheresultsfromthisgageareincludedfortheverificationoftheloadingsonthecolumns.8.5.3.2Verificationofthe~SecifiedPressureA~mlitudesandVerticalPressureProfilesafterVentClearingThemeasuredpeakpressureamplitudesforthe125ventclearingtestsaretabulatedinTables8.9and810.Section8.4.1alsopresentsanumberofFigures(8.27to8.34)whichshowthatthepressureamplitudesmeasuredinthetestshadnosignificantdependenceontheinitialreactorpressure.Therefore,nomodificationtothemeasuredpressureswillbemadetoaccountfordifferencesinthereactorpressurebetweenSSESandtheKarlsteinteststand.Inaddition,asexplainedintheprevioussection,thepositivepressuremeasurementsaP5.5willnotbeconsideredwhenverifyingthedesignspecificationfortheoverallcontainmentanalysis.8532.1OverpressuresThemaximumoverpressureamplitudemeasuredontheboundaryoftheKarlsteintesttankwas1.0barThatpressurewasmeasuredattheconcretewall(p5.4)intest20.R1.10.Amaximumpressureamplitudeofl.2barisspecifiedinsection4.1.3(KKBPressureTraceNo.35withthe1.5multiplier).Themaximumspecifiedoverpressureamplitudeof1.2bar.evelopsthemeasuredmaximumoverpresureamplitudeof10bar.8.5.32.1.1VerticalPressureProfileItcanbeassumedthatthemaximumdynamicpressurevilloccurinaspherewhichsurroundsthequencherandhasapproximatelytheradiusofaquencherarm,(5'-0").Atsomedistancefromit,themaximumvaluewillbeattenuatedinaccordancewithadistancelaw.Foraninfinitewaterspace,the1/Rlawisapplicableforthedecreaseofthepressurewithdistancefromthesource.Thatlawappliesinalldirections,i.e,intheverticaldirectionalso.Thevalidityofthe1/Rlawisbasedontheassumptionofastationary(i.e.,fixedposition)oscillatingbubbleintheinfinitewaterspace.Thatidealcasedoesnotholdfortheclearingofthereliefsystem.Alreadyshortlyaftertheexpulsionoftheair-steammixture,BEV1,3/798-84 PROPRIETARYsmallairparticlesmovetothesurfaceofthepoolbecauseofbuoyancy.Evenmoreimportant,however,isthefactthatthewatersurfaceandthetankboundarysurfacesinfluencethedistancelawandthatthepressureamplitudemustvanishatthewatersurfaceitself.Accordingly,specifiedin6e0(183m)<hatheight,surface.apressureprofileintheverticaldirectionisSection4.1.3.4providingforaconstant'pressureatabovethesuppressionpoolsbottomand,startingata,lineardecreaseofpressureuptothewaterFigure8-165showsthatthemaximumspecifiedpressuredistributionveryconservativelyenvelopsthemeasuredmaximumpressureamplitudes.Theconservativenessbecomesclearlyevidentif,basedonthemeasuredmaximumvalueofwallpressureamplitudeof1baratpressuretransducerP5.4,weassumealineardecreaseofpressurefromthatmeasuringpointtothewatersurface.Thatassumedlinearpressuredecrease(depictedinFigure8-165byadashedline)alsoenvelopsthemaximumpressureamplitudesmeasuredintheverticaldirection.Incomparisonwiththeassumedlinearpressuredecreaseandthespecifiedpressuredistribution,theconservativenessofthespecificationbecomesobvious.~'-5-3-~2..2VerticalPressureProfileIucluainulocalEffactsatP5.5Fortheevaluationoftheunpertubedpressuredistributionintheverticaldireciton,themeasuringpointP5.5wasomitted,eventhoughitliesinadirectlinewiththepressuretransducersP5.4,P5.6andP5.7.BecauseofthelocaleffectforP5.5,aseparateanalysisshallbeperformedhere.ThatanalysisstartswithanestimationoftheverticalzoneofinfluenceassociatedwiththepressurepeakmeasuredatP5.5.Thelateralholesinthequencherarmsextendoverananglerangeof72~oneachside.Theholesaredrilledradially,sothatinfirstapproximationwecanassumeasourceflowoftheemergingfluid.Thehigh-frequencypressurepeakatP5.5occursatamuchlatertimethantheventclearing.Itcanbesupposedthatatthattimethereisasteam-airmixtureflowingoutofthequencher.Thesteam-airjetsemergingfromtheholeshaveahighdegreeofturbulence.Thus,theedgesareverysoonmixedwiththesurroundingwater.Furthermore,theemergingsteamiscondensedimmediatelyandtheexpelledairiscooleddownquickly,sothattheexpelledcompactvolumeisreducedrapidly.Thereforetoestimatetherangeofaction,itisassumedthatthesourceflowactsoverameananglerangeof8=e/2=720/236o.ThetotalrangeofactionisthenREVli3/798-85 PROPRIETARYb=xtan36~x=1.575m{distancefromcenterlineofb=1.14mquencherarmtoconcretewall)Thisrangeofactionof'1.14misdividedintoequalpartsaboveandbelowthemeasuringpointP5.5,sothatweobtainarangeofactionof10.57mrelativetothemeasurementlocationBasedonthisrangeofactinthemeasuredverticalpressuredistributionconsideringthelocaleffectiscomparedwiththespecifiedpressuredistributioninFigure8-166.ThebasepointsofthepressureelevationatP5.5wereplacedonthestraightlineofthelinearpressuredropsymmetricallywithrespecttothequencher'scenterplane.FromFigure8-166itcanbeseenthatthemaximumspecifiedpressuredistributionresultsinalargerresultantforceonthecontainmentboundaryandcolumnsthandoesthemeasuredpressuredistributionincludingconsiderationofthelocaleffectThismeansthattheoverallspecifiedpressuredistributrionintheverticaldirectionalsoenvelopesthelocalpressureelevationatp5.5.8.5322Unde~rressuresThemaximumunderpressureamplitudemeasuredontheboundaryofKarlsteintesttankwas-0.68bar.Thatpressurewasmeasuredatheconcretewall.{P5.10)intest25.R2.Amaximumunderpressureamplitudeof-0.56barisspecifiedinSection4.1.3{KKBPressureTraceNo.76withthe1.5multiplier).The'nextlargestunderpressurerecordedduringtest25.R2was-050bar.Thenextlargestunderpressurerecordedanywhereduringtheventclearingtestswas-058baratP5.2intest25.1.Exceptforthetwomeasurementvaluescalledoutaboveallothermeasuredunderpressureswerehounded.bythemaximumspecifiedvalueof-0.56har.8.5.322.1VerticalPressureProfileFigure8.167showsaplotofthemaximumspecifiedunderpressuredistributionandthemaximummeasuredunderpressurevaluesfortheKarlsteintests.Itcanheseenthat,exceptfortheonevalueatP510fortest2S.R2,themaximumspecifiedpressuredistributionenvelopsthemaximummeasuredpressureamplitudes.REV.1,3/798-86 PROPRIETARYInaddition,forSSES,themostunfavorableboundaryconditioninthiscomparisonisthelowliquidlevelof22ft=6.70minthesuppressionpool.ThehydrostaticpressuredistributionwithrespecttothatliquidlevelisindicatedbyadashedlineinFigure8-167.Thecomparisonofthemeasuredworstunderpressuredistributionviththehydrostaticwaterloadresultingfromtheworstboundaryconditionforthiscomparison(lowestwaterlevelinthesuppressionpool)showsthatthecompressiveforcesfromthewaterloadandthetensileforcesfromtheunderpressuredistribution=maintaintheequilibrium."Thus,theKarlsteintestshave,inaddition,demonstratedthattheblowdownoftheSSESreliefsystemviththequencherdoesnotresultinanyresultanttensileforcesonthesteelliner,evenfortheworstpossiblesuperposition.85.33VerificationofthePressureTimeHistoriesUsedfortheSSESContainmentAnalysisXnordertoverifythatthepressuretimehistoriesusedfortheSSESdynamicanalysisduetoSRVactuationarebounding,thePowerSpectralDensities(PSDs)ofthespecifiedtimehistories(withtheappropriateamplitudeincreaseandfrequencyrangefromSection4.1.3)arecomparedwiththePSD'softheappropriatetimehistoriesrecordedintheKarlsteintesttankandtransposedtotheSSES"suppressionpool.Statementsconcerningtheclearingofparallelquenchersarebasedontheunrealisticandextremelyconservativeassumptionthattheexpelled,airbubblesareequallylargeandoscillateinphase.Aquantificationofthatconservativenessisnotgiven.Mevillfirstdiscussandverifythetheorytobeusedtotransposetheoscillationfrequenciesmeasuredinthetesttanktothesuppressionpool.Then,theappropriatemultipliersforthisfrequencytranspositionwillbeestablished.Adiscussionisalsoprovidedfortransposingthemeasuredpressureamplitudestothesuppressionpool.Finally,theactualverificationispresented.85.33.1Tr~ansositionmethodfortheOscillationFrequencyThetheoreticalbasisforthetranspositionofthepressuretimehistoriesmeasuredintheKarlsteinteststotheSSESsuppressionpoolisprovidedbytheKMUcomputercodesVELPOTandKOVIBlAByusingthetestresultsfromthePPGLquenchertestsinKarlstein,theGKMquenchertests,andthenon-nuclearhottestsintheBrunsbuttelnuclearpowerplant(KKBhottests),weshallfirstconfirmexperimentallythecorrectnessofthetranspositionBEV1,3/798-87 PROPRIETARYtheory.Thatisfollowedbyacalculationofthefrequenciesforthefollowingthreeblowdowncases:(1)Simultaneousblowdownofall16quenchers(2)Simultaneousblowdownofthe6quenchersrelatedtotheautomaticdepressurizationsystem(ADS)(3)BlowdownofoneouterquencherForeachcase,acomparisonofthetheoreticallycalculatedfrequencieswiththefrequenciesmeasuredintheteststand)providesanumber(frequencymultiplier)bywhichafrequencymeasuredintheteststandmustbemultipliedinordertogetthecorrespondingfrequencyintheSSESsuppressionpoolAfactorfortheinfluenceofthesuppressionpooloverpressureisalsodeterminedinthesameway.Thecorrespondingmeasuredpressuretimehistoryistransposedtotheplantbydividingbythisfactor853.3.1.1CalculationofMeasuredOscillation~r~cruencies85.33.1.11PPGLTestsatKarlsteinSinceitwasfoundthatFluid-StructureInteractionintheKarlsteintesttankhasnosignificantinfluenceonthemeasuredpressuretimehistories,itissufficienttocarryouttheanalysisforarigidtank.Thecomparisonofcalculatedandmeasuredoscillationfrequencieswillbebasedon'theassumptionofequalbubblevolumes.ThemeasuredoscillationfrequenciesaretakenfromTables8.9and8.10.Theassociatedbubblevolumeswerecalculatedfromthetestdata,usingtheformula:pp-piieeee'e)]~eo>[P-cP(7T[p-P(T'satpool)]TpipepipePpipePsatCTpoolpipefreepipevolume(ms)pressureinpipe(bar)hydrostaticpressureatthequencherlocation(bar)saturationsteampressure(bar)relativehumidity(s=1at1005)watertemperature(oC)meantemperatureinpipe(oC)TheaveragingofthetemperatureinthepipeisperformedbyusingtheformulaEi1NpipeiwherethepipewasdividedintoNequalsections.ThetemperatureTintheithsectionwasobtainedbyinterpolationbetweenthemeasuredtemperatures.REV.li3/798-88 PROPRIETARYThecomparisonbetweenthemeasuredandcalculatedbubblefrequencyisshowninFigures8-168and8-169inwhichthebubblepulsationfrequencyisplottedversustheequilibriumvolumeatstaticpressure.PorthemeasurementpointsinFigure8-168itwasassumedthatdryairwasinthepipepriortotheteststart,whilewetair(100%humidity)wasassumedinFigure8-169.Ingeneral,goodagreementisfoundbetweenthetheoryandmeasuredfrequency.However,wecannotoverlookthefactthatthemeasuredfrequenciesinfigure8-168)arehigherthanthecalculatedones,especiallyforsmallbubblevolumes.Thismayberelatedtothefactthattheactivevolumeofairunderwaterisactuallysmallerthanthevolumefoundfordryairfromthetestdata.Thisishintedatbythecalculationofthebubblevolumeundertheassumptionof100%humidityinthepipe.Therethemeasurementpointsareclosertothecalculatedcurve(Pigure8.169).Inordertokeeptheuncertaintiesassociatedwithsucheffectsassmallaspossible,onlytestsforwhichtheinitialpipetemperaturewasbelow700Cwerechosenforthecomparisonwiththetheoreticalcase.8.5.3.3.1.1.2GKMMode~luencherTestsAnothersorceusedtoverifythetheoryisofferedbytheGKNquenchertests(Ref1).Sincethepipetemperaturestherewereinthevicinityof300Corbelow,uncertaintiesinthebubble.volumeunderwateraredistinctlysmallerthanintheKarlsteintests.Inaddition,theGKMtestswerealsorunwithbackpressureinthesuppressionchamber,sothatinformationderivedfromthecomputercodesforblowdownofthequencherduringaloss-of-coolantaccidentcanalsobeverified.TheresultscanbefoundinFigures8-.170and8-171.Figure8-170showsthecalculatedandmeasureddependenceofthepulsationfrequencyonthebubblevolumeforvarioussubmergences(2m,4mand6m)withatmosphericpressureinthesuppressionchamber.ThetheoryandmeasuredfrequencyagreeevenbetterherethanintheKarlsteinquenchertests.Thisisprobablyduetothefactthatthebubblevolumesdeterminedfromthemeasurementvalueshaveamuchsmallerscatterduetothelowtemperaturesinthepipe.TheinfluenceofbackpressureonthepulsationfrequencyisshowninFigure8-171.Hereagain,thetheoryisverifiedbythetestdata.85.33113KKBHotTestsInordertodemonstratethecorrectnessofthetheoryforin-plantconditionsalso,calculationswereperformedfortheblowdowntestswithonevalveinthenon-nuclearhottestsintheBrunsbuttelBMRplant(Ref.3).Pigure8-172showstheresults.Theagreementbetweenthecalculatedandmeasured'requencyissimilartothatintheKarlsteintests.Thesameistrueforthescatterrangeofthemeasurementvalues.Sincethepipetemperatureherewasatabout90OC,alargerscatteractuallyREYlt3/798-89 PROPRIETARYwouldhavebeenexpected,butdidnotoccurbecausethepipewascarefullyflushedwithairpriortothebeginningofthesetests.8.5.3.3.1.1.4ConclusionfromtheFrequencyCalculationsThetestcalculationsdescribedaboveshowthatthetheory{VELPOTandKOVIB1Acomputerprograms)describesthemeasuredfrequenciesnotonlyinonespecialcase,butalsoforabroadrangeofgeometriesandbackpressure:(1)Thesizeofthewaterspacevariesfromapproximately7m>(GKN)toapproximately23m~(testtankatKarlstein)toapproximately400m~(suppressionchamberinBrunsbuttelnuclearpowerplant).(2)Thequenchersubmergencerangedfromapproximately2mto6m.(3)Thebubbleequilibrium.volumevariedbetweenapproximately015m~to37m~.(4)Thesuppressionchamberpressurevariedfrom1barto3bar.(5)Thewatertemperatureinthesuppressionpoolvariedbetweenapproximatley16~Cto800C.Thus,thetheorycanbeconsideredverifiedandcanbeusedtotransposethepulsationfrequenciesmeasuredintheKarlsteinteststandtotheSSESsuppressionpool.8.5.33.2Nuit~iliersforConversionoftheBubbleFrequenciesfromtheTestStandtoSSESUsingtheVELPOTandKOVIBlAcomputercodes,thefollowingthreeblowdowncasesareanalyzed:(1)'Simultaneousblowdownofall16quenchers(2)SimultaneousblowdownofthequenchersA,B,G,K,M,PwhichareincludedintheADS{3)Blowdownofonequencher(quencherB)TheresultsareillustratedinFigure8-173whichshowsthepulsationfrequencyasafunctionofbubblevolume(bubbleinhydrostaticequilibrium).Thebehaviorofthefrequencycurveforthe16-quenchercaseintheplantispracticallythesameasfortheteststand(Figure168),therebyconfirmingonceagainthesuitabilityoftheteststandgeometrythatwaschosen.Inthecaseofthe6quenchersintheADScase,thefrequenciesarehigherduetothelargersinglecellcorrespondingtothesmallerREVli3/79~8-90 PROPRIETARYhydrodynamicbubblemass.Theyareevenhigherinthecaseofonequencher.BasedontheresultsshowninFigures8-168and.8-173,asimpleformulacanbegivenforconvertingfromthemeasuredbubblefrequenciestothesefrequenciesfoundintheplantbyasking:Bywhatfactort"multiplier")mustabubblefrequencymeasuredintheteststandbemultipliedtogetacorrespondingfrequencyintheplant'?ThismultiplierisplottedinFigure8-174versusthe(measured)startingfrequency.Thus,wehave:v=f(v).vplantvtest'estinwhichthemuliiplierfforagiveninitialfreguencycanhereadofffromPigure8-173.ThegraphinFigure8-173isapplicableonlyforcaseswithapressureof1barinthesuppressionpoolairspace.However,theblowdownfortheADScaseduringaloss-of-coolantaccidentisassociatedwithasuppressonpooloverpressure.p~>lbarAnadditionalmultiplierfpKK(pKK)isnecessaryforsuchcases,sothatthefrequencyconversionmustbewritteninamoregeneralmanner:V=f(P).S(V).VplantPkk'testtestkkThemultiplierfpKK(pKK')canbetakenfromFigure8-175.Forasuppressionchamberpressureof7bar,ithastheavalueof1,asitmustbe.Themultipliersforthefrequencyalsofixthemultipliersfortheoscillationperiodwhentransposingthepressuretimehistoriesmeasuredintheteststandtotheplant:ttesttlPkkvtestkkI85.33.3TranspositionMethodforthePressureA~mlitudesAsalreadydescribedindetailinSection8.51,theteststandwassodesignedandthe.pressuretransducersweresoarrangedthatthemeasuredpressureamplitudescanbetransposedtotheplantwithoutchangeCorrespondingly,a1:1transpositionismade.Becauseofitsobviousconservativeness,sucha1:1REV1,3/798-91 PROPRIETARYaiplitudetranspositionofferstheadvantagethatmoreexactquantitativeproofsdonothavetobeprovided.Themostsignificantconservativefeaturesarethefolloving:(1)Inblowdowncase'swithseveralquenchers,itisassumedthatallbubblesareequallylargeandoscillateinphase.Deviationsfromthisassumption{suchasactuallyoccurintheplant)resultonlyinloverpressureamplitudes.(2)Blowdowncaseswithlessthan16quenchersareassignedthesamepessureamplitudeasthe16-quenchercase.Inreality,suchcaseshavealoweramplitudeduetothegeometry(largersinglecell).Theconservativenessdescribedin(1)hasnotyetbeenprovenexperimentallyinanyquenchertests,butitisalreadyobviousfromatheoreticalviewpoint,sinceatime-shiftedsuperpositionoftvotemporalmaximaalwaysyieldssmallervaluesthananadditionofthemaximumvalues.Concerningtheconservativenessof(2),thereareanumberqualitativeindicationsfromtheKarlsteinteststhemselves,fromcorrespondingmodelstudiesattheKarlsteinmodelteststand(Ref.1),andfromcalculationswiththeVELPOTandKOVIB1Aprograms.Theinformationobtainedfromallthreeoftheseinvestigationsshallbedescribedinthefollowingsections.Inaddition,wewillalsoexaminewhethertheconservativefeaturesareaffectedbyapossiblebackpressureinthesuppressionpoolairspace.85.33.31PPGL~uencherTestsatKarlsteinIndicationsoftheconservativenessdiscussedin(2)aboveareobtainedfromtheKarlsteintestsonthebasisofFigure8-176vhichillustratesthemeasuredrelationshipbetweenexcitation(relativeamplitude)andpressure-oscillationfrequencyfortheKarlsteintests.Thefrequencyanalysisforeachpressuretimehistoryhasatleasttwomaximaofthepowerdensity.Onepowerdensitymaximumliesatlowfrequenciesandtheotheratsomevhathigherfrequencies.Thereisafactorofapproximatelytwobetweenthetvofreqeuncies.Thefirstpeakofthepowerdensity(lowfrequency).isalwayslargerthanthesecondpeakofthepowerdensity(higherfrequency).Accordingly,thelovfrequencyisalvaysdesignatedasthedominantfrequencyForpressuretransducerP510,thepowerdensitiesofallanalyzedtestsareevaluatedinFigure8-176.DifferentanalysistimesvereselectedfortestshavingdifferentpressureoscillationfrequenciesThetimevassochosenthatREV.1,3/798-92 PROPRIETARYapproximatelythesameoscillationperiodscouldalvaysbeevaluated.Thefollowinganalysistimeswereselectedfortheevaluation:3HzTime:5HzTime:9HzTime:0-1.8seconds0-1.3seconds0-0.6secondsTheareabeneaththefrequencyspectrumvasdeterminedandthenthesquarerootofthatnumericalvaluewastaken.Thatresultsinvalueshavingthedimension>>har>>.Thosenumericalvalueswerenormalizedtothemaximumvalue.Theresultsarethen"relativepressures>>withrespecttothecalculatedmaximumpressurefromthefrequencyspectra.Sincenodominantfrequencieshigherthan6.5HzweremeasuredintheKarlsteintests,thesecondpeakswerealsousedtoevaluatethehigherfrequencies.Hence,thepowerdensitiesofboththedominantfrequencyandthenexthigherfrequencyareevaluatedinFigure8-176.Basedonanempiricalevaluation,it.followsfromFigure8-176thatthepressureoscillationsvithhigherfrequencieshavesmallerenergycontentthanthepressureoscillationswithloverfrequencies.Znaddition,asshowninFigure8-169,thehubblefrequencyincreaseswithdecreasinghubblevolume.Butdecreasingbubblevolumewithconstantsingle-cellsizemeans,accordingtothelawsofsimilarity,thesamethingasincreasingthecellsizewithconstantbubblevolumeTherefore,fromtheKarlsteintestdata,itcanbesaidthatthepressureamplitudesdecreasewithincreasingcellsize.8.5.33.32KM~UuencherTestsintheModelTestStandin~Ka1steinDuringthedevelopmentoftheKWUquencher,testsvereperformed.toexaminetheinfluenceofthesizeofthewaterspace(specifically:freewatersurface)inthemodelteststandinKarlstein(Ref.1).TheresultsareillustratedinFigure8-177,whichvastakenfromRefence1.Itshowsdirectlyhovthebottompressureamplitudesdecreasewithincreasingsizeofthewaterspace(singlecell).8.53.3.33AnalyticalCalculationsTheconservativenessdescribedin(2)aboveisalsoconfirmedfromresultsofcalculationswiththeVELPOTandKOVIB1AREV.1,3/798-93 PROPRIETARYprograms.Asforthefrequencyconversion,appropriatemultiplierscanbedeterminedalsofortheconversionofthepressureamplitudesfromtheteststandtotheplant.Theydependontheinfluenceofthewaterspaceonthestationaryvelocitypotential(spatialpressuredistributionnormalizedtounitsourcestrength)andonthehydrodynamicsourcestrengthassociatedwiththebubbledynamics.Thesourcestrengthitselfisdependentinturnonthepressureinthebubble,.whichisdeterminedbytheinterplay-of.bubblevolumeandairsupplyintothebubble.Sincetheairsupplyvariesaccordingtothedifferentoperatingconditionsduringtheblowdown,onlyaconservativeestimatecanbegivenwithintheframeworkofthepresentinvestigationsT'econversion.fromteststandtotheplantforonequenchermayserveasanexamplehere.Meobtainforthebottompressurebeneaththequencher:P(1quencher)<0.7Pplanttestasuppervalue.8.5.333.4InfluenceofBackgressureonthePressureAmplitudesAsforthebubbleoscillationfrequency,thequestionoftheeffectofbackpressureinthesuppressionpoolairspacemustbeinvestigated.Figure8-178showsthebottompressureamplitudesmeasuredintheGKNmodelquenchertestsforasuppressionpoolairspacepressuresof1and3barAscanbeseen,thepressureamplitudesdonotdependonthesuppressionpoolairspacepressure.8.533.4VerificationofDes~in~SecificationInthetranspositionofthepressureoscillationsmeasuredinKarlsteintotheSSES,theextremelyconservativeassumptionthatthesamepressuretimehistoriesareactingatallquencherssimultaneouslyisused.Differencesinthepressuretimehistoriesoriginatingfromthedifferentdischargelinesareneglected.Therefore,eachmeasuredpressureoscillationintheKarlsteinventclearingtestsisarepresentativecontainment,loadforallloadcases:symmetricalloadcase(simultaneousresponseofall16SRV'sunsymmetricalloadcase(responseofoneorthreeadjacentSRV'sautomaticdepressurizationinloss-of-coolantaccidentREV.1,3/798-94 PROPREETARYAtranspositionofthemeasurementresultstotheplantisperformedfortheseloadcases.TheKarlsteintesttankformsaconservativesinglecell.Therefore,conservativeenvelopingpressureamplitudesweremeasuredinthatteststand.Mhentransposingthepressureoscillationsfromthesinglecelltotheplant,thereisanincreaseofthepressureoscillationfrequenciesasdiscussedinSection8.5.3.3.2.Asstatedpreviously,theincreaseofthepressureoscillationfrequenciesisaccompaniedbyadecreaseoftheamplitudes.Thedecreaseoftheamplitudesisneglectedforthisevaluation,Theamplitudesofthemeasuredpressureoscillationsremainconstantforallfrequencies.Thatisanadditionalconservativefeature,asalreadydiscussedinSection8533385.33.4.1~foe~nencAnalysesofSelectedTestsThepressuretimehistoriesforselectedKarlsteintestsareillustratedinFigures8-41to8-65Thefreqeuncyanalyseswerecarriedout.withtheFourierAnalyzer5451madebyHewlettPackard.Thefrequencyanalysesweregeneratedaspowerspectraldensities.Thefrequenciesatwhichastructureisexcitedintooscillationcanbereadofffromthepowerspectraldensities.FreqeuncyanalyseswereperformedforpressuretransducersP5.2,P5.4,P5.5,andP5.10andforthefollowingtests:4el.l,4.16,12.1,llel,19R27~20Rl1,20eRle10,2lel~21.2,25.R2Pressureoscillationsatboththewallandthebottomareconsideredinthefreqeuncyanalyses.AlsoconsideredwasthefrequencyanalysisforpressuretransducerP5.5,whichshowsthefrlocaleffectThelimitationofthemeasuredfrequenciesofthepressureoscillationswasdeterminativeinselectingtheteststobeanalyzed.Thetestsselectedwerethosewhichexhibitedpressureamplitudes>0.3barbothatlowfrequencyandalsoathigherfrequencies.ThefrequencyspectraforseveralKarlsteintestsareillustratedinFigures8-179to8-182forpressuretransducersP5.10andp5.4.ThefrequencyspectrafortwotestswiththelongdischargelineandloweredwaterlevelareshowninFigure8-179.TheprincipalREV.1,3g798-95 PROPRIETARYfrequencyofthepressureoscillationsisat2-.3Hzforthesetests.TheyarethelowestpressureoscillationfrequenciesthatveremeasuredintheKarlsteintests.Figure8-180showsthedifferenceinthepressureoscillationfrequenciesfromclean-conditionteststoreal-conditionand/ormultiple-actuationtestsforthelonglineThepressureoscillationshaveaprincipalfreqeuncyof3.5Hzintest4.1.1(cleancondition)and5Hzintest4.1.6{realcondition)Fortheshortdischargeline,thefrequencyshiftsfromcleantorealconditionareillustratedinFigure8-181fortests21.1and21.2.Theresultfortheshortlineis:cleancondition:pressureoscillationfrequency5Hzrealcondition:pressureoscillationfrequency6.5HzThefollowingcanbesaidaboutthemeasuredgrin~cialfrequenciesfortheKarlsteintests:Thelowestpressureoscillationfrequencyvasmeasuredinthetestswiththelonglineandadischargelinewaterlevelloweredto2.5mabovethemiddleofthequencher.Itwas2.0-3Hz.2)Fortheclean-conditiontests,pressureoscillationfrequenciesof3.5-4Hzweremeasuredwiththelongdischargeline.3)Fortheclean-conditiontests,pressureoscillationfrequenciesof4.5-5Hzweremeasuredwiththeshortdischarge1ine.4)ThehighestfrequencyfortheKarlsteintestsvasmeasuredforthereal-conditionand/ormultiple-actuationtests.Themeasuredfrequenciesvere6-6.5Hz.Figure8-183shovsfrequencyanalysesfordifferentpressuretransducersforonetest.P5.2-sitsonthebottombeneaththemiddleofaquencherarm.P54-ismountedontheconcretewallattheintersectionofwallandbottom.P5.10-sitsontheconcretewalloppositethecenterpointof'heballofthequencher.Thefrequencyspectraofthepressuretransducersalldisplayapowermaximumatthesamefrequency(3Hz).Therefore,theREV13/798-96 PROPRIETARYlocationofthemeasurementandthestructureofthemountingpositioninthewaterregionoftheKarlsteinteststandhavenoinfluenceonthemeasuredfrequencyofthepressureoscillations.85.3.34e2ShiftingofthePSD'intheTra~nsositionfromtheTestStandtoSSESThecomparisonofthepressuretimehistoriesmeasuredintheKarlsteinquenchertestswiththepressuretimehistoriesspecifiedinSection4.13isaccomplishedbyusingthefrequencypowerspectra.ThefrequencyspectraoftheKKBtracesformingthebasisofthespecificationinSection4.1.3andareillustratedinFigures4-31to4-33Thespecifiedpressureoscillationshavetheirdominantfrequencyintherangeof6.5-8Hz.TocoverthepressureoscillationfrequenciesforSSES,thefollowingrulefortreatmentofthetraceswasgiven:Thethreetracesshouldbetime-expandedbyafactorintherangefrom0.9to1.8.Thepressureamplitudesshouldbemultipliedbyafactorof1.5.Tobeabletomakeacomparisonwiththemeasuredpressureoscillations,itisnecessarythatthefrequencyspectraofthethreetracesbeshiftedinfrequencyandstretchedinamplitude.InthisSection,weillustrateamethodbywhichthoseoperationsonthefrequencyspectracanbeperformed.8.5.3.3.4.2.1Pte5uencfshiftTheamplitudesarepreservedinthefrequencyshift.Toensurethat,theareaunderthepowerspectrummustbeheldconstant.Sincetheanalysistimerangeforthefrequencyanalysisisfinite,itmustbemadecertainthatthecomparisoninvolvesonlyspectrainwhichapproximatelythesamenumberofoscillationperiodswereanalyzedThetracesareexpandedorcompressedbythefactorf<,whilekeepingthezeropointfixedLetusdesignatetheexpandedorcompressedfrequencybyf'ndtheoriginalfrequencybyf.ApowerspectrumcanalwaysbesubdividedapproximatelyintotriangleswhosebaseisthefrequencyandwhosealtitudeisthepowerdensityIntheoriginalspectrum,theareabeneathatriangleis:f-fA21~h2REVli3/798-97 PROPRIETARYForthenewfrequency:fl=fxflfxfp2Therefore,wehaveforthenewarea:A''utsinceA'A,h=f~h''h~rThepowerdensityoftheshiftedspectrumisinverselyproportionaltothefrequencymultiplier.Inthisdefinition,thefrequencymultipliersaretobetakenfromSection4.1.3.Fromthefactor1.8wegetfV=1/1.8andfromthefactor0.9wegetfV=1/0.9.Ifthefrequencyisreducedtohalf,thepowerdensityisdoubled.85.3.34.2.2~AmlitudeStretchingThefollowingrelationprevailsbetweentheamplitudeofaload-vs.-timefunctionandthepowerdensity:a=k-bf'2k=correctionfactorForthestretchedamplitude,wehavea'fa.Therelationbetweenpowerdensityandamplitudeispreservedbythestretching,sothatthesamecorrectionfactorisalsovalidafterthestretching.Therefore:h''k-bf'andthus:hah)h'=f.hh2a8-98 PROPRIETARYThepowerdensityratiointheamplitudestretchingisproportionaltothesquareof'heamplitudemultiplier.8.53.3.43SymmetricalLoadCase~SimultaneousBlowdownofall16SRV's)AlltheKarlsteinclean-conditionandreal-conditiontestsareusedtoevaluatethisloadcase.Themultipleactuationtestsareconsideredasirrelevanttotheplantforthisloadcase.Theoneexceptionisthe10thblovdowntestofanentiremultipleactuationtestwiththeshortdischargeline.Thosetestsarestarted10,minutesaftercompletionofthe9thblowdowntest.Theyarethussubjecttothesameconditionsasthereal-conditiontests.Accordingly,the10thblowdovntestsofamultipleactuationtestwiththeshortdischargelinearetreatedasreal-conditiontests.ThetesttankinKarlsteinrepresentsthesmallestsinglecellwithrespecttothewaterspace.ThatmeansthatthemaximumpossiblepressureamplitudesforSSESweremeasured.AccordingtoSection8.5.3.2,themeasuredpressureamplitudesarecoveredbythespecificationForthisloadcase,themeasuredfrequenciesofthepressureoscillationscanalsobetransposeddirectlyfromtheKarlsteinteststandtoSSES(seeSection8.53.2).Thus,allthepressuretimehistorycanbetransposeddirectlyfromtheteststandtoSSES.Inordertoshowthatthemeasuredtimehistoriesarealsoenvelopedbythespecification,thefrequencyspectraofthemeasuredpressureoscillationsarecomparedwiththefrequencyspectraofthespecifiedtraces.Sincethemeasuredfrequenciesdifferfrom.thefrequenciesofthespecifiedtraces,thespectramustbetreatedbythemethodillustratedinSection8.5.3.3.42andbroughtintocoincidenceatthedominantfrequency.ThepressureoscillationsmeasuredatpressuretransducerP5.2areusedforthiscomparison,since,thepressuretransducerP5.2exhibitsthehighestpowerspectrumofallthepressuretransducersthatareuseablefortheoverallloadingofthecontainment(P5.5isnotconsidered-seeSection8.5.3.1).PressuretransducerP5.2ismountedonthebottomofthetesttank,directlybeneathaquencherarm.ThatpositionisalsopresentinSSES.Therefore,thispressuretransducermeasurespressureoscillationshavingthegreatestrelevancetoSSES.Purthermore,thespecifiedtracesarealsoresultsofameasurementmadewitha.bottompressuretransducerwhoselocationwassimilartothatofP5.2.REV1,3/798-99 PROPRIETARYThecomparisonofthefrequencypowerspectraisshowninFigures8-184to8-188WeseethatthefrequencyspectraoftheKKBtraces,whichwerefrequency-shiftedandamplitude-stretchedasdescribedinSection8.5.3.3.4.2envelopthefrequencyspectraofthemeasuredpressureoscillations.Therefore,itcanbestatedthat:a)theKarlsteinmeasurementresultsareconservativefortheloadcaseofsimultaneousclearingofall16quenchers{single-celleffect);'Ib)forthisloadcase,thepressureoscillationsareenvelopedbythespecificationwithrespecttotheiramplitude,theirfrequencypowerspectra,'andtheirspatialdistribution.8.5.3.3.44UnsymmetricalLoadCaseslowdownViaOneSRV)Forthisloadcase,alldeterminativeparameters,exceptforthe-,watersurfacearea,weresimulatedintheKarlsteinteststandaccordingtotheiractualvaluesforSSES.Fortheloadcaseofventclearingwithonequencher,alargerwatersurfaceareaisavailabletothequencherinSSESthaninthetestintheKarlsteinteststand.Accordingly,thepressureoscillationfrequenciesareraisedandthepressureamplitudesarelowered.Inthisverification,weconservativelymakenoallowancefortheamplitudedecreasewithincreasingwatersurfacearea.ThefrequenciescalculatedaccordingtoSection8.5.3.3.2fortheloadcaseofblowdownviaoneSRVarecompiledinthefollowingtable:Frequencyofthepressureoscillations(Hz)MeasuredFrequencymultiplierPlantSpecifiedfrequency.bandCLEANCONDITION3.5-41.54-1.485.4-5.9REALCONDITIONS5CLEANCONDITION1.421.427.17.13.75-8.9480RWREALCONDITIONS6.51.378.9REV.1,3/798-100 PROPRIETARYThefrequenciestransposedtotheplantareallenvelopedbythespecifiedfrequencyband.Fortheloadcaseofventclearingofonequencher,themultipleactuationtestsmustalsobeconsidered(theywereincludedunder"realconditions"intheTableabove)Fortheloadcaseofsimultaneousblowdownof16quenchers,itwasshownthatthemeasuredpowerspectraareenvelopedbythespecifiedpowerspectra.Thatstatementappliesforallfrequencyranges.Iftwopowerspectraarebroughtintocoincidenceatonefrequencyandifbothspectraaresubjectedtothesamefrequencyshift,thenthereisnochangeintherelationofthetwospectratoeachother.Therefore,thepowerspectraoftheclean-conditionandreal-conditiontestsarealsocoveredbythespecificationintheloadcaseofventclearingofonequencher,since,asstatedabove,thetransposedfrequenciesfromthetestareallenvelopedbythespecificationfrequencyrange.Forthemultipleactuationtests,test4.1.6isconsideredtobeenvelopingforthelongdischargeline,sinceitprovidedthehighestpressureamplitudes.Fortheshortdischargeline,test20.R1.10(whichformallycanbeclassifiedasamultipleactuationtest)isconsideredtobeenvelopingforthesamereason..Classifiedasareal-conditiontest,itwasshownintheprecedingSectionthatthespecifiedtracesenvelopthepressuretimehistoriesforthattest.InFigure8-189itisshownthatthepowerspectrumoftest4.16isalsoenvelopedbythespecifiedKKBtraces.EvenundertheveryconservativeassumptionthatthepressureamplitudesmeasuredinKarlsteincanbetransferredwithoutchangefortheloadcaseofvent,clearingofonequencher,thepressuretimehistoriesareenvelopedbythespecifiedtraces.85.3.34.5Un~smmetricalLoadCaseslowdownviaThreeA~d'acentSRV~sgThisloadcaseisboundedbytheloadcasesofsimultaneousventclearingof16quenchersandventclearingofonequencher.8.5.3.3.4-6AutomaticDeDzessuzizatio~asstem~A~DSLoadCaseInthissectionwediscusstheloadcasethatconsidersthefiringofthesixquenchersassociatedwiththeADSunderLOCAconditions.REVli3j'798-101 PROPRIETARYAsshowninFigure8-190,thefollowingconditionsprevailinthesuppressionchamberwhentheautomaticdepressurizationsystemisactuatedduringIBA:Absolutepressureinthewetwellairspace,approximatelyPressuredifference.betweendrywell'andsuppressionchamber2.55bar0.42barTheKarlsteintestswithloweredwaterlevelinthedischargelineareusedtoverifytheADScase.Thesetestsareusedastheycorrectlysimulatethedischargelineasitwouldbewithapositivepressuredifferentialofapproximately0.42barinthedrywell.Thispositivepressuredifferentialwouldresultintheloweringofthewaterlevelinthedischargelinetotheelevationofthebottomofthedowncomersaswassimulatedfortests10.3,ll1,12.1and13.1.Of'thosetests,thetest11.1(envelopinginamplitudeandpowerdensity)isusedasthebasisfortheverification.Theamplitude-reducinginfluenceofthelargerwatersurfaceareaassignedtotheindividualquencherintheADScaseisconservativelyneglected.Also,sinceearlierKMUtestsprovedthatthebackpressureinthesuppressionchamberhasnoinfluenceonthepressureamplitudes,themeasuredpressureamplitudesaretakenunalteredfromthecorrespondingKarlsteintests,inwhichthemeasurementsweremadeatatmosphericpressure.Thepredominantfrequencyintest11.1isat3Hz.AccordingtoSection8.5.3.3.2,Figures8-174and8-175,thefollowingfrequencymultipliersareobtainedfortheADScasefortranspositionofthepressureoscillationsfromtest11.1totheplant:InfluenceofthelargerwatersurfaceareaInfluenceofthe2.55barbackpressureTotalfrequencyfactorDomi.nantfrequency135141957HzNote:Themeasuredlowestdominantpressureoscillationfrequencywasmeasuredintests12.1and13.1,whichfallintothesamecategoryastestll1.Miththetotalmultiplier1.9,thefrequenciesareraisedto3.8Hzandthusliewithinthespecifiedfrequencyband(seeSection8.5.3.3.5).ThedominantfrequencyiswithinthespecifiedfrequencybandREVlg3/798-102 PROPRIETARYThecomparisonbetweenthepreparedtracefrompressuretransducerP5.2fortest11.1andthespecificationisshowninFigure8-191.Asfortheotherloadcases,thecomparisonismadeinthepowerspectra'ofthepressuretimehistories.Thespectrumoftest11.1wasshiftedfromthedominant.frequencyof3Hztothedominantfreqeuncyof5.7Hzwhilepreservingthearea(amplitude).TheKKBtraceoftest76wasshiftedfrom8Hzto5.7Hzwhilepreservingthearea,andthenstretchedbyafactorofl.5inamplitude.Figure8-191showsthatthetracefromthespecification,treatedinthismanner,envelopsthetraceofKarlsteintest11.1transformerdtotheADScasesincethetotalenergyrepresentedbytheareaunderthepowerspectrumcurvefromthespecificationisgreaterthanthatfromtheKarlsteintestll.185.33.47SummaryIthasbeendemonstratedthatthe,frequencypowerspectrumofthepressureoscillationsinthesuppressionchamberareenvelopedbythefrequencypowerspectrumspecifiedinSection4.1.3forallloadcases.Thus,thedesignspecificationprovidesenvelopingloadsalsoforthedynamicexcitationoftheSSFScontainmentbyventclearingofthereliefsystemwiththequencher.8.5.33.5EvaluationoftheNeasuredPressureOscillationsDuringCondensationAsdiscussedinSection8.4.2,threeregimescanbedistinguisedinthecondensationprocess:a)Thequencherisclearedcontinually.b)Thequencherisnotclearedcontinually.c)Onlytheslidingjointiscleared,andthesteamcondensesinthedischargeline.8.533.51TheguencherisClearedContinuallyThesteamiscondensedcontinuallyinthewaterpooloutsidethe-quencher.Calmcondensationprevailsforcoldwaterandalsoforhotwaterintheblowdowntank(seeFigures8-78and8-79)Themeasured.maximumpressureamplitudeiss0.13bar.Thiscondensationphasewasmeasuredforreactorpressuresuptoabout4bar.Thefrequenciesofthepressureoscillationsare70-120Hzforacoldpooland20-45Hzforahotpool.REV1,3/798-103 PROPRIETARY8.5.3.3.52TheQuencherisnotClearedContinu~allThiscondensatonphasebeginsvhenthecondensationrateoutsidethequencheris.greaterthanthesteammassflovthroughtheline.Thepressureinthequencherdropsbelovthehydrostaticpressureofthesurroundingvater.Thewaterpenetratesintothequencher.Thecondensationsurfaceareaistherebydecreasedandsoisthecondensationrate.Theresultisapressureriseinthedischargeline,sothatthewaterthathasflovedinisexpelledagain.Theinflowofvaterfromthesuppressionchamberintothequencherandthesubsequentbrakingandre-expulsionofthewaterisanonstationaryprocessvhichoccursperiodically.Forthatreason,thiscondensationphaseisalsocalledintermittentcondensation.Thephenomenonofintermittentcondensatonisdependentonthewatertemperature.Forcoldvaterthereisahigherrateofcondensationoutsidethequencher,resultinginalargergenerationofnegativepressureinsidethequencherandthereforeamorevigorousflowofwaterintothequencher.Foracoldwaterpool,theprofileofthedynamicpressuresissimilartotheprofilevhichisfamiliarfromthechuggingphaseofthecondensationattheventpipes;seeFigure8-76.Forheatedvaterinthesuppressionchamber,thecondensationrateoutsidethequencherissmaller,sothattheentireprocesstakesontheformofalow-frequencypressureoscillation(SeeFigure8-80)Thetestsin"Karlsteinyieldedasmaximummeasurementresultforthedynamicpressure:+0.28,-0.18bar,foracoldpool.Thetimebetweentwoeventsisabout1.0second.Foraheatedpool,themeasuredmaximumamplitudeis+0.12,-0.07,bar.8.533.5.3CondensationintheDischargeLineandThrutheSlidi~nJointIfthesteamflovdecreasesfurther,aconditionisfinallyreachedinwhichthequencherisnolongercleared,butratherremainscontinuallyfilledwithvater.Thenthereissteady-statecondensationofsteaminsidethedischargelineThiscondensationphaseproceedsverycalmlyandbeginsatreactorpressuresbelow2bar.Inthiscondensationphase,maximumdynamicpressuresof+0.08,-0.04barweremeasuredinthewaterpoolduringtheKarlsteintests.REVl,3/798-t04 PROPRIETARY853.3.54Tr~ansositionoftheMeasurementResultstoSSESInregardtosteamcondensation,theconditionsoftheKarlsteinteststandaredirecltytransposabletotheconditionsofSSES.Onthewhole,thepressureamplitudesduringcondensationaresmallcomparedtot'hoseduringventclearingandthereforearecoveredbythelatter.85.4PoolMixinqDurincnSRVActuationandThermalPerformanceofthe~uencher85.41IntroductionWhenanSRVresponds,steamiscondensedinthewaterofthesuppressionpoolviaaquencherAsthishappens,thewatermustabsorbtheheatofvaporizationofthesteam,andsoitisheated.Whenthereisalong-lastingdischargeofsteamviaaquencher,allthewaterinthesuppressionchambershouldparticpateintheheating,soastolimitthelocalheatinginthevicinityofthedischargingquencherInordertoobtaingoodmixingofthehotterandcolderwaterinthepool,allquenchersarepositionedatasmalldistancefromthebottom(3~6"=1.07m)(seeFigure8-192)).Thewaterheatednearaquencherisspecificallylighterthanthecolderwaterlyingaboveit.Therefore,thewarmerwaterwillriseandmixwiththecolderwater.Toobtainanadditionalmixingeffect,theholeoccupancyofthequenchersweremadeslightlyunsymmetrical(approximately8%).Mhereasthequencherarmshavethesameholeoccupanciesonthesides,onlyonearmofeachquencherhasholesontheendcap.Inthatway,aunilateralthrustcanbeexertedonthewaterinthesuppressionpool.Inthetopviewofthequencherarrangement(Figure8-193),weseethatthequenchersarearrangedintwograduatedcircles.Alongtheinnergraduatedcircle,thequencherarmsallpointinthecircumferentialdirection,andtheendcapwithholesallpointinthesamecircumferentialdirection.Ontheoutergraduatedcircle,thecolumnswouldpracticallypreventathrusteffectifthequencherswerearrangedinthesamemanner.Therefore,thequenchersweredirectedmoreradially,butturnedbyanangleofgf=300inthecircumferentialdirectionfromtheradii.Inthisway,50%ofthethrusttillactsinthecircumferentialdirection(equidirectionallywiththethrustofthequenchersontheinnergraduatedcircle).ItshouldbenotedthatthisnewarrangementsupersedestheoriginalarrangementshowninFigure1-4.Inthefollowing,weshallestimatetheaccelerationofthewaterpoolforthecaseinwhichonequencherontheoutergraduatedREVli3/798-105 PROPRIETARYcircleisoperatedforalongperiodoftimeatareactorpressureof70bar{valvefailureinopenposition).Thenweshallpresentsomemeasurementresultsfromatestwitha4-armquencherintheBrunsbuttelnuclearpowerplantandsomeinformationfromtheGEMNodelquenchertestsrelatedtosteamcondensationwithaquencher.85.42EquationofNotionoftheRotationPoolZtisassumedthatthewaterflowintherotatingpoolcanbeconsideredasastraight-linechannelflowduetothesmallcurvatureofthegraduatedcircleandthelowcircumferentialvelocity.Ifweplacetheoriginofthecoordinatesystematthecenterofthedischargingquencher,thentheequationofmotionoftherotatingpoolreads:5'.2mx+c2xFffmWcWeffThismassofwatertobeacceleratedinthesuppressionchambersumofallflowresistanceseffectivedrivingforcedifferentialequationhasthegeneralform:x+ax=bSubstitutingx=u,thedifferentialequationtakestheform:u+au~=bThisdifferentialequationisaspecialformoftheRiccatidifferentialequationThegeneralsolutionofthedifferentialequationsreadsR'ef.53:na.b+bTanha.b(t-K)u(t(n)=-.ga.b+a.q'Tanhga'b(t-c)Theinitialconditionfort=0reads:0n/ab+bTanh/ab-g)/ab=an.Tanh/ab(-g)REVl,3/798-106 PROPRIETARYThisconditionalequationissatisfiedonlyif6andn=0.Theinitialconditionthenleadstothesolution:b.Tanha.bu(t)=ga.bSinceu(t)=X(t),theequationforthevelocityoftherotatingpoolreads:x(t)=b)a.bTaab)a.btForthedistancecovered,wehave:x(t)=pJ'(v)dvThesolutionreads:X(t)=-ln[coshia.b.t(085.03'eterminationofthePlowResistancesThefollowingresistancesareconsidered:a)Wallresistanceofthechannelb)Resistanceforflowaroundthedischargelineswithquenchersandbottomsupportc)Resistanceforflowaroundtheventpipesd)ResistanceforflowaroundthecolumnsThechannelhasthefollowingdimensions:REV.li3/798-107 PHOPRIETARYThehydraulicdiameterofthechannelis:73(26~822-8~84Rect226.822-8.84(27.3)+,22.8cnFortheReynoldsnumber,wehave:Re!W.RRAccordingtoReference36,thekinematicviscosityforwaterat40oCisv=0.65lx10-~m~/s.Xfweassumeavelocityof10-2m/ssoas'tocoverthestart-upphasealso,weget:R+10x2.8.43x1024.651x10TheSSESsuppessionpoolislinedwithasteellinerwhichcannotbeconsideredhydraulicallysmooth.Forsuchlargesteelstructuresitmustbeassumedthattheindividualplatesarenotjoinedtogetherwiththeiredgesparallel,sothattheflowresistanceisincreasedbyprojectingedges.Wethereforeconservativelyassumeanabsoluteroughnessofk=2mm.Thenwehave:Kdh2.8x10-47.1x10Thiscorrespondstoafrictioncoefficientof>=0.022.Theresistancecoefficientisthen:~.1mWh26.844+8.84m'2~7fRR56mr-.022w'28)Cylindricalbodiesareimmersedintothewaterofthesuppressionchamber.Theyarethedischarge.lineswithguenchers,theventpipes,andthesteelcolumns.REV1,3/798-108 PROPRIETARYOutsidediametermSubmergencemQuantityDischargelinesVentpipesSteelcolumns03240.61106733357-3168712Fortheindividualstructuralcomponents,wethenhavethefollowingReynoldsnumber:v=0.01m/s(seeabove)WRe=-dmFortheroughness,weassumek=0.2mm.Then,accordingtoReference39:ReynoldsnumberSubmergencedDischargelinewithquenchervi.thbottomsupportVentpipe5x10~9.4x10~6.17x10-i6.28x1022.555073073Columnl63x10'9x10-+6.9073Theresistanceforceist,hen:p92Thesurfaceareaonwhichthewallresistanceactsis:24Furthermore:c6.16x.44+.73x50+.73x177.8+.73x93WAA16x0.324x9.650m.2c=238mA87x.61x3.35177.8.m2AS-12xl.06x7.3~93m2Sincethewaterregionofthesuppressionchamberalsocontainsafewstructuralcomponentswhichverenotconsideredhere,anadditionalallowanceshallbemade.Mechoose:2c300mWREV.13/798-109 PROPRIETARY8.5.4.4DeterminationoftheForceNovi~nthePoolForcesonthewatermassinthesuppressionpoolareproducedbythrustfromtheboreholesononeoftheendcapswhicharepresentoneachofthequenchers.Thesmallestthrustforceisproducedbythequenchersalongtheoutergraduatedcircle,sincetheydonothavetheirthrustboreholesarrangedinthecircumferentialdirection.Thequenchersalongtheoutergraduatedcircleareturnedbyanangle4=30orelativetotheradialdirection.g=50'DV=differencebetweenpressureinthequencherandambientpressureThethrustforceresultsfromtheimpulseoftheoutflowingsteam.F~APxA-+PxMDxA~UeffectiveoutletareaofquencherPD=densityoftheoutflowingsteamW=velocityoftheoutflowingsteamDAsaneffectiveoutletareaofaquencherendcap,thereisavailable:A~ga<xADUgeomeDC0.8(Section8.5.2.3)+0geom(~l2(4)6.9x10m-32ADOgeom5'~2x10m-32Aconstantreactorpressureof70barischosenfortheestimateof.theeffectivenessoftherotatingpool.AccordingtoReference37,themassflowthroughthereliefvalveatareactorpressureof70baris:REVlr3/798-110 PROPRIETARYm=illkg/sTheresultingstagnationpressureinthequencheris:p=llbarandthesteampressureinthequencherholesis:pD=64barTherefore,PD"=34kg/m>WD462m/sTheforceactinginthecircumferentialdirectionisthen:FeffFsinFeffTherefore:(AP+P+Wj))A"xsinQwithQ=30'DUFeff2~O'N+lo5KN3~5KN85.45WorkingEquationsfortheRotation'PoolofSSESTheequationofmotionfortherotatingpoolreads:5'.2mx+c2effwwThisdifferentialequationwassolvedingeneralforminSection85.4-2.Todetermineconsiderthehave:themassofwaterwhichistobemoved,wemustinternalstructureswhichreducethewatermass.MeI,4(26.822)-(8.84))x.73--x(.324)x7.3x164--x(61)x3.35x87--x(1.06)x12x7.3]43.5x10KgForthetotalresistancecoefficientwehaveaccordingtoSection8543:C<=300mandfortheeffectivelyactingforcewehaveaccordingtoSection8544:F=35KNeffREVl,3/798-111 PROPRIETARYTherefore,theeguationofmotionreads:3.5x10xX+1.5x10xX6-52or3.5x10Therefore,for:a=4.3x104b=9.9x102X+aX=bVab6.55x10tTheequationforthevelocityoftherotatingpoolreads:-1-3X(t)=,1.52x10Yanh6.55x10tTheequationforthedisplacementreads:X(t)=23.21nicosh6.55x10tiTheresultsareillustratedinFigures8-194and8-195.854.6EstimateoftheHeati~noftheSuppressionChamberMaterThelocalheatingofthesuppressionchambervaterresultsfromthebalanceoftheheatbroughtinbythecondensingsteamandtheheatdissipatedbytheflowingwater.Astimepasses,hovever,thepoolissetintomotionbytheimpulseof.theinflowingsteamandreachesavelocitysuchthatmostoftheheatbroughtinisdistributedoveralargervolumeofwaterthantheassumedlocalvolume.,Thedifferencebetweenthelocalandmeanwatertemperaturedecreases.85.47ExperimentalProofs854.71NodelTankTestsThrustmeasurementsonasteamjetveremadeintheKarlsteinmodeltankintheSpringof1973(Ref40).Thetestset-upisillustratedinFigure8-196Thesteampipeisconnectedby.aspringtothesidewallofthemodeltank.Theexcursionofthespringwiththesteampipeismeasuredbyadisplacementtransducer.Themeasurementsystemvascalibratedbydeterminingtheexcursionofthesteampipeforadefinedforce.Thesteamoutletopeninghadadiameterof10mm.Themassflowdensityvas600to630kg/m<s.Themeasuredreactionforceswere20-28N.REV1,3/798-112 PROPRIETARYAshortcalculationyields:OutletareaRestpressurebeforetheoutletopeningPressureaftertheoutletopeningSteamdensity(at2.6bar)=7.854x10-~m~4.5bar2.6bar.=1.44kg/m~Theresultingoutletvelocityis:W=gK-2.6x10~~1.135W-452.7m/sandthethrustforceis:F=(PW+hP)AffAff=08xAegeomF~(1.44x(452.72)+1.6x10)x0.8x7.854x10F~284NThemeasuredvaluesareloverthanthecalculatedvalues.Themeasurementshaveprovedclearlythattheimpulseoftheemergingsteamjetbecomesactiveasathrustandthat,vithrespecttothevelocitybuildupoftherotatingpool(andthusforthemaximumlocalheating),itisconservativelyboundedbythecalculatedvalues.85.47.2KKBTestDuringtheNuclearCommissioningThepressurereliefsystemwastestedduringthecommissioningphaseoftheBrunsbuttelnuclearpoverplant.Inonesuchtest,areliefvalvewasheldopenforatimeofabout270seconds.Thesuppressionchambercoolingsystemvasswitchedonduringthetest.Waterwasdrawnoffinthelowerpartofthepool,cooled,andsprayedfrompipesprovidedwithholesandlocatedunderthetopofthesuppressionchamber.12measuringpointsaremountedinthewaterregionofthesuppressionchamber.Theyarearrangedatthreedifferentelevations(14m,16.5m,18.2m)andatfourdifferentcircumferentialpositions(5o,75o,195o,245o).Thewaterlevelisataheightof18.89m.Figure8-197showsathreedimensionalspatialrepresentationofthemeasuredtemperaturefieldinthevaterjustbeforeteststart(curve1)andat228secondsafterteststart(curve2).InFigure8-197,theverticalpositionofthetransducerisrepresentedontheordinateandthecircumferentialpositionontheabscissaThetemperatureaxispointstotherear.Theheatingofthepoolisindicatedasthedifferenceofcurves2and1atthreeelevationpositions.Themeanwatertemperaturewasapproximately32.3oCbeforethetestandapproximately42.8oCREV.1,3/798-113 PROPRIETARYat228slater..Themaximummeasuredtemperaturewas500C,sothatthemaximumdeviationfromthemeanwas7.20C.Thedischargingquencherwaslocatedat285'tanelevationof14915mandacceleratedthewatertowardtheleftintheFigure.Correspondingly,the.watertemperatureishigheraboveandtotheleftofthequencher.Fromthatwecanseetheeffectivenessofthequencher'sarrangementnearthebottomandoftheunsymmetricalholearrangementwithre.,pecttouniformutilizationoftheheatsinkofthewaterpool.85473GKNHalfScaleQuencherCondensationTestAseriesofintermediatescale(1:2)condensationtestswereperformedintheGKMteststandtodemonstratethehightemperatureperformanceoftheguenchers(Ref.27).Condensationtestswererunonsevendifferentversionsofthequencherdevice.Thelastthreeversionshad10-mmdiameterholeson'hequencherarmsThespacingoftheholecenterlineswas1.5diameterscircumferentiallyand5.0diametersaxially.ThisholepatternisalsoadoptedintheactualSSESquencherdesign.Thesetestswererunatawatertemperaturerangingfrom13oCto100oC(56oF-2120F)andasteammassflux(withrespecttotheholearea)rangeof8to495kg/m~(1.6to101ibm/ft~s).Matertemperaturesashighas1070C(225~F)weremeasuredatcertainlocationsinthesetests.85.48SummaryTheKarlstein'uenchertestsandpreviousGKMhalfscalequenchertestsshowclearlythatsmoothsteamcondensationcanbeachievedatelevatedtemperatureswhichapproachthe.localsaturationlimit.XnadditionthecalculationsandKKB,inplanttestsprovideinformationwhichsuggestthatpoolmixingisenhancedbysteamdischargethroughtheholesintheendcapsofthequencher."85.5VerificationofSubmergedStructuresLoad~SecificationDueToSQVActuationSection4.1.3.7givesthedesignspecificationfortheloadsonsubmergedstructuresduetoSRVactuation.Thebasisforthespecificationisthethreepressuretimehistoriesusedforthecontainmentanalysisbutinsteadofaconstantamplitudemultiplierof1.5variousmultipliers,relatedtothecrossectionalareaoftheobject,areused.(seeTable4-15).TheloadingonthecolumnsincludingthelocalizedefectatP5.5hasbeendiscussedinSection8.5.3.2.1.2REV1,3/798-114 PROPREETARYInadditiontheeffectsofairbubbleoscillationloadsonthequenchershavebeendiscussedinSection8.5.23.6.ThefollowingsectionwilldiscusstheloadingsontheventpipesasmeasuredintheKarlsteintesttankandprovideadescriptionoftheinfluencefortheexpelledwaterduri'ngventclearing.8.5.51LoadsontheVentP~ie85.511MeasurementoftheLoadsXnordertodeterminetheloadingoftheventpipenearaquencher,aventpipehavingthesameoutsidediameterandwallthicknessasthatinSSESwasinstalledintheKarlsteinteststandandsupportedbytypicalbracing.(seeFigure8-10).Underneaththebracing,bendingstrainsweremeasuredintwomutuallyperpendicularplanesbymeansofstraingauges(SGS1andSGS2)(seeFigures8-11and8-12).Thestraingaugesweremountedabout100mmbelowthebracing.Theoutsidediameteroftheventpipeis:D=0609mandtheinsidediameteris:D;=0589mThus,thecross-sectionalareais:A~0.0188m2andthemomentofresistanceis:4Dl320Wehave:-332.77x10mxW~'M~GxExWTherefore:M~2.77x10'0'-3,11'ndhence;M~057cREV.1,3/798-115 PROPRIETARYIfweinsertcinmicrometerspermeterintothisequation,weobtainthebendingmomentinkN-mThebendingmomentscalculatedinthismannerarestaticequivalentloads.5-5-5.1.2MeasnnedBendi~nsaeentsFigures8-198to8-200showthedependenceofthemeasuredresultantbendingmomentsonthereactorpressure,ventclearingpressure,and.pressureoscillationamplitudethatweremeasuredneartheventpipeontheconcretewall.Onlythetestswithcleanconditionswereusedfortheplotofthemeasuredbendingmomentsversusreactorpressure,whereasalltestsinthereactorpressurerangeof60-81barwereusedfortheplotsofthebendingmomentversusventclearingpressureandpressureoscillationamplitude.Themeasurementsofthebendingstrainsattheventpipewereperformedonlyforthetestswiththelongdischargeline.Themeasuredmaximumbendingmomentwas14.6kN-mata74barreactorpressureanda13.8barventclearingpressure.85513Extr~aolationoftheMeasurementResultsand~ComarisonwiththeSpecifiedValueIfthemeasurementvaluesareextrapolatedtotheextremeconditionsintheplantonthebasisofFigures8-198and8-199,wegetthefollowingextrapolatedmaximumvalues:16.5kN-mwithrespecttoan88barreactorpressure,19.0kN-mwithrespecttotheventclearingpressureof16.5barforthelongdischargepipe,asextrapolatedinSection8.4fortheextremeboundaryconditionsintheplant.Inthespecification,amaximumpressuredifferenceof0.75x0.8=0.6barwasspecifiedfortheventpipewiththedistributionillustratedinFigure4-24.ThepressuredistributionfortheventpipeinstalledintheKarlsteinteststandisshowninFigure8-201Thefollowingrelationappliesforthepressureattheendofaventpipe:~dPdP7.3-1.837.3-3.65hPssOe4barREVli3/798-116 PROPRIETARYAttheclampingpointoftheventstrut,wehave:AP07.3-1.837.3-6.3AP=01barThepressuredistributionfromtheendoftheventpipetotheclampingpointofthevent-pipestrutistrapezoidal.L=0.1x-'-'-x2.652.65(0.4-0.1)2S'(2)3Theleverarmoftheactingforcewithrespecttotheclampingpointis:0.1+.0.42S1'59Forthebendingmomentattheclampingpointweget:M5=(~2~'2.65x0.6x1.59)10SP~SP63kNmRelativetothestraingauges,wehave:MBSP57kNmTheextrapolatedmaximummomentwas19kN-m.Itisthusdemonstratedthatthespecificationenvelopsthemeasurementvaluesandtheirextrapolation.Theproofthatthespecificationenvelopsthemeasurementvaluesandtheirextrapolationisbasedonapurelystaticanalysis.Suchananalysisispermissiblebecausetheexcitingpressureoscillationshaveafrequencyof4-6Hz.However,thestraingaugesindicateanaturaloscillationfrequencyof17-20HzfortheventpipewhichisveryclosetothenaturalfrequencyoftheventinSSES(19Hz)(seeFigure8-202).Hence,itcanbeassumedthatthedynamicloadfactorisclosetoone.8552InfluenceofExpelledWaterDuringVentClearinc[AreviewofthehighspeedfilmsandpressuretracesatP5.5fromtheKarlsteintestsshowsnegligableinfluenceoftheexpelledwateratthisgage.Inadditionthetotalpenetrationoftheexpelledvaterappearstobeapproximately3feetfora70barinitialsystempressure.Therefore,noadditionalloading,otherthanthatalreadyincludedinthepressuretracesvillheconsidered.REV1,3/798-117 PROPRIETARY(AtimecorrelationofahighspeedfilmtopressuretraceatP5.5willbesuppliedlater.}85.53SummaryTheloadsmeasuredonthedummyventpipearestaticequivalentloads,butloadswhichareasumofindividualcomponents.Inthespecification,thetransverseloadsoninternalstructuresoriginatingfromtheblowdownofthereliefsystemare.formulatedasdifferentialpressuresacrosstheinternalstructures.ThedifferentialpressureshavethesamepressuretimehistoryasthedynamicpressuresinthewaterregionofthesuppressionchamberThisformulationofthetransverseloadsontheventpipe(moregenerallyontheinternalstructuresinthewaterregionofthesuppressionpool)yieldstheenvelopingstaticeg'uivalentload.ThiswasalsoverifiedbytheKKBtestswiththeactualreliefsystem(Ref.38).Themaximumdifferentialpressurescalculatedfromthemeasurementresultsarep=0.16baratthequencherarm,andp=0.11barattheprotectivepipeonthedischargeline.TheyarebothconservativelyboundedbytheKKBspecifiedvalueofp=0.2bar.TheKKBtestresultsshowsthatthereisaclearseparationbetweenthespecifiedloadsandthemaximummeasuredloadsforboththelateralandverticalloadsoninternalsinthepoolofthesuppressionpool.BasedontheverificationofthetransverseloadsbytheKKBtestsandbasedonthecomparisonbetweenspecificationandmeasurementfortheKarlsteintests(seeSection8.5.5.1),itcanbestatedthatthevaluesformulatedinthespecificationforthetransverseloadsoninternalstructuresinthewaterregionyieldenvelopingstaticequivalentloads.REV1g3/798-118 KEY:1.Reliefvalve2.Compressedair3.HPsteamline4.Heating5.SRV6.CondensateeP,$$...'CIA,(o!>.d'1$$'4(C$!PACmli0a!0CUmmZCh0Zzzgzo$mLmZZr~Dmmnfh0Z | ||
1$3t'4P<Pg~$4.$1716.))),O~15OO$$X(4DO~$)3KEY)l.Reliefvalve2.Compressedair3~HPsteamline4.Heating5.SRV6.CondensateI~~Ii.Ispo'gA9ggl~P8~%1,,C0Cpmxmgm)NyIIIgr~'Ngoc.)pE$$$$eIPE$'4)7Q3gy$~pill)yOH e | 1$3t'4P<Pg~$4.$1716.))),O~15OO$$X(4DO~$)3KEY)l.Reliefvalve2.Compressedair3~HPsteamline4.Heating5.SRV6.CondensateI~~Ii.Ispo'gA9ggl~P8~%1,,C0Cpmxmgm)NyIIIgr~'Ngoc.)pE$$$$eIPE$'4)7Q3gy$~pill)yOH e | ||
Line 56: | Line 56: | ||
~DlIT88LJOFCONTENTSChapterlGENERALINFORMATION1.1PurposeofReport1.2HistoryofProblem1.3QuencherDischargeDevice14tlKIISupportinqProgram1.5PlantDescription1.6Fiqures1.7TablesChapter2SUN'RY2.1LoadDefinitionSummary2.2DesiqnAssessmentSummaryChapter3SRVDISCHARGEANDLOCATRANSIENTDESCRIPTIONChapter4LOADDEFINITION4.1LoadsfromSafetyReliefValveDischarge4.2LoadsfromLoss-of-CoolantAccident4.3AnnulusPressurization4.4Fiqures4.5Tables3.1DescriptionofSafetyReliefValve{SRV)Discharge3.2DescriptionofLoss-of-CoolantAccident(LOCA)Chapter5LOADCONBINATIONSFORSTRUCTUR~ESPIPING~ANDEOUIPi'IENT5.1ConcreteContainmentandReactorBuildingLoadCombinations5.2STructuralSteelLoadCombinations5.3LinerPlateLoadCombinations54DovncomerLoadCombinations5.5Piping,Quencher,andQuencherSupportLoadCombinations5.6NSSSLoadCombinations5.7EquipmentLoadCombinations5.8Figures5.9TablesChapter6DESIGNCAPABILITYASSESSMENT6.1ConcreteContainmentandReactorBuildingCapabilityAssessmentCriteria6.2StructuralSteelCapabilityAssessmentCriteria6.3LinerPlateCapabilityAssessmentCriteria | ~DlIT88LJOFCONTENTSChapterlGENERALINFORMATION1.1PurposeofReport1.2HistoryofProblem1.3QuencherDischargeDevice14tlKIISupportinqProgram1.5PlantDescription1.6Fiqures1.7TablesChapter2SUN'RY2.1LoadDefinitionSummary2.2DesiqnAssessmentSummaryChapter3SRVDISCHARGEANDLOCATRANSIENTDESCRIPTIONChapter4LOADDEFINITION4.1LoadsfromSafetyReliefValveDischarge4.2LoadsfromLoss-of-CoolantAccident4.3AnnulusPressurization4.4Fiqures4.5Tables3.1DescriptionofSafetyReliefValve{SRV)Discharge3.2DescriptionofLoss-of-CoolantAccident(LOCA)Chapter5LOADCONBINATIONSFORSTRUCTUR~ESPIPING~ANDEOUIPi'IENT5.1ConcreteContainmentandReactorBuildingLoadCombinations5.2STructuralSteelLoadCombinations5.3LinerPlateLoadCombinations54DovncomerLoadCombinations5.5Piping,Quencher,andQuencherSupportLoadCombinations5.6NSSSLoadCombinations5.7EquipmentLoadCombinations5.8Figures5.9TablesChapter6DESIGNCAPABILITYASSESSMENT6.1ConcreteContainmentandReactorBuildingCapabilityAssessmentCriteria6.2StructuralSteelCapabilityAssessmentCriteria6.3LinerPlateCapabilityAssessmentCriteria | ||
TABLEOFCONTENTS~Continue~d6.4DowncomerCapabilityAssessmentCriteria6.5Pipinq,Quencher,andQuencherSupportCapabilityAssessmentCriteria6.6NSSSCapabilityAssessmentCriteria6.7EquipmentCapabilityAssessmentCriteriaChapter7DESIGNASSESSMENT7.1AssessmentMethodology7.2DesignCapabilityMargins7.3FiguresChapter88~SEEOENCHERVERIVICETIONTEST8.1UnitCellApproach8.2SimulationofSSESParameters8.3InstrumentationArrangement8.4TestMatrix8.5AnalysisofData8.6FiguresChapter9RESPONSESTONRC~UESTIONS9.1IdentificationofQuestionsUniquetoSSES9.2QuestionsUniquetoSSESandResponsesThereto9.3FiguresChapter10REFERENCESAppendixACONTAINMENTDESIGNASSESSMENTA.IContainmentStructuralDesignAssessmentA.2SubmergedStructuresDesiqnAssessmentAppendixBCONTAINMENTRESPONSESPECTRADUETOSRVANDLOCA)LOADSAppendixCREACTORBUILDINGRESPONSESPECTRADUETOSRVANDLOCALOADS It.'II,it)''+lrIii''''tt' AppendixDTABLEOPCONTENTS~Continue~dPROGRAMVERIFICATIOND.1PoolswellModelVerificationD.2FiquresD.3TablesAppendixEREACTORBUILDINGSTRUCTURALDESIGNASSESSMENTAppendixFPIPINGDESIGNASSESSNENTAppendixGNSSSDESIGNASSESSNENTAppendixHEQUIPMENTDESIGNASSESSHENT CHAPTER1GENERALINFORMATIONTABLEOFCONTENTS1.1PURPOSE'OFREPORT12HISTORYOFPROBLEM13QUENCHERDISCHARGEDEVICE1.4MARKIX,SUPPORTINGPROGRAM15PLANTDESCRIPTION1.5.1PrimaryContainment1.5.1.1Penetrations1.5.1.2InternalStructures16FIGURES1.7TABLES1-1 NumberTitle1-11-21-0CrossSectionofContainmentSuppressionChamber,PartialPlanSuppressionChamber,SectionViewQuencherDistribution CHAPTER1TABLESNumber1-2TitleSSESX,icensingBasis"SSESContainment'Dimensions1-3SSESContainmentDesignParameters" 3 10GQNERAIINFOBHATEON11PURPOSEANDOBGANXZATXONOFBEPQRTThepurposeofthisreportistopresentevidencethattheSusguehannaSteamElectricStation(SSES)designmarginsareadequateshouldtheplantbesubjectedtotherecentlydefinedthermohydrodynamicloadswhich-'r'esult'romsafetyreliefvalve(SBV)operationsand/ordischargesduringaloss-of-coolantaccident(LOCA)..inaGE.boilinqwaterreactor(BWB)1-4 Thecriteriausedfors'electionoftheSRVdischargedeviceforSSESwereminimizationofpressureoscillationloadsinthesuppressionpoolandstablecond'ensationofsteamfortherangeofsuppressionpooltemperaturesoverwhichsafetyreliefvalvescanbeexpectedtooperateTheoptionsconsideredforsatisfyinqthesecriteriaweretherams-headtee,thequencherdischargedevice,andvariationsonthesedesigns.Evaluationofthetwoprincipaldevicesindicatedthatthequencherofferedsignificantadvantagesovertherams-head,includingimprovedthermalperformanceathigherpooloperatinqtemperatures,aswellasreducedloads.Athermohydraulicquencherdesignforthesafe'tyreliefsystemoftheSSZSisbeingengineeredbyKraftwerkOnion{KMU)tosatisfytheabovecriteria.TheSSESquencherdesignisdifferentfromthatpresentedintheMarkIIDFFRinthatithasbeenoptimizedbasedonparametricteststudieswhichwereconductedbyKMUinordertominimizeSRVdischargeloadsKraftwerkUnionhassuppliedtoPPGLapackageofsignificantdesiqnandtestreportspertainingtothequencherdevelopmenttodemonstratedesignadequacyandgualityoftheirdevice{refertoTable1-1).Mithreqardtothe<<secondpop<<phenomenon,KMUtestshaveindicatedthat,duetothequencherflowresistance,thewaterlevelintheSRVdischargepipefollowinginitialdischargedoesnotriseabovethewaterlevelofthesuppressionpool.RefertoSubsection0.1.3.6forafurtherdiscussion.ToverifyKMU'sdesignapproachafull-scaleSSESuniqueunitce11test,asdescribedinChapter8,isbeingperformedbyKMUforPPSL.Section01presentstheanalysismethodsoftheSRVdischarqeloading1-7 1.4HARKIISUPPORTINGPROGRAMPPGLisamemberoftheMarkIIovnersgroupthatvasformedinJune,1975todefineandinvestigatethedynamicloadsduetoSRVdischargeandLOCA.TheNarkIIovnersgroupcontainmentprogramconcentratedinitiallyonthetasksrequiredforthelicensingoftheleadplants(Zimmer,LaSalle,andShoreham).Thisphaseof,work,calledtheshorttermprogram,isessentiallycomplete(asofJanuary,1978)andalonqertermprogramisundervay.ThefinalgoaloftheMarkIIprogramistoevolveacompleteDPFRwhichvillsupporttheplant-uniqueDARssubmittedbyeachplantforitslicensetooperate.AfterqainingsomeunderstandingofthecontainmentloadsthroughtheinitialMarkIIwork,PPGLdecidedtofindaqualifiedconsultanttosupplementin-housetechnicalresourcesandassistinthedeterminationofarealisticcourseofactionforSusquehanna.InNovember,1976,StanfordResearchInstitute,novcalledStanfordResearchInstituteInternational(SRI),wasselected,andaninformationexchangebetweenSHIandPPGLensuedtodeterminewhatcausedthegreatestloadsonthecontainmentstructure.AfterconductingacompletereviewofknowndatafromtheMarkIIprogramandotherknowledqeablepersonsandorganizations,PPGXandSRIdecidedthattheloadsfrommainsteamsafetyreliefvalve(SHV)dischargewerethekey1oadstobecontrolled.AstudyofpossiblemethodsofcontrollingtheloadandareviewofvhatactivitieswereoccurringinEuropeledPPGLandSHItotheconclusionthatanSRVdischargemitigatingdevice{quencher)shouldbeemployedtoreducethisloadingontheSusquehannacontainment.AlthoughtheHarkIIovnersgrouphadquencher-relatedtasksintheirprogram,thesetaskswerenotsufficientlytimelytosatisfySSES-constructionscheduleneeds.PromreviewinqtheworkdoneinEuropebysuchfirmsasASEATOM,MARVIKEN,andKraftwerkUnion,PPGLdiscoveredthatallknownquencherdesignswerebasedondatafromKraftwerkUnion(KMU).Thus,inMarch,1977,SRI,Bechtel(theSSESArchitect/Engineer)andPPGLvisitedKWUfordiscussionandtourofquencher-relatedfacilities.InlateJuly,1977,PPGLemployedtheservicesofKMUtodesignaSSES-uniquequencherdevice(seeSection1.3).ThedefinitionofLOCAloads(Section4.2)isinaccordancewiththeNarkIIprogramDuetotheschedulerestrictionsforSusquehanna.PPGLwilldefinethethermo-hydrodynamicloadsresultinqfromSRVdischargeusinqanapproachdevelopedbyKMU.Thisapproach(presentedinSection4.1)differsfromthatoftheHarkIIproqram.SeeTablel-lforasummaryofthedocumentationsupportinqSSESlicensinq.1-8 15PLANTDESCRIPTIONTheSSES,Units1and2,isbeingbuiltinSalemTownship,Luzecne,County,about5milesnoctheastoftheBorouqhofBerwick.Twogeneratingunitsofapproximately1,100megawattseachare.scheduledforoperation.Unit'1forNovembec1,1980,andUnit2forMay1,1982.GeneralElectricissupplying'henuclearsteamsupplysystems;Bechtelpowercorporationisthearchitect-engineerandconstructor.Thereactorbuildingcontainsthemajornuclearsystemsandequipment.ThenuclearreactorsforUnits1and2areboilingwater,directcycletypeswitharatedheatoutputof11.2x10~Btu/hr.Eachreactorsupplies134x10~lb/hrofsteamtothetandemcompound,doubleflowtucbines.15.'IP~cimanContainmentaThecontainmentisareinforcedconcretestructureconsistingofacylindricalsuppressionchamberbeneathatruncatedconicaldrywell.Piqure1-1showsthegeometry'fthe.containmentandinternalstructures.Theconicalportionoftheprimarycontainment(drywell)enclosesthereactorvessel,reactorcoolantrecirculationloops,andassociatedcomponents'ofthereactorcoolantsystem.Thedcywellisseparatedfromthewetwell,ie,thepressuresuppressionchamberandpool,bythedrywellfloor,.alsonamedthediaphragmslabMajorsystemsandcomponentsinthecontainmentincludetheventpipesystem(downcomers)connectingthedrywellandwetwell,isolationvalves,vacuumreliefsystem,containmentcoolingsystems,andotherserviceequipment.Theconeandcylinderformastructurallyinteqratedreinforcedconcretevessel,linedwithsteelplateandclosedatthetopofthedrywellwithasteeldomedhead.Thecarbonsteellinerplateisanchoredtotheconcretebystru'cturalsteelmembersembeddedintheconcreteandweldedto-theplate.Theentire-containmentisstructurallyseparatedfromthesurroundingreactorbuildingexceptatthebasefoundationslabfareinforcedconcretemat,toplinedwithacarbonsteellinerplate)whereacoldjointbetweenthetwoadjoiningfoundationslabsisprovided.ThecontainmentstructuredimensionsandparametersarelistedinTables1-2and1-3.AdetailedplantdescriptioncanbefoundintheSSESPSAH,Section3.81.5.l.1PenetrationsServicesandcommunicationbetweentheinsideandoutsideofthecontainmentaremadepossiblebypenetrationsthroughthecontainmentwallThebasictypesofpenetrationsarethedrywellhead,accesshatches(equipmenthatches,personnellock,suppressionchamberaccesshatches,CRDremovalhatch),electricalpenetrations,andpipepenetcations.Thepiping1-9 penetrationsconsistbasicallyofapipewithplateflangeweldedtoit.Theplateflange.isembeddedintheconcretewallandprovidesananchorageforthepenetrationtoresistnormaloperatingandaccidentpipereaction.loads.Theinternalstructuresconsistofreinforcedconcreteandstructuralsteelandhavethemajorfunctionsofsupportingandshieldingthereactorvessel,supportingthepipingandeguipment,andformingthepressuresuppressionboundary.Thesestructuresinclude'hedrywellfloor(diaphragmslab),thereactorpedestal(aconcentriccylindricalreinforcedconcreteshellrestingonthecontainmentbasefoundationslabandsupportinqthereactorvessel),thereactorshieldwall,thesuppressionchambercolumns(hollowsteelpipecolumnssupportingthediaphragmslab),thedrywellplatforms,theseismictrusses,thequenchersupports,andthereactorsteamsupplysystemsupports.SeeFigures1-1through1-4aridTables1-2and1-31-10 dyCt"dp~rIsl.400106PC'CrCONTAINMENTUp(gpSRVDIAPHRAGMSLABPENETRATIONFROMDIAPHRAGMSLABPENETRATIONO~,Ce,i03.0>~~dp~136o~fp<'~u~e~Criy160P0r~drg1sr~irIdQ8O~'EDESTALTOQUENCHERSSRVDIAPHRAGMSLABPENETRATIONNOTE:BRACINGISNOTSHOWNSUSQUEHANNASTEAMELECTRICSTATION.UNITS1AND2DESIGNASSESSMENTREPORTSUPPRESSIONCHAMBERPARTIALPLAN/FIGURE1-2 3454001So315000'PP,<<PO.<<<<~~~~<<'<<454300088FT'"K"-ur,25822mm~~2850P100FT"30480mmH~h,BGo0~a'~.A<<<<MP~,<<I'O7500<<2700d.a2550b<<0RSOFT"18288mm<<~pP3eFT<mP'<<4<<a'1050240V5r~P<<0~~C22542FT'"F12802mmD~<<<<<<~~L'<<<<~~.''e0<<~4'0PP~k'3519541CS4NOTE:INDICATESADS-ASSOCIATEDQUENCHERSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTQUENCHERDISTRIBUTIONFIGURE1R TABLE1-1SSESLICENSINGBASISI..MarkIIContainment-SupportingProgramA.LOCh-RelatedTasksh.2.h.3.PoolSwellModelReportImpactTestsh.4.ImpactModelEPRI1/13ScaleTestsTaskeeetet~ettivetA.l."4T"PhasesI,II,IIXActivitePhaseITestReportPhaseIApplicationMemorandumPhaseII6IIITestReportPhaseII&IIIApplicationMemorandumModelReportPSTF1/3ScaleTestsMarkI1/2ScaleTestsPSTF1/3ScaleTestsMarkI1/2ScaleTestsEPRIReportTarget~teeletteCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedDocumentationNEDO/NEDE13442-P-01-5/76ApplicationMemo-6/76NEDO/NEDE13468-P-12/76ApplicationMemo-1/77HEDO/NEDE21544-P-12/76HEDE13426-P-8/75NEDC20989-2P-9/75NEDE13426-P-8/75HEDC20989-2P-9/75EPRINP-441-4/77UsedforSSESLicensinYesYesYesYesYesYesYesYesNoYesA.5.LoadsonSubmergedStructuresLOCA/RHAirBubbleModel12/77LOCA/RHHaterJetModel12/77ApplicationsMethods12/77TestReports1Q/78NEDE21471NEDE21472NEDE21730ReportUndecidedUndecidedUndecidedUndecidedh.6.h.7~A.8.ChuggingAnalysisandTestingChuggingSingleVentEPRITestEvaluationSingleCellReport4TFSIReportMultiventModelCREAREReportEPRI-4TComparisonCompleted1/7812/774Q/77CompletedHEDE23703-P-ll/77NEDE23710-PNEDE21669-PReportNEDO21667<<8/77YesHoHoYesh.9.MultiventSubscaleTestingandAnalysisFacilityDescriptionand4Q/77TestPlanTestReport1979ReportFinalReportUndecidedUndecided TaskNumberA.10.~eetteitSingleVentLateralLoadsActiviteAnalysisReportTarget~Contestee4Q/77DocumentationReportUsedforSSESLicensinUndecidedB.SRVRelatedTasksB.l.QuencherModelB.2.RamsheadModelB.3.MonticelloIn-PlantSRVTestsDFFRModelConfirmatoryTestsAnalysisPreliminaryTestReportHydrodynamicReportCompleted3Q/78CompletedCompletedCompletedNEDO/NEDE21061-P-9/76ReportNEDO/NEDE21061-P-9/76NEDC21465-P-12/76NEDC215&1-P-8/77..NoNoNoNoNoB.4.B.5.B.6.Be7.B.S.B.9vB.10.B.11.B.12.B.13.ConsecutiveActuationTransientAnalysisSRVQuencherIn-PlantCaorsoTestsThermalMixingModelSRVWaterClearingQuencherAirBubbleFrequencyMonticelloFluidStructureInteraction(FSI)DFFRRamsheadModelComparisontoMonticelloDataRamsheadSRVMethodologySurnLaryStructuralResponsetoSRVDischargeQuencherEmpiricalModelUpdateAnalyticalModelsTestPlanAdvanceTestReportPinalReportAnalyticalModelAnalysisAnalyticalModelAnalysisData/ModelComparisonAnalyticalMethodsAnalyticalReportAnalyticalModelandCorrelation4Q/77Completed1Q/784QI784Q/783Q/784Q/771Q/78CompletedCompleted4Q/77lQ/79ReportNEDM20988-12/76ReportReportNEDC23689ReportReportReportNSC-GEN0394-10/77NEDO24070-11/77ReportReportNoNoNoNoNoNoNoNoNoNoNoNo TaskNnett~ActivitC.MiscellaneousTasksActiviteTarget.~ConlotionDocumentationUsedforSSESLicensinC.l.C.2.C.3.DFFR,Rev.3MassandEnergyReleaseReportNRCRound1QuestionsRevisionAnalyticalReportDFFRAmendment110/783/77CompletedGB-77-65NEDO/NEDE21061Amendment1-12/76YesYesNEDO/NEDE21061Revision3NotyetavailableC.4.C.5.C.6.DecouplingChuggingandSRVLoadsSRSSJustificationNRCRound2QuestionsDFFRAmendment1,Supplement1SRSSReportDFFRAmendment212/77OnholdCompletedCompletedNEDO/NEDE21061Amendment1,Supplement2NEDO/NEDE24010-7/77NEDO/NEDE21061Amendment2-6/77YesYesYesDFFRAmendment2,Supplement1DFFRAmendment2,Supplement2Supplement3CompletedCompleted4q/77NEDO/NEDE21061Amendment2,Supplement1-8/77NEDO/NEDE21061Amendment2,.Supplement2Supplement3YesYesYesC.7.Justificationof"4T"BoundingLoadsChuggingLoadsJustificationCompletedNEDO/NEDE23617"P-8/77NEDO/NEDE24013-P-8/77NEDO/NEDE24104-P-8/77NEDO/NEDE24015-P-8/77NEDO/NEDE24016-P'-8/77NEDO/NEDE24017-P-8/77NEDO/NEDE23627-P-8/77UndecidedUndecidedUndecidedUndecidedUndecidedUndecidedUndecidedC.g.FSIEffectsinMarkIIContainmentsC.9.MonitorWorldTestsII.1NUTestsandReports(suppliedtoPP&L)EvaluationofFSIEffectslg/78MonitoringWorldPressureSuppressionTestsReportReports(Quarterly)UndecidedNoDocumentNumberTitleFormationandoscillationofasphericalgasbubbleStatusCompletedDocumentationAEG-Report2241UsedforSSESLicensinYesAnalyticalmodelforclarificationofpressurepulsationinthewetwellafterventcleaningCopletedAEG-Report2208 | TABLEOFCONTENTS~Continue~d6.4DowncomerCapabilityAssessmentCriteria6.5Pipinq,Quencher,andQuencherSupportCapabilityAssessmentCriteria6.6NSSSCapabilityAssessmentCriteria6.7EquipmentCapabilityAssessmentCriteriaChapter7DESIGNASSESSMENT7.1AssessmentMethodology7.2DesignCapabilityMargins7.3FiguresChapter88~SEEOENCHERVERIVICETIONTEST8.1UnitCellApproach8.2SimulationofSSESParameters8.3InstrumentationArrangement8.4TestMatrix8.5AnalysisofData8.6FiguresChapter9RESPONSESTONRC~UESTIONS9.1IdentificationofQuestionsUniquetoSSES9.2QuestionsUniquetoSSESandResponsesThereto9.3FiguresChapter10REFERENCESAppendixACONTAINMENTDESIGNASSESSMENTA.IContainmentStructuralDesignAssessmentA.2SubmergedStructuresDesiqnAssessmentAppendixBCONTAINMENTRESPONSESPECTRADUETOSRVANDLOCA)LOADSAppendixCREACTORBUILDINGRESPONSESPECTRADUETOSRVANDLOCALOADS It.'II,it)''+lrIii''''tt' AppendixDTABLEOPCONTENTS~Continue~dPROGRAMVERIFICATIOND.1PoolswellModelVerificationD.2FiquresD.3TablesAppendixEREACTORBUILDINGSTRUCTURALDESIGNASSESSMENTAppendixFPIPINGDESIGNASSESSNENTAppendixGNSSSDESIGNASSESSNENTAppendixHEQUIPMENTDESIGNASSESSHENT CHAPTER1GENERALINFORMATIONTABLEOFCONTENTS1.1PURPOSE'OFREPORT12HISTORYOFPROBLEM13QUENCHERDISCHARGEDEVICE1.4MARKIX,SUPPORTINGPROGRAM15PLANTDESCRIPTION1.5.1PrimaryContainment1.5.1.1Penetrations1.5.1.2InternalStructures16FIGURES1.7TABLES1-1 NumberTitle1-11-21-0CrossSectionofContainmentSuppressionChamber,PartialPlanSuppressionChamber,SectionViewQuencherDistribution CHAPTER1TABLESNumber1-2TitleSSESX,icensingBasis"SSESContainment'Dimensions1-3SSESContainmentDesignParameters"-1-3 10GQNERAIINFOBHATEON11PURPOSEANDOBGANXZATXONOFBEPQRTThepurposeofthisreportistopresentevidencethattheSusguehannaSteamElectricStation(SSES)designmarginsareadequateshouldtheplantbesubjectedtotherecentlydefinedthermohydrodynamicloadswhich-'r'esult'romsafetyreliefvalve(SBV)operationsand/ordischargesduringaloss-of-coolantaccident(LOCA)..inaGE.boilinqwaterreactor(BWB)1-4 Thecriteriausedfors'electionoftheSRVdischargedeviceforSSESwereminimizationofpressureoscillationloadsinthesuppressionpoolandstablecond'ensationofsteamfortherangeofsuppressionpooltemperaturesoverwhichsafetyreliefvalvescanbeexpectedtooperateTheoptionsconsideredforsatisfyinqthesecriteriaweretherams-headtee,thequencherdischargedevice,andvariationsonthesedesigns.Evaluationofthetwoprincipaldevicesindicatedthatthequencherofferedsignificantadvantagesovertherams-head,includingimprovedthermalperformanceathigherpooloperatinqtemperatures,aswellasreducedloads.Athermohydraulicquencherdesignforthesafe'tyreliefsystemoftheSSZSisbeingengineeredbyKraftwerkOnion{KMU)tosatisfytheabovecriteria.TheSSESquencherdesignisdifferentfromthatpresentedintheMarkIIDFFRinthatithasbeenoptimizedbasedonparametricteststudieswhichwereconductedbyKMUinordertominimizeSRVdischargeloadsKraftwerkUnionhassuppliedtoPPGLapackageofsignificantdesiqnandtestreportspertainingtothequencherdevelopmenttodemonstratedesignadequacyandgualityoftheirdevice{refertoTable1-1).Mithreqardtothe<<secondpop<<phenomenon,KMUtestshaveindicatedthat,duetothequencherflowresistance,thewaterlevelintheSRVdischargepipefollowinginitialdischargedoesnotriseabovethewaterlevelofthesuppressionpool.RefertoSubsection0.1.3.6forafurtherdiscussion.ToverifyKMU'sdesignapproachafull-scaleSSESuniqueunitce11test,asdescribedinChapter8,isbeingperformedbyKMUforPPSL.Section01presentstheanalysismethodsoftheSRVdischarqeloading1-7 1.4HARKIISUPPORTINGPROGRAMPPGLisamemberoftheMarkIIovnersgroupthatvasformedinJune,1975todefineandinvestigatethedynamicloadsduetoSRVdischargeandLOCA.TheNarkIIovnersgroupcontainmentprogramconcentratedinitiallyonthetasksrequiredforthelicensingoftheleadplants(Zimmer,LaSalle,andShoreham).Thisphaseof,work,calledtheshorttermprogram,isessentiallycomplete(asofJanuary,1978)andalonqertermprogramisundervay.ThefinalgoaloftheMarkIIprogramistoevolveacompleteDPFRwhichvillsupporttheplant-uniqueDARssubmittedbyeachplantforitslicensetooperate.AfterqainingsomeunderstandingofthecontainmentloadsthroughtheinitialMarkIIwork,PPGLdecidedtofindaqualifiedconsultanttosupplementin-housetechnicalresourcesandassistinthedeterminationofarealisticcourseofactionforSusquehanna.InNovember,1976,StanfordResearchInstitute,novcalledStanfordResearchInstituteInternational(SRI),wasselected,andaninformationexchangebetweenSHIandPPGLensuedtodeterminewhatcausedthegreatestloadsonthecontainmentstructure.AfterconductingacompletereviewofknowndatafromtheMarkIIprogramandotherknowledqeablepersonsandorganizations,PPGXandSRIdecidedthattheloadsfrommainsteamsafetyreliefvalve(SHV)dischargewerethekey1oadstobecontrolled.AstudyofpossiblemethodsofcontrollingtheloadandareviewofvhatactivitieswereoccurringinEuropeledPPGLandSHItotheconclusionthatanSRVdischargemitigatingdevice{quencher)shouldbeemployedtoreducethisloadingontheSusquehannacontainment.AlthoughtheHarkIIovnersgrouphadquencher-relatedtasksintheirprogram,thesetaskswerenotsufficientlytimelytosatisfySSES-constructionscheduleneeds.PromreviewinqtheworkdoneinEuropebysuchfirmsasASEATOM,MARVIKEN,andKraftwerkUnion,PPGLdiscoveredthatallknownquencherdesignswerebasedondatafromKraftwerkUnion(KMU).Thus,inMarch,1977,SRI,Bechtel(theSSESArchitect/Engineer)andPPGLvisitedKWUfordiscussionandtourofquencher-relatedfacilities.InlateJuly,1977,PPGLemployedtheservicesofKMUtodesignaSSES-uniquequencherdevice(seeSection1.3).ThedefinitionofLOCAloads(Section4.2)isinaccordancewiththeNarkIIprogramDuetotheschedulerestrictionsforSusquehanna.PPGLwilldefinethethermo-hydrodynamicloadsresultinqfromSRVdischargeusinqanapproachdevelopedbyKMU.Thisapproach(presentedinSection4.1)differsfromthatoftheHarkIIproqram.SeeTablel-lforasummaryofthedocumentationsupportinqSSESlicensinq.1-8 15PLANTDESCRIPTIONTheSSES,Units1and2,isbeingbuiltinSalemTownship,Luzecne,County,about5milesnoctheastoftheBorouqhofBerwick.Twogeneratingunitsofapproximately1,100megawattseachare.scheduledforoperation.Unit'1forNovembec1,1980,andUnit2forMay1,1982.GeneralElectricissupplying'henuclearsteamsupplysystems;Bechtelpowercorporationisthearchitect-engineerandconstructor.Thereactorbuildingcontainsthemajornuclearsystemsandequipment.ThenuclearreactorsforUnits1and2areboilingwater,directcycletypeswitharatedheatoutputof11.2x10~Btu/hr.Eachreactorsupplies134x10~lb/hrofsteamtothetandemcompound,doubleflowtucbines.15.'IP~cimanContainmentaThecontainmentisareinforcedconcretestructureconsistingofacylindricalsuppressionchamberbeneathatruncatedconicaldrywell.Piqure1-1showsthegeometry'fthe.containmentandinternalstructures.Theconicalportionoftheprimarycontainment(drywell)enclosesthereactorvessel,reactorcoolantrecirculationloops,andassociatedcomponents'ofthereactorcoolantsystem.Thedcywellisseparatedfromthewetwell,ie,thepressuresuppressionchamberandpool,bythedrywellfloor,.alsonamedthediaphragmslabMajorsystemsandcomponentsinthecontainmentincludetheventpipesystem(downcomers)connectingthedrywellandwetwell,isolationvalves,vacuumreliefsystem,containmentcoolingsystems,andotherserviceequipment.Theconeandcylinderformastructurallyinteqratedreinforcedconcretevessel,linedwithsteelplateandclosedatthetopofthedrywellwithasteeldomedhead.Thecarbonsteellinerplateisanchoredtotheconcretebystru'cturalsteelmembersembeddedintheconcreteandweldedto-theplate.Theentire-containmentisstructurallyseparatedfromthesurroundingreactorbuildingexceptatthebasefoundationslabfareinforcedconcretemat,toplinedwithacarbonsteellinerplate)whereacoldjointbetweenthetwoadjoiningfoundationslabsisprovided.ThecontainmentstructuredimensionsandparametersarelistedinTables1-2and1-3.AdetailedplantdescriptioncanbefoundintheSSESPSAH,Section3.81.5.l.1PenetrationsServicesandcommunicationbetweentheinsideandoutsideofthecontainmentaremadepossiblebypenetrationsthroughthecontainmentwallThebasictypesofpenetrationsarethedrywellhead,accesshatches(equipmenthatches,personnellock,suppressionchamberaccesshatches,CRDremovalhatch),electricalpenetrations,andpipepenetcations.Thepiping1-9 penetrationsconsistbasicallyofapipewithplateflangeweldedtoit.Theplateflange.isembeddedintheconcretewallandprovidesananchorageforthepenetrationtoresistnormaloperatingandaccidentpipereaction.loads.Theinternalstructuresconsistofreinforcedconcreteandstructuralsteelandhavethemajorfunctionsofsupportingandshieldingthereactorvessel,supportingthepipingandeguipment,andformingthepressuresuppressionboundary.Thesestructuresinclude'hedrywellfloor(diaphragmslab),thereactorpedestal(aconcentriccylindricalreinforcedconcreteshellrestingonthecontainmentbasefoundationslabandsupportinqthereactorvessel),thereactorshieldwall,thesuppressionchambercolumns(hollowsteelpipecolumnssupportingthediaphragmslab),thedrywellplatforms,theseismictrusses,thequenchersupports,andthereactorsteamsupplysystemsupports.SeeFigures1-1through1-4aridTables1-2and1-31-10 dyCt"dp~rIsl.400106PC'CrCONTAINMENTUp(gpSRVDIAPHRAGMSLABPENETRATIONFROMDIAPHRAGMSLABPENETRATIONO~,Ce,i03.0>~~dp~136o~fp<'~u~e~Criy160P0r~drg1sr~irIdQ8O~'EDESTALTOQUENCHERSSRVDIAPHRAGMSLABPENETRATIONNOTE:BRACINGISNOTSHOWNSUSQUEHANNASTEAMELECTRICSTATION.UNITS1AND2DESIGNASSESSMENTREPORTSUPPRESSIONCHAMBERPARTIALPLAN/FIGURE1-2 3454001So315000'PP,<<PO.<<<<~~~~<<'<<454300088FT'"K"-ur,25822mm~~2850P100FT"30480mmH~h,BGo0~a'~.A<<<<MP~,<<I'O7500<<2700d.a2550b<<0RSOFT"18288mm<<~pP3eFT<mP'<<4<<a'1050240V5r~P<<0~~C22542FT'"F12802mmD~<<<<<<~~L'<<<<~~.''e0<<~4'0PP~k'3519541CS4NOTE:INDICATESADS-ASSOCIATEDQUENCHERSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTQUENCHERDISTRIBUTIONFIGURE1R TABLE1-1SSESLICENSINGBASISI..MarkIIContainment-SupportingProgramA.LOCh-RelatedTasksh.2.h.3.PoolSwellModelReportImpactTestsh.4.ImpactModelEPRI1/13ScaleTestsTaskeeetet~ettivetA.l."4T"PhasesI,II,IIXActivitePhaseITestReportPhaseIApplicationMemorandumPhaseII6IIITestReportPhaseII&IIIApplicationMemorandumModelReportPSTF1/3ScaleTestsMarkI1/2ScaleTestsPSTF1/3ScaleTestsMarkI1/2ScaleTestsEPRIReportTarget~teeletteCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedDocumentationNEDO/NEDE13442-P-01-5/76ApplicationMemo-6/76NEDO/NEDE13468-P-12/76ApplicationMemo-1/77HEDO/NEDE21544-P-12/76HEDE13426-P-8/75NEDC20989-2P-9/75NEDE13426-P-8/75HEDC20989-2P-9/75EPRINP-441-4/77UsedforSSESLicensinYesYesYesYesYesYesYesYesNoYesA.5.LoadsonSubmergedStructuresLOCA/RHAirBubbleModel12/77LOCA/RHHaterJetModel12/77ApplicationsMethods12/77TestReports1Q/78NEDE21471NEDE21472NEDE21730ReportUndecidedUndecidedUndecidedUndecidedh.6.h.7~A.8.ChuggingAnalysisandTestingChuggingSingleVentEPRITestEvaluationSingleCellReport4TFSIReportMultiventModelCREAREReportEPRI-4TComparisonCompleted1/7812/774Q/77CompletedHEDE23703-P-ll/77NEDE23710-PNEDE21669-PReportNEDO21667<<8/77YesHoHoYesh.9.MultiventSubscaleTestingandAnalysisFacilityDescriptionand4Q/77TestPlanTestReport1979ReportFinalReportUndecidedUndecided TaskNumberA.10.~eetteitSingleVentLateralLoadsActiviteAnalysisReportTarget~Contestee4Q/77DocumentationReportUsedforSSESLicensinUndecidedB.SRVRelatedTasksB.l.QuencherModelB.2.RamsheadModelB.3.MonticelloIn-PlantSRVTestsDFFRModelConfirmatoryTestsAnalysisPreliminaryTestReportHydrodynamicReportCompleted3Q/78CompletedCompletedCompletedNEDO/NEDE21061-P-9/76ReportNEDO/NEDE21061-P-9/76NEDC21465-P-12/76NEDC215&1-P-8/77..NoNoNoNoNoB.4.B.5.B.6.Be7.B.S.B.9vB.10.B.11.B.12.B.13.ConsecutiveActuationTransientAnalysisSRVQuencherIn-PlantCaorsoTestsThermalMixingModelSRVWaterClearingQuencherAirBubbleFrequencyMonticelloFluidStructureInteraction(FSI)DFFRRamsheadModelComparisontoMonticelloDataRamsheadSRVMethodologySurnLaryStructuralResponsetoSRVDischargeQuencherEmpiricalModelUpdateAnalyticalModelsTestPlanAdvanceTestReportPinalReportAnalyticalModelAnalysisAnalyticalModelAnalysisData/ModelComparisonAnalyticalMethodsAnalyticalReportAnalyticalModelandCorrelation4Q/77Completed1Q/784QI784Q/783Q/784Q/771Q/78CompletedCompleted4Q/77lQ/79ReportNEDM20988-12/76ReportReportNEDC23689ReportReportReportNSC-GEN0394-10/77NEDO24070-11/77ReportReportNoNoNoNoNoNoNoNoNoNoNoNo TaskNnett~ActivitC.MiscellaneousTasksActiviteTarget.~ConlotionDocumentationUsedforSSESLicensinC.l.C.2.C.3.DFFR,Rev.3MassandEnergyReleaseReportNRCRound1QuestionsRevisionAnalyticalReportDFFRAmendment110/783/77CompletedGB-77-65NEDO/NEDE21061Amendment1-12/76YesYesNEDO/NEDE21061Revision3NotyetavailableC.4.C.5.C.6.DecouplingChuggingandSRVLoadsSRSSJustificationNRCRound2QuestionsDFFRAmendment1,Supplement1SRSSReportDFFRAmendment212/77OnholdCompletedCompletedNEDO/NEDE21061Amendment1,Supplement2NEDO/NEDE24010-7/77NEDO/NEDE21061Amendment2-6/77YesYesYesDFFRAmendment2,Supplement1DFFRAmendment2,Supplement2Supplement3CompletedCompleted4q/77NEDO/NEDE21061Amendment2,Supplement1-8/77NEDO/NEDE21061Amendment2,.Supplement2Supplement3YesYesYesC.7.Justificationof"4T"BoundingLoadsChuggingLoadsJustificationCompletedNEDO/NEDE23617"P-8/77NEDO/NEDE24013-P-8/77NEDO/NEDE24104-P-8/77NEDO/NEDE24015-P-8/77NEDO/NEDE24016-P'-8/77NEDO/NEDE24017-P-8/77NEDO/NEDE23627-P-8/77UndecidedUndecidedUndecidedUndecidedUndecidedUndecidedUndecidedC.g.FSIEffectsinMarkIIContainmentsC.9.MonitorWorldTestsII.1NUTestsandReports(suppliedtoPP&L)EvaluationofFSIEffectslg/78MonitoringWorldPressureSuppressionTestsReportReports(Quarterly)UndecidedNoDocumentNumberTitleFormationandoscillationofasphericalgasbubbleStatusCompletedDocumentationAEG-Report2241UsedforSSESLicensinYesAnalyticalmodelforclarificationofpressurepulsationinthewetwellafterventcleaningCopletedAEG-Report2208 | ||
DocumentNumber3~4,5.6.7.8.TitleTestsonmixedcondensationwithmodelquenchersCondensationandventclearingtestsatGKMwithquenchersConceptanddesignofthepressurereliefsystemwithquenchersKKBventclearingwithquencherTestsoncondensationwithquencherswhensubmergenceofquencherarmsisshallowKKB-ConceptandtaskofpressurereliefsystemStatusCompletedCompletedCompletedCompletedCompletedCompletedDocumentationKWV-Report2593KWV-Report2594KWV-Report2703KWV-Report2796KWV-Report2840KWV-.Report2871UsedforSSESLicensinYesYesYesYesYesYes9.ExperimentalapproachtoventclearinginamodeltankCompletedKMV-Report3129Yes10.KKB-Specificationofblowdowntestsduringnon-nuclearhot.functionaltest-Rev.IdatedOctober4,1974CompletedKWU/V822ReportYesAnticipateddataforblowdowntestswith-pressurereliefsystemduringthenon-nuclearhotfunctionaltestatnuclearpowerstationBrunsbuttel(KKB)CompletedKWU-Report3141Yes12.13.Resultsofthenon-nuclearhotfunctionaltestswiththepressurereliefsysteminthenuclearpowerstationBrunsbuttelAnalysisoftheloadsmeasuredonthepressurereliefsystemduringthenon-nuclearhot,functionaltestatKKBCompletedCompletedKWU-Report3267KWU-Report3346YesYes14.KKB-Listingoftestparametersandimportanttestdataofthenon-nuclearhotfunctionaltestswiththepressurereliefsystemCompletedKWU-WorkingReportR521/40/77Yes15.KKB-Specificationofadditionaltestsfortestingofthepressurereliefvalvesduringthenuclearstart-up,Rev.1CompletedKWU/V822TAYes16.KKB-Resultsfromnuclearstart-uptestingofpressurereliefsystemCompletedKNJ-WorkingReportR142-136/76Yes17'uclearPowerStationPhillipsburg-Unit1HotFunctionalTest:SpecificationofpressurereliefvalvetestsaswellasemergencycoolingandwetwellcoolingsystemsCompletedKWU/V822/RF13Yes DocumentNumberTitleStatusDocumentationUsedforSSESLicensin18.Resultsofthenon-nuclearhotfunctionaltestswiththepressurereliefsysteminthenuclearpowerstationPhillipsburgCompletedKWU-WorkingReportR142-38/77Yes19.KKPI-Listingoftestparametersandimportanttestdataofthenon-nuclearhotfunctionaltestswiththepressurereliefsystemCompletedKWU-WorkingReportR521/41/77Yes20.AiroscillationsduringventclearingwithsingleanddoublepipesCompletedAEG-Report2327Yes616715/cak TABLE1-3SSESCONTAINMENTDESIGNPAKQKTERSA.DrellandSuressionChamber1.InternalDesignPressure2.ExternalDesignPressure3.DrywellFloorDesignDifferentialPressureUpwardDownward4.DesignTemperature5.DrywellFreeVolume(Minimum)(includingvents)(Normal)(Maximum)6.SuppressionChamberVolumeFree(Minimum)(Normal)(Maximum)7.SuppressionChamberWaterVolume(Minimum)(Normal)(Maximum)8.PoolCross-SectionArea~Drell53psig5psid340'F239,337ft33239,593ft3239,850ft.28psid28psidSuressionChamber45psig5psid220'F148,590ft3153,860ft3159,130ft122,410ft3131,550ftGross(OutsidePedestal)TotalGross(IncludingPedestalWaterArea)Free(OutsidePedestal)TotalFree5379ft5679ft5065ft5365ft CHAPTER2SUMMARYTABLEOPCONTENTSLOADDEFINITIONSUMMARY21.1SRULoadDefinitionSummary2.1.2LOCAI.oadDefinitionSummaryDESIGNASSESSMENTSUMMARY2.2.1ContainmentStructureandReactorBuildingAssessmentSummary2.2.1.1ContainmentStructureAssessmentSummary.2.2.1.2ReactorBuildingAssessmentSummary222223ContainmentSubmergedStructuresAssessmentSummaryPipingSystemsAsessmentSummary 20SUMMARYThisDesignAssessmentReportcontainstheSSESadequacyevaluationfordynamicloadsduetoLOCAandSRVdischarge.2-2 21LOADDEFINITIONSUMNARY2.1.1SRVLoadDefinitionS~nmmarHydrodynamicloadsresultingfromSRVactuationfallintotvodistinctcategories:loadsontheSRVsystemitself(thedischargelineandthedischargequencherdevice),andtheairclearingloadsonthesuppressionpoolwallsandsubmergedstructures.LoadsontheSRVsystemduringSRVactuationincludeloadson,theSRVpipingduetoeffectsofsteadybackpressure,transientvaterslugclearing,andSRVlinetemperature.Determinationofloadingonthequencherbody,arms,andsupportisbasedontransientsresultingfromvalveopening(waterclearingandairclearing),valveclosing,andoperationofanadjacentquencher.Airclearingloadsareexaminedforfourloadingcases:symmetric(allvalve)SRVactuation,asymmetricSBVactuation,singleSRVactuation,andAutomaticDepressurizationSystem{ADS)actuation.Dynamicforcingfunctionsforloadingofthecontainmentwalls,pedestal,basemat,andsubmergedstructuresaredevelopedusingt'echniquesdevelopedinSection4.1.LoadsontheSRVsystemduetoSRVactuationarediscussed.inSubsection4.1.2,andloadsonsuppressionpoolstructuresduetoSRVactuationarediscussedinSubsection4.1.3.Afullscale,unitcelltestprogramisheingemployedtoverifySSESuniqueSRVloadingasdescribedinChapter8.2.12LOCALoadDefinitionSummaryThespectrumofLOCA-inducedloadsontheSSEScontainmentstructureischaracterizedbyLOCAloadsassociatedwithpoolsvell,condensationoscillationandchuggingloads,aswellaslongtermLOCAloads.TheLOCAloadsassociatedvithpoolsvellresultfromshortdurationtransientsandincludedowncomerclearingloads,w'aterjetloads,poolsvellimpactanddragloads,poolfallbackdragloads,poolswellairbubbleloads,andloadsduetodryvellandvetwelltemperatureandpressuretransients.TechniquesusedtoevaluatetheseloadsaredescribedinSubsection4.21.Condensationoscillationsresultfrommixedflow{air/steam)andpuresteamfloveffectsinthesuppressionpool.Chuggingloadsresultfromlovmassfluxpuresteamcondensation.TheloaddefinitionsforthesephenomenaarecontainedinSubsection4.2.2.LongtermLOCAloadsresultfromthosevetvellanddryvellpressureandtemperaturetransientswhichareassociatedwithdesignbasisaccidents(DBA),intermediateaccidents(IBA),andsmallbreakaccidents(SBA).TheirloaddefinitionsarecontainedinSubsection4.2.3. | DocumentNumber3~4,5.6.7.8.TitleTestsonmixedcondensationwithmodelquenchersCondensationandventclearingtestsatGKMwithquenchersConceptanddesignofthepressurereliefsystemwithquenchersKKBventclearingwithquencherTestsoncondensationwithquencherswhensubmergenceofquencherarmsisshallowKKB-ConceptandtaskofpressurereliefsystemStatusCompletedCompletedCompletedCompletedCompletedCompletedDocumentationKWV-Report2593KWV-Report2594KWV-Report2703KWV-Report2796KWV-Report2840KWV-.Report2871UsedforSSESLicensinYesYesYesYesYesYes9.ExperimentalapproachtoventclearinginamodeltankCompletedKMV-Report3129Yes10.KKB-Specificationofblowdowntestsduringnon-nuclearhot.functionaltest-Rev.IdatedOctober4,1974CompletedKWU/V822ReportYesAnticipateddataforblowdowntestswith-pressurereliefsystemduringthenon-nuclearhotfunctionaltestatnuclearpowerstationBrunsbuttel(KKB)CompletedKWU-Report3141Yes12.13.Resultsofthenon-nuclearhotfunctionaltestswiththepressurereliefsysteminthenuclearpowerstationBrunsbuttelAnalysisoftheloadsmeasuredonthepressurereliefsystemduringthenon-nuclearhot,functionaltestatKKBCompletedCompletedKWU-Report3267KWU-Report3346YesYes14.KKB-Listingoftestparametersandimportanttestdataofthenon-nuclearhotfunctionaltestswiththepressurereliefsystemCompletedKWU-WorkingReportR521/40/77Yes15.KKB-Specificationofadditionaltestsfortestingofthepressurereliefvalvesduringthenuclearstart-up,Rev.1CompletedKWU/V822TAYes16.KKB-Resultsfromnuclearstart-uptestingofpressurereliefsystemCompletedKNJ-WorkingReportR142-136/76Yes17'uclearPowerStationPhillipsburg-Unit1HotFunctionalTest:SpecificationofpressurereliefvalvetestsaswellasemergencycoolingandwetwellcoolingsystemsCompletedKWU/V822/RF13Yes DocumentNumberTitleStatusDocumentationUsedforSSESLicensin18.Resultsofthenon-nuclearhotfunctionaltestswiththepressurereliefsysteminthenuclearpowerstationPhillipsburgCompletedKWU-WorkingReportR142-38/77Yes19.KKPI-Listingoftestparametersandimportanttestdataofthenon-nuclearhotfunctionaltestswiththepressurereliefsystemCompletedKWU-WorkingReportR521/41/77Yes20.AiroscillationsduringventclearingwithsingleanddoublepipesCompletedAEG-Report2327Yes616715/cak TABLE1-3SSESCONTAINMENTDESIGNPAKQKTERSA.DrellandSuressionChamber1.InternalDesignPressure2.ExternalDesignPressure3.DrywellFloorDesignDifferentialPressureUpwardDownward4.DesignTemperature5.DrywellFreeVolume(Minimum)(includingvents)(Normal)(Maximum)6.SuppressionChamberVolumeFree(Minimum)(Normal)(Maximum)7.SuppressionChamberWaterVolume(Minimum)(Normal)(Maximum)8.PoolCross-SectionArea~Drell53psig5psid340'F239,337ft33239,593ft3239,850ft.28psid28psidSuressionChamber45psig5psid220'F148,590ft3153,860ft3159,130ft122,410ft3131,550ftGross(OutsidePedestal)TotalGross(IncludingPedestalWaterArea)Free(OutsidePedestal)TotalFree5379ft5679ft5065ft5365ft CHAPTER2SUMMARYTABLEOPCONTENTSLOADDEFINITIONSUMMARY21.1SRULoadDefinitionSummary2.1.2LOCAI.oadDefinitionSummaryDESIGNASSESSMENTSUMMARY2.2.1ContainmentStructureandReactorBuildingAssessmentSummary2.2.1.1ContainmentStructureAssessmentSummary.2.2.1.2ReactorBuildingAssessmentSummary222223ContainmentSubmergedStructuresAssessmentSummaryPipingSystemsAsessmentSummary 20SUMMARYThisDesignAssessmentReportcontainstheSSESadequacyevaluationfordynamicloadsduetoLOCAandSRVdischarge.2-2 21LOADDEFINITIONSUMNARY2.1.1SRVLoadDefinitionS~nmmarHydrodynamicloadsresultingfromSRVactuationfallintotvodistinctcategories:loadsontheSRVsystemitself(thedischargelineandthedischargequencherdevice),andtheairclearingloadsonthesuppressionpoolwallsandsubmergedstructures.LoadsontheSRVsystemduringSRVactuationincludeloadson,theSRVpipingduetoeffectsofsteadybackpressure,transientvaterslugclearing,andSRVlinetemperature.Determinationofloadingonthequencherbody,arms,andsupportisbasedontransientsresultingfromvalveopening(waterclearingandairclearing),valveclosing,andoperationofanadjacentquencher.Airclearingloadsareexaminedforfourloadingcases:symmetric(allvalve)SRVactuation,asymmetricSBVactuation,singleSRVactuation,andAutomaticDepressurizationSystem{ADS)actuation.Dynamicforcingfunctionsforloadingofthecontainmentwalls,pedestal,basemat,andsubmergedstructuresaredevelopedusingt'echniquesdevelopedinSection4.1.LoadsontheSRVsystemduetoSRVactuationarediscussed.inSubsection4.1.2,andloadsonsuppressionpoolstructuresduetoSRVactuationarediscussedinSubsection4.1.3.Afullscale,unitcelltestprogramisheingemployedtoverifySSESuniqueSRVloadingasdescribedinChapter8.2.12LOCALoadDefinitionSummaryThespectrumofLOCA-inducedloadsontheSSEScontainmentstructureischaracterizedbyLOCAloadsassociatedwithpoolsvell,condensationoscillationandchuggingloads,aswellaslongtermLOCAloads.TheLOCAloadsassociatedvithpoolsvellresultfromshortdurationtransientsandincludedowncomerclearingloads,w'aterjetloads,poolsvellimpactanddragloads,poolfallbackdragloads,poolswellairbubbleloads,andloadsduetodryvellandvetwelltemperatureandpressuretransients.TechniquesusedtoevaluatetheseloadsaredescribedinSubsection4.21.Condensationoscillationsresultfrommixedflow{air/steam)andpuresteamfloveffectsinthesuppressionpool.Chuggingloadsresultfromlovmassfluxpuresteamcondensation.TheloaddefinitionsforthesephenomenaarecontainedinSubsection4.2.2.LongtermLOCAloadsresultfromthosevetvellanddryvellpressureandtemperaturetransientswhichareassociatedwithdesignbasisaccidents(DBA),intermediateaccidents(IBA),andsmallbreakaccidents(SBA).TheirloaddefinitionsarecontainedinSubsection4.2.3. | ||
StructuresdirectlyaffectedbyLOCAloadsincludethedrywellwallsandfloor,wetwellwalls,RPVpedestal,basemat,linerplate,columns,downcomers,downcomerbracingsystem,guenchers,andwetwellpiping.TheirloadingconditionsaredescribedinSubsection424.2-4 22DESIGNASSESSMENTSUNKARYDesignassessmentoftheSSESstructuresandcomponentsisachievedbyanalyzingtheresponseofthestructuresandcomponentstotheloadcombinationsexplainedinChapter5.InChapter7,predictedstressesandresponses(fromtheloadsdefinedinChapter4andcombinedasdescribedinChapter5)arecomparedviththeapplicablecodeallovablevaluesidentifiedinChapter6;theSSESdesignvillbeassessedasadequatebyvirtueofdesigncapabilitiesexceedingthestressesorresponsesresultingfromSRVdischargeorJ.OCAloads.2.2.1ContainmentStructureandReactorBuildingAssessmentSum~mar2.2.1.1ContainmentStructureAssessmentSummaryTheprimarycontainmentvalls,baseslab,diaphragmslab,reactorpedestal,andreactorshieldareanalyzedfortheeffectsofSRVandLOCAinaccordancevithTable5-1.TheANSYSfiniteelementprogramisusedforthedynamicanalysisofstructures.Responsespectracurvesaredevelopedatvariouslocationswithinthe~containment>>structure.toassesstheadequacyof,components.,Stressresultantsduetodynamicloadsarecombinedwithother1oadsinaccordancewithTable5-1to'evaluaterebarandconcretestresses.'esign'safetymarginswillaredefinedbycomparingtheactualconcreteandrebarstressesatcriticalsectionsvith,thecode,allowablevalues.2.2.'1.2ReactorBuildingAssessmentSummaryThereactorbuildingisassessedfortheeffectsofSRVandLOCAloadsinaccordancewithTable5-1.ContainmentbasemataccelerationtimehistoriesareusedtoinvestigatethereactorbuildingresponsetotheSRVandLOCAloads.Responsespectracurvesatvariousreactorbuildingelevationsareusedtoassesstheadequacyofcomponentsinthereactorbuilding.2.2.2ContainmentSubmergedStructuresAssessmentSummaryDesignassessmentofthesuppressionchambercolumnsanddovncomerpipesisbeingperformed.Baseduponanapproximate,equivalentstaticanalysiscarriedouttodate,strengtheningofthesestructuresshouldnotberequired.ThisconclusionvillbeconfirmedwhenthedynamicanalysisiscompletePreliminaryresultsfromthedynamicanalysisofthesuppressionpoollinerplateindicatethatnostructuralmodificationsarerequiredThisconclusionwillbeconfirmedwhenthefinalanalysisiscomplete.2-5 Theoriginaldowncomerbracinghasbeenredesignedwithpipesectionstominimizebracinqdragloadsduetopoolsvellandfallback.Therevisedbracingsystemisdesignedusingasimplifiedequivalentstaticapproach.ContainmentandreactorbuildingpipingsystemsarebeingdesignedtowithstandtheeffectsofLOCAandSRVinduceddynamicloads.TheloadcombinationsforpipingaredefinedinTable6.1ofRef.10.2-6 31DESCRIPTXOMOFSAFETYRELIEFVALVEDISCHARGESusquehannaUnits1and2areequippedwithasafetyreliefsystemwhichcondensesreactorsteaminasuppressionchamberpool.Bythisarrangement,reactor,steamisconductedtothewetwell.viafastactingsafetyreliefvalvesandquencherequippeddischargelines..ThissectiondiscussesthecausesofSRVdischarge,describestheSRVdischargeproce'ss,andidentifiestheresultantSRVdischargeactuationcases31.1CausesofSVDischa~reDuringcertainreactoroperatingtransients,theSRVsmaybeactuated(bypressure,byelectricalsignal,orbyoperatoraction)forrapidreliefofpressureinthereactorpressurevessel.Thefollowingreactoroperatingtransientshave,beenidentifiedasthosewhichmayresultinSRVactuation:\a.Turbineqeneratortrip(withbypassorwithout)b.Hainsteamlineisolationvalve(NSIV)closurec.Lossofcondenservacuumd.Feedwatercontrollerfailuree.Pressureregulatorfailure-openf.Generatorloadrejection(withandwithoutbypass)q.LossofacpowerhLossoffeedwaterflowTripoftworecirculationpumpsRecirculationflowcontrolfailure-decreasingflowk.InadvertentsafetyreliefvalveopeningAdetaileddescriptionofthesetransientsisprovidedinSection15.2oftheFSAR3.1.2DescriptionoftheSRVDischargePhenomenaandSRV~t.aadinCasesBeforeanindividualsafetyreliefvalveopens,thewaterlevelinthedischargelineisapproximatelyequaltothewaterlevelinthepoolAsavalveopens,steamflowsintothedischargelineairspacebetweenthevalveandthewatercolumnandmixeswiththeair(seedetailedevaluationinChapter3ofRef1,pages6-12through6-14).Sincethedownstreamportionofthedischargelinecontainsawatersluganddoesnotallowan3-3 immediatesteamdischargeintothepool~thepressureinside-theline'increases.Theincreasedpres'sureexpelsthewaterslugfromtheSRVdischargelineandquencher.Themagnitudeofthewaterclearinqpressureisprimarilyinfluencedbythesteamflowratethroughthevalve,thedeqreetowhichenteringsteamiscondensedalongthedischargeline;,walls,thevolumeofthedischargelineairspace,andthelengthofthewaterslugtobeaccelerated.Theclearingofwaterisfollowedbyanexpulsionoftheenclosedair-steamvolume.Theexhaustedgasformsanoscillatingsystemwiththesurroundingwater,wherethegasactsasthespringandthewateractsasthemass.Thisoscillatingsystemisthesourceofshorttermairclearingloads.Whiletheair-steammixtureoscillatesinthepoolitrisesbecauseofbuoyancyandeventuallybreaksthroughthepoolwatersurfaceatwhichtimeairclearingloadscease.Whenall'theairleavesthesafetyreliefsystem,steamflowsintothesuppressionpoolthrough'hequencherholesandcondenses.TheSSESquencherdesignassuresstablecondensationevenwithelevatedpoolwatertemperature.TheSBVactuationcasesresultingfromthetransientslistedinSubsection3.1.1areclassified,asbeingoneofthefollowingcases:a.Symmetric(allvalve,orAOT)dischargeb.Asymmetricdischarge,includingsinglevalvedischargec.AutomaticDepressurizationSystem(ADS)discharqeAlsoconsideredinthecontainmentdesignistheeffectofsubsequentSRVactuations(second-pop),discussedinSubsection4.1 6.Thesymmetricdischargecase(otherwisetermedtheall-valve,orAOT,case)isclassifiedasthetypeofSRVdischargethatwouldfollowrapidisolationofthevesselfromtheturbinesuchasturbinetrip,closureofallMSXVs,lossofcondenservacuum,etc.Aspressurebuildsupfollowingisolationofthe'vessel,theSBVsactuatesequentiallyaccordingtothepressuresetpointsofthevalves.ThismayormaynotresultinactuationofalltheSRVs,butforconservatisminloadingconsiderationsallvalvesareassumedtoactuateRefertoSubsection4.1.3.1fordiscussionoftheloadsresultingfromthisall-valvecase.AsymmetricdischarqeisdefinedasthefiringoftheSRVsforthe~threead'jacentquencherdeviceswhichresultsinthegreatestasymmetricpressureloadinqonthecontainment.Thissituationishypot'hesizedwhen,followingareactor'cramandisolationoftheVessel,decayheatraisesvesselpressuresothatlowsetpointvalvesactuate.Xf,duringthistimeofdischargeofdecayheatenergy,manualactuationofthetwootheradjacentSRVsthat3-4 comprisetheasymmetriccaseisassumed,thisactuationwouldresultinthemaximumsymmetricpressure'loadonthecontainment.Subsection4.1.3.2givesadiscussionoftheloadsresultingfromtheasymmetricdischarqecase.ThesinglevalvedischargecaseisclassifiedasthefiringoftheSRVwhichqivesthesinglelargesthydrodynamicload.TransientsthatcouldpotentiallyinitiatesuchacaseareaninadvertentSRVdischarqeorDesignBasisAccident(DBA).RefertoSubsection3.2.3foradiscussioriofthelatterpossibilitySubsection4.1.3.2.1providesadiscussionoft'eloadsresultingfromthesinglevalvecase.TheADSdischargeisdefinedasthesimultaneousactuationofthesixSRVsassociatedwiththeADS.SeePigure1-4forthelocationofthequencherdevicesassociatedwiththeADSvalves.TheADSisassumedtoactuatedurinqanIntermediateBreakAccident(IBA)orSmallBreakAccident'(SBA).IfanADSdischargeishypothesized'oincidenttoanIBAorSBA(describedinSubsections3.2.2and3.2.1,respectively),theeffectsofanincreasedsuppressionpooltemperature(resultingfromsteamcondensationduringtheLOCAtransient)andincreasedsuppressionchamberpressure(resultinqfromclearingofthedryvellairintothepooldurinqthetransient)areconsideredinthecalculationofpressureloadingsfortheADSdischargecase.SeeSubsection4.1.3.'3forfurtherdiscussionoftheloadsresultingfromtheADScase.I3-5 32DESCRIPTIONOFLOSS-OF-COOLANTACCIDENT'hiseventinvolvesthepostulationofaspectrumofpipingbreaksinsidethecontainmentvaryinginsizetype,andlocationofthebreak.Fortheanalysisofhydrodynamicloadingsonthecontainment,thepostulatedLOCAeventisidentifiedas.aSmallBreakAccident(SBA),anIntermediateBreakAccident(IBA),oraDesignBasisAccident(DBA).32.1SmallBreakAccientSB~AThissubsectiondiscussesthe.containmenttransientassociatedwithsmallprimarysystemblowdowns.TheprimarysystemrupturesiathiscategoryarethoserupturesthatwillnotresultinreactordepressurizationfromeitherlossofreactorcoolantorautomaticoperationoftheECCSequipment,ie,thoseruptureswithabreaksize.lessthan01sqftThefollowinqsequenceofeventsisassumedtooccurWiththereactorandcontainmeatoperatingatthemaximumnormalconditions,asmallbreakoccursthatallowsblowdownofreactorsteamorwatertothedrywell.Theresultingpressureincreaseinthedrywellleadstoahiqhdrywellpressuresignalthatscramsthereactorandactivatesthecontainmentisolationsystem.Thedrywellpressurecontinuestoincreaseataratedependentuponthesizeofthesteamleak.Thepressureincreaselowersthewaterlevelinthe-downcomers.At.thistime,airandsteamenterthesuppressionpoolataratedependentuponthesizeoftheleak.Onceallthedrywellairiscarriedovertothesuppressionchamber,pressurizationofthesuppressionchamberceasesandthesystemreachesanequilibriumcondition.The'drywellcontainsonlysuperheatedsteam,andcontinuedblowdownofreactorsteamcondensesinthesuppressionpool.Theprincipalloadinqconditioninthiscaseisthegraduallyincreasinqpressureinthedrywellandsuppressionpoolchamberandtheloadsrelatedtothecondensationofsteamattheendofthevents.3.2.2IntermediateBeakAccidentIBAThissubsectiondiscussesthecontainmenttransientassociatedwithintermediateprimarysystemblowdowns.ThisclassificationcoversbreaksforwhichtheblowdownwillresultinlimitedreactordepressurizationandoperationoftheECCS,ie,thebreaksizeisequaltoorslightlyqreaterthan0.1sqft.Followingthebreak,thedrywellpressureincreasesatapproximately1.0psi/sec.Thisdrywellpressuretransientissufficientlyslowsothatthedynamiceffectofthewaterintheventsisnegligibleandtheventswillclearwhenthedrywell-to-suppressionchamberdifferentialpressureisequaltothehydrostaticpressurecorrespondingtotheventsubmergence.The3-6 CHAPTER4LOADDEFINITIONTABLEOFCONTENTS4.1SAFETYRELIEFVALVE(SRV)DISCHARGELOADDEFINITION42LOSS-OF-COOLANTACCIDENT(LOCA)LOADDEFINITION42.142114212421342.1442.1.54.2.1.64.2.1.74.2.24.2.2.1422.242.34.23.14.2.3.2423-342.44.2.4142.424.2.4.342444.2.4.542.46LOCALoadsAssociatedwithPoolswellMetwell/DrywellPressuresduringPoolswellPoolswellImpactI.oadPoolswellDragLoadDowncomerClearingIoadsDowncomerWaterJetLoadPoolswellAirBubbleLoadPoolswellFallbackLoadCondensationOscillationsandChuggingLoadsCondensationOscillationLoadDefinitionChuggingLoadDefinitionLongTermI.OCALoadDefinitionDesignBasisAccident(DBA)TransientsIntermediateBreakAccident(IBA)TransientsSmallBreakAccident(SBA)TransientsLOCA.LoadingHistoriesforSSESContainmentComponentsLOCALoadsontheContainmentMallandPedestalLOCALOadsontheBasematandLinerPlateLOCALoadsontheDrywellandDrywellFloorLOCALoadsontheColumnsLOCALoadsontheDowncomersLOCALoadsonthedowncomerBracing4-1 4.247LOCALoadsonMetwellPiping43ANNULUSPRESSURIZATION44FIGURES45TABLES4-2 CHAPTER4Mum~beTitleP~XGURS4-1through4-37k4-384-394-404-414-424-434-444-454-464-474-48Thesefiguresareproprietaryandare,foundintheproprietarysupplementtothisDAR.SSESShortTermSuppressionPoolHeightSSESShortTermWetwellPressureSSESPoolSurfaceVelocityvsElevationBasemat-SSESWaterClearingJetSSESJetImpingementArea(WaterClearing)SSESPoolswellAirBubblePressureAirBubblePressureonSuppressionPoolWallsSymmetricandAsymmetricSpatial'LoadingSpecificationSSESDrywellPressureResponsetoDBALOCASSESWetwellPressureResponsetoDBALOCASSESSuppressionPoolTemperatureResponsetoDBALOCA4-494-504-514-524-534-544-55SSESDrywellTemperatureResponsetoDBALOCASSESSuppressionPoolTemperatureResponsetoIBATypicalNarkIIContainmentResponsetotheIBATypicalNarkIIContainmentResponsetotheSBASSESComponentsAffectedbyLOCALoadsSSESComponentsAffectedbyLOCALoadsLOCALoadingHistory.fortheSSESContainmentWallandPedestalLocalLoadingHistoryfortheSSESBasematandLinerPlate4-57LOCALoadingHistoryfortheSSESDrywellandDrywellFloor4-3 4-584-594-604-614-62LOCALoadingHistoryfortheSSESColumnsLOCALoadingHistoryfortheSSESDovncomersLOCALoadingHistoryfortheSSESDovncomerBracingSystemLOCALoadingHistoryforSSESQetvellPipingThisfigureisproprietary4-4 CHAPTER4Num~beTitleTABLES4-1through4-15ThesetablesareproprietaryandarefoundintheproprietarysupplementtothisDAR4-164-174-184-194-204-21LOCALoadsAssociatedwithPoolsvellSSESDryvellPressureSSESPlantUniquePoolsvellCodeInputDataInputDataforSSESLOCATransientsComponentLOCALoadChartforSSESHetvellPipingLOCALoadingSItuations 0LOADDEPXNTTION4.1SAFETYRELIEFVALVESB~VDISCRARGELOADDEFINITIONSeetheProprietarySupplement-forthissection.4-6 42LOCALOADDEFINITIONSubsections4.2.1,4.2.2and4.2.3villdiscussthenumericaldefinitionofloadsresultingfromaLOCAintheSSEScontainment.TheLOCAloadsaredividedintothreegroups.V(1)ShorttermLOCAloadsassociatedwithpoolsvell(Subsection4.2.1)(2)Condensationoscillationsandchuggingloads(Subsection4.2.2)(3)LongtermLOCAloads(Subsection4.2.3).TheapplicationoftheseloadstothevariouscomponentsandstructuresintheSSEScontainmentisdiscussedinSubsection4.2.4.421LOCALOADSASSOCIATEDWITHPOOLSWELLAdescriptionoftheLOCA/PoolswelltransienthasbeengiveninSection3.2ofthisDesignAssessmentReport.TheLOCAloadsassociatedvithpoolsvellare-listedinTable4-16.TheappropriateMarkXIgenericdocumentfromwhichSSESplantuniqueloadsarecalculatedisalsoshowninTable4-16.Adiscussionoftheseloa'dsandtheirSSESuniquevaluesfollows.ThedrywellpressuretransientusedforthepoolswellportionoftheLOCAtransient(<2.0seconds)isgiveninTableIV-D-3ofRef7.AportionofthistableisreproducedhereinasTable4-17.Thisdrywellpressuretransientincludestheblovdowneffectsofpipeinventoryandreactorsubcoolingandisthehighestpossibledrywellpressurecaseforpoolsvell.Theshorttermpoolsvellwetwellpressuretransientresultingfromthis'dr@wellpressuretransientiscalculatedbyapplyingthepoolswellmodelcontainedinRef8.TheequationsandassumptionsinthepoolsvellmodelwerecodedintoaBechtelcomputerprogramandverifiedagainsttheClass1,2and3testcasescontainedinRef9.ThisverificationisdocumentedinAppendixDtothisreport.OtherinputsusedforthecalculationoftheSSESplantuniguepoolswelltransientareshowninTable4-18.Theshorttermsuppressionpoolsurfaceelevationandcorrespondingwetwellpressuretransientcalculatedwiththe.poolsvellcodeareshowninPigures4-38and4-39respectively.Theshorttermwetvellpressurepeakis561psia{41.4psig).The(drywellminuswetvell)pressuredifferentialisalsoplottedonthiscurve.TheminimumAPoccurringdurinqpoolswellis-9.2psidat0.893secondsafterventclearing(1.58secondsafterthebreakoccurs)4-7 4.212PoolswellZmactLoa'dAnystructurelocatedbetweenthe.initialsuppressionpoolsurface(el.672')andthepeakpoolswellheight(el690',seefiqure4-38)issubjecttothepoolswellwaterimpactloadThereareonlyminorstructures(suchasmiscellaneouswetwellpipinq)inthisportionoftheSSESwetwall.ThisloadiscalculatedasspecifiedinRef10,S'ubsection44.6.ASSESplant-uniquevelocityyselevationcurvehasbeenqeneratedwiththepoolswellmodel(Figure4-40).I<isusedinconjunctionwithimpactpressurevsvelocitycurvesforvarioussizeandshapecomponents(Ref10,Figures4-34,4-35and4-36)todevelopapeakimpactpressureatthecomponen-t'selevation.Thepe'akimpactpressureiscombinedwithageneralizedimpactpressuretimehistorycurve(Ref10,Figure4-37)tospecifythestructuralload.Allstructures,subjecttopoolswellimpactloadsintheSSEScontainmentareclassifiedas>>smallstructures>>.2.1.3PoolaeellD~aaLoadThepoolswelldragloadappliestoanystructurelocatedbetweentheelevationoftheventexit(el.660')andthepeakpoolswell,heiqht(el.690').Theloadiscalculated'forallcomponentsintheregionbaseduponthemaximumpoolsurfacevelocity(29.35fps),regardlessofelevation.ThedragloadpressureiscalculatedfromRef10,Equation4-24usingVf=29.35fpsforthevelocityandp=62.41bmjft~.forthedensityofwater,fP=(1/2)CDpfVf~(4-1)P(psi)=5.8C'4-2)TheappropriatedragcoefficientforthestructureinvolvedisselectedfromRef10,Figure4-29.Thepoolswelldragloadisappliedineitherthehorizontalorverticaldirection(Subsection4.45.2ofRef10).Forthecaseofacomponentorientedverticallywithitsaxisparalleltothevelocityofthepoolsurface,*theskinfrictioncoefficient,AC~,usedinRef10,Subsection44.8isappliedinplaceofCD.I'hismethodwouldapply,forexample,totheverticalloadsondowncomers,columns,orsafetyrelieflinesinthewetwell.UsinqCf=0.0023,theverticaldragforcepsonaverticallyorientedcomponentisrecalculatedusingEquation4-26ofRef10.F(lbf)=0.0133Af(in~).V(4-3)HereAfistheskinfrictionarea(wettedsurfacearea)subjectto,theverticaldragforce.LOCAloadsonthedowncomerbracingaredescribedinSubsection4246 Verticalloadsonthedowncomersduringdowncomer'learingcanbeestimatedbyusingadragloadformulasimilartoEquation4-3.Inthiscasetheventclearingvelocityis60fps(Ref10,Subsection4.4.5.1)andAfisthewettedinsideareaofthedowncomer,conservativelycalculatedto.beAf={12ft}(m)(2ft)=75~4ft<FromEquation4-3theverticalclearingloadonthedowncomerforSSESis,P=0.6kips.V'hisisofsimilarmagnitudetotheverticalthrustloadof0.7kipsonthedowncomerdurinqsteamblowdown(Ref10,Subsection4.2.3).LateralloadsonthedowncomersduringclearingareestimatedfromRef11,Table3-4tobelessthan3kips.4215DowncomerHaterJetLoadThewaterclearinqjetloadiscalculatedbasedontheapproachdevelopedinthedesignguides{Befs12and13).Thisloadisexperienceasadragloadbystructureslocatedwithinthe)etconebeneaththedowncomersandasa,jetimpingementloadbythebasemat.ThejetimpingementloadonthebasematiscalculatedfromRef10,Equation4-25,p.4-43.Pg=pfAvf2(4-4)Herepisthedensityofwater{takentobe62.4ibm/ft~),Aisthetotaljetimpingementareaandvistheattenuatedwatervelocitycorrespondingtothemaximumventclearinggetvelocity(Ref10).Figures4-41and4-42showelevationandplanviewsoftheSSESdowncomersandtheirassociatedjetcones.Theradiusofthejetconeatthebasematis2.69ft.andthetotalareainterceptedbythe87downcomersintheSSESwetwellis1978ft~.AsseeninFigure4-42thereisnosignificantoverlapofadjacentjetsonthebasemat.Theventclearingvelocityof60fpsisattenuatedbyafactorof0.68usinqthemethoddescribedinRef.10,'ubsection4.4.5.1toyieldavalueof40.8fpsatthebasemat.ThejetimpingementpressureiscalculatedfromRef10,Equation4-26,p.4-43tobeP=pfvIP=22m4ps'.~IUsingthevalueforAof1978ft~fortheSSESdesignthetotaldowncomerwaterjetimpingementloadonthebasematis4-9 F=2848.3kips.Thisloadactsverticallydownwardonthebasematfromthetimethebreakoccursuntil'thedovncomershavecleared,at0.6863sec(Ref7).4.2.1.6PoolswellAirBubbleLoadThepoolsvellairbubblepressureload'asitappliestothecontainmentwallsisdescribedinRef10,Subsection4.4.5.3.ThisloadisviewedasanincreaseinthehydrostaticpressureonthesuppressionpoolwallsbelovtheventexitplaneandiscausedbytheairbubblewhichhasbeenpurgedfromthedrywellintheinitialstagesoftheLOCA.Theairbubblepressuretransientcalculatedwiththepoolsvellmodel(describedinSubsection4.2.1.2)isshovninFigure4-43.Figure4-44shovsthenormalizedtotal.pressuredistribution(hydrostaticplusairbubble)toheappliedtothecontainmentasaresultofthisload.Thepressureonthewetvellwallsbetweentheventexitandthewatersurfacecontainsalineardecreaseto0.0psigatthewatersurface(Ref10,Subsection4.4.5.3).'hisloadasitappliestosubmergedstructuresisdescribedinRefs13and14.4.2.1.7PoolswellFallbackLoadThepoolswellfallback3.oadisadragloadvhichappliestoallstructuresbetweenthepeakpoolswellheight(el.690')andtheventexit(el.660').ThisloadiscalculatedforcomponentsinthisregionusingtheanalysisofSubsection4.4.54ofRef10Sincetheverticalstructuresareparalleltothefallbackflow,theyaresubjectedtonegligiblefallbackloads.(Forafallbackvelocityof30fpstheloadissignificantlylessthan1kip).Thedowncomerbracingstructureatelevation668'-0"is,however,perpendiculartothefallbackflowandvillundergoafallbackloadappliedverticallydovnvard.Thefallbackdragvelocityiscalculatedusingtheequationonpage4-45ofRef10.VFB='.82(8)</~(4-6)FortheSSESdesign,themaximumdowncomersubmergence,Ho,is12feetsothefallbackvelocityis34.05fps.ThedragpressureduetothisvelocityiscalculatedfromRef10,Equation4-24.tobePFB(psi)=78(4-7)whereCpistheappropriatedragcoefficientforthestructurebeingloaded.4-10 PallbackloadsarecalculatedusingRefs12and13.2.2Condensationoscillationsa~adchuin'oadsCondensationoscillationandchugginqloadsfollowthepoolswellloadsintime.Therearebasicallythreeloadsinthistimeperiod,i.e.,fromabout4to60secondsafterthebreak.Condensationoscillationisbrokendownintotwophenomena,amixedflowregiemeandasteamflowregieme.Themixedflowreqiemeisarelativelyhighmassfluxphenomenonvhichoccursduringthefinalperiodofairpurgingfromthedrywelltothevetvell.Thus,themixedflowthroughthedovncomerventscontainssomeairaswellassteam.Thesteamflovportionofthecondensationoscillationphenomenaoccursafteralltheairhasbeencarriedovertothevetwellandarelativelyhighmassfluxofpuresteamflowisestablished.Chuggingisapulsatingcondensationphenomenonwhichcanoccureither.follovingtheintermediatemassfluxphaseofaLOCA,orduringtheclassofsmallerpostulatedpipebreaksthatresultinsteamflowthroughtheventsystemintothesuppressionpoolAnecessaryconditionforchuggingtooccuristhatpuresteam.flowsfromtheLOCAvents.Chuqqingimpartsaloadingconditiontothesuppressionpoolboundaryandallsubmergedstructures.4.2.2.1CondensationOscillationLoadDefinitionTheloadspecificationforthemixedandsteamflowphasesofcondensationoscillationistakenfromAppendixAtoRef20.Themixedflovportionofthecondensationoscillationloadisspecifiedasasinusoidalloadatthecontainment'scriticalfrequencies,between2and7Hzvithanamplitudeofx1.75psi.Thisloadistobeapplieduniformlytothevettedportionofthesuppressionpoolboundarybelowtheventexitwithalinearattenuationtothefreesurfaceo'fthesuppressionpool.Thedurationofthisloadisfrom4to15secondsafterthebreakhasoccurred.Thesteamflowportionofthecondensationoscillationloadisspecifiedasasinusoidalloadatthecontainment'scriticalfreguenciesbetveen2and7Hzvithanamplitudeofa5.0psi.Theloadistobeapplieduniformlyto,thewettedportionofthe:suppressionpoolboundarybelowtheventexitwithalinear.attenuationtothesuppressionpoolfreesurface.Alsoasinusoidalloadofamplitudea0.5psiisapplieduniformlytothedrywellboundaryatcriticalfrequenciesbetween2and7Hz.Thedurationofboththedryvellandsuppressionpoolsteamflovcondensationoscillationloadisthetimeperiodfrom15to25secondsfollowinqtheinitial.break.Condensation'oscillationloadsonsubmergedstructuresarecalculatedusinqRefs12and13.4-11 ThepoolboundarychuggingloadisspecifiedinRef15Tvoloadinqconditionsaredescribed:symmetricandasymmetric.Thesymmetricloadinqconditionisspecifiedas+4.8psig/-4.0psigandistobeapplieduniformlyaroundtheentirepoolboundaryasshovn.inFigure4-45(extractedfromRef15).eTheasymmetricloadingconditionhasaspecifiedmaximumpositive/negativepressureof+20psig/-14psiqandhasthecircumfezentialspatialdistributiondepictedinFigure4-45.ChugqingloadsonsubmergedstructuresvillbeevaluatedwhenthedesiqnguidedealingviththeseloadsiscompletedThechuggingloadimpartedtothedowncomerwillbespecifiedwhentheappropriatedynamicforcingfunctionbecomesavailable4-23LONGTERMLOCALOADDEFINITIONTheloss-of-coolantaccidentcausespressureandtemperaturetransientsinthedrywellandwetvellduetomassandenergyreleasedfromthelinebreak.Thedryvellandwetvellpressureandtemperaturetimehistoriesarerequiredtoestablishthestructuralloadingconditionsi.ntheconta'inmentbecausetheyarethebasisforothercontainmenthydrodynamicphenomena.Theresponsemustbedeterminedforarangeofparameterssuchasleaksize,reactorpressureandcontainmentinit'ialconditions.TheresultsofthisanalysisaredocumentedinRef7.TheDBALOCAforSSESisconservativelyestimatedtobea3.53ft~breakoftherecirculationline(Ref7).'heSSESplantuniqueinputsforthisanalysisareshowninTable4-19.DrywellandvetvellpressureresponsesareshovninFigures4-46and4-47(extractedfromRef7)Thesetransientdescriptionsdonot,hovever,containtheeffectsofreactorsubcoolingSuppressionpooltemperatureresponseisshovninFigure4-48(Ref7~).Thistransientdescriptionalsodoesnotcontaintheeffectofreactorsubcoolinq.Dry'veiltemperaturezesponseisshowninFigure4-49andsimilarlydoesnotcontaintheeffectsofpipeinventoryorreactor'subcooling4.2.3.2IntermediateBreakaccident~IBA)TransientsTheworst-caseintermediatebreakfortheMarkXIplantsisamainsteamlinebreakontheorderof0.05to0.1ft~.AtthistimeplantuniqueIBA,data.forSSESisavailableonlyforthesuppressionpooltemperatureresponsetoa0.05ft>break(Ref7).ThisdataisshowninPiqure4-50.DrywelltemperatureandwetwellanddryvellpressuresfortheSSESXBAareestimatedfromcurvesforatypicalMarkIZcontainmentshowninFigure4-51(extractedfromRef10)4-12 Atthistimeplant-uniqueSBAdataforSSESisnotavailable.Thewetwellanddrywellpressureandtemperaturetransients'foratypicalNarkIIcontainmentareusedtoestimateSSEScontainmentresponsetotheseaccidents.ThesecurvesareshowninFigure4-52{extractedfromRef10).424LOCALOADINGHISTORIESFORSSESCONTAINNENTCONPONENTSThevariouscomponentsdirectlyaffectedbyLOCAloadsareshownschematicallyinFigures4-53and4-54.ThesecomponentsmayinturnloadothercomponentsastheyrespondtotheLOCAloads.Forexample,lateralloadsont'edowncomerventsproduceminorreactionloadsinthedryvellfloorfromwhichthe:downcomer'saresupported.ThereactionloadinthedrywellfloorisanindirectloadresultingfromtheLOCAandisdefinedbytheappropriatestructuralmodelofthedowncomer/drywellfloorsystemOnlythedirectloadinqsituationsaredescribedexplicitlyhere.Table4-20isaLOCAloadchartforSSES.ThischartshowswhichLOCAloadsdirectlyaffectthevariousstructuresintheSSEScontainmentdesignDetailsoftheloadingtimehistoriesarediscussedinthefollovingsubsections..424.1LOCALoadsontheContainmentWallandPedestalFigure4-55showstheLOCAloadinghistoryfortheSSEScontainmentwallandtheRPVpedestal.Thewetvellpressureloadsapplytotheunwettedelevationsinthewetwell;theappropriatehydrostaticpressureadditionismadeforloadsonthewettedelevations.Condensationoscillationandchuggingloadsareappliedtothewettedelevationsinthevetwellonly.Thepoolswellairbubbleloadappliestothevetwellbo'undariesasshowninFigure4-44.42.4.2LOCALoadsontheBasematandLinerPlateFigure4-56showstheLOCAloadinghistoryfortheSSESbasematandlinerplate.WetwellpressuresareappliedtothewettedandunwettedportionsofthelinerplateasdiscussedinSubsection4.2.4.1.Thedowncomerwaterjetimpactsthebasematlinerplateasdoesthepoolsvellairbubbleload.Chuggingandcondensationoscillationloadsareappliedtothewettedportionofthelinerplate.42.4.3LOCALoadsontaheDcwellandD~rwellFloorFiqure4-57showstheLOCAloadinghistoryfortheSSESdrywellanddrywellfloor.Thedrywellfloorundergoesaverticallyapplied,continuouslyvaryingdifferentialpressure,theupwardcomponentofwhichisespeciallyprominentduringpoolswellvhenthevetwellairspaceishighlycompressed4-13 Figure4-58showstheLOCAloadinghistoryfortheSSEScolumns.Poolswelldragandfallbackloadsareveryminorsincethecolumnsurfaceisorientedparalleltothepoolswellandfallbackvelocities.Thepoolswellairbubble,condensationoscillationsandchugqinqwillprovideloadsonthesubmerged{wetted)portionofthecolumns.4.2.4.5LOCALoadsontheDowncomersFiqure4-59showstheLOCAloadinghistory.fortheSSESdowncomers.Thedowncomerclearingloadisalateralloadappliedatthedowncomerexit{inthesamemannerasthechugginglateralload)plusaverticalthrustload.Poolswelldragandfallbackloadsareveryminorsincethedowncomersurfacesareorientedparalleltothepoolswellandfallbackvelocities.Thepoolswellairbubbleloadisappliedtothesubmergedportionofthedowncomerasarethechuggingandcondensationoscillationloads.4.2.4.6LOCALoadsontheDowncomerBracingFigure4-60showstheLOCAloadinqhistoryfortheSSESdowncomerbracinqsystem.Thissystemisnotsubjecttoimpactloadssinceitissubmergedatelevation668'sasubmergedstructureitissubjecttopoolswelldrag,fallbackandairbubbleloads.Condensationoscillationsandchuggingattheventexitwillalsoloadthebracingsystemboththroughdowncomerreaction{indirectload)anddirectlythroughthehydrodynamicloadinginthesuppressionpool.I4.2.47LOCALoadsonMetwellPipingFigure4-61showstheLOCAloadinghistoryforpipingintheSSESwetwell.SincethewetwellpipingoccursatavarietyofelevationsintheSSESwetwell,sectionsmaybecompletelysubmerqed,partiallysubmerged,orinitiallyuncovered.pipingmayoccurparalleltopoolswellandfallbackvelocitiesaswiththemainsteamsafetyreliefpiping.For,thesereasonsthereareanumberofpotentialloadinqsituationswhichariseasshowninTable4-21..Znaddition,thepoolswellairbubbleloadappliesto,thesubmergedportionofthewetwellpipingasdothecondensationoscillationandchuggingloads.'-14 43ANNULUSPRESSURIZATIONTheRPVshieldannulushastherecirculationpumpssuctionlinespassingthroughit{forlocationincontainmentseeFigure1-1).Themassandenergyreleaseratesfrom-apostulatedrecirculationlinebreakconstitutethemostseveretransientinthereactor,shieldannulus.Therefore,thispipebreakisselectedforanalyzingloadingoftheshieldwallandthereactorpressurevesselsupportskirtforpipebreaksinsidetheannulusThereactorshieldannulusdifferentialpressureanalysisandanalyticaltechniquesarepresentedinAppendices6Aand6BoftheSSESFinalSafetyAnalysisReport(FSAB)4-15 17.71ft0.883sec150:J'2wCO~10ZI0L,5IKCgfLwDXT30766.1104770.00.250.50.0.751.0TIMEAFTER'VENTCLEARING(SEC)SUSQUEHANNASTEANQLIKCTRICSTATIONUNITS1ANQ2DESIGNASSESSMENTREPORTSSESSHORTTERMSUPPRESSIONPOOLSURFACEHEIGHTFIGURE4-38 | StructuresdirectlyaffectedbyLOCAloadsincludethedrywellwallsandfloor,wetwellwalls,RPVpedestal,basemat,linerplate,columns,downcomers,downcomerbracingsystem,guenchers,andwetwellpiping.TheirloadingconditionsaredescribedinSubsection424.2-4 22DESIGNASSESSMENTSUNKARYDesignassessmentoftheSSESstructuresandcomponentsisachievedbyanalyzingtheresponseofthestructuresandcomponentstotheloadcombinationsexplainedinChapter5.InChapter7,predictedstressesandresponses(fromtheloadsdefinedinChapter4andcombinedasdescribedinChapter5)arecomparedviththeapplicablecodeallovablevaluesidentifiedinChapter6;theSSESdesignvillbeassessedasadequatebyvirtueofdesigncapabilitiesexceedingthestressesorresponsesresultingfromSRVdischargeorJ.OCAloads.2.2.1ContainmentStructureandReactorBuildingAssessmentSum~mar2.2.1.1ContainmentStructureAssessmentSummaryTheprimarycontainmentvalls,baseslab,diaphragmslab,reactorpedestal,andreactorshieldareanalyzedfortheeffectsofSRVandLOCAinaccordancevithTable5-1.TheANSYSfiniteelementprogramisusedforthedynamicanalysisofstructures.Responsespectracurvesaredevelopedatvariouslocationswithinthe~containment>>structure.toassesstheadequacyof,components.,Stressresultantsduetodynamicloadsarecombinedwithother1oadsinaccordancewithTable5-1to'evaluaterebarandconcretestresses.'esign'safetymarginswillaredefinedbycomparingtheactualconcreteandrebarstressesatcriticalsectionsvith,thecode,allowablevalues.2.2.'1.2ReactorBuildingAssessmentSummaryThereactorbuildingisassessedfortheeffectsofSRVandLOCAloadsinaccordancewithTable5-1.ContainmentbasemataccelerationtimehistoriesareusedtoinvestigatethereactorbuildingresponsetotheSRVandLOCAloads.Responsespectracurvesatvariousreactorbuildingelevationsareusedtoassesstheadequacyofcomponentsinthereactorbuilding.2.2.2ContainmentSubmergedStructuresAssessmentSummaryDesignassessmentofthesuppressionchambercolumnsanddovncomerpipesisbeingperformed.Baseduponanapproximate,equivalentstaticanalysiscarriedouttodate,strengtheningofthesestructuresshouldnotberequired.ThisconclusionvillbeconfirmedwhenthedynamicanalysisiscompletePreliminaryresultsfromthedynamicanalysisofthesuppressionpoollinerplateindicatethatnostructuralmodificationsarerequiredThisconclusionwillbeconfirmedwhenthefinalanalysisiscomplete.2-5 Theoriginaldowncomerbracinghasbeenredesignedwithpipesectionstominimizebracinqdragloadsduetopoolsvellandfallback.Therevisedbracingsystemisdesignedusingasimplifiedequivalentstaticapproach.ContainmentandreactorbuildingpipingsystemsarebeingdesignedtowithstandtheeffectsofLOCAandSRVinduceddynamicloads.TheloadcombinationsforpipingaredefinedinTable6.1ofRef.10.2-6 31DESCRIPTXOMOFSAFETYRELIEFVALVEDISCHARGESusquehannaUnits1and2areequippedwithasafetyreliefsystemwhichcondensesreactorsteaminasuppressionchamberpool.Bythisarrangement,reactor,steamisconductedtothewetwell.viafastactingsafetyreliefvalvesandquencherequippeddischargelines..ThissectiondiscussesthecausesofSRVdischarge,describestheSRVdischargeproce'ss,andidentifiestheresultantSRVdischargeactuationcases31.1CausesofSVDischa~reDuringcertainreactoroperatingtransients,theSRVsmaybeactuated(bypressure,byelectricalsignal,orbyoperatoraction)forrapidreliefofpressureinthereactorpressurevessel.Thefollowingreactoroperatingtransientshave,beenidentifiedasthosewhichmayresultinSRVactuation:\a.Turbineqeneratortrip(withbypassorwithout)b.Hainsteamlineisolationvalve(NSIV)closurec.Lossofcondenservacuumd.Feedwatercontrollerfailuree.Pressureregulatorfailure-openf.Generatorloadrejection(withandwithoutbypass)q.LossofacpowerhLossoffeedwaterflowTripoftworecirculationpumpsRecirculationflowcontrolfailure-decreasingflowk.InadvertentsafetyreliefvalveopeningAdetaileddescriptionofthesetransientsisprovidedinSection15.2oftheFSAR3.1.2DescriptionoftheSRVDischargePhenomenaandSRV~t.aadinCasesBeforeanindividualsafetyreliefvalveopens,thewaterlevelinthedischargelineisapproximatelyequaltothewaterlevelinthepoolAsavalveopens,steamflowsintothedischargelineairspacebetweenthevalveandthewatercolumnandmixeswiththeair(seedetailedevaluationinChapter3ofRef1,pages6-12through6-14).Sincethedownstreamportionofthedischargelinecontainsawatersluganddoesnotallowan3-3 immediatesteamdischargeintothepool~thepressureinside-theline'increases.Theincreasedpres'sureexpelsthewaterslugfromtheSRVdischargelineandquencher.Themagnitudeofthewaterclearinqpressureisprimarilyinfluencedbythesteamflowratethroughthevalve,thedeqreetowhichenteringsteamiscondensedalongthedischargeline;,walls,thevolumeofthedischargelineairspace,andthelengthofthewaterslugtobeaccelerated.Theclearingofwaterisfollowedbyanexpulsionoftheenclosedair-steamvolume.Theexhaustedgasformsanoscillatingsystemwiththesurroundingwater,wherethegasactsasthespringandthewateractsasthemass.Thisoscillatingsystemisthesourceofshorttermairclearingloads.Whiletheair-steammixtureoscillatesinthepoolitrisesbecauseofbuoyancyandeventuallybreaksthroughthepoolwatersurfaceatwhichtimeairclearingloadscease.Whenall'theairleavesthesafetyreliefsystem,steamflowsintothesuppressionpoolthrough'hequencherholesandcondenses.TheSSESquencherdesignassuresstablecondensationevenwithelevatedpoolwatertemperature.TheSBVactuationcasesresultingfromthetransientslistedinSubsection3.1.1areclassified,asbeingoneofthefollowingcases:a.Symmetric(allvalve,orAOT)dischargeb.Asymmetricdischarge,includingsinglevalvedischargec.AutomaticDepressurizationSystem(ADS)discharqeAlsoconsideredinthecontainmentdesignistheeffectofsubsequentSRVactuations(second-pop),discussedinSubsection4.1-3-6.Thesymmetricdischargecase(otherwisetermedtheall-valve,orAOT,case)isclassifiedasthetypeofSRVdischargethatwouldfollowrapidisolationofthevesselfromtheturbinesuchasturbinetrip,closureofallMSXVs,lossofcondenservacuum,etc.Aspressurebuildsupfollowingisolationofthe'vessel,theSBVsactuatesequentiallyaccordingtothepressuresetpointsofthevalves.ThismayormaynotresultinactuationofalltheSRVs,butforconservatisminloadingconsiderationsallvalvesareassumedtoactuateRefertoSubsection4.1.3.1fordiscussionoftheloadsresultingfromthisall-valvecase.AsymmetricdischarqeisdefinedasthefiringoftheSRVsforthe~threead'jacentquencherdeviceswhichresultsinthegreatestasymmetricpressureloadinqonthecontainment.Thissituationishypot'hesizedwhen,followingareactor'cramandisolationoftheVessel,decayheatraisesvesselpressuresothatlowsetpointvalvesactuate.Xf,duringthistimeofdischargeofdecayheatenergy,manualactuationofthetwootheradjacentSRVsthat3-4 comprisetheasymmetriccaseisassumed,thisactuationwouldresultinthemaximumsymmetricpressure'loadonthecontainment.Subsection4.1.3.2givesadiscussionoftheloadsresultingfromtheasymmetricdischarqecase.ThesinglevalvedischargecaseisclassifiedasthefiringoftheSRVwhichqivesthesinglelargesthydrodynamicload.TransientsthatcouldpotentiallyinitiatesuchacaseareaninadvertentSRVdischarqeorDesignBasisAccident(DBA).RefertoSubsection3.2.3foradiscussioriofthelatterpossibilitySubsection4.1.3.2.1providesadiscussionoft'eloadsresultingfromthesinglevalvecase.TheADSdischargeisdefinedasthesimultaneousactuationofthesixSRVsassociatedwiththeADS.SeePigure1-4forthelocationofthequencherdevicesassociatedwiththeADSvalves.TheADSisassumedtoactuatedurinqanIntermediateBreakAccident(IBA)orSmallBreakAccident'(SBA).IfanADSdischargeishypothesized'oincidenttoanIBAorSBA(describedinSubsections3.2.2and3.2.1,respectively),theeffectsofanincreasedsuppressionpooltemperature(resultingfromsteamcondensationduringtheLOCAtransient)andincreasedsuppressionchamberpressure(resultinqfromclearingofthedryvellairintothepooldurinqthetransient)areconsideredinthecalculationofpressureloadingsfortheADSdischargecase.SeeSubsection4.1.3.'3forfurtherdiscussionoftheloadsresultingfromtheADScase.I3-5 32DESCRIPTIONOFLOSS-OF-COOLANTACCIDENT'hiseventinvolvesthepostulationofaspectrumofpipingbreaksinsidethecontainmentvaryinginsizetype,andlocationofthebreak.Fortheanalysisofhydrodynamicloadingsonthecontainment,thepostulatedLOCAeventisidentifiedas.aSmallBreakAccident(SBA),anIntermediateBreakAccident(IBA),oraDesignBasisAccident(DBA).32.1SmallBreakAccientSB~AThissubsectiondiscussesthe.containmenttransientassociatedwithsmallprimarysystemblowdowns.TheprimarysystemrupturesiathiscategoryarethoserupturesthatwillnotresultinreactordepressurizationfromeitherlossofreactorcoolantorautomaticoperationoftheECCSequipment,ie,thoseruptureswithabreaksize.lessthan01sqftThefollowinqsequenceofeventsisassumedtooccurWiththereactorandcontainmeatoperatingatthemaximumnormalconditions,asmallbreakoccursthatallowsblowdownofreactorsteamorwatertothedrywell.Theresultingpressureincreaseinthedrywellleadstoahiqhdrywellpressuresignalthatscramsthereactorandactivatesthecontainmentisolationsystem.Thedrywellpressurecontinuestoincreaseataratedependentuponthesizeofthesteamleak.Thepressureincreaselowersthewaterlevelinthe-downcomers.At.thistime,airandsteamenterthesuppressionpoolataratedependentuponthesizeoftheleak.Onceallthedrywellairiscarriedovertothesuppressionchamber,pressurizationofthesuppressionchamberceasesandthesystemreachesanequilibriumcondition.The'drywellcontainsonlysuperheatedsteam,andcontinuedblowdownofreactorsteamcondensesinthesuppressionpool.Theprincipalloadinqconditioninthiscaseisthegraduallyincreasinqpressureinthedrywellandsuppressionpoolchamberandtheloadsrelatedtothecondensationofsteamattheendofthevents.3.2.2IntermediateBeakAccidentIBAThissubsectiondiscussesthecontainmenttransientassociatedwithintermediateprimarysystemblowdowns.ThisclassificationcoversbreaksforwhichtheblowdownwillresultinlimitedreactordepressurizationandoperationoftheECCS,ie,thebreaksizeisequaltoorslightlyqreaterthan0.1sqft.Followingthebreak,thedrywellpressureincreasesatapproximately1.0psi/sec.Thisdrywellpressuretransientissufficientlyslowsothatthedynamiceffectofthewaterintheventsisnegligibleandtheventswillclearwhenthedrywell-to-suppressionchamberdifferentialpressureisequaltothehydrostaticpressurecorrespondingtotheventsubmergence.The3-6 CHAPTER4LOADDEFINITIONTABLEOFCONTENTS4.1SAFETYRELIEFVALVE(SRV)DISCHARGELOADDEFINITION42LOSS-OF-COOLANTACCIDENT(LOCA)LOADDEFINITION42.142114212421342.1442.1.54.2.1.64.2.1.74.2.24.2.2.1422.242.34.23.14.2.3.2423-342.44.2.4142.424.2.4.342444.2.4.542.46LOCALoadsAssociatedwithPoolswellMetwell/DrywellPressuresduringPoolswellPoolswellImpactI.oadPoolswellDragLoadDowncomerClearingIoadsDowncomerWaterJetLoadPoolswellAirBubbleLoadPoolswellFallbackLoadCondensationOscillationsandChuggingLoadsCondensationOscillationLoadDefinitionChuggingLoadDefinitionLongTermI.OCALoadDefinitionDesignBasisAccident(DBA)TransientsIntermediateBreakAccident(IBA)TransientsSmallBreakAccident(SBA)TransientsLOCA.LoadingHistoriesforSSESContainmentComponentsLOCALoadsontheContainmentMallandPedestalLOCALOadsontheBasematandLinerPlateLOCALoadsontheDrywellandDrywellFloorLOCALoadsontheColumnsLOCALoadsontheDowncomersLOCALoadsonthedowncomerBracing4-1 4.247LOCALoadsonMetwellPiping43ANNULUSPRESSURIZATION44FIGURES45TABLES4-2 CHAPTER4Mum~beTitleP~XGURS4-1through4-37k4-384-394-404-414-424-434-444-454-464-474-48Thesefiguresareproprietaryandare,foundintheproprietarysupplementtothisDAR.SSESShortTermSuppressionPoolHeightSSESShortTermWetwellPressureSSESPoolSurfaceVelocityvsElevationBasemat-SSESWaterClearingJetSSESJetImpingementArea(WaterClearing)SSESPoolswellAirBubblePressureAirBubblePressureonSuppressionPoolWallsSymmetricandAsymmetricSpatial'LoadingSpecificationSSESDrywellPressureResponsetoDBALOCASSESWetwellPressureResponsetoDBALOCASSESSuppressionPoolTemperatureResponsetoDBALOCA4-494-504-514-524-534-544-55SSESDrywellTemperatureResponsetoDBALOCASSESSuppressionPoolTemperatureResponsetoIBATypicalNarkIIContainmentResponsetotheIBATypicalNarkIIContainmentResponsetotheSBASSESComponentsAffectedbyLOCALoadsSSESComponentsAffectedbyLOCALoadsLOCALoadingHistory.fortheSSESContainmentWallandPedestalLocalLoadingHistoryfortheSSESBasematandLinerPlate4-57LOCALoadingHistoryfortheSSESDrywellandDrywellFloor4-3 4-584-594-604-614-62LOCALoadingHistoryfortheSSESColumnsLOCALoadingHistoryfortheSSESDovncomersLOCALoadingHistoryfortheSSESDovncomerBracingSystemLOCALoadingHistoryforSSESQetvellPipingThisfigureisproprietary4-4 CHAPTER4Num~beTitleTABLES4-1through4-15ThesetablesareproprietaryandarefoundintheproprietarysupplementtothisDAR4-164-174-184-194-204-21LOCALoadsAssociatedwithPoolsvellSSESDryvellPressureSSESPlantUniquePoolsvellCodeInputDataInputDataforSSESLOCATransientsComponentLOCALoadChartforSSESHetvellPipingLOCALoadingSItuations 0LOADDEPXNTTION4.1SAFETYRELIEFVALVESB~VDISCRARGELOADDEFINITIONSeetheProprietarySupplement-forthissection.4-6 42LOCALOADDEFINITIONSubsections4.2.1,4.2.2and4.2.3villdiscussthenumericaldefinitionofloadsresultingfromaLOCAintheSSEScontainment.TheLOCAloadsaredividedintothreegroups.V(1)ShorttermLOCAloadsassociatedwithpoolsvell(Subsection4.2.1)(2)Condensationoscillationsandchuggingloads(Subsection4.2.2)(3)LongtermLOCAloads(Subsection4.2.3).TheapplicationoftheseloadstothevariouscomponentsandstructuresintheSSEScontainmentisdiscussedinSubsection4.2.4.421LOCALOADSASSOCIATEDWITHPOOLSWELLAdescriptionoftheLOCA/PoolswelltransienthasbeengiveninSection3.2ofthisDesignAssessmentReport.TheLOCAloadsassociatedvithpoolsvellare-listedinTable4-16.TheappropriateMarkXIgenericdocumentfromwhichSSESplantuniqueloadsarecalculatedisalsoshowninTable4-16.Adiscussionoftheseloa'dsandtheirSSESuniquevaluesfollows.ThedrywellpressuretransientusedforthepoolswellportionoftheLOCAtransient(<2.0seconds)isgiveninTableIV-D-3ofRef7.AportionofthistableisreproducedhereinasTable4-17.Thisdrywellpressuretransientincludestheblovdowneffectsofpipeinventoryandreactorsubcoolingandisthehighestpossibledrywellpressurecaseforpoolsvell.Theshorttermpoolsvellwetwellpressuretransientresultingfromthis'dr@wellpressuretransientiscalculatedbyapplyingthepoolswellmodelcontainedinRef8.TheequationsandassumptionsinthepoolsvellmodelwerecodedintoaBechtelcomputerprogramandverifiedagainsttheClass1,2and3testcasescontainedinRef9.ThisverificationisdocumentedinAppendixDtothisreport.OtherinputsusedforthecalculationoftheSSESplantuniguepoolswelltransientareshowninTable4-18.Theshorttermsuppressionpoolsurfaceelevationandcorrespondingwetwellpressuretransientcalculatedwiththe.poolsvellcodeareshowninPigures4-38and4-39respectively.Theshorttermwetvellpressurepeakis561psia{41.4psig).The(drywellminuswetvell)pressuredifferentialisalsoplottedonthiscurve.TheminimumAPoccurringdurinqpoolswellis-9.2psidat0.893secondsafterventclearing(1.58secondsafterthebreakoccurs)4-7 4.212PoolswellZmactLoa'dAnystructurelocatedbetweenthe.initialsuppressionpoolsurface(el.672')andthepeakpoolswellheight(el690',seefiqure4-38)issubjecttothepoolswellwaterimpactloadThereareonlyminorstructures(suchasmiscellaneouswetwellpipinq)inthisportionoftheSSESwetwall.ThisloadiscalculatedasspecifiedinRef10,S'ubsection44.6.ASSESplant-uniquevelocityyselevationcurvehasbeenqeneratedwiththepoolswellmodel(Figure4-40).I<isusedinconjunctionwithimpactpressurevsvelocitycurvesforvarioussizeandshapecomponents(Ref10,Figures4-34,4-35and4-36)todevelopapeakimpactpressureatthecomponen-t'selevation.Thepe'akimpactpressureiscombinedwithageneralizedimpactpressuretimehistorycurve(Ref10,Figure4-37)tospecifythestructuralload.Allstructures,subjecttopoolswellimpactloadsintheSSEScontainmentareclassifiedas>>smallstructures>>.2.1.3PoolaeellD~aaLoadThepoolswelldragloadappliestoanystructurelocatedbetweentheelevationoftheventexit(el.660')andthepeakpoolswell,heiqht(el.690').Theloadiscalculated'forallcomponentsintheregionbaseduponthemaximumpoolsurfacevelocity(29.35fps),regardlessofelevation.ThedragloadpressureiscalculatedfromRef10,Equation4-24usingVf=29.35fpsforthevelocityandp=62.41bmjft~.forthedensityofwater,fP=(1/2)CDpfVf~(4-1)P(psi)=5.8C'4-2)TheappropriatedragcoefficientforthestructureinvolvedisselectedfromRef10,Figure4-29.Thepoolswelldragloadisappliedineitherthehorizontalorverticaldirection(Subsection4.45.2ofRef10).Forthecaseofacomponentorientedverticallywithitsaxisparalleltothevelocityofthepoolsurface,*theskinfrictioncoefficient,AC~,usedinRef10,Subsection44.8isappliedinplaceofCD.I'hismethodwouldapply,forexample,totheverticalloadsondowncomers,columns,orsafetyrelieflinesinthewetwell.UsinqCf=0.0023,theverticaldragforcepsonaverticallyorientedcomponentisrecalculatedusingEquation4-26ofRef10.F(lbf)=0.0133Af(in~).V(4-3)HereAfistheskinfrictionarea(wettedsurfacearea)subjectto,theverticaldragforce.LOCAloadsonthedowncomerbracingaredescribedinSubsection4246 Verticalloadsonthedowncomersduringdowncomer'learingcanbeestimatedbyusingadragloadformulasimilartoEquation4-3.Inthiscasetheventclearingvelocityis60fps(Ref10,Subsection4.4.5.1)andAfisthewettedinsideareaofthedowncomer,conservativelycalculatedto.beAf={12ft}(m)(2ft)=75~4ft<FromEquation4-3theverticalclearingloadonthedowncomerforSSESis,P=0.6kips.V'hisisofsimilarmagnitudetotheverticalthrustloadof0.7kipsonthedowncomerdurinqsteamblowdown(Ref10,Subsection4.2.3).LateralloadsonthedowncomersduringclearingareestimatedfromRef11,Table3-4tobelessthan3kips.4215DowncomerHaterJetLoadThewaterclearinqjetloadiscalculatedbasedontheapproachdevelopedinthedesignguides{Befs12and13).Thisloadisexperienceasadragloadbystructureslocatedwithinthe)etconebeneaththedowncomersandasa,jetimpingementloadbythebasemat.ThejetimpingementloadonthebasematiscalculatedfromRef10,Equation4-25,p.4-43.Pg=pfAvf2(4-4)Herepisthedensityofwater{takentobe62.4ibm/ft~),Aisthetotaljetimpingementareaandvistheattenuatedwatervelocitycorrespondingtothemaximumventclearinggetvelocity(Ref10).Figures4-41and4-42showelevationandplanviewsoftheSSESdowncomersandtheirassociatedjetcones.Theradiusofthejetconeatthebasematis2.69ft.andthetotalareainterceptedbythe87downcomersintheSSESwetwellis1978ft~.AsseeninFigure4-42thereisnosignificantoverlapofadjacentjetsonthebasemat.Theventclearingvelocityof60fpsisattenuatedbyafactorof0.68usinqthemethoddescribedinRef.10,'ubsection4.4.5.1toyieldavalueof40.8fpsatthebasemat.ThejetimpingementpressureiscalculatedfromRef10,Equation4-26,p.4-43tobeP=pfvIP=22m4ps'.~IUsingthevalueforAof1978ft~fortheSSESdesignthetotaldowncomerwaterjetimpingementloadonthebasematis4-9 F=2848.3kips.Thisloadactsverticallydownwardonthebasematfromthetimethebreakoccursuntil'thedovncomershavecleared,at0.6863sec(Ref7).4.2.1.6PoolswellAirBubbleLoadThepoolsvellairbubblepressureload'asitappliestothecontainmentwallsisdescribedinRef10,Subsection4.4.5.3.ThisloadisviewedasanincreaseinthehydrostaticpressureonthesuppressionpoolwallsbelovtheventexitplaneandiscausedbytheairbubblewhichhasbeenpurgedfromthedrywellintheinitialstagesoftheLOCA.Theairbubblepressuretransientcalculatedwiththepoolsvellmodel(describedinSubsection4.2.1.2)isshovninFigure4-43.Figure4-44shovsthenormalizedtotal.pressuredistribution(hydrostaticplusairbubble)toheappliedtothecontainmentasaresultofthisload.Thepressureonthewetvellwallsbetweentheventexitandthewatersurfacecontainsalineardecreaseto0.0psigatthewatersurface(Ref10,Subsection4.4.5.3).'hisloadasitappliestosubmergedstructuresisdescribedinRefs13and14.4.2.1.7PoolswellFallbackLoadThepoolswellfallback3.oadisadragloadvhichappliestoallstructuresbetweenthepeakpoolswellheight(el.690')andtheventexit(el.660').ThisloadiscalculatedforcomponentsinthisregionusingtheanalysisofSubsection4.4.54ofRef10Sincetheverticalstructuresareparalleltothefallbackflow,theyaresubjectedtonegligiblefallbackloads.(Forafallbackvelocityof30fpstheloadissignificantlylessthan1kip).Thedowncomerbracingstructureatelevation668'-0"is,however,perpendiculartothefallbackflowandvillundergoafallbackloadappliedverticallydovnvard.Thefallbackdragvelocityiscalculatedusingtheequationonpage4-45ofRef10.VFB='.82(8)</~(4-6)FortheSSESdesign,themaximumdowncomersubmergence,Ho,is12feetsothefallbackvelocityis34.05fps.ThedragpressureduetothisvelocityiscalculatedfromRef10,Equation4-24.tobePFB(psi)=78(4-7)whereCpistheappropriatedragcoefficientforthestructurebeingloaded.4-10 PallbackloadsarecalculatedusingRefs12and13.2.2Condensationoscillationsa~adchuin'oadsCondensationoscillationandchugginqloadsfollowthepoolswellloadsintime.Therearebasicallythreeloadsinthistimeperiod,i.e.,fromabout4to60secondsafterthebreak.Condensationoscillationisbrokendownintotwophenomena,amixedflowregiemeandasteamflowregieme.Themixedflowreqiemeisarelativelyhighmassfluxphenomenonvhichoccursduringthefinalperiodofairpurgingfromthedrywelltothevetvell.Thus,themixedflowthroughthedovncomerventscontainssomeairaswellassteam.Thesteamflovportionofthecondensationoscillationphenomenaoccursafteralltheairhasbeencarriedovertothevetwellandarelativelyhighmassfluxofpuresteamflowisestablished.Chuggingisapulsatingcondensationphenomenonwhichcanoccureither.follovingtheintermediatemassfluxphaseofaLOCA,orduringtheclassofsmallerpostulatedpipebreaksthatresultinsteamflowthroughtheventsystemintothesuppressionpoolAnecessaryconditionforchuggingtooccuristhatpuresteam.flowsfromtheLOCAvents.Chuqqingimpartsaloadingconditiontothesuppressionpoolboundaryandallsubmergedstructures.4.2.2.1CondensationOscillationLoadDefinitionTheloadspecificationforthemixedandsteamflowphasesofcondensationoscillationistakenfromAppendixAtoRef20.Themixedflovportionofthecondensationoscillationloadisspecifiedasasinusoidalloadatthecontainment'scriticalfrequencies,between2and7Hzvithanamplitudeofx1.75psi.Thisloadistobeapplieduniformlytothevettedportionofthesuppressionpoolboundarybelowtheventexitwithalinearattenuationtothefreesurfaceo'fthesuppressionpool.Thedurationofthisloadisfrom4to15secondsafterthebreakhasoccurred.Thesteamflowportionofthecondensationoscillationloadisspecifiedasasinusoidalloadatthecontainment'scriticalfreguenciesbetveen2and7Hzvithanamplitudeofa5.0psi.Theloadistobeapplieduniformlyto,thewettedportionofthe:suppressionpoolboundarybelowtheventexitwithalinear.attenuationtothesuppressionpoolfreesurface.Alsoasinusoidalloadofamplitudea0.5psiisapplieduniformlytothedrywellboundaryatcriticalfrequenciesbetween2and7Hz.Thedurationofboththedryvellandsuppressionpoolsteamflovcondensationoscillationloadisthetimeperiodfrom15to25secondsfollowinqtheinitial.break.Condensation'oscillationloadsonsubmergedstructuresarecalculatedusinqRefs12and13.4-11 ThepoolboundarychuggingloadisspecifiedinRef15Tvoloadinqconditionsaredescribed:symmetricandasymmetric.Thesymmetricloadinqconditionisspecifiedas+4.8psig/-4.0psigandistobeapplieduniformlyaroundtheentirepoolboundaryasshovn.inFigure4-45(extractedfromRef15).eTheasymmetricloadingconditionhasaspecifiedmaximumpositive/negativepressureof+20psig/-14psiqandhasthecircumfezentialspatialdistributiondepictedinFigure4-45.ChugqingloadsonsubmergedstructuresvillbeevaluatedwhenthedesiqnguidedealingviththeseloadsiscompletedThechuggingloadimpartedtothedowncomerwillbespecifiedwhentheappropriatedynamicforcingfunctionbecomesavailable4-23LONGTERMLOCALOADDEFINITIONTheloss-of-coolantaccidentcausespressureandtemperaturetransientsinthedrywellandwetvellduetomassandenergyreleasedfromthelinebreak.Thedryvellandwetvellpressureandtemperaturetimehistoriesarerequiredtoestablishthestructuralloadingconditionsi.ntheconta'inmentbecausetheyarethebasisforothercontainmenthydrodynamicphenomena.Theresponsemustbedeterminedforarangeofparameterssuchasleaksize,reactorpressureandcontainmentinit'ialconditions.TheresultsofthisanalysisaredocumentedinRef7.TheDBALOCAforSSESisconservativelyestimatedtobea3.53ft~breakoftherecirculationline(Ref7).'heSSESplantuniqueinputsforthisanalysisareshowninTable4-19.DrywellandvetvellpressureresponsesareshovninFigures4-46and4-47(extractedfromRef7)Thesetransientdescriptionsdonot,hovever,containtheeffectsofreactorsubcoolingSuppressionpooltemperatureresponseisshovninFigure4-48(Ref7~).Thistransientdescriptionalsodoesnotcontaintheeffectofreactorsubcoolinq.Dry'veiltemperaturezesponseisshowninFigure4-49andsimilarlydoesnotcontaintheeffectsofpipeinventoryorreactor'subcooling4.2.3.2IntermediateBreakaccident~IBA)TransientsTheworst-caseintermediatebreakfortheMarkXIplantsisamainsteamlinebreakontheorderof0.05to0.1ft~.AtthistimeplantuniqueIBA,data.forSSESisavailableonlyforthesuppressionpooltemperatureresponsetoa0.05ft>break(Ref7).ThisdataisshowninPiqure4-50.DrywelltemperatureandwetwellanddryvellpressuresfortheSSESXBAareestimatedfromcurvesforatypicalMarkIZcontainmentshowninFigure4-51(extractedfromRef10)4-12 Atthistimeplant-uniqueSBAdataforSSESisnotavailable.Thewetwellanddrywellpressureandtemperaturetransients'foratypicalNarkIIcontainmentareusedtoestimateSSEScontainmentresponsetotheseaccidents.ThesecurvesareshowninFigure4-52{extractedfromRef10).424LOCALOADINGHISTORIESFORSSESCONTAINNENTCONPONENTSThevariouscomponentsdirectlyaffectedbyLOCAloadsareshownschematicallyinFigures4-53and4-54.ThesecomponentsmayinturnloadothercomponentsastheyrespondtotheLOCAloads.Forexample,lateralloadsont'edowncomerventsproduceminorreactionloadsinthedryvellfloorfromwhichthe:downcomer'saresupported.ThereactionloadinthedrywellfloorisanindirectloadresultingfromtheLOCAandisdefinedbytheappropriatestructuralmodelofthedowncomer/drywellfloorsystemOnlythedirectloadinqsituationsaredescribedexplicitlyhere.Table4-20isaLOCAloadchartforSSES.ThischartshowswhichLOCAloadsdirectlyaffectthevariousstructuresintheSSEScontainmentdesignDetailsoftheloadingtimehistoriesarediscussedinthefollovingsubsections..424.1LOCALoadsontheContainmentWallandPedestalFigure4-55showstheLOCAloadinghistoryfortheSSEScontainmentwallandtheRPVpedestal.Thewetvellpressureloadsapplytotheunwettedelevationsinthewetwell;theappropriatehydrostaticpressureadditionismadeforloadsonthewettedelevations.Condensationoscillationandchuggingloadsareappliedtothewettedelevationsinthevetwellonly.Thepoolswellairbubbleloadappliestothevetwellbo'undariesasshowninFigure4-44.42.4.2LOCALoadsontheBasematandLinerPlateFigure4-56showstheLOCAloadinghistoryfortheSSESbasematandlinerplate.WetwellpressuresareappliedtothewettedandunwettedportionsofthelinerplateasdiscussedinSubsection4.2.4.1.Thedowncomerwaterjetimpactsthebasematlinerplateasdoesthepoolsvellairbubbleload.Chuggingandcondensationoscillationloadsareappliedtothewettedportionofthelinerplate.42.4.3LOCALoadsontaheDcwellandD~rwellFloorFiqure4-57showstheLOCAloadinghistoryfortheSSESdrywellanddrywellfloor.Thedrywellfloorundergoesaverticallyapplied,continuouslyvaryingdifferentialpressure,theupwardcomponentofwhichisespeciallyprominentduringpoolswellvhenthevetwellairspaceishighlycompressed4-13 Figure4-58showstheLOCAloadinghistoryfortheSSEScolumns.Poolswelldragandfallbackloadsareveryminorsincethecolumnsurfaceisorientedparalleltothepoolswellandfallbackvelocities.Thepoolswellairbubble,condensationoscillationsandchugqinqwillprovideloadsonthesubmerged{wetted)portionofthecolumns.4.2.4.5LOCALoadsontheDowncomersFiqure4-59showstheLOCAloadinghistory.fortheSSESdowncomers.Thedowncomerclearingloadisalateralloadappliedatthedowncomerexit{inthesamemannerasthechugginglateralload)plusaverticalthrustload.Poolswelldragandfallbackloadsareveryminorsincethedowncomersurfacesareorientedparalleltothepoolswellandfallbackvelocities.Thepoolswellairbubbleloadisappliedtothesubmergedportionofthedowncomerasarethechuggingandcondensationoscillationloads.4.2.4.6LOCALoadsontheDowncomerBracingFigure4-60showstheLOCAloadinqhistoryfortheSSESdowncomerbracinqsystem.Thissystemisnotsubjecttoimpactloadssinceitissubmergedatelevation668'sasubmergedstructureitissubjecttopoolswelldrag,fallbackandairbubbleloads.Condensationoscillationsandchuggingattheventexitwillalsoloadthebracingsystemboththroughdowncomerreaction{indirectload)anddirectlythroughthehydrodynamicloadinginthesuppressionpool.I4.2.47LOCALoadsonMetwellPipingFigure4-61showstheLOCAloadinghistoryforpipingintheSSESwetwell.SincethewetwellpipingoccursatavarietyofelevationsintheSSESwetwell,sectionsmaybecompletelysubmerqed,partiallysubmerged,orinitiallyuncovered.pipingmayoccurparalleltopoolswellandfallbackvelocitiesaswiththemainsteamsafetyreliefpiping.For,thesereasonsthereareanumberofpotentialloadinqsituationswhichariseasshowninTable4-21..Znaddition,thepoolswellairbubbleloadappliesto,thesubmergedportionofthewetwellpipingasdothecondensationoscillationandchuggingloads.'-14 43ANNULUSPRESSURIZATIONTheRPVshieldannulushastherecirculationpumpssuctionlinespassingthroughit{forlocationincontainmentseeFigure1-1).Themassandenergyreleaseratesfrom-apostulatedrecirculationlinebreakconstitutethemostseveretransientinthereactor,shieldannulus.Therefore,thispipebreakisselectedforanalyzingloadingoftheshieldwallandthereactorpressurevesselsupportskirtforpipebreaksinsidetheannulusThereactorshieldannulusdifferentialpressureanalysisandanalyticaltechniquesarepresentedinAppendices6Aand6BoftheSSESFinalSafetyAnalysisReport(FSAB)4-15 17.71ft0.883sec150:J'2wCO~10ZI0L,5IKCgfLwDXT30766.1104770.00.250.50.0.751.0TIMEAFTER'VENTCLEARING(SEC)SUSQUEHANNASTEANQLIKCTRICSTATIONUNITS1ANQ2DESIGNASSESSMENTREPORTSSESSHORTTERMSUPPRESSIONPOOLSURFACEHEIGHTFIGURE4-38 | ||
WETWELLPRESSURE(PSIA)OOOo0QQOVOl4CPOQCmPfllIIOQmZJHfllgmCXfm0QIllmXQZzZpC>COCOm~mCoCOCO~zZc+amrvfltll0OCO.Ozrm(mzoUl0IIIRQOooooo(0CO4l~0AtOMDK4oO(DRYWELL-WETWELL)QP(PSID) fP~yh'1lf,Hk' DOWNCOMERB.O.VENTPIPEEL.660'-0"PEDESTAL"iDIAPHRAGMSLABSUPPORTCOLUMN12'4"EL.648'-0<'ASEMATSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2:.DESIGNASSESSMENTREPORTBASEMATSSESWATERCLEARINGJETFIGURE441 tIIAA-,I1WA~"tt | WETWELLPRESSURE(PSIA)OOOo0QQOVOl4CPOQCmPfllIIOQmZJHfllgmCXfm0QIllmXQZzZpC>COCOm~mCoCOCO~zZc+amrvfltll0OCO.Ozrm(mzoUl0IIIRQOooooo(0CO4l~0AtOMDK4oO(DRYWELL-WETWELL)QP(PSID) fP~yh'1lf,Hk' DOWNCOMERB.O.VENTPIPEEL.660'-0"PEDESTAL"iDIAPHRAGMSLABSUPPORTCOLUMN12'4"EL.648'-0<'ASEMATSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2:.DESIGNASSESSMENTREPORTBASEMATSSESWATERCLEARINGJETFIGURE441 tIIAA-,I1WA~"tt |
Revision as of 22:38, 1 May 2018
ML18017A245 | |
Person / Time | |
---|---|
Site: | Susquehanna |
Issue date: | 06/09/1980 |
From: | CURTIS N W PENNSYLVANIA POWER & LIGHT CO. |
To: | YOUNGBLOOD B J Office of Nuclear Reactor Regulation |
Shared Package | |
ML17138B347 | List: |
References | |
PLA-491, NUDOCS 8006110222 | |
Download: ML18017A245 (325) | |
Text
ARKGULATOYINFORHATIONDISTRIBUTIONSTEM(RIDS)ACCESSIONNBR:80064'10292DOC~DATE!80/06/09NOTARIZED~NOFACIL:50-387SusquehannaSteamKlectricStationiUnitli'Pennsylva50388SusquehannaSteamElectricStationiUnit2iPennsylvaAUTH'AMEAUTHORAFFILIATIONCURTISrN,H~PennsylvaniaPo~er8LightCo..RKCIP,NAMERECIPIENTAFFILIATIONYOUNGBLOODiB~J'I.icensingBranchr"JvDOCKET¹053-030389
SUBJECT:
Forwards"DesignAssessmentReptiRevision2'Proprietaryversionwithheld(ref10CFR2~790)~yP.C7DISTRIBUTIONCOOK:PBOlSCOPIESRECEIVED:LTRKNCLSIZEiTITLE!Proprietar'yInfoRePSAR/FSARNOTES'04KW4CV5Q5R0'tACCAWglf'gRECIPIKNTCOPIKSIDCODE/NAMELTTRENCLACTION:02PM~+~~11'+~BC~~~10INTERNAL:Ol-I1104QAB~1106STRUCENGBR+1109REACSYSBRM1111COREPERFBR1113CONTHNTSYS4f1115PARSYSBRW1118ACDKNTANL01120RADASHTBR~1123KIRK'NOOD~Sf11ADPLANTSYS10AD/CORE8CONT10HPA10OKLD10EXTERNAL:2AACREsr+1gLwR¹3LA'rr/AFF10RKCIPIKNTIDCOOK/NAMEADAO)4A~LA03OPERALIC'R05MECHENGBR07HATL,ENGBR10ANI.BR12AUXSYSBR14ICCSYSBR16ADSITETECH19EFLTTRTSYS21I8EAOFORENGADSITEAAlYSDIRECTORNRR"NRCPORLPDRNSICr"hre~irzIWgrETiverIiEIbriber(LOW,)'re@(L.+0,COPIESLTTRENCL101011W~ig22~W-I<1~l1Mi1Mj4I22eea10100101010<IAAAAStfg7g]9~~@0TOTALNUMBEROFCOPIESRKQUIRKDR-LTTRMENCL e~PPaLTW6NORTHNINTHSTREETALLENTOWNsPA18101PHONEst215)8215151June9,1980Mr.B.J.Youngblood,ChiefLightWaterReactorsBranchNo.3DivisionofProjectManagementU.S.NuclearRegulatoryCommissionWashington,D.C.20SSSSSESDOCKETNOS.50-387850-388DESIGNASSESSMENTREPORT,REVISION2ER100450FILES172-1,840-2PLA491
DearMr.Youngblood:
Transmittedherewithare40copiesofRevision2totheSusquehannaSESDesignAssessmentReport.BothVolume1andtheProprietarySupplementhavebeenrevised.Listedbelowarethemajormodifications.1.RevisionofSection4.2,"LOCALoadDefinition",toreflectthechangesinloadmethodologyrequiredtocomplywiththeOctober,1978NUREG-0487,aswellastheadditionofSubsection4.2.3,"ResponsetoNRCCriteriaforLoadsonSubmexgedStructure".2.UpdateofSection7.0,"DesignAssessment".3.Preparationofanon-proprietaryandproprietarySection9.0,"SSESLOCASteamCondensationVerificationTestGKM-IIM".4.CompletionofAppendixA,"ContainmentDesignAssessment",andAppendixE,"ReactorBuildingStructuralDesignAssessment".S.UpdateofAppendixD,"ProgramVerification",toincludeverificationoftheKWUcomputercodeVELPOT.Lo,6.Rewriteofsubsection8.8.4,"ThermalperformanceofQuenchers".pm8,JP<gPq0p5q.PENNSYLVANIAPOWER8LIGHTCOMPANY8o06110ggr}
Mr.B.J.YoungblocdJune9,1980Page2Inaddition,anumberofeditorialandsyntacticalsentencemodificationshavebeenincluded.Pursuantto10CFR2.790andtheaffidavitsubmittedwithourApril14,1978letter(PLA-244),werequestthatthosepagesmarkedproprietarybewithheldfrompublicdisclosure.Verytrulyyours,,N.i'.CurtisVicePresident-Engineering5ConstructionPAF:JLIPAF128:3 DOCKETEO.SISDATE.'8A'OTETONRCAÃD/ORE,OCAL?UBt'iCDOCENTROOMSfromThefollowingitemsubmitedwithletrdatedisbeingwithheldfromoublicdiscosureinacccrdancewithSecton2.790.PROPRIETARYLVEORCATiONK4stribctionServ-ce'B~ch
~p,flRK0((40ss~*<<<<+UNITEDSTATESNUCLEARREGULATORYCOMMIS&4WASHINGTON,D.C.20555OBQIDL~iFOR:TERA.Corp.FROM:
SUBJECT:
US%5C/TIDC/DistributionServicesBranchSpecialDocumentHandlingRequirementsO1.Pleaseusethefollowingspecialdistributionlistfortheattacheddocument.n2.-Theattached.documentrequiresthefollowingspecialconsiderations:Donotsendoversi.seenclosuretotheHRCPDR.nOnlyoneoversizeenclosurewasreceived-pleasereturnforRegulatoryFilestorage.Proprietaryinformation-sendaff'davitonlytotheHRCPDRQOther:(speciiy)cc:DSBFilesTMC/DSBAuthori"edSignature PROPRIETARYSECTION4TABLESNuebeuTITLE4-1DesignParametersAffectingSRVLoadingQuencherHole:ieldData4-54-74-9HOGE!fInputDataLineLoadsDuringSRVOpeningLineLoadsDuringSRVClosingLineLoadsDurinqIrregularCondensationTotalQuencherLoadsDuringSRVOpeningTotalQuencherLoadsDuringSRVClosinq'TotalQuencherLoadsDuringIrregularCondensation4-104-114-1'2QuencherArmLoadsDuingSRVOpeningQuenche.ArmLoadsDuringSRVClosingQuencherArmLoadsDurinqIr"egula-Condensation4-13measuredParametersRelativetoFigures4-28to4-304-144-15SubmergedStructurePressureDiferenceasaFunctionofBodyDimensionSubmerqedStructure.'luitipliers80061102>R4p-5 i'
PROPRIETARY4.0LOADDZP'I';AITEO'.tA.VSAPETVRPLI:.PVALVE~SRV!DISCHARGELOADDEPINITIORIThissection'providesaqeneraldiscussionoftheapproachusedfordesiqnoftheSSFSSafetyReliefValvesystem(Subsection4.1.1)aswellasthemethodsuedtocalculatesuppressionpoolboundaryandsubmergedstructu"eloads.Forclaritytheloadingconditionshavebeendividedintotwocateqories:'a~b.SRV'ischa"qehydrodynamicloads..exertedontheSRVsystem(pipe,quencher,andsupport)itself(Subsection4.1.2)'SRV'ischargeloadsonthesuppressionpoolboundaryandsubmergedstructures(Subsection4.1.3).4.1..1GeneralDiscussiono"theSSESA2oroachTheSRVsystemused.forSS"=Shasbeo..n,designedbasedonthefollowinqcriteria:a.Redutiontothemaximumextentp"acticableor.ti."'etwellwatersoacedynaicpressu=esassociatedwithSBVdischargeb.Avoidanceofcondensationins-abilitiesassociatedwithhighmassfluxSRVsteamdiscna"gesinno'upto2000~)suppressionpools.'Tosatisfythesecriteria,quenche=shavebeendeveloped~,specificallyforthePennsylvaniaPowerandLightComoany(?PGL)byi<raftwerkUnion(KMU).ASSZS-uniquedynamicloadsDecificationha=-beenpreoa"edbyK~3for-hisdeviceandisdescribedinSubsections4.1.2and4.1.3.Du"inqanextensivequencherdevelor!mentprogram(Ref1),AUhasdeterminedthedeqreeofinfluenceo>>variousSRVsystemdesignparametersonthedynamicPressureswhichresultfromSRVdischarqeandhasconcludedthefollowinq:b.P'oolpressureamplitudesdecreasewithdecreasingpooltemperature.Thisisaconsequenceoftherelationshipbetweenbubblesteamcontentandsaturationconditions.Poolpressureamplitudes"decreasewithincreasingpoolfreewaterarea.TheeffectofeccentricSRVdischargelocationsonpoolpressureamplitudesisnegligible.=4P-7 r
PROPRIETARY.DHIHIT"g"C~Poolpressureamplitudesdecreasewithdecreasingquencherexhausta"ea.Fordecreasingexhaus-areas,theenergy'nputtotheoscillatinqauoble-watersystemisspreadoveralonqertime,withacorrespondingdecreaseinexcitationofthe'oscillatinqsystem.dTheinfluenceofSBVdischargelinelengthovertherangemeasuredbyKMU,9to19m(29.5to62.3ft),isinsignificantforaconstantdischargelineairmass.(4hentwosetsofunit"(Enqlishandmetric)aregiven,thefirstvalueistheoriqinalone;thesecondisanapproximationprovidedforconvenience.)Detailedinformationconcerningeffectsduetolongdischarg".lineswithnumerousbendswillbeobtainedduringtheSusquehannaunitcelltestsdescribedinChapter8.Poolpressureamplitudesdecreasewithdecreasingexpelledairmass,ie,totalenergyinputtothesystem'dec"easeswithdecreasingairvolume.However,theai"volumeinthepipeshouldnotbeconsideredasanabsolutequanticyofinfluence,but=atherasarelativeef"ect,highlydeoendentuponthemassofwateroverwhichtheinputenergyisdistributedandtherateatwhicheno.rqyi"addedtothesystem.fhefollowinqpa"amete=safeetpoo'ressuresbecauseof.heirin"luenceonS'.?V.ischagelineclearingpressures,butarelessimportantthanthosemen"ionedabove:valveopeningtime,steammassflux,SRV0'he.,ai~cnarqlineemperature,andsumerence.morec"mpletelistingofajorandsomelocalizedparametersiscontainedinTableQ-l.TheeffectsofdifferencesinphysicalparametersbetweenSSESandKNUBNRshavebeenaccountedforinthequencherdesignshownonFigureo-1andinTablea-2.~oco"rectprimarilyfo;"hereducedsteammassfluxperS?Vandincreasedlineairvolumes,theSSESquenchershavebeendesignedwithanoutletareaapproximately50percentoftharwhichhasbeenusedforGeman3/Rs.Thisassuresthatoptimumusehasbeenmadeofthedischarqeareaeffectonpressu"eamplitudereduction.FutherdecreasesinoutletareaarenotfeasibleduetotheadverseeffeetonSRVbac<pressuresandSRVdischargelinedesignpressures.TheeffectofSRYdischarqelinelength(thelonqestSSESSRVdischargelineisabouttwiceaslongastnelongestlinepreviouslytestedbyKWU)onpressu"eamplitudewillbestudied;duringtheSSZSuniquetestinqprogram,aswilltheSSBScurved/
PROPRIETARYEXHIBIT"g",'ipe,arranqementwithcespecttoinhibitingsteam-airmixinqpriortoandduringventclearinq.4.1.1.1ThermalPerformariceOneofthekeystotheKRUquencherdeviceisitsabilitytocondensestably(withoutlarqepressureamplitudes)thesteamfractioninexhausts"earn-airmixturesaswellaspuresteamdischarges.Themotrestcictiveconditions,whichinvolvehighsteammassfluxesa"..delevatedpooltemperatues,areofp"imacyimportance.Disc'narqehclepattecnsarearrangedtoenableaninfluxofpoolwaterbetweenadjacentrowsunderalloperatingconditions.Thisarranqement.ensuresthatimmediatecontaciisestablishedbetweenthecoolerpoolwaterandthewarmergas0ischarqed.TheoptimumquencherholepatternverifiedduringtheGKHquencherdevelopmentprogramisusedfortheSSESdesign(Figure4-1andTable4-2).Th10mmdischacqeholesaresDaced15mmoncenters'ndacearrangedinrowswhichareseparatedby50mm.The50mmcenter-to-centerspacingprovidesthepathwayforsupplyinqwaterothes:earn(seePiquce4-2),therebyenablingthepoolobheatedalmos"totheboilingpointwithouta"iseinthepress>reampl;tul.sassociatedwi:hSHVDischarge.Verificationofauencher=hermalpe=focmancemaybefoundinHef(onPiquce513an.page5-34)4.1.2Loadsor.theSRVSystemduetoSHVAetna"ionTheloadingconditionswhichared'esccibed.nthefollowinqsubsectionsa~plytotheSHVpiping,quencherbodyandarms,andquenchersupport.4.1.2.1SP.VLin.BackoressureLoadThemaxi=urnSHVbacknressureduringstealysta-eblowdownwasinvestigatedanalyticallyforthequencherdlscnarqedevice.Thelongestlinegeometrywasusedintheanalysis.ZtwasdeterminedthatthemaximumSHVdischargelineinternalpressureislessthan550psiq.4.1.2.2SHVSystemRaterClearingP=essuce-LoadThissubsectionsummarizestheanalyticaltechniquesemployed.ocalculateinternalpressuresandverticalloadsactingon"heSHVdischargepipingasacesultofwatersluqclearing.Safetyreliefvalvesteamflowwasassumedsatucatedforallcalculations.TheKAU'omputercodeHOGE.'twasusedtocomputethepressure.riseinanSHVdischargelinethcouqhoutthewatecclearingphasewhichfollowstheliftinqofanSHVThecode, tt PROPRIHTARYDIHIHITdocumentedinRef1,hasbeenverifiedwithsubscale(model)and.in-planttestdata.Cl))TheSSES'niqueparameteslistedinTable4-3.have.beenusedas.inputdatatotheHOGEifcomputerproqram.Theflowresistance.coefficientforquenchers.whichhad-beenoptimizedtoparametersuniquetoKNU-designedplantswasfoundtobe=1.5.richencalculatinqtheSSES-uniqueflowresistance'oefficients,particular.conside"ationwasgiventotheSSZS-uniquequencher.Duetodiferentparameters(comparedtoKRUplants),approximatelyone-halfthedischargeareaofearlierKdUdesiqnswasrequied.Usinqanareareductionfactorof0.6,theeffectivedischargeareaoftheSusquehannaguen'cheriscalculatedas:Aeffq0.5Aqeom=0.522m~(5.617ft~)Sincethecross-sectionalaeaofan,SRV-dischaqelineis:0.073m~thearearatiobecomes:A>fcq=072ADTheHOGZIcoderelatesthequencherflowresistancecoefficient,(,tothesquareoftheflowvelocityinsidetheSBVdischargelino,necessitatingthecalculationofavelocityratiobetweenthequencherdischaqeandthepipeflowvelocities.An=1=139MDAeffq0.72where:flowvelocityForquencherstypicalofthoseusedinKMUdesignedplants,thisratioisequaltoone.4Asa.significantpo"tionofthepressu=ereductionmechanismisrelatedtothequencherdischargearea,anappropriateresistance 49pp (139)~Hence,theSusquehannaquencherflowresistancecoefficientis:PBOPBI-TABYSl'pcoefficientwasusedfortheSusquehannaquencrer,basedona.valuewhichhadbeenpreviouslyverifiedforKWUplants.ConsistentwiththeHOGZNcodemethodology,theSSES-uniquevaluewascalculatedbymultiplyingtheKWUvaluebythesquareoftheSSESvelocityratio.-M~~1.93orapproximately2.W~DgSSES=2x1.5=3where"1.5=4forKWVplantsThefollowinqclearinqpressureswerecalculatedfor.thelongestandtheshortest.SRVdischarqelines,respectively,basedonthe-KWUHOGEM'analysis:LencethofBiecl.n~t'line48.3m(158.5ft)349m(1145ft)Calculatedclearingpre-suxe22.7bar(314.5psig)=27.1bar(378.3psiq)'TheclearingSubsequerittothesteadystz"essureti"..ehistoriesaeshownonFigure4-3.waterclearirq,tne'r."erna'ressurechangestotessamflowcondition.'orcalculatinat'everticalloadimposedonthequencherdue<<othedirectionalchanqeinflowvelocity-ithinthequencher(verticalSRVdischarqe,linetohorizonta'lquenc.".erarms),aconserv'tiveesis"ancecoefficient,g=0,wasused(=a<<herthanthevalueE=3.0describedinthepaagraphsabove).Thefollowinaverticalloadsactinqon'anSBVdischarqelineresultf=omachanqeindirectionofthewaterlegdurinqwatercl'earinq:T.engthoftheSBVdischargeline483m34.9mVerticalload490kN(110.2kips)620kH(139.4kips)Thetimehistoriesoftheseverticalloads(with=0)areshownonFiqu=e4-4.4P-11 IL1 PROPRIETARY4123SRVDisch~a".eLineLoadsDurinqwatersluqclearing,thedifferentpiperunsof.theSRVlinearesubjectedtodynamicloadsduetoflowchangeswithinthepipe(pŽessureandmomentumchanges).ThepipinganalysiscontainedinSection5.5includesthese.loads.Figure4-5representstheverticalloadonthelastpipecun(endingwiththequencher).Tables4-4,4-5and4-6listthemaximumloadsexperiencedbyanSRVdischarqelineduringSRVopening,SRVclosinqandirregularcondensation,respectively.4~1..4quencherBodyLoadsOscillatinq.bubblesfromSRVdischargeintothesuppressionpoolproduceexternalloadsonthequenchers.Anoperatingquencherisaffectedbybubblescaused.byitsowndischargeaswellasby'bubblefromadjacentquenchers.Xthasbeenshownexperimentally(Ref-23)thatthemaximumexternalloadingconditiononanindividualquencheroccursdurinqoperationofthequencheitself..heoperationofoneormoreadjacentquenchersdoesriotpcoduceincreasedloads.">>xternalloadsonquencherswhicharenotopeatinqaceevaluatedusingloadingconditionsdescciredinSubsection4.1.3'ccordingtotheiclocationinthepool.Theloadsactingonthequenche:bodya"eshownonFigure4-6.Tables4-7I4-8and4-9listth>>a"imiimloadsexperiencedbyanSSKSquencherduringSR'1opening,SRVclosingandir=egularcondensation,resoectively.Theloadtimehistociesacereferencedinthesametables.Seven"hous'.ndvalveopenings,seventhousandvalveclosingsandonezillionirceqularcondensaticnloadcycleshavebeenassumed.4.1.2.5~~uenchecArmLoads,Theloadsactinqoneacnauencheca=mareshownonFigure4-14.Tables4-10,4-11and4-12listthemaximumloadsexperiencedbyanSS"-Squenche"armdurinqSRVopeninq',SRVclosingandirreqularcondensation,respectively.1heloadtimehisto"iesarerefecencedinthesametables.Seventhousandvalveopeninqs,seventhou=andvalveclosingsandone-millionirregularcondensaticnloadcycleshavebeenassumed.41.2.6quencherSuyoo"tLoadsThequenchersupportshavebeendesignedforthefollowingloads:Loadsactingonthequenche"duetoSRVdischargeasdiscussedinSubsections4.1.2.4and4.1.2.5.
t PROPBI"TABBYb.Loadsfrom'heSBVdischarqe'lineExHIBIT"A"c.Loadsfromflowdeflectionwithinthed'ischargelineLoadsduetooscillatingdischargebubbles4.12.7quencherFatigueLoadsAlthouqheachclearingeventisfollowedbynearlycontinuoussteamflow,steamcondensationdoesnotexhibitaunifombehaviorthrouqhoutthe,entirerangeofsteammassflowratesandwetwellwatertemperaturs.ThevariousregionsofcondensationbehaviorareshownonFigure4-22.Thequencherexperiencesmaximumhydrodynamicandthermalfatigueloadsduring.discontinuousfloworirregularcondensation(transitionregion,Fiqure4-22).TheirregularcondensationloadsfromTable4-9areusedforfatigueconsiderations.Onemilliontotalstresscycles(associatedwiththeirreqularcondensationareassumedfortheanalysis.4.1.3LoadsonSuppressionPoolSt=uctu"esduetoSBVAct>ationThissubsec-iond.sc=ibesloadsonwettedpor"ionsothesuppress'onoolboundarvandsubmergedstructures.Subsections4.1.3.1th=o)gh4.1.3.3aivthecircumferentialpressuredistributionsonth~suppressionpoolboundaries"orthevariousSHVactuationcases.Theverticalpressuredistributionontheboundariesis,discussedinSubsection4.1.3.4.Subsection4.1.3.5givesthepressureti".ehistoriesusedrortheanalysis.4.1..3.'ymmetricLoadinaCon'it'on+SBVAll}Theassumptionthatallqasbubblesar'inqfroSRVdischargeoscillateinphasewiththesamesrena:h(hiqhestpossible)leadsto..hewostloadingcaseasi~scribe.lfor"henormalizedconditiononFigure<<-23.Fortheentireregion(vase."at,containmentwettedwall,andpedestal.wettedwall),themostrestrictivepressuretimehistoriesasdescribedinSubsection4.1.3.5havebeenusedfortheanalysistoensureconservatism.InthelowerregiontheamplitudemultiplierhasbeenchosentobeconsistentwiththeanalysispresentedinSubsection4.1.3.5,whileintheupperregionthesamemultiplierdecreaseslinearlytozeroatthewatersurfaceshownonFiqu"e4-24.4.1.3.2As~mmetricLoadin~Condition.Themostrestrictiveasymmetricalloadingconditionoccurswhenaqroupofadjacentvalvesisoperating.heanalysiswasmadefor,thecasein<<hichthreeadjacentvalvesareoperatinq.The.normalizedpressuredistributionisshownonFigure4-25./
,JIÃ PROPBI""TARKEXHIBIT"P"Thepressuredistributionwasdefinedcircumferentiallyfora180osegment.Onbothsidesofa90~rangewithaconstantpressurelevelthepressuredecreaseslinearlytozeroover450.Ontheother180~segmentofthepool,thepressureswereassumedtobezeroTheverticalpressuredistributionwasassumedtobethesameasforthesymmetricalcase.ThemultiplierdescribedinSubsection41.31i.salsoappliedtothiscase.4.1331Si~n1eValvaActuationLoadingConditionAsymmetric,loadinqalsooccurswhenasinglevalveactuates.Thenormalizedpressure"distributionforthiscaseisshownonFiqure4--26.Thepressurelevelinthecircumferentialdi"ectionremainsconstantoverarangeof300and,onbothsidesofthisranqe,decreaseslinearlyove"47.50to20percentofthemaximumvalue.Outsideofthisreqion,thepressureequals20percentofthemaximuarpressurevalve.Forcomparison,apressuredecreaserelatedtothelaw1/Risshownonthesanefigure.ThepressuredistributionintheradialdirectionisalsoincludedinFigure0-264.1.3.3AutomaticDepressurizationSystem(ADS)LoadingCondition'Assuminghat~he,sixADSvalves("o"locat'on,seeFiqu=e1-4)areactinqinphase,thereisnogreatdifferencebetweenthsym=e"-icand~'-eADSloadingconditions..=.gu=e4-27depictsthenormalizedpressuredistributionusedforthiscase.4.1.3.47ertic=lP"essureDi=tribuionOnceth.@asbubbleshavebeenexpelledfro.a=quencher,theycoa'le-.ceand"heresultingbubbleaqqlomerationrisesduetobuovancyeffectswhileoscillatinq.Becauseofthefreesufacepresence,pr-su"esoncontainmentan"peestalwallsnearthewatersu.facesarelowerthanthepressuresonthebasemat.Zorsuchconfigurationstheobservedvertic~velocityco"ponen"isintheorderof2m/sec.However,theoubhl.oscillationisnearlyd'ampe'doutafterapproxiately1secondascanbeseenonFiqures4-.28to4-30.:Therefore,theassundpressuredecreasewithelevationasshownonFIgure4-24-isconservative.4.1.3.5pressureTimeHistoriesThedefinitionofSBVloadsonsuppressionpoolwetedboundari'esandsub.=erqedinternalscanbelimitedtaloadsresultingfromqasbubbleoscillationfollowingventclea=ing,as:heseloadshavebeenshowntobeboundingwhencomparedtothoseassociatedwiththeotherphasesofSRVdischarge(Ref3).ThissectioncontainsaDiscussionof'ndividualpressuretimehistoriesaswellasspatialeffects EXHIBIT"4"ImmediatelyfollowinqtheliftingofanSRV,amixtureofsteamandairisdischarqedintothesuppressionpool.Thepressuretimehistoriesexperiencedbythesuppressionpoolwettedboundariesandsubmergedstructu"esdifferwithrespecttoamplitudefrequencyanddampinqforeachactuationevent(Ref21).~ToobtainaboundingloadingconditionforSSHScontainmentanalysis,conservatismwithrespecttofrequency,dampinq,andpressureamplitudeisrequired.Theresultingloadsareappliedtothecontainmentinaccordancewiththespatialpressure'distributionsdescribedinSubsections4.1.3.1through4.1.3.4.~Inordertoobtainavalidfrequencyspread,measuredtracesfrompreviousK'2Ufullscaletestingprogramswereselectedand,analyzed.Approximately200runsfromvariousKraftwerkUnionBWRpowerplantswereavailable.Fromthese,threetraceswerechosenfromtheBrunsbuttelnon-,nuclearhot.functionaltestingproqramforuseinSSESdesignverification(forconservatism,subsequentactuationcaseshavebeenused).Thethreetracesare=showninFiqu=es4-28,4-29,and4-30andthetestconditionsaredescribedinReference21.~amorparacletesa"elis=edinTable4-13.DuringtheBrunsbut-el:e=tmeasuredatal'1wallcositiDistanceto"'."enearest'ct-1m(3.28f").The.":.eau=eexoectedtoinclu'eall"a"oscillatione6":ects.'ngproqraŽ,alprossu=eswereonadjacenttotheopera"inqquencher.uatingGuncherarmwasapproxl~'ateiyp"ossu"et"aresa"ethereforeecleai~a/water~et}andai=bubble'Shet"acesusedwereselectednot.onlyfotheirfrequ'encyvariationbu"also"ortheirrelativelylargep=essureamplitudesof0.5to0.8bar(7.25to11.6psia).Figure4-28containsthenighestpressurea..plitudeevermeasuredduringin-planttestingfolio"inqthewaterslugclearingphaseofaKiUquencherequippedsa".~tyreiefsystem.~i!i'theoscillationshowninPiqure4-28isdampedoutrapidly,theothetwotracesexhibitlessdamping.AcomparisonbetweenFiqures4-29and4-30indicatesthat,peakpressureamplitudescanbeexperiencedatdifferenttimes.Figures4-31to4-33containpowerspectraldensityfunctionsfortheinitial0.6sec.oftnemeasuredpressuretraces.Forpu"posesofcomparisonitshoulihementionedthatthetracescontainvariation"inordinatescaling.Inallcasesanair<bubbleoscillaticnfrequencybe"ween6and8cpsisdominant.Althouqthepressureamplitudeofrun435hasthehighestmagnitude(refertoFigure4-28),themaqnitudeofthepowerspectraldensityofthedominant.bubblef"equencyissmall(Ziqure4-31)whencomparedtothetwoothercases(Figures4-324P-15
~'lI'>N' PROPRIETARYIBIT"g"and4-33).rIhentherapiddampingFigure4-28istakenintoaccount,canbepresumedtohaveoccurred.EXHoftheoscillationasshowninauniquebubbleoscillationInadDitiontothemostimportantfirs"0.6secondsofeachtrace,Fiqures4-34to4-36showthepowerspectraldensityfunctionsQurinqlongerperiodsofthesametraces.Thebubble'requencyremainsthedominantfrequencyeventhoughthepressureamplitudesareinpracticedampedoutbeforetheanaly'zertraceends.This-prevailinqfequencyshowsthatthetracesdonotcontainqeometricaleffectssuchaseiqenfrequenciesofthestructure.Therefore,thepressuretimehistoriesareusedaspureforcinqfunctions.Ztshouldbeaddedthatthepressuretransducersusedwerefastenedtoastiffsandwichwall.structuretominimieinteractioneffects.Inordertoobtainaconservativefrequencycontent,'thevariationinairmassbetweentheSusquehannaSRVdischargelinsandthoseusedfocBrunsbuttelweretakenintoconsideration.ThelonqestSS"=Sdischacqelinehasaconservativelyestimatedenclosedairvolue0<<3.1m~(109.5ft~)(refectoTable1-3)whiletheBcunsbutteldi.-eh~roe.lineshaveanenclosedaicvolumeof145"~(v1.2f"~)(ceertoRef2)Thera<<ioofthesevolumesis2.14to1.'ssum1.ilg3.Dverse1ascanbsectionainvecselcanbesshapetofrequen"air-voluasohericalai"bubble,theaicbubblefrequencyisyp00ctlcna1t0thec1beroot0f<<heairvolu2era10eseeninRef4.raflatbubblewithaconstanteros"1areaisassumed,theairbubblef"equencywilloe/protor<<iona1tothesquareccoto=thevolumeratioaseeninRef5.Thisanalysisasuncstherealbubblebbetween'esetwolimit.s,anD-theresultinqvsNifttobebe<<ween<<hetwomodels'redictionofthemeratiopropoctionality.Inordertoobtainaconsrvatiyefrequencycontent,thethreetraces(K'iqures4-28to4-30)whichwereusedasnormalizedforcinqfuncti'orlsweceexpandedintimebya<<zczocl.8(anexpansion)andreducedintimebyafactor0.9(a-contraction).NIthinagivenfre'quencyrangeoneofthethreetracesaffectsanindividuallocationinthecontainmentstructuremoreadverse'.ythantheothers.TheSusquehannaSZSquenchecsweredesignedtocompensateforthefactthatsomeoftheSusquehannaparametersweedifferentfromtheseoftheBrunsbuttel-pl.nt.Toadjusto-'love"valuesofsteammassfluxperSRV,androrthe.greaterinitialerlclosedaicmass,.theexitareaoftheSusquehannaquencne=wasreducedoapproximatelyonehalfofthatofexistinqKMUpowerplants.Any~furtherreductioninquencnerdischarqearea,regardlessofitsdesirability,isunfeasibleduetolimitationsimposedonSRV4P-16 5~~+*~4+&
PROPR1FTARYEXHIBIT"A"dischargelineinternalprssuresaswellasSRVbackpcssuces.Basedontheexperienceobtainedduringthesu'bscaletestinqphaseofKNU'squencherdevelopentprogram(cexertoRef1),itisunlikelythemaximumSS"=Spressureampli-.udeswilleverexceedanormalizedvalueof1.5whenappliedtotheBcunsbuttelpressureamplitudes.Therefore,thisevaluationisbasedonaconservativenormalizedvalueof1.5;thisvalu.;willbeverifieddurinqt~~unxqtcolltestingprogram..whichisexplaineChapter8,ndhasbeenusedinco,;unctionwithpressure-te..(Figures4-28through4->0)forthesuppressionp-wall,pedestal,andbasematadequacyassessments.1.3.7LoadsonSubmergedStructuresduetoSRVActuationThenormalizedpressuretimehistoriespresentedonFigures4-28,4-29,and4-30(refertoSubsection4.l.3.5)arealsousedfortheanalysisofloadsonsubmerqedstructures.TheverticalpressuredistributionofFigure4-24isadopted.Theloadsarecalculatedusinqthepcessuevaluesandthesubmergedsructureprojectedarea.Thecomputedloadsvereassumedtobeactinginthe:lateraldirectionexceptfocthedowncomecbracingandthe'owncomecstiffenerringloads.Thedovnco.".e"b"acingloa'sa"eassumedtobeacinginlateralan"verticaldirectionssimul:aneously..helaeralloadiscalc>latedusinqthereducedpressurevalueaccordingtoFigure4-24.TheverticalloaDiscalculatedusinqtheullpressurevalue.Thedowncomecrinqplateloadsaceassumedtobeactinginthevrticaldi"ection.Thisvecicalloadisalsocalculatedusinqthefullpressurevalue.F~~Similarto"heload=onthesuppressionpoolwettedwalls,a,multipliecwasadoptedwhenapplyingthenoc=alizedpressu"etimehistoriestoaccountfordif<<ecencesbetweenSSFSandBcuns~uttelquenche"s.Thevalueof-hemultipliecwastakendifferen"lydependinqonthesize(diameter)o=t'submergedstcucuce.Discussicn-pertainingtothechoiceofthismultiplieca=eprovidedbelow.Forthecaseofasinglesphericaloscillatinggasbubble,thepressureamplitudesrelativetothesuc"oundinqwate-pressurecanbecalculatedbythesimplerelation:Pressuredifferentialattenuation=whereRo=bubbleradius4P-17
SectionA-AOETKILXlllsl+4CKsaTI'IarrC-e---o-I-""""-$4-~I4'Iss4WLrsaaCassssaalsasa~I,IaaiIIIIsa'II4,esCV~)illaia(sasIICss4saaasal<<TCcsaIIItsFIIIICSIWCraCVSCIdIaIIOITaa74Jl'saalarsaasAgaSrrSCsaA0ETAILYIllslIC~IIt$5Naaaf5/+CasINCIsafafaalI~I.ta/o.SiI5~assIrI~II~5J~Ia~ssssssI--~104sa(sa'IMT~MZ-l44IIOaf5II~4*Lrr,sLll.v'ItiollView4aaaaasraaalalfacasrasafTOTALNO.OFHOLESCa?SIHOLESrIxdlHOI.ES~IlddHOLESlR~0cUmmoc)czmmOQRmgQlCCraDr.mCaaC)ZZ2c+CaaZChIIICalCaaCllAm>IHZZI--I0mC)-C'Clfll0Ch0PlanView
PROPRIETARYEXHIBIT:"A".TABLE4-7TOTAL~U='.)CHERLOADSDURTi'IGSRVOPENING~1>7LoadHaximumValueDirectionTimeHisto~rinternaloverpressure27bars(377psiq)SeeFigure4-7ExternalloadhHaterdeflectionloadinsidethequencherTorque44kn<2>Simultaneouslyin(9891lb)thehorizontalandverticalquencherplanes620knVertical(139,376lb)40knmEnhorizontal.(29,501quencherplaneft-lb)SeeFigure4-8SeeFigure4-5SeeFigure4-9Externalloadduetobubbleoscillation(SeeSubsection4.1.2.4)<1>.Forthecaseofaslidingjointinthedischa"qelineclosetothequencher(Fiqure4-10),tnepressureinsidethepipeactsasanexternalforce.Thiscasisshownin,Fiqure4.11.,<2>Effectsofasymmetricholearrangementareincluded.
PROPRIETARYTABLE-4-8.EXHIBIT"A"~TOTALURNCHRRLOADSDURINGSRVCLOS1NGLoadMaximumValueDirectionTimeHistoryExternalLoadTorqu~4.5kn(1012lb)6knm(4425ft-lb)Simultaneouslyin-.hehorizontalandverti-calquencherplanesXnhorizontalquencherplaneSeePigure4-12SeeFigure.4-'12
PROPRIETARYTABLE4-9TOTALQUENCHERLOADSDURINGIRREGULARCONDENSATIONEXHIBIT."g".1LoadMaximumValueDirectionTimeHistoryExternalload317.5kn(3934lb)Simultaneouslyinthehoizontaland,verticalquencherplanesSeeFigure4-13Torque19knmInhorizontal(14,013quencheplane.Ct-lb)SeeFigure4-13 1
PROPRIETARYTABLE4-11~UENCHERARSLOAQSDURINGSRVCLOSINGEXHIBIT"A"'oad-'HaximumValueDirectionTimeHistorvinternaloverpressureExternalloadBendingmomentonmeldingseamatintersectionbetweenquencnera"mandquencherball22bars(304psiq)4.5Kn{1012lb)3Knm(2213tt-lb)SeeFigure4-18Simultaneouslyinthehorizontalandvertica1planesSeePigure4-19SimultaneouslySeeFigure4-19inthehorizonta1.andverticalplanesTherma1load-2190C(Internaltemper-=(426~-:)atu"-)SeeFigure4-18 f,11 PBOPREETABYTABLE4-12.EXHlHIT"g"QUENCHERARMLOADSDURIMGIRREGULARCOMDENSAIIOMLoadmaximumValueDirect.ionTimeHistoryEnternalpressureExternalload3.0bars(28.8psig)14.5KnSimultaneously(6638inthehorizontalft-lb)andverticaldirectionSeeFigure4-20SeeFigure4-21Bendingmoment,onweldingseamatintersectionbetweenquenche:armandquencherball9Knm(66380t-lb)Simultaneouslyinthehorizons.alandverticalplaneSeeFigure4-21Thermalload(Entlnaemuature)1330C(271.4~F)SeFigure4-20
PROPRIETARYCHAPTER8SSESQUENCHERVERIFICATIONTESTTABLEOFCONTENTS81INTRODUCTION8118.1.281.218.12.1.181.21.2812.281.2218.12.2.281.2.2.38.122-48.1-2.25PurposeofTestsTestConceptSingleCellApproachSingleCellTheoryApplicationofSingleCellApproachSimulationofSSESParameters'rimarySystemPressureSafetyBeliefValve(SBV)DischargeLineVacuumBreakersQuencherBEV1,3r79 82TESTFACILITYANDINSTRUMENTATION8.2.182.1.18.2.1118.2.1128.2.1138211.48.21.158.21168228.2.21822.282.238.2.231822.328.2.2.48.22.4.182.24282.2.5TestFacilityMechanicalSet-UpSteamBoilerSteamAccumulatorSteamLineandBufferTankSafety/RelicfValve(SRV)DischargeLineandQuencherTestTankInstrumentationGeneralDescriptionInstrumentationIdentificationOperatingInstrumentationDisplayonControlConsoleAcquisitionbyComputerTestInstrumentationMeasuringPointsSet-UpofMeasuringInstrumentsVisualRecordingREVli3/798-2 83TESTPARAiiETEHSANDMATRIX83I83.2VentClearingTestsCondensationTestsREV1,3/798-3 84'ZESTRESULTS84.1841.184128.413VentClearingTestResultsTestParametersBehavioroftheSRVandSystemPress'uresDynamicPressureLoadsonthePoolBoundaries841.48428.42.18.4.2.2842218.4.221.18.42.2.12842.2.2LoadsontheQuencherandBottomSupportSteamCondensationTestResultsTestParametersPresentationofTestResultsSurveyofObservedCondensationPhasesBlowdownatLowWaterTemperatureBlowdownatHighMaterTemperatureStatisticalEvaluationoftheDynamicPressureLoadsonthePoolBoundaries8422.21DependenceofDynamicBottomandWallPressuresonSystemPressureandWaterTemperature842222842223OccurrenceFrequencyDistributionsoftheDynamicBottomandMallPressuresStatisticalCharacteristicsoftheDynamicBottomandMallPressures84223TemperatureVariationsintheMaterRegionoftheTestTank8422.4843MaterLevelintheDischargeLineWhenOpeningandAfterClosingtheSRVCheckingandCal.ibrationoftheMeasuringInstrumentation844845AnalysisofMeasurementErrorsRepetitionTestsandReproducibilityoftheResultsREVl,3/798-4 85DATAANALYSISANDVERIFICATIONOFLOADSPECIFICATION8.5.18.5.11EvaluationofTestTankEffectsonBoundaryPressureMeasurementsEffectsofFreeWaterSurfaceandRigidWalls8.5128513851.4MethodofImagesTheTestStandasaSingleCellSpatialDistributionofPressureintheTestTank8.5.1.58515185152851.5.38515.4InvestigationoftheInfluenceofMovableWallsontheMeasurementResults(Fluid-StructureInteraction)GeneralRemarksExperimentalInvestigationoftheTank'sNaturalOscillationsExperimentalInvestigationoftheTank'sResponsetoVentclearingLoadsTheoreticalInvestigationsandModelCalculationsoftheInfluenceofFSI85.1.5.41851.5428.51543ComputationModelsModelParametersandInputforCalculationsWithoutFSI(RigidTank)ModelParametersandInputforCalculationsWithFSI85.1.544ResultsoftheFSICalculations8.5.285.21VerificationofSRVSystemLoadSpecificationDuetoSRVActuationPressuresDuringtheVentClearingProcess852118.5212VentClearingPressuresfortheLongLineVentClearingPressuresfortheShortlineREV.l.3/798-5 8521.3Transposition-oftheMeasurementValuestoSSESandComparisonwiththeDesignSpecification852.2PressuresDuringtheStationaryCondensationofSteam8522185222LongLineShortLine8.522.3TranspositionoftheMeasurementValuestoSSESandComparisonwiththeDesignSpecification8.52.3ExternalLoadsontheQuencherandBottomSupport8.5.2318.5.2.3.1.185231.2852312.185.231.2.2VerticalForceMeasurementoftheVerticalForceMeasuredVerticalForcesLongLineShortLine85231.3TranspositionoftheMeasurementValuestoSSES8523131852.31,.3.28.5.2.313-3852328523218.5.2.3-2.285232.2185232.22852323LongLineShortLineSummaryTorsionalMomentMeasurementoftheTorsionalMomentMeasuredTorsionalMomentslongLineShortLineTranspositionoftheMeasurementValuesto'SSES85-2338.5.2.3.31BendingMomentsattheQuencherArmsMeasurementoftheBendingMomentsREV1,3/798-6 8.5.2-33.28.5.2.3.3.3MeasuredBendingMomentsTranspositionoftheMeasurementResultsIntotheWeld8.5.233.4852.33.58.5-2.3.48523418.5234.28.5.2.3438.523448.5.2358-52-3.6SpecifiedStaticEquivalentLoadsEvaluati.onoftheMeasurementResultsBendingMomentsattheBottomSupportMeasurementoftheBendingMomentsMeasuredBendingMomentsSpecifiedStaticEquivalentLoadEvaluationoftheMeasurementResultsForcesontheQuencherInfluenceofanAdjacentQuencher~8-5.237LoadsontheQuencherDuringSteamCondensation852371ManifestionFormsofIntermittentCondensationintheKarlsteinTests8523728.5.237.3IllustrationoftheMeasurementValuesEvaluationoftheMeasurementResultsforthe.QuencherArm85237.4EvaluationoftheMeasurementResultsfortheBottomSupport8.5-2.375EvaluationoftheMeasuredTorsionalMoments852376EvaluationoftheMeasuredMaximumMomentsattheQuencherArmDuringIntermittentCondensation85.3VerificationofSuppressionPoolBoundaryLoadSpecificationDuetoSRVActuation8531EvaluationoftheLocalEffectsSeenatPressureTransducerP5.58532Veri.ficationoftheSpecifiedPressureAmplitudesandVerticalPressureProfilesafterVentClearingREVl~3/798-7 8532185321.185.3212OverpressuresVerticalPressureProfileVerticalPressureProfileIncludingLocalEffectsatP5.58532285322.185.3385.3.3.185331.18.5.33.11.1853311.28.5.3.3.1.1.385.331.14UnderpressuresVerticalPressureProfileVerificationofthePressureTimeHistoriesUsedfortheSSESContainmentAnalysisTranpositionMethodfortheOscillationFrequencyCalculationofMeasuredOscillationFrequenciesPPGLTestsatKarlsteinGKMModelQuencherTests'KBHotTestsConclusionfromtheFrequencyCalculations8.5.3.328.5.3.3.38533.3.18.53332MultipliersforConversionoftheBubbleFrequenciesFromtheTestStandtoSSESTranspositionMethodforthePressureAmplitudesPPGLQuencherTestsatKarlsteinKMUQuencherTestsintheModelTestStandinKarlstein8.53.3.338533.34853348533.41853.34.2AnalyticalCalculationsInfluenceofBackpressureonthePressureAmplitudesVerificationofDesignSpecificationFrequencyAnalysesofSelectedTestsShiftingofthePSD'sintheTranspositionFromtheTestStandtoSSESREV.li3/798-8 853.34.2l85.334.228533.4385.3.3.4485334585.33.4.6FrequencyShiftAmplitudeStretchingSymmetricalLoadCase(SimultaneousBlowdownofall16SRV's)UnsymmetricalLoadCase(BlowdownViaOneSRV)UnsymmetricalLoadCase(BlowdownViaThreeAdjacentSRV')AutomaticDepressurizationSystem(ADS)LoadCase853347853.3.58.533.518533.5285.33.53853.354SummaryEvaluationoftheMeasuredPressureOscillationsDuringCondensationTheQuencherisClearedContinuallyTheQuencherisNotClearedContinuallyCondensationintheBlowdownPipeandThrutheSlidingJointTransportationoftheMeasurementResultstoSSES854854.18.5.4285.438544PoolMixingDuringSRVActuationandThermalPerformanceoftheQuencherIntroduction'EquationofMotionoftheRotatingPoolDeterminationoftheFlowResistancesDeterminationoftheForceMovingthePool85.4.5WorkingEquationsfortheRotatingPoolofSSES854.68.5.4.7854.7185472EstimateoftheHeatingoftheSuppressionChamberWaterExperimentalProofsModelTankTestsKKBTestDuringtheNuclearCommissioningREVl,3/798-9 8.5.4.7-3GKMHalfScaleQuencherCondensationTest8.5488.5.5SummaryVerificationoftheSubmergedStructuresLoadSpecificationDuetoSRVActuation8.5.5.1855118551285513LoadsontheVentPipeMeasurementoftheLoadsMeasuredBendingMomentsExtrapolationoftheMeasurementResultsandComparisonwiththeSpecifiedValue8.55.2InfluenceofExpelledHaterDuringVentClearing8553SummaryREVl,3f'798-10 SECTION8.0FIGURESNumberTitle8-1MathematicalDesscriptionofaSingleCellConfigurationwithSolidWalls;SolidBottomandFreeWaterSurface8-28-38-48-58-68-7EguivalenceofaSingleCellConfigurationandaParallelBubbleFieldOscillatinginPhaseGeometricSingleCellPartitionoftheSuppressionPoolTestStandSchematicDiagramLongDischargeLineConfigurationShortDischargeLineConfigurationKarlsteinTestTankPlanVievTypicalVentClearingInstrumentation8-8KarlsteinTestTankC-DVievTypicalVentClearingInstrumentation8-9KarlsteinTestTankA-BVievTypicalVentClearingInstrumentation8-10KarlsteinTestTankPlanVievTypicalCondensationTestInstrumentation8-11KarlsteinTestTankC-DViewTypicalCondensationTestInstrumentation8-12KarlsteinTestTankA-BViewTypicalCondensationTestInstrumentation8-13T-QuencherShowingTypicalVentClearingInstrumentation8-14T-QuencherShovingTypicalCondensationTestInstrumentation8-158-168-178-188-19TestMatrigforVentClearingTestLocationofTestGroupNo.1intheOperationFieldLocationofTestGroupNo.2intheOperationFieldLocationofTestGroupNo.3intheOperationFieldLocation'fTestGroupNo.4intheOperationFieldREV1,3/798-11 8-208-218-228-23LocationofTestGroupNo.5intheOperationFieldLocationofTestGroupNo.6intheOperationFieldLocationofCondensationTestsintheOperationFieldValveOpeningTimeVersusAccumulatorPressureLongPipeVentClearingTests8-24ValveOpeningTimeVersusAccumulatorPressureShortPipeVentClearingTests8-25VentClearingPressureVersusSystemPressureLongLineVentClearingTests8-26VentClearingPressureVersusSystemPressureShortLineVentClearingTests8-27PeakPositiveWallandBottomPressuresVersusSystemPressure-Long.Line,CleanConditions,ColdPool8-28PeakPositiveMallandBottomPressuresVersusSystemPressure-ShortLineCleanConditions,ColdPool8-29PeakPositiveWallandBottomPressuesVersusSystemPressure-LongLineRealConditions,ColdPool8-30PeakPositiveWallandBottomPressuresVersusSystemPressure-ShortLine,RealConditions,ColdPool8-31PeakPositiveMallandBottomPresssuresVersusSystemPressure-LongLine,CleanConditions,Heatedpool8-32PeakPositiveWallandBottomPressuresVersusSystemPressure-ShortLine,CleanConditions,HeatedPool8-33PeakPositiveWallandBottomPressuresVersusSystemPressure-LongLine,RealConditions,HeatedPool8-34PeakPositiveMallandBottomPressuresVersusSystemPressure-ShortLine,RealConditions,HeatedPool8-35PeakPositiveMallandBottomPressuresVersusValveActuation-LongPipeTest148-36PeakPositiveWallandBottomPressuresVersusValveActuation-LongPipeTest58-37PeakPositiveMallandBottomPressuresVersusValveActuation-LongPipeTests4and4R8-38PeakPositiveWallandBottomPresuresVersusValveActuation-LongPipeTests15and15RREV1,3/798-12 8-3'98-408-418-428-438-448-458-468-478-488-498-508-518-528-538-548-558-568-578-588-598-608-618-628-638-648-65PeakPositiveMallandBottomPressureVersusValveActuation-ShortPipeTests19and19RPeakPositiveMallandBottomPressuresVersusValveActuation-ShortPipeTests20and20RVisicorderTraceP51-P5.10Test41.1.VisicorderTraceP5.1-P5.10Test4R.l.1VisicorderTraceP5.1-P5.10Test4.1.6VisicorderTraceP5.1-P5.10Test11.1VisicorderTraceP5.1-P5.10Test12.1VisicorderTraceP5.1-P5.10Test15.1.1VisicorderTraceP5.1-P5.10Test15.Rl.1VisicorderTraceP5.1-P5.10Test19.1.1VisicorderTraceP5.1-P5.10Test19.R2.1VisicorderTraceP5.1-P5.10Test19.R2.2VisicorderTraceP5.1-P5.10Test19.R2.3VisicorderTraceP5.1-P5.10Test19.R2.4UisicorderTraceP5.1-P5.10Test19.R2.5VisicorderTraceP5.1-P5.10Test19.R26VisicorderTraceP51-P510Test19.R2.7VisicorderTraceP5.1-P5.10Test19R2.8VisicorderTraceP5.1-P5.10Test19.R29VisicorderTraceP5.1-P5.10Test19.32.10VisicorderTraceP5.1-P5.10Test20.1.1VisicorderTraceP5.1-P5.10Test20.Rl.lVisicorderTraceP5.1-P5.10Test20.R1.10VisicorderTraceP5.1-P5.10Test21.1VisicorderTraceP5.1-P5.10Test21.2VisicorderTraceP5.1-P5.10Test25.1VisicorderTraceP5.1-P5.10Test25.R2REVlg3/798-13 8-66MaximumResultantBendingMomentatQuencherArm1-,LongPipeVentClearingTests8-67MaximumResultantBendingMomentatQuencherArm2-LongPipeVentClearingTests8-688-69MaximumResultantBendingMomentatQuencherArm1-ShortPipeVentClearingTestsMaximumResultantBendingMomentatQuencherArm2-ShortPipeVentClearingTests8-70MaximumResultant'BendingMomentattheQuencherSupport-LongPipeVentClearingTests8-718-728-73MaximumResultantBendingMomentattheQuencherSupport-ShortPipeVentClearingTestsObservedCondensationPhasesDuringTestsTypicalVisicorderTraceofStationaryOperationofQuencherTest33.2-10SecondsafterStart8-74TypicalVisicorderTraceofStationaryOperationofQuencherTest35.1-20-SecondsafterStart8-75VisicorderTraceShovingIntermittentOperationoftheQuencher-Test36.1SysemPressure-6.2-1.0barPoolWaterTemp-26~C-300C8-76VisicorderTraceShovingExcerptfromIntermittentOperationofQuencherTest36.1-280SecondsafterStart8-778-78VisicorderTraceShovingSingleEventOutofIntermittentCondensationTest36.1TypicalVisicorderTraceofStationaryOperationofQuencherTest37.2-13SecondsafterStart8-79TypicalVisicorderTraceofStationaryOperationofQuencherTest39.1-10SecondsafterStart8-80VisicorderTraceShovingIntermittentOperationofQuencher-Test40.1SystemPressure-2.5barPoolWaterTemp.-89~C-91~C8-81DynamicBottomPressuresduringtheBlowdownAlongtheUpperandLoverBoundaryoftheOperationField'-82DynamicWallPressuresDuringtheBlowdownAlongtheUpperandLoverBoundaryoftheOperationFieldREV.1,3/798-14 8-83OccurrenceFrequencyDistributionPositiveandNegativeDynamicAmplitudesfortheCondensationTestsPoolTemp.22~C-300C8-84OccurrenceFrequencyDistributionPositiveandNegativeDynamicAmplitudesfortheCondensationTestsPoolTemp.59~C-91>C8-85OccurrenceFrequencyDistributionPositiveandNegativePressureAmplitudeforCondensationTestsPoolTemp.22C-30C8-86OccurrenceFrequencyDistributionPositiveandNegativePressureAmplitudeforCondensationTestsPoolTemp59C-91C8-878-888-89NeanValuesoftheBottomDynamicPressuresDuringtheBlowdownsAlongtheUpperandLowerBoundaryoftheOperationFieldMeanValuesoftheWallDynamicPressuresDuringtheBlowdownsAlongtheUpperandLowerBoundaryoftheOperationFieldWaterTemperatureTimeHistoriesOnPoolWallCondensationTest33.28-90WaterTemperatureTimeHistoriesOnPoolWallCondensationTest35.18-91MaterTemperatureTimeHistoriesOnPoolMallCondensationTest37.28-92RaterTemperatureTimeCondensationTest39.1HistoriesOnPoolMall8-93WaterTemperatureTimeCondensationTest33.2HistoryOnQuencherArm18-94RaterTemperatureTimeHistoryonQuencherArm1CondensationTest35.18-95MaterTemperatureTimeHistoryonQuencherArm1CondensationTest3728-96MaterTemperatureTimeHistoryon.QuencherArm1CondensationTest-39.18-978-98CalibrationofSensorsandRegistrationInstrumentsIntervalsforCalibrationChecksandAdjustmentsofInstrumentation8-99CalibrationSystemREV.1>>3/79 8-100CalibrationResultsDeviationsfromNominalValue-P5.1-P5-108-1018-1028-1038-104MaterLevelinDischargeLineTest15.1MaterLevelinDischargeLineTest20.1WaterLevelinDischargeLineTest32EffectsofFreeSurfaceandRigidTankWallsonDynamicFluidPressure8-1058-1068-1078-1088-109MethodofImagesSSESSmallestUnitCellandtheKarlsteinTestTankPressureProfilesforDifferentBubbleLocationsPressureProfileforaOneandFourBubbleArrangementComparisonofMeasuredandCalculatedNormalizedPressureProfiles8-110ComparisonofPressureProfilesCalculatedfortheKarlsteinTestTankandtheSSESSuppressionPool8-111ComparisonofCalculatedandSpecifiedPressureProfiles8-1128-113TankArrangementShowingInstrumentationandExplosiveChargeLocationsforMeasuringTankReponseConfigurationofExplosiveContainerUsedtoGenerateUnderwaterPressureImpulse8-1148-1158-1168-1178-1188-1198-1208-1218-122TypicalTankResponseDuetoPressureImpulseFrequencyAnalysisofGageMA2FrequencyAnalysisofGageMA7FrequencyAnalysisofGageMA8FrequencyAnalysisofGageP5.10DisplacementCorrelationsfor13Hz-EigenmodeDisplacementforthe13HzEigenmodeTestTankArrangementforShakedownTestsTankDisplacementsand.PressureTraceDuringShakedownTest08.18-123FrequencyAnalysisofGageMA2ShakedownTest08.1REVli3/798-16 8-1248-1258-1268-127FrequencyAnalysisofGageWA7ShakedownTest08.1FrequencyAnalysisofGageWA8ShakedownTest081FrequencyAnalysisofP510ShakedownTest08.1AirNassPlowusedforKOVlBlComputerCodeCalculations8-128UnitWallDisplacementof13HzNodeUsedinKOVlB1ComputerCodeCalculations8-129BoundryPressureDistributionCalculatedforUnitDisplacementof13HzNode8-1308-1318-1328-133WallPresureCalculationwithKOVlB1ComputerCodeEffectsofFSZonBubblePrequencyTypicalPressureTraceinSRVDischargeLineTest4.1.4TypicalPressureTraceinSRVDischargeLineTest20.Rl78-134PressureinSteamLinebeforeSRVVersusPressureinBufferTankatValueOpening8-1358-136PressureinDischargeLineVersusReactorPressureatVentClearing-P4.1LongLineTestsPressureinDischargeLineVersusReactorPressureatVentClearing-P44LongLineTests8-137PressureinDischargeLineVersusReactorPressureatVentClearing-P4.1ShortLineTests8-138PressureinDischargeLineVersusReactorPressureatVentClearing-P4.4ShortLineTests8-1398-140VentClearingPressureVersusValveOpeningTimeSteadyStatePressureVersusReactorPressure-P4.1LongLineTests8-141SteadyStatePressureVersusReactorPressure-P4.4LongLineTests8-l42SteadyStatePressureVersusReactorPressure-P4.1ShortLineTests8-143SteadyStatePressureVersusReactorPressure-P4.4ShortLineTests8-144SteadyStatePressuresatDifferentLocationsAlongtheDischargeLineExtrapolatedto88BarReactorPressureREV1,3/798-17 8-1458-146.8-147TypicalTraceforVerticalLoadLongLineTestsVerticalLoadVersusClearingPressureLongLineTestsVerticalLoadVersusVentClearingPressureShortLine-Tests8-1488-149TypicalTraceforTorqueonBottomSupportLongLineTestBottomSupportTorqueVersusVentClearingPressureLongLineTests8-150BottomSupportTorqueVersusVentClearingPressureShortl.ineTests8-151TypicalTraceforBendingMomentsonQuencherArmsLongLineTests8-152ResultantQuencherArmBendingMomentVersusVentClearingPressureShortLineTests8-153FrequencyDistributionofMaximumResultant'BendingMomentonQuencherArmsandatWeldSeam8-154ResultantBottomSupportBendingMomentVersusVentClearingPressureShortLineTests8-155FrequencyDistributionofMaximumResultantBendingMomentonBottomSupportSG4.5-468-156FrequencyDistributionofMaximumResultantBendingMomentsOnQuencherArmsgtStrainGagesIntermittentCondensation8-157FrequencyDistributionofMaximumResultantBendingMomentatWeldSeamonQuencherArm-IntermittentCondensation8-158FrequencyDistributionofMaximumResultantBendingMomentsatBottomSupport-IntermittentCondensation-0.5mbelowQuencherCenter8-1598-160PowerSpectralDensitiesTestll.1-P5.5PowerSpectralDensitiesTestll1-P528-161PowerSpectralDensitie'sTest4.1.6-P5.58-1628-1638-164PowerSpectralDensitiesTest4.1.6-P5.2PowerSpectralDensitiesTest20.R1.10-P55PowerSpectralDensitiesTest20-R1.10-P5.2REV13/798-18 8-165MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValues-Overpressures8-166MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValuesConsidering.LocalEffects8-167MaximumSpecifiedVerticalPressureProfileandMeasuredMaximumValues-UnderPressures8-168KarlsteinTests-ComparisonofMeasuredandCalculatedBubbleFrequency-0$Humidity8-169KarlsteinTests-ComparisonofMeasuredandCalculatedBubbleFrequency-100%Humidity8-170GKMTests-ComparisonofMeasuredandCalculatedBubbleFrequency8-171GKMTests-ComparisonofMeasuredandCalculatedBubbleFrequency-Overpressure8-172KKBIn-PlantTests-ComparisonofMeasuredandCalculatedBubbleFrequency8-1738-174SSESCalculatedBubbleFrequenciesMultipliersforConversionofBubbleFrequenciestheKarlsteinTesttoSSES8-175OverpressureMultiplierforConversionofBubbleFrequencies8-176NormalizedAmplitudeSpectrumVersusBubbleFrequency-KarlsteinTests8-177KarlsteinModelTests-InfluenceofMaterSurfaceonPressureAmplitude8-178GKMTests-InfluenceofOverpressureonBubblePressure8-1798-1808-1818-1828-1838-1848-185PSDofKarlsteinTests-11.1and12.1-P5.10PSDofKarlsteinTests4.1.1and4.1.6-PS.10PSDofKarlsteinTests21.1and21.2-P5.10PSDofTest20.R110-P5.4PSD'sofTest11.1-P5.2,5.4and5.10PSDComparison-Test20Rl1andDesignSpecificationPDSComparisonTest4.1.1andDesignSpecificationREVlg3/798-19 8-1868-1878-1888-1898-1908-1918-1928-1938-194PSDComparisonTest20.R1.10andDesignSpecificationPSDComparisonTest211andDesignSpecificationPSDComparisonTests21.2and25.R2andDesignSpecificationPSDComparisonTest0.1.6andDesignSpecificationIBADrywellandWetvellPressureHistoryPSDComparisonTest11.1andDesignSpecificationTypicalCrossSectionofSSESSuppressionPoolRevisedQuencherArrangementVelocityofRotatingPoolforOneActuatingValveinOuterRow8-1958-1968-1978-1988-1998-2008-2018-202WaterMotionoftheAcceleratedPoolTestStandforMeasuringThrustMeasuredTemperatureDistributionintheKKBSuppressionPoolResultantBendingMomentonDummyVentVersusReactorPressureResultantBendingMomentonDummyVentVersusClearingPressure0~ResultantBendingMomentonDummyVentVersusPressureAmplitudeatP5.7'JSpecifiedPressureDistributiononDummyVentTypicalVisicorderTraceforBendingMomentonDummyVentREVli3/798-20 PROPRIETARYNumberSECTION8TABLESTitle818.28.3TypicalOperatingInstrumentationTypicalVentClearingTestInstrumentationTypicalCondensationTestInstrumentation8.4ParametersatTestStart-LongPipeVentClearingTestSeries85ParametersatTestStart-ShortPipeVentClearingTestSeries8.6ParametersatTestStart-CondensationTestSeries8.7BehavioroftheSRVandSystemPressures-LongPipeVentClearingTestSeries8.8BehavioroftheSRVandSystemPressures-ShortPipeVentClearingTestSeries8.9PeakDynamicPressuresonthePoolBoundaryDuringVentClearing-LongPipeVentClearingTests8.10PeakDynamicPressuresonthePoolBoundaryDuringVentClearing-ShortPipeVentClearingTests8.11Maximum,Strains,MomentsandVerticalLoadsontheQuencherArmsandSupportDuringVentClearing-LongPipeTests8.12MaximumStrains,MomentsandVerticalLoadontheQuencherArmsandSupportDuringVentClearing-ShortPipeTests8.13SystemPressuresandPoolWaterTemperaturesoftheCondensationTests810PeakDynamicPressuresAmplitudesDuringtheDifferentCondensationPhases8.15StatisticalCharacteristicsoftheBottomDynamicPressures(P5.2)8.16StatisticalCharacteristicsoftheWallDynamicPressures(P5.10)8.178.18REV.1,3/79RepetitionTests-ComparisonofRecordedValvesRepetitionTests-MeanValuesandDeviations8-21 PROPRIETARY80SSE~SUENCHERVERIFICATIONTEST81INTRODUCTION8.l.1Pu~roseoftheTestsTheoptimizedquencherdesignforSSESandtheloadspecificationonthewettedboundariesofthesuppressionpool,onthesubmer'gedstructuresandonthepressurereliefsystem,arebasedonparametricmodelteststudiesandfullscaleinplanttestresultsfromasimilarquencherdesign.TheloadspecificationsfortheSSESquencheraredescribedindetailinSection4.1.Inordertoverifytheseloadspecificationsandfurtherverifythequencher'ssteamcondensingcharacteristics,fullscalesinglecelltestswereconductedattheKraftwerkUnionlaboratoriesinKarlstein,WestGermany.8.1.2TestConceptTheconceptsusedtodesignandperformthetestswere:1)Useofaconservativelydefinedsinglecell2)Theclosesimulationofthemainsafetyreliefvalvesystemparameters81.2.1UnitCellApproach81.21.1SingleCellTheoryForagasbubbleoscillationinafreewaterspace,thewatermasscoupledtothebubbleisalternatelyacceleratedanddeceleratedDuringthisprocesstheoverpressureandunderpressureamplitudesdecreasewithincreasingdistancefromthebubble.Whenasolidwallisplacedneartheoscillatingbubble,thewateraccelerationisrestrictedinthedirectionofthewallandthedecreaseinpressureamplitudeinthedirectionofthewallisless.Thiseffectcanbeexpressedmathematicallybyreplacingthe.bubblebyapotentialsourceandaccountingforthewallbythemethodofimages.Theeffectsoftherealsourceandtheimagesourceareaddedforeachpointoftheflowfield.Forthecaseinwhichabubbleisenclosedinanarrowwaterspace,closelysurroundedbysolidwallsandasolidbottomwithafreewatersurfaceatthetop,thewaterspacebelowthebubbleisforallpracticalpurposesunmoved.Onlythewatervolumeabovethebubbleisfreetooscillate..Consequently,thepressuregradientinthelowerwaterspaceisnearlyzero,whilethepressureamplitudeabovethebubbledecreaseswithincreasingproximitytothewatersurface.Thepressureamplitudesarezeroatthewatersurfaceandthemethodofimagesapplies.REV.1,3/798-22 PROPRIETARYAnalytically,thecaseinwhichaplanarfieldofuniformstrengthbubblesarealloscillatinginphaseisthesameasthecaseinwhichsolidwallsexistbetweeneachoftheindividualbubbles.ThesinglecelltestconfigurationusedatKarlsteinsimulatesthisextremelyconservativecaseofparallelbubblesoscillatinginphasewiththesamesourcestrength.Adescriptionoftheequivalenceofthesinglecellconfigurations,usingthemethod-ofimages,iscontainedinFigures8.1and8.2.ForamoredetailedevaluationoftheKarlsteintesttanksinglecell,seeSection8.5.1.81.21.2A~licationofSingleCell~AroachThesubmergenceofthequencherinthetesttankisequaltothehighestvalueintheplant.Astothewatercross-sectionareathesinglecelltheorydescribedaboveisusedFigure8.3showsageometricalpartitionofwaterspace.Thewatercross-sectionareasrelatedtothedifferentquenchersarelistedbelow:QuencherAQuencherBQuencherCQuencherDQuencherEQuencherFQuencherGQuencherHQuencherJQuencherKQuencherLQuencherQuencherNQuencherPQuencherRQuencherSAverageWaterSurface3147mz(33862ftz)3147mz31.47mz3147mz31475231~47AI231.47mz31.47mz31.47mz3147mz3147mz3147mz31.47mz31.47mz31.47mz31.47mzRelatedWaterSurface21.4mz(230-26ftz)214mz31.3mz{336.79ftz)42mz(45192ftz)31~3mz31'mz42mz31m3mz31~3mz42mz31~3mz31a3mz214mz21.4mz31~3mz42mzREV1,3/798-23 PROPRIETARYThesmallestwatersurface(approximately21.4m~)issimulatedinthetests.Therefore,thedynamicpressureamplitudesatthewallsandthebottomaremeasuredunderconservativeboundaryconditions.8.1.22SimulationofSSESParametersThefo'llowingsectionprovidesadescriptionofthoseparametersthatweresimulatedintheKarlsteintestfacilityTheseparametersaretypicalofmostMKIIplants.FormoredetailonthetestfacilityseeSection8.28l.22.1Prima~rSgsternPressureThereactoroperatingpressureforSSESisapproximately1000psig(69bar)whilethehighestpressuresetpointforanySSESSafetyReliefvalveis1205psig(83bar),whichisclosetothehighestprimarypressurethatcanbesimualtedintheKarlsteintestfacility(82bar).Thisallowedthetestsimulationtoverycloselymatchtherangeofinitialprimarysystempressuresthat.canbeexpectedintheoperatingplant.8.1222Safet~ReliefVal'vegSRVJInordertomatchthecharacteristicsoftheSafetyReliefValve,anoriginalCrosbySRV,shippeddirectlyfromtheplantsite,wasinstalledintheteststandandusedinalltests.81223DischargeJ.ineInordertocovertherangeofdischarge'linelengthsandthereforeairvolumesthatexistinSSES,twoventclearingtestserieswererun;onewithadischargelinethatsimulatesthelongestSSESdischargeline{48m)andonethatsimulatestheshortestSSESdischargeline(35m).Inaddition,thenumberofbendsineachline,thei'nnerdiameterofthemainpartofthe'ine(303.9mm),andtheinnerdiameterofthelastverticalruntothequencher(2889mm)arecloselysimulatedtothatwhichexistsintheSSESplant.(schedule40pipeandschedule80pipe,respectively).Inadditiona24ft.submergence,correspondingtothehighestwaterlevelinthesuppressionpool,wasusedforalltests8.12.24VacuumBreakersInordertocloselysimulatetheeffectsofvacuumbreakeroperationonthetests,twosix-inchdiameterCrosbyvacuumbreakerswereshippedtoGermanyandinstalledintheteststandatthesamerelativelocationasplannedfortheSSFSplant.REVli3/798-24 PROPRIETARY81.22.5OuencherAfullsizeprototypeofthequencherinstalledintheSSESplantwasinstalledinthetestfacilityandusedforalltests.Figure8.13showsthequencherwithinstrumentationforventclearingtestswhilefigure8.14showsthequencherwithinstrumentationforthecondensationtests.8-2TESTFACILITYANDINSTRUHENTATION8.21TestFacility8.2.1.1MechanicalSet-upThetestconfigurationasconstructedistypicallyillustrateddiagrammaticallyinFigure8.4.Theteststandconfigurationcanbedividedinto:thesteamboiler,thesteamaccumulator,thesteamlinebeforetheSRVandthebuffertank,theSRV,thedischargelinebetweentheSRVandthewaterpoolwiththequencheraspipetermination,andthelargetankaswaterpool.8.2111SteamboilerThesteamboilerisanoil-fired,once-through,forced-flowboilerwithanoutputofapproximately20HWatamaximumsteampressureof170bar(2499psig)andamaximumsteamtemperatureof520~C(968~F).The.boilerisdesignedforaclosedoperatingmodeinnormaloperation.Afractionoftheboiler'soutputisrecoveredfromthecondensateviathehigh-pressurecooler.Whenthereisanopenloop{ie.,lostcondensate),theoutputisreduced.Thesteamflowavailableinthismodeisapproximately8to9kg/s(17.6to19.8ibm/s).Thelostcondensateresultsinatimelimitationoncontinuousoutput.Thefeedwatersupplyoftheboilerisabout20m3(705ft~).Oncethatamountisusedup,furthersteamsupplyascontinuousoutputispossibleonlyuptotheoutputofthefeedwaterconditioningsystem.Thatamountsto5m~/h(176ft3/h).Forlongertestperiodsitisnecessarytointerruptoperationfor4hoursinordertorefillthefeedwaterstoragetank.82.1.1.2SteamAccumulatorAsdescribedin8.2.1.1.1theamountsofsteamsuppliedcontinuouslybytheboileraretoosmalltotestanSRV.REV.1,3/798-25 PROPRIETARYToprovideawaytotestvalvesatflowratesofuptoapproximately22kg/s(484ibm/s),avalvetestfacilitywasbuiltusingtheboilerplantandapressurevesselconnectedtoit.Thisvesselischargedwithasteam/watermixturebytheboilerandisusedasasteamaccumulator.Fromthissteamaccumulator,highersteamflowratescanbedeliveredforashortperiodoftimeThedimensionsofthepressurevesselare1.5mdiameterand12mhigh,whichresultsinanaccumulatorvolumeofapproximately22m3.Adaptedtotherequiredsteamoutput,theaccumulatorisfilledwithsaturatedwaterandsaturatedsteamatthespecifiedratio.Thesteamisdrawndownwardthroughastandpipe.Thehighsteam,flowtobeextractedtransientlyfromtheaccumulatorresultsinarapiddecreaseofpressureandtemperature.Forstrengthreasons,thetemperaturedifferencebetweentheinsideandoutsideoftheaccumulatorvesselmustnotexceedacertainvalue.Thislimitsthemaximumpressuredropandthustheavailabletesttime.82.1.1.3SteamLineandBufferTankTheconnectionbetweenthesteamaccumulatorandthevalveteststandconsistsofanND250pipeline.ThislinecontainsisolatingdevicesforemergencyisolationandameasurementsectionconstructedasaVenturinozzle.Theexistingeguipmeatprovideforadirecthorizontalconnectionofthevalvebeingstudied.ThiscorrespondstothedesignoftheSRVsusedinGermanBMRplantsandtotheirarrangementattheendofataplinecomingfromthemainsteamline.ThesteamsupplylinewasrebuilttomatchthedesignfeaturesofSSES.ThepreviouslydescribedpipelinenowendsinaT-piece.XnordertosimulatetheSSESmainsteamlineandtokeepthesteamsupplyflowtothevalveasuniformaspossible,abuffertankhavingavolumeof5.2m3wasconnectedtothesecondhorizontaloutletoftheT-piece.Theverticaloutletoftheabove-describedT-pieceleadstothevalve.8.2.114Safet~ReliefValveQSRVQTheSRVusedinthetestsistheactualversionbeingusedforSSES.Thesevalvesarearrangedvertically,haveasteaminletfrombelowandanoutlettotheside.Asdescribedin8.211.3,thesteamsupplylinewasrebuiltinsuchawaythatthesamearrangementwaspossibleintheteststand.ThevalvewasmountedontheT-piece,usingthesameconnectiondimensionsasintheactualplants.REV1,3/798-26 PROPRIETARYOperationofthevalveduringthetestsrequirestheconnectionofpowersupplylines,controllinesandmeasurementlines.Theexistingequipmentatthevalvetestfacilitywasusedtosatisfymostofthoserequirements.Somemodificationsbecamenecessaryinordertoadapttotheconstructionofthevalve.TheSRVsinGermanBMRplantsareoperatedbyanelectricallyactuatedpilotvalvewithitsownoperatingmedium.Incontrast,theSS'>>Svalveusedinthetestwasopenedpneumatically.Accordingly,thecompressed-airconnectionwasrebuiltsothattheopeningconditionsintheactualplantcouldbesimulatedintheteststand.8.2~11.5DischargeLineandQuencherTheSRVdescribed.in8.2.1.1.4dischargesontheexhaust-steamsideintoapipewhichrepresentstheSRVdischargeline.ThelengthoftheSRVdischargelineandthenumberofbendsaredifferentforthe16SRV'sforSSES.Twolinelengthswereusedforthetests,correspondingtothelongestandshortestlengthsoftheSRVdischargelinesintheplant.IsometricdrawingsofthetwodischargelinesareshowninFigure8.5(longline)andFigure8.6(shortline).Pipesupportsandvibrationdampersweremountedattherequiredplaces.Theseplaceswerenotidenticaltothecorrespondingonesintheplant,becausethemountingsituationsandespeciallytheconcreteconstructionoftheplantcannotbe-simulateddirectlyinthetestfacility.Topreventthebuildupofalargeunderpressureinthepipe,twoactualvacuumbreakerswereinstalledinaverticalpartofthepipeline,asintheplant.ThequencherformstheterminationoftheSRVdischargeline(seeFigure8ll).Thesteamisconductedintothewaterthroughalargenumberofholeshavingadiameterof10mm.ThedesignofthequencherisdescribedindetailinSection4.1.AbottomsupportisprovidedtoholdthequencherinplaceinthetesttankItconnectsthequencherrigidlytothebottomofthetankandisconstructedinsuchawayastomakeitpossibletomeasuretheloadsexertedonthequencherduetoventclearingprocessesandsteamcondensation.Theslidingjointprovidedbetweenthequencherandthedischargelineintheplantissimulatedintheteststandhgdraulicallgbyacorrespondingannulargap8211.6TestTankForSSES,theexhauststeamfromthereliefvalvesisconductedintothesuppressionpoolandiscondensedthere.Xnthetestfacility,asectionofthatpoolissimulatedbyastiffenedREV-lg3/798-27 PROPRIETARYsteeltank(seeFigures8.7,88,89).Intheplant,thesuppressionpoolcanbesubdi'videdconceptuallyintosuhspaces,eachofwhichisassociatedwithasteamsupplyline(seeFigure8.3).Inordertoadapttheconditionsinthetesttanktothedimensionsofthesmallestgeometricalsinglesell,concreteshapedblockswereinsertedintothetesttank.TheconcreteshapedblocksareclearlyillustratedinFigure8.7.Theexposedcross-sectionalareaofthewaterspaceis7.2mx3.15m=22.7m~.Itcorrespondsconservativelytothesmallestindividualcellintheplant.Illuminatingdevicesandviewingportsmadepossiblethedirectobservationandalsophotographicrecordingoftheunderwaterprocesses.8.2.2InstrumentationInstrumentationisprovidedforcontrollingthetestprocedure,determiningtheprescribedmeasurementquantities,andrecordingthem.82.21GeneralDescriptionTheinstrumentationusedintheKarlsteintestfacilityconsistsofoperatinginstrumentationandtestinstrumentation.Operatinginstrumentationassuresthecontrolofthetestfacilityanditsenvironmentcorrelation.ThetestinstrumentationrecordstheloaddatawhichisusedtoverifytheconservatisminthedesignloadsasspecifiedfortheSSESinsection4.1ofthisDesignAssessmentReport.DetailsontheoperatinginstrumentationaregiveninSection8.2.2.3.AdetaileddescriptionofthetestinstrumentationcanbefoundinSection8.2.248.2.22InstrumentationIdentificationForidentification,themeasuringsensorsaredesignatedaccordingtoasystemoflettersandfigures.Thefirstoneortwocharactersareletterswhichidentifythetypeofinstrument:PTFLDGSGILPPressureTransducerTemperatureSensor(Thermocouple)FlowRateMeasurementsRaterLevelMeasurementsDisplacementGageStrainGageElectricalImpulseSignalLevelProbeREV.l,3/798-28 PROPRIETARYTheselettersarefollowedbyanumberwhichcharacterizesthelocationwithinthetestfacilitywheretheinstrumentissituated.Thefacilitywasdividedintosectionsasfollows:Section1containsthesteamsupply,includingtheaccumulator{onlytransducersoftheteststandinstrumentationsystemarecontainedinthissection).Section2containsthesteamlineuptothesafetyreliefvalveandincludesthebuffertank.Section3containsthesafetyreliefvalve.Section4containsthedischargelineandquencher.Section5containsthetesttank.Thesensordesignationiscompletedbyaddingadecimalpointandasequentialnumber.Forexample,"P5.6"means:thenumber6pressuretransducerinthetesttank.Additionalabbreviationsusedareasfollows:DPSCTCDCACFAHT-SGSRVPGRTDDataProcessingSystemCoatedThermocoupleDirectCurrentAmplifierCarrierFrequencyAmplifierHighTemperatureStrainGageSafetyReliefValvePressureGageResistorTemperatureDetector82.2.3Operating1nstrumentationTheoperatinginstrumentationisprovidedformeasurementofparametersinrelationtothesteamaccumulator,thesteamlinesandtheSRV'sAtotalof30sensorscanberecordedbyaprocesscomputerwhichispartoftheoperatinginstrumentationsystem.ThedataarestoredonamagneticdiskandcanbeprintedoutTherecordingfrequencyoftheprocesscomputerwasadaptedtoalignwiththeinstrumentationchanriels,coveringarangefrom0.5Hz,forthosesensorswhereonlysmalltransientsaretobeexpected,uptoabout200Hzforthesensorswherehigherfrequencysignalsareexpected(e.g.forpipevibrations)Theoperatinginstrumentationcomprisesthemeasuringdevicesusedtomonitorandcontrolthesystemandalsothedataacquisitiondevicesneededforthatpurpose.TypicalmeasuringlocationsforthetestsareillustratedinFigure8.4andlistedinTable8.1.BEV1,3/798-29 PROPRIETARYAccordingtothetypeofacquisitionanddisplay,themeasurementsensorscanbeclassifiedintotwogroups="DisplayonControlConsole"and"AcquisitionbyComputer".82.2.3.1Disp~la'nControlConsoleToenabletheoperatingpersonneltocontrolthetestequipment,anumberofquantitieswhichcharacterizetheoperatingconditionofthesystemaredisplayedcontinuously.Inparticular,theyare:Waterlevelin:Steamaccumulator,steamline,buffertank,dischargeline,testtankPressureinSteamaccumulator,buffertank,controlline,dischargelineTemperaturein:Steamaccumulator,buffertank,dischargeline,testtanka822.3.2A~cuisition~bComputerMostofthedatasensorscomprisingtheoperatinginstrumentationareinterrogatedbyacomputeratprescribedtimeintervalsbefore,duringandafterthete.t.Thevaluesarestoredonadisk.Thedataareprintedoutatprogrammedintervals.Ataninterrogaticnfrequencyof200Hz,thecapacityofthestoragedeviceissufficientforarecordingtimeof2minutes.Thefollowingmeasurementvaluesareinterrogated:WaterlevelSteamaccumulator,buffertankdischargeline,testtankPressureSteamaccumulator,buffertank,steamline,controlline,dischargelineTemperatureBeforeSRV,afterSRV,surfaceofSRV,dischargeline,testtankVibrationsValvetravelSwitchingtimeSteamlinebeforeSRV,dischargelineSRV,vacuumbreakersElectricalenergizationofSRVREV1g3/798-30 PROPRIETARY822.4Test,InstrumentationMesurementvaluesusedtoverifythetesttasksaredeterminedbythetestinstrumentation.Itisnecessarytoincludehereafewtypicalmeasuringpointsthatarealreadyusedformonitoringpurposesintheoperatinginstrumentationonthepipes.andSRV.Sincemostoftheseprocessesareofahigh-frequencynature,thedataisacquiredinanalogformbymeansofcarrier-frequencymeasuringamplifiersanddcamplifiersonanalogmagnetictape,andtoalargeextentalsoonvisicorders.Thevisicordertracesallowaninitialreviewandapre-evaluationofthetestdata.82.24.1MeasuringPointsMeasurementsaremadeofthepressureonthesteamlinebeforetheSRV;valveactuationandvalvetravel;pressurevariationinthedischargelineatfourpointsbetweentheSRVandquencher;temperatureinthedischargelineatthreepointsbetweentheSRVandquencher;waterlevelinthedischargelinebeforethequencherinletatfourpositionsforthelonglineandfivepositionsfortheshortline;bending,axialandtorsionalstrainsonthebottomsupport;bendingstrainsonthequencher;bendingstrainonadummyventpipe;temperaturedistributioninthetesttank;temperaturedistributionatthequencherforthecondensationtest;wallpressuresandbottompressuresinthetesttank.TypicalmeasurementpointsfortheventclearingtestsareillustratedinFigures8.7,8.8,8.9andlistedinTable8.2.TypicalmeasurementpointsforthecondensationtestsareillustratedinFigures8.10,8.11,8.12andlistedinTable8.3.8.2.24.2Set~uofMeasuringInstrumentsAllinstrumentationischannelledtoonecentralstationsituatedinthecontrolroomofthelaboratory.Eachinstrumentationchannelconsistsoftheindividualsensor,connectingcable,amplifier(carrierfrequencyamplifierordirectcurrentamplifier),attenuator;andarerecordedonmagnetictapesandvisicorders,mostchannelsbeinginparallelonbothsystems.Threemagnetictaperecordersandthreevisicorderswereusedinthecontrolroom.Eachunitallowstherecordingof12channelsand,inaddition,atimereferencesignalandaphysicalcorrelationtrace.REV1,3/798-31 PROPRIETARYThesensorsareconnectedbyshieldedcabletotheamplifiersvhicharelocatedneartherecordersinthecontrolroom.Forthestraingages,displacementgagesandpressuretransducers,carrierfrequencyamplifiersvereusedwhichallowafrequencyresolutionofupto1KHz.Fortemperaturemeasurements,directcurrentamplifiers(10Hz)vereusedtogethervitha10Hzlovpassfilter.82.25Visual-RecordingThreehigh-speedcamerasvereusedtofilmtheprocessesinthepoolduringtheblowdovnthroughthequencher.KMUusesa"HYCAM120m~'orthatpurpose.TvoLOCAMcameras(model51-0003)werebeingmadeavailablebytheStandfordResearchInstitute(SRI)Thepositioningofthecameraswasasfollovs:HYCAMcamerainfrontofonebull'seyeatquencherheight;LOCAMcamera1infrontofonebull'seyeatatankheightofapproximately4m;LOCAMcamera2ontheserviceplatformabovethetankataheightofapproximately9m.AcorrelationbetweenthemovingpicturesandthedatarecordingsontheVisicorderandmagnetictapevasaccomplishedbymeansofatimingmarkonthefi'lms.83TESTPARAMETERSANDMATRIX8.31VentClearingTestsThetestmatrixfortheventclearingtestsispresentedinFigure8.15.Thisfigureshowsthetestnumberandparameterconditionsusedforeachtest..Thenumberofbasictestswas25.These25testsweresplitinto5groupsoftestswherebyeachgroupcoveredasetoftestparameters.Testsnumbered26to32wereadditionaltestsvhichwerenotrequiredtoverifythequencherdesignbutwhichcouldproveusefulinevaluatingtheperformanceofthesafetyreliefsystem.Testsnumber27,28,30and31weretoinvestigateshorterthannormalSRVopeningtimes,but,asvalveopeningtimesverefoundtobequitefast,thesetestswerenotaddedtotherequiredtests.Testsnumber26and32,withonelockedvacuumbreaker,wereincludedintothetestmatrix.Theresultsshovedtheeffectofthelockedvacuumbreakertobeminimalsotestnumber29wasnotadded.REV1,3/798-32 PROPRIETARYTheallocationofeachtestgroupwithintheoperationrangeofthesafetyreliefsystemisshowninFigures8.16to8.21bytestpoints.BaseparametersinGroup1(Figure816)arelongdischargeline'length,normaldischargelineairtemperature,normalinitialwaterlevelinsidethedischargelineandnormalvalveopeningtime.EachofthefollowinggroupsvaryoneormoreoftheseGroup1baseparameters;Group2(Figure817)usesalowinitialwaterlevelinsidetheSRVpipe;Group3(Figure8.18)usesahighdischargelinetemperature;Group4(Figure8.19)usesashortdischargelinelengthandGroup5(Figure8.20)usesashortdischargelinelengthandahighdischargelinetemperature.Eachofthebasic25testswascomprisedoftwoormorevalveactuationswherebyonlythefirstactuationismadeatt,hespecifiedconditionsofthedischargeline(so-calledcleancondition).Anyotheractuationwasmadeattheprevailingdischargelinetemperatureandwaterlevel(so-calledRealCondition).Inthecaseofonlytwoactuationsatatestpointthetimeintervalbetweentheactuationswasapproximately10minutes.Inthecaseofmultipleactuationsatatestpointthetimeintervalsbetweenactuationswerevariedasfollows:Fortestpoints4,5,14,15thetimebetweensuccessiveactuationswasl.5/5/15/30/60/120seconds,accountingforsevenvalveactuations.Fortestpoints19and20thetimebetweensuccessiveactuationswas15/5/15/30/60/120/5/15/600seconds,accountingfortenvalveactuations.ForventclearingtestswithonlytwoSRVactuations,thehold-opentimefortheSRVwas2secondswhileforthemultiplevalueactuationteststhehold-opentimewas1.5seconds.'Fivetestpointswererepeated,theseweretestpoints4,15,19'0and25.RepeattestsatadesignatedtestpointareindicatedwithaletterRinthetestnumberi.e.Testnumber20.Rl.listhefirstvalueactuationoftherepeattestattestpoint20.11.AcompilationofactualparametersatthestartofeachtestistabulatedinTable8.4forthelongpipetestseriesandTable85fortheshortpipetestseries.8.32CondensationTestsInordertofurtherverifythesteamcondensationcapabilitiesofthequencherdeviceandprovidespecificinformationregardingitssteamcondensationcapabilitiesforthesafetyreliefsystemlREV.1,3/798-33 PROPHIETARYoperationrangeaseriesofeightextendedblowdowntestswereperformed.Thesetestsaredesignatedastestnumbers33to40.Eachtestwasperformedwiththeshortdischargelineconfigurationasdescribedinsection8.2.1.1.5andwithaninitialdischargelinetemperatureofapproximately90~C.ThelocationoftheinitialsystemconditionsforeachtestpointisplottedonthesafetyreliefsystemoperationrangeinFigure822InordertoinitiateeachtesttheSRVwasactuatedaswasdoneintheventclearingtests.Thevalvethenremainedopen-untilthesystempressurereachedthepredesignatedvalueforthattest.Atthistimethevalvewasclosedandthetestwascompleted.Thetotalallowablepressuredropintheaccumulatortankforeachinitialsystempressuredictatedthedurationofeachbiowdown.AcompilationofactualparametersatthestartofeachtestpointinthecondensationtestsmatrixistabulatedinTable8.6.84TESTRESULTSThissectionprovidesacompilationofthetestresultsfortheventclearingandsteamcondensationtestsconductedattheKraftwerkUnionlaboratoriesinKarlstein,WestGermanyinordertoverifytheloadspecificationandsteamcondensingcharacteristicsofthequencherdesignfortheSusquehannaSteamElectricStation.Includedinthissectionisinformationabouttheboundaryconditionsatthebeginningofeachtest,the'esultsofthebehavioroftheSRV,primarysystempressures,dynamicpressureloadsonthepoolboundariesandtheirprimaryfreguencyandtheloadsonthequencherandbottomsupportThisinformationisprovidedintheformoftables,figuresandactualvisicorderrecordings.841VentCleari~nTestResultsNineteentestswithatotalof67ventclearingprocesseswereperformedwiththelongdischargelineintheperiodfromMay8,1978toJune7,1978and13testswithatotalof58ventclearingprocesseswereperformedwiththeshortdischargelineintheperiodfromJune27,1978toJuly7,1978.841.1TestParametersThemostimportantoftheparametersbeinginvestigatedwasdescribedinSection8.3.AdetailedlistoftestparametersforeachvalveactuationisgivenforthelongdischargelinetestsinTable8.4andfortheshortdischargelinetestsinTable8.5.ThisincludesREV.1,3/798-34 PROPRIETARYtypeoftestlengthofdischargelineaccumulatorpressurewatertemperatureinthetesttankwaterlevelindischargelineairtemperatureindischargelineTheaccumulatorpressureP1.1AandthebuffertankpressureP2.6Aarethedeterminativevaluesforthesystempressureatthestartofeachtest.Thevalueswerereadbycomputerjustpriortothestartofthetest.Inadditionthesepressureswerestoredcontinuouslyonmagnetictape.IfalongperiodpassedbetweenthelastcomputerreadingandtheactualteststartthentheinitialvaluesfortheaccumulatorpressureweretakenfromthecorrespondingcomputerplotsTheinitialaccumulatorpressureswerealsoreadfromthoseplotsforthemultiplevalveactuationtests.Foraccumulatorpressuresbelow30bar(435psi),measuringpointP2.5wasusedtodeterminethesystempressure,sincemeasuringpointsPl.1AandP2.6Awereoutsidethemeasuringrange.Thewatertemperatureatthestartofthetestwastakeneitherfromthecomputerlistingsor,inthemultiplevalveactuationtests,fromthecomputerplotsDuetotheinertiaoftheBartoncell,themeasurementvalueforwaterlevelinthedischargeline(measuringpointL4.1)inthemultipleactuationtests,especiallyforthe2nd,3rdandifapplicable,the8thactuation,mustbedisregardedorconsideredonlyasanindicativevalue.Thetemperatureinthedischargelineatthestartofeachtestwastakenfromthecomputerlistingsorthecomputerplotsforthemultipleactuationtests841.2BehavioroftheSRVandSystemPressuresToevaluatethevalvebehavior,thevalveopeningtime,t,wasdeterminedfromtherecordedvalveliftvariationforalltests.0Thisinvolvesthetimefromthebeginningofvalveopeninguntilattainmentofthesteadystatelift(seesketchbelow).Theseopeningtimesarelisted,forthelongdischargelinetests,inTable8.7and,fortheshortdischargelinetests,inTable8.8.Theassociatedsteadystateliftsarealsoindicated.AplotofthemeasuredvalveopeningtimesasafunctionofaccumulatorpressureatthestartofeachtestisshowninFigure8.23forthelongdischargelinetestsandFigure8.24fortheshortdischargelinetests.Theso-calledventclearingtimestpzarealsogiveninTables8.7and88ThisisthetimefromthebeginningofvalveREV.1,3/798-.35 PROPRIETARYopeninguntiltheinstant'ofmaximumpressureatmeasuringpointP4.4inthedischargeline.(seesketchbelow)tsvalveliftventclearingpressurepressurebeforequencher1'wovaluesareindicatedinTables8.7and8.8forsystempressuresmeasuredin:buffertank-P2.6beforetheSRV-P2.5inthedischargeline-P4.1toP4.4Thesetwovaluesarethepressureattheventclearingtime(ventclearingpressure)andthepressureapproximately1.5secondsafterthestartoftest(steadypressure)Theinitialparametersofrelevancefortheclassificationof-testsareindicatedintherowheadings.Theventclearingpressureinthedischargelinebeforethequencherinlet(measuringpointP4.4)isplottedversussystempressure(measuringpointP2.6)underCleanConditionsinFigure8.25forthelongdischargelinetestsandinFigure8.26fortheshortdischargelinetests.SeeSection8.5.2.1foradiscussionoftheventclearingpressuresandtheirdependenceonreactorpressure.84.1.3DynamicPressureLoadsonthePoolBoundariesAsreadofftheVisicordertraces,thepeakpositiveandpeaknegativepressureamplitudesduringventclearingformeasuringpointsP5.1-P5.3(bottompressures)andP5.4-P5.10{wallpressures)arecompiledinTable8.9forthelongdischargelinetestsandinTable8.10fortheshortdischargelinetests.Ina'ddition,approximatevaluesforthepredominatefrequencyofthepressureoscillationsareindicated.Thesefrequencieswerereadfromthevisicordertraces.Figures8.27and8.28showthemeasuredpeakpositivepressureamplitudesatthetankbottomdirectlybeneaththequencher(P5.2)andontheconcretewallatthequencher'smid-heightREV1,3/798-36 PROPRIETARY(P5.10)asafunctionofsystempressureforthelongdischargelineandshortdischargelinetests.ThetestpointsplottedareallCleanConditiontestswithcoldwaterinthetesttank{approximately25~C)anddischargelinecold(approximately50~C)(Longdischargelinetests1.1,2.1,3.1,4.11,4.81.1and32.1andshortdischargelinetests16.1,171,J.8.1,19.1.1and19.R1.1)AsacomparisonFigures8.29and8.30representcorrespondingmeasuringpointsfortestsperformedunderRealCondition(Longdischargelinetestsl.2,2.2,3.2,.10.4and32.2andshortdischargelinetets16.2,17.2and18.2).AscanbeseenthepressureamplitudesareslightlyhigherfortheCleanConditiontestsandnosignificantchangewithsystempressureisobserved.Figures8.31and8.32showthemeasuredpeakpositivepressureamplitudesatmeasuringpointsP5.2andP5.10forCleanConditiontestswithheatedwater{45C-80~COinthetesttankforthelongdischargelinetestsandshortdischargelinetestsrespectively.(Longdischargelinetests5.1.1,6.1,71,8.1,9.1,151.1and15.R1.1andshortdischargelinetests20.1.1,20.Rl.l,22.1,23.1,24.1).Again,asacomparison,Figures8.33and8.34representcorrespondingmeasuringpcintsfortestsperformedunderRealConditions(Longdischargelinetests6.2,7.2,8.2,92,11.2and12.2andshortdischargelinetests20.R1.7,22.2,23.2and24.2)Incontrasttothetestswithcoldwaterinthetesttank,thepressureamplitudesareslightlyhigherfortheRealConditiontests,butaswiththecoldwatertests,nosignificantchangewithsystempressureisobserved.Figures8.35to8.40showthemeasuredpeakpositivepressureamplitudesatmeasuringpointsP5.2andP5.10foranumberofmultiplevalveactuationtestsplottedagainstthecorrespondingvalveactuation.Figures8.41to865showthefirstsecondofvisicorderpressurestraces(forthepoolboundarypressures,P5.1-P5.10)fromvarioustests.8~414LoadsOnThequencherandBottomSupportThebendingstrainsonthetwoarmsofthequencherandatthebottomsupportwereeachmeasuredintwomutuallyperpendiculardirections.Theresultantbendingstrainsandbendingmomentswerecalculatedfromtheseindividualvalues.Thestrain-versus-timevariationsstoredonmagnetictapewerereadforthemaximumresultantduringventclearing.Ahigh-passfilterhavingacutofffrequencyof2HzwasinsertedinordertoruleoutanyfalsificationoftheevaluationduetoslowdriftingofthezeropointTheupperfrequencylimitwasat400Hzduetothemechanica1conditions.REV1~3/798-37 PROPRIETARYThemaximumresultantbendingstrainsdeterminedinthismannerandthebendingmomentscalculatedfromthemarecompiledinTables8.11and8.12forthelongandshortdischargelinetestsrespectively.Toclarifythedirectiondistributionoftheresultingbendingmomentsonthequencherarms,thecomponentsofthemaximumresultantbendingmomentsaredepictedinpolarcoordinatesinFigures8.66and8.67forthelongdischargelinetestsandFigures8.68and8.69fortheshortdischargelinetests.AsshowntheresultantbendingmomentsonthequencherarmsoccurprincipallyintheverticaldirectionFigures870and8.71forthelongandshortdischargelinetestsshowacorrespondingdistributionofthemaximumresultantbendingmomentsatthebottomsupport.Tables8.11and8.12alsoindicatethemaximumtorsionalstrainsandtorsionalmomentsmeasuredatthebottomsupportandthemaximumverticalstrainsandverticalforcesmeasuredatthebottomsupportduringventclearing.Thisdataisbasedonasevaluationofthevisicordertraces.842SteamCondensationTestResultsEightcondensationtestswiththeshortdischargelinewereperformedintheperiodfromJuly18,1978toJuly.21,1978.8.4.21TestParametersThemostimportantoftheparametersbeing-investigatedwasdescribedinSection8.3.AdetailedlistoftestparametersisgiveninTable8.6.CompiledinthatTablearetheparametersatthebeginningofthetests,suchas:typeoftestlengthofdischargelineaccumulatorpressurewatertemperatureintesttankwaterlevelindischargelinewaterlevelintesttankairtemperatureindischargelineTheaccumulatorpressurePl.lAandbuffertankpressureP2.6Aarethedeterminativevaluesforthesystempressureatthestartofeachtest.Thevalueswerereadbycomputerjustpriortothestartofthetest.Inaddition,thesepressureswerestoredcontinuouslyontapebutonlyupto360secondsafterthestartoftests36.1and40.1.Thiswasdictatedbythelimitedstoragecapacityoftheoperatinginstrumentationcomputer'smagneticdisk.Thisdatawascontinuouslystoredonthevisicordertracesandthetestinstrumentationmagnetictapes.REV1i3/798-38 PROPRIETARYForaccumulatorpressuresbelow30bar(435psi),measuringpointP2.5was,usedtodeterminethesystempressure,sincemeasuringpointsPl.1AandP2.6Awereoutsidethemeasuringrange.ThewatertemperatureatthestartofatestwastakenfromthecomputerlistingsandattheendofatestfromthecomputerplotsThevaluesforthewaterlevelsandairtemperaturesinthedischargelineatthestartofatestweretakenfromthecomputerlistings.Table8.13showstherelationbetweentheteststep,testnumber,andrangesofpressureandwatertemperatureastheyactuallyoccurred.842.2PresentationofTestResultsFirstwewillpresentasurveyoftheobservedcondensationphases.Thatisfollowedbyapresentationofthedynamicpressureamplitudesinthewaterregionofthetesttank.Finallythetemperaturevariationsinthewaterregionaredescribed.8.4.221Surve~ofObservedCondensationPhasesIntheoperationfieldofthequencherasgivenbythetestmatrix,theobservedcondensationphasesareindicatedinFigure8.71forblowdownsalongtheupperandlowerboundarylinesoftheoperationfield.8422.1.1BlowdownatlowMaterTe~meratureFortheblowdownalongthelowerboundaryline,thefollowingcondensationphaseswereobservedforthetestedpressurerange:AbsolutesystemPressureinBarCondensationPhaseTests70-25Stationary33.2,34.1,35.1,andinitialsectionof36.125-2IntermittentMiddlesectionof35.12-1Inthepipe(1)Endsectionof36.1(1)Itshouldbenotedherethatatthebeginningofthisphaseaportionofthesteamflowhasemergedthroughtheannulargapabovethequencherinlet.AsnotedinSection8.2.1.1.5,REV1,3/798-39 PROPRIETARYthisannulargapsimulateshydraulicallytheslidingfitofthequencherinstalledatSSES.Figure8.73showsatypicalexampleofthemeasurementtracesobtainedwiththebottomandwallpressuresensorsforstationaryoperationofthequencherintheupperpressurerange(test33.2).Figure874showsatypicalexampleofthelowerpressure'ange(test35.1).High-frequencypressureoscillationsoccurwithverylowamplitude,andwithoutanyfixedfrequency.Toillustratetheintermittentoperation,thevariationofthebottomandwallpressuresandtwopipepressuresthroughouttheentiredurationoftest361isshowninanextremelytime-compressedforminFigure8.75.Theintermittentcondensationphaseisclearlyrecognizableinthemiddlesectionofthetest.Figure8.76showsamoretime-expandedexcerptfromthatphase.Supplementarily,Figure877showsatypicalpowerfulindividualeventinanextremelytime-expandedform.Thehigh-frequencypressurepeakssuperimposedonthelow-frequencysinusoidalpressurepulsationsareclearlydiscernibleinbothFigures875and8.76.Forthephaseofcondensationinthepipe,thetesttracesexhibitnegligiblylowamplitudes,whichareclosetotheresolutionlimitofthemeasuringchain.Therefore,noexampleofsuchatraceisshown.8.4.2212-BlowdownatHighWaterTe~meratureForblowdownalongtheupperboundaryline,thephasesdescribedin8.4.2.2.1.1wereobservedinpracticallythesamepressureranges.However,theappearanceofthepressureoscillationsdifferstosomeextentfromthatofthepressureoscillationsatlowwatertemperature.First,hereistheobservedrelationbetweenpressurerangeandcondensationphase:AbsolutesystemPressureinBarCondensationphaseTests>70-4.Stationary372e38ls39leandinitialsectionof4014-2IntermittentMiddlesectionof40-12-1Inthepipe<>>Endsectionof40.1REV1,3/798-40 PROPRIETARY(1)Itshouldbenotedherethatatthebeginningofthisphaseaportionofthesteamflowhasemergedthroughtheannulargapabovethequencherinlet.AsnotedinSection8.2.1.1.5,thisannulargapsimulateshydraulicallytheslidingfitofthequencherinstalledatSSES.Forstationaryoperationintheupperrangeofpressure,Figure8.78showsatypicalexamplefortest37.2.1helowerrangeofpressureforthisphaseisrepresentedbyanexamplefromtest391(Figure8.79).Therearealsohigher-frequencypressureoscillationswithlowandverylowamplitude,respectively,andwithoutanyfixedfrequency.Atypicalexampleofintermitten~toerationisshowninFigure8.80byanexcerptfromtest40.1.Comparedtothisphaseatlowwatertemperature(seeespeciallyFigure8.76),adistinctattenuationofthestrengthofthepressurepulsationsisobservableathighwatertemperature.Superimposedhigh-frequencypressurepeaksdonotoccur.Forthephaseofcondensationintheprie,the.testtracesexhibitnegligiblylowamplitudesevenatextremelyhighwatertemperatureofmorethan90oC.8.4.2.2.2StatisticalEvaluationofthe~DnamicPressureLoadsonthePoolBoundariesAsdescribedinSection8.4.22.1,thesteamcondensationdoesnothaveanyuniformformthroughouttheentirerangeofsystempressureandwatertemperature.Tonowbeabletoquantifythedistributionofdynamicpressureamplitudesduringablowdownfrom70bartoapproximately1bar,therecordingsfromarepresentativebottompressuresensorandwallpressuresensorforallthetestswerestatisticallyevaluated.Thisalsoallowedustoinvestigatetheinfluenceofsystempressureandwatertemperatureonthedynamicpressureamplitudes.B.a2.22.1Dependenceof~DnamicBottomandIlailPressuresonSystemPressureandMaterTemperatureThepressure-timehistoriesstoredonmagnetictapeforpressuresensorsP5.2(bottompressure)andP5.10(wallpressure)wereeachreadformaximumvalueatuniformtimeintervals.Ahigh-passfilterwithafrequencycutoffof2Hzandalow-passfilterwithafrequencycutoffof500Hzwereinsertedintothecircuit.In'hismanner,afalsificationoftheevaluationduetoslowREV1,3/798-41 PROPHIETARYdriftingofthezeropointorduetoelectricalinterferencewaslargelyexcluded.Portests33.2,34.1,351,37.2,38.1and39.1,.auniformintervalof1secondwaschosenbecauseoftherelativelyshorttestdurationofamaximumof64secondsintest39.1.Intests36.1and40.1withtestdurationsofover800seconds,theuniformintervalwas4seconds.Inthesetwotests,thephasesofstationaryandintermittentcondensationandcondensationinthepipewerecoveredseparatelyatthesametime.Noerrorwasintroducedintotheevaluationbythedifferentchoiceofintervals,sincethemaximumvalueswerecoveredineachcaseTheextremevaluesdeterminedforthepositiveandnegativedynamicpressureamplitudesatthebottomandonthewallareplottedversusthetransientvariationofthesystempressureinFigures881and8.82.Duetothelargenumberofextremevalues,aselectionwasmadewiththeaimofconsideringonlythehighervalues.ThetophalfoftheFigureshowsthemeasuredmaximumpressureamplitudesfortheblowdownathigherandhighwatertemperaturealongtheupperboundarylineoftheoperationfield.Thebottomhalfshowsthemfortheblowdownatlowwatertemperaturealongthelowerboundaryline.AsimilarillustrationforthemeasuredmaximumwallpressureamplitudesisgiveninFigure8.82.Thepeakbottom-pressureandwall-pressureloadsmeasuredduringtheindividualcondensationphasesareindicatedasafunctionofwatertemperatureinTable8.14.Promthesepeakvalues,wecanascertainaslightdecreaseofthepressurelevelwithahotpoolforthestationaryandintermittentcondensationphases.Porthephaseofcondensationinthepipe,ofcourse,therearepracticallynodiffe'rencesinthepressurelevelsforcoldandhotpool-84.22.2.2Occurrenc~efreuenceDistributionsofthe~DnanicBottomandMallPressuresInparallelwiththedeterminationofextremevaluesasdescribedinSection8.4.2.2.2.1allpositiveandnegativepeakvaluesbetweenthezeropassagesofthepressure-vs.-timevariationsweredetermined.ineachtimeintervalandclassifiedaccordingtomagnitude.Thiscountingmethod,knownas<<peakcountbetweenzeropassages"or"meancrossingpeakcountmethod>>,avoidstheinclusionandconsequentialoverassessmentofsmallintermediateoscillations.Onlytheabsolutemaximabetweentwozeropassages'reincludedinthecount.REVli3/798-42 PROPRIETARXThecountresultsuppliestheclassoccurrencefrequencydistributionatonce.Positiveandnegativepeakvaluesweretreatedseparately.Anyerrorinthecountresultsbythenoiselevelonthemagnetictapeswaslargelyeliminatedbymeansofa.prescribedamplitudesuppressionof10mV=0.015bar.Auniformclassintervalof0.025barwaschosenforthehistograms.Inthatway,thehistogramsoftheindividualtestswereabletobecombinedintoanoveralldistributionforblowdownswithcoldandhotpool.ThehistogramsofthepositiveandnegativeamplitudesofthedynamicbottompressuresatmeasuringpointP5.2areillustratedinFigures8.83and884forblowdownswithcoldandhotwater,respectively.AnalogoushistorgramsforthewallpressuresatmeasuringpointP5.10areshowninFigures8.85and886.8.4.2.2.2.3StatisticalCharacteristicsoftheDynamicBottomandMallPressuresInfluencesoftestparameterscanbereadofffromthestatisticallydeterminedmeanvalues,sincethosevaluesareobviouslymuchmoretypicalthanthemagnitudesofindividualandveryraremaximumvalues.Themeanvaluesweredeterminedbythegroupvaluemethodsusingthefollowingequation:PkZn.Pi~1KZn<illwherePG=meanvalue;Fifrequency.classmeanvalue;n=classThegroupvaluemethodwasindividualhistogramsofadistributions.Thosemeanto8.86alsousedforthecombiningoftheblowdowntogetthemutualfreqeuncyvaluesareindicatedinFigures8.83Ingeneral,thetrendsaresupportedbythemaximumvalues.Theunavoidablescatterofthemaximumvaluesisallowedforbyformingtheaveragevalueofthe10highestamplitudesineachtest.Duetothesmallnumber,theyweredeterminedbythesingle-valuemethod:wherePE=NZPi~1NREVli3/798-43 PROPRIETARYPE=meanvalue;P.=singleextremevalue;N=numberofextremevaluesTables815and8.16provideanoverviewoftheabovementionedmostimportant.statisticalcharacteristicsofthepressure-timehistoriesatthebottomandatthewall,respectivelyfortests33.2to40.1.Indicatedare:maximumvaluerelativetotheentiretest,meanvaluerelativetotheentiretest,-lowerlimitvalueofthe10highestvalues,meanvalueofthe10highestvalues.*Besidethedataconcerningthesystempressuresandwatertemperatures,thecondensationphasesarealsolisted.Intests36.1and40.1,thephasesofstationaryandintermittentcondensationandcondensationinthepipeweretreatedseparately.'igures8.87and8.88showplotsofthemeanvaluesrelativetotheentiretestortestsectionandthemeanvaluesofthe10highestvalues,asfunctionsofsystempressure.Themeanvaluesofthebottomandwallpressuresareslightlyhigherfortheblowdownwithacoldpool.Thistrend,alreadyalludedtoinSection8.4.2.2.2.1onthebasisoftheabsoluteextremevalues,isthereforeverifiedstatistically.,Thelevelofthemeanvaluesfromthe10highestvaluesishigherbya-factorofapproximately3-4thanthelevelofthemeanvaluesrelativetotheentiretestortestsection.8.4223Te~meratureVariationsintheRaterRegionoftheTestTankFourtestswereselectedtoillustratethetemperaturevariationsinthewaterregionofthetesttank:test33.2forhighsystempressureandcoldpool,test35.1forlowsystempressureandcoldpool,-test37.2forhighsystempressureandhotpool,test39.1forlowsystempressureandhotpool.Figures8.89to892showtheverticaltemperaturedistributionobtainedfromthemeasuringpointsT5.5,T5.2,T5.3andT5.4arrangedaboveoneanotherontheconcretewallIneachcase,themeasuredtemperaturesarescatteredaboutameancurve.ThescatterisgreatestformeasuringpointT52(approximatemax.a8oC).Thatmeasuringpointisattheheightofthequencherarmandisimpingedupondirectlybythesidewardsdirectedflowimpulse.ThescatterisleastformeasuringpointT5.4(approximatemax.a50C).Thescattercanbeexplainedbythehighdegreeofturbulenceinthe.pool.REV~1i3/798-44 PROPRIETARYFigures8.93to8.96showthetemperaturevariationsatquencherarm1forthesametests.AtmeasuringpointT5.8locatedinthemiddleoftheholearray(seefigure8.14)adistincttemperatureincreaseofapproximately15-200C,ontheaverage,wasrecordedrelativetothepooltemperature.Incontrast,thetemperaturesattheupperedgeoftheholearray(T5.9)andattheupperedgeoftheguencherarm(T5.10)aresomewhatlowerthanthepooltemperatureatT5.1duetoasufficient"coldwatersupply".Thisisanindicationofthegoodcirculationofwaterneartheguencher.This.confirmedtheexpectedcondensationbehaviorofthequencherasrelatedtothelayoutoftheholearray.(SeeSection4.1.1.1).84.2.24WaterLevelintheDischargeLineWhenOpeningandAfterClosi~ntheSRVInthetestswiththelongdischargeline,thewaterlevelinthepipewasmeasuredbythe"LevelProbes"LP4.1thruLP4.4atfourpositions,oneaboveanother.Inthetestswiththeshortdischargeline,thisinstrumentaitonwasextendedbythemeasuringpointLP4.5abovethemeasuringpointLP4.4;seeFigure8.8ThemeasurementsignalsfromtheseLevelProbeswererecordedonvisicordersandmagnetictape.ABartoncell,measuringpointL4.1inFigure8.4,wasusedtosetandmeasurethewaterlevelinthedischargelinebeforeteststart.ThereadingofthatmeasuringpointwasinterrogatedbythecomputerbeforeandduringthetestandwasstoredTheindicationsoftheLevelProbesandalsotheindicationsoftheBartoncellwereusedtodepictthetimevariationofthewater.levelinthedischargeline.ItmustbetakenintoconsiderationthattheresponsespeedoftheBartoncellistooslowfortherapidchangesofthewaterlevelduringventclearingandaftertheclosingoftheSRV.Themeasuringpointwasusedessentiallytodeterminethesteady-state.waterlevelsinthedischargeline.Figures8101and8.102showtwotypicalexamplesofthevariationofthewaterlevelinthepipefortheintervaltest15.1withthelongdischargelineand20.1withtheshortdischargeline.Itwasfoundthatintwoinstancesinintervaltest15.1(Figure8.101),thewatercolumnbrieflyexceededtheexternalwaterlevel,butfellbackimmediately.Thesetwotestpointsrepresentthemaximumwatercolumnrisemeasuredintheventclearingtests.Intheintervaltest20.1,thewatercolumndidnotreachtheleveloftheexternalwatersurfaceinanyinstanceafterclosingoftheSRV.Themaximumwaterlevelrisewasgenerallyfound,inalltests,tooccurafterthethirdvalveactuation.REV1,3/798-45 PROPRIETARYToevaluatetheeffectofvacuumbreakeroperationonthewatercolumnrefloodfollowingventclearing;Test32,withonelockedvacuumbreakerandatimeintervalof3secondsbetweentheclosingofthevalveafterthefirstactuationandthenextactuation,wasincluded.Figure8-105showsthevariationofthemovementofthewatercolumninTest32.Ascanbeseennoadverseeffectswererecorded.8.43CheckingandCalibrationoftheNeasuringInstrumentationThecalibrationandtheelectricalandphysicalcheckingofallsensorsbefore,duringandafterthetestswereperformedinaccordancewiththeTestandCalibrationSpecifications.Fig.897showsdiagrammaticallythephysicalcalibrationofthesensors,thesettingandcalibrationoftheamplifiersandrecordinginstruments,andthequalityinspectionofthesensors.Pig.8.98showsthetimeintervalsstiplatedforthechecksandcalibrationsintheTestandCalibrationSpecifications.Fig.8.99clarifiesthechainofthecalibrationsystemfromthenationalstandardsofthePhysikalisch-TechnischeBundesanstalt(PTB)tothemeasuringinstruments.ThepressuresensorsP5.1thruP5.10usedinthetestswerefullyoperableuntiltheendofthetests.Thelowestinsulationresistanceof1.2x10~0measuredatP5.1afterthetestscanbeclassifiedas"good".ThepipepressuresensorP4.1failedon31Nay1978Itwasreplacedbyanewsensorforthesubsequenttests.WiththisnewsensorP4.1~thelowestinsulationresistanceforthegroupofpipepressuresensorsafterthetestswas3x10~~,whichwasverygoodTherewerenofailuresforthestraingaugesSG41thruSG4.8,SG5.1andSG5.2Herealso,averygoodinsulationresistancelevelwasrecordedwithalowestvalueof3x10~uatSG4.6afterthetests.J.ikewise,noneofthetemperaturemeasuringpiontsT5.1thruT5.10failed.Thelowestinsulationresistanceof1.3x10~~wassufficientlyhigh.844AnalysisofNeasurementErrorsBasedoninformationfromthemanufacturersofthemeasuringinstruments,KWUsowninvestigations,andtakingintoconsiderationtheexperienceaccumulatedinsimilartestprojects,themaximummeasurementerrorsfortheindividualsensorscanbeindicatedasfollows:REV.1,3/798-46 PROPRIETARYPressuresensorsP5.1thruP5.10Linearityerrorofthesensor2.5%ofmeasuredvalueinrangeof0to2barError2.5'5Reproductionerrorofthesensor0.2%of5barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder0.01bar05%0.5%Max.totalerrorx[0.01bar+3.5%ofthemeasurementvalue]PressuresensorsP4.1thruP4.5ErrorLinearityerrorofthesensor0.5%ofmeasuredvalueinrangeof0to20bar0.5%Reproductionerrorofthesensor0.1%of35bar0.035barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder05%05%Max.totalerrora[0.035bar+1.5%ofthemeasurementvalue]PressuresensorsP2.3andP25ErrorLinearityerrorofthesensor1%ofmeasuredvalueinrangeof0to40barReproductionerrorofthesensor0.1%of140bar.0.14barErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder0.5%05%Max.totalerrora[0.14bar+2%ofthemeasurementvalue]StraingaugesSG4.1thruSG48~SG5.1~andSG5.2ErrorToleranceoftheguagefactorInfluenceoftemperatureontheguagefactorREV1,3/798-47 PROPRIETARYErrorofthemeasuringamplifierErrorofthebalancingunitanrecorder05%0.5%Max.totalerrora5%ofthemeasurementvalueTemperaturemeasur~in~pintsT5.1thruT5.10ErrorofthesensorlocErrorofthemeasuringamplifierErrorofthebalancingunitandrecorder05%0.5SMax.totalerrorx[l~C+1$ofthemeasurementvalue]AfterthefirsttestsonMay10,1978andafterconclusionofthetestsonAugust2,1978,additionalphysicalchecksofthepressuresensorsinthewaterregionwereperformedbyincrementalloweringofthewaterlevelinthetesttank.Themax.deviationsfromthenominalvaluewereapproximately+0.01and-0.02bar.Fig.8.100illustratesafrequencydistributionofthesedeviationscombinedfrombothchecksandforallpresuresensors.ItshowsatypicalGaussiandistribution.Inordertorecordthehigh-frequencyprocessescorrectlyinfrequencyandamplitude,thedatawasacquiredinanalogformonmagnetictape.Porasensoreigenfrequencyofapproximately30kHz,thedynamicrangewaslimitednotbythesensorsbutratherbythecarrier-frequencymeasuringamplifierslocatedfurtheroninthecircuit.Thefrequencycutoffofthemeasuringamplifierswasat1.5kHzandthatofthemagnetictaperecorderswasat2.5kHz.ThefrequencycutoffsofthevisicordersweredeterminedbytheutilizedgalvanometersThesefrequencycutoffsareapproximately1kHz.Thefrequencyresponseofeachindividualgalvanometerwascheckedpriortothetests.8.45RepetitionTestsand~ReroducibilityoftheResultsToverifythereproducibilityofthemeasurementresults,arepetitionof5testswasspecifiedintheTestMatrix.Basedonapreliminaryassessmentoftheresultsafterconclusionofthetestserieswiththelongandshortdischargelines,thefollowingtestswererepeated(asmentionedpreviously):Longline:4.1through4.Rl15.1through15.RlIntervaltestsIntervaltestsShortline:REV1,3/798-48 PROPRIETARY19.1through19.R2201through20.R125.1through25.R2IntervaltestsIntervaltestsSingleActuationtestsInadditiontotherelevantinitialconditions,Table8.17alsogivesthemeasuredventclearingpressure(measuringpointP4.4),max.dyn.bottompressures(measuringpointP5.2)imardyn.wallpressures{measuringpointP5.10)andfrequenciesofthepressureoscillationsforthefirstSRVactuationineachoftherepetitiontests(>>CleanConditionstests").Acomparisonoftheabove-citedvaluesfortherepetitiontestsassociatedwitheachotherdemonstratesthegoodreproducibilityunderCleanConditions.Themaximumdeviationsfromthemeanvalueforeachpairofrepetitiontestsare(seeTable8.18):fortheventclearingpressureforthebottomandwallpressures10.75barora6%a0.05barora7Xforthefrequencyofthepressureoscillations105Hzora7%Themeandeviationsfromthemeanvalueofrepetitiontests,averagedforall5pairsfortheventclearingpressureforthebottomandwallpressuresforthefrequencyofthepressoscillationseachpairofoftests,are:10.37baror13K1002barora6%RO2HzOrX5%Figures8.37and8.38illustratesthemax.dynamicpressuresinthepoolduringtheventclearingforthemultiplevalveactuationrepetitiontestswiththelongline.Figures8.39and840showsthesamethingforthemultipleactuationrepetitiontestswiththeshortlineIncomparisonwiththefirstSRVactuationsunderCleanConditions,somelargerdeviationsareexhibitedhereinthetestsunderRealConditions(2ndto7thand10thSRVactuations).Thereasonforthesedeviationsisthattheinitialconditionsdiffersignificantlyfromeachother.Thevisicordertracesforeach"cleancondition"actuationatarepetitiontestpointisprovided:Tests4.1.1and4.Rl.l-Figures8-41and8-42REV1i3/798-49 PROPRIETARYTests15.1.1and15Rl1-Figures8-46and8-47Tests19.1.1and19.R2.1-Figures8-48and8-49Tests20.1.1and20.Rl.l-Figures8-59and8-60Tests25.land25R2-Figures8-64and8-65Avisualcomparisonofthetracesfromeachrepititiontestalsoshowsgoodreproducibility.Accordingly,itcanbesaidthat:Iftheinitialconditionsofthetestsaresetinacontrolledmanner(CleanConditions),thenthetestresultsarereproducible.Iftheinitialconditionscorrespondtotherandomlyprevailingoperatingstates(RealConditions),thenthemeasurementvalueslieinalargerscatterrange.85DATAANALYSISANDVERIFICATIONOFLOADSPECIFICATION8.5.1EvaluationofTestTankEffectsonBoundaryPressureNeasurementsInthisSection,vepresenttheoreticalandexperimentalinvestigationswhichshowthattheKarlsteintesttankrepresentsagoodsimulationofthehydraulicconditionsoftheSSESsuppressionpool.Meareconcernedprimarilywiththeeffectsexertedontheprocessesinthevaterbytheexistingboundarysurfacessuchasthewatersurface,tankbottom,movableorimmovabletankwalls.TheresultsoftheinvestigationfacilitatetheevaluationandtranspositioncftheboundaryloadsmeasuredintheteststoSSES.85.1.1EffectsofFreeRaterSurfaceandR~iidWallsTheeffectsofthefreewatersurfaceandtherigidwallsofthetankonthefluidpressurewillbeexplainedfirstbymeansoftheexamplesillustratedinFigure8-104.ThetophalfoftheFigureshowsthevelocitypotentialandflowfieldofasphericalbubblesubjectedtooverpressureorunderpressureinaninfinitelyextended,incompressiblefluid.Thepotentialfieldisdescribedbyasimple1/rlaw(Reference35).If,forexample,thesamebubbleislocatedinacylindricalrigidtankwhichispartiallyfilledwithfluid,thenthepotentialfieldandflovfieldhaveavisiblydifferentappearance(Figure8-100,bottom).Thedifferences.inthenonstationaryfluidpressure,whichisproportionaltothevelocitypotentialforsufficientlylovflowvelocity(pressurefield=potentialfield;seeReference4forexample),areclearlyevidentinthepressureprofilesontherightsideoftheFigure8-104.Thefreewatersurfaceconstrainsthepressuretozero,vhilethecylindricalwallcausesanincreaseinglymorepoverfulpressurerisewithREV1,3/798-50 PROPRIETARYincreasingdepth.Thenarrowerthetank,thegreateristhepressurerise.ThecalculationsrelatingtoFigure8-104wereperformedbythefinite-elementsmethod{Reference34)foratankdiameterof3mandawaterdepthof6m.Thebubblewas2.8mdeepand0.8mindiameter.Besidesthepressurefield,thereisalsoaneffectonthewatermasswhichiseffectivelyentrainedbythebubbleduringpulsationmotions(pressureoscillations)andthusalsotheoscillationfrequency.InthecaseshowninFigure8-104,thebubbleinthetankhasalargercoupledmassthanintheinfinitelyextendedmedium.Thisismanifestedbythefactthatthepulsationfrequencyofthebubbleiscorrespondinglylower(seeSection8.5.3.2).8.512MethodofImagesThemethodofimagesisanimportantaidwhichmakesitpossibletoclearlyunderstandthehydraulicactionsofthewatersurfaceandrigidwallsandtocalculatethemquantitativelyinasimpleway(Reference35).Ztisbasedonthefactthattheinfluenceofaplanerigidwallontheflowfieldofahydrodynamicpointsourcecanberepresentedbyasuperpositionoftheflowfieldwithoutthewall(infinitelyextendedfluid)andtheflowfieldofanimagesourceofidenticalsignandidenticalstrengthlocatedbehindthewall(Fig.8-105).Thesameholdsforaplanefreewatersurface,exceptthattheimagesourcehastheoppositesign.Usingthismethodofimages,theflowfieldofapointsourceinarectangular,vesselisobtainedfinallybyrepeatedapplicationofsuitableimagingoperations(Figure8-105dandFigure8-2).Theimmediatesignificanceofthemethodofimagesliesinthefactthatapulsatingbubblecanbeconceivedofasahydrodynamicsource,thusprovidingasimplemethodtocalculatethepressurefield.Ofspecialimportancefortheperformanceoftestsistheconsequencederivedbyinversionofthemethodofimages:Aconfigurationofbubblesoscillatinginparallelcanbesimplifiedinatestbysurroundingonebubblewithrigidwalls.Thiswillbeclarifiedfurtherinthefollowing.851.3TheTestStandasaSincCleCellBasedontheabovediscussion,anoscillatingbubbleinarectangularvesselisequivalenttoaplanefieldofsimultaneouslyoscillatingbubbles{Figure8-2).FromFigure8-2itfollowsfurtherthatvesselswithseveralbubblesarealsoequivalent,sincebetweeneachpairofbubblestheimagingwallsectioncanalsobeomitted.REV1,3f'798-51 PROPR1ETARYApplicationofthemethodofimagestothetranspositionofasystemofvalvesblowingdownsimultaneouslyinaplanttoateststandwithaquencherleadstothecelldivisionillustratedinFigure8-3..Asdiscussedinsection8.1,thewaterspaceoftheteststandwasformedaccordingtotheinteriorsinglecellsC,F,KandN(Figures8-3and8-108),sincetheyarethenarrowestandwillthereforeexhibitthehighestwallandbottompressures.Thatcanbeseenbyobservingthat,accordingtotheimagingprinciple,theyconservatively'simulatemorequencherslyingclosertogetherthanisactuallythecaseintheSSESsuppressionpool8.51.4SpatialDistributionsofPressureintheTestTankTogetmeaningfultestresults,pressuresensorshavetobemountedatsuitablepointsinthetesttank.Aseriesoftheoreticalinvestigationswasperformedinordertobetterassesstheirarrangement.Theyconsistedofcalculatingthespatialdistributionofpressurealongthetankwallsforvariousbubbleconfigurationsunderwater.TheKRUcomputercodeVELPOTwasusedforthisinvestigation.Abubblewassimulatedbyapoint,sourcenormalizedtounitsourcestrength.TheresultsareillustratedinFigures8-107to8-109.Figure8-107showsthecalculatedwallpressuredistributionforabubbleinthreedifferentpositionsnearthequencher:Case1Sourceonthetankaxis,0.7mabovethequencheraxisCase2Sourceonthetankaxis,atquencherelevationCase3Sourceatcenterofthequencher(eccentric).Theresultsshowthat.,theeccentricarrangementofthequencherwhichbecamenecessarybecauseofspacelimitationsinthetank,includingthecorrespondingpositioningofthepressuresensors(blacksquaresinFigure8-107),results,theoretically,inslightlyhighermeasurementvaluesforthepressures.Thenextcalculation(case4,Figure8-108)servestoanswerthequestionastohowthebubble'sforminfluencesthepressuredistribution.Todothat,thesinglesourcefromcase3,figure8-107,wasreplacedbyfouridenticalsourceswiththesametotalsourcestrength.Figures8-108and8-109showthattherearenomajordifferences.Notealsothegoodagreementseenbetweenthemeasuredpressuresfromshakedowntest081andthecalculatedvaluesi'nFigure8.109.Themodelcases3and4(singlebubbleatcenterofquencherand4-bubblearrangement)arebestadaptedtotheteststandgeometry.Sincetheassociatedpressuredistributionshardlydifferatall{Figure8-109),itisdemonstratedthatanexactREV1,3/798-52 PROPRIETARYknowledgeoftheairdistributionunderwaterisnotnecessaryforacorrectarrangementofthepressuresensorsInordertodemonstratetheconservativenatureofthechosensinglecell,asalreadyexplainedinSection8.5.1.3,thepressuredistributionformodelcase4iscomparedtothedistributioncalculatedfortheSusquehannaplantinFigure8-110.ThepressuredistributionintheteststandenvelopsthepressuredistributionintheSSES.Furthermore,thepressuredistributionintheteststandisenvelopedbythespecifieddistribution{FigureB-ill).8.5.1.5InvestigationoftheInfluenceofSavableSalleontheMeasurementResults/Fluid-StructureInteracti~on8.5.151GeneralRemarksIntheprecedingdiscussion,itwasassumedthatthesinglecellhasrigidandimmovablewalls.TheconstructionoftheKarlstein.testtankissuchthatthetank,despiteaseriesofstiffeningribs(seeFigures8-10to8-12),stillhasaresidualcompliance.Thetime-varyingloadsactingduringtheblowdownofthequenchercanthereforeexcitethetankintooscillationduetoFluid-StructureInteraction(FSI).Usingexperimentalandtheoreticalinvestigations,itwillbeshownthatinfluencesoftankoscillationsonthemeasuredboundaryloadscanbeneglected.Theexperimentalinvestigationsconsisted,firstly,ofmeasuringthetank'sresponsetoashortpressureimpulsewhichwasproducedbyanexplosivechargedetonatednearthequencher(Section8.5.1.5.2).Measurementsmadeduringthestart-uptestsontheteststandthensuppliedthetank'sresponsetotheloadsoccurringduringventclearing(Section85.1.5.3).Takingintoconsiderationtheinpulseresponse,itturnsoutthateffectsoftankoscillationsattheeigenfrequenciesarenegligible.Thisstatementislaterconfirmedbycalculationsandalsoisextendedtoforcedoscillations.85.1.52Ex2erimentalInvestigationoftheTank'sNaturalOscillationsTheexperimentalinvestigationofthetank'snaturaloscillationswasperformedwithimpulsiveexcitationbyanexplosivechargeinthewaterandsimultaneousmeasurementofthedisplacementsofthewallandbottomsectionsandofthefluidpressure.ThearrangementofthechargeandsensorsinthetankisillustratedinFigure8-112.Thepositionofthechargewaschosensuchthatthespatialloadprofileinthetankmatchestheprofileoftheblowdownloadsaswellaspossible.ThechargeitselfwasastoichiometricmixtureofhydrogenandoxygenwhichREV1,3/798-53 PROPRIETARYwasignitedinaplasticallydeformableflatcontainer(Figure8-113).Eightdisplacementtransducers(WA1toWA8)wereavailableforthedisplacementmeasurements.Theywerepositionedwiththeaimofobtainingthemostusefulinformation.Thearrangementofthepressuremeasuringpointsinthewater(P5.1toP5.10,Figures8-10to8-12)wasthesameasinthelaterblowdowntests.AsfortheevaluationofthepressuretracesinSection8.5.3,transducerP5.10waschosenasreferencepressuretransducerThechargewaslocatedatdifferentpositionsnearthequencherasshowninfigure8-112,i.nordertoobtainenvelopingloadprofiles.AtypicalresultisillustratedinFigure8-114,whichshowstherecordingsfromdisplacementtransducersWAltoMA8andpressuretransducerP5.10fortestno.2(chargeinposition2).Thelowestoccurringfrequenciesarebelow1Hz,buthavenothingtodowiththetank'sresponse,butratherrepresentsashiftofthezeropointThelowesteigenfrequencyofthetankisatapproximately13HzandisseenclearlyintheresponsefromtransducersWA2andWA3oscillatinginphase.Bothgagesareseatedonthebox-shapedstiffeningringsasshowninfigure8-112.Atthewallsectionsbetweenthestiffeners(WA4andMA6)andatthebottom(WA8),thefrequenciesthatoccuraremainlybetween30and60Hz.Theoscillationsoftheflatlowerstiffenerrings(WA1andWA5)arelesspronounced.Thesmallestdisplacementsarefoundattheconcretesections(WA7),wheresomeoftheamplitudesaresmallerhyanorderofmagnitude.ThepressuresignalfromP5..10showsdistinctexcursionsonlyduringthefirst100ms.Tobeabletobetterevaluatethetank'sfrequencyresponse,themeasuredtimevariationswereZourieranalyzedandpowerspectrawereformed.Thespectraassociatedwiththedisplacementtransducersonthesteelwall(MA2),concretewall(MA7)andbottom(WA8)andthepressuretransducerPS10areshowninFigures8-115to8-118.Itturnsoutthatthepreviouslymentioned13Hzoscillationinthelow-frequencyrangeisofgreatestimportance.Theassociatedtankdeformation(eigenmode)canbederivedfromthepointcorrelationsshowninFigure8-119.There,thedisplacementsofthedisplacementtransducersWA2,WA3andMA7,filteredbyabandpassfilterat13Hz,areplottedagainsteachotheratthesametimes.Thefitlinethroughthesetofpointshasapositiveslopeinthetopgraphandanegativeslopeinthebottomgraph.Therefore,displacementtransducerWA3(steelwallaboveMA2;seeFigure8-106)-oscillatesinphasewithWA2,whiledisplacementtransducerWA7(concretewall)oscillatesoutofphase.Thismeansthatthe13Hzoscillationcorrespondstoanovalizingmotionofthewall(seeFigure8-120).REV1,3/798-54 PROPRIETARY8.5.1.53ExperimentalInves~tiationoftheTank'sRe~sonsetoVentClear~inLoadsTheinvestigationsofthetank'sresponsetoventclearingloadswereperformedduringtheteststandshakedowntests.Tomeasurethetank'sresponse,thechoicewasmadetouseonedi,splacementtransducereachonthesteelwall(WA2),ontheconcretewall(WA7)andonthebottom(WA8).TheinstrumentationisshowninFigure8-121.Test08.1representsatypicalexampleoftheshakedownteststhatwererun.ThemeasuredtimehistoriesofthewallandbottomdisplacementsandofthereferencepressureP5.10areshowninFigure8-122.Thezero-pointdriftmentionedabovewaseliminatedbyusinga2Hzhigh-passfilter.Itcanbeseenthatboththepressureandthedisplacementsoscillateatthesameprincipalfrequencyof5.1Hz.Thesteelwall(WA2)andbottom(QA8)moveinphase.Theverysmallmovementoftheconcretewall(WA7)isalmostoutofphasecomparedtothepressureP5.10.Inaddition,thedisplacementtransducerWA8recordsahigher-frequencyoscillationat30Hz.Ithasalreadybegunweaklyatteststart,thendevelopsstronglyataboutthetimeoftheventclearing',andthendecaysagainabout300mslaterThephysicalinterpretationofthe5Hzoscillationisobvious.Thepressureoscillationiscausedbythepulsationoftheairbubblewhichiscreatedduringventclearing.Atthesametime,thetankcarriesoutforcedoscillationsatthefrequencyoftheforcingforce(5Hzpulsationoftheairbubble).Thesometimesphase-opposednatureofthedisplacementsofthesteelwallandbottom,ontheonehand,andtheconcretewall,ontheotherhand,makesitevidentthattheabove-discussedovalizingeigenmodeplaysadominantrole.Theoriginoftherapidlydecaying30HzoscillationseenatWA-8attheteststartisattributedtolocalforcestransmittedthroughthedischargelineandthequenchersupportduringventclearing.'Figures8-123todisplacementtimeduringshakedownspectraldensityduringshakedownlittleinfluence30HzlocaleffecfromP510showseffects.8-125showthepowerspectraldensitiesofthehistoriesforgagesWA2,WA7andWA8measuredtest081.Figure8-126showsthepowerofthepressuretimehistoryforP5.10measuredtest08l.A,reviewofthesefiguresshowsveryfromthe13HztankeigenfrequencyorfromthetseenatWA8.Figure8-126showingthersultspracticallynoinfluencefromeitheroftheseREVli3/798-55 PROPRIETARYPromthisitcanbeconcludedthatforallpracticalpurposestheKarlsteintesttankisrigidandhasnoinfluenceonthepoolboundarypressuremeasurementsmadeduringthetests.8.5.1.54TheoreticalInvestigations.andModelCalculationsoftheInfluenceofFluid-StructureInteraction851.5.4.1ComputationModelsTheanalysisdescribedbelowtocomputetheFSIonthemeasuredpressuresintheKarlsteintesttankwasperformedbyusingtheKWUcomputercodeKOVIBlAwhichwasdevelopedoriginallyandusedsuccessfullyfortheanalysisoffluid'-structureinteractioninthewaterpoolofKWU's69ProductLineBWRPlant.Theunderlying-theoryfollowsfromauniformformulationofthemechanicalprocessesbasedonpotentialtheoryandclassicalLagrangeandynamics.ItunifiesthedynamicsofthebubbleandtheFSIbyusingtheresultsofmodalanalyses.Inparticular,thefeedbackeffectsbetweenbubbleandstructureviathefluidareinc1uded.8.5.1542ModelParametersand~InutforCalculationsWithoutThemodelparametersandinputquantitiesforcalculationsoftheairbubbleoscillationsintherigidtankare:airmassflowintothebubble,watertemperature(=airtemperatureinstationaryequilibrium),hydrostaticpressureatbubbleposition,hydrodynamicmassparameterofthebubble,spatialpressuredistribution,initialvalues,(hubbleradius,etc.).Thetotalairmass(integratedairmassflow),watertemperatureandstaticpressureatthebubblepositionareobtainedfromthetestdata.Thehydrodynamicmassconstantofthebubbleandthespatialpressuredistributionareobtainedfromthecorrespondingpotentialcalculations(Figure8-107,case1).Thetimevariationoftheairsupplyintothebubblewasadjustedheuristicallybymeansofsystematictrialanderror,inparallelwiththeinitialvalues,insuchawaythatthecalculatedandmeasuredtimevariationsofthepressureattransducerP5.10exhibitedoptimalagreementThestart-uptest08.1wasusedasreferencetestforthesecalculations.TheairmassflowdeterminedinthismannerisillustratedinFigure8-127REVli3/79~8-56 PROPRIETARY8.5.15.43NodelParametersandXnautforCalculationswithPdjJustasforthedeterminationoftheairsupplyintothewaterpool,asemiempiricalmethodisusedforthestructuraldynamicsdata.Theyaredeterminedonthebasisoftheeigenfrequencymeasurementsdescribedpreviously.Inputdataforthecalculationare:eigenfrequency,modalmass,modalweight,dynamicpressuredistribution.Basedontheimpulseresponseofthetank(Figures8-115to8-117),itisplausibletoselecttheoscillationmodelyingat13Hz.Thatfixesthefrequency.Themodalmasscannotbetakendirectlyfromtheexperiment,butrathercanbedeterminedindirectlyviathemeasuredunitdisplacementsofthewall.TheunitwalldisplacementisillustratedinFigure8-128.Itisobtainedfromdisplacementsatthedisplacementtransducersbybandpassfilteringat13Hzandplottingsimultaneousvaluesofdisplacementwhicharenormalizedto1atthewatersurface.Thedisplacementdirectionisdefinedaspositiveiftherelevantwallsectionmovesinward.Thehydrodynamiccomponentofthemodalmass,(coupledwatermass)isthencalculatedbymethodsofpotentialtheory.Themodalweight,whichisequaltotheintegralloadrelativetothemodalmassandaveragedovertheunitdisplacement,isbasedontheloaddistributioncalculatedforcase1(Figure8-107,centeredbubble).Thedynamicpressuredistribution(seeFigure8-129)isobtainedfromtheunitdisplacement.bymeansofpotentialcalculations.85.1.5.4.4ResultsoftheFSIcalculationsTheresultsofthecalculationsconcerningtheinfluenceofFSIareshowninFigures8-130and8-131.Figure8-130showsthecalculatedtimevariationofthepressureatpressuretransducerP5.10,firstintherigidtank(withoutFSI)andthenintheelastictankwiththe13Hzeigenfrequency.Thereisaveryslightreductioninthepressureamplitudes,butitiscertainlynegligibleincomparisontothescatterofthemeasurementvaluesthemselves.AsisevidentfromFigure8-131,thefrequencyinfluenceofFSIalsocanbeneglected.InthatFigure,theoscillationfrequencyofthebubbleisplottedagainstthebubblevolume.ThebubblehasaslightlylowerfrequencywithFSIeffectsincludedthanwithout.REV1,3/798-57 PROPRIETARYAphysicallyclearexplanationoftheveryslightFSIeffectsfoundintheKarlsteinTestTankcanbeobtained.bycomparingthevolumesoffluidwhicharemovedbytheoscillatingwallandbottomandbythepulsatingbubble.Forabubblevolume{longline)of2.2m~andpressurefluctuationsofx0.4bar{seeFigure8-126),thevolumechangeofthebubbleisapproximately1m~isentropically.Incontrasttothis,fordisplacementslikethosefoundinFigures8-124and8-125thewallsandbottomuseuponlyabout0.05m~,whichisonly5%ofthewatervolumecomingfromthebubble.Therefore,duetothecomplianceofthetank,95%ofthewaterflowsupwardinsteadof100%(rigidtank).Thus,theresultoftheexperimentalandtheoreticalFSIinvestigationsisthateffectsofthecomplianceoftheKarlsteintesttankwallsandbottomonthepressureloadsmeasuredontheboundariesofthetankduringthetestscanbeneglected.852VerificationofSHVSystemLoadSpecificationDuetoSRVActuationThepressuresinsidetheSRVdischargelineweremeasuredatfourmeasuringpoints:justbehindtheSRVatmeasuringpointP4.1,inthecenteroftheblowdownpipeatmeasuringpointP4.2(measuringpointP4.5fortheshortdischargeline),justabovethenormalwaterlevelatmeasuringpointP4.3,andjustbeforetheinletofthequencheratmeasuringpointP4.4(seeFigure84).ThelongandshortdischargelinesareillustratedinFigures8-5and8-6.ThemeasuredpressuresinthedischargelinearedocumentedinSection8.4.1.8.5.21PressuresDuri~ntheVentClearincCProcessTypicalmeasurementtracesofthepressuresinthedischargelineareshowninFigures8-132and8-133.TheventclearingpressureisreadoffatP4.4.AsdiscussedinSection8.41,theventclearingpresureisdefinedasthepressurewhichisreadoffatthefirstpressuremaximumatP4.4.Atypicalfeatureofthispressurevariationisthedynamicovershootofthepressureabovethestationaryvalue.This-phenomenondoesnotoccurinsuchapronouncedmannerattheotherpressuretransducersalongthedischargeline.Thisdynamiceffectindicatesthatthepressurerequiredtoexpelthe~atercolumnisgreaterthan.thepressurenecessarytobringthesteammassflowthroughthequencher.Theexpulsionofthewatercolumn,isalsoclearfromthedifferenttimevariationsatP4.3andP4.4.Thepressureat8EVli3/798-58 PROPRIETARYmeasuringpointP4.3(abovethevatercolumn)risesmuchmoresteeplythanthepressureatmeasuringpointP4.4(insidethevatercolumn)Thedifferencebetveenthetvopressurosisthepressurewhichisnecessaryfortheaccelerationofthevatercolumn.Atthetimeoftheventclearing,thetwopressureshaveapproximatelyequalvalues.Butaftertheventclearingtheydifferagain,thistimeduetothedifferentpressurelossescausedbyflowresistancesinthepipe.8.521.1VentClearingPressuresfortheL~onLineThesteammassflowthroughtheSRVisapracticallylinearfunctionofthestagnation'pressure(reactorpressure).Sincethesteammassflowisoneofthemainparametersforthepressurebuild-upintheairregionofthedischargelineandthusfortheaccelerationofthewatercolumn,wewillplotthepressuresinthedischargelineasafunctionofreactorpressure.Thepressureinthebuffertank(P2.6)andnotthepressureinthesteamlinebeforetheSRVisusedasthereactorpressureforthetestssincethepressureinthebuffertankmorecloselysimulatestherepresentativestagnationpressureinthereactor.(seeFigure8-134).Todescribethedependenceoftheventclearingpressureonthereactorpressure,onlythosetestsforwhichtheinitialconditionsweresetandthusknovnexactlywereused.Thosearethetestsvithso-called<>cleanconditions".FromFigures8-135and8-136,itcanbeseenthatthemeasurementresultshavegoodreproducibilityforthetestswithcleanconditions.Thepressuresinthepipeincreasepracticallylinearlywithreactorpressure.Thefollowingtrendscanbeobserved:1)Aloweredwaterlevelinthedischargelineresultsinloverpressuresduringtheventclearing.2)Ahotpiperesultsinhigherpressuresduringtheventclearing.Thisisduetothesmallerpercentageofcondensationonthepipewall.3)Thepressure(atthetimeofventclearing)behindtheSBVisalwayshigherthantheventclearingpressureclosetothequencher.Thedifferenceisattributabletotheflowlossalongtheline.REV1,3/798-59 PROPRXETARY4)*Thepressure(atthetimeofventclearing)behindtheSRVincreases-withincreasingreactorpressure{orincreasingsteamflowratethroughthereliefvalve).Besidestheclean-conditiontests,thereisalargenumberofreal-conditiontestsandintervaltests.SincetheinitialconditionsinthemwererandomandwerenotvariedinacontrolLedmanner,themeasurementvaluesarescatteredoveramuchwiderbandthanintheclean-conditiontests.Hence,thesetestsarenotusablefortrendanalyses,butmaybeusedforverificationofmaximumspecificationvalues.Themeasuredmaximumvaluesare:PressurebehindtheSRV(atventclearingtime):19baratareactorpressureof72barVentclearingpressurebeforethequencher:14.5baratareactorpressureof72bar852.12VentClearingPressuresfortheShortLineFigures8-137and8-138showthemeasuredpipepressuresplottedagainstreactorpressureforcleanconditiontestswiththeshortdischargeline.Thesametrendsasseenwiththelonglineareseenhere.'Sincetheshortlinehasasmallerairvolumethanthelongline,whilethewatercolumntobeclearedandotherparametersremainthesame,thepressuresintheshortlinearehigherthanthoseinthelongline.Themeasuredmaximumvaluesare:Pressurebehindthereliefvalve{atventclearingtime):22baratareactorpressureof73barVentclearingpressurebeforequencher:18baratareactorpressureof73bar.85.213Tran~sositionoftheMeasurementValuestoSSESandComparisonwiththeDesicCndecificationTheverificationtestsinKarlsteinwererunwiththeactualgeometryofthereliefsystem,theactualSRV,andthehighestwaterlevelinthedischarge-line(6.2mabovecenterofquencher)thatoccursforSSES.Themeasuredventclearingtimesforthatwaterlevelandahighreactorpressure(69-81bar)wasbetween250and400msREV.1,3/798-60 PROP3IETARYFortheseventclearingtimes,theopeningtimeoftheSRV(measuredopeningtimes:29-60ms)hasnonoticableeffectontheventclearingpressure(seeFigure8-139).Hence,inregardtotheventclearingpressure,theonlyvariablewhosemaximumvalueforSSESwasnotcompletelycoveredwasthereactorpressure.Thefollowingextrapolationappliesforthat:a)PressurebehindthevalveatventclearingtimeTheMeasuredmaximumvalueforthelonglineis19baratareactorpressureof72barASlopeof25%isseeninfigure8-135.Extrapolatingto88bar,theresultis:Pox=23barforthelonglineTheMeasuredmaximumvaluefortheshortlineis22baratareactorpresssureof73barSlopeof25%isseeninfigure8-137.Extrapolatingto88bar,theresultis:Pax=26barfortheshortlineThedesignvaluegiveninSection4.1.2.1is550psi=37.93bar.TheKarlsteintestsdemonstratethatthedesignvalueisveryconservativefortheventclearingcase.b)Ventcleari~npressureThemeasuredmaximumvalueforthelong.lineis14.5barforreactorpressureof72barASlopeof12.5%isseeninfigure8-136.Extrapolatingtoareactorpressureof88barresultsinPmax=16.5barforthelongline.Themeasuredmaximumvaluefortheshortlineis18baratareactorpressureof73bar.ASlopeof12.57isseeninfigure8-138.Extrapolatingtoareactorpressureof88barresultsinPmax=20barfortheshortline.ThespecificationvaluegiveninSection4.1.1.2isPmax27barTheKarlsteintestsdemonstratethatthespecificationvaluefortheventclearingpressureisveryconservative.85.2.2PressuresDuri~ntheStationaryCondensationofSteamAboutonesecondaftertheopeningoftheSRV,theventclearingprocessiscompletedandthephaseofsationarysteamcondensationbegins.Inthisphase,thepressuresinthedischargelinearedeterminedbythesteammassflowandtheflowresistance.SincethesteamREV1,3/798-61 PROPRIETARYmass.flowisproportionaltothereactorpressure,hereagainwewillinvestigatethedependenceofthepipepressurestothereactorpressure.85.221Lo~nLineFigures8-140and8-141showthedependenceofthesteadystatepressureonthereactorpressure.Meseethattherelationcanberepresentedverywellbyastraightline.Asaresultofpipefriction,thestationarypressurebehindtheSRVhashighervaluesthanthepressuresjustbeforethequencher.Italsoexhibitsafasterincreasewithreactorpressure.Themeasuredmaximumvaluesare:17.5baratreactorpressureof72barforthepressurebehindtheSRV(P41)>l0baratreactorpressureof70barforthepressurebeforetheinlettothequencher(P4.4)852.22ShortLineFigures8-142and8-143showthedependenceofthesteadystatepressureonthereactcrpressure.Thebehaviorofthepressurebeforethequencher(P4.4)ispracticallyidenticalfortheshortlineandlongline.Thisisnotsurprising,sincethispressuredependsonlyontheflowresistanceofthequencher.ThepressuresbehindtheSRVarelowerthanthoseforthelongline,butdisplaythesameincrease,withreactorpressure.Thedifferentflowresistancesofthetwodischargelinesaremanifestedhere.Toclarifythiseffect,thevariationofthestationarypressureatthemeasuringpointsalongthedischargelineareplottedinFigure8-144fortheshortandlonglines.Theaveragepressureswereused,i.e.,thepressureswerereadofffromtheinterpolation1'inesat88bar(seeFigures8-190to8-143).Themeasuredmaximumvaluesfortheshortlineare:PressurebehindtheSRV(P4.1)16baratareactorpressureof72bar,and15baratareactorpressureof63barREV1,3/798-62 PROPRIETARYPresurebeforeinlettothequencher(P4.4)9.5baratareactorpressureof71bar,and9.0baratareactorpressureof65bar.85.2.2.3TranspositionoftheMeasurementValuestoSSFSandcomparisonwiththeDesignSpecificationAswasthecasewiththeventclearingpressure,theonlyvariablewhosemaximumvalueintheSSESwasnotcompletelycoveredbytheteststandwasthereactorpressure.Anextrapolationofthemeasuredmaximumvaluestoareactorpressureof88baryieldsthefollowingresults:a)LongLineThemeasuredmaximumvaluebehindtheSRVis17.5baratareactorpressureof72bar.ASlopeof22%isseeninfigure8-140Extrapolatingto88bar,theresultis:Pm~~--21barThemeasuredmaximumvaluebeforequencherinletis10baratareactorpressureof70bar.ASlopeof16%isseeninfigure8-141.Extrapolatingto88bar,theresultis:Pmmx=13bar.b)ShortLineThemeasuredm'aximumvaluebehindtheSRVis16baratareactorpressureof72barand15baratareactorpressureof63bar.ASlopeof22%isseeninfigure8-142.Extrapolatedto88bartheresultis:Pmax=19.6barand20.5bar,respectively.Themeasuredmaximumvaluebeforequencherinletis9.5baratareactorpressureof71barand9baratareactorpressureof65bar.ASlopeof16%isseeninfigure8-143.Extrapolatedto88bar,theresultis:P=12.5barand13.0bar,respectively.maxItcanbestatedthatthedesignvalueof550psi=37.93barforthestationarypressurebehindthevalveisveryconservative.85.23ExternalLoadsonthe~uencherandBottomSup2ortInthisSectionweshalldiscussthemeasurementresultswhichprovideinformationabouttheexternalloadsonthequencherandREV1,3/798-63 PROPRIETARYbottomsupport.ThemeasuringpointsprovidedforthatpurposeareshowninFigure8-13,andareasfollows:SG41/42SG43/4SG45/46SG47SG48Bendingatquencherarm1Bendingatquencherarm2BendingatthebottomsupportLongitudinalstrainatthebottomsupport'orsionatthebottomsupportStrainsweremeasuredatallmeasuringpoints.Themeasuredstrainswereusedtocalculatetheloadswhichproducedthestrains.Theloadsthuscalculatedarestaticequivalentloadswhichcontainhydraulicandalsostructural-dynamicaleffects.85.2.31VerticalForce8.5.23.11MeasurementoftheVerticalForceTomeasuretheverticalforce,twostrainconnectedinsuchawaythattheymeasureverticalforces.Thefollowingrelationexistsbetweenthegauges,SG4.7,werestrainsresultingfromloadandstrain:F~A~E~eBF=33~ekNBwhereA~.016m252F~2.06x10N/mmIfweinserteinpm/m,wethengettheverticalforceinkN.Thisequationwasusedtoconvert'hemeasuredstrainsintoverticalforces.8.5.2.3.1.2MeasuredVerticalForcesFigure8-145showsatypicalmeasurementtracefortheverticalforceItincreasesrapidlyduring'theexpulsionofthewatercolumnand,afterreachingthemaximumvalue,returnsquicklytozero.8523.1.21Lo~nLineTheverticalforceexhibitsastrongrelationshipwithventclearingpressureasshowninFigure8-146Thisholdstrueforalltests,eventhosewithrandominitialconditionssuchastherealconditionsandmultipleactuationtest.AsdiscussedinSection8.5.21.3,theventclearingpressureisinturninfluencedbythereactorpressure,initialwatercolumninthe'dischargeline,dischargelinetemperature,etc.andwasREV13/798-64 PROPRIETARYextrapolatedouttoamaximumreactorpressureof88bar.Therefore,themaximumverticalloadwillbeextrapolatedtothemaximumventclearingpressurefromSection8.5.2.1.3.Themeasuredmaximumvaluefortheverticalforceis:149kNata128barventclearingpressure.8.5.2.3.1.22ShortLineFigure8-147illustratesthedependenceoftheverticalforceontheventclearingpressure.Inprinciple,thesamediscussionasinSection8.5.2.3.1.2.1forthelonglineappliesherealso.Themeasuredmaximumvaluefortheverticalforceis:192kNata168barvent-clearingpressure.Theverticalforcesrelativetotheventclearingpressurearepracticallythesame.85.2.3.1.3TranspositionoftheMeasurementValuestoSSESAswasdiscussedpreviouslyfortheextrapolationoftheventclearingpressures,themeasurementvaluesfortheverticalforcecanalsobetransposeddirectlytotheplant.Forverificationofextremeconditionsintheplant,themeasurementvaluesareextrapolatedtoareactorpressureof88bar.Theextrapolationcanbeperformeddirectlyviatheventclearingpressure.8.5.23.1.31LongLineThemeasuredmaximumvaluewas:149kNata12.8barvent-clearingpressureSlope=13kN/bar(Figure8-146)AccordingtoSection8.5213,theextrapolatedvent-clearingpressureforthelonglinewas16barExtrapolationoftheverticalforceto16baryields:Fymax=190kN85.23.1..3.2ShortlineThemeasuredmaximumvaluewas:192kNat168barvent-clearingpressureSlope=13kN/bar(Figure8-147)AccordingtoSection8.5.2.1.3theextrapolatedventclearingpressurefortheshortlinewas20bar.REV1,3/798-65 PROPRIETARYExtrapolationoftheverticalforceto20baryields:FvmxInadditionFigure8-147,showsameasuredvalueof149kNata12barvent-clearingpressure.Thisleadstoamaximumextrapolatedverticalforceof:F~~y=252kN8523.13.3SummaryTheextrapolationofthemeasurementresultsfortheverticalforceyieldsama'ximumvalueof:F~~~~=,252kN>>InFigure4-11,thespecifiedverticalforceisgivenas860kN.Dnthebasisofthemeasurementresults,thespecificationvaluecanbeviewedasextremelyconservative,bothinthemaximumvalueandalsointheload-versus-timefunction.8.5.232'ore'ionalMoment8~2~3.2lMeas~nementoftheTorsionalMomentTomeasurethetorsionalmoment,twostraingauges(SG4.8-Figure8-13)wereconnectedinsuchawaythattheymeasurestrainresultingfromtorsionalmomentonly.AccordingtoReference41,thereisaverysimplerelationbetweenthetorsionorshearstrainandthemeasuredstrain,whenthestraingaugesaremountedata45oanglerelativetotheprincipalshearstressdirection.Qehave:YmsshearstrainThereforeisincethestraingaugesSG4.8weremountedata45oinclinationtotheverticalaxis,vehave:Gz=shearstressGmsshearmodulusY=2.eandr~Da2REV1,3/798-66 PROPBIETAHYIpg=torsionalmomentI=polarmomentofinertiaPr~outsideradiusofthetwistedcylindricalbarYrG'PQethusobtaintherelationbetweentorsionalmomentandmeasuredstrax.nTheshearmodulusisdefinedasG2(1+p)MithE=2.06x10sN/mm~andDap=poisson'sratioMeget:'p=0~3G~7.9x10'/mmeThepolarmomentofinertiaisdefinedas7f~D(1-D/D)4p32Therefore:I4.64x10mInsertingthevariousnumericalvalues,weget:0.41',InsertingE.inMm/m,thisequationgivesusthetorsionalmomentinkN-mHEVlg3/798-67 PROPRIETARYThisequationwasusedtoconvertthemeasuredstrainsatSG4.8intotorsionalmoments.Thetorsionalmomentsobtainedinthismannerrepresentstaticequivalentloads.85.232.2MeasuredTorsionalMomentsFigure8-148showsatypicalmeasurementtraceforthetorsionalmoments.Aftertheendoftheventclearingprocess,(approximately1secondafterteststart)theamplitudesofthemeasuredtorsionalmomentsareverysmallcomparedtothemaximumamplitudeduringtheventclearingprocessThereisafactorof6-7differencebetweenthetwoofthem.Themaximumamplitudeofthetorsionalmomentoccursmuchlaterthantheexpulsionofthewatercolumn.8.5.2.32.2.1L~onLineThetorsionalmomentatthebottomsupporthas,itsoriginonlyinunsymmetricalprocessesatthequencherduringtheventclearingandduringthetransitiontostationarycondensation.iFigure8-149showsthedependenceofthetorsionalmomentontheventclearingpressure.Sincetheventclearingpressureisadirectinfluencingparameter(seeSection8.5.2.3.1.2.1)wewillcorrelatethetorsionalmomentwiththatvalue.Thesharplypronouncedscatterbandisanindicationthatarandomprocessissuperimposedonthatdependence.Thatisexpressedbythefactthatthetorsionalmomentisbroughtaboutbyrandomunsymmetry.Themeasuredmaximumvalueofthetorsionalmomentis:M=55.8kN-mata14barvent-clearingpressure.TmaX$.523.2.g2ShortLinePigure8-150againshowsthedependencesofthetorsionalmomentontheventclearingpressure.Inprinciple,thesituationisthesameasintheprecedingSectionfo-thelongline.Themeasuredmaximumvalueofthetorsionalmomen'tis:39.2kN-mata18barvent-clearingpressure.8.523.2.3Tran~sositionoftheMeasurementValuestoSSESShentransposingthemeasurementresultstoSSES,weshallconsiderinaconservativemannertheloadcarriedbythedischargeline,whichintheteststandisconnectedrigidly(butREVli3/798-68 PROPRIETARYnotinaleaktightmanner)tothequencherandbottomsupportbymeansofweldbrackets(seeFigure8-13and8-14)incontrasttothefreemovingslidingjointatSSES.Todothat,wemaketheassumptionthatthedischargelineisfixedinatorsionresistingmanneratthefirstbendabovethequencher.Thatresultsinthefollowingpicture:DischargeLineQuencherBottomsupport////ThetorsionalmomentN~actsatthequencher.ThetorsionalmomentN~~wasmeasured.atthebottomsupport.ThedischargelinecarriesthetorsionalmomentM><.Therefore:+M2Promtheequalityoftherotation,weget:Therefore:"TVYGIpGTl11G~Ipl="T2'2'2G~Ip2Tl=P1~22T2P211Mehavethefollowingdimensions:r=0.1775mlar1g0.125m0.45m1r=0.162m2a2g~0.3.445m~11.313m2REV1,3//798-69 PROPRIETARYTherefore:-444.64.10mZ4.0,.10"m4Therefore:-26.6Tl4.640.16211.313M240'17750'4526.6TTlŽT2Tl(1+1)26.6M1.0376M1Thus,theloadtransmittedtothedischargelineislessthan4gofthattransmittedtothebottomsupport.If,withouttakingintoconsiderationthedischargeline,wefirstusePigures8-149and8-150asthebasisforanextrapolationofthemeasuredmaximumvaluestomaximumvent-clearingpressureforthecorrespondingdischargeline,thenwegetthefollowingmaximumvalues:a)longlineMr,~=598kN-mb)short.lineMr)~ax=43.2kN-mIfwenowconsiderthetorsioncarriedbythedischargeline,thenthisvalueisincreasedtoamaximumof:"ri~x=62kN-mThetorsionalmomentspecifiedin4.1.2.6fortheguenchersupportwas40kN-mtobeappliedasastepfunctionAtorsionalmomentstepfunctionappliedtoanundampedonemassREVl,3/798-70 PROPRIETARYoscillator(quencheractingasinertialmassandbottomsupportasatorsionalspring)correspondstoamaximumresponseof:M<<=2(40)kN-m=80kN-mSincethemaximumtorsionalmomentderivedfromtheKarlsteintestsisM<=62kN-m,thespecificationisconservative./8-5.2.33Bean~inncnenteattheguenchecAten85.2.33.1MeasurementoftheBendingMomentsIntheKarlsteintests,thebendingmomentsveremeasuredinthehorizontalplane(paralleltothetank'sbottom)andalsointheverticalplane,atbothofthequencnerarms.Toaccomplishthat,twostraingaugeseachvereconnectedinsuch.awaythattheymeasuredunsymmetricalstrainsresultingfromnormalstresses(unsymmetricalcomponent).Thefollowingstraingaugesweremountedforthatpurpose(seeFigure8-13:SG4.1)MomentsinverticaldirectionSG43)SG4.2)MomentsinhorizontaldirectonSG44)Thestraingaugesweremountedapproximately150mmfromtheweldbetweenthequencherarmandthecentralball.Thesectionmodulusofonequencherarmis:3W+D(1--)aQehave:'D~0.4064maD=0.3744ma=cE=M/WM=cEWThisleadstotheequationbetveenquantities:M=0.38-c8-71 PROPRIETARYThisgivesthebendingmomentinkN-m,ifcisinsertedinpm/m.Withthisequation,allthemeasuredbendingstrainswereconvertedintobendingmoments.Thebendingmomentsthuscalculatedarestaticequivalentloads.8.5.23.3.2MeasuredBendingMomentsFigure8-151showsatypicalmeasurementtraceofthemeasuredbendingmomentsatthequencherarms.Weseeclearlythatthemaximumvaluesoccurmuchlaterthantheclearingofthequencher.Theevaluationoftheindividualbendingmomentsrelatestothetotalresultantbendingmoment,ie.,thebendingmomentwhichactuallyloadsthe.quencherarm.Theresultantbendingmomentisobtainedbyusingtherelationship:M~'gM+M2x'esyzThebendingmomentsMgarereadoffatSG4.2and4;4.ThebendingmomentsMzarereadoffatSG4.1and43Theresultantbendingmomentsexhibitnodeterministicdependenceontheventclearingpressure,asshovn.inFigure8-152.Therefore,theresultantbendingmomentsonthequencherarmsmustbeconsideredasstatisticalvalues.Themeasuredmaximumvalueofthereultantbendingmomentis63kN-m.e8.52.33.3TranspositionoftheMeasurementResultsintotheWeldInSection4l.2.5,thebendingmomentsintheweldwerespecified.IntheKarlsteinteststand,thestraingaugesweremountedabout150mmfromtheweldinordernottomeasurelocalizedstressesduetotheweldandtheintersectionbetweentheballcentralbodyandthequencherarm.Availableexperienceindicatesthatthisdistanceissufficienttomeasureastressprofilewhichisindependentofshapefactors.Fromthespecifiedforceandmoment(Table4-10),weobtainforthedistancebetweentheweldandtheforceproducingthebendingmoment:lp~~Oo6551929Bytreatingthequencherarmasacantileverbeam,weobtainforthemaximumstressandthusforthemaximumbendingmoment:g0.655=Mg(0.655-0.15).REV~1>>3/798-72 PROPRIETARYM=bendingmomentintheveldBmaxM-=measuredbendingmomentBmaxTherefore:=1.297MBmaxBmeasThus,basedonthemeasuredmaximumresultantbendingmomentof62KN-m(seeSection852.3.3.2),weobtainthefollowingmaximumbendingmomentintheweld:'aximumresultantbendingmoment:81kN-m852.33.4~SecifiedStaticEguivalentLoadsAsalreadynotedabove,themeasuredbendingmomentsaretobeconsideredasstaticeguivalentloadsInSection4.1.2.5Table4-10,two'ontributionswerespecifiedwithrespecttothebendingmomentintheweld:a)astepfunctionhavingastepheightof19kN-mb)amaximumdifferentialpressurevhich,accordingtoSection4.1.3.7,is08barfromKKBtraceNo.35witha0.5multiplier.Thisresultsinamaximumdifferentialpressureof0.4bar.Thecontributionofthedifferentialpressureistobeviewedstatically,since,accordingtoSection413.5,thefreguencyofthedifferentialpressureisapproximately6Hz.Thebendingeigenfreguencyoftheguencherarmisontheorderof100Hz.Thecontributionofthedifferentialpressuretothebendingmomentintheweldisthus:11.4kN-mThecontributionofthestepfuncionistobevieweddynamically.Therefore,thesameconsiderationsareapplicableasthosemadeforthetorsionalmomentsinSection8.5.2.3.2.3.Accordingly,wehavethefollowingstaticeguivalentloads:ComponentinoneDirectionContributionfromstepfunction=2X19=38KN-mContributionfromdifferentialpressure=11.4KN-mTotal=49.4KN-mREVli3/79'-73 PROPRIETARYResultantMomentContributionfromstepfunction=38x~2=53.7KN-mContributionfromdifferentialpressure=11.4KN-mTotal=65.1KN-m8.5.23.3.5FvaluationoftheMeasurementResultsAsalreadymentionedinSection8.5.23.3.2,thebendingmo'mentsonthequencherarmaretobetreatedasstatisticalvalues.Figure8-153showsthefrequencydistributionofthemeasuredmaximumbendingmomentsineachtestsandtheresultingfrequencydisrihutionofthevaluestransposedintotheweld.Thefrequencydistributionsarebasedonthepeakmaximumvalueofeachindividualtest,whichweremeasuredeitheratSG4.1/4.2oratSG4.3/4.4.Thespecifiedstaticequivalentloads(seeSection8.5.2.3.3.4areintroducedfor7000responsesofthereliefvalve.Therefore,theloadsaretobeevaluatedinafatigueanalysis.ItfollowsfromFigure8-153thatthemeanvalueofthemeasuredmaximumvaluestransposedintotheweldis35kN-m.Exceptforthreecases,thespecifiedresultantbendingmomentsalsocoverthemaximummeasuredvalues.Thequencherisbeingevaluatedforthesemeasuredmaximumvalues.Itshouldbenotedthatboththespecifiedstationaryinternalquencherpressureof22.0barandtheresultingthermalloadof219~Cwerefoundtobeveryconservativewhencomparedtothemaximumextrapolatedvaluesof13.0barandtheresultingsaturatedsteamtemperatureof195~Cmeasuredduringthetests.(Section8.5.2.23).8.5.23.4BendingMomentsattheBottomS~ugort852.3.4.1MeasurementoftheBending.MomentsTomeasurethebendingmomentsathebottomsupport,twostraingaugescapableofmeasuringthebendingstrainsweremounted.In,themeasurementarrangement,thebendingstrainscouldbemeasuredintwomutuallyperpendiculardirections(seeFigure8-13).Thestrainsformomentsaboutthex-axisweremeasuredwiththestraingaugesSG4.5.Thestrainsformomentsaboutthey-axisweremeasuredwiththestraingaugeSG46.REVlg3/798-74 PHOPRIETARYThesectionmodulusofthebottomsupportis:D4W=-D(1--)332a4aW~1.307x10m-33Wehavea~E~@~M/WThisleadstotheequation:M~0.27~cThisequationgivesthebendingmomentinkN-m,ifcisinsertedinpm/m.Thisequationwasusedtoconvertallmeasuredbendingstrainsofthebottomsupportintobendingmoments.Thebendingmomentsthuscalculatedarestaticequivalentloads.8.52.3.4.2MeasuredBendingMomentsInFigure8-151,thebendingmomentsatthebottomsupportcanbeseenunderthetracesofthebendingmomentsatthequencherarms.Themaximumvaluesoccuratalatertimethantheventclearing.Buttheyoccuratthesametimeasthemaximumvaluesofthebendingstrainsatthequencherarms.Themaximumstrainresultingfromtorsiondoesnotoccuratthetimeofthemaximumbendingstrain(seeFigure8-151,SG4.8).Theevaluationofthebendingmomentsrelatestotheresultantbendingmoment,i.e.,thebendingmomentwhichactuallyloadsthebottomsupport.Theresultantbendingmomentisobtainedbyinterconnectingtheactualload-versus-timefunctionsoftheindividualcomponentsthroughtherelation:ThebendingmomentsMzarereadoffatSG4.5andthebendingmomentsM>atSG4.6Themaximumresultantbendingmomentwas54.5kN-mTheresultantbendingmomentsdisplaynodependenceontheventclearingpresure,asshowninFigure8-154.Hence,thesameconclusionsthatweredrawnforthebendingmomentsatthequencherarmsareapplicablehere,also.BEV.1,3/798-75 PBOPRIETABY8.5.234.3SpecifiedStaticEquivalentLoad.Asalreadymentioned,themeasuredbendingmomentsaretobeviewedasstaticequivalentloads.Thebendingmomentsatthebottomsupportareintroducedthroughthequencher.Section4.12.4andTable4-7specifyatransverseforceof44kNonthequencherwasusedasstepfunction.Inaddition,amaximumdifferentialpressureof0.4baronthequencherwasspecified.Thecontributionresultingfromthedifferentialpressureistobeviewedasastaticallyactingload.Itamountsto48kN.Note:Thedischargelineandthebottomsupportwerenotconsideredhere.Thepresssuredifferencewasformulatedonlyovertheprojectedareaofthequencher.Thespecificationthenyieldsthefollowingtransverseforcesonthequencher:Contributionfromstepfun'ction=2x44=88kNContributionfromdifferentialpressure=48kNTotal=136kNStraingaugesSG4.5andSG4.6weremountedapproximately0.5mbelowthecenterofthequencher.Transposedtothislocation,thespecificationyields:68kN-m85.23.44EvaluationoftheMeasurementResultsFigure8-155showsthefrequencydistributionofthemeasuredmaximumbendingmomentsatthebottomsupport.Themeasuredmaximumvaluesarealsocoveredbythespecification.Thus,theKarlsteintestshavedemonstratedthatthespecifiedtransverseforcesonthequenchercanbeviewedasveryconservative.85235ForcesontheQuencherIntheKarlsteinQuencherTests,onlybendingmomentswereabletobedeterminedforthequencheritself.InSection4.1.2,forcesandmomentsonthequencherwerespecified.Thespecifiedmomentswerecalculatedfromtheforces.Themeasuredmomentsarewithinthespecification.Therefore,wecanconcludethattheforcesarealsoverified.REVli3/798-76 PROPRIETARY85.23.6InfluenceofanAdgacentQuencherDuringtheclearingofthequencher,strongturbulencesandeddiesoftheexpelledandambientwaterdeveloparoundthedischargingquencher.Inparticular,aftertheventclearingthequencherissurroundedbyalargenumberofairbubbleswhichrepresentalocallycompressiblevolumeinthewater.Thisstate,whichformsaroundthedischargingq<<niche>ipreventseffectsfromtheblowdownofanadjacentquencherfrompenetratingtothequencherunderconsideration.Itisthereforeunderstandablethat,intheKMUinplanttestswithintheBrunsbuttelandPhilippsburgnuclearpowerplants,noincreaseoftheloadonthequencherandbottomsupportwasfoundfortheresponseofseveralquenchersincomparisontotheresponseofonequencher(Reference6).Aneffectofaloadononequencherduetothefiringofanadjacentquencheristobeobservedonlywhentheadjacentquencherblowsdownalone.Inthatcase,adetailedevaluationwasmadefortheBrunsbuttelblowdowntests(Reference38).Theresultoftheinvestigationwasthatthemeasuredloadsareenvelopedbyapressuredifferenceof0.2barappliedovertheadjacentinternalstructuresinthepoolatthequencherlevel,i.e.,alsooverthequencher.Amaximumpressuredifferenceof0.4baroverthequencherarmswasspecifiedforSSES.TheventclearingpressuresanddynamicpressuresinthewaterpoolobtainedforSSESfromtheKarlsteintestsareofthesameorderofmagnitudeasthecorrespondingmeasurementresultsinBrunsbuttel.Therefore,thespecifieddifferentialpressureof0.4baroverthequencherarmscanbeviewedasconservativelyenveloping.8.5237LoadsontheQuencherDuringSteamCondensationThemaximummechanicalandthermalloadsonthequencherduringthecondensationphaseoccurduringthephaseofintermittentcondensation.InSection4.1.2.7,theloadsresultingfromintermittentcondensationweretakenasthebasisforthefatiguedesignofthequencher.TheevaluationoftheloadsonthequencherduringsteamcondensationintheKarlsteinteststhereforerelatesprimarilytothephaseofintermittentcondensation.REV1,3/798-77 PROPRIETARY8.5.2.371Manifestation/ormsofIntermittentCondensationintheKarlsteinTestsAsdiscussedinSection8.1.3,thecondensationtestsvereperformedalongthelowerandupperboundarylinesoftheoperationfieldforwatertemperatures<30~Candalsoforwatertemperatures>590C.Inbothregions,theintermittentcondensationphaseoccursforverylowreactorpressures(approximatelybetween2and4bar).InSection84.2itisshownthatthemaximumvaluesforthedynamicpressuresinthevaterregionoccur.duringintermittentcondensationincoldvaterThesameistruealsofortheloadsonthequencher.Foritheevaluationandcomparisonwiththespecification,weusethemeasurementvaluesofthebendingmomentsatthequencherduringtheintermittentcondensationinthecoldpool.ThemeasurementvaluesaredocumentedinSection8.4.2.85.23.7.2IllustrationoftheMeasurementValuesThetimedurationoftheintermittentcondensationinthecoldpoolwasabout100seconds.Thetotalnumberofcondensationeventsatthequencherwas52.ThemaximummeasurementvaluesoccurredintheverticaldirectionatSG4.3.Thefrequencydistributionoftheresultantbendingmoments(SG4.'3/4.4)atthequencherarmisshowinFigure8-156.Themeanvalueofthemaximummeasurementvaluesofeacheventis11.8kN-m.Themaximummeasuredvaluewas66.5kN-m.Thefrequencydistributionoftheresultantbendingmoments(SG4.5/4.6)atthebottomsupportisshowninfigure8-158.Themeanvalueofthemeasurementvaluesis89kN-m.Themaximumvaluewasapproximately30kN-mThemeasuredmaximumvalueofthetorsionalmomentduringtheintermittentcondensationis6.2kN-m.8.52373EvaluationoftheMeasurementResultsfortheOuencherArmFigure8-157showsthefrequencydistributionoftheresultantbendingmoments,whichweretransformedfromthemeasuringpointintotheweld(seeSection8.5.2.3.3.3.Themeanvalueofthesebendingmomentsis15.2kN-m.Themaximumvalueis86kN-m.Themeasuredbendingmomentsrepresentstaticequivalentloads.,InSection4.1.2.7andTable4-12,avalueof25.4kN-mwasspecified.fortheequivalentloadfortheresultantbendingmomentintheweldduringintermittentcondensation.Theloadsspecifiedareformulatedforanoccurrencefrequencyof106.REV.li3/798-78 PROPRIETARYInthefatigueanalysis,themechanicalloadsrepresentonlyoneloadcomponent.Anotherpartofthefatigueloadingisproducedbythealternatingthermalloading.Theassumptionmadeinthespecificationwas106temperaturestepsfrom35~Cto133~Candfrom133~Cto35oC.Thelow-frequencyoscillationsofthepipe'sinternalpressuremeasuredatP4.4areusedasabasisforthemeasuredtemperaturealternation.Thesaturated-steamtemperaturesarethencorrelatedwiththosepressures.Thepressureoscillationshaveanoscillationfrequencyofabout0.5Hzandamaximumamplitudeof05baroverpressure=approx.2barabsolutepressure.Thispressureliesbelowthespecifiedvalueof3bar.Themeasuredmaximumpressureof2barcorrespondstoasaturated-steamtemperatureof120~C.AssumingthattheinflowingwaterinSSESisatatemperaturecfatleast35~C,thenthetemperaturestepis85C.Atemperaturestepof98~Cisassumedinthespecification,sothatthereisareserveof13~C.Themeasurementvaluesformingthebasisfortheevaluationandcomparisonwiththespecificationwereobservedonlyduringthephaseofintermittentcondensationwithcoldwaterinthetesttank.Aswiththeboundarypressuresinthetesttank(Section8.4.2),theloadsonthequencherwereconsiderablylowerduringtheintermittentcondensationphasewithwarmwaterthanduringintermittentcondensationwithcoldwater.Themeasuredmaximumbendingmomentduringthiscondensationphasewas(1kN-mrelativetotheweldseam.Inaddition,KMUinplanttestsintheBrunsbuttelnuclearpowerplantshowedthat,forapoolwatertemperatureofapproximately35~andabove,intermittentcondensationloadsonaquencherweresmaller.Thisindicatesthattheregionwh'ereintermittentcondensationloadsofanyconsequencecanbeexpectedislimitedtothatofverylowpooltemperatures(approximately25~C)andverylowsteammassflowsandthatheatingofthepoolasmallamountresultsinareductioninloading8.52.3.7.4EvaluationoftheMeasurementResultsfortheBottomS~uportAnimpulsivelyactingtransverseforceof17.5kNwasspecifiedonthequencherforintermittentcondensation.REVl~3/798-79 PROPRIETARYThedistancefromthemiddleofthequenchertothemeasuringpointforthebendingmomentsatthebottomsupportis0.5m,sothatthespecifiedbendingmomentwithrespecttothebottomsupportis:(175kNx2)x0.5m=17.5KNm(staticequivalentload)Themaximumresultantbendingmomentfromthetestsisapproximately30KN-m.l85.2375EvaluationoftheMeasuredTorsionalMomentsAnimpulsivelyactingtorsionalmomentof19kN-mwasspecifiedfortheintermittentcondensation.Thisstepfunctionyieldsatorsionalmomentof:38kN-masthestaticequivalentloadThespecifiedtorsionalmomentsconservativelyenvelopthemeasuredmaximumvalueof6.2kN-m.852.3.76EvaluationoftheMeasuredMaximumMomentsattheQuencherArmduringIntermittentCondensationAmaximumresultantbendingmomentof665kN-matthequencherarmwasmeasuredintheintermittentcondensationphase,whichresultsinamomentof86kN-mintheweld.Themeasuredmaximumvaluesoftheresultantbendingmomentsatthequencherarmduringintermittentcondensationareontheorderofmagnitudeofthemeasuredmaximumvlauesduringtheventclearingphase(Section8.5.2.3.3.2).Fortheventclearing,atemperaturedifferenceof184~Cwasspecified.Fortheintermittentcondensation,atemperaturedifferenceof98~Cwasspecified.Thetotalstressesloadingthequencherarmarecomposedofmechanicalandthermalstresses.Thethermalstressesaredistinctlylargerthanthemechanicalstresses.Themaximumresultantbendingmomentatthequencherarmforintermittentcondensationexceedthevaluespecifiedfortheventclearingbyabout40%.,However,theassociatedtemperaturejumpisonlyabouthalfaslargeasfortheventclearing.REV-1,3/798-80 PROPRIETARY853VerificationofSuppressionPoolBoundaryLoadSpecificationDuetoSRVActuationInSection4l.3,threepressuretimehistoriesarespecifiedasthebasisforthecontainmentanalysisduetoSRVactuation.ThethreetracesveretakenfromalargenumberofbottompressuretimehistoriesfromvariousKKBinplanttests.TheevaluationofthepressureoscillationmeasurementsintheKarlsteinventclearingtestswillthereforeconcentrateondemonstratingthatthepressuretimehistoriesspecifiedareenveloping.Accordingly,analysisandassessment.oftheindividualmeasuredpressuretimehistoriesisrestrictedtoaminimum.8.5.3.1EvaluationoftheLocalEffectsSeenatPressureTransducerP5.5AsshowninFigures8-10to8-12,thepressuretransducerP5.5ismountedontheconcretewalloppositethemiddleoftheholearrayonthequencherarm.About0.25secondsafterexpulsionofthewatercolumn,P5.5,incomparisonwiththeotherpressuretransducers,exhibitshigh-frequencypositivepressurepeakswhicharenotobservedattheneighboringpressuretransducers.Thiseffectisfromthelocalturbu1ences.Thesehighfrequencypressurepeakshaveasmallenergycontentsothattheirrangeofactionislimitedtotheimmediatevicinityofthepressuretransducer.'hefollowingTableshouldmakethisclearInthisTable,theratioofthemeasuredpressureamplitudesoftheneighboringpressuretransducers{P5.10andP5.4)tothepressuremaximumatP5.5isindicatedforalltestsvhichexhibitedamaximumpressureamplitude>1baratpressuretransducerP5.5.REV1,3/798-81 PROPRIETARYp5.lOPS'+TestP54P5.10P5.5P5.4/55P5.10/P5.5(bar)(bar)(bar)4165.1.710R1.720Rl920Rl1025125R20,60,551,00,450,4li00,730,551,70,450,4.lr01,00,651,730,550,61,00,85081,550,60,450,430,450,580~550,550,550,40,320,40,380,60,52FromthisTablewecanseethatthemeasurementvaluehasdecayedbyhalfatabout1mfromthemeasuringpointP5.5.ThecomparisonmeasurementpointsP5.4andP5.10areintheregionoforiginationoftheairbubbleoscillation,sothatnoattenuationeffectduetodistanceeffectscouldoccuratthatmeasuringpointTherefore,thesharpdecreaseofthepressureamplitudewhichismeasuredneverthelessshowsclearlythatthepressuremeasuredatpressuretransducerP5.5islimitedtoitslocalvicinity.AsfurtherverificationthatthiseffectislimitedtotheareaaroundpressuretransducerP5.5,acomparisonismadebetweenthepowerspectraldensitiesfromP5.5andthebottompressuretransducerP5.2.REV.1,,3/798-82 PROPRIETARYThefollowingtestsvereselected:Test11.1ThistestexhibitedthehighestpowerspectrumatthedominantfrequencyTest4.1.6ThistestexhibitedthehighestpressureamplitudeatP5.5forthelongdischargelineTest20.R1.10ThistestexhibitedthehighestpressureamplitudeatP55fortheshortdischargeline.Thecomparisoncanbesummarizedasfollovs:Atthedominantfrequency,thepowerdensitiesarethesamemagnitudeforthepressureoscillationsatthebottompressuretransducerP5.2andat.pressuretransducerP5.5.Thedifferencesatthehigherfrequenciesissignificant.Fortests4.1.6and20.R1.10thefrequencyspectrumofP5.5exhibitssignificantlyhigherpowerdensitiesathigherfrequenciesthanthecorrespondingfrequencyspectrumatpressuretransducerP52.Thissignificantfactorisnotnotedforthefrequencyspectrumoftest11.1(seeFigures8-159and8-160).Inthattest,thedifferencebetweenthemaximumpressureamplitudesforpressuretransducersP5.5andP5.2was013bar.ThepressureratioisP55/P52=0'8/065=12.Intest4.1.6,thedifferenceinthepowerdensitiesatthehigherfrequenciesisalreadymorestronglyevident(seeFigures8-161and8-162).Inthattest,thedifferencebetweenthemaximumpressureamplitudesforP5.5andP5.2was0.5bar.ThepressureratioisP5.5/P52=1/0.5=2.Thedifferenceinthepowerdensitiesatthehigherfrequenciesisquitestronglypronouncedintests20.Rl.10(seeFigures8-163and8-164).ThedifferenceinthemaximumpressureamplitudesforP5.5andP5.2was1.1barinthattest.ThepressureratioisP5.5/P5.2=1.73/0.63=2.75.Thepressuredifferencesorpressureratiosarenotdiscernibleinthepowerspectraforthedominantfrequencies,butareatthehigherfrequenciesFromthatwecanconcludethatthepressureoscillationwhichwasmeasuredatpressuretransducerP5-5hasapproximatelythesameamplitudeatthedominantfrequencyasthepressureoscillationsvhichweremeasuredelsewhereinthevicinityofthequencher,eg.,atP5.2Inaddition,higherfrequencypressureoscillationcomponentshavingahighamplitudeareoccasionallysuperimposedonthefundamentaloscillationinthepressureoscillationsatP55.Thehigherfrequencycomponents,vhichoccuratpressureREV1,3/798-83 PROPRIETARYtransducerP5.5,decayrapidlyintimeandspace,sothattheeffectofthehighfrequencypressureoscillationsremainslimitedtotheimmediatevicinityofmeasuringlocationP5.5Therefore,asstatedbefore,themeasurementresultsforthedynamicpressuresatP5.5representlocaleventshavingnoglobaleffectonthecontainment.WewillthereforenotconsiderthepositivepressuremeasurementsatP5.5whenverifyingthedesignspecificationfortheoverallcontainmentanalysistheresultsfromthisgageareincludedfortheverificationoftheloadingsonthecolumns.8.5.3.2Verificationofthe~SecifiedPressureA~mlitudesandVerticalPressureProfilesafterVentClearingThemeasuredpeakpressureamplitudesforthe125ventclearingtestsaretabulatedinTables8.9and810.Section8.4.1alsopresentsanumberofFigures(8.27to8.34)whichshowthatthepressureamplitudesmeasuredinthetestshadnosignificantdependenceontheinitialreactorpressure.Therefore,nomodificationtothemeasuredpressureswillbemadetoaccountfordifferencesinthereactorpressurebetweenSSESandtheKarlsteinteststand.Inaddition,asexplainedintheprevioussection,thepositivepressuremeasurementsaP5.5willnotbeconsideredwhenverifyingthedesignspecificationfortheoverallcontainmentanalysis.8532.1OverpressuresThemaximumoverpressureamplitudemeasuredontheboundaryoftheKarlsteintesttankwas1.0barThatpressurewasmeasuredattheconcretewall(p5.4)intest20.R1.10.Amaximumpressureamplitudeofl.2barisspecifiedinsection4.1.3(KKBPressureTraceNo.35withthe1.5multiplier).Themaximumspecifiedoverpressureamplitudeof1.2bar.evelopsthemeasuredmaximumoverpresureamplitudeof10bar.8.5.32.1.1VerticalPressureProfileItcanbeassumedthatthemaximumdynamicpressurevilloccurinaspherewhichsurroundsthequencherandhasapproximatelytheradiusofaquencherarm,(5'-0").Atsomedistancefromit,themaximumvaluewillbeattenuatedinaccordancewithadistancelaw.Foraninfinitewaterspace,the1/Rlawisapplicableforthedecreaseofthepressurewithdistancefromthesource.Thatlawappliesinalldirections,i.e,intheverticaldirectionalso.Thevalidityofthe1/Rlawisbasedontheassumptionofastationary(i.e.,fixedposition)oscillatingbubbleintheinfinitewaterspace.Thatidealcasedoesnotholdfortheclearingofthereliefsystem.Alreadyshortlyaftertheexpulsionoftheair-steammixture,BEV1,3/798-84 PROPRIETARYsmallairparticlesmovetothesurfaceofthepoolbecauseofbuoyancy.Evenmoreimportant,however,isthefactthatthewatersurfaceandthetankboundarysurfacesinfluencethedistancelawandthatthepressureamplitudemustvanishatthewatersurfaceitself.Accordingly,specifiedin6e0(183m)<hatheight,surface.apressureprofileintheverticaldirectionisSection4.1.3.4providingforaconstant'pressureatabovethesuppressionpoolsbottomand,startingata,lineardecreaseofpressureuptothewaterFigure8-165showsthatthemaximumspecifiedpressuredistributionveryconservativelyenvelopsthemeasuredmaximumpressureamplitudes.Theconservativenessbecomesclearlyevidentif,basedonthemeasuredmaximumvalueofwallpressureamplitudeof1baratpressuretransducerP5.4,weassumealineardecreaseofpressurefromthatmeasuringpointtothewatersurface.Thatassumedlinearpressuredecrease(depictedinFigure8-165byadashedline)alsoenvelopsthemaximumpressureamplitudesmeasuredintheverticaldirection.Incomparisonwiththeassumedlinearpressuredecreaseandthespecifiedpressuredistribution,theconservativenessofthespecificationbecomesobvious.~'-5-3-~2..2VerticalPressureProfileIucluainulocalEffactsatP5.5Fortheevaluationoftheunpertubedpressuredistributionintheverticaldireciton,themeasuringpointP5.5wasomitted,eventhoughitliesinadirectlinewiththepressuretransducersP5.4,P5.6andP5.7.BecauseofthelocaleffectforP5.5,aseparateanalysisshallbeperformedhere.ThatanalysisstartswithanestimationoftheverticalzoneofinfluenceassociatedwiththepressurepeakmeasuredatP5.5.Thelateralholesinthequencherarmsextendoverananglerangeof72~oneachside.Theholesaredrilledradially,sothatinfirstapproximationwecanassumeasourceflowoftheemergingfluid.Thehigh-frequencypressurepeakatP5.5occursatamuchlatertimethantheventclearing.Itcanbesupposedthatatthattimethereisasteam-airmixtureflowingoutofthequencher.Thesteam-airjetsemergingfromtheholeshaveahighdegreeofturbulence.Thus,theedgesareverysoonmixedwiththesurroundingwater.Furthermore,theemergingsteamiscondensedimmediatelyandtheexpelledairiscooleddownquickly,sothattheexpelledcompactvolumeisreducedrapidly.Thereforetoestimatetherangeofaction,itisassumedthatthesourceflowactsoverameananglerangeof8=e/2=720/236o.ThetotalrangeofactionisthenREVli3/798-85 PROPRIETARYb=xtan36~x=1.575m{distancefromcenterlineofb=1.14mquencherarmtoconcretewall)Thisrangeofactionof'1.14misdividedintoequalpartsaboveandbelowthemeasuringpointP5.5,sothatweobtainarangeofactionof10.57mrelativetothemeasurementlocationBasedonthisrangeofactinthemeasuredverticalpressuredistributionconsideringthelocaleffectiscomparedwiththespecifiedpressuredistributioninFigure8-166.ThebasepointsofthepressureelevationatP5.5wereplacedonthestraightlineofthelinearpressuredropsymmetricallywithrespecttothequencher'scenterplane.FromFigure8-166itcanbeseenthatthemaximumspecifiedpressuredistributionresultsinalargerresultantforceonthecontainmentboundaryandcolumnsthandoesthemeasuredpressuredistributionincludingconsiderationofthelocaleffectThismeansthattheoverallspecifiedpressuredistributrionintheverticaldirectionalsoenvelopesthelocalpressureelevationatp5.5.8.5322Unde~rressuresThemaximumunderpressureamplitudemeasuredontheboundaryofKarlsteintesttankwas-0.68bar.Thatpressurewasmeasuredatheconcretewall.{P5.10)intest25.R2.Amaximumunderpressureamplitudeof-0.56barisspecifiedinSection4.1.3{KKBPressureTraceNo.76withthe1.5multiplier).The'nextlargestunderpressurerecordedduringtest25.R2was-050bar.Thenextlargestunderpressurerecordedanywhereduringtheventclearingtestswas-058baratP5.2intest25.1.Exceptforthetwomeasurementvaluescalledoutaboveallothermeasuredunderpressureswerehounded.bythemaximumspecifiedvalueof-0.56har.8.5.322.1VerticalPressureProfileFigure8.167showsaplotofthemaximumspecifiedunderpressuredistributionandthemaximummeasuredunderpressurevaluesfortheKarlsteintests.Itcanheseenthat,exceptfortheonevalueatP510fortest2S.R2,themaximumspecifiedpressuredistributionenvelopsthemaximummeasuredpressureamplitudes.REV.1,3/798-86 PROPRIETARYInaddition,forSSES,themostunfavorableboundaryconditioninthiscomparisonisthelowliquidlevelof22ft=6.70minthesuppressionpool.ThehydrostaticpressuredistributionwithrespecttothatliquidlevelisindicatedbyadashedlineinFigure8-167.Thecomparisonofthemeasuredworstunderpressuredistributionviththehydrostaticwaterloadresultingfromtheworstboundaryconditionforthiscomparison(lowestwaterlevelinthesuppressionpool)showsthatthecompressiveforcesfromthewaterloadandthetensileforcesfromtheunderpressuredistribution=maintaintheequilibrium."Thus,theKarlsteintestshave,inaddition,demonstratedthattheblowdownoftheSSESreliefsystemviththequencherdoesnotresultinanyresultanttensileforcesonthesteelliner,evenfortheworstpossiblesuperposition.85.33VerificationofthePressureTimeHistoriesUsedfortheSSESContainmentAnalysisXnordertoverifythatthepressuretimehistoriesusedfortheSSESdynamicanalysisduetoSRVactuationarebounding,thePowerSpectralDensities(PSDs)ofthespecifiedtimehistories(withtheappropriateamplitudeincreaseandfrequencyrangefromSection4.1.3)arecomparedwiththePSD'softheappropriatetimehistoriesrecordedintheKarlsteintesttankandtransposedtotheSSES"suppressionpool.Statementsconcerningtheclearingofparallelquenchersarebasedontheunrealisticandextremelyconservativeassumptionthattheexpelled,airbubblesareequallylargeandoscillateinphase.Aquantificationofthatconservativenessisnotgiven.Mevillfirstdiscussandverifythetheorytobeusedtotransposetheoscillationfrequenciesmeasuredinthetesttanktothesuppressionpool.Then,theappropriatemultipliersforthisfrequencytranspositionwillbeestablished.Adiscussionisalsoprovidedfortransposingthemeasuredpressureamplitudestothesuppressionpool.Finally,theactualverificationispresented.85.33.1Tr~ansositionmethodfortheOscillationFrequencyThetheoreticalbasisforthetranspositionofthepressuretimehistoriesmeasuredintheKarlsteinteststotheSSESsuppressionpoolisprovidedbytheKMUcomputercodesVELPOTandKOVIBlAByusingthetestresultsfromthePPGLquenchertestsinKarlstein,theGKMquenchertests,andthenon-nuclearhottestsintheBrunsbuttelnuclearpowerplant(KKBhottests),weshallfirstconfirmexperimentallythecorrectnessofthetranspositionBEV1,3/798-87 PROPRIETARYtheory.Thatisfollowedbyacalculationofthefrequenciesforthefollowingthreeblowdowncases:(1)Simultaneousblowdownofall16quenchers(2)Simultaneousblowdownofthe6quenchersrelatedtotheautomaticdepressurizationsystem(ADS)(3)BlowdownofoneouterquencherForeachcase,acomparisonofthetheoreticallycalculatedfrequencieswiththefrequenciesmeasuredintheteststand)providesanumber(frequencymultiplier)bywhichafrequencymeasuredintheteststandmustbemultipliedinordertogetthecorrespondingfrequencyintheSSESsuppressionpoolAfactorfortheinfluenceofthesuppressionpooloverpressureisalsodeterminedinthesameway.Thecorrespondingmeasuredpressuretimehistoryistransposedtotheplantbydividingbythisfactor853.3.1.1CalculationofMeasuredOscillation~r~cruencies85.33.1.11PPGLTestsatKarlsteinSinceitwasfoundthatFluid-StructureInteractionintheKarlsteintesttankhasnosignificantinfluenceonthemeasuredpressuretimehistories,itissufficienttocarryouttheanalysisforarigidtank.Thecomparisonofcalculatedandmeasuredoscillationfrequencieswillbebasedon'theassumptionofequalbubblevolumes.ThemeasuredoscillationfrequenciesaretakenfromTables8.9and8.10.Theassociatedbubblevolumeswerecalculatedfromthetestdata,usingtheformula:pp-piieeee'e)]~eo>[P-cP(7T[p-P(T'satpool)]TpipepipePpipePsatCTpoolpipefreepipevolume(ms)pressureinpipe(bar)hydrostaticpressureatthequencherlocation(bar)saturationsteampressure(bar)relativehumidity(s=1at1005)watertemperature(oC)meantemperatureinpipe(oC)TheaveragingofthetemperatureinthepipeisperformedbyusingtheformulaEi1NpipeiwherethepipewasdividedintoNequalsections.ThetemperatureTintheithsectionwasobtainedbyinterpolationbetweenthemeasuredtemperatures.REV.li3/798-88 PROPRIETARYThecomparisonbetweenthemeasuredandcalculatedbubblefrequencyisshowninFigures8-168and8-169inwhichthebubblepulsationfrequencyisplottedversustheequilibriumvolumeatstaticpressure.PorthemeasurementpointsinFigure8-168itwasassumedthatdryairwasinthepipepriortotheteststart,whilewetair(100%humidity)wasassumedinFigure8-169.Ingeneral,goodagreementisfoundbetweenthetheoryandmeasuredfrequency.However,wecannotoverlookthefactthatthemeasuredfrequenciesinfigure8-168)arehigherthanthecalculatedones,especiallyforsmallbubblevolumes.Thismayberelatedtothefactthattheactivevolumeofairunderwaterisactuallysmallerthanthevolumefoundfordryairfromthetestdata.Thisishintedatbythecalculationofthebubblevolumeundertheassumptionof100%humidityinthepipe.Therethemeasurementpointsareclosertothecalculatedcurve(Pigure8.169).Inordertokeeptheuncertaintiesassociatedwithsucheffectsassmallaspossible,onlytestsforwhichtheinitialpipetemperaturewasbelow700Cwerechosenforthecomparisonwiththetheoreticalcase.8.5.3.3.1.1.2GKMMode~luencherTestsAnothersorceusedtoverifythetheoryisofferedbytheGKNquenchertests(Ref1).Sincethepipetemperaturestherewereinthevicinityof300Corbelow,uncertaintiesinthebubble.volumeunderwateraredistinctlysmallerthanintheKarlsteintests.Inaddition,theGKMtestswerealsorunwithbackpressureinthesuppressionchamber,sothatinformationderivedfromthecomputercodesforblowdownofthequencherduringaloss-of-coolantaccidentcanalsobeverified.TheresultscanbefoundinFigures8-.170and8-171.Figure8-170showsthecalculatedandmeasureddependenceofthepulsationfrequencyonthebubblevolumeforvarioussubmergences(2m,4mand6m)withatmosphericpressureinthesuppressionchamber.ThetheoryandmeasuredfrequencyagreeevenbetterherethanintheKarlsteinquenchertests.Thisisprobablyduetothefactthatthebubblevolumesdeterminedfromthemeasurementvalueshaveamuchsmallerscatterduetothelowtemperaturesinthepipe.TheinfluenceofbackpressureonthepulsationfrequencyisshowninFigure8-171.Hereagain,thetheoryisverifiedbythetestdata.85.33113KKBHotTestsInordertodemonstratethecorrectnessofthetheoryforin-plantconditionsalso,calculationswereperformedfortheblowdowntestswithonevalveinthenon-nuclearhottestsintheBrunsbuttelBMRplant(Ref.3).Pigure8-172showstheresults.Theagreementbetweenthecalculatedandmeasured'requencyissimilartothatintheKarlsteintests.Thesameistrueforthescatterrangeofthemeasurementvalues.Sincethepipetemperatureherewasatabout90OC,alargerscatteractuallyREYlt3/798-89 PROPRIETARYwouldhavebeenexpected,butdidnotoccurbecausethepipewascarefullyflushedwithairpriortothebeginningofthesetests.8.5.3.3.1.1.4ConclusionfromtheFrequencyCalculationsThetestcalculationsdescribedaboveshowthatthetheory{VELPOTandKOVIB1Acomputerprograms)describesthemeasuredfrequenciesnotonlyinonespecialcase,butalsoforabroadrangeofgeometriesandbackpressure:(1)Thesizeofthewaterspacevariesfromapproximately7m>(GKN)toapproximately23m~(testtankatKarlstein)toapproximately400m~(suppressionchamberinBrunsbuttelnuclearpowerplant).(2)Thequenchersubmergencerangedfromapproximately2mto6m.(3)Thebubbleequilibrium.volumevariedbetweenapproximately015m~to37m~.(4)Thesuppressionchamberpressurevariedfrom1barto3bar.(5)Thewatertemperatureinthesuppressionpoolvariedbetweenapproximatley16~Cto800C.Thus,thetheorycanbeconsideredverifiedandcanbeusedtotransposethepulsationfrequenciesmeasuredintheKarlsteinteststandtotheSSESsuppressionpool.8.5.33.2Nuit~iliersforConversionoftheBubbleFrequenciesfromtheTestStandtoSSESUsingtheVELPOTandKOVIBlAcomputercodes,thefollowingthreeblowdowncasesareanalyzed:(1)'Simultaneousblowdownofall16quenchers(2)SimultaneousblowdownofthequenchersA,B,G,K,M,PwhichareincludedintheADS{3)Blowdownofonequencher(quencherB)TheresultsareillustratedinFigure8-173whichshowsthepulsationfrequencyasafunctionofbubblevolume(bubbleinhydrostaticequilibrium).Thebehaviorofthefrequencycurveforthe16-quenchercaseintheplantispracticallythesameasfortheteststand(Figure168),therebyconfirmingonceagainthesuitabilityoftheteststandgeometrythatwaschosen.Inthecaseofthe6quenchersintheADScase,thefrequenciesarehigherduetothelargersinglecellcorrespondingtothesmallerREVli3/79~8-90 PROPRIETARYhydrodynamicbubblemass.Theyareevenhigherinthecaseofonequencher.BasedontheresultsshowninFigures8-168and.8-173,asimpleformulacanbegivenforconvertingfromthemeasuredbubblefrequenciestothesefrequenciesfoundintheplantbyasking:Bywhatfactort"multiplier")mustabubblefrequencymeasuredintheteststandbemultipliedtogetacorrespondingfrequencyintheplant'?ThismultiplierisplottedinFigure8-174versusthe(measured)startingfrequency.Thus,wehave:v=f(v).vplantvtest'estinwhichthemuliiplierfforagiveninitialfreguencycanhereadofffromPigure8-173.ThegraphinFigure8-173isapplicableonlyforcaseswithapressureof1barinthesuppressionpoolairspace.However,theblowdownfortheADScaseduringaloss-of-coolantaccidentisassociatedwithasuppressonpooloverpressure.p~>lbarAnadditionalmultiplierfpKK(pKK)isnecessaryforsuchcases,sothatthefrequencyconversionmustbewritteninamoregeneralmanner:V=f(P).S(V).VplantPkk'testtestkkThemultiplierfpKK(pKK')canbetakenfromFigure8-175.Forasuppressionchamberpressureof7bar,ithastheavalueof1,asitmustbe.Themultipliersforthefrequencyalsofixthemultipliersfortheoscillationperiodwhentransposingthepressuretimehistoriesmeasuredintheteststandtotheplant:ttesttlPkkvtestkkI85.33.3TranspositionMethodforthePressureA~mlitudesAsalreadydescribedindetailinSection8.51,theteststandwassodesignedandthe.pressuretransducersweresoarrangedthatthemeasuredpressureamplitudescanbetransposedtotheplantwithoutchangeCorrespondingly,a1:1transpositionismade.Becauseofitsobviousconservativeness,sucha1:1REV1,3/798-91 PROPRIETARYaiplitudetranspositionofferstheadvantagethatmoreexactquantitativeproofsdonothavetobeprovided.Themostsignificantconservativefeaturesarethefolloving:(1)Inblowdowncase'swithseveralquenchers,itisassumedthatallbubblesareequallylargeandoscillateinphase.Deviationsfromthisassumption{suchasactuallyoccurintheplant)resultonlyinloverpressureamplitudes.(2)Blowdowncaseswithlessthan16quenchersareassignedthesamepessureamplitudeasthe16-quenchercase.Inreality,suchcaseshavealoweramplitudeduetothegeometry(largersinglecell).Theconservativenessdescribedin(1)hasnotyetbeenprovenexperimentallyinanyquenchertests,butitisalreadyobviousfromatheoreticalviewpoint,sinceatime-shiftedsuperpositionoftvotemporalmaximaalwaysyieldssmallervaluesthananadditionofthemaximumvalues.Concerningtheconservativenessof(2),thereareanumberqualitativeindicationsfromtheKarlsteinteststhemselves,fromcorrespondingmodelstudiesattheKarlsteinmodelteststand(Ref.1),andfromcalculationswiththeVELPOTandKOVIB1Aprograms.Theinformationobtainedfromallthreeoftheseinvestigationsshallbedescribedinthefollowingsections.Inaddition,wewillalsoexaminewhethertheconservativefeaturesareaffectedbyapossiblebackpressureinthesuppressionpoolairspace.85.33.31PPGL~uencherTestsatKarlsteinIndicationsoftheconservativenessdiscussedin(2)aboveareobtainedfromtheKarlsteintestsonthebasisofFigure8-176vhichillustratesthemeasuredrelationshipbetweenexcitation(relativeamplitude)andpressure-oscillationfrequencyfortheKarlsteintests.Thefrequencyanalysisforeachpressuretimehistoryhasatleasttwomaximaofthepowerdensity.Onepowerdensitymaximumliesatlowfrequenciesandtheotheratsomevhathigherfrequencies.Thereisafactorofapproximatelytwobetweenthetvofreqeuncies.Thefirstpeakofthepowerdensity(lowfrequency).isalwayslargerthanthesecondpeakofthepowerdensity(higherfrequency).Accordingly,thelovfrequencyisalvaysdesignatedasthedominantfrequencyForpressuretransducerP510,thepowerdensitiesofallanalyzedtestsareevaluatedinFigure8-176.DifferentanalysistimesvereselectedfortestshavingdifferentpressureoscillationfrequenciesThetimevassochosenthatREV.1,3/798-92 PROPRIETARYapproximatelythesameoscillationperiodscouldalvaysbeevaluated.Thefollowinganalysistimeswereselectedfortheevaluation:3HzTime:5HzTime:9HzTime:0-1.8seconds0-1.3seconds0-0.6secondsTheareabeneaththefrequencyspectrumvasdeterminedandthenthesquarerootofthatnumericalvaluewastaken.Thatresultsinvalueshavingthedimension>>har>>.Thosenumericalvalueswerenormalizedtothemaximumvalue.Theresultsarethen"relativepressures>>withrespecttothecalculatedmaximumpressurefromthefrequencyspectra.Sincenodominantfrequencieshigherthan6.5HzweremeasuredintheKarlsteintests,thesecondpeakswerealsousedtoevaluatethehigherfrequencies.Hence,thepowerdensitiesofboththedominantfrequencyandthenexthigherfrequencyareevaluatedinFigure8-176.Basedonanempiricalevaluation,it.followsfromFigure8-176thatthepressureoscillationsvithhigherfrequencieshavesmallerenergycontentthanthepressureoscillationswithloverfrequencies.Znaddition,asshowninFigure8-169,thehubblefrequencyincreaseswithdecreasinghubblevolume.Butdecreasingbubblevolumewithconstantsingle-cellsizemeans,accordingtothelawsofsimilarity,thesamethingasincreasingthecellsizewithconstantbubblevolumeTherefore,fromtheKarlsteintestdata,itcanbesaidthatthepressureamplitudesdecreasewithincreasingcellsize.8.5.33.32KM~UuencherTestsintheModelTestStandin~Ka1steinDuringthedevelopmentoftheKWUquencher,testsvereperformed.toexaminetheinfluenceofthesizeofthewaterspace(specifically:freewatersurface)inthemodelteststandinKarlstein(Ref.1).TheresultsareillustratedinFigure8-177,whichvastakenfromRefence1.Itshowsdirectlyhovthebottompressureamplitudesdecreasewithincreasingsizeofthewaterspace(singlecell).8.53.3.33AnalyticalCalculationsTheconservativenessdescribedin(2)aboveisalsoconfirmedfromresultsofcalculationswiththeVELPOTandKOVIB1AREV.1,3/798-93 PROPRIETARYprograms.Asforthefrequencyconversion,appropriatemultiplierscanbedeterminedalsofortheconversionofthepressureamplitudesfromtheteststandtotheplant.Theydependontheinfluenceofthewaterspaceonthestationaryvelocitypotential(spatialpressuredistributionnormalizedtounitsourcestrength)andonthehydrodynamicsourcestrengthassociatedwiththebubbledynamics.Thesourcestrengthitselfisdependentinturnonthepressureinthebubble,.whichisdeterminedbytheinterplay-of.bubblevolumeandairsupplyintothebubble.Sincetheairsupplyvariesaccordingtothedifferentoperatingconditionsduringtheblowdown,onlyaconservativeestimatecanbegivenwithintheframeworkofthepresentinvestigationsT'econversion.fromteststandtotheplantforonequenchermayserveasanexamplehere.Meobtainforthebottompressurebeneaththequencher:P(1quencher)<0.7Pplanttestasuppervalue.8.5.333.4InfluenceofBackgressureonthePressureAmplitudesAsforthebubbleoscillationfrequency,thequestionoftheeffectofbackpressureinthesuppressionpoolairspacemustbeinvestigated.Figure8-178showsthebottompressureamplitudesmeasuredintheGKNmodelquenchertestsforasuppressionpoolairspacepressuresof1and3barAscanbeseen,thepressureamplitudesdonotdependonthesuppressionpoolairspacepressure.8.533.4VerificationofDes~in~SecificationInthetranspositionofthepressureoscillationsmeasuredinKarlsteintotheSSES,theextremelyconservativeassumptionthatthesamepressuretimehistoriesareactingatallquencherssimultaneouslyisused.Differencesinthepressuretimehistoriesoriginatingfromthedifferentdischargelinesareneglected.Therefore,eachmeasuredpressureoscillationintheKarlsteinventclearingtestsisarepresentativecontainment,loadforallloadcases:symmetricalloadcase(simultaneousresponseofall16SRV'sunsymmetricalloadcase(responseofoneorthreeadjacentSRV'sautomaticdepressurizationinloss-of-coolantaccidentREV.1,3/798-94 PROPREETARYAtranspositionofthemeasurementresultstotheplantisperformedfortheseloadcases.TheKarlsteintesttankformsaconservativesinglecell.Therefore,conservativeenvelopingpressureamplitudesweremeasuredinthatteststand.Mhentransposingthepressureoscillationsfromthesinglecelltotheplant,thereisanincreaseofthepressureoscillationfrequenciesasdiscussedinSection8.5.3.3.2.Asstatedpreviously,theincreaseofthepressureoscillationfrequenciesisaccompaniedbyadecreaseoftheamplitudes.Thedecreaseoftheamplitudesisneglectedforthisevaluation,Theamplitudesofthemeasuredpressureoscillationsremainconstantforallfrequencies.Thatisanadditionalconservativefeature,asalreadydiscussedinSection8533385.33.4.1~foe~nencAnalysesofSelectedTestsThepressuretimehistoriesforselectedKarlsteintestsareillustratedinFigures8-41to8-65Thefreqeuncyanalyseswerecarriedout.withtheFourierAnalyzer5451madebyHewlettPackard.Thefrequencyanalysesweregeneratedaspowerspectraldensities.Thefrequenciesatwhichastructureisexcitedintooscillationcanbereadofffromthepowerspectraldensities.FreqeuncyanalyseswereperformedforpressuretransducersP5.2,P5.4,P5.5,andP5.10andforthefollowingtests:4el.l,4.16,12.1,llel,19R27~20Rl1,20eRle10,2lel~21.2,25.R2Pressureoscillationsatboththewallandthebottomareconsideredinthefreqeuncyanalyses.AlsoconsideredwasthefrequencyanalysisforpressuretransducerP5.5,whichshowsthefrlocaleffectThelimitationofthemeasuredfrequenciesofthepressureoscillationswasdeterminativeinselectingtheteststobeanalyzed.Thetestsselectedwerethosewhichexhibitedpressureamplitudes>0.3barbothatlowfrequencyandalsoathigherfrequencies.ThefrequencyspectraforseveralKarlsteintestsareillustratedinFigures8-179to8-182forpressuretransducersP5.10andp5.4.ThefrequencyspectrafortwotestswiththelongdischargelineandloweredwaterlevelareshowninFigure8-179.TheprincipalREV.1,3g798-95 PROPRIETARYfrequencyofthepressureoscillationsisat2-.3Hzforthesetests.TheyarethelowestpressureoscillationfrequenciesthatveremeasuredintheKarlsteintests.Figure8-180showsthedifferenceinthepressureoscillationfrequenciesfromclean-conditionteststoreal-conditionand/ormultiple-actuationtestsforthelonglineThepressureoscillationshaveaprincipalfreqeuncyof3.5Hzintest4.1.1(cleancondition)and5Hzintest4.1.6{realcondition)Fortheshortdischargeline,thefrequencyshiftsfromcleantorealconditionareillustratedinFigure8-181fortests21.1and21.2.Theresultfortheshortlineis:cleancondition:pressureoscillationfrequency5Hzrealcondition:pressureoscillationfrequency6.5HzThefollowingcanbesaidaboutthemeasuredgrin~cialfrequenciesfortheKarlsteintests:Thelowestpressureoscillationfrequencyvasmeasuredinthetestswiththelonglineandadischargelinewaterlevelloweredto2.5mabovethemiddleofthequencher.Itwas2.0-3Hz.2)Fortheclean-conditiontests,pressureoscillationfrequenciesof3.5-4Hzweremeasuredwiththelongdischargeline.3)Fortheclean-conditiontests,pressureoscillationfrequenciesof4.5-5Hzweremeasuredwiththeshortdischarge1ine.4)ThehighestfrequencyfortheKarlsteintestsvasmeasuredforthereal-conditionand/ormultiple-actuationtests.Themeasuredfrequenciesvere6-6.5Hz.Figure8-183shovsfrequencyanalysesfordifferentpressuretransducersforonetest.P5.2-sitsonthebottombeneaththemiddleofaquencherarm.P54-ismountedontheconcretewallattheintersectionofwallandbottom.P5.10-sitsontheconcretewalloppositethecenterpointof'heballofthequencher.Thefrequencyspectraofthepressuretransducersalldisplayapowermaximumatthesamefrequency(3Hz).Therefore,theREV13/798-96 PROPRIETARYlocationofthemeasurementandthestructureofthemountingpositioninthewaterregionoftheKarlsteinteststandhavenoinfluenceonthemeasuredfrequencyofthepressureoscillations.85.3.34e2ShiftingofthePSD'intheTra~nsositionfromtheTestStandtoSSESThecomparisonofthepressuretimehistoriesmeasuredintheKarlsteinquenchertestswiththepressuretimehistoriesspecifiedinSection4.13isaccomplishedbyusingthefrequencypowerspectra.ThefrequencyspectraoftheKKBtracesformingthebasisofthespecificationinSection4.1.3andareillustratedinFigures4-31to4-33Thespecifiedpressureoscillationshavetheirdominantfrequencyintherangeof6.5-8Hz.TocoverthepressureoscillationfrequenciesforSSES,thefollowingrulefortreatmentofthetraceswasgiven:Thethreetracesshouldbetime-expandedbyafactorintherangefrom0.9to1.8.Thepressureamplitudesshouldbemultipliedbyafactorof1.5.Tobeabletomakeacomparisonwiththemeasuredpressureoscillations,itisnecessarythatthefrequencyspectraofthethreetracesbeshiftedinfrequencyandstretchedinamplitude.InthisSection,weillustrateamethodbywhichthoseoperationsonthefrequencyspectracanbeperformed.8.5.3.3.4.2.1Pte5uencfshiftTheamplitudesarepreservedinthefrequencyshift.Toensurethat,theareaunderthepowerspectrummustbeheldconstant.Sincetheanalysistimerangeforthefrequencyanalysisisfinite,itmustbemadecertainthatthecomparisoninvolvesonlyspectrainwhichapproximatelythesamenumberofoscillationperiodswereanalyzedThetracesareexpandedorcompressedbythefactorf<,whilekeepingthezeropointfixedLetusdesignatetheexpandedorcompressedfrequencybyf'ndtheoriginalfrequencybyf.ApowerspectrumcanalwaysbesubdividedapproximatelyintotriangleswhosebaseisthefrequencyandwhosealtitudeisthepowerdensityIntheoriginalspectrum,theareabeneathatriangleis:f-fA21~h2REVli3/798-97 PROPRIETARYForthenewfrequency:fl=fxflfxfp2Therefore,wehaveforthenewarea:AutsinceA'A,h=f~hh~rThepowerdensityoftheshiftedspectrumisinverselyproportionaltothefrequencymultiplier.Inthisdefinition,thefrequencymultipliersaretobetakenfromSection4.1.3.Fromthefactor1.8wegetfV=1/1.8andfromthefactor0.9wegetfV=1/0.9.Ifthefrequencyisreducedtohalf,thepowerdensityisdoubled.85.3.34.2.2~AmlitudeStretchingThefollowingrelationprevailsbetweentheamplitudeofaload-vs.-timefunctionandthepowerdensity:a=k-bf'2k=correctionfactorForthestretchedamplitude,wehavea'fa.Therelationbetweenpowerdensityandamplitudeispreservedbythestretching,sothatthesamecorrectionfactorisalsovalidafterthestretching.Therefore:hk-bf'andthus:hah)h'=f.hh2a8-98 PROPRIETARYThepowerdensityratiointheamplitudestretchingisproportionaltothesquareof'heamplitudemultiplier.8.53.3.43SymmetricalLoadCase~SimultaneousBlowdownofall16SRV's)AlltheKarlsteinclean-conditionandreal-conditiontestsareusedtoevaluatethisloadcase.Themultipleactuationtestsareconsideredasirrelevanttotheplantforthisloadcase.Theoneexceptionisthe10thblovdowntestofanentiremultipleactuationtestwiththeshortdischargeline.Thosetestsarestarted10,minutesaftercompletionofthe9thblowdowntest.Theyarethussubjecttothesameconditionsasthereal-conditiontests.Accordingly,the10thblowdovntestsofamultipleactuationtestwiththeshortdischargelinearetreatedasreal-conditiontests.ThetesttankinKarlsteinrepresentsthesmallestsinglecellwithrespecttothewaterspace.ThatmeansthatthemaximumpossiblepressureamplitudesforSSESweremeasured.AccordingtoSection8.5.3.2,themeasuredpressureamplitudesarecoveredbythespecificationForthisloadcase,themeasuredfrequenciesofthepressureoscillationscanalsobetransposeddirectlyfromtheKarlsteinteststandtoSSES(seeSection8.53.2).Thus,allthepressuretimehistorycanbetransposeddirectlyfromtheteststandtoSSES.Inordertoshowthatthemeasuredtimehistoriesarealsoenvelopedbythespecification,thefrequencyspectraofthemeasuredpressureoscillationsarecomparedwiththefrequencyspectraofthespecifiedtraces.Sincethemeasuredfrequenciesdifferfrom.thefrequenciesofthespecifiedtraces,thespectramustbetreatedbythemethodillustratedinSection8.5.3.3.42andbroughtintocoincidenceatthedominantfrequency.ThepressureoscillationsmeasuredatpressuretransducerP5.2areusedforthiscomparison,since,thepressuretransducerP5.2exhibitsthehighestpowerspectrumofallthepressuretransducersthatareuseablefortheoverallloadingofthecontainment(P5.5isnotconsidered-seeSection8.5.3.1).PressuretransducerP5.2ismountedonthebottomofthetesttank,directlybeneathaquencherarm.ThatpositionisalsopresentinSSES.Therefore,thispressuretransducermeasurespressureoscillationshavingthegreatestrelevancetoSSES.Purthermore,thespecifiedtracesarealsoresultsofameasurementmadewitha.bottompressuretransducerwhoselocationwassimilartothatofP5.2.REV1,3/798-99 PROPRIETARYThecomparisonofthefrequencypowerspectraisshowninFigures8-184to8-188WeseethatthefrequencyspectraoftheKKBtraces,whichwerefrequency-shiftedandamplitude-stretchedasdescribedinSection8.5.3.3.4.2envelopthefrequencyspectraofthemeasuredpressureoscillations.Therefore,itcanbestatedthat:a)theKarlsteinmeasurementresultsareconservativefortheloadcaseofsimultaneousclearingofall16quenchers{single-celleffect);'Ib)forthisloadcase,thepressureoscillationsareenvelopedbythespecificationwithrespecttotheiramplitude,theirfrequencypowerspectra,'andtheirspatialdistribution.8.5.3.3.44UnsymmetricalLoadCaseslowdownViaOneSRV)Forthisloadcase,alldeterminativeparameters,exceptforthe-,watersurfacearea,weresimulatedintheKarlsteinteststandaccordingtotheiractualvaluesforSSES.Fortheloadcaseofventclearingwithonequencher,alargerwatersurfaceareaisavailabletothequencherinSSESthaninthetestintheKarlsteinteststand.Accordingly,thepressureoscillationfrequenciesareraisedandthepressureamplitudesarelowered.Inthisverification,weconservativelymakenoallowancefortheamplitudedecreasewithincreasingwatersurfacearea.ThefrequenciescalculatedaccordingtoSection8.5.3.3.2fortheloadcaseofblowdownviaoneSRVarecompiledinthefollowingtable:Frequencyofthepressureoscillations(Hz)MeasuredFrequencymultiplierPlantSpecifiedfrequency.bandCLEANCONDITION3.5-41.54-1.485.4-5.9REALCONDITIONS5CLEANCONDITION1.421.427.17.13.75-8.9480RWREALCONDITIONS6.51.378.9REV.1,3/798-100 PROPRIETARYThefrequenciestransposedtotheplantareallenvelopedbythespecifiedfrequencyband.Fortheloadcaseofventclearingofonequencher,themultipleactuationtestsmustalsobeconsidered(theywereincludedunder"realconditions"intheTableabove)Fortheloadcaseofsimultaneousblowdownof16quenchers,itwasshownthatthemeasuredpowerspectraareenvelopedbythespecifiedpowerspectra.Thatstatementappliesforallfrequencyranges.Iftwopowerspectraarebroughtintocoincidenceatonefrequencyandifbothspectraaresubjectedtothesamefrequencyshift,thenthereisnochangeintherelationofthetwospectratoeachother.Therefore,thepowerspectraoftheclean-conditionandreal-conditiontestsarealsocoveredbythespecificationintheloadcaseofventclearingofonequencher,since,asstatedabove,thetransposedfrequenciesfromthetestareallenvelopedbythespecificationfrequencyrange.Forthemultipleactuationtests,test4.1.6isconsideredtobeenvelopingforthelongdischargeline,sinceitprovidedthehighestpressureamplitudes.Fortheshortdischargeline,test20.R1.10(whichformallycanbeclassifiedasamultipleactuationtest)isconsideredtobeenvelopingforthesamereason..Classifiedasareal-conditiontest,itwasshownintheprecedingSectionthatthespecifiedtracesenvelopthepressuretimehistoriesforthattest.InFigure8-189itisshownthatthepowerspectrumoftest4.16isalsoenvelopedbythespecifiedKKBtraces.EvenundertheveryconservativeassumptionthatthepressureamplitudesmeasuredinKarlsteincanbetransferredwithoutchangefortheloadcaseofvent,clearingofonequencher,thepressuretimehistoriesareenvelopedbythespecifiedtraces.85.3.34.5Un~smmetricalLoadCaseslowdownviaThreeA~d'acentSRV~sgThisloadcaseisboundedbytheloadcasesofsimultaneousventclearingof16quenchersandventclearingofonequencher.8.5.3.3.4-6AutomaticDeDzessuzizatio~asstem~A~DSLoadCaseInthissectionwediscusstheloadcasethatconsidersthefiringofthesixquenchersassociatedwiththeADSunderLOCAconditions.REVli3j'798-101 PROPRIETARYAsshowninFigure8-190,thefollowingconditionsprevailinthesuppressionchamberwhentheautomaticdepressurizationsystemisactuatedduringIBA:Absolutepressureinthewetwellairspace,approximatelyPressuredifference.betweendrywell'andsuppressionchamber2.55bar0.42barTheKarlsteintestswithloweredwaterlevelinthedischargelineareusedtoverifytheADScase.Thesetestsareusedastheycorrectlysimulatethedischargelineasitwouldbewithapositivepressuredifferentialofapproximately0.42barinthedrywell.Thispositivepressuredifferentialwouldresultintheloweringofthewaterlevelinthedischargelinetotheelevationofthebottomofthedowncomersaswassimulatedfortests10.3,ll1,12.1and13.1.Of'thosetests,thetest11.1(envelopinginamplitudeandpowerdensity)isusedasthebasisfortheverification.Theamplitude-reducinginfluenceofthelargerwatersurfaceareaassignedtotheindividualquencherintheADScaseisconservativelyneglected.Also,sinceearlierKMUtestsprovedthatthebackpressureinthesuppressionchamberhasnoinfluenceonthepressureamplitudes,themeasuredpressureamplitudesaretakenunalteredfromthecorrespondingKarlsteintests,inwhichthemeasurementsweremadeatatmosphericpressure.Thepredominantfrequencyintest11.1isat3Hz.AccordingtoSection8.5.3.3.2,Figures8-174and8-175,thefollowingfrequencymultipliersareobtainedfortheADScasefortranspositionofthepressureoscillationsfromtest11.1totheplant:InfluenceofthelargerwatersurfaceareaInfluenceofthe2.55barbackpressureTotalfrequencyfactorDomi.nantfrequency135141957HzNote:Themeasuredlowestdominantpressureoscillationfrequencywasmeasuredintests12.1and13.1,whichfallintothesamecategoryastestll1.Miththetotalmultiplier1.9,thefrequenciesareraisedto3.8Hzandthusliewithinthespecifiedfrequencyband(seeSection8.5.3.3.5).ThedominantfrequencyiswithinthespecifiedfrequencybandREVlg3/798-102 PROPRIETARYThecomparisonbetweenthepreparedtracefrompressuretransducerP5.2fortest11.1andthespecificationisshowninFigure8-191.Asfortheotherloadcases,thecomparisonismadeinthepowerspectra'ofthepressuretimehistories.Thespectrumoftest11.1wasshiftedfromthedominant.frequencyof3Hztothedominantfreqeuncyof5.7Hzwhilepreservingthearea(amplitude).TheKKBtraceoftest76wasshiftedfrom8Hzto5.7Hzwhilepreservingthearea,andthenstretchedbyafactorofl.5inamplitude.Figure8-191showsthatthetracefromthespecification,treatedinthismanner,envelopsthetraceofKarlsteintest11.1transformerdtotheADScasesincethetotalenergyrepresentedbytheareaunderthepowerspectrumcurvefromthespecificationisgreaterthanthatfromtheKarlsteintestll.185.33.47SummaryIthasbeendemonstratedthatthe,frequencypowerspectrumofthepressureoscillationsinthesuppressionchamberareenvelopedbythefrequencypowerspectrumspecifiedinSection4.1.3forallloadcases.Thus,thedesignspecificationprovidesenvelopingloadsalsoforthedynamicexcitationoftheSSFScontainmentbyventclearingofthereliefsystemwiththequencher.8.5.33.5EvaluationoftheNeasuredPressureOscillationsDuringCondensationAsdiscussedinSection8.4.2,threeregimescanbedistinguisedinthecondensationprocess:a)Thequencherisclearedcontinually.b)Thequencherisnotclearedcontinually.c)Onlytheslidingjointiscleared,andthesteamcondensesinthedischargeline.8.533.51TheguencherisClearedContinuallyThesteamiscondensedcontinuallyinthewaterpooloutsidethe-quencher.Calmcondensationprevailsforcoldwaterandalsoforhotwaterintheblowdowntank(seeFigures8-78and8-79)Themeasured.maximumpressureamplitudeiss0.13bar.Thiscondensationphasewasmeasuredforreactorpressuresuptoabout4bar.Thefrequenciesofthepressureoscillationsare70-120Hzforacoldpooland20-45Hzforahotpool.REV1,3/798-103 PROPRIETARY8.5.3.3.52TheQuencherisnotClearedContinu~allThiscondensatonphasebeginsvhenthecondensationrateoutsidethequencheris.greaterthanthesteammassflovthroughtheline.Thepressureinthequencherdropsbelovthehydrostaticpressureofthesurroundingvater.Thewaterpenetratesintothequencher.Thecondensationsurfaceareaistherebydecreasedandsoisthecondensationrate.Theresultisapressureriseinthedischargeline,sothatthewaterthathasflovedinisexpelledagain.Theinflowofvaterfromthesuppressionchamberintothequencherandthesubsequentbrakingandre-expulsionofthewaterisanonstationaryprocessvhichoccursperiodically.Forthatreason,thiscondensationphaseisalsocalledintermittentcondensation.Thephenomenonofintermittentcondensatonisdependentonthewatertemperature.Forcoldvaterthereisahigherrateofcondensationoutsidethequencher,resultinginalargergenerationofnegativepressureinsidethequencherandthereforeamorevigorousflowofwaterintothequencher.Foracoldwaterpool,theprofileofthedynamicpressuresissimilartotheprofilevhichisfamiliarfromthechuggingphaseofthecondensationattheventpipes;seeFigure8-76.Forheatedvaterinthesuppressionchamber,thecondensationrateoutsidethequencherissmaller,sothattheentireprocesstakesontheformofalow-frequencypressureoscillation(SeeFigure8-80)Thetestsin"Karlsteinyieldedasmaximummeasurementresultforthedynamicpressure:+0.28,-0.18bar,foracoldpool.Thetimebetweentwoeventsisabout1.0second.Foraheatedpool,themeasuredmaximumamplitudeis+0.12,-0.07,bar.8.533.5.3CondensationintheDischargeLineandThrutheSlidi~nJointIfthesteamflovdecreasesfurther,aconditionisfinallyreachedinwhichthequencherisnolongercleared,butratherremainscontinuallyfilledwithvater.Thenthereissteady-statecondensationofsteaminsidethedischargelineThiscondensationphaseproceedsverycalmlyandbeginsatreactorpressuresbelow2bar.Inthiscondensationphase,maximumdynamicpressuresof+0.08,-0.04barweremeasuredinthewaterpoolduringtheKarlsteintests.REVl,3/798-t04 PROPRIETARY853.3.54Tr~ansositionoftheMeasurementResultstoSSESInregardtosteamcondensation,theconditionsoftheKarlsteinteststandaredirecltytransposabletotheconditionsofSSES.Onthewhole,thepressureamplitudesduringcondensationaresmallcomparedtot'hoseduringventclearingandthereforearecoveredbythelatter.85.4PoolMixinqDurincnSRVActuationandThermalPerformanceofthe~uencher85.41IntroductionWhenanSRVresponds,steamiscondensedinthewaterofthesuppressionpoolviaaquencherAsthishappens,thewatermustabsorbtheheatofvaporizationofthesteam,andsoitisheated.Whenthereisalong-lastingdischargeofsteamviaaquencher,allthewaterinthesuppressionchambershouldparticpateintheheating,soastolimitthelocalheatinginthevicinityofthedischargingquencherInordertoobtaingoodmixingofthehotterandcolderwaterinthepool,allquenchersarepositionedatasmalldistancefromthebottom(3~6"=1.07m)(seeFigure8-192)).Thewaterheatednearaquencherisspecificallylighterthanthecolderwaterlyingaboveit.Therefore,thewarmerwaterwillriseandmixwiththecolderwater.Toobtainanadditionalmixingeffect,theholeoccupancyofthequenchersweremadeslightlyunsymmetrical(approximately8%).Mhereasthequencherarmshavethesameholeoccupanciesonthesides,onlyonearmofeachquencherhasholesontheendcap.Inthatway,aunilateralthrustcanbeexertedonthewaterinthesuppressionpool.Inthetopviewofthequencherarrangement(Figure8-193),weseethatthequenchersarearrangedintwograduatedcircles.Alongtheinnergraduatedcircle,thequencherarmsallpointinthecircumferentialdirection,andtheendcapwithholesallpointinthesamecircumferentialdirection.Ontheoutergraduatedcircle,thecolumnswouldpracticallypreventathrusteffectifthequencherswerearrangedinthesamemanner.Therefore,thequenchersweredirectedmoreradially,butturnedbyanangleofgf=300inthecircumferentialdirectionfromtheradii.Inthisway,50%ofthethrusttillactsinthecircumferentialdirection(equidirectionallywiththethrustofthequenchersontheinnergraduatedcircle).ItshouldbenotedthatthisnewarrangementsupersedestheoriginalarrangementshowninFigure1-4.Inthefollowing,weshallestimatetheaccelerationofthewaterpoolforthecaseinwhichonequencherontheoutergraduatedREVli3/798-105 PROPRIETARYcircleisoperatedforalongperiodoftimeatareactorpressureof70bar{valvefailureinopenposition).Thenweshallpresentsomemeasurementresultsfromatestwitha4-armquencherintheBrunsbuttelnuclearpowerplantandsomeinformationfromtheGEMNodelquenchertestsrelatedtosteamcondensationwithaquencher.85.42EquationofNotionoftheRotationPoolZtisassumedthatthewaterflowintherotatingpoolcanbeconsideredasastraight-linechannelflowduetothesmallcurvatureofthegraduatedcircleandthelowcircumferentialvelocity.Ifweplacetheoriginofthecoordinatesystematthecenterofthedischargingquencher,thentheequationofmotionoftherotatingpoolreads:5'.2mx+c2xFffmWcWeffThismassofwatertobeacceleratedinthesuppressionchambersumofallflowresistanceseffectivedrivingforcedifferentialequationhasthegeneralform:x+ax=bSubstitutingx=u,thedifferentialequationtakestheform:u+au~=bThisdifferentialequationisaspecialformoftheRiccatidifferentialequationThegeneralsolutionofthedifferentialequationsreadsR'ef.53:na.b+bTanha.b(t-K)u(t(n)=-.ga.b+a.q'Tanhga'b(t-c)Theinitialconditionfort=0reads:0n/ab+bTanh/ab-g)/ab=an.Tanh/ab(-g)REVl,3/798-106 PROPRIETARYThisconditionalequationissatisfiedonlyif6andn=0.Theinitialconditionthenleadstothesolution:b.Tanha.bu(t)=ga.bSinceu(t)=X(t),theequationforthevelocityoftherotatingpoolreads:x(t)=b)a.bTaab)a.btForthedistancecovered,wehave:x(t)=pJ'(v)dvThesolutionreads:X(t)=-ln[coshia.b.t(085.03'eterminationofthePlowResistancesThefollowingresistancesareconsidered:a)Wallresistanceofthechannelb)Resistanceforflowaroundthedischargelineswithquenchersandbottomsupportc)Resistanceforflowaroundtheventpipesd)ResistanceforflowaroundthecolumnsThechannelhasthefollowingdimensions:REV.li3/798-107 PHOPRIETARYThehydraulicdiameterofthechannelis:73(26~822-8~84Rect226.822-8.84(27.3)+,22.8cnFortheReynoldsnumber,wehave:Re!W.RRAccordingtoReference36,thekinematicviscosityforwaterat40oCisv=0.65lx10-~m~/s.Xfweassumeavelocityof10-2m/ssoas'tocoverthestart-upphasealso,weget:R+10x2.8.43x1024.651x10TheSSESsuppessionpoolislinedwithasteellinerwhichcannotbeconsideredhydraulicallysmooth.Forsuchlargesteelstructuresitmustbeassumedthattheindividualplatesarenotjoinedtogetherwiththeiredgesparallel,sothattheflowresistanceisincreasedbyprojectingedges.Wethereforeconservativelyassumeanabsoluteroughnessofk=2mm.Thenwehave:Kdh2.8x10-47.1x10Thiscorrespondstoafrictioncoefficientof>=0.022.Theresistancecoefficientisthen:~.1mWh26.844+8.84m'2~7fRR56mr-.022w'28)Cylindricalbodiesareimmersedintothewaterofthesuppressionchamber.Theyarethedischarge.lineswithguenchers,theventpipes,andthesteelcolumns.REV1,3/798-108 PROPRIETARYOutsidediametermSubmergencemQuantityDischargelinesVentpipesSteelcolumns03240.61106733357-3168712Fortheindividualstructuralcomponents,wethenhavethefollowingReynoldsnumber:v=0.01m/s(seeabove)WRe=-dmFortheroughness,weassumek=0.2mm.Then,accordingtoReference39:ReynoldsnumberSubmergencedDischargelinewithquenchervi.thbottomsupportVentpipe5x10~9.4x10~6.17x10-i6.28x1022.555073073Columnl63x10'9x10-+6.9073Theresistanceforceist,hen:p92Thesurfaceareaonwhichthewallresistanceactsis:24Furthermore:c6.16x.44+.73x50+.73x177.8+.73x93WAA16x0.324x9.650m.2c=238mA87x.61x3.35177.8.m2AS-12xl.06x7.3~93m2Sincethewaterregionofthesuppressionchamberalsocontainsafewstructuralcomponentswhichverenotconsideredhere,anadditionalallowanceshallbemade.Mechoose:2c300mWREV.13/798-109 PROPRIETARY8.5.4.4DeterminationoftheForceNovi~nthePoolForcesonthewatermassinthesuppressionpoolareproducedbythrustfromtheboreholesononeoftheendcapswhicharepresentoneachofthequenchers.Thesmallestthrustforceisproducedbythequenchersalongtheoutergraduatedcircle,sincetheydonothavetheirthrustboreholesarrangedinthecircumferentialdirection.Thequenchersalongtheoutergraduatedcircleareturnedbyanangle4=30orelativetotheradialdirection.g=50'DV=differencebetweenpressureinthequencherandambientpressureThethrustforceresultsfromtheimpulseoftheoutflowingsteam.F~APxA-+PxMDxA~UeffectiveoutletareaofquencherPD=densityoftheoutflowingsteamW=velocityoftheoutflowingsteamDAsaneffectiveoutletareaofaquencherendcap,thereisavailable:A~ga<xADUgeomeDC0.8(Section8.5.2.3)+0geom(~l2(4)6.9x10m-32ADOgeom5'~2x10m-32Aconstantreactorpressureof70barischosenfortheestimateof.theeffectivenessoftherotatingpool.AccordingtoReference37,themassflowthroughthereliefvalveatareactorpressureof70baris:REVlr3/798-110 PROPRIETARYm=illkg/sTheresultingstagnationpressureinthequencheris:p=llbarandthesteampressureinthequencherholesis:pD=64barTherefore,PD"=34kg/m>WD462m/sTheforceactinginthecircumferentialdirectionisthen:FeffFsinFeffTherefore:(AP+P+Wj))A"xsinQwithQ=30'DUFeff2~O'N+lo5KN3~5KN85.45WorkingEquationsfortheRotation'PoolofSSESTheequationofmotionfortherotatingpoolreads:5'.2mx+c2effwwThisdifferentialequationwassolvedingeneralforminSection85.4-2.Todetermineconsiderthehave:themassofwaterwhichistobemoved,wemustinternalstructureswhichreducethewatermass.MeI,4(26.822)-(8.84))x.73--x(.324)x7.3x164--x(61)x3.35x87--x(1.06)x12x7.3]43.5x10KgForthetotalresistancecoefficientwehaveaccordingtoSection8543:C<=300mandfortheeffectivelyactingforcewehaveaccordingtoSection8544:F=35KNeffREVl,3/798-111 PROPRIETARYTherefore,theeguationofmotionreads:3.5x10xX+1.5x10xX6-52or3.5x10Therefore,for:a=4.3x104b=9.9x102X+aX=bVab6.55x10tTheequationforthevelocityoftherotatingpoolreads:-1-3X(t)=,1.52x10Yanh6.55x10tTheequationforthedisplacementreads:X(t)=23.21nicosh6.55x10tiTheresultsareillustratedinFigures8-194and8-195.854.6EstimateoftheHeati~noftheSuppressionChamberMaterThelocalheatingofthesuppressionchambervaterresultsfromthebalanceoftheheatbroughtinbythecondensingsteamandtheheatdissipatedbytheflowingwater.Astimepasses,hovever,thepoolissetintomotionbytheimpulseof.theinflowingsteamandreachesavelocitysuchthatmostoftheheatbroughtinisdistributedoveralargervolumeofwaterthantheassumedlocalvolume.,Thedifferencebetweenthelocalandmeanwatertemperaturedecreases.85.47ExperimentalProofs854.71NodelTankTestsThrustmeasurementsonasteamjetveremadeintheKarlsteinmodeltankintheSpringof1973(Ref40).Thetestset-upisillustratedinFigure8-196Thesteampipeisconnectedby.aspringtothesidewallofthemodeltank.Theexcursionofthespringwiththesteampipeismeasuredbyadisplacementtransducer.Themeasurementsystemvascalibratedbydeterminingtheexcursionofthesteampipeforadefinedforce.Thesteamoutletopeninghadadiameterof10mm.Themassflowdensityvas600to630kg/m<s.Themeasuredreactionforceswere20-28N.REV1,3/798-112 PROPRIETARYAshortcalculationyields:OutletareaRestpressurebeforetheoutletopeningPressureaftertheoutletopeningSteamdensity(at2.6bar)=7.854x10-~m~4.5bar2.6bar.=1.44kg/m~Theresultingoutletvelocityis:W=gK-2.6x10~~1.135W-452.7m/sandthethrustforceis:F=(PW+hP)AffAff=08xAegeomF~(1.44x(452.72)+1.6x10)x0.8x7.854x10F~284NThemeasuredvaluesareloverthanthecalculatedvalues.Themeasurementshaveprovedclearlythattheimpulseoftheemergingsteamjetbecomesactiveasathrustandthat,vithrespecttothevelocitybuildupoftherotatingpool(andthusforthemaximumlocalheating),itisconservativelyboundedbythecalculatedvalues.85.47.2KKBTestDuringtheNuclearCommissioningThepressurereliefsystemwastestedduringthecommissioningphaseoftheBrunsbuttelnuclearpoverplant.Inonesuchtest,areliefvalvewasheldopenforatimeofabout270seconds.Thesuppressionchambercoolingsystemvasswitchedonduringthetest.Waterwasdrawnoffinthelowerpartofthepool,cooled,andsprayedfrompipesprovidedwithholesandlocatedunderthetopofthesuppressionchamber.12measuringpointsaremountedinthewaterregionofthesuppressionchamber.Theyarearrangedatthreedifferentelevations(14m,16.5m,18.2m)andatfourdifferentcircumferentialpositions(5o,75o,195o,245o).Thewaterlevelisataheightof18.89m.Figure8-197showsathreedimensionalspatialrepresentationofthemeasuredtemperaturefieldinthevaterjustbeforeteststart(curve1)andat228secondsafterteststart(curve2).InFigure8-197,theverticalpositionofthetransducerisrepresentedontheordinateandthecircumferentialpositionontheabscissaThetemperatureaxispointstotherear.Theheatingofthepoolisindicatedasthedifferenceofcurves2and1atthreeelevationpositions.Themeanwatertemperaturewasapproximately32.3oCbeforethetestandapproximately42.8oCREV.1,3/798-113 PROPRIETARYat228slater..Themaximummeasuredtemperaturewas500C,sothatthemaximumdeviationfromthemeanwas7.20C.Thedischargingquencherwaslocatedat285'tanelevationof14915mandacceleratedthewatertowardtheleftintheFigure.Correspondingly,the.watertemperatureishigheraboveandtotheleftofthequencher.Fromthatwecanseetheeffectivenessofthequencher'sarrangementnearthebottomandoftheunsymmetricalholearrangementwithre.,pecttouniformutilizationoftheheatsinkofthewaterpool.85473GKNHalfScaleQuencherCondensationTestAseriesofintermediatescale(1:2)condensationtestswereperformedintheGKMteststandtodemonstratethehightemperatureperformanceoftheguenchers(Ref.27).Condensationtestswererunonsevendifferentversionsofthequencherdevice.Thelastthreeversionshad10-mmdiameterholeson'hequencherarmsThespacingoftheholecenterlineswas1.5diameterscircumferentiallyand5.0diametersaxially.ThisholepatternisalsoadoptedintheactualSSESquencherdesign.Thesetestswererunatawatertemperaturerangingfrom13oCto100oC(56oF-2120F)andasteammassflux(withrespecttotheholearea)rangeof8to495kg/m~(1.6to101ibm/ft~s).Matertemperaturesashighas1070C(225~F)weremeasuredatcertainlocationsinthesetests.85.48SummaryTheKarlstein'uenchertestsandpreviousGKMhalfscalequenchertestsshowclearlythatsmoothsteamcondensationcanbeachievedatelevatedtemperatureswhichapproachthe.localsaturationlimit.XnadditionthecalculationsandKKB,inplanttestsprovideinformationwhichsuggestthatpoolmixingisenhancedbysteamdischargethroughtheholesintheendcapsofthequencher."85.5VerificationofSubmergedStructuresLoad~SecificationDueToSQVActuationSection4.1.3.7givesthedesignspecificationfortheloadsonsubmergedstructuresduetoSRVactuation.Thebasisforthespecificationisthethreepressuretimehistoriesusedforthecontainmentanalysisbutinsteadofaconstantamplitudemultiplierof1.5variousmultipliers,relatedtothecrossectionalareaoftheobject,areused.(seeTable4-15).TheloadingonthecolumnsincludingthelocalizedefectatP5.5hasbeendiscussedinSection8.5.3.2.1.2REV1,3/798-114 PROPREETARYInadditiontheeffectsofairbubbleoscillationloadsonthequenchershavebeendiscussedinSection8.5.23.6.ThefollowingsectionwilldiscusstheloadingsontheventpipesasmeasuredintheKarlsteintesttankandprovideadescriptionoftheinfluencefortheexpelledwaterduri'ngventclearing.8.5.51LoadsontheVentP~ie85.511MeasurementoftheLoadsXnordertodeterminetheloadingoftheventpipenearaquencher,aventpipehavingthesameoutsidediameterandwallthicknessasthatinSSESwasinstalledintheKarlsteinteststandandsupportedbytypicalbracing.(seeFigure8-10).Underneaththebracing,bendingstrainsweremeasuredintwomutuallyperpendicularplanesbymeansofstraingauges(SGS1andSGS2)(seeFigures8-11and8-12).Thestraingaugesweremountedabout100mmbelowthebracing.Theoutsidediameteroftheventpipeis:D=0609mandtheinsidediameteris:D;=0589mThus,thecross-sectionalareais:A~0.0188m2andthemomentofresistanceis:4Dl320Wehave:-332.77x10mxW~'M~GxExWTherefore:M~2.77x10'0'-3,11'ndhence;M~057cREV.1,3/798-115 PROPRIETARYIfweinsertcinmicrometerspermeterintothisequation,weobtainthebendingmomentinkN-mThebendingmomentscalculatedinthismannerarestaticequivalentloads.5-5-5.1.2MeasnnedBendi~nsaeentsFigures8-198to8-200showthedependenceofthemeasuredresultantbendingmomentsonthereactorpressure,ventclearingpressure,and.pressureoscillationamplitudethatweremeasuredneartheventpipeontheconcretewall.Onlythetestswithcleanconditionswereusedfortheplotofthemeasuredbendingmomentsversusreactorpressure,whereasalltestsinthereactorpressurerangeof60-81barwereusedfortheplotsofthebendingmomentversusventclearingpressureandpressureoscillationamplitude.Themeasurementsofthebendingstrainsattheventpipewereperformedonlyforthetestswiththelongdischargeline.Themeasuredmaximumbendingmomentwas14.6kN-mata74barreactorpressureanda13.8barventclearingpressure.85513Extr~aolationoftheMeasurementResultsand~ComarisonwiththeSpecifiedValueIfthemeasurementvaluesareextrapolatedtotheextremeconditionsintheplantonthebasisofFigures8-198and8-199,wegetthefollowingextrapolatedmaximumvalues:16.5kN-mwithrespecttoan88barreactorpressure,19.0kN-mwithrespecttotheventclearingpressureof16.5barforthelongdischargepipe,asextrapolatedinSection8.4fortheextremeboundaryconditionsintheplant.Inthespecification,amaximumpressuredifferenceof0.75x0.8=0.6barwasspecifiedfortheventpipewiththedistributionillustratedinFigure4-24.ThepressuredistributionfortheventpipeinstalledintheKarlsteinteststandisshowninFigure8-201Thefollowingrelationappliesforthepressureattheendofaventpipe:~dPdP7.3-1.837.3-3.65hPssOe4barREVli3/798-116 PROPRIETARYAttheclampingpointoftheventstrut,wehave:AP07.3-1.837.3-6.3AP=01barThepressuredistributionfromtheendoftheventpipetotheclampingpointofthevent-pipestrutistrapezoidal.L=0.1x-'-'-x2.652.65(0.4-0.1)2S'(2)3Theleverarmoftheactingforcewithrespecttotheclampingpointis:0.1+.0.42S1'59Forthebendingmomentattheclampingpointweget:M5=(~2~'2.65x0.6x1.59)10SP~SP63kNmRelativetothestraingauges,wehave:MBSP57kNmTheextrapolatedmaximummomentwas19kN-m.Itisthusdemonstratedthatthespecificationenvelopsthemeasurementvaluesandtheirextrapolation.Theproofthatthespecificationenvelopsthemeasurementvaluesandtheirextrapolationisbasedonapurelystaticanalysis.Suchananalysisispermissiblebecausetheexcitingpressureoscillationshaveafrequencyof4-6Hz.However,thestraingaugesindicateanaturaloscillationfrequencyof17-20HzfortheventpipewhichisveryclosetothenaturalfrequencyoftheventinSSES(19Hz)(seeFigure8-202).Hence,itcanbeassumedthatthedynamicloadfactorisclosetoone.8552InfluenceofExpelledWaterDuringVentClearinc[AreviewofthehighspeedfilmsandpressuretracesatP5.5fromtheKarlsteintestsshowsnegligableinfluenceoftheexpelledwateratthisgage.Inadditionthetotalpenetrationoftheexpelledvaterappearstobeapproximately3feetfora70barinitialsystempressure.Therefore,noadditionalloading,otherthanthatalreadyincludedinthepressuretracesvillheconsidered.REV1,3/798-117 PROPRIETARY(AtimecorrelationofahighspeedfilmtopressuretraceatP5.5willbesuppliedlater.}85.53SummaryTheloadsmeasuredonthedummyventpipearestaticequivalentloads,butloadswhichareasumofindividualcomponents.Inthespecification,thetransverseloadsoninternalstructuresoriginatingfromtheblowdownofthereliefsystemare.formulatedasdifferentialpressuresacrosstheinternalstructures.ThedifferentialpressureshavethesamepressuretimehistoryasthedynamicpressuresinthewaterregionofthesuppressionchamberThisformulationofthetransverseloadsontheventpipe(moregenerallyontheinternalstructuresinthewaterregionofthesuppressionpool)yieldstheenvelopingstaticeg'uivalentload.ThiswasalsoverifiedbytheKKBtestswiththeactualreliefsystem(Ref.38).Themaximumdifferentialpressurescalculatedfromthemeasurementresultsarep=0.16baratthequencherarm,andp=0.11barattheprotectivepipeonthedischargeline.TheyarebothconservativelyboundedbytheKKBspecifiedvalueofp=0.2bar.TheKKBtestresultsshowsthatthereisaclearseparationbetweenthespecifiedloadsandthemaximummeasuredloadsforboththelateralandverticalloadsoninternalsinthepoolofthesuppressionpool.BasedontheverificationofthetransverseloadsbytheKKBtestsandbasedonthecomparisonbetweenspecificationandmeasurementfortheKarlsteintests(seeSection8.5.5.1),itcanbestatedthatthevaluesformulatedinthespecificationforthetransverseloadsoninternalstructuresinthewaterregionyieldenvelopingstaticequivalentloads.REV1g3/798-118 KEY:1.Reliefvalve2.Compressedair3.HPsteamline4.Heating5.SRV6.CondensateeP,$$...'CIA,(o!>.d'1$$'4(C$!PACmli0a!0CUmmZCh0Zzzgzo$mLmZZr~Dmmnfh0Z
1$3t'4P<Pg~$4.$1716.))),O~15OO$$X(4DO~$)3KEY)l.Reliefvalve2.Compressedair3~HPsteamline4.Heating5.SRV6.CondensateI~~Ii.Ispo'gA9ggl~P8~%1,,C0Cpmxmgm)NyIIIgr~'Ngoc.)pE$$$$eIPE$'4)7Q3gy$~pill)yOH e
bi<<'seyesronJle'e/rs+cecrco~fttprnConcretesthtcttrralelementspaA0o<O040O4nchoa40OIiCl~0~I+ern,/iP5.30POwsleywntppe~O0ociOon00D'nXPS.p0oV/~~+YgPP59'P5.4np55P5.6P5.7p4IfR'~RT5.3vus.cXi0Oron/Ctop0n~0BBrooreDischargelineforttWUSRvtnnlustedforIh+testsinqwstionlREV13/79SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTKARLSTEINTESTTANKPLANVIERTYPICALVENTCLZARIl6INSTRU-MENTATIONFIGURE'.7 LPc.5Ciscctce!cr(("IU5((ar(hata>seatla>Iheleeleahaa>estaahl~~'0o0cteo'0ct'c>00orIIII.IIIILPr3IStJ5.(Ic>G5.2I"0't'.0o~000~00a~"~P5.70~~~~00J0,~a7...-.~T5.c0~a0LP42-gt~a>I.ooI~00oa~7o00iJiWJ~iLt.;,*:;'.Cl~~pa90rJ~Ja>~Ph:iI>thesBracingOummymentptpet>>at>'se>esIipali5.5K.S.prototypequencherP54P,S.3Pc2REVISUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTKARLSTEIHTESTTAHKC-DVIENTYHCALVEHTCLEARINGIsSSTRU-MEHTATIOHFIGURE8.8 LPC>d,dLP4,4~7,2m~6,2emSG52~oI~~~0~0~~0.00~~rLP4.2SG5.t0T56PST0Jg'h20000Cl02.65m~~p0Od~g4000000LprIaooT53nP56ppeP5.9~o00~rIOdd~~o4~~JdJo20JddCrooorPJ~dJCdT5IIT52PSoIP5!0P6iSS.ESprototypequencherREV13/79SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTKhRLSTEINTE8TThEKh-BVISfvrncar.errchemi;aeraU-MEm,'hTIOSFIGURE8.9
100Oc908093C=200FIa0TwoActuations(firstclean,secondreal)~MultipleActuations(firstclean,subsequentreal)OO70CDCD6050e<03020-24"C=75F55C=130FOperationFieldaVlID.CDCDCD6a~cn<(IaCD.CDCD34I10III203040I506070,80bar90ReactorPressureREV1SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLOCATIONOFTESTCROUPN).1INTHEOPERATIONFIELOFIGUREey6 100908093C=200FIaCO130TwoActuations(firstclean,secondreal)OO70CDCD6055C=130FSO403024C=7SFOperationFieldUJDO)COCIIClCLCDCDCD110Cll10I":02030II4050607080bar90RecctorPresure3/79SUSQUEHANNASTEAMELECTRICSTATIONUNITS'IAND2DESIGNASSESSMENTREPORTLOCATIONOFTESTGROUPNO.2INTHEOPERATIONFIELDFIGURE817 100'CU9080OO7093C=200FIaChCDaVlCLC)C)~MultipleActuations(firstclean,subsequentreal)6055C=130F50403024C=75FOperationFieldLaJDCACDaVlCLC)15~cr'(IatDC)g020304050I607080bar90ReaciorPressUreREF1SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLOCATIONOFTESTGROUPIOo3INTHEOPERATIONFIELDFIGURE
1009080OO7093C=200FaIhCL(D240TwoActvations(firstclean,secondreal1~MultipleActvations(firstclean,subsequentreal)6050403055C=130FOperationField22LUChlDCIIt5CLlDC3CD21~oi0UlGLC)EDPl24DC~750F1617181920I10II2030II4050II607080bar90ReactorPressure3/V9SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLOCATIONOFTESTGROUPNO.4INTHEOPKRhTIONFIELDFIGURE8.19 100'C9080OO7093C=200FlaalhCL.lDC)0TwoActuations(firstcleanIsecondreal)6050403024C=75F55C=130FOperationFieldLaJDChCDCIIaCLC)C)C)L25alACLCIC)201020304050607080bcr90RBQclorPressureREV1SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLOCATIONOFTESTGROUPNO~5INTHEOPERATIONFIELDFIGURE8.20 100'CP9080OO7093C=200FlaClCIlCLCDCD0OneActuatinnundercleancondition6TwoActuations(firstclean,secondreal)6050403024C=75F55C=130FOperationFieldIaJDCDCDCIIaUlO.C)CD26/32laA3CDCD<IIaVlID.CD20-I1020304050607080bar90ReactorPressureaZV13/79SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLOChTIONOFTESTGROUPIO.6IlfTHEOPERATIONFIELDFIGURE8.21 bor129Q.8cn6acn5o4CD32102030405060?08090bQrAbs.SystemPressureP2.6-~REV13/V9SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTVESTCLEARIRrHIESSUREVERSUSSYSTEMPRESSUREM?FLIlKV%ITCLEARINGTESTSFIGURE8.25 0
bar1413121098UlVlOl0OcA6cn5Co432102030405060708090barAbs.SystemPressureP2.6-5>>3/79SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTVENTCLEARINGPRESSUREVERSUSSYSTEMPRESSURESHORTLINEVENTCLFAHIlSTESTSFIGUREB.p6 50Height2019CO501TemperatureDistribution0beforeTest-Beginnigrr2TemperatureDistribution228SecondsafterTest-BeginnigCJe18,2co18cr0ooocoo17g16,5oc16Iclo)0COCOCO4CgagC.,o~C(C0'~~O)@0iN)IpMMMOI0CC7mmXCO922yC+m3'm22r~Um~uOOI02153(fC14~04080'ocationoftheQuencher195751201602002402803203604080CircumferentialLocationoftheCondensationChamber QRj+~l~+~q~~ge7PREFACF.hisReportcontainsdata,descriptionsandanaylsisre1ativetoeadequacyoftheSusquehannaSteamElectricStationsdesigntoacmmodateloadsresultinqfromasafetyrelief,valve(SRV)discrgeand/oraloss-of-coolantaccident(LOCA)./h IC=
~DlIT88LJOFCONTENTSChapterlGENERALINFORMATION1.1PurposeofReport1.2HistoryofProblem1.3QuencherDischargeDevice14tlKIISupportinqProgram1.5PlantDescription1.6Fiqures1.7TablesChapter2SUN'RY2.1LoadDefinitionSummary2.2DesiqnAssessmentSummaryChapter3SRVDISCHARGEANDLOCATRANSIENTDESCRIPTIONChapter4LOADDEFINITION4.1LoadsfromSafetyReliefValveDischarge4.2LoadsfromLoss-of-CoolantAccident4.3AnnulusPressurization4.4Fiqures4.5Tables3.1DescriptionofSafetyReliefValve{SRV)Discharge3.2DescriptionofLoss-of-CoolantAccident(LOCA)Chapter5LOADCONBINATIONSFORSTRUCTUR~ESPIPING~ANDEOUIPi'IENT5.1ConcreteContainmentandReactorBuildingLoadCombinations5.2STructuralSteelLoadCombinations5.3LinerPlateLoadCombinations54DovncomerLoadCombinations5.5Piping,Quencher,andQuencherSupportLoadCombinations5.6NSSSLoadCombinations5.7EquipmentLoadCombinations5.8Figures5.9TablesChapter6DESIGNCAPABILITYASSESSMENT6.1ConcreteContainmentandReactorBuildingCapabilityAssessmentCriteria6.2StructuralSteelCapabilityAssessmentCriteria6.3LinerPlateCapabilityAssessmentCriteria
TABLEOFCONTENTS~Continue~d6.4DowncomerCapabilityAssessmentCriteria6.5Pipinq,Quencher,andQuencherSupportCapabilityAssessmentCriteria6.6NSSSCapabilityAssessmentCriteria6.7EquipmentCapabilityAssessmentCriteriaChapter7DESIGNASSESSMENT7.1AssessmentMethodology7.2DesignCapabilityMargins7.3FiguresChapter88~SEEOENCHERVERIVICETIONTEST8.1UnitCellApproach8.2SimulationofSSESParameters8.3InstrumentationArrangement8.4TestMatrix8.5AnalysisofData8.6FiguresChapter9RESPONSESTONRC~UESTIONS9.1IdentificationofQuestionsUniquetoSSES9.2QuestionsUniquetoSSESandResponsesThereto9.3FiguresChapter10REFERENCESAppendixACONTAINMENTDESIGNASSESSMENTA.IContainmentStructuralDesignAssessmentA.2SubmergedStructuresDesiqnAssessmentAppendixBCONTAINMENTRESPONSESPECTRADUETOSRVANDLOCA)LOADSAppendixCREACTORBUILDINGRESPONSESPECTRADUETOSRVANDLOCALOADS It.'II,it)+lrIii'tt' AppendixDTABLEOPCONTENTS~Continue~dPROGRAMVERIFICATIOND.1PoolswellModelVerificationD.2FiquresD.3TablesAppendixEREACTORBUILDINGSTRUCTURALDESIGNASSESSMENTAppendixFPIPINGDESIGNASSESSNENTAppendixGNSSSDESIGNASSESSNENTAppendixHEQUIPMENTDESIGNASSESSHENT CHAPTER1GENERALINFORMATIONTABLEOFCONTENTS1.1PURPOSE'OFREPORT12HISTORYOFPROBLEM13QUENCHERDISCHARGEDEVICE1.4MARKIX,SUPPORTINGPROGRAM15PLANTDESCRIPTION1.5.1PrimaryContainment1.5.1.1Penetrations1.5.1.2InternalStructures16FIGURES1.7TABLES1-1 NumberTitle1-11-21-0CrossSectionofContainmentSuppressionChamber,PartialPlanSuppressionChamber,SectionViewQuencherDistribution CHAPTER1TABLESNumber1-2TitleSSESX,icensingBasis"SSESContainment'Dimensions1-3SSESContainmentDesignParameters"-1-3 10GQNERAIINFOBHATEON11PURPOSEANDOBGANXZATXONOFBEPQRTThepurposeofthisreportistopresentevidencethattheSusguehannaSteamElectricStation(SSES)designmarginsareadequateshouldtheplantbesubjectedtotherecentlydefinedthermohydrodynamicloadswhich-'r'esult'romsafetyreliefvalve(SBV)operationsand/ordischargesduringaloss-of-coolantaccident(LOCA)..inaGE.boilinqwaterreactor(BWB)1-4 Thecriteriausedfors'electionoftheSRVdischargedeviceforSSESwereminimizationofpressureoscillationloadsinthesuppressionpoolandstablecond'ensationofsteamfortherangeofsuppressionpooltemperaturesoverwhichsafetyreliefvalvescanbeexpectedtooperateTheoptionsconsideredforsatisfyinqthesecriteriaweretherams-headtee,thequencherdischargedevice,andvariationsonthesedesigns.Evaluationofthetwoprincipaldevicesindicatedthatthequencherofferedsignificantadvantagesovertherams-head,includingimprovedthermalperformanceathigherpooloperatinqtemperatures,aswellasreducedloads.Athermohydraulicquencherdesignforthesafe'tyreliefsystemoftheSSZSisbeingengineeredbyKraftwerkOnion{KMU)tosatisfytheabovecriteria.TheSSESquencherdesignisdifferentfromthatpresentedintheMarkIIDFFRinthatithasbeenoptimizedbasedonparametricteststudieswhichwereconductedbyKMUinordertominimizeSRVdischargeloadsKraftwerkUnionhassuppliedtoPPGLapackageofsignificantdesiqnandtestreportspertainingtothequencherdevelopmenttodemonstratedesignadequacyandgualityoftheirdevice{refertoTable1-1).Mithreqardtothe<<secondpop<<phenomenon,KMUtestshaveindicatedthat,duetothequencherflowresistance,thewaterlevelintheSRVdischargepipefollowinginitialdischargedoesnotriseabovethewaterlevelofthesuppressionpool.RefertoSubsection0.1.3.6forafurtherdiscussion.ToverifyKMU'sdesignapproachafull-scaleSSESuniqueunitce11test,asdescribedinChapter8,isbeingperformedbyKMUforPPSL.Section01presentstheanalysismethodsoftheSRVdischarqeloading1-7 1.4HARKIISUPPORTINGPROGRAMPPGLisamemberoftheMarkIIovnersgroupthatvasformedinJune,1975todefineandinvestigatethedynamicloadsduetoSRVdischargeandLOCA.TheNarkIIovnersgroupcontainmentprogramconcentratedinitiallyonthetasksrequiredforthelicensingoftheleadplants(Zimmer,LaSalle,andShoreham).Thisphaseof,work,calledtheshorttermprogram,isessentiallycomplete(asofJanuary,1978)andalonqertermprogramisundervay.ThefinalgoaloftheMarkIIprogramistoevolveacompleteDPFRwhichvillsupporttheplant-uniqueDARssubmittedbyeachplantforitslicensetooperate.AfterqainingsomeunderstandingofthecontainmentloadsthroughtheinitialMarkIIwork,PPGLdecidedtofindaqualifiedconsultanttosupplementin-housetechnicalresourcesandassistinthedeterminationofarealisticcourseofactionforSusquehanna.InNovember,1976,StanfordResearchInstitute,novcalledStanfordResearchInstituteInternational(SRI),wasselected,andaninformationexchangebetweenSHIandPPGLensuedtodeterminewhatcausedthegreatestloadsonthecontainmentstructure.AfterconductingacompletereviewofknowndatafromtheMarkIIprogramandotherknowledqeablepersonsandorganizations,PPGXandSRIdecidedthattheloadsfrommainsteamsafetyreliefvalve(SHV)dischargewerethekey1oadstobecontrolled.AstudyofpossiblemethodsofcontrollingtheloadandareviewofvhatactivitieswereoccurringinEuropeledPPGLandSHItotheconclusionthatanSRVdischargemitigatingdevice{quencher)shouldbeemployedtoreducethisloadingontheSusquehannacontainment.AlthoughtheHarkIIovnersgrouphadquencher-relatedtasksintheirprogram,thesetaskswerenotsufficientlytimelytosatisfySSES-constructionscheduleneeds.PromreviewinqtheworkdoneinEuropebysuchfirmsasASEATOM,MARVIKEN,andKraftwerkUnion,PPGLdiscoveredthatallknownquencherdesignswerebasedondatafromKraftwerkUnion(KMU).Thus,inMarch,1977,SRI,Bechtel(theSSESArchitect/Engineer)andPPGLvisitedKWUfordiscussionandtourofquencher-relatedfacilities.InlateJuly,1977,PPGLemployedtheservicesofKMUtodesignaSSES-uniquequencherdevice(seeSection1.3).ThedefinitionofLOCAloads(Section4.2)isinaccordancewiththeNarkIIprogramDuetotheschedulerestrictionsforSusquehanna.PPGLwilldefinethethermo-hydrodynamicloadsresultinqfromSRVdischargeusinqanapproachdevelopedbyKMU.Thisapproach(presentedinSection4.1)differsfromthatoftheHarkIIproqram.SeeTablel-lforasummaryofthedocumentationsupportinqSSESlicensinq.1-8 15PLANTDESCRIPTIONTheSSES,Units1and2,isbeingbuiltinSalemTownship,Luzecne,County,about5milesnoctheastoftheBorouqhofBerwick.Twogeneratingunitsofapproximately1,100megawattseachare.scheduledforoperation.Unit'1forNovembec1,1980,andUnit2forMay1,1982.GeneralElectricissupplying'henuclearsteamsupplysystems;Bechtelpowercorporationisthearchitect-engineerandconstructor.Thereactorbuildingcontainsthemajornuclearsystemsandequipment.ThenuclearreactorsforUnits1and2areboilingwater,directcycletypeswitharatedheatoutputof11.2x10~Btu/hr.Eachreactorsupplies134x10~lb/hrofsteamtothetandemcompound,doubleflowtucbines.15.'IP~cimanContainmentaThecontainmentisareinforcedconcretestructureconsistingofacylindricalsuppressionchamberbeneathatruncatedconicaldrywell.Piqure1-1showsthegeometry'fthe.containmentandinternalstructures.Theconicalportionoftheprimarycontainment(drywell)enclosesthereactorvessel,reactorcoolantrecirculationloops,andassociatedcomponents'ofthereactorcoolantsystem.Thedcywellisseparatedfromthewetwell,ie,thepressuresuppressionchamberandpool,bythedrywellfloor,.alsonamedthediaphragmslabMajorsystemsandcomponentsinthecontainmentincludetheventpipesystem(downcomers)connectingthedrywellandwetwell,isolationvalves,vacuumreliefsystem,containmentcoolingsystems,andotherserviceequipment.Theconeandcylinderformastructurallyinteqratedreinforcedconcretevessel,linedwithsteelplateandclosedatthetopofthedrywellwithasteeldomedhead.Thecarbonsteellinerplateisanchoredtotheconcretebystru'cturalsteelmembersembeddedintheconcreteandweldedto-theplate.Theentire-containmentisstructurallyseparatedfromthesurroundingreactorbuildingexceptatthebasefoundationslabfareinforcedconcretemat,toplinedwithacarbonsteellinerplate)whereacoldjointbetweenthetwoadjoiningfoundationslabsisprovided.ThecontainmentstructuredimensionsandparametersarelistedinTables1-2and1-3.AdetailedplantdescriptioncanbefoundintheSSESPSAH,Section3.81.5.l.1PenetrationsServicesandcommunicationbetweentheinsideandoutsideofthecontainmentaremadepossiblebypenetrationsthroughthecontainmentwallThebasictypesofpenetrationsarethedrywellhead,accesshatches(equipmenthatches,personnellock,suppressionchamberaccesshatches,CRDremovalhatch),electricalpenetrations,andpipepenetcations.Thepiping1-9 penetrationsconsistbasicallyofapipewithplateflangeweldedtoit.Theplateflange.isembeddedintheconcretewallandprovidesananchorageforthepenetrationtoresistnormaloperatingandaccidentpipereaction.loads.Theinternalstructuresconsistofreinforcedconcreteandstructuralsteelandhavethemajorfunctionsofsupportingandshieldingthereactorvessel,supportingthepipingandeguipment,andformingthepressuresuppressionboundary.Thesestructuresinclude'hedrywellfloor(diaphragmslab),thereactorpedestal(aconcentriccylindricalreinforcedconcreteshellrestingonthecontainmentbasefoundationslabandsupportinqthereactorvessel),thereactorshieldwall,thesuppressionchambercolumns(hollowsteelpipecolumnssupportingthediaphragmslab),thedrywellplatforms,theseismictrusses,thequenchersupports,andthereactorsteamsupplysystemsupports.SeeFigures1-1through1-4aridTables1-2and1-31-10 dyCt"dp~rIsl.400106PC'CrCONTAINMENTUp(gpSRVDIAPHRAGMSLABPENETRATIONFROMDIAPHRAGMSLABPENETRATIONO~,Ce,i03.0>~~dp~136o~fp<'~u~e~Criy160P0r~drg1sr~irIdQ8O~'EDESTALTOQUENCHERSSRVDIAPHRAGMSLABPENETRATIONNOTE:BRACINGISNOTSHOWNSUSQUEHANNASTEAMELECTRICSTATION.UNITS1AND2DESIGNASSESSMENTREPORTSUPPRESSIONCHAMBERPARTIALPLAN/FIGURE1-2 3454001So315000'PP,<<PO.<<<<~~~~<<'<<454300088FT'"K"-ur,25822mm~~2850P100FT"30480mmH~h,BGo0~a'~.A<<<<MP~,<<I'O7500<<2700d.a2550b<<0RSOFT"18288mm<<~pP3eFT<mP'<<4<<a'1050240V5r~P<<0~~C22542FT'"F12802mmD~<<<<<<~~L'<<<<~~.e0<<~4'0PP~k'3519541CS4NOTE:INDICATESADS-ASSOCIATEDQUENCHERSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTQUENCHERDISTRIBUTIONFIGURE1R TABLE1-1SSESLICENSINGBASISI..MarkIIContainment-SupportingProgramA.LOCh-RelatedTasksh.2.h.3.PoolSwellModelReportImpactTestsh.4.ImpactModelEPRI1/13ScaleTestsTaskeeetet~ettivetA.l."4T"PhasesI,II,IIXActivitePhaseITestReportPhaseIApplicationMemorandumPhaseII6IIITestReportPhaseII&IIIApplicationMemorandumModelReportPSTF1/3ScaleTestsMarkI1/2ScaleTestsPSTF1/3ScaleTestsMarkI1/2ScaleTestsEPRIReportTarget~teeletteCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedCompletedDocumentationNEDO/NEDE13442-P-01-5/76ApplicationMemo-6/76NEDO/NEDE13468-P-12/76ApplicationMemo-1/77HEDO/NEDE21544-P-12/76HEDE13426-P-8/75NEDC20989-2P-9/75NEDE13426-P-8/75HEDC20989-2P-9/75EPRINP-441-4/77UsedforSSESLicensinYesYesYesYesYesYesYesYesNoYesA.5.LoadsonSubmergedStructuresLOCA/RHAirBubbleModel12/77LOCA/RHHaterJetModel12/77ApplicationsMethods12/77TestReports1Q/78NEDE21471NEDE21472NEDE21730ReportUndecidedUndecidedUndecidedUndecidedh.6.h.7~A.8.ChuggingAnalysisandTestingChuggingSingleVentEPRITestEvaluationSingleCellReport4TFSIReportMultiventModelCREAREReportEPRI-4TComparisonCompleted1/7812/774Q/77CompletedHEDE23703-P-ll/77NEDE23710-PNEDE21669-PReportNEDO21667<<8/77YesHoHoYesh.9.MultiventSubscaleTestingandAnalysisFacilityDescriptionand4Q/77TestPlanTestReport1979ReportFinalReportUndecidedUndecided TaskNumberA.10.~eetteitSingleVentLateralLoadsActiviteAnalysisReportTarget~Contestee4Q/77DocumentationReportUsedforSSESLicensinUndecidedB.SRVRelatedTasksB.l.QuencherModelB.2.RamsheadModelB.3.MonticelloIn-PlantSRVTestsDFFRModelConfirmatoryTestsAnalysisPreliminaryTestReportHydrodynamicReportCompleted3Q/78CompletedCompletedCompletedNEDO/NEDE21061-P-9/76ReportNEDO/NEDE21061-P-9/76NEDC21465-P-12/76NEDC215&1-P-8/77..NoNoNoNoNoB.4.B.5.B.6.Be7.B.S.B.9vB.10.B.11.B.12.B.13.ConsecutiveActuationTransientAnalysisSRVQuencherIn-PlantCaorsoTestsThermalMixingModelSRVWaterClearingQuencherAirBubbleFrequencyMonticelloFluidStructureInteraction(FSI)DFFRRamsheadModelComparisontoMonticelloDataRamsheadSRVMethodologySurnLaryStructuralResponsetoSRVDischargeQuencherEmpiricalModelUpdateAnalyticalModelsTestPlanAdvanceTestReportPinalReportAnalyticalModelAnalysisAnalyticalModelAnalysisData/ModelComparisonAnalyticalMethodsAnalyticalReportAnalyticalModelandCorrelation4Q/77Completed1Q/784QI784Q/783Q/784Q/771Q/78CompletedCompleted4Q/77lQ/79ReportNEDM20988-12/76ReportReportNEDC23689ReportReportReportNSC-GEN0394-10/77NEDO24070-11/77ReportReportNoNoNoNoNoNoNoNoNoNoNoNo TaskNnett~ActivitC.MiscellaneousTasksActiviteTarget.~ConlotionDocumentationUsedforSSESLicensinC.l.C.2.C.3.DFFR,Rev.3MassandEnergyReleaseReportNRCRound1QuestionsRevisionAnalyticalReportDFFRAmendment110/783/77CompletedGB-77-65NEDO/NEDE21061Amendment1-12/76YesYesNEDO/NEDE21061Revision3NotyetavailableC.4.C.5.C.6.DecouplingChuggingandSRVLoadsSRSSJustificationNRCRound2QuestionsDFFRAmendment1,Supplement1SRSSReportDFFRAmendment212/77OnholdCompletedCompletedNEDO/NEDE21061Amendment1,Supplement2NEDO/NEDE24010-7/77NEDO/NEDE21061Amendment2-6/77YesYesYesDFFRAmendment2,Supplement1DFFRAmendment2,Supplement2Supplement3CompletedCompleted4q/77NEDO/NEDE21061Amendment2,Supplement1-8/77NEDO/NEDE21061Amendment2,.Supplement2Supplement3YesYesYesC.7.Justificationof"4T"BoundingLoadsChuggingLoadsJustificationCompletedNEDO/NEDE23617"P-8/77NEDO/NEDE24013-P-8/77NEDO/NEDE24104-P-8/77NEDO/NEDE24015-P-8/77NEDO/NEDE24016-P'-8/77NEDO/NEDE24017-P-8/77NEDO/NEDE23627-P-8/77UndecidedUndecidedUndecidedUndecidedUndecidedUndecidedUndecidedC.g.FSIEffectsinMarkIIContainmentsC.9.MonitorWorldTestsII.1NUTestsandReports(suppliedtoPP&L)EvaluationofFSIEffectslg/78MonitoringWorldPressureSuppressionTestsReportReports(Quarterly)UndecidedNoDocumentNumberTitleFormationandoscillationofasphericalgasbubbleStatusCompletedDocumentationAEG-Report2241UsedforSSESLicensinYesAnalyticalmodelforclarificationofpressurepulsationinthewetwellafterventcleaningCopletedAEG-Report2208
DocumentNumber3~4,5.6.7.8.TitleTestsonmixedcondensationwithmodelquenchersCondensationandventclearingtestsatGKMwithquenchersConceptanddesignofthepressurereliefsystemwithquenchersKKBventclearingwithquencherTestsoncondensationwithquencherswhensubmergenceofquencherarmsisshallowKKB-ConceptandtaskofpressurereliefsystemStatusCompletedCompletedCompletedCompletedCompletedCompletedDocumentationKWV-Report2593KWV-Report2594KWV-Report2703KWV-Report2796KWV-Report2840KWV-.Report2871UsedforSSESLicensinYesYesYesYesYesYes9.ExperimentalapproachtoventclearinginamodeltankCompletedKMV-Report3129Yes10.KKB-Specificationofblowdowntestsduringnon-nuclearhot.functionaltest-Rev.IdatedOctober4,1974CompletedKWU/V822ReportYesAnticipateddataforblowdowntestswith-pressurereliefsystemduringthenon-nuclearhotfunctionaltestatnuclearpowerstationBrunsbuttel(KKB)CompletedKWU-Report3141Yes12.13.Resultsofthenon-nuclearhotfunctionaltestswiththepressurereliefsysteminthenuclearpowerstationBrunsbuttelAnalysisoftheloadsmeasuredonthepressurereliefsystemduringthenon-nuclearhot,functionaltestatKKBCompletedCompletedKWU-Report3267KWU-Report3346YesYes14.KKB-Listingoftestparametersandimportanttestdataofthenon-nuclearhotfunctionaltestswiththepressurereliefsystemCompletedKWU-WorkingReportR521/40/77Yes15.KKB-Specificationofadditionaltestsfortestingofthepressurereliefvalvesduringthenuclearstart-up,Rev.1CompletedKWU/V822TAYes16.KKB-Resultsfromnuclearstart-uptestingofpressurereliefsystemCompletedKNJ-WorkingReportR142-136/76Yes17'uclearPowerStationPhillipsburg-Unit1HotFunctionalTest:SpecificationofpressurereliefvalvetestsaswellasemergencycoolingandwetwellcoolingsystemsCompletedKWU/V822/RF13Yes DocumentNumberTitleStatusDocumentationUsedforSSESLicensin18.Resultsofthenon-nuclearhotfunctionaltestswiththepressurereliefsysteminthenuclearpowerstationPhillipsburgCompletedKWU-WorkingReportR142-38/77Yes19.KKPI-Listingoftestparametersandimportanttestdataofthenon-nuclearhotfunctionaltestswiththepressurereliefsystemCompletedKWU-WorkingReportR521/41/77Yes20.AiroscillationsduringventclearingwithsingleanddoublepipesCompletedAEG-Report2327Yes616715/cak TABLE1-3SSESCONTAINMENTDESIGNPAKQKTERSA.DrellandSuressionChamber1.InternalDesignPressure2.ExternalDesignPressure3.DrywellFloorDesignDifferentialPressureUpwardDownward4.DesignTemperature5.DrywellFreeVolume(Minimum)(includingvents)(Normal)(Maximum)6.SuppressionChamberVolumeFree(Minimum)(Normal)(Maximum)7.SuppressionChamberWaterVolume(Minimum)(Normal)(Maximum)8.PoolCross-SectionArea~Drell53psig5psid340'F239,337ft33239,593ft3239,850ft.28psid28psidSuressionChamber45psig5psid220'F148,590ft3153,860ft3159,130ft122,410ft3131,550ftGross(OutsidePedestal)TotalGross(IncludingPedestalWaterArea)Free(OutsidePedestal)TotalFree5379ft5679ft5065ft5365ft CHAPTER2SUMMARYTABLEOPCONTENTSLOADDEFINITIONSUMMARY21.1SRULoadDefinitionSummary2.1.2LOCAI.oadDefinitionSummaryDESIGNASSESSMENTSUMMARY2.2.1ContainmentStructureandReactorBuildingAssessmentSummary2.2.1.1ContainmentStructureAssessmentSummary.2.2.1.2ReactorBuildingAssessmentSummary222223ContainmentSubmergedStructuresAssessmentSummaryPipingSystemsAsessmentSummary 20SUMMARYThisDesignAssessmentReportcontainstheSSESadequacyevaluationfordynamicloadsduetoLOCAandSRVdischarge.2-2 21LOADDEFINITIONSUMNARY2.1.1SRVLoadDefinitionS~nmmarHydrodynamicloadsresultingfromSRVactuationfallintotvodistinctcategories:loadsontheSRVsystemitself(thedischargelineandthedischargequencherdevice),andtheairclearingloadsonthesuppressionpoolwallsandsubmergedstructures.LoadsontheSRVsystemduringSRVactuationincludeloadson,theSRVpipingduetoeffectsofsteadybackpressure,transientvaterslugclearing,andSRVlinetemperature.Determinationofloadingonthequencherbody,arms,andsupportisbasedontransientsresultingfromvalveopening(waterclearingandairclearing),valveclosing,andoperationofanadjacentquencher.Airclearingloadsareexaminedforfourloadingcases:symmetric(allvalve)SRVactuation,asymmetricSBVactuation,singleSRVactuation,andAutomaticDepressurizationSystem{ADS)actuation.Dynamicforcingfunctionsforloadingofthecontainmentwalls,pedestal,basemat,andsubmergedstructuresaredevelopedusingt'echniquesdevelopedinSection4.1.LoadsontheSRVsystemduetoSRVactuationarediscussed.inSubsection4.1.2,andloadsonsuppressionpoolstructuresduetoSRVactuationarediscussedinSubsection4.1.3.Afullscale,unitcelltestprogramisheingemployedtoverifySSESuniqueSRVloadingasdescribedinChapter8.2.12LOCALoadDefinitionSummaryThespectrumofLOCA-inducedloadsontheSSEScontainmentstructureischaracterizedbyLOCAloadsassociatedwithpoolsvell,condensationoscillationandchuggingloads,aswellaslongtermLOCAloads.TheLOCAloadsassociatedvithpoolsvellresultfromshortdurationtransientsandincludedowncomerclearingloads,w'aterjetloads,poolsvellimpactanddragloads,poolfallbackdragloads,poolswellairbubbleloads,andloadsduetodryvellandvetwelltemperatureandpressuretransients.TechniquesusedtoevaluatetheseloadsaredescribedinSubsection4.21.Condensationoscillationsresultfrommixedflow{air/steam)andpuresteamfloveffectsinthesuppressionpool.Chuggingloadsresultfromlovmassfluxpuresteamcondensation.TheloaddefinitionsforthesephenomenaarecontainedinSubsection4.2.2.LongtermLOCAloadsresultfromthosevetvellanddryvellpressureandtemperaturetransientswhichareassociatedwithdesignbasisaccidents(DBA),intermediateaccidents(IBA),andsmallbreakaccidents(SBA).TheirloaddefinitionsarecontainedinSubsection4.2.3.
StructuresdirectlyaffectedbyLOCAloadsincludethedrywellwallsandfloor,wetwellwalls,RPVpedestal,basemat,linerplate,columns,downcomers,downcomerbracingsystem,guenchers,andwetwellpiping.TheirloadingconditionsaredescribedinSubsection424.2-4 22DESIGNASSESSMENTSUNKARYDesignassessmentoftheSSESstructuresandcomponentsisachievedbyanalyzingtheresponseofthestructuresandcomponentstotheloadcombinationsexplainedinChapter5.InChapter7,predictedstressesandresponses(fromtheloadsdefinedinChapter4andcombinedasdescribedinChapter5)arecomparedviththeapplicablecodeallovablevaluesidentifiedinChapter6;theSSESdesignvillbeassessedasadequatebyvirtueofdesigncapabilitiesexceedingthestressesorresponsesresultingfromSRVdischargeorJ.OCAloads.2.2.1ContainmentStructureandReactorBuildingAssessmentSum~mar2.2.1.1ContainmentStructureAssessmentSummaryTheprimarycontainmentvalls,baseslab,diaphragmslab,reactorpedestal,andreactorshieldareanalyzedfortheeffectsofSRVandLOCAinaccordancevithTable5-1.TheANSYSfiniteelementprogramisusedforthedynamicanalysisofstructures.Responsespectracurvesaredevelopedatvariouslocationswithinthe~containment>>structure.toassesstheadequacyof,components.,Stressresultantsduetodynamicloadsarecombinedwithother1oadsinaccordancewithTable5-1to'evaluaterebarandconcretestresses.'esign'safetymarginswillaredefinedbycomparingtheactualconcreteandrebarstressesatcriticalsectionsvith,thecode,allowablevalues.2.2.'1.2ReactorBuildingAssessmentSummaryThereactorbuildingisassessedfortheeffectsofSRVandLOCAloadsinaccordancewithTable5-1.ContainmentbasemataccelerationtimehistoriesareusedtoinvestigatethereactorbuildingresponsetotheSRVandLOCAloads.Responsespectracurvesatvariousreactorbuildingelevationsareusedtoassesstheadequacyofcomponentsinthereactorbuilding.2.2.2ContainmentSubmergedStructuresAssessmentSummaryDesignassessmentofthesuppressionchambercolumnsanddovncomerpipesisbeingperformed.Baseduponanapproximate,equivalentstaticanalysiscarriedouttodate,strengtheningofthesestructuresshouldnotberequired.ThisconclusionvillbeconfirmedwhenthedynamicanalysisiscompletePreliminaryresultsfromthedynamicanalysisofthesuppressionpoollinerplateindicatethatnostructuralmodificationsarerequiredThisconclusionwillbeconfirmedwhenthefinalanalysisiscomplete.2-5 Theoriginaldowncomerbracinghasbeenredesignedwithpipesectionstominimizebracinqdragloadsduetopoolsvellandfallback.Therevisedbracingsystemisdesignedusingasimplifiedequivalentstaticapproach.ContainmentandreactorbuildingpipingsystemsarebeingdesignedtowithstandtheeffectsofLOCAandSRVinduceddynamicloads.TheloadcombinationsforpipingaredefinedinTable6.1ofRef.10.2-6 31DESCRIPTXOMOFSAFETYRELIEFVALVEDISCHARGESusquehannaUnits1and2areequippedwithasafetyreliefsystemwhichcondensesreactorsteaminasuppressionchamberpool.Bythisarrangement,reactor,steamisconductedtothewetwell.viafastactingsafetyreliefvalvesandquencherequippeddischargelines..ThissectiondiscussesthecausesofSRVdischarge,describestheSRVdischargeproce'ss,andidentifiestheresultantSRVdischargeactuationcases31.1CausesofSVDischa~reDuringcertainreactoroperatingtransients,theSRVsmaybeactuated(bypressure,byelectricalsignal,orbyoperatoraction)forrapidreliefofpressureinthereactorpressurevessel.Thefollowingreactoroperatingtransientshave,beenidentifiedasthosewhichmayresultinSRVactuation:\a.Turbineqeneratortrip(withbypassorwithout)b.Hainsteamlineisolationvalve(NSIV)closurec.Lossofcondenservacuumd.Feedwatercontrollerfailuree.Pressureregulatorfailure-openf.Generatorloadrejection(withandwithoutbypass)q.LossofacpowerhLossoffeedwaterflowTripoftworecirculationpumpsRecirculationflowcontrolfailure-decreasingflowk.InadvertentsafetyreliefvalveopeningAdetaileddescriptionofthesetransientsisprovidedinSection15.2oftheFSAR3.1.2DescriptionoftheSRVDischargePhenomenaandSRV~t.aadinCasesBeforeanindividualsafetyreliefvalveopens,thewaterlevelinthedischargelineisapproximatelyequaltothewaterlevelinthepoolAsavalveopens,steamflowsintothedischargelineairspacebetweenthevalveandthewatercolumnandmixeswiththeair(seedetailedevaluationinChapter3ofRef1,pages6-12through6-14).Sincethedownstreamportionofthedischargelinecontainsawatersluganddoesnotallowan3-3 immediatesteamdischargeintothepool~thepressureinside-theline'increases.Theincreasedpres'sureexpelsthewaterslugfromtheSRVdischargelineandquencher.Themagnitudeofthewaterclearinqpressureisprimarilyinfluencedbythesteamflowratethroughthevalve,thedeqreetowhichenteringsteamiscondensedalongthedischargeline;,walls,thevolumeofthedischargelineairspace,andthelengthofthewaterslugtobeaccelerated.Theclearingofwaterisfollowedbyanexpulsionoftheenclosedair-steamvolume.Theexhaustedgasformsanoscillatingsystemwiththesurroundingwater,wherethegasactsasthespringandthewateractsasthemass.Thisoscillatingsystemisthesourceofshorttermairclearingloads.Whiletheair-steammixtureoscillatesinthepoolitrisesbecauseofbuoyancyandeventuallybreaksthroughthepoolwatersurfaceatwhichtimeairclearingloadscease.Whenall'theairleavesthesafetyreliefsystem,steamflowsintothesuppressionpoolthrough'hequencherholesandcondenses.TheSSESquencherdesignassuresstablecondensationevenwithelevatedpoolwatertemperature.TheSBVactuationcasesresultingfromthetransientslistedinSubsection3.1.1areclassified,asbeingoneofthefollowingcases:a.Symmetric(allvalve,orAOT)dischargeb.Asymmetricdischarge,includingsinglevalvedischargec.AutomaticDepressurizationSystem(ADS)discharqeAlsoconsideredinthecontainmentdesignistheeffectofsubsequentSRVactuations(second-pop),discussedinSubsection4.1-3-6.Thesymmetricdischargecase(otherwisetermedtheall-valve,orAOT,case)isclassifiedasthetypeofSRVdischargethatwouldfollowrapidisolationofthevesselfromtheturbinesuchasturbinetrip,closureofallMSXVs,lossofcondenservacuum,etc.Aspressurebuildsupfollowingisolationofthe'vessel,theSBVsactuatesequentiallyaccordingtothepressuresetpointsofthevalves.ThismayormaynotresultinactuationofalltheSRVs,butforconservatisminloadingconsiderationsallvalvesareassumedtoactuateRefertoSubsection4.1.3.1fordiscussionoftheloadsresultingfromthisall-valvecase.AsymmetricdischarqeisdefinedasthefiringoftheSRVsforthe~threead'jacentquencherdeviceswhichresultsinthegreatestasymmetricpressureloadinqonthecontainment.Thissituationishypot'hesizedwhen,followingareactor'cramandisolationoftheVessel,decayheatraisesvesselpressuresothatlowsetpointvalvesactuate.Xf,duringthistimeofdischargeofdecayheatenergy,manualactuationofthetwootheradjacentSRVsthat3-4 comprisetheasymmetriccaseisassumed,thisactuationwouldresultinthemaximumsymmetricpressure'loadonthecontainment.Subsection4.1.3.2givesadiscussionoftheloadsresultingfromtheasymmetricdischarqecase.ThesinglevalvedischargecaseisclassifiedasthefiringoftheSRVwhichqivesthesinglelargesthydrodynamicload.TransientsthatcouldpotentiallyinitiatesuchacaseareaninadvertentSRVdischarqeorDesignBasisAccident(DBA).RefertoSubsection3.2.3foradiscussioriofthelatterpossibilitySubsection4.1.3.2.1providesadiscussionoft'eloadsresultingfromthesinglevalvecase.TheADSdischargeisdefinedasthesimultaneousactuationofthesixSRVsassociatedwiththeADS.SeePigure1-4forthelocationofthequencherdevicesassociatedwiththeADSvalves.TheADSisassumedtoactuatedurinqanIntermediateBreakAccident(IBA)orSmallBreakAccident'(SBA).IfanADSdischargeishypothesized'oincidenttoanIBAorSBA(describedinSubsections3.2.2and3.2.1,respectively),theeffectsofanincreasedsuppressionpooltemperature(resultingfromsteamcondensationduringtheLOCAtransient)andincreasedsuppressionchamberpressure(resultinqfromclearingofthedryvellairintothepooldurinqthetransient)areconsideredinthecalculationofpressureloadingsfortheADSdischargecase.SeeSubsection4.1.3.'3forfurtherdiscussionoftheloadsresultingfromtheADScase.I3-5 32DESCRIPTIONOFLOSS-OF-COOLANTACCIDENT'hiseventinvolvesthepostulationofaspectrumofpipingbreaksinsidethecontainmentvaryinginsizetype,andlocationofthebreak.Fortheanalysisofhydrodynamicloadingsonthecontainment,thepostulatedLOCAeventisidentifiedas.aSmallBreakAccident(SBA),anIntermediateBreakAccident(IBA),oraDesignBasisAccident(DBA).32.1SmallBreakAccientSB~AThissubsectiondiscussesthe.containmenttransientassociatedwithsmallprimarysystemblowdowns.TheprimarysystemrupturesiathiscategoryarethoserupturesthatwillnotresultinreactordepressurizationfromeitherlossofreactorcoolantorautomaticoperationoftheECCSequipment,ie,thoseruptureswithabreaksize.lessthan01sqftThefollowinqsequenceofeventsisassumedtooccurWiththereactorandcontainmeatoperatingatthemaximumnormalconditions,asmallbreakoccursthatallowsblowdownofreactorsteamorwatertothedrywell.Theresultingpressureincreaseinthedrywellleadstoahiqhdrywellpressuresignalthatscramsthereactorandactivatesthecontainmentisolationsystem.Thedrywellpressurecontinuestoincreaseataratedependentuponthesizeofthesteamleak.Thepressureincreaselowersthewaterlevelinthe-downcomers.At.thistime,airandsteamenterthesuppressionpoolataratedependentuponthesizeoftheleak.Onceallthedrywellairiscarriedovertothesuppressionchamber,pressurizationofthesuppressionchamberceasesandthesystemreachesanequilibriumcondition.The'drywellcontainsonlysuperheatedsteam,andcontinuedblowdownofreactorsteamcondensesinthesuppressionpool.Theprincipalloadinqconditioninthiscaseisthegraduallyincreasinqpressureinthedrywellandsuppressionpoolchamberandtheloadsrelatedtothecondensationofsteamattheendofthevents.3.2.2IntermediateBeakAccidentIBAThissubsectiondiscussesthecontainmenttransientassociatedwithintermediateprimarysystemblowdowns.ThisclassificationcoversbreaksforwhichtheblowdownwillresultinlimitedreactordepressurizationandoperationoftheECCS,ie,thebreaksizeisequaltoorslightlyqreaterthan0.1sqft.Followingthebreak,thedrywellpressureincreasesatapproximately1.0psi/sec.Thisdrywellpressuretransientissufficientlyslowsothatthedynamiceffectofthewaterintheventsisnegligibleandtheventswillclearwhenthedrywell-to-suppressionchamberdifferentialpressureisequaltothehydrostaticpressurecorrespondingtotheventsubmergence.The3-6 CHAPTER4LOADDEFINITIONTABLEOFCONTENTS4.1SAFETYRELIEFVALVE(SRV)DISCHARGELOADDEFINITION42LOSS-OF-COOLANTACCIDENT(LOCA)LOADDEFINITION42.142114212421342.1442.1.54.2.1.64.2.1.74.2.24.2.2.1422.242.34.23.14.2.3.2423-342.44.2.4142.424.2.4.342444.2.4.542.46LOCALoadsAssociatedwithPoolswellMetwell/DrywellPressuresduringPoolswellPoolswellImpactI.oadPoolswellDragLoadDowncomerClearingIoadsDowncomerWaterJetLoadPoolswellAirBubbleLoadPoolswellFallbackLoadCondensationOscillationsandChuggingLoadsCondensationOscillationLoadDefinitionChuggingLoadDefinitionLongTermI.OCALoadDefinitionDesignBasisAccident(DBA)TransientsIntermediateBreakAccident(IBA)TransientsSmallBreakAccident(SBA)TransientsLOCA.LoadingHistoriesforSSESContainmentComponentsLOCALoadsontheContainmentMallandPedestalLOCALOadsontheBasematandLinerPlateLOCALoadsontheDrywellandDrywellFloorLOCALoadsontheColumnsLOCALoadsontheDowncomersLOCALoadsonthedowncomerBracing4-1 4.247LOCALoadsonMetwellPiping43ANNULUSPRESSURIZATION44FIGURES45TABLES4-2 CHAPTER4Mum~beTitleP~XGURS4-1through4-37k4-384-394-404-414-424-434-444-454-464-474-48Thesefiguresareproprietaryandare,foundintheproprietarysupplementtothisDAR.SSESShortTermSuppressionPoolHeightSSESShortTermWetwellPressureSSESPoolSurfaceVelocityvsElevationBasemat-SSESWaterClearingJetSSESJetImpingementArea(WaterClearing)SSESPoolswellAirBubblePressureAirBubblePressureonSuppressionPoolWallsSymmetricandAsymmetricSpatial'LoadingSpecificationSSESDrywellPressureResponsetoDBALOCASSESWetwellPressureResponsetoDBALOCASSESSuppressionPoolTemperatureResponsetoDBALOCA4-494-504-514-524-534-544-55SSESDrywellTemperatureResponsetoDBALOCASSESSuppressionPoolTemperatureResponsetoIBATypicalNarkIIContainmentResponsetotheIBATypicalNarkIIContainmentResponsetotheSBASSESComponentsAffectedbyLOCALoadsSSESComponentsAffectedbyLOCALoadsLOCALoadingHistory.fortheSSESContainmentWallandPedestalLocalLoadingHistoryfortheSSESBasematandLinerPlate4-57LOCALoadingHistoryfortheSSESDrywellandDrywellFloor4-3 4-584-594-604-614-62LOCALoadingHistoryfortheSSESColumnsLOCALoadingHistoryfortheSSESDovncomersLOCALoadingHistoryfortheSSESDovncomerBracingSystemLOCALoadingHistoryforSSESQetvellPipingThisfigureisproprietary4-4 CHAPTER4Num~beTitleTABLES4-1through4-15ThesetablesareproprietaryandarefoundintheproprietarysupplementtothisDAR4-164-174-184-194-204-21LOCALoadsAssociatedwithPoolsvellSSESDryvellPressureSSESPlantUniquePoolsvellCodeInputDataInputDataforSSESLOCATransientsComponentLOCALoadChartforSSESHetvellPipingLOCALoadingSItuations 0LOADDEPXNTTION4.1SAFETYRELIEFVALVESB~VDISCRARGELOADDEFINITIONSeetheProprietarySupplement-forthissection.4-6 42LOCALOADDEFINITIONSubsections4.2.1,4.2.2and4.2.3villdiscussthenumericaldefinitionofloadsresultingfromaLOCAintheSSEScontainment.TheLOCAloadsaredividedintothreegroups.V(1)ShorttermLOCAloadsassociatedwithpoolsvell(Subsection4.2.1)(2)Condensationoscillationsandchuggingloads(Subsection4.2.2)(3)LongtermLOCAloads(Subsection4.2.3).TheapplicationoftheseloadstothevariouscomponentsandstructuresintheSSEScontainmentisdiscussedinSubsection4.2.4.421LOCALOADSASSOCIATEDWITHPOOLSWELLAdescriptionoftheLOCA/PoolswelltransienthasbeengiveninSection3.2ofthisDesignAssessmentReport.TheLOCAloadsassociatedvithpoolsvellare-listedinTable4-16.TheappropriateMarkXIgenericdocumentfromwhichSSESplantuniqueloadsarecalculatedisalsoshowninTable4-16.Adiscussionoftheseloa'dsandtheirSSESuniquevaluesfollows.ThedrywellpressuretransientusedforthepoolswellportionoftheLOCAtransient(<2.0seconds)isgiveninTableIV-D-3ofRef7.AportionofthistableisreproducedhereinasTable4-17.Thisdrywellpressuretransientincludestheblovdowneffectsofpipeinventoryandreactorsubcoolingandisthehighestpossibledrywellpressurecaseforpoolsvell.Theshorttermpoolsvellwetwellpressuretransientresultingfromthis'dr@wellpressuretransientiscalculatedbyapplyingthepoolswellmodelcontainedinRef8.TheequationsandassumptionsinthepoolsvellmodelwerecodedintoaBechtelcomputerprogramandverifiedagainsttheClass1,2and3testcasescontainedinRef9.ThisverificationisdocumentedinAppendixDtothisreport.OtherinputsusedforthecalculationoftheSSESplantuniguepoolswelltransientareshowninTable4-18.Theshorttermsuppressionpoolsurfaceelevationandcorrespondingwetwellpressuretransientcalculatedwiththe.poolsvellcodeareshowninPigures4-38and4-39respectively.Theshorttermwetvellpressurepeakis561psia{41.4psig).The(drywellminuswetvell)pressuredifferentialisalsoplottedonthiscurve.TheminimumAPoccurringdurinqpoolswellis-9.2psidat0.893secondsafterventclearing(1.58secondsafterthebreakoccurs)4-7 4.212PoolswellZmactLoa'dAnystructurelocatedbetweenthe.initialsuppressionpoolsurface(el.672')andthepeakpoolswellheight(el690',seefiqure4-38)issubjecttothepoolswellwaterimpactloadThereareonlyminorstructures(suchasmiscellaneouswetwellpipinq)inthisportionoftheSSESwetwall.ThisloadiscalculatedasspecifiedinRef10,S'ubsection44.6.ASSESplant-uniquevelocityyselevationcurvehasbeenqeneratedwiththepoolswellmodel(Figure4-40).I<isusedinconjunctionwithimpactpressurevsvelocitycurvesforvarioussizeandshapecomponents(Ref10,Figures4-34,4-35and4-36)todevelopapeakimpactpressureatthecomponen-t'selevation.Thepe'akimpactpressureiscombinedwithageneralizedimpactpressuretimehistorycurve(Ref10,Figure4-37)tospecifythestructuralload.Allstructures,subjecttopoolswellimpactloadsintheSSEScontainmentareclassifiedas>>smallstructures>>.2.1.3PoolaeellD~aaLoadThepoolswelldragloadappliestoanystructurelocatedbetweentheelevationoftheventexit(el.660')andthepeakpoolswell,heiqht(el.690').Theloadiscalculated'forallcomponentsintheregionbaseduponthemaximumpoolsurfacevelocity(29.35fps),regardlessofelevation.ThedragloadpressureiscalculatedfromRef10,Equation4-24usingVf=29.35fpsforthevelocityandp=62.41bmjft~.forthedensityofwater,fP=(1/2)CDpfVf~(4-1)P(psi)=5.8C'4-2)TheappropriatedragcoefficientforthestructureinvolvedisselectedfromRef10,Figure4-29.Thepoolswelldragloadisappliedineitherthehorizontalorverticaldirection(Subsection4.45.2ofRef10).Forthecaseofacomponentorientedverticallywithitsaxisparalleltothevelocityofthepoolsurface,*theskinfrictioncoefficient,AC~,usedinRef10,Subsection44.8isappliedinplaceofCD.I'hismethodwouldapply,forexample,totheverticalloadsondowncomers,columns,orsafetyrelieflinesinthewetwell.UsinqCf=0.0023,theverticaldragforcepsonaverticallyorientedcomponentisrecalculatedusingEquation4-26ofRef10.F(lbf)=0.0133Af(in~).V(4-3)HereAfistheskinfrictionarea(wettedsurfacearea)subjectto,theverticaldragforce.LOCAloadsonthedowncomerbracingaredescribedinSubsection4246 Verticalloadsonthedowncomersduringdowncomer'learingcanbeestimatedbyusingadragloadformulasimilartoEquation4-3.Inthiscasetheventclearingvelocityis60fps(Ref10,Subsection4.4.5.1)andAfisthewettedinsideareaofthedowncomer,conservativelycalculatedto.beAf={12ft}(m)(2ft)=75~4ft<FromEquation4-3theverticalclearingloadonthedowncomerforSSESis,P=0.6kips.V'hisisofsimilarmagnitudetotheverticalthrustloadof0.7kipsonthedowncomerdurinqsteamblowdown(Ref10,Subsection4.2.3).LateralloadsonthedowncomersduringclearingareestimatedfromRef11,Table3-4tobelessthan3kips.4215DowncomerHaterJetLoadThewaterclearinqjetloadiscalculatedbasedontheapproachdevelopedinthedesignguides{Befs12and13).Thisloadisexperienceasadragloadbystructureslocatedwithinthe)etconebeneaththedowncomersandasa,jetimpingementloadbythebasemat.ThejetimpingementloadonthebasematiscalculatedfromRef10,Equation4-25,p.4-43.Pg=pfAvf2(4-4)Herepisthedensityofwater{takentobe62.4ibm/ft~),Aisthetotaljetimpingementareaandvistheattenuatedwatervelocitycorrespondingtothemaximumventclearinggetvelocity(Ref10).Figures4-41and4-42showelevationandplanviewsoftheSSESdowncomersandtheirassociatedjetcones.Theradiusofthejetconeatthebasematis2.69ft.andthetotalareainterceptedbythe87downcomersintheSSESwetwellis1978ft~.AsseeninFigure4-42thereisnosignificantoverlapofadjacentjetsonthebasemat.Theventclearingvelocityof60fpsisattenuatedbyafactorof0.68usinqthemethoddescribedinRef.10,'ubsection4.4.5.1toyieldavalueof40.8fpsatthebasemat.ThejetimpingementpressureiscalculatedfromRef10,Equation4-26,p.4-43tobeP=pfvIP=22m4ps'.~IUsingthevalueforAof1978ft~fortheSSESdesignthetotaldowncomerwaterjetimpingementloadonthebasematis4-9 F=2848.3kips.Thisloadactsverticallydownwardonthebasematfromthetimethebreakoccursuntil'thedovncomershavecleared,at0.6863sec(Ref7).4.2.1.6PoolswellAirBubbleLoadThepoolsvellairbubblepressureload'asitappliestothecontainmentwallsisdescribedinRef10,Subsection4.4.5.3.ThisloadisviewedasanincreaseinthehydrostaticpressureonthesuppressionpoolwallsbelovtheventexitplaneandiscausedbytheairbubblewhichhasbeenpurgedfromthedrywellintheinitialstagesoftheLOCA.Theairbubblepressuretransientcalculatedwiththepoolsvellmodel(describedinSubsection4.2.1.2)isshovninFigure4-43.Figure4-44shovsthenormalizedtotal.pressuredistribution(hydrostaticplusairbubble)toheappliedtothecontainmentasaresultofthisload.Thepressureonthewetvellwallsbetweentheventexitandthewatersurfacecontainsalineardecreaseto0.0psigatthewatersurface(Ref10,Subsection4.4.5.3).'hisloadasitappliestosubmergedstructuresisdescribedinRefs13and14.4.2.1.7PoolswellFallbackLoadThepoolswellfallback3.oadisadragloadvhichappliestoallstructuresbetweenthepeakpoolswellheight(el.690')andtheventexit(el.660').ThisloadiscalculatedforcomponentsinthisregionusingtheanalysisofSubsection4.4.54ofRef10Sincetheverticalstructuresareparalleltothefallbackflow,theyaresubjectedtonegligiblefallbackloads.(Forafallbackvelocityof30fpstheloadissignificantlylessthan1kip).Thedowncomerbracingstructureatelevation668'-0"is,however,perpendiculartothefallbackflowandvillundergoafallbackloadappliedverticallydovnvard.Thefallbackdragvelocityiscalculatedusingtheequationonpage4-45ofRef10.VFB='.82(8)</~(4-6)FortheSSESdesign,themaximumdowncomersubmergence,Ho,is12feetsothefallbackvelocityis34.05fps.ThedragpressureduetothisvelocityiscalculatedfromRef10,Equation4-24.tobePFB(psi)=78(4-7)whereCpistheappropriatedragcoefficientforthestructurebeingloaded.4-10 PallbackloadsarecalculatedusingRefs12and13.2.2Condensationoscillationsa~adchuin'oadsCondensationoscillationandchugginqloadsfollowthepoolswellloadsintime.Therearebasicallythreeloadsinthistimeperiod,i.e.,fromabout4to60secondsafterthebreak.Condensationoscillationisbrokendownintotwophenomena,amixedflowregiemeandasteamflowregieme.Themixedflowreqiemeisarelativelyhighmassfluxphenomenonvhichoccursduringthefinalperiodofairpurgingfromthedrywelltothevetvell.Thus,themixedflowthroughthedovncomerventscontainssomeairaswellassteam.Thesteamflovportionofthecondensationoscillationphenomenaoccursafteralltheairhasbeencarriedovertothevetwellandarelativelyhighmassfluxofpuresteamflowisestablished.Chuggingisapulsatingcondensationphenomenonwhichcanoccureither.follovingtheintermediatemassfluxphaseofaLOCA,orduringtheclassofsmallerpostulatedpipebreaksthatresultinsteamflowthroughtheventsystemintothesuppressionpoolAnecessaryconditionforchuggingtooccuristhatpuresteam.flowsfromtheLOCAvents.Chuqqingimpartsaloadingconditiontothesuppressionpoolboundaryandallsubmergedstructures.4.2.2.1CondensationOscillationLoadDefinitionTheloadspecificationforthemixedandsteamflowphasesofcondensationoscillationistakenfromAppendixAtoRef20.Themixedflovportionofthecondensationoscillationloadisspecifiedasasinusoidalloadatthecontainment'scriticalfrequencies,between2and7Hzvithanamplitudeofx1.75psi.Thisloadistobeapplieduniformlytothevettedportionofthesuppressionpoolboundarybelowtheventexitwithalinearattenuationtothefreesurfaceo'fthesuppressionpool.Thedurationofthisloadisfrom4to15secondsafterthebreakhasoccurred.Thesteamflowportionofthecondensationoscillationloadisspecifiedasasinusoidalloadatthecontainment'scriticalfreguenciesbetveen2and7Hzvithanamplitudeofa5.0psi.Theloadistobeapplieduniformlyto,thewettedportionofthe:suppressionpoolboundarybelowtheventexitwithalinear.attenuationtothesuppressionpoolfreesurface.Alsoasinusoidalloadofamplitudea0.5psiisapplieduniformlytothedrywellboundaryatcriticalfrequenciesbetween2and7Hz.Thedurationofboththedryvellandsuppressionpoolsteamflovcondensationoscillationloadisthetimeperiodfrom15to25secondsfollowinqtheinitial.break.Condensation'oscillationloadsonsubmergedstructuresarecalculatedusinqRefs12and13.4-11 ThepoolboundarychuggingloadisspecifiedinRef15Tvoloadinqconditionsaredescribed:symmetricandasymmetric.Thesymmetricloadinqconditionisspecifiedas+4.8psig/-4.0psigandistobeapplieduniformlyaroundtheentirepoolboundaryasshovn.inFigure4-45(extractedfromRef15).eTheasymmetricloadingconditionhasaspecifiedmaximumpositive/negativepressureof+20psig/-14psiqandhasthecircumfezentialspatialdistributiondepictedinFigure4-45.ChugqingloadsonsubmergedstructuresvillbeevaluatedwhenthedesiqnguidedealingviththeseloadsiscompletedThechuggingloadimpartedtothedowncomerwillbespecifiedwhentheappropriatedynamicforcingfunctionbecomesavailable4-23LONGTERMLOCALOADDEFINITIONTheloss-of-coolantaccidentcausespressureandtemperaturetransientsinthedrywellandwetvellduetomassandenergyreleasedfromthelinebreak.Thedryvellandwetvellpressureandtemperaturetimehistoriesarerequiredtoestablishthestructuralloadingconditionsi.ntheconta'inmentbecausetheyarethebasisforothercontainmenthydrodynamicphenomena.Theresponsemustbedeterminedforarangeofparameterssuchasleaksize,reactorpressureandcontainmentinit'ialconditions.TheresultsofthisanalysisaredocumentedinRef7.TheDBALOCAforSSESisconservativelyestimatedtobea3.53ft~breakoftherecirculationline(Ref7).'heSSESplantuniqueinputsforthisanalysisareshowninTable4-19.DrywellandvetvellpressureresponsesareshovninFigures4-46and4-47(extractedfromRef7)Thesetransientdescriptionsdonot,hovever,containtheeffectsofreactorsubcoolingSuppressionpooltemperatureresponseisshovninFigure4-48(Ref7~).Thistransientdescriptionalsodoesnotcontaintheeffectofreactorsubcoolinq.Dry'veiltemperaturezesponseisshowninFigure4-49andsimilarlydoesnotcontaintheeffectsofpipeinventoryorreactor'subcooling4.2.3.2IntermediateBreakaccident~IBA)TransientsTheworst-caseintermediatebreakfortheMarkXIplantsisamainsteamlinebreakontheorderof0.05to0.1ft~.AtthistimeplantuniqueIBA,data.forSSESisavailableonlyforthesuppressionpooltemperatureresponsetoa0.05ft>break(Ref7).ThisdataisshowninPiqure4-50.DrywelltemperatureandwetwellanddryvellpressuresfortheSSESXBAareestimatedfromcurvesforatypicalMarkIZcontainmentshowninFigure4-51(extractedfromRef10)4-12 Atthistimeplant-uniqueSBAdataforSSESisnotavailable.Thewetwellanddrywellpressureandtemperaturetransients'foratypicalNarkIIcontainmentareusedtoestimateSSEScontainmentresponsetotheseaccidents.ThesecurvesareshowninFigure4-52{extractedfromRef10).424LOCALOADINGHISTORIESFORSSESCONTAINNENTCONPONENTSThevariouscomponentsdirectlyaffectedbyLOCAloadsareshownschematicallyinFigures4-53and4-54.ThesecomponentsmayinturnloadothercomponentsastheyrespondtotheLOCAloads.Forexample,lateralloadsont'edowncomerventsproduceminorreactionloadsinthedryvellfloorfromwhichthe:downcomer'saresupported.ThereactionloadinthedrywellfloorisanindirectloadresultingfromtheLOCAandisdefinedbytheappropriatestructuralmodelofthedowncomer/drywellfloorsystemOnlythedirectloadinqsituationsaredescribedexplicitlyhere.Table4-20isaLOCAloadchartforSSES.ThischartshowswhichLOCAloadsdirectlyaffectthevariousstructuresintheSSEScontainmentdesignDetailsoftheloadingtimehistoriesarediscussedinthefollovingsubsections..424.1LOCALoadsontheContainmentWallandPedestalFigure4-55showstheLOCAloadinghistoryfortheSSEScontainmentwallandtheRPVpedestal.Thewetvellpressureloadsapplytotheunwettedelevationsinthewetwell;theappropriatehydrostaticpressureadditionismadeforloadsonthewettedelevations.Condensationoscillationandchuggingloadsareappliedtothewettedelevationsinthevetwellonly.Thepoolswellairbubbleloadappliestothevetwellbo'undariesasshowninFigure4-44.42.4.2LOCALoadsontheBasematandLinerPlateFigure4-56showstheLOCAloadinghistoryfortheSSESbasematandlinerplate.WetwellpressuresareappliedtothewettedandunwettedportionsofthelinerplateasdiscussedinSubsection4.2.4.1.Thedowncomerwaterjetimpactsthebasematlinerplateasdoesthepoolsvellairbubbleload.Chuggingandcondensationoscillationloadsareappliedtothewettedportionofthelinerplate.42.4.3LOCALoadsontaheDcwellandD~rwellFloorFiqure4-57showstheLOCAloadinghistoryfortheSSESdrywellanddrywellfloor.Thedrywellfloorundergoesaverticallyapplied,continuouslyvaryingdifferentialpressure,theupwardcomponentofwhichisespeciallyprominentduringpoolswellvhenthevetwellairspaceishighlycompressed4-13 Figure4-58showstheLOCAloadinghistoryfortheSSEScolumns.Poolswelldragandfallbackloadsareveryminorsincethecolumnsurfaceisorientedparalleltothepoolswellandfallbackvelocities.Thepoolswellairbubble,condensationoscillationsandchugqinqwillprovideloadsonthesubmerged{wetted)portionofthecolumns.4.2.4.5LOCALoadsontheDowncomersFiqure4-59showstheLOCAloadinghistory.fortheSSESdowncomers.Thedowncomerclearingloadisalateralloadappliedatthedowncomerexit{inthesamemannerasthechugginglateralload)plusaverticalthrustload.Poolswelldragandfallbackloadsareveryminorsincethedowncomersurfacesareorientedparalleltothepoolswellandfallbackvelocities.Thepoolswellairbubbleloadisappliedtothesubmergedportionofthedowncomerasarethechuggingandcondensationoscillationloads.4.2.4.6LOCALoadsontheDowncomerBracingFigure4-60showstheLOCAloadinqhistoryfortheSSESdowncomerbracinqsystem.Thissystemisnotsubjecttoimpactloadssinceitissubmergedatelevation668'sasubmergedstructureitissubjecttopoolswelldrag,fallbackandairbubbleloads.Condensationoscillationsandchuggingattheventexitwillalsoloadthebracingsystemboththroughdowncomerreaction{indirectload)anddirectlythroughthehydrodynamicloadinginthesuppressionpool.I4.2.47LOCALoadsonMetwellPipingFigure4-61showstheLOCAloadinghistoryforpipingintheSSESwetwell.SincethewetwellpipingoccursatavarietyofelevationsintheSSESwetwell,sectionsmaybecompletelysubmerqed,partiallysubmerged,orinitiallyuncovered.pipingmayoccurparalleltopoolswellandfallbackvelocitiesaswiththemainsteamsafetyreliefpiping.For,thesereasonsthereareanumberofpotentialloadinqsituationswhichariseasshowninTable4-21..Znaddition,thepoolswellairbubbleloadappliesto,thesubmergedportionofthewetwellpipingasdothecondensationoscillationandchuggingloads.'-14 43ANNULUSPRESSURIZATIONTheRPVshieldannulushastherecirculationpumpssuctionlinespassingthroughit{forlocationincontainmentseeFigure1-1).Themassandenergyreleaseratesfrom-apostulatedrecirculationlinebreakconstitutethemostseveretransientinthereactor,shieldannulus.Therefore,thispipebreakisselectedforanalyzingloadingoftheshieldwallandthereactorpressurevesselsupportskirtforpipebreaksinsidetheannulusThereactorshieldannulusdifferentialpressureanalysisandanalyticaltechniquesarepresentedinAppendices6Aand6BoftheSSESFinalSafetyAnalysisReport(FSAB)4-15 17.71ft0.883sec150:J'2wCO~10ZI0L,5IKCgfLwDXT30766.1104770.00.250.50.0.751.0TIMEAFTER'VENTCLEARING(SEC)SUSQUEHANNASTEANQLIKCTRICSTATIONUNITS1ANQ2DESIGNASSESSMENTREPORTSSESSHORTTERMSUPPRESSIONPOOLSURFACEHEIGHTFIGURE4-38
WETWELLPRESSURE(PSIA)OOOo0QQOVOl4CPOQCmPfllIIOQmZJHfllgmCXfm0QIllmXQZzZpC>COCOm~mCoCOCO~zZc+amrvfltll0OCO.Ozrm(mzoUl0IIIRQOooooo(0CO4l~0AtOMDK4oO(DRYWELL-WETWELL)QP(PSID) fP~yh'1lf,Hk' DOWNCOMERB.O.VENTPIPEEL.660'-0"PEDESTAL"iDIAPHRAGMSLABSUPPORTCOLUMN12'4"EL.648'-0<'ASEMATSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2:.DESIGNASSESSMENTREPORTBASEMATSSESWATERCLEARINGJETFIGURE441 tIIAA-,I1WA~"tt
/II///-/IIjlCOLUMNS///8X//ql/I//t/(////lfJETIMPINGEMENTAREA(22.73SQ.FT./VENT)(FROMDOWNCOMERWATERCLEARING)4~~~4~lj~hz~~++I4Pv~~'~eg~e.O4'~~~5~~ygI~ICONTAINMENTWALL~~~~I~r~~~SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTSSESJETIMPINGEMENTAREA(WATERCLEARING)FIGURE442
~IP.IAA'Il'\'~\C~rE CONTAINMEATWALLCOLUMNPEDESTAL0.0PSIG0.0PSIGEL.672'-0"PB(t)+HYDRO.STATICPB(t)+HYDRO.STATICEL.660'0"PB(t)+HYDROSTATICPB(t)+HYDROSTATICPB(t)+HYDROSTATICEL.6480BASEMATSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTAIRBUBBLEPRESSUREONSUPPRESSIONPOOLWALLSFIGURE444 30DRYWELL20CllCOLQscc10WETWELL-1OO10102TIME(seel103,104105(a)CONTAINMENTPRESSURERESPONSEFORINTERMEDIATEBREAKAREA300-200ss:DIIZ100IDRYWELL1OO10110TIME(sec)1031O4105(b)DRYWELLTEMPERATURERESPONSEFORINTERMEDIATEBREAKAREASUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTTYPICALMARKIICONTAINMENTRESPONSETOTHEIBAFIGURE4-51 CONTAINMENTWALL0K0ooo00000000~04A00~0:o0"ooooac,ooORoy.o0DOWNCOMERSCOLUMNS,0000~0q00'Ieoo~oooooo0ooooWETWELLPIPINGNOTE:DOWNCOMERBRACINGISONLYPARTIALLYSHOWNINTHEINTERESTOFCLARITY.LETTERSINDICATESRVQUENCHERSSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTSSESCOMPONENTSAF-FECTEOBYLOCALOADSFIGURE4-S3
~~'\~r~,IINI I~1.:~~,B.O.SLABEL.700'-3"B.O.HYDROGENRECOMBINEREL.691'-0"VACUUMBREAKEREl'.692'-1"IIT.O.PLATFORMEL.691'-0"MAXIMUMPOOLSWELLEL.690'-0"MAXIMUMPOOLSWELLHEIGHT~1.5XMAXVENTSUBMERGENCE~18'-0'C~~'RACINGEL.668'-0"HIGHWATERLEVELEL.672'.0"INORMWATERLEVEL'L.671'-0"IMMAXUMVENTSUBMERGENCE0tiB.O.VENTPIPEEL.660'-0"12'IAPHRAGMSLABSUPPORTCOLUMN0IIWETWELLPIPING3t6ttT.O.SLABEL.648'-0"SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORT.SSES.COMPONENTSAFFECTEDBYLOCALOADSFIGURE4-54
TABLE4-16LOCALOADSASSOCIATEDMITHPOOLSMFLLLoadReference1.Hetwell/DrywellPressuresduringPoolswell2.PoolswellImpactLoadsRef7,TableIV-D-3;Ref10,Subsec-tion4.,4.1.5Ref10,Subsec-tion4.4.6~3.PoolswellDragLoadsRef10,Subsections4452,4.4.7,4484..DowncomerClearingLoadsRef10,Subsection4.3.1,Reference11,Subsection3.3.1.25.DowncomerHaterJetLoad6.PoolswellAirBubbleLoadRef10,Sub-section4.4.5.1Ref10,Sub-section4.45.37.PoolswellFallbackLoadRef10,Sub-section4.4.5.4 TAB.LE4-18SSESPLANTUNIQUEPOOLSWZLLCODEINPUTD'ATADowncomerArea(each)SuppressionPoolFreeSurfaceAreaMaximumDowncnmerSubmerqenceDowncomerOverallLossCoefficientNumberofDowncomersInitialMetwellPressureHetwellFreeAiVolumeVentClearingTimePoolVelocityatVentClearingInitialDrywellTemperatureInitialDrywellRelativeHumidity2.96ft~506503ft~12.002.58715.45psia149,000ft~0.6863sec3.0ft/sec135oF020 0
TABLE4-19INPUTDATAFORSSESLOCATRANSIENTSDrywellfreeairvolume(includinqventsjMetwellfreeairvolumeNaximumdowncomersubmerqenceDowncomerflowarea(total)Downcomerlosscoefficient'Initialdrywe11pressureInitialwetwellpressureInitialdrywellhumidityInitialpooltemperatureEstimatedDBAbreaksizeNumberofventsInitialmassofsteaminvesselInitialmassofsaturatedwaterinvessel239,600fthm149r000ft>12.0ft256.7ft2.515.45psia15.V5psia205900P353ft~8724,5001bm674,0001bmMinimumsuppressionpoolmassInitialvesselpressureVessel6internalsmassVessel6internalsoverallheattransfercoefficient7.6x10~ibm1,055psia2,940,300ibm484.9Btu/secoFVesselandinternalsspecificheatInitialcontrolroddriveflowInitia1steamflowto'ainturbineRCIC6HPCI(HPCS)flowinitiationlevel,distancefromvessel"0"0123Dtu/1bmF1083ibm/sec3931.51bm/sec489.5in Tahle4-19~ContinuedgRCIC6HPCI(HPCS)flowshutofflevel(normalwaterlevel),distancefromvessel"0<<564.0inRatedRCICflowratetovesselRatedHPCI(HPCS)flowratetovesselRCICshutoffpressureHPCI(HPCS)shutoffpressureCondensatestoragetankentha1py'CRDenthalpyInitialpowerlevelPeedwaterenthalpyCleanupsystemflowCleanupsystemreturnenthalpyInitialvesselfluidenthalpyRHRheatexchanqer>>K~~inpoolcoolingmodeRHRheatexchangersteamflowincondensingmodeRHRheatexchangerflowinpoolcoolinqmodeRHRheatexchangeroutletenthalpyincondensinqmodeServicewatertemperature83.4ibm/sec695.ibm/sec165psia165psia48Btu/ibm48Btu/ibm3.23x10~Btu/sec78Btu/ibm36.94ibm/sec413.2Btu/ibm573.1Btu/1bm306Btu/sec~F25lbs/sec1390lbs/sec108Btu/lhm90~F CHAPTER5LOADCOMBINATIONSFORSTRUCTURES'IPING'NDEQUIPMENTTABLEOFCONTENTS51CONCRETE.CONTAINMENTANDREACTORBUILDINGLOADCOMBINATIONS52STRUCTURALSTEELLOADCOMBINATIONS53LINERPLATELOADCOMBINATIONS54DOMNCOMERLOADCOMBINATIONSPIPING'UENCHER'NDQUENCHERSUPPORTLOADCOMB'INATIONS5.5.1LoadConsiderationsforPipingInsidethe'Dryvell5.5.2LoadConsiderations.forPipingInsidetheMetwell5.5.3QuencherandQuencherSupportLoadConsiderations5.5.4LoadConsiderationsforPipingintheReactorBuilding5'NSSSLOADCOMBINATIONS57EQUIPMENTLOADCOMBINATIONS58FIGURES59TABLES5-1 CHAPTER5FIGURES~umber5-15-25-35-4TitlePipingStressDiagramsandTablesPipingStressDiagramsandTablesPipingStressDiagramsandTablesPipingStressDiagramsandTables5-2 CHAPTER5TABLESTitle5-1LoadCombinationsforContainmentandReactorBuildinqConcreteStructuresConsideringHydrodynamicLoads5-2LoadCombinationsandAllowableStre'ssesforStructuralSteelComponents5-3LoadCombinationsandAllowableStressesforDowncomers5-3 50LQRU~CQBINILTIO~SFORSTRUCTURESPIPINGINDEQUIPNENTToverifytheadequacyofmechanicalandstructuraldesign,itisnecessaryfirsttodefinetheloadcombinationstowhichstructures,piping,andequipmentmaybesubjected.Inadditiontotheloadsduetopressure,weight,thermalexpansion,seismic,andfluidtransients,hydrodynamicloadsresultingfromLOCAandSRVdischargeareconsideredinthedesignofstructures,piping,andequipmentinthedrywellandsuppressionpool.ThischapterspecifieshowtheLOCAandSRVdischargehydrodynamicloadswillhecombinedwiththeotherloadingconditions.Zortheloadcombinationsdiscussedinthischapter,seismicandhydrodynamicresponsesarecombinedhythemethodsspecifiedinRef.10Subsection5.2.2andRef10Section'.3.5-4 53LINERPLATELOADCOMBINATIONSThelinerplateandanchoragesystemaredesignedfortheloadcombinatiomslistedinTable5-1exceptthatallloadfactorsaretakenasunity.5-7
~5.4DIIRHCO~NRLOANCOMBINATIONSI.oadcombinationsforthedowncomersaregiveninTable5-3.'heseloadcombinationsarebasedontheloadcombinationsgiveninTable6-1ofBef10.5-8 55PTPIN~G~UEQCH~E~A~NDUgNCHERSUPPORTLOADCOMBINATIONST.OCAloadsconsideredonpipingsystemsincludepoolswellimpactloads,poolswelldragloads,downcomerwaterjetloads,poolswellairbubbleloads,,fallbackdragloads,condensationoscillationloads,chuqginqloads,'ndinertialloadingduetoaccelerationofthecontainmentstructureproducedbyLOCAloads.LoadsduetoSRVdischargeonpipingsystemsincludewaterclearingloads,airclearing.loads,fluidtransientloadsonSRVdischargepiping,reactionforcesatthequencher,andinertialloadingduetotheacclerationofthecontainmentstructureproducedbySRVdischargeloads.TheloadcombinatioasandtheacceptancecriteriaforpipingsystemsaregiveninTab1e6-1ofRef10.55.1LoadConsiderationsforPi~in@InsidetheDrywellPipingsystemsinside%hedrywellaresubjectedtoinertialloadinqduetotheacceleratioaofthe.containmentproducedbyLOCAandSRVdischargeloadsinthewetwell.'heSRVdischargepipinginthedrywellisalsosubjectedtofluidtransientforcesduetoSRVdischarqe.55.2LoadConsiderationsfor~piincn'nsidetheMetwellAllpipinginthewetwellissubjecttotheinertialloadingduetoLOCAandSRVdischarge.DragandimpactloadsduetoLOCAandSRVdischargeonindividualpipesinthewetwelldependonthephysicallocationofthepipinq.OtherSRVdischargeandLOCAloadsapplicabletopiping-inthewetwellarediscussedintheparagraphsthatfollow.PipingsystemslocatedbelowthesuppressionchamberwaterlevelareshownonFigures5-1and5-2.Theselinesarelocatedoutsideofthejetimpingementconeofthedowncomer.Inadditiontotheinertialloads'hesepipingsystemsaresubjecttoairbubbleloads,condensationoscillationldads,andchuggingloadsduetoLOCAandSRVoperation.TheSRVpiping,quencher,andquenchersupportarealsosubjecttofluidtransientforcesduetoSRVdischarge.PipingsystemswithinthepoolswellvolumeareshowncnFigures5-2,5-3and5-4.Allhorizontalrunsofthesepipesareabovethesuppressionchamberwaterlevel.Thefollowingloads,inadditiontoinertialloads,actonthesesystems:a.'Thehorizontalrunsofpipebelowelevation690',experiencepoolswellimpact,poolswelldrag,andfallbackdraqloads.5-9 b.Thevertical,portionsofpipeinthewaterbelowelevation690'-experience"poolswelldragandfallbackdragloads.~55.3nenc~hrandquenchersn~ortLoadconsiderationsThequencherandquenchersupportsaresubjectedtothe'ollowinqhydrodynamicloadsinadditiontothepressure,weight,thermal,andseismicloads:a.UnbalancedloadsonthequencherduetoSRVwaterclearinqandairclearingtransients,irregularcondensation,andsteadystateblowdownb.DragloadsduetoSRVdischargeandLOCAc.SRVpipinqendloadsd.InertialloadingduetotheaccelerationofthecontainmentproducedbySRVdischargeandLOCA.5.5.4LoadConsiderationsforPipinginthetheReactorBuildingTheeffectsoftheinertialloadingduetoaccelerationofthecontainmentproducedbySRVdischarqeandLOCAloadswillbeevaluatedforthispiping.5-10 56NSSSLOADCOMBINATIONSTobeprovidedlater.5-11 57EQUXPMZNTL'OADCONBINATZONS'1Loadcombinationsforsafety-relatedequipmentlocatedwithinthereactorbuildingandcontainmentwillbeassessedanddescribedinarevisiontothis'DesignAssessmentReport("Safety-related"isdefinedinTable1.8-1ofthePSAR).5-12 FIG.NO.LINENO.12"-GBC-10112"-GBC-101QTY10SYSTEMM.S.R.V.DISCHARGEPIPINGR.F.C.M.E.L.S.J.B.D.P.N.G.K.A.H.ELACS1'-0"CS1'4I"SI.EEVEPENETRATIONEL704'-0"~~IIII~4.~I'1.0.EL694'-0"HIGHWATERLEVEL!TWODIRHORZ84TORSIONALREST.EL672'4"ANCHOREL694'4"HEIGHTEL,680'L668'-0"TWODIRHORZREST.s0'a,o.'AELAEL649'4I"TWODIRHORZ64TORSIONALREST.FIGUREAFIGUREBBOTTOMSUPPRESSIONPOOLEL040'4I"SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTPIPINGSTRESSDIAG1'A@8ANDTABLESFIGURE6-3" FIG.NO.QTYLINENO.SYSTEMELAEl.BEl.CRADYDIM.XREST.EL6"-GBB-120RHRMS'-61/2"M6'4"897'4"12'43/6"1S6/8"697'-0"696'4"QRPVRADYELC24VERTICAL5iAXIALREST.ELB~'ERTICALREST.POOLSWELLEL690'4"ELAIHIGHWATEREL672'A"DIM.Xa%~~4aEL048'4I"SUSQUEHANNASTEAMELECTRICSTATIONUNITt1AND2DESIGNASSESSMENTREPORT=IMPINGSTRESSDIAGRAMSANDTABLES'IGURE$4
TABLE5-1LOADCOMBINATIONSFORCOHTAINMEHTANDREACTORBUILDIHGCONCRETESTRUCTURES{CONSIDERINGLoad.EquationConditionDLPTR000SSpARSRV<>>AOTVADSASYMSingleValveLOCA~>>1Normalw/oTemp2Normalw/Temp.3NormalSev.Env.141710101310101010101010101251513125X(tl4Abnormal1.01.04aAbnormal1.01.01.25-1010125101012510XX5AbnormalSev.Env,1.01.01010XISaAbnormal"SevEnv6NormalExt.Env.101010101010101011101010107AbnormalExt.Env.7aAbnormalExt.Env.10101010010101.01.01.01.01010101010XILoadDescritionD-"DeadLoadsL.=LiveLoadsT=OperatingTemperatureLoads0R=OperatingPipeReactions0P.=OperatingPressureLoads0SRV=SafetyReliefValveLoadsE=Operating-BasisEarthquake0ES=SafeShutdownEarthguakeSSPB=SBAorIBA(LOCA)PressureLoadP=DBA(LOCA)PressureLoadAT=PipeBreakTemperatureLoadAR=PipeBreakTemperaturesReactionLoadsAR=Reactionandjetforcesassociatedwiththepipebreak~Noe1)Xindicatesapplicabilityforthedesignatedloadcombination.
2)porthecoluansdesignatedAOT,ADS,ASYH,andSingleValve,onlyoneofthefourpossiblecolunnsnaybeincludedintheloadconbinationforanyoneequation.Forexanple,inequation1eitherAOTorASYNnaybeconsideredwiththeotherloadsbutnotbothAOTandASY5sinultaneously.3)LOCAchuggingandcondensationoscillationloadswillbeincludedinasubsequentrevisiontothistable TABLE5-2LOADCOMBINATIONSANDALLOWABLESTRESSESFORSTEELSTRUCTURALCOMPONENTS{SuppressionChamberColumns,DouncoaerB~racin~andReactorBuildingStructuralSteel)~cCuation.ConditionNormalw/oTemp.Normalw/TempNormal/SevereLoadCombinationD+L+SRVD+L+T+SRV0D+L+T+E+SRV0Stress,'LimitF15FSNormal/ExtremeD+L+T+E'+SRV01.5FSAbnormalD+L+P+(T+T)+R+SRV(Note1)6Abnormal/SevereD+L+P+(T+T)+R+E(Note1)+SRVAbnormal/ExtremeD+L+P+(To+Ta)+R+E+SRV(Note1)Note1:Innocaseshalltheallowablestressexceed0.90Finbendinq,0.85Finaxialtensionorcompression,5nd0.50F>inshear.Wherethedesignisgovernedbyrequirementsofstability(localorlateralbuckling),theactualstressshallnotexceed1.5FS.
TABLE5-2~ContinuedgNotations:Fq=AllowablestressaccordingtotheAISC,"SpecificationfortheDesign,Fabrication,andErectionofStructuralSteelforBuildings",dated1969,Part10DeadloadLiveloadTpTaThermaleffectsduringnormaloperatingconditionsincludingtemperaturegradientsandequipmentandpipereactions.Addedthermaleffects(overandaboveop'cratingthermaleffects)whichoccurduringadesignaccident.DesignBasisAccidentpressureloadLocalforceorpressureonstructureduetopostulatedpiperuptureincludingtheeffectsofsteam/watergetimpingement,pipe.whip,'andpipereaction.ELoadduetoOperatingBasisEarthquake.LoadduetoSafeShutdownEarthquake.SHVFySafetyreliefvalveloads.Ninumumspecifiedyieldstength TABLE5-3,LOADCOMBINATIONSANDALLOMABLESTRESSESFORDOMNCOMERS~EuatgonConditionUpsetEmerqencyEmergencyFaultedFaultedLoadCombinationD+P+SRVALL0D+.PSRVALL+E0D'PSBA'SRADS'E'BAD~P+SRVALL+E0D+PIBA+SRVADS+E+IEAPrimaryStressLimit1SSm2.25S225Sm3Smm7Notations:FaultedFaultedFaulted0+PS/A[orPjBA)(orBA)D+PA+E'+DBA1D+PA+E'DBA2m3S3SSmDP0MaximumallowablestressaccordingtoTableI-10.1,Ref28.DeadweightofthedowncomerPressuredifferentialbetweendrywellandsuppressionchamberduringnormaloperating,condition.SBAIBA'ASRVALLSRVADSPressuredifferentialbetweendrywellandsuppressionchamberdurinqSBA.PressuredifferentialbetweendrywellandsuppressionchamberdurinqIBA.PressuredifferentialbetweendrywellandsuppressionchamberduringDBA.Dynamiclateralpressureandinertialoadduetothedischargeofall16safetyreliefvalvessequentially.Dynamiclateralpressureandinertialoadduetothedischargeofall6ADSsafetyreliefvalvessimultaneously.LoadduetoOperatingBasisEarthquake ElSBALoadduetoSaeShutdownEarthquakeChuqqingloadsduetoSBAasfollows:.1.Horizontalloadatbottomofdowncomer,and2.Horizontalandverticalinertialloads.ZBAChuqqingloadsduetoIBAasfollows:1.Horizontalloadatbottomofdowncomer,and2.Horizontalandverticalinertialloads.DBA1Verticalloadsdueto:/1.Viscousandpressureforcesexertedbytheflowingsteam,andDEA22.InertialloadduetoDBAChuqqingloadsduetoDBAasfollows:1.Horizontalloadat'bottomofdowncomer.and2.Horizontalandverticalinertialloads.
CHAPTER6DESIGNCAPABILITY'ASSESSMENTCRITERIATABLEOFCONTENTS61CONCRETECONTAINMENTANDREACTORBUILDINGCAPABILITYASSESSMENTCRITERIA6.16.1.2ContainmentStructureCapabilityAssessmentCriteriaReactorBuildingCapabilityAssessmentCriteria.62STRUCTURALSTEELCAPABILITYASSESSMENTCRITERIA63LINERP'LATECAPABILITYASSESSMENTCRITERIA6'DOWNCOMERCAPABILITYASSESSMENTCRITERIA65PIPENGiQUENCHERANDQUENCHERSUPPORTCAPABILITYASSESSMENTCRITERIA66NSSSCAPABILITYASSESSMENTCRITERIA67EQUIPMENTCAPABILITYASSESSMENTCRITERIA6-1 60DESIGNCAPABILITYASSESSMENTCRITERIAThecriteriabywhichthedesigncapabilityisdeterminedarediscussedin.thischapterDesignoftheSSESisassessedasadequatewhenthedesigncapabilityofthestructures,piping,andequipmentisgreaterthantheloads(includingLOCAandSRVdischarge)towhichthestructures,piping,andeguipmentaresubjected.LoadingcombinationsarediscussedinChapter5.ThemarginsbywhichdesigncapabilitiesexceedtheseloadingsarediscussedinChapter7,DesignAssessment.6-2 63LENEPLATECAPABILITYASSESSMENTCRITERIAThestrainsinthelinerplateandanchoragesystem(weldsandanchors)fromself-limitingloadssuchasdeadload,creep,.shrinkage,andthermaleffectsarelimitedtotheallowablevaluesspecifiedinTableCC-3720-1ofRef29,andthedisplacementsofthelineranchoragearelimitedtothedisplacementvaluesofTableCC-3730-'IofRef29.Primarymembranestressesinthelinerplateandanchoragesystem(weldsandanchors)frommechanicalloadssuchasSRVdischargeandchuggingarecheckedaccordingtoSubsectionNE-3221.1ofRef28.PrimaryplussecondarymembraneplusbendingstressesarecheckedaccordingtoSubsectionNE-3222.2ofthesamecode.ZatiguestrengthevaluationisbasedonSubsectionNE-3222.4.Allowabledesignstressintensityvalues,designfatiguecurves,andmaterialpropertiesusedconformtoSubsectionNA,AppendixofRef28.Thecapacityofthelinerplateanchorageislimitedbyconcretepull-outtotheserviceloadallowablesofconcreteasspecifiedinRef30.6-5 TheallowablestressesforthedovncomersaregiveninTable5-3.TheseallowablestressesareinaccordancewithRef28;SubsectionNE.Aspermittedby,SubsectionNE-1120forMCcomponents,thedowncomersareanalyzedinaccordancevithSubsectionNB-3650ofRef28;however,theloverallovablestresses,Sm~fromTableI-10.1forNCcomponentsare"usedvhenperformingtheanalysis.
65PIPINGQUENCHERANDQUENCHERSUPPORTCAPABILITYASSESSNENTCRITERIAPipinginthecontainmentandreactorbuildingisanalyze'dinaccordancewithRef28SubsectionsNB3600,NC3600,andND3600fortheloadingdescribedinSection5.5.ThequencherisdesignedinaccordancewithRef28,SubsectionNC3200,forloadingdiscussedinSubsection5.5.3.The,quenchersupportisdesignedinaccordancewithSubsectionNF3000-ofRef28.6-7 Tobeprovidedlater.6-8 67EUIPMENTCAPABILITYASSESSMENTCRITERIAAssessmentcriteriaforsafety-relatedequipmentsubjecttoLOCAandSRVdischargeloadingwhichislocatedwithinthecontainmentandreactorbuildingwillbedescribedinarevisiontothisDesignAssessmentReport("Safety-related".isdefinedinTable1.8-1ofthePSAH)4 CHAPTER7DESIGNASSESSMENTTABIEOPCONTENTS71117.1.127.127-12171.227'.1.37.1.471.S'1671772DESIGNCAPABILITYMARGINS73PIGURES71ASSESSMENTMETHODOLOGY7.1.1ContainmentandReactorBuildingAssessmentMethodologyContainmentStructureAssessmentMethodologyReactorBuildingAssessmentMethodologyStructuralSteelAssessmentMethodologySuppressionChamber.ColumnsAssessmentMethodologyDovncomerBracingAssessmentMethodlogyLinerPlateAssessmentMethodlogyDowncomerAssessmentMethodologyPipingandSRVSystemsAssessmentMethodologyNSSSAssessmentMethodlogyEguipmentAssessmentMethodology7-1 CHAPTER7PIGURES'Nu~be7-17-27-37-47-57-6Title.GeometryPlotofContainmentStructureModelEquivalentModelDampingRatiovs.Model'FrequencyStructuralStiffness-Proportional-DampingFiniteElementModelofColumn1LinerPlateLoads-NormalConditionLinerPlateLoads-AbnormalConditionDowncomerAnalyticalModel7-2 70DESIGN'ASSESSNENTJ.oadsonSSESstructures,piping,andequipmentaredefinedinChapter4.ThemethodsbywhichtheseloadsarecombinedarediscussedinChapter5ThecriteriaforestablishingdesigncapabilityarestatedinChapter6.Thischapterdescribestheassessmentoftheadequacy.oftheSSESdesiqnbycomparingdesigncapabilitieswiththeloadingsto.whichstructures,piping,andcomponentsaresubjectedand~demonstratingtheextentofthedesignmargin.Thefirstsection,ofthischapterdiscussesthemethodologybywhichdesign'capabilityandloadsarecompared.Thesecondsectionindicates'theresultsofthesecomparisons.7-3 71ASSESSMENTMETHODOLOGY7.11ContainmentandReactorBuildinAssessmentMethodolo~'ThedynamicanalysisforthestructuralresponseofthecontaiamentandinternalstructuresduetotheSRVdischargeloadsandLOCArelatedloadsisperformedusingthefiniteelementmethod.TheANSYSfiniteelementcomputerprogramischosenforthetransientdynamicanalysis.Figure7-1showstheANSYSfiniteelementmodel.Platshellelementsareusedtomodelthereinforced-concretecontainmentstructureandthereactorvessel.Pipeelementsareusedtomodelthecolumnssupportinqthediaphragmslab.The.soilstructureinteractionistakenintoconsiderationbymodelling'hesoilusingaseriesofdiscretespringsanddampers.inthreedir'ectionsasshownonFigure7-1.ThesediscretesprinqsanddampersarespecifiedbasedontheformulaeforlumpedparameterfoundationsfoundinRef.32.TheANSYSprogramusesstiffness-proportional-damping,implyingastructuraldampingmatrixinthefollowingform:{C)=g{K)~whereCKDampingMatrixastiffness-proportionaldampingconstantStiffnessMatrix,Fiqure7-2showstheequivalentmodaldampingratioversusthemodalfrequencyforstructuralstiffness-proportional-dampingAvalueofgequaling000063isusedintheANSYSmodelwhichcorrespondstoastructuralmodaldampingofapproximately4percentofcriticalat20Hz.Twocomputerprogramshavebeendeveloped,oneasapreprocessorandtheotherasapostprocessortotheANSYScomputerprogram.Thepreprocessortransformsthepressureforcingfunctionsactingonthesuppressionpoolwalls,basemat,andpedestalintoa~concentratedforceactiagattheassociatednodesoftheANSYSmodel.ThepostprocessorcalculatestheaccelerationtimehistoryfromthedisplacementtimehistoryobtainedbyANSYSandscansforthemaximum.displacementsandaccelerations.Accelerationtimehistories,maximumstructuraldisplacements,accelerations,andbroadenedaccelerationresponsespectraatselectednodesanddirectionsaredevelopedfortheanalysisofthepiping,equipment,andNSSSsystems.ResponsespectracurvesaredevelopedforallthepreviouslymentionedSRVdischargeandLOCAloads.7-4 Theresponsespectraarefurnishedforfourdifferentspectraldampingvaluesie,0.5,,1,2,and5percentofcriti'cal.Eachspectrumhasbeenbroadenedtoaccountfortheuncertaintiesinthestructuralmodelingtechniquesandmaterialproperties;Allspectralaccelerationsareexpressedinunitsofg(thegravitationalconstant).AppendixBcontainsexamplesofthebroadenedresponse"spectracurvesdevelopedforthedifferentloadingcasesofSRVdischargeandLOCArelatedloads.(Thepressuretimehistorysho~non'iqure4-29isusedasthebasisfortheexamplesgiven)TheANSYSprogram(stresspass)isalsousedtocomputetheforcesandmomentsduetotheSRVdischargeandLOCArelatedloads.TheseforcesandmomentsarethencombinedwiththenonhydrodynamicloadsinaccordancewithTable5-1.Material.stressesatthecriticaldesignsectionsintheprimarycontainme'ntandinternalconcretestructuresareanalyiedusing.theCECAPcomputerproqram(RefertoAppendixAtoFSARSection3.Q).Concretecrackingisconsideredintheanalysisofreinforcedconcretesections.TheconstructionoftheSSESreactorbuildingissuchthatnodirectcouplingwiththecontainmentoccurs.A2in.separation)ointiskeptbetweenthecontainmentstructureandthereactorbuildinqatallpointswherethetwostructuresabut,exceptatthebaseslabswhereacoldjointexists.Thisarrangementminimizesthetransferofanydirectdynamicresponsetothereactorbuildingfromthecontainment,wheretheSRVdischarge,andLOCAhydrodynamicloadsoriginate.Theaveragehorizontalandverticalbaseaccelerationsfromthecontainmentdynamicanalysisarecomputedandusedasinputmotionsonthereactorbuildingfoundations.Thisresultsintwohorizontalmotionsandoneverticalmotion.Theinputmotionsareusedintheformofaccelerationtimehistoriesatthebaseslab.Reactorbuildinqseismicmodels(horizontalnorth-southandeast-vestandvertical),asshownonPSARFigures3.7-9throuqh3.7-11andexplainedindetailinSubsection3.7.2.1boftheFSAR,areusedinthestructuralresponseanalysisduetoSRVdischargeandLOCAloads.AppendixCprovideseXamplesofthebroadenedresponsespectracurvesforthereactorbuildingduetoSRVdischargeloadsfortheabnormaloperatingtransient(AOT)caseatselectedlocations.ThepressuretimehistoryshowninPiqure4-29isusedasthebasisfortheexamplesgiven).TheresponsespectracurvesaredevelopedforuseinthedesignofpipingandNSSSsystems.Theresponsespectraarefurnishedforfourdifferentspectraldampinqvalues,ie,0.5,,1,2,and5percentofcritical.Eachspectrumhasbeenbroadenedtoaccountforthe7-5 uncertaintiesinthestructuralmodellingtechniquesandmaterialpropeties.Al'lspectralaccelerationsareexpressedinunitsofq(thegravitationalconstant).TheforcesandmomentsduetoSRVdischargeandLOCAloadsarecombinedwiththenon-hydrodynamicloadsinaccordancewithTable5-1.7.-12SructuralSteelAssessmentNethodolo~471.21SuressionChamberColumnsAssessmentNethodol~oTheassessmentmethodsusedfornon-hydrodynamicloadssuchasdead,live,pressure,temperature,seismic,andpipe,ruptureloadsaredescribedintheFSAR,Section3.8..3.45.Fortheanalysisofthecolumnsforhydrodynamicloads,theAHSYScomputerprogramisusedAtypicalcolumnismodelledasa,fixed-endedbeamasshownonFigure7-3.Thetotallengthofthecolumnis'ividedintobeamfiniteelements)oinedatnodepoints.Aneffectivewatermassduetosubmerqenceisconsidered.Dynamichorizontalforcesareappliedtothecolumnatthenodepointsbelowwaterlevel.Time-varyingforcesandmomentsinthecolumnarecalculatedforeachfiniteelement.Theseresultsarecombinedwiththosefornon-hydrodynamicloadstodeterminethet'otalforcesandmomentsinthecolumn.Axialloadsareproducedinthebracingduetolateralloadingonthedowncomers.SeeSubsection7.1.4foradescriptionoftheanalysisofthedowncomersforlateralloads.Todeterminethemaximumaxialloadinthebracing,lateralloadsareassumedtooccuronalldowncomerswithina90degreeinfluencezonein.ei.thertheradialortanqentialdirections.Bracingforthe16SRVdischargepipesisincludedwiththedowncomerbracing.Aslidingsupportisprovidedattheconnectionofthebracingtothedischargepipetoallowthedischargepipetomoveverticallywithoutproducingareactionloadonthebracing.Sincetheselateralloadsonthedowncomersduetoseismicandhydrodynamicloadsarerandomlyoriented,variouscombinationsofloaddirectionsareconsideredinordertodeterminethemaximumaxialloadinthebracinq.Xnadditiontotheaxialload,therearelateralpressuresappliedalongthelengthofthebracingmembersduetodirecthydrodynamicloadingSincethebracingmembersareofvaryinglengths,severaldifferentlengthsofbracingmembersareconsideredfortheanalysis.Stressesinthedowncomerbracingduetoequivalentstaticlateralpressuresarecalculatedusingclassicalbeamtheoryequations.StressesinthedowncomerbracinqduetodynamiclateralpressuresarecalculatedusingtheANSYScomputerprogram..Thetotallengthofthebracingmemberisdividedintobeamfiniteelementsjoinedatnodepoints.Aneffectivewatermassduetosubmergencewillbeconsidered.Dynamiclateralforcesareappliedtothebracingatthenode points.,Time-varyingforcesandmomentsinthebracingmemberarecalculatedforeachfiniteelement.Maximumstressesarecalculatedfromtheseresultsusingclassicalbeamtheoryequations.7.1.3LinerPlateAssessment~ethodolocCyPSARSubsection3.8.1providesadescriptionofthe.linerplateandanchoragesystemforthecontainment.Theanalysisofthelinerplateandanchoragesfornon-hydrodynamicloadsisinaccordancewithRef18Fortheanalysisofthelinerplateandanchorages-forhydrodynamicsuctionpressureloads,theloadonthelineristhenetnegativepressureload.ThenetnegativepressureloadequalsthedynamicnegativepressureviuetoSRVactuationorLOCAchuggingminusthestaticpositivepressureduetohydrostaticpressureorLOCA..Pigures7-4and7-5describetheloadsonthebasematandsuppressionchamberealllinerplateforthenormalandabno'rmalloadcombinationsrespectively.Porthenormalcondition,thehydrostaticpressureonthebasematis10.4psiandthemaximumnegativepressureduetotheactuationofallSRV'sis7.8psi.ThedistributionofthesepressuresonthesuppressionchamberwallisshowninFigure.7-4.Portheabnormalcondition,thetotalpositivepressureonthebasematis35.4psiwhichconsistsof10.4psifromhydrostaticpressureplus25.0fromLOCA(smallorintermediatebreakaccident).Thetotalmaximumnegativepressureonthebasematis21.8psiduetotheasymmetricchuggingload.ThemaximumnegativepressuresfromSRVactuationandchuggingarecombinedforconservatism.Itisrecognizedthattheprobabilityofthesetoophenomenaproducingpeak,negativepressuresatthesametimeisverylosThedistributionofpressuresonthesuppressionchamberwallisshowninFigure7-5.Sincethenegativepressureismorethanbalancedbythepositivepressure,theliner-platedoesnotexperienceanynetnegativepressure.Therefore,therearenoflexuralstressesinducedinthe'inerplate.7.1.4DovncomerAssessmentMethodolo~Stressesinthedovncomerpipesduetostaticloads,suchasdead@eightandpressure,arecalculatedusingclassicaleguations.Stressesinthe'dosncomerpipesduetoinertialloadscausedbyseismicandhydrodynamicloadsarecalculatedusingtheresponsespectrummethod.TheANSYScomputerprogramisusedtosolvefor7>>7 themodeshapesandfrequenciesofthedovncomersandthedovncomerbracing.Agroupofdovncomerpipesandbracingmembersisrepresentedbyalumpedmassmodel.Theinertiaeffectofthewatersurroundingthesubmergedportionofthedovncomersisapproximatedbytheadditionofaneffectivevatermass.ThemassofwaterinsidethedowncomersisincludedinthemodelforalldynamicloadingsexceptLOCA.FortheLOCAconditions,thewaterhasbeenventedfromthedovncomersandthereforeitisnotincludedinthemodel.TheANSYScomputerprogramisusedtocalculatethestressesinthedowncomerpipesduetohydrodynamiclateralloads.AtypicaldovncomerpipeismodelledasshowninFigure7-6.PointAatthetopofthedovncomerisrestqainedto'epresentthefixityofthedowncomeratthedryvellfloor.PointBislaterally.restrai'nedtorepresentth'elateralsupportfurnishedbythedowncomerbracing.Thetotallengthofthedowncomerisdividedintobeamfiniteelements)oinedatnodepoints.Dynamichorizontal-forcesareappliedtothedowncomeratthenodepointsbelowwaterlevel.Time-varyingforcesandmomentsinthedovncomerarecalculatedforeachfiniteelement.Maximum.stressesarecalculatedfromtheseresultsusingclassical.beamtheoryequations.ThepipingandSRVsystemsvillbeanalyzedfortheloadsdiscussedinSection5.5usingBechtelcomputerprogramsHE101andME632.TheseprogramsaredescribedinFSARSection3.9-StaticanddynamicanalysesofthepipingandSRVsystemsareperformedasdescribedintheparagraphsbelow.Staticanalysistechniquesareusedtodeterminethestressesduetosteadystateloadsand/ordynamicloadshavingequivalentstaticloads.Thedragandimpactloadsareappliedasequivalentstaticloads.ResponsespectraatthepipinganchorsareobtainedfromthedynamicanalysisofthecontainmentsubjectedtoLOCAandSRVloading.PipingsystemsarethenanalyzedfortheseresponsespectrafollowingthemethoddescribedinRef19.
TimehistorydynamicanalysisoftheSRVdischargepipingsubjectedtofluidtransientforcesinthepipeduetoreliefvalveopeningisperformedusingBechtelcomputercodeHE632.71.6SSSAssessmentNethodol~ogTobeprovidedlater~71.7.simesmsesseemtNe~hodolocpyAnalysismethodologiesforsafety-relatedequipmentwithinthecontainmentandreactorbuildingsubjecttoLOCAandSBVdischargeloadingwillbedescribedinarevisiontothisDAR(<<Safetyrelated<<isdefinedinTable1.8-1oftheFSAR)7-9 72DESIGNCAPABZLXTYMARGINSStressesatthecriticalsectionsforeachoftheab'ovestructures,piping,andeguipmentvillbeevaluatedforalltheloadinqcombinationspresentedinChapter5.'TheresultsofthestructuralassessmentofthecontainmentandsubmergedstructuresvillbesummarizedinAppendixA.-'FigureA-2showsthedesignsectionsinthebasemat,containmentwalls,reactorpedestal,andthediaphragmslabconsideredinthestructuralassessment)ThetablesofAppendixAatpresentgivethecalculateddesignmarginsforloadcombinationEguation1ofTable5-1whichappliestothepreviouslymentionedstructural>components.Similartablesvillheincludedinafuturerevisionofthisreportinordertopresentthefullassessmentofthedesigncapabilitymarginforalltheotherloadcombinations.Thereinforcinqsteelandconcretegualitycontroltestresultsshovthatmaterialstrengthsarehigherthantheminimumspecifiedvaluesusedincomputingthesemargins.Thisconservatism,alongwiththeoverloadfactorsintheloadcombinationsgiveninTable5-1andthematerialunderstrengthfactorsbuiltintotheallovablestresscriteria,resultsinactualsafetymarginsqreaterthatthosegiveninthetablesofApp'endixA.TheresultsofthestructuralassessmentofthereactorbuildingvillbesummarizedinAppendixE.TheresultsoftheanalysisofthepipingsystemsvillbesummarizedinAppendixFintheformoftables.Thesetablesvillprovidethemaximumstressforthecriticalloadcombination,theallowablestress,andthedesignmarginTheresultsoftheassessmentoftheNuclearSteamSupplySystem(NSSS)villbesummarizedinAppendixG.TheresultsoftheassessmentofequipmentvillhesummarizedinAppendixH.7-10 RPVNOTE:X-AXISISINPLANTEWANDY-AXISINPLANTNSDIRECTIONRPVSHIELDCONTAINMENTRPVPEDESTALSUSQUEHANNASTEAMELECTRICSTATION'UNITS1AND2DESIGNASSESSMENTREPORT.GEOMEYRYPLOTOFCONTAINMENTSTRUCTUREMODELFIGURE7-1 DRYWELLFL.Ogp.~vrrriv~j5sQ-DECK3lt42"PSTEELPIPECOL.FINITEELEMENTSL~52'-3"NODEPOINTSHIGHWATERLEVEL24'-0"BASEMAT(VdpiI0io44.jgiCMODELSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTFINITEELEMENTMODELOFCOLUMNFIGURE7-3 PEDESTALCONTAINMENTWALLHYDROSTATIC24'10.4psi+10.4psi.BASEMATSRVIII18'7.8psi-7.8ps<TOTAL18'2.6psi+2.6psiSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLINERPLATEPRESSURESNORMALCONDITIONFIGURE7C POSITIVEPEDESTALHYDROSTATICCONTAINMENTWALL+10.4PSI+10.4PSIBASEMATWETWELLPRESSUREDUETOSBAORIBAD+25PSINOTE:WETWELLPRESSUREDUETODBA=34PSI+25PSITOTAL+35,4PSI+35.4PSISUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLINERPLATEPRESSURESABNORMALCONDITIONFIGURE7-5SHEET1OF3
RPVNOTEX~AXISISINPLANTEWANDY-AXISINPLANTNSDIRECTIONRPVSHIELDCONTAINMENTRPVPEDESTALSUSQUEHANNASTEAMEI.ECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORT3-DCONTAINMENTFINITEELEMENTMODEL(ANSYSMODEL)fIGURE7 DAMPINGRATIOP~0.000630.080.070.060.050.040.030.020.01]02040FREQUENCYSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTEQUIVALENTMODALDAMPINGRATIOVSMODALFREQ.FORSTRUCTURALSTIFFNESSPROPORTIONALDAMPINGFIGURE7-2
NEGATIVEPEDESTALCONTAINMENTWALL<<VADS-7.8PSI-7.8PSIBASEMAT12'HUGGINGASyM-14PSI-14PSITOTAL-21.8PSI-19,2PSI-21.8PSISUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTLINERPLATEPRESSURESABNORMALCONDITIONFIGURE7-5'SHEET2OF3 PEDESTAL+11PSICONTAINMENTWALL+13.6PSIBASEMATTOTAL=POSITIVE+NEGATIVESUSQUEHANNASTEAMELECTRICSTATION'NITS1AND2DESIGNASSESSMENTREPORTLINERPLATEPRESSURESABNORMALCONDITIONFIGURE7-5SHEET3OF3 SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTMNINCOMERANALYTICALMODELFIGURE7%
CHAPTER9RESPONSESTONRCQUESTIONSTABLEOFCONTENTS91929.3,IDENTIFICATIONOPQUESTIONSUNIQUETOSSES4QUESTIONSUNIQUETOSSESANDRESPONSESTHERETO,FIGURES ThischaptervillprovideresponsestothoseNuclearRegulatoryCommission(HRC)questionswhichhavebeendesignatedbyRef10(asamended)tobefoundintheplant-uniqueDesignAssessmentReportandtothosequestionsforvhichtheresponseinRef10isinapplicable.TheNRCquestionsforvhichresponsesvillbeprovidedarej.dentifiedinSection9.1,anddetailedresponsestothequestionsarefoundinSection-9.2.9-2 1IDENTIFICATION~OFUESTIONSUNIQUETOSSESThebelowlistedquestionsaddressconcernsuniquetoSSES.ThesequestionsareansweredindetailinSection9.2.NRC~ueetiauNumberM02026N020.27N020.44N02055M02058(1)i(2),(3)M020.59(1),(3),(4)MP20.60M02061M1301M1302M1304N1305N130.6QuestionTopicPrimaryandSecondaryLOCALoadsInventoryEffectsonBlowdownPoolswellWavesandSeismiCSloshSRVLoadsonSubmergedStructuresPlantUniquePoolswellCalculationsDowncomerLateralBracesWetwellPressureHistoryPoolswellInsidePedestalPressureLoadingDuetoSRVDischargeLoadCombinationHistorySoilNodellinqLinerandAnchorageNathematicalModelContainmentStructuralModel-AsymmetricLoadsN130.12SRVStructuralResponse9-3
~92UESTIONSUNIQUETOSSESANDRESPONSESTHERETO~UESTIONM02026TheDFFRpresentsadescriptionofanumberofLOCArelatedhydrodynamicloadswithoutdifferentiatingbetveenprimaryandsecondaryloads.ProvidethisdifferentiationbetveentheprimaryandsecondaryLOCA-relatedhydrodynamicloads".Werecognizethatthisdifferentiationmayva'ryfromplanttoplant.WevoulddesignateasaprimaryloadanyloadthathasorvillresultinadesignmodificationinanyNarkIXcontainmentsincethepooldynamicconcernswereidentifiedinourApril1975genericletters.~ESPONSENO2026ThetablebelowshowstheLOCA-relatedhydrodynamicloadsontheSSEScon'tainment.Thoseloadswhichhaveresultedincontainmentdesiqnmodificationsaredesignatedas"PrimaryLoads"Theseprimaryloadsresultfromthepoolswelltransient.DryvellfloorupliftpressuresduringthewetwellcompressionphaseofpoolswellleadtothedecisiontoincreasetheSSESdrywellfloordesignsafetymarginforupliftpressuresbyrelocatingdrywellfloorshearties.Poolsvellimpact,drag,andfallbackloadsresultedintherelocationofequipmentintheSSESwetwelltoapositionabovethepeakpoolsvellheight.Furthermore,thedowncomerbracingsystemvasredesigned.AllotherLOCA-relatedhydrodynamic1oadsaredesignatedas"SecondaryLoads"sincenodesignmodificationhasresultedfromtheirpresence.LOCALoad"PrimarvLoad>>SecondaryLoad'R1.Wetwell/DrywellPressures(DuringPoolswell)XC1)2.PoolswellImpactLoad3.PoolsvellDraqLoad4.DowncomerClearingLoad5.DowncomerJetLoadx<>>x~>>6.PoolsvellAirBubbleLoad7.PoolswellFallback.Loadxc+>8.MixedFlovCondensationOscillationLoad9-4 9.PureSteamCondensationOscillationLoad10.Chugging.11.Wetwell/DrywellPressureandTemperatureduringDBALOCA(LongTerm)12.Wetwell/DrywellPressureandTemperatureduringXBALOCA(LonqTerm)13.Wetwell/DrywellPressureandTemperatureduringSBALOCA(LonqTerm)Footnote's:(3,)Sheartieschangedindrywellfloor.(2)Equipmentmovedinwetwell.(3)Equipmentmovedinwetwell.Bracingsystemredesign.(4)Equipmentmovedinwetwell.QUESTIONM02027Thecalculateddrywellpressuretransienttypicallyassumesthatthemassflowratefromtherecirculationsystemorsteamlineisequaltothesteady-statecriticalflowratebasedonthecriticalflowareaofthejetpumpnozzleorsteamlineorifice.However,forapproximatelythefirstsecondafterthebreakopening,therateofmassflowfromthebreakwillbegreaterthanthesteady-statevalue.IthasbeenestimatedthatforaNarkIcontainmentthiseffectresultsinatemporaryincreaseinthedrywellpressurizationrateofabout20percentabovethevaluebasedsolelyonthesteady-statecriticalflowrate.ThedrywellpressuretransientusedfortheLOCApooldynamicloadevaluation,foreachNarkIIplant,shouldincludethisinitiallyhigherblowdownrateduetotheadditionalfluidinventoryintherecirculationline.RESPONSEM02027ThedrywellpressuretransientshavebeenrecalculatedbyGE(Ref7)withtheadditionalblowdownflowrateproducedbytheinventoryeffectsincludedintheanalysis.TheLOCAloadspresentedinSection4.2havebeencalculatedusingtheserecalculateddrywellpressuretransients.Specifically,thedrywellpressuretransientresultingfromtheDBALOCAincludingtheeffectsofpipeinventoryhasbeenusedasinputtothepoolswellmodel.
UESTIN02044Table5-1andFigures5-1through5-16intheDFPRprovidealistingoftheloadsandtheloadcombinationstobeincludedintheassessmentofspecificMarkIIplants.Thistableandthesefiguresdo'notincludeloadsresultingfrompoolswellwavesfollowinqthepoolswellprocessorseismicslosh.MerequirethatanevaluationoftheseloadsbeprovidedfortheMarkIIcontainmentdesign.RESPONSEM02044ThisinformationwillbesuppliedinasubsequentrevisiontothisDARgUESTZONN020.55ThecomputationalmethoddescribedinDFFRSection34'forcalculatingSRVloadsonsubmergedstructuresisnotacceptable.ItisourpositionthattheMarkIIcontainmentapplicationsshouldcommittooneofthefollowingtwoapproaches:(1)DesignthesubmergedstructuresforthefullSRVpressureloadsactingononesideofthestructures;thepressureattenuationlawdescribedinSection3.4.1ofNEDO-21061fortheramsheadandSectionA10.3."1ofNEDO-11314-08forthequenchercanbeappliedforcalculatingthepressureloads.(2)FollowtheresolutionofGESSAH-238NJonthisissue.TheapplicantforGESSAR-238NIhasproposedamethodpresentedintheGEreport,"UnsteadyDragonSubmergedStructures,"whichisattachedtotheletterdatedMarch24,1976fromG.L.GyoreytoR.L.Tedesco.Thisreportisactively,underrev,iew.RESPONSEM02055LoadsonsubmergedstructuresduetoSRVactuation.arediscussedinSubsection4.1.3.7.~OESTZONM020.58Relatinqtothepool.swellcalculations,werequirethefollowinginformationforeachMarkIIplant:(1)ProvideadescriptionofandjustifyalldeviationsfromtheDPPRpoolswellmodel.Identifythepartyresponsibleforconductingthepoolswellcalculations(ie,GEortheAGE).Providetheprograminputandresultsofbenchmarkcalculationstoqualifythepoolswellcomputerproqra,m.9-6 (2)Providethepoolswellmodelinputincludingailinitialandboundaryconditions.Showthatthembd01injectrepresentsconservativevalueswithresPecttoobtainingmaximumpoolswellloads.Inthecaseofcalculatedinput,(ie,drywellpressureresponse,ventclearingtime),thecalculationalmethodsshould.bedescribedandjustified.Xnaddition,thepartyresponsibleforthecalculation(ie,GEortheAGE)shouldbeidentified.(3)PoolswellcalculationsshouldbeconductedforeachNarkIIplantThefollowingpoolswellresultsshouldbeprovidedingraphicformforeachplant,:(a)Poolsurfacepositionversustime(b)Poolsurfacevelocityversustime(c)Poolsurfacevelocityversusposition(d)Pressureofthesuppressionpoolairslugandthewetwellairversustime.RESPONSEN02058AspecificresponsetothisquestioncanbefoundinSubsection4.2.1.1.VerificationoftheSSESpoolswellmodelisprovidedinAppendixSectionD.l(2)(3)Inputanddiscussionofthepoolswellmod@1inputcanbefoundinTables4-17,4-18,andSection4.2.1.1.TherequestedgraphicresultsoftheSSESpoolswellcalculationcanbefoundinFigures4-38,4-39,4-40,and4-43.QUESTIONM02059Xnthe4TtestreportNEDE-13442P-01Section3.3thestatementismadethatforthevariousNarkIIplantsawidediversityexistsinthetypeandlocationoflateralbracingbetweendowncomersandthatthebracinginthe4Ttestswasdesignedtominimizetheinterferencewithupwardflow.ProvidethefollowinginformationforeachNarkIIplant:(3)Adescriptionofthedowncomerlateralbracingsystem.Thisdescriptionshouldincludethebracingdimensions,methodofattachmenttothedowncomersandwalls,elevationandlocationrelativetothepoolsurface.Asketchofthebracingsystemshouldbeprovided.Thebasisforcalculatingtheimpactordragloadonthebracingsystemordowncomerflanges.Thdmagnitudeanddurationofimpactordragforcesonthebracingsystemordowncomerflangesshouldalsobeprovided.9-7 (4)Anassessmentoftheeffectofdowncomerflangesonventlateralloads.RESPONSEN020.59.Adowncomerbracingsystemisfurnishedtoresistlateralloadsonthedowncomers.Theoriginaldowncomerbracingwasdesignedtoresistseismicinertialoads.Areviseddowncomerbracingsystemhasbeendesignedto.resisthydrodynamicloadsaswellasseismicinertialoads.Therevised,bracingsystemconsistsofhorizontal6in.diametersteelpipesspanningbetweenthedowncomersandembedsinthesuppressi'onchamberwallortheRPVpedestal.ThepatternofbracingmembersformsahorizontaltrussasshownonFigure9-1.ThebracingmembersareboltedorweldedtothedowncomersandembedsinthesuppressionchamberwallasshownonFigure9-2.Thebracingsystemislocated8ftfromthebottomendofthedowncomerwhichis.3ftbelowthenormalwaterlevel.(3)Thebasisforcalculatingtheimpactordragloadsonthedowncomerbracingsystem(el.668')anddowncomerstiffenerrinqs(el.668'ndel.682')isgiveninSection42.Themagnitudeanddurationofimpact'rdragforcesonthebracingsystemanddowncomerstiffenerringsisalsoqiveninSection4.2.(4)ThisitemisnotapplicabletotheSSESdesign.QUESTIONM02060Inthe4TtestreportNEDE-13442P-01Section5.4.3.2thestatementismadethatanunderpressuredoesoccurwithrespecttothehydrostaticpressurepriortothechug.However,thepressurizationoftheairspaceabovethepoolissuchthattheoverallpressureisstillpositiveatalltimesduringthechug.MerequirethateachNark,IIplantprovidesufficientinformationreqardinqtheboundaryunderpressure,thehydrostaticpressure,theairspaceandtheSRVloadpressuretoconfirmthisstatementoralternativelyprovideaboundingcalculationapplicabletoallNarkIIplants.RESPONSEN020.60Thisinformation-willbesuppliedinasubsequentrevisiontothisDAB.~UMNTZONM020.61SiqniticantvariationsexistintheNarkIIplantswithregardtothedesiqnofthewetwellstructuresintheregionenclosedbythereactorpedestalThesevariationsoccurintheareasof{1)concretebackfillofthepedestal,{2)placementofdowncomers,(3)wetwellairspacevolumes,and(4)locationofthediaphragm9-8 relativetothepoolsurface.Inadditiontovariationbetweenplants,foragivenplant,variationsexistinsomeoftheseareaswithinagivenplant.Asaresult,foragivenplant,significantdifferencesinthepoolswellphenomenacanoccurinthesetworegions.Mewill.reguirethateachplantprovideaseparateevaluationofpoolswellphenomenaandloadsinsideofthereactorpedestal.RESPONSEN020.61TheSSESpedestalandvetwellareaisshownonFigures1-1and9.3.Duetotheabsenceofdovncomersinthepedestalinterior,nopoolswellwouldbeexpectedinthisregion.Thereare12holesinthepedestal,hovever,eightofwhichwouldallowtheflowofwaterfromthesuppressionpooltothepedestalduringaLOCA.Somedovncomersarenearthepedestalflowholes,leadingtothepossibilitythataircouldbeblownthroughthepedestalholes,whichwouldleadtoagreaterpedestalpoolswellthanvouldbeexperiencedbyincompressiblevaterflovalone.Onewouldexpectthepedestalpoolswelltobemuchreducedfromthesuppressionpoolswellduetoitsrelativeseparationfromthesuppressionpoolandthelackofdirectchargingfromdcwncomervents.Indeed,l/13.3scalemodeltestsoftheSSESpedestaldesiqnconductedattheStanfordResearchInstituteunderthesponsorshipofEPRIshowthatthepedestalpoolswellislesst'han20percentofthepoolswellinthesuppressionpool(Ref31).ThereisnopipingorequipmentinsidetheSSESpedestaland,sincethepedestalpoolswellisverysmall,theonlyloadinvolvedduetopedestalpoolswellvouldbeasmall~Pacrossthepedestalduetodifferentwaterlevelsbetweenthesuppressionpoolandthepedestal.ThisloadisconsideredinthedesignoftheSSESpedestal9~USTIONN1301ProvideinSection5adescriptionof.thepressureloadingsonthecontainmentwall,pedestalwall,basemat,andotherstructuralelementsinthesuppressionpool,duetothevariouscombinationsofSRVdischarges,includingthetimefunctionandprofileforeachcombination.Ifthisinformationisnotgeneric,eachaffectedutilityshouldsubmittheinformationasdescribedabove.RESPONSEN130-1Chapter4describesthepressureloadingsandtimehistoriesduetoSRVdischargeand'otherhydrodynamicloads.~UESTIOHN1302InDFFRSection5.2itisstatedthattheloadcombinationhistoriesarepresentedintheformofbarchartsasshovnonFigures5-1through5-16.Itisnotindicatedhovtheseload9-9 combinationhistoriesareused.Inparticular,itisnotclearwhetheronlyloadsrepresentedbyconcurrentbarswillbecombinedanditshouldbenotedthatdependingonthedynamicpropertiesofthestructuresandtherisetimeanddurati6noftheloads,astructuremayrespondtotwoormoregivenloadsatthesametimeeventhoughtheseloadsoccurat.differenttimes.Also,althoughcondensationoscillationsaredepictedasbarsonthebarchart's,theprocedurefortheanalysisofstructuresduetotheseloadshasnotbeenpresented.Accordingly,thedescriptionofthemethodshouldincludeconsideration"ofsuchconditions.Also,forcondensationoscillationloadsandforSRVoscillatoryloads,includelowcyclefatigueanalysis.RESPONSEM1302TheloadswillbecombinedaccordingtoTabl'es5-1and5-2ofthisDARtoassessthecontainment'tructuralcomponentsChapters5.and7explaintheloadcombinationmethodsusedincontainmentanalysis.Thestructuralanalysisprocedur'etoaccountforcondensationoscillationloadwillbepresentedinasubsequentrevisiontothisDAR.QUESTIONM1304Throughtheuseoffigures,describeindetailthesoilmodellingasindicatedinDPFRSubsection5.4.3anddescribethe'olidfiniteelementswhichyouintendtouseforthesoil.RESPONSEM1304SoilmodellingisexplainedinSubsection7.1.1.1andZiqure7-1.QUESTZOE5130.5Describethemathematicalmodelwhichyouwilluseforthe'linerandtheanchoragesystemintheanalysisasdescribedin.DPFRSubsection5.6.3.RESPONSEM1305ThemathematicalmodelwhichwillbeusedforanalysisofthelinerandtheanchorageforhydrodynamicsuctionpressuresisdescribedinSubsection7.1.3QUESTION6130.6InDPPRSubsection5.1.1.1itwasstatedthattheSRVdischargecouldcauseaxisymmetricorasymmetricloadsonthecontainment.InSubsection5.4.1anaxisymmetricfiniteelementcomputerprogramisrecommendedtordynamicanalysisofstructuresduetoSRVloads,andnomentionismadeoftheanalysisforasymmetricloads.Describethestructuralanalysisprocedureusedtoconsiderasymmetricpooldynamicloadsonstructuresandthrough9-100 theuseoffigures,describeinmoredetailthestructuralmodelwhichyouintendtouse.RESPONSE8130.t3ThedynamicanalysesandmodelsusedareexplainedinChapter7gUPSTZONN13012ReferenceismadeinDFFRSubsection5.4.3tostudiesofstructuralresponsetoSRVload.Providecitationsforthisreferenceandwheresuchstudiesarenotreadilyavailable,copiesarerequested.RESPONSEM130.12StudiesmentionedinDFFRSubsection5.4.3aretheresultsofanalysiscompletedforaspecificplantatthetimeofwritingoftheDFFR.Referencetothestudieswasintendedtoindicatetheneedforconsideringstraindependentsoilproperties.For.theSSESanalysis,Ref32isusedtodeterminethesoilconsta<Ltsintheanalysis9-11 g0CONTAINMENTWALLc.b+>iSy-ij+o0000MSRVDISCHARGEPIPE(TYPICAL)0000000DOWNCOMER(TYPICAL)COLUMN(TYPICAL)BRACINGMEMBER(TYPICAL)e'ye4r.~rtIegto4~RPVPEDESTALSUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTDOWNCOMERBRACINGSYSTEMFIGURE9-'I
DIAPHRAGMSLAB~~~oio'O3-1"5H.S.BOLTSDOWNCOMERDOWNCOMER6"$PIPE11/4"Q1/2"ItTOP&BOTTOM.,DETAIL1:-.aEMBED/gO,P4rei4~0o>o5HIGHWATERLEVELEL672'4"BRACINGEL668'4"6"j5PIPEDETAILCONTAINMENTWALLBASEMATb'+~igO)r+SUSQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORT-DOWNCOMERBRACINGDETAILSFIGURE9-2 DOWNCOMERS~~'IHIGHWATERLEVELEL.672'-0"PEDESTAL.HOLES12'USQUEHANNASTEAMELECTRICSTATIONUNITS1AND2DESIGNASSESSMENTREPORTSPATIALRELATIONSHIPOFDOWNCOMERSANDPEDESTALHOLESFIGURE9-3 100REFERENCES2\Dr.N.BeckerandDr.E.Koch,<<KKB-VentClearingwiththePerforated-PipeQuencher"(translatedbyAd-ExpWatertown,Nassachusetts),,KWU/E3-2796,Kraftwez'kUnion,October1973.IDr.M-BeckerandDr.E.Koch,"ConstructionandDesignoftheReliefSystemwithPerforated-PipeQuencher>>(translatedbyAd-Ex),E3/E2-2703,KraftwerkUniori,July1973.3.Dr.N.Becker,<<ResultsoftheNon-NuclearHotTestswiththeReliefSystemintheBrunsbuttelNuclearPowerPlant"(translatedbyAd-Ex),KWU/R113-3267,KraftwerkUnion,,December1974.4.Dr.H.Weisshaupl,<<FormationandOsci1lationsofaSphericalGasBubbleUnderWater<<(translatedbyAd-Ex),AEG-TelefunkenReportNo.2241,KraftwerkUnion,December19725.Dr.HWeisshauplandSchall,"CalculationModeltoClarify'hePressureOscillationsintheSuppressionChamberAfterVentClearing>>(translatedbyAd-Ex),AEG-'elefunkenReportNo.2208,KraftwerkUnion,March1972.6.Dr.M.Becker,FeistandM.Burro,"AnalysisoftheLoadsMeasuredontheReliefSystemDuringtheNon-NuclearHotTestinKKB<<(translatedbyAd-Ex),R113/R213/R314/R521-3346'raftwerkUnionsApril1975-7.Letter,J.W.Mi1lardtoN.J.Lidl,"Susquehanna162:MassandEnergyReleaseforSuppressionPoolTemperatureAnalysisduringSafetyReliefValveandLOCATransients,"GB-77-65,March14,1977.8.R.J.ErnstandM.G.Ward,"NarkIIPressureSuppressionContainmentSystems:AnAnalyticalModelofthePoolSwellPhenomenon,<<NEDE-21544P,GeneralElectricCo.,December1976.9.Letter,F.C.RallytoNarkIITechnicalSteeringCommitteeMembers,"PoolSwellNodelTestCases,"MKII-301-'E,August22,.1977.10.>>DynamicForcingFunctionsInformationreport(DFFR),<<Rev.2,NED0-21061,GeneralElectricCo.andSarqentandLundyEngineers,September1976.11.T.Y.Fukushima,etal,"TestResultsEmployedbyGEforBWRContainmentandVertica1VentLoads,<<NEDE-21078-P,Table3-4,GeneralElectricCo.,October1975.10-1 12.F.J.Moody,AnalyticalModelforLiquidJetPropertiesforPredictingForcesonRigidSubmergedStructures,AEDE-21472,GeneralElectricCo.,(tobepublished).13.R.J.Ernst,etal.,MarkIIPressureSuppressionContainmentSystems:LoadsonSubmergedStructures-AnapplicationMemorandum,NEDE-2)730,GeneralElectricCo.,(tobepubLished).14.F.-J..Moody,AnalyticalModelforEstimatingDragForcesonRigidSubmergedStructuresCausedbyLOCAandSafetyReliefValveRamsheadAirDischarges,NEDE-21471,GeneralElectricCo.,(tobepublished)15.MarkII-PhaseI,4TTestsApplicationsMemorandum,LetterandReporttoM.R'utler(NRC)fromJ.FQuirk(GE),June14,1976.16.M.J.Bilanin,et.al.,MarkIILeadPlantTopicalReport:PoolBoundaryandMainVentChuggingLoadsJustification,NEDE-23617P,July1977.17.Warmeatlas(HeatTransferData),VDI(SocietyofGermanEngineers),Dusseldorf,1974.1'8.T.E.Johnson,etal.,<<ContainmentBuildingLinerPlhtbDesignReport,>>BC-TOP-1,BechtelCorporation;SanFrancisco,December1972.19.<<Sei'smicAnalysisofPipingSystems,"BP-TOP"1,Rev2,BechtelPowerCorporation,SanFrancisco,Januhry1975.20.Letter,J.R.MartintoMarkIIOwnersGroupandTSC,MKXI-,'50-E,
Subject:
CondensationOscillationExcerptstoApplicationsMemorandum,July1,197721.D.HoffmanandESchmid,>>BrunsbuttelNuclearPowerPlantListofTestParametersandMostImportantMeasurementResultsoftheNon-NuclearHotTestswiththePresureReliefSystem<<(translatedbyAd-Ex),R521/40/77,KraftwerkUnion,August197722.DGobel,"ResultsoftheNon-NuclearHotTestswiththeReliefSysteminthePhilippsburqNuclearPowerplant<<(translatedbyAd-Ex),R142-38/77,KraftwerkUnion,March1977.23.D.HoffmanandE.Schmid,>>PhilippsburgINuclearPowerPlantListofTestParametersandMostImportantMeasurementResultsoftheNon-NuclearHotTestswiththePressureReliefSystem"(translatedbyAd-Ex),R521/41/77,KraftwerkUnion,August1977.10-2 24.Klans-D.Werner,>>ExperimentalStudiesofVentCleavingintheNodelTestStand<<(translatedbyAd-Ex)',KWU/R521-3129,KraftwerkUnion,July1975.25.D.Gobel,>>KKB-NuclearStart-UpResultsoftheTestswiththePressureReliefSystem>>(translatedbyAd-Ex),R142-136/76,KraftwerkUnion,September1976.426.D.HoffmanandDr.KMelchior,<<CondensationandVentClearingTestsinGKNwithPerforatedPipes"(translatedbyAd-Ex),KWU/E3-2594,KraftwerkUnion,Nay1973.27.GEDrawing761E579,BechtelNo.8856-N1-B11-8928.ASNEBoilerandPressureVesselCode,SectionIII,Division1,197429.ASNEBoilerandPressureVesselCode,SectionIII,Division2,197430ACI318-7131.R.L.KiangandB.J.Grossi,<<DynamicNodellingofaNarkIZPressureSuppressionSystem,"EPRI-NP-441,PaloAlto~April1977.32."SeismicAnalysesofStructuresandEquipmentforNuclearPowerPlants,<<BC-TOP-4A,BechtelPowerCorporation,November197410-3 APPENDIXACONTAINMENTDESIGNASSESSMENTTABLEOPCONTENTSA1CONTAINMENTSTRUCTURALDESIGNASSESSMENTA2CONTAINMENTSUBMERGEDSTRUCTURESDESIGNASSESSMENTA3FIGURES APPENDIXAFIGURESNumber.TitleConcreteandReinforcementStressElementsA-2TypicalSectionShovingSectionLocationReinforcedBarArrangementContainmentContainmentmargins-DryvellMallmargins-ShieldMallandRPVPedestalA-6A-8ContainmentMargins-MetvellMallContainmentNargins-RPVPedestalContainmentMargins-BaseSlabContainmentNargins-Diaphragm-SlabA-2 APPENDIXAContainmentDesignAssessmentThisappendixindicatesthecontainmentelementsandcross-sectionswherestressesaretobedeterminedandcontainsatabulationof.thepredictedstresses,allowedstressed,phddesignmarginsforeachloadingcombinationconsidered.Thestructuralassessmentofthecontainmentiscoveredi>iSectionA.1;thesubmergedstructuresareassessedinSectionA.2.A1CONTAINMENTSTRUCTURALDESIGNASSESSMENTTypicalexamplesofthismaterialareincludedinthede.port(FiguresA-1throughA-9);acompleteSectionA.1willbeincludedinafuturerevisiontothisreportA2CONTAIN'MENTSUBMERGEDSTRUCTURESDESIGNASSESSMENTTobeincludediaafuturerevisiontothisreport.A-3 CECAPOUTPUTLOADCOMBINATIONEQN.1=1.4D+1.5SRV(ASYM)STRESSESINKSISTRUC.TURALCOMPONENTANSYSELEMENTNUMBERSECTIONNUMBERVERT.HOOPINSIDEFACEREBAR4VERT.OUTSIDEFACEREBAR4HOOPSPIRALISPIRAL2SHEARTIESPRINCIPALCONC.STRESS86-0.017-0.067-0.1230.1230.027-0.027-0.233-0.039103-0.099-0.054~0.1450.052-0.018~0.0764.103-0.026231~0.264~0.017~0.3730.080~0.126~0.166.4I.127-0.0630Cmmrnzzmzpc+Chcn~+cnmmen~+<+mmz~ZOm+~0mAg7co0z311-0.3500.409315~0.6360.570"ALLOWABLEREINFORCINGSTEELSTRESS=54KSI-0.480-0.5950.5860.0070.6180.0970.044~0.015-0.140~0.304-0.076-0.090
CECAPOUTPUTLOAD'COMBINATIONEQN.1=1.4D+1.5SRV(ASYM)STRESSESINKSIANSVSTURALELEMENT*NUMBERNENTSECTIONNUMBERVERT.HOOPINSIOEFACEREBAR%VERT.HOOPSPIRAL1OUTSIDEFACEREBAR%SPIRAL2SHEARTIESPRINCIPALCONC.STRESS165120Z134)Chczmm~rpLpr2r<mpg0mB02mcn"-IDcmCOmzZZLc+>zenm+mCOCom+m2ZI~0mAm0COOZI-K36215"ALLOWABLEREINFORCINGSTEELSTRESS=54KSI CECAPOUTPUTLOADCOMBINATIGNEQN.-1='1.4D+1.5SFIV(ASYM)STRESSESIN-KSISTRUC.ANSySELEMENTCOMPO-NUMBERNENTSECTIONNUMBERVERT.HOOPINSIDEFACEREBAR"VERT,HOOPSPIRALISPIRAL2OUTSIDEFACEREBAR"SHEARTIESPRINCIPALCONC.STRESS441-0.996.113.601.391.36-0.080~0.145455-0.943.76~0.952.520.990.59-0.140-0.140473~0.922.91~0.872.080.620.59-0.078~0.13100hlZIIllIllIZrIrzraZ0Cmmnzzzac+gzc,m-ImCOCOm>mZZI~0m0illOOXIMOz47510-1.236.10495-1.244.09"ALLOWABLEREINFORCINGSTEELSTRESS=54KSI~0.703~0.833.111.691.121.651.174.052~0.32-0.191~0.19 CECAPOUTPUTLOADCOMBINATIONEQN.1=1.4D+1.5SRV(ASYM)STRESSESINKSISTRUC-TURALCOMPO.NENTANSYSELEMENTNUMBERSECTIONNUMBERVERT.HOOPINSIDEFACEREBAR%VERT.OUTSIDEFACEREBAR"HOOPSPIRAL1SPIRAL2SHEARTIESPRINCIPALCONC.STRESS484161.710.472-3.150.6870.00.01.68~0.55255017~0.9530.926-1.782.570.00.00.160-0.264l-cc:59518-1.120.306-1.730.480.00.0-0.030~0.25760619-1.07-0.038-2.190.2380.00.00.234~0.337O0zx-azCgmmmz0+lllg0xlQz0CmmcnXazzzcn+PyZcillfllCOm~mZZgIOmg7~Am1l0AOZ20-1.28-0.031"ALLOWABLEREINFORCINGSTEELSTRESS=54KSI-2.170.1960.00.00.281-0.330 CECAPOUTPUTLOADCOMBINATIONEQN.1=1.4D+1.5SRV(ASYM)STRESSESINKSISTRUC.TURALCOMPO.NENTANSYSELEMENTNUMBERSECTIONNUMBERRADIALTANGENTIALTOPFACEREBAR~~iBOTTOMFACEREBAR"""RADIALTANGENTIALSHEARTIESPRINCIPALCONC.STRESS551308.512.102.430.920.501-0.129260.4311.28-0.26-0.134.2774.082QcmO0Z+'ZgmCOr~m2QZD0mmMpc+ez~en+mcn~~zm+m~Zc0mX~OmACO0z710.72729'.552.411.42270.5540.6803.115.890.54iNORTH-SOUTHBARS4%EAST-WFSTBARS""~ALLOWABLEREINFORCINGSTEELSTRESS=54KSI0.821.960.02'I0.0950.2000.481-0.075-0.1024.105 CECAPOUTPUTLOADCOMBINATIONEQN.1=1.4D+1.5SRV(ASYM1STRESSESINKSISTRUC.TURALCOMPO.NENTANSYSELEMENTNUMBERSECTIONNUMBERRAOIALTANGENTIALTOPFACEREBAR%BOTTOMFACEREBAR"RADIALTANGENTIALSHEARTIESPRINCIPALCONC.STRESS253.603.712.90-0.321-0.050ACoOUZxz33K)OIllG)ZcoEI~Qz0CITImcnzzzc.+~zengWmCh<+mmzzom+MOITI0Og7COI0zClUIT:KCLO41144024211.933.33452220.844470230.392'ALLOWABLEREINFORCINGSTEELSTRESS=54KSI3.401.553.623.873.232.931.922.185.551.694.424.2204.3504.614.564.048-0.0734.031 APPENDIXDPROGRAMVERIPICATXONTABLEOPCONTENTSD1D2D3POOLSMELL5ODELVERIFICATIONPIGURESTABLES APPENDIXD~FZORES~uebeoD-1D-2D-3D-4D-5CodeVerification-PoolsvellHeightforClassl'lantCOdeVerification-PoolssellVelocityforClass1PlantCodeVerification-PoolssellHeightforClass2Plant'CodeVerification-PoolsvellVelocityforClass2PlantCodeVerification-PoolswellHeightforClass3PlantCodeVerification-PoolsvellVelocityforClass3PlantD-2 APPBNDIXD~urbeD-1D-2D-3~T~te4DrywellPressureTransientsfortheTestCases4PLantSpecificParametersfortheTest,CasesComparisonofHaximumPoolSmellVelocityforClasses,l,2,and3TestCasesD-4CommonAssumptionsfortheTestCasesD-3 APPENDIXDROGAVERIXCAIONThepurposeofthisappendixistoprovideinformationwhichverifiestheaccuracyofthecomputerprogramsusedinconjunctionwithSSESdesignassessment.ADOOLSLL-NODLVIICAIONThissubsectiondemonstratestheaccuracyoftheSSESDARpoolswellmodelbycomparingitwiththemodeldevelopedbytheGeneralElectricCompany.Thelattermodelhaspredictedconservativelytheresultsofthe4Tpoolswelltests{Ref8).ToevaluatetheagreementbetweentheGEpoolswellcodeandthepoolswellcodeusedfortheSSESDAR,threetestcaseswereselected.ThetestcasesusedweretheClasses1,2,and3plantsdescribedinRef10TheinputdataforthesethreeproblemsaregiveninTablesD-1andD-2(takenfromRef9).(Intheverificationofthemodel,theboundaryconditionsassumedbyGEinRef9wereused..TheseassumptionsareshowninTableD-4.),ThesedataarerepresentativeoftypicalU.S.NarkIIBWRs.ThepoolswellcodeusedinthisDARwasreviseduntiltheresultswereincloseagreementwithGE'sresultsasgiveninRef9.Agreementwasjudgedbyexaminingthepeakswellvelocitypredicted,sincethisisoneofthemostimportantpoolswellparametersandonethatisfairlysensitivetohowthephenomenon'smodelled.Thedegreeofagreementfinallyachievedbetweenthepoolswell'codeusedinthisDARandtheGEcodeisshowninTableD-3wherepeakswellvelocitiesarecompared.TransientcomparisonsforClasses1,2,and3plantsareshownonFiguresD-1throughD-6wherethetransientpredictionsofthetwocodesareshowntobeessentiallyidenticalPromthegoodagreementshowninthecheckcases,thepoolswellcodeusedinthisDARisverifiedtobethesameastheGEcodeforevaluationofpoolswell.
APPENDIXETheresultsofanalysisofthereactorbuildingstructurewillbesummarizedinthisappendix.Thisappendixwillbeprovidedinafuturerevisiontothisreport f,t