ML18038A726

Technical Response to SE by NRR Related to Proposed Deferment of Torus Mods, Nmp,Nmpns,Unit 1 Docket 50-220

ML18038A726
Person / Time
Site:	Nine Mile Point
Issue date:	11/19/1992
From:	TELEDYNE ENGINEERING SERVICES
To:
Shared Package
ML17058B068	List: ML17058B067 ML18038A726
References
NUDOCS 9211300182
Download: ML18038A726 (54)

Text

i(TELEDYNEENGINEERINGSERVICESADNISIONOFTELEDYNEBROWNENGINEERING10ForbesRoadWoburn,Massachusetts01801-2103617-932-9000FAX617-932-9970TECHNICALRESPONSETOSAFETYEVALUATIONBYTHEOFFICEOFNUCLEARREACTORREGULATIONRELATEDTOPROPOSEDDEFERMENTOFTORUSNODIFICATIONSNIAGARANOHAWKPOWERCORPORATIONNINENILEPOINTNUCLEARSTATIONUNIT1DOCKETNO.50-220NOVENBER19,1992cy211g00182921123PDRADQCK05000220PDR I

TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992TABLEOFCONTENSUMMARYi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~o~~~~o~~~~~~~S1.0BACKGROUND..................................................2.0COLOADDEFINITIONCONSERVATISMS............................2.1FSTFCONSERVATISMS2.1.1FSTFENDCAPS................................2.1.2FSTFSTRUCTURALDAMPING......................2.2LDRCONSERVATISMS......................................2.2.1HARMONICCOMPONENTAMPLIFICATION.............2.2.2SUMMINGOFHARMONICCOMPONENTS...............3.0TESCOLOADREDUCTION-STRUCTURALMODEL/SHELLANALYSIS.....4.0ADDITIONALCONSERVATISMS....................................5.0APPLICABILITYOFTHECONCLUSIONSPROVIDEDINTHEREVIEWOFTHEVALIDITYOFRANDOMPHASINGRULESASAPPLIEDTOCOTORUSLOADS..........6.0NMP-1TORUSSHELLCMTR'S7.0RELEVANCETOSAFETY..8.0REFERENCE...................................................~Pae121820236TELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING PI0 TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992SUMMARYTheNRCSafetyEvaluationReport(SER)discussestwomethodsofobtainingreliefintheevaluationforcondensationoscillation(CO)loads.Thefirstmethodistocombinethe31stressharmonicsbytakingtheabsolutesumofthe4peakresponsesandaddingtothisthesquarerootofthesumofthesquares(SRSS)oftheremaining27harmonics.ThismethodhasbeenpreviouslyacceptedbytheNRC.ThesecondmethodistoincorporatetheanalyticallydeterminedCOpressurereductionfactorspresentedinNiagaraMohawkPowerCorporation'srecent,andsubjectsubmittal,reference1.TheSERadvisesthatthereviewtheyhavedonesofarsupportsthepositionthatoneortheotherofthetwoapproachesmaybeused,butnotboth.Thatis,ifcreditistakenfortherecentlydevelopedCOpressurereduction,thenall31stressharmonicsmustbecombinedbyabsolutesummation,ratherthanusingtherandomphasingrulespreviouslyapprovedfortheMarkITorusProgram.TheCOpressurereductionrepresentsnewandadditionalreliefanditsuseshouldnotprohibitusingthealreadyestablishedreliefgainedfromtherandomphasingofthestressharmonicssincethetwoanalysesareindependentofeachother.Theworkpresentedinreference1usestheContinuumDynamics,Inc.COpressurereductionresultsandcombinesthe31stressharmonicsbyabsolutesummingthe4peaksandaddingtothistheSRSSoftheremaining27harmonics.TheCOpressurereductionworkreducestheCOstressesbyapproximately17%and36%fortheeightandfourdowncomerbays,respectively.Absolutesummingthe4peakstressharmonicsandaddingtothistheSRSSoftheremaining27stressharmonicsresultsinanestimated33%lowerstressthanabsolutesummingalltheindividualharmonics.ConverselyATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING e"4d',f

-2-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992stated,absolutesummingalltheindividualstressharmonicsoverpredictsmeasuredstressesbyabout50%.Themethodofcombiningthestressharmonicsbyabsolutesummingthe4peaksandaddingtothistheSRSSoftheremaining27stressharmonicswastheresultofextensivestructuralanalysisdonefortheoriginaltorusprogramandwasrecognizedastheappropriatewaytocorrelateFullScaleTestFacility(FSTF)measuredstresseswithFSTFanalyticallypredictedstresses.Thismethodwasavailableforallplantstouse,notjustNineMilePointUnit1(NMP-1),anditisstillvalid.TheCOpressurereductionaccountsfortheendcapeffectoftheFSTFinmakingitappearthatallbayshaveeightdowncomersandthehydrodynamicsineachbayareinphase.Thisisadifferent,andindependent,phenomenonthantheunderlyingbasistothemethodofcombiningstressharmonicsdiscussedintheprecedingparagraph.TheBrookhavenattachmenttotheSERagreesthattheendcapeffectisanFSTFconservatism.Theattachmentdoesnotquantifytheconservatism.NRC-acceptanceof-thereduced-CO.loadingaddressedhereingains-NMP-1areliefof573psiintheeightdowncomerbays,and1565psiinthefourdowncomerbays.Toseethatthisstressreductionisnominal,thesevaluesaretobecomparedtoatotalcombinedstressfromallsourcesontheorderof16,025psi.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERINGC I\

TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19921.0BACKGROUNDTheMarkIProgramGeneralElectric(GE)determinedthemagnitudeoftheCondensationOscillation(CO)loadingbasedonthetestresultsfromtheFullScaleTestFacility(FSTF).TheFSTFfacilitywasonebaywithendcapstocontainthefluid,andasaresultofthecompromisesintestfacilitydesign,theseendcapscausedconservativeCOshellpressurestobemeasured.TheCOloaddefinition,basedonthesemeasuredpressures,wasconservativeontheorderof15to30percentandthiswasrecognizedatthattime.TheMarkIOwners'roupdetermineditwouldnotbecosteffectivetofundtheanalysisanddocumentationeffortnecessarytoachievefurtherreductionintheCOloaddefinitionsincemostoftheMarkIplantshadadequatemarginonCodestressallowablesfortheCOfrequencydomaineventcombinationloading.However,theNineMilePointUnit1(NHP-1)torushasathinshell(0.460in.)comparedwithmostofHarkIplants,andasaresult,thepostulatedeventcombinationwhichincludesDesignBreak/Accident(DBA)pressureandCO(eventcombination20)controlsthemarginontorusshellthickness.TeledyneEngineeringServices(TES)andNiagaraMohawkPowerCorporation(NHPC)recognizedthisproblemasbeingcriticalearlyintheHarkIprogram,andjointlytookthenecessarystepstomitigateloadsfromthiseventcombination.First,TESrefinedtheTorusAnalysisforDBApressureandCOincludingthepostprocessingofresults.Then,TESandNMPCinitiatedaseriesofthinshellmeetingsatGEforNHP-1andOysterCreek.ThesemeetingsidentifiedareasofconservatismintheloaddefinitiontobefurtherexploredbyGE.The,reductioninNHP-1DBApressureresultingfromthesemeetingswasessentialtothesuccessfulcomplianceofNMP-1totheHarkIProgramStructuralAcceptanceCriteriafortheCOeventcombination.TheDBApressure,ratherthantheCOloadingconservatisms,wereaddressedbasedoncostandtimeconsiderations.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

~I~P~+

-4-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19922.0COLOADDEFINITIONCONSERVATISMSTheloadsonwhichtheTESstructuralanalysisisbasedarepresentedprimarilyinG.E.ReportNED0-21888,MarkIContainmentProgramLoadDefinitionReport(LDR),datedNovember1981(Reference3).TheseloadsweredevelopedfromtheFSTFduringtheHarkIProgramandhaveinherentconservatisms.2.1FSTFCONSERVATISHSTherearetwomajorconservatismsinherenttothegeometryoftheFSTF,theyaretheFSTFbayendcapsandthestructuraldampingassociatedwiththelowlevelofstressintheFSTFshell.2.1.1FSTFENDCAPSIn1979,ContinuumDynamics,Inc.(CDI)wasaskedbytheMarkIownersgroup,throughG.E.,toassesstheconservatismintheCondensationOscillationtorusloadsmeasuredduringtheFSTFblowdowntests.Thiseffortconfirmedgenerallyacceptedconservatisminthetestswithregardtotestinitialconditionthermodynamics,andidentifiedasignificantconservatismwhichwasnotidentifiedduringtestdesign.Thisconservatismwasintroducedbythegeometryofthetestfacility,one-sixteenthsectorwhichisequivalentlya22-1/2'egmentoftheMarkIPressureSuppressionPoolTorus.Thetestfacility,althoughfull-scaleincrosssection,attemptedtosimulateatfull-scalethecondensationphenomenoninonebayonly.Endcapswererequiredtocontainthepoolwaterandtheairspaceabovethepoolinthebay.Theanalysis,whichanalyzesthehydrodynamicconsequencesoftheseendcaps,waspresentedtotheHarkIownersin1980.Toexpeditecompletionofthisissue,theMarkIownersdecidednottopursuereducingthisconservatismatthattime.ThisworkisrevisitedforthiseffortanddevelopedspecificallyforNineMilePointUnitl.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING P0l>>

-5-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992ThejointTESandCDIeffortpresentedinReferences1810andsummarizedhereinconsistsofananalyticalreductionintheHarkITorusProgramCondensationOscillationLoadDefinition.TheanalysisshowsthattheeightdowncomerbayshavebayaveragedCOloadswhichareconservativebyatleast19%atfrequenciesotherthan5-6Hzandforfourdowncomerbays,thebayaveragedCOloadsareconservativebyatleast38%atfrequenciesotherthan5-6Hz.Theloadconservatismsinthe5-6Hzfrequencybandare6%and28%fortheeightandfourdowncomerbays,respectively.Takingallfrequencybands(0to31Hz)intoaccountresultsinanetCOloaddefinitionconservatismof17.1%and36.1%fortheeightdowncomer(non-vent)andfourdowncomer(vent)bays,respectively.2.1.2FSTFSTRUCTURALDAMPINGDampingequalto2percentwasusedintheevaluationsperformedtodevelopthephasingrules.Theassumptionof2percentdamping,whichisappropriatefordesignresponsesfromcombinedloadsnearone-halfyieldstress,maybetoohighforthelowlevelFSTFresponse.Ifthedampingusedwereonly1.5percent,the-harmonic,responseamplitudesusedtodevelopthephasingruleswouldbesignificantlylargersuchthatthecombinedcalculatedresponseusingtheabsolutesumofthe4peakresponsesandaddingtothistheSRSSoftheremainder,wouldbeconservativecomparedtothemeasuredresponse.Inotherwords,usingtheabsolutesumoflessthanthe4peakresponses,perhapsthe3peakresponses,plustheSRSSoftheremainder,mayhaveboundedthemeasuredresponse.Inourjudgement,dampingontheorderof1.5percentwouldbemoreappropriateforFSTFresponselevelsthan2percent.Itmustbeemphasizedthatdampingof2percentismoreappropriatefordesignanalysesinwhichstresslevelsofone-halfyieldareallowable.O'TELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING J~I1

~i'ECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19922.2LDRCONSERVATISMSTherearetwoconservatismsassociatedwiththedevelopmentoftheLDR,G.E.ReportNEDO-21888(Reference3).Thefirstconservatismisrelatedtotheamplificationoftheindividualharmoniccomponentsbeforecombiningthecomponentsandthesecondconservatismisrelatedtothemethodusedtocombineorsumtheindividualharmonicstresscomponents.2.2.1HARMONICCOMPONENTAMPLIFICATIONForFSTFtheresponseamplificationfactorsateach0.914Hzfrequencyintervalwereusedinlieuoftheresponseamplificationfactorsatthestructuralnaturalfrequenciesineach1-Hzwindow(References4,758).Thephasingrulesweredevelopedwiththisreducedresponse.Suchanapproachdoesnotintroducetheconservatisminresponsecalculationwhichisobtainedwhenaccountingfortheresponseamplificationfactorsatthestructuralnaturalfrequenciesineach1-Hzwindow.Nocreditisgivenforthelatterapproach.2.2.2SUMMINGOFHARMONICCOMPONENTSTheLDR(Reference3)statesthatthecombinationof'individualharmonicstresscomponentsshallbesummed.Threeacceptablemethodsareavailable:1.Absolutesumofallharmoniccomponents.2.Absolutesumofthe3highestpeaksaddedtotheSRSSoftheremainingcomponentsandapplya1.15factor.3.Absolutesumofthe4highestpeaksaddedtotheSRSSoftheremainingcomponents,providedthereportedshellstressesarenotwithinafewpercentoftheallowables.iTELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

~I~1JI1g,,1I TECHNICALRESPONSETODOCKETNO.50-220NOVEHBER19,1992TheconservatismsassociatedwiththesemethodsarebasedontheboundingofthemeasuredFSTFtestshellstressresults(TestNo'sH-8,H-llBandH-12).Hethods1and2boundallthreetests,whilemethod3fallsjustshortofboundingtestH-lR.IITELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

~rII

-8-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19923.0TESCOLOADREDUCTION-STRUCTURALMODELSHELLANALYSISOscillatingloadsonthesubmergedportionofthetorusshellduringtheCOphenomenonarecausedbyperiodicoscillationssuperimposedontheprevailinglocalstaticpressures.PlantuniqueloadsarederivedfromFSTFdata.FlexiblewallloadsweremeasureddirectlyintheFSTFwhichisprototypicalofMarkIplantconfigurationswiththeexceptionoftherigidendcaps.PressuremeasurementsobtainedfromvariouslocationsonthetorusshellshowthatthelongitudinalpressureoscillationamplitudedistributionalongthetorusbottomcenterlineisessentiallyuniformfortheFSTF.Specificationofabaselinerigidwallloadisgivenaspressureoscillationamplitudeasafunctionoffrequency.ThisloadhasbeenderivedfromthemeasuredFSTFflexiblewallloadbyanalysiswithacoupledfluid-structuraldynamicmodeloftheFSTFtorus.Thederivationofthebaselinerigidwallloadisdescribedbelow:a.Afiniteelementcoupledfluid-structuraldynamicmodeloftheFSTFtoruswasexcitedatvaryingfrequencieswithaunitamplitudepressuresourceattheventexits.Thetorusshellpressureamplitudesrelativetothesourcepressure(amplificationfactors)weredeterminedasafunctionoffrequency.b.Usingtheserelativeamplitudes(amplificationfactors),theFSTFventexitsourcepressureswerederivedfromthemeasuredtorusshellpressuresatthevariousfrequencies.c.ThebaselinerigidwallloadwasderivedfromthecomputedFSTFventexitsourcepressuresbyhydrodynamicanalysis.IIITELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING 1~I4,l TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992TheCOshellloadisspecifiedasadistributionofharmonicpressureamplitudesin1Hzbands(Reference3).Theanalysisforthisloadwasperformedbyconsideringtheeffectofunitloadsateachloadfrequency(harmonicanalysis)andthenscalingandcombiningtheindividualfrequencyeffectstodeterminetotalstressatthecriticalelement.ThethreevariationsintheCOspectrum(Reference3)wereevaluatedbyrescalingtheresultsoftheunitloadanalysis.100%ofwatermasswasusedforallCOanalysis.ThereductionfactorspresentedinTable1ofReference10wereappliedtotheindividualharmonicpressures.Thecombinationofindividualharmonicstressesintototalelementstresswasdonebyconsideringfrequencycontributionsat31Hzandbelow.TheactualcombinationwasdonebyaddingtheabsolutevalueofthefourhighestharmoniccontributorstotheSRSScombinationoftheothersforshellstress.Thiscombinationmethodanduseofthe31Hzcutoffaretheresultofextensivestructuralevaluationoffullscaletestdata,whichisreportedanddiscussedinReferences4and7.Includingthefrequencycontributionoutto50HzwouldincreasetheCOstressbyabout20psi,orlessthan1/2%.Themethodofcombiningthestressharmonicsbytheabsolutesumofthe4peaksplustheSRSSoftheremainderwastheresultofextensivestructuralanalysisdonefortheoriginaltorusprogramandwasrecognizedastheappropriatewaytocorrelateFSTFmeasuredstresseswithFSTFanalyticallypredictedstressesusingtheloaddefinition.Touseabsolutesumofallthecomponentswouldbetoignorethephasingbetweenfrequencydependentpressurecomponentsandbetweenfrequencydependentstructuralresponses,andtoassumephasingbetweentheharmoniccomponentstoproducethehighestpossiblestressresponse.Themethodofcombiningthestressharmonicsbytheabsolutesumofthe4peaksplustheSRSSoftheremainderwasavailableforallplantstouse,notjustNMP-1,andtodayremainsvalid.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING l~'I,,~)I

-10-TECHNICALRESPONSETODOCKETNO.50-220NOVENBER19,1992Thepresentanalysis(References1510)investigatestheconservatismofthehydrodynamictorusCOloaddefinitionderivedfromdatatakenintheHarkIFSTF.ItisshownthatduringCO,thecondensationeventsatthedowncomerexitsare,asafunctionoffrequency,randominphaseformostharmoniccomponents.Asaconsequenceofthisobservation,andthegeometricalconstraintsbuiltintotheFSTF,theCOloadsdefinitionappliedtoNHP-Iisconservativefortworeasons.~AlternatedowncomerbaysinNHP-Ihavefour-eight-four-eight,etc.,downcomersperbay.TheFSTFfacility,byconstruction,assumesthatallbayshaveeightdowncomersperbay.TheFSTFmodeleda22I/2'ectorofaprototypicalHarkIsuppressionchamber.Thewaterwascontainedinthesectorbytwoveryrigidendcapswhichwouldnotexistinafullsuppressionchamber.Theseendcapshydrodynamicallyactasmirrors.Thisresultsinameasuredload,asifallbaysinafulltorushadcondensationphenomenonidenticalinphaseandamplitude,totheinstrumentedFSTFbay.TheanalysisshowsthatforNHP-I:~EightdowncomerbayshavebayaveragedCOloadswhichareconservativebyatleast19%atfrequenciesotherthan5-6Hz.~FourdowncomerbayshavebayaveragedCOloadswhichareconservativebyatleast38%atfrequenciesotherthan5-6Hz.Thepresentwork(Reference1)accountsforboththerandomphasingofthestressharmonicsandthereductioninFSTFmeasuredpressures,twoindependentphenomena.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING rll~A v,TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992InclusionofthereducedCOloadingaddressedhereingainsNHP-Iareliefof573psiintheeightdowncomerbays,and1565psiinthefourdowncomerbays.Thesevaluesaretobecomparedtoatotalcombinedstressfromallsourcesontheorderof16,025psi.TheCOloadreductionaccountsfortheendcapeffectoftheFSTFinmakingitappearthatallbayshaveeightdowncomersandthehydrodynamicsineachbayisinphasewithadjoiningbays.IntheBrookhavenattachmenttotheSER(Reference15)itisstatedthattheendcapsintroduceanFSTFconservatism.made:Onpage3oftheBrookhavenattachment,thefollowingpointsareTheFSTFdatasupportthenotionthattheCOprocessisrandomovermostofthefrequencyspectrumconsideredintheloadmethods.Becauseofthegeometricdifferences,particularlythe4-8-4downcomerarrangement,thepressureloadsduringCOblowdownwilltendtobegreaterintheFSTFrelativetotheNHPtorusforthesamehydrodynamicflowconditions.~Theprocedureusedtoquantifytheseeffectsrepresentsastraightforwardapplicationofaconventionalhydrodynamicmethod.Theresultsarereasonableandprobablyconservativebecauseofthehighsoundspeedusedinthenumerics.Wealsoconsidertheassumptionthatacorrelationexistsbetweenbaystobeasignificantconservatism.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING 0'i~

-12-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992Theattachmentitselfdoesnotquantifytheseconservatisms.However,GeorgeBienkowski'sreviewoftheRandomPhasingRules,datedAugust25,1983(Reference14),alsoincludessomequantificationoftheseeffects.ThisisdiscussedinSection5.0herein.IIITELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING It.JI

-13-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19924.0ADDITIONALCONSERVATISHSThefollowingisadelineationofadditionalconservatismsforwhichnocredithasbeentaken.1.UniformcorrosionoftheshellwouldcausethefrequenciestodropandtheresponsetoCOloadtodecrease.Ifthefrequenciesdrop1Hz,thenitisestimatedthatNMP-1wouldgainabout600psireliefinshellstress.2.Theanalysishasbeenperformedusing2%damping.Itisestimatedthatincreasingthedampingto4%,toaccountforwater/structuredamping,wouldgainNHP-1morethan900psirelief.Even4%dampingmaybelowforathinshelledwaterfilledstructure.3.ThehighsoundspeedusedinthedeterminationoftheCOloadreductionfactorsisconservativeasisshowninReference10.4.Curvatureeffectsofthetorus.5.TheASHECodeallowsauniform10%reductioninwallthicknessforClassMCComponents,1990Addenda,SectionXI,ParagraphIWE-3519.3.Thisisthesameasallowinga10%increaseinthematerialallowables.6.TheASMECode,SectionIII,ParagraphNf-3213.10,permitsuptoI/16inchlocalcorrosionasisexplainedinTESTechnicalReportTR-6801-2(Referencell).7.Class1allowablesforthismaterialaregreaterthanClassHCallowablesby21%,Sm=20ksiforClass1vs.Sm=16.5ksiforClassHC.HTELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

]tI~i4VjW

-14-TECHNICALRESPONSETODOCKETNO.50-220NOVEHBER19$19928.CertifiedHaterialTestReports(CHTR's)fortheNHP-1shellmaterialindicatehigherallowablesthanthoseusedintheHarkIProgramAnalysis,17.6ksivs.16.5ksi(seeSection6.0).IfitbecomesnecessarytousetheCHTR's,hardnesstestscouldbeperformedonthetorusshelltoincreaseconfidenceinthematerialCHTR's.9.Thereisinherentconservatisminthefactorofsafety(FS)associatedwiththecode,FS=1/(1.lxl/4)=3.64onultimatematerialstrength.ItTELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

~I4k,

-15-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19925.0APPLICABILITYOFTHECONCLUSIONSPROVIDEDINTHEREVIEWOFTHEVALIDIYOFRANDOMPHASINGRULESASAPPLIEDTOCOTORUSLOADSGeorgeBienkowskiissuedtheresultsofhisreviewoftheValidityofRandomPhasingRulesasAppliedtoCOTorusLoads,onAugust25,1983(Reference14).Thereviewconcludedthefollowing:l.A/E'scouldeliminatethe1.15responsefactoronshellstressesiftheyuse4harmonicssummedabsolutelyaddedtotheremainingsummedSRSS(inlieuof1.15factoron3harmonicssummedabsolutelyaddedtotheremainingsummedSRSS),providedthereportedshellstressesarenotwithinafewpercentoftheallowables,otherwisetheissueshouldberevisited.Theadditionof1harmonic,tobesummedabsolutely,providesonlyabouta10%increaseintheresponsesratherthanthe15%neededtoboundtheFSTFmeasurements.2.A/E'scouldneglecttheharmoniccomponentsabove30HertzforstructureswithsimilarnaturalfrequencycontenttotheFSTForOysterCreekifspecificjustificationintheformoftorusresponsefrequencycharacteristicispresented.3.A/E'scoulduseanyvariationthatproducesatleastashigharatioofresponsetothatproduced,byabsolutesumasthehighestobservedintheFSTFandOysterCreekanalyses(63%).TherefinedshellanalysisoftheNMP-1torusshell,performedbyTES,usingthereducedCOloaddefinitionincludesthefollowing:1.TheCOloadreductionresultingfromtheFSTFendcaps,17.1%and36.1%fornon-ventandventbays,respectively.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING v.r,jl4

-16-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19922.Thecombinationofindividualharmonicstressesintototalelementstresswasdonebyconsideringfrequencycontributionsat31Hzandbelow.TheactualcombinationwasdonebyaddingtheabsolutevalueofthefourhighestharmoniccontributorstotheSRSScombinationoftheothersforshellstress.Thiscombinationmethodanduseofthe31Hzcutoffaretheresultofextensivenumericalevaluationoffullscaletestdata.Includingthefrequencycontributionoutto50HzwouldincreasetheCOstressbyabout20psi,orlessthanI/2%.GeorgeBienkowski'sreviewoftheRandomPhasingRulesalsoaddressedthe,impactoftheendcapsonproducingconservativelymeasuredCOloads.Thisdiscussionbeginsatthebottomofpage3ofthereview.Here,thereviewreferstoanunreferencedcommunicationinReportsSMA12101.04-R002D(Reference7)andSMA12101.04-R003D(Reference8),fromDr.AlanBilanin,thatthepresenceofthebulkheadsintroducesafactorofconservatismofatleast1.33tothemeasuredCOloadingandresponsesfromwhichtheLDRamplitudes-werederived.Thisisseparatefromthephasingrulesgoverningthesummationofstressharmonics.Inaddition,an"AppendixA"ismentionedwhereinthiseffectisexamined.This"AppendixA"isassumedtobeanappendixtoGeorgeBienkowski'sreviewandithasnotbeenreviewedbyTESorCDI.ItisstatedthattheAppendixconcludesthatforfrequenciesthatarenotcorrelatedbetweenbays,theFSTFshouldproduce32%to35%higherloadsthanwouldexistinarealfacilityandthatonlythefundamentalfrequencynear6Hzshowsanycorrelationbetweendowncomers.Thereviewfurtherstatesthatifoneassumescorrelationbetweenbaysatthatfrequencyandrandomphasing(thisisnottherandomphasingofresponsestressharmonics)atallotherfrequencies,theoverallconservatismforIITELEDYNEENGlNEERlNGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING 1~ck

-17-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992theaveragepressure~abeaslowas17%(notethereviewsays"may"),whileattheresponseleveltheFSTFconservatismwillrangefrom18%forhoopstressto38%fortheaxialstress.TheseindependentBrookhavengeneratedresultsarenotunlikethosethatNNPCiscurrentlypresenting.RobertKennedy,inhisreportSMA12101.04-R003D(Reference8),usesthisadditional33%conservatism,advancedbyDr.AlanBilanin,toaccountfortheuncertaintyofusingonlythethreepeakresponsesfromFSTFtestsH-8,H-llBandM-12.Ontheotherhand,GeorgeBienkowskiarguesthatthisuncertaintyestimateisprobablyexcessivelyconservativeandthat7%ratherthan33%uncertaintywouldprovideahighconfidencelevelofnon-exceedance.UsingtheabovenumbersfromGeorgeBienkowski'swork,wehaveasaminimumthatthemeasuredstressesareconservativebyatleast:(1-[(1-.18)(1+.07)])100=(1-,88)100=12%Also,nomentionhasbeenmadeinGeorgeBienkowski'sreviewregardingtheeffectofthe4-8-4downcomerconfiguration.ThejointTES/CDIwork(Reference1)evaluatesthiseffectaswell.Sowearelookingfor17%reliefinthe8downcomerbaysandtheBrookhavenworkshowsatleast12%relief.Thedifferencemaybeattributedtothe4downcomerbaystoeithersideofthe8downcomerbay.Inaddition,the36%reductionthatwehaveevaluatedforthe4downcomerbaysalsoincludesthefactthatonlyhalfasmanydowncomersarepresentasareintheFSTF,aneffectnotmentionedinGeorgeBienkowski'sreview.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

,2~gp'f

-18-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,1992TheconservatismsrelatedtotheendcapeffectdiscussedhereinwerenotincludedinGeorgeBienkowski'sconclusionsregardingthephasingrules.Oncetheendcapeffectisincluded,thereisnoreasonwhytheresultingresponseshouldboundFSTFdata.Indeed,byitsverynature,itshouldnotboundFSTFdata.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING r1V~wg/

-19-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,19926.0NAP-ITORUSSHELLCNTR'SCertifiedMaterialTestReortReviewforTorusShell(Reference11)TEShasreviewedtheTorusShellCertifiedHaterialTestReports(A201GrBFBX).Astatisticalanalysiswasperformedusingthis(large)sampledatatodeterminethe.99confidenceintervalestimateofthemeanyieldandultimatestrengthofthismaterial.TheCoderequiresthattheminimumyieldandminimumultimatestrengthofthematerialbeusedtodeterminetheallowablestressintensity(SgC)asfollows:SgCat70'Fisthelessorof1'15SYor1.11SUTherefore,TEShasassumedthattheminimumyieldandultimatestrengthsofthematerialareboundedbyusingtwosamplestandarddeviationsfromthestatisticallyestimatedminimummeanvalues.Basedonthecalculationsusingtheabovestatedcriteria,theCodeallowablestressintensitywouldbeestimatedat:SNC=17600at70'FUseofthisestimatedallowablestressintensitywillprovideanadditional1100psireliefascomparedtothepresentCodeallowableof16500psiwhichwasusedduringtheHarkIContainmentProgramanalysisforATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING S',fI

-20-TECHNICALRESPONSETODOCKETNO.50-220NOVEHBER19,1992thetorusshellmaterialforthefullrangeofanticipatedeventtemperaturesfrom70to350'F.Intermsofreliefontheshellthicknessrequirements,theincreasedallowablewillprovidejustunder1/32inchesor6/.additionalmargin.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING qXmy<~),+LC

-21-TECHNICALRESPONSETODOCKETNO.50-220NOVEHBER19,19927.0RELEVANCETOSAFETYWedonotviewthisissueasadverselyaffectingsafety.ItcanberegardedastheeffortundertakentoremoveknownconservatismsthroughrefinementofthepostulatedappliedCOloads.Inaddition,resolutionofCodeissuespertainingtotheallowablematerialstrengthandallowablenominalcorrosion,delineatedinSection4.0,"AdditionalConservatisms",ifintroduced,wouldprovideforaddedmarginforthetorusshellstresses.Also,themanyotherconservatismsdiscussedthroughoutthisresponse,ifaccountedfor,wouldfurtherreducetheshellstress.TorusCorrosionRate(>)NiagaraHohawkundertookanewcorrosionmonitoringprograminAugust,1989.Underthisprogram1'3'ridsonall40midbaybottomplateswereUTinspected.HPRAssociates,Inc.ReportHPR-1152delineatestheresultsofthisinspection.Thesemeasurementsdidnotshowanysignificantlossduetocorrosionorpittingevenatthenormalwaterlevelregion;andtherewerenowallthicknessmeasurementsthatwouldrequireapplicationofthemethodsdescribedinTeledyneReportTR-6801-2,reference11.HPRquantifiedtheshellthicknessloss,over20years,bycomparingthemeasuredshellthicknessvaluestothecalculatedoriginalplatethickness.Thirtyfourshellplates,traceabletotheoriginalmilcertifications,wereusedinthiscomparison.Originalplatethicknesseswerecalculatedusingplatedimensions,weightanddensityofthesteel.ThesethicknesseswerecomparedtotheUTthicknessobtainedinAugust,1989,onthesame34plates.Theresultsindicatedanaveragecorrosionlossof0.8milsperyear.Thisratetranslatestoatotallossof32milsoraboutI/32"overtheoriginalprojected40yearplantlife;andcomparescloselytotheratepredictedbyRadiological8ChemicalTechnology,Inc.(RCT),basedonanalysisofsludgesamplesin1979.Duetovariationsin(1)ThisinformationprovidedbyNHPC.ATELEDYNEENGlNEERlNGSERVlCESADIVISIONOFTELEDYNEBROWNENGINEERING Vrl-k

-22-TECHNICALRESPONSETODOCKETNO.50-220NOVENBER19,1992theoriginalplatedimensionsandweights,andmeasurements,onestandarddeviationwasaddedtothe0.8milsperyear.Thisresultedinaconservativepredictionof1.26milsperyearcorrosionrate.Additionally,NiagaraMohawkPowerCorporationcommittedtoperformUTmeasurementsonasixmonthbasisandprovidetheNRCwiththeresults(Ref.November22,1989,letterC.TerrytoNRC).Sincebaselineestablishmentofthenewcorrosionprogramof1989,six(6)six-monthmeasurementshavebeenconducted,themostrecentofwhichwerejusttakeninSeptember,1992.Furtheranalysisandtrendingofthesemeasurementsindicatethataconservativecorrosionrateof1mi-l/yrincludingonestandarddeviationisamorerealisticcorrosionratethanthebaselineestimateof1.26mils/yr.Themostprobablepredictionofcorrosionrateisstill0.8mils/yr,butthelaterresultshavereducedthestandarddeviationto+0.2mils/yr.Thelatest(September,1992)UTmeasurements,onthethinnesttorusplatesindicatestheaveragethicknessoftheworstplateis0.453inches.Thisthicknessrepresentsanaverageof63individualmeasurements(calibrationadjusted)ona1'3'ridonthatworstplate.Worstcaseindividualmeasurementsonthisorotherplateshavebeenaslowas0.445inches(calibrationadjusted).Overthenextyear,theprojectedworstcaselossofthicknesswouldbeonemilor0.001".Thiswouldtheoreticallyreducetheworstgridtoa0.452"averagethicknessandtheworstindividualpointto0.444".TheaveragethicknesswouldcomplywithminimumthicknessallowedbyTeledyneReportTR-6801-2,reference2,andtheoreticallyprovideforaboutanother5yearsofoperationbeforereachingthecurrentminimumthicknessallowed.However,the0.444"individualpointwouldhavetobeanalyzedforcompliancetoTR-6801-2bythemethodsoutlinedtherein.TheanalysissubmittedunderNHPC'sMay14,1991coverletter(Reference1),totheNRCsupportsanyaveragethicknessorindividual6TELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

-23-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19i1992pointdownto0.431".Inaddition,themanyotherconservatismsdiscussedherein,forwhichnocredithasbeenrequestedortakenintoaccount,wouldsupportreductionsinexcessofthis.Therefore,lossofthicknessoccurringduetoprojectedcorrosionoverthenextyearwouldbewithinlimitsofsupportinganalyses.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

~jlV~Aall~

-24-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,199

28.0REFERENCES

1.TESTechnicalReportTR-7353-1,Revision2,"NineMilePointUnit1,ReductioninHarkITorusProgramCondensationOscillationLoadDefinitionandResultingEffectonMinimumShellThicknessRequirements,"datedJanuary14,1992.2.-TESReportTR-5320-1,Rev.1,"HarkIContainmentProgram,Plant-UniqueAnalysisReportoftheTorusSuppressionChamberforNineMilePointUnit1NuclearGeneratingStation,"datedSeptember21,1984.3.G.E.ReportNED0-21888,Rev.2,"MarkIContainmentProgramLoadDefinitionReport,"datedNovember1981.4.G.E.ReportNEDE-24840,"HarkIContainmentProgram-EvaluationofHarmonicPhasingforMarkITorusShellCondensationOscilla-tionLoads,"datedOctober1980.5.G.E.ReportNED0-24574,Rev.1,"HarkIContainmentProgram-Plant-Unique.LoadDefinition-NineMilePoint1NuclearGen-eratingPlant,"datedJuly1981.6.ASMEB&PVCode,SectionIII,Division1throughSummer1977.7.StructuralMechanicsAssociatesReportSHA-12101.04-R002D,"ResponseFactorsAppropriateforUsewithCOHarmonicResponseCombinationDesignRules,"datedMarch1982.8.StructuralMechanicsAssociatesReportSHA-12101.04-R003D,"AStatisticalBasisforLoadFactorsAppropriatefo}UsewithCOHarmonicResponseCombinationDesignRules,"datedMarch1982.ATELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING

~Q~f':trl<~0

-25-TECHNICALRESPONSETODOCKETNO.50-220NOVEMBER19,199

28.0REFERENCES

(Cont'd)9.G.E.SupplementarySupportEffort(SSE)ResponseNumber310,datedFebruary8,1982.10.ContinuumDynamicsTechnicalNoteNo.90-11,"ReductionofTorusShellCondensationOscillationHydrodynamicLoadsforNineMilePointUnit1,"datedNovember1990.ll.TESTechnicalReportTR-6801-2,"MarkITorusShellandVentSystemThicknessRequirements,"NineMilePointUnit1NuclearStation,January29,1988,Rev.l.12.HarkIContainmentProgram,StructuralAcceptanceCriteria,PlantUniqueAnalysisApplicationGuide,NED0-24583-1,October1979.13.TESCalculationPackage7353-1,Revision2,"NineNilePointUnit1,ReductioninHarkITorusProgramCondensationOscillationLoadDefinitionandResultingEffectonMinimumShellThicknessRequirements,"datedJanuary14,1992.14."ReviewoftheValidityofRandomPhasingRulesasAppliedtoCOTorusLoads,"byGeorgeBienkowski,datedAugust25,1983,ContainmentSystemsGroup,DepartmentofNuclearEnergy,BrookhavenNationalLaboratory.15.UnitedStatesNuclearRegulatoryCommission,"SafetyEvaluationbytheOfficeofNuclearReactorRegulationRelatedtoProposedDefermentofTorusModifications,NiagaraMohawkPowerCorporation,NineMilePointNuclearStationUnit1,DocketNo.50-220,"datedAugust25,1992.IIITELEDYNEENGINEERINGSERVICESADIVISIONOFTELEDYNEBROWNENGINEERING gg~g>.g4: