ML17258A520

From kanterella
Revision as of 12:39, 29 June 2018 by StriderTol (talk | contribs) (Created page by program invented by StriderTol)
Jump to navigation Jump to search
Forwards Info Re Seismic Capacity of Reactor Coolant Pump. Action Proposed to Close Item
ML17258A520
Person / Time
Site: Ginna Constellation icon.png
Issue date: 09/11/1981
From: NELSON T A
LAWRENCE LIVERMORE NATIONAL LABORATORY
To: RUSSELL W T
Office of Nuclear Reactor Regulation
Shared Package
ML17258A195 List:
References
CON-FIN-A-0415, CON-FIN-A-415 SM-81-247, NUDOCS 8202090379
Download: ML17258A520 (19)


Text

LawrenceLivermore NationalLaboratory September ll,1981SM81-247Docket850-244FINA0415Mr.WilliamT.Russell,BranchChiefSystematic Evaluation ProgramBranchDivisionofLicensing OfficeofNuclearReactorReg.Washington, D.C.20555

DearBill:

Ihaveenclosedinformation regarding theseismiccapacityofareactorcoolantpumpattheGinnaplant.Whilethisopenitemhasnotbeencompletely

resolved, arathersimpleactionisproposedwhichshouldclosethisitem.Sincerely, TAN/mg0131mEnclosure ThomasA.NelsonStructural Mechanics GroupNuclearTestEngineering Divisionsooaoqoo79 saoiisPDRADOCK05000244IPDRAnEndleaf~tnffyEl~NiVeetyef CaifemiaPQBaX808LimmesCaifemia94550~TelephOne(415)422-1100

~TWX9f0-38&8339 UCLLLLWR ii11IEPP.'I STRUCTURAL IECHAfllCS ASSOClATESACollf.Coro.5160BirchStreet,NewportBeach,Caiif.S2660(714)833-7552SMA12205.20August27,1981Mr.ThomasA.Nelson(L-9Q}LawrenceLivermore Laboratory NuclearTestEngine'ering DivisionP.0.Box08Livermore, California 94550P

Subject:

Resolution ofOpenItemsonGinnaEquipment l

DearTom:

!nmyJune22letter,therewerestilltwoopenitemsforGinnaequipment whichwere:1.ControlRodDrivesand.Supports 2.PrimaryCoolantPumpThisletteristoadviseyouofthecurrentstatusoftheseitems..GeorgeWrobelofRochester Gas&Electric(RGE}CompanyarrangedformetotalkdirectlytoRobertKellyof'Westinghouse, whowasresponsible forproviding muchoftheGinnaequipment seismicdocumentation.

Regarding loadingdocumentation ontheCRDassemblies, R.KellywillsendhisanalysisofCRDhousingandseismicsupportloadingfor'a0.8gstaticcoefficient loading.Aquickreviewoftheanalysisshouldresolvealloutstanding issues.ontheCRDsystem.Alegiblecopyoftheprimarycoolantpumpreportwastransmitted tomefromRGE.Conclusions fromthereviewofthatreportarecontained intheattachment, TomChengsuggested thatIbepresentatanupcomingopenitemsmeetingforGinnainSeptember.

InviewofthecurrentstatusofopenitemsforwhichIhavebeenresponsible, Idon'tfeelitnecessary formetoattend;the=

meeting.Aftermyconversation withR.KellyofWestinghouse, IfeelthattheCRDandprimarycoolantpumpissueswillbe.resolvedvery'quickly andcanbehandledwithoutameeting.Itis.therefore recommended thatthese'temsnotbeatopicofthemeeting.rYerytrulyyours,STRUCTURAL MECHANICS ASSOCIATES, INC.rZWRobertD.CampbellProjectManagerRDC:lcacc:T.Cheng{NRC)

~~ATTACHMENT REACTORCOOLANTPUMPSEISMICDESIGNREVIEMTheoriginalstressreportsubmitted forreview,wasillegible.

Amorelegib1ecopywassubmitted toSMAbyRGEonAugust17,1981,andthefollowing conclusions canbereachedfromreviewofthatsubmittal.

I~1.Thepumpwasanalyzedfora0.8ghorizontal staticcoef-ficientanda0.54g.vertical staticcoefficient.

Asreportedin.NUREG/CR-1821, the7Xdampedpeakspectralacceleration forbothhorizontal directions is0.55g'sresulting inavectorsumof0.78g's.Thus,theequiva-Ilentstaticcoefficient usedintheoriginalanalysisisconservative byasmallmargin.2.'llstressescalculated fdrthe0.8gHand0,54gVstaticcoefficients aiewithinallowables.

designated fortheoriginaldesignbasis.3.Analytical methodsusedinthedesignanalysisarereason-ableexceptintheca'seofpumpnozzles.4.Thepumpnozzlesanalysisisunrealistic andinadequate forthefollowi.ng reasons:a.Onlythestraightpipeportionofthenozzleswereevaluated.

Localmembranestressesinthepumpcasingwerenotcomputed.

ATTACHHENT Continued b.Thederivedpumpnozzleloadshavenoresemblance toactualachievable loads.Thepumpwasassumedtobesupported bythepipingforpurposesofderivingnozzleloads.Thisisprobablyhighlyconservative butnotnecessarily so.Actualpipingreactions areavailable andshouldbeusedinanevaluation ofthepumpcase.Aconversation betweenR.KellyofMestinghouse andR.CampbellofSMArevealedthat:.i1.TheSanOnofrepumpsareverysimilartotheGinnapumpsandthatadetailedfiniteelementanalysiswasconducted fortheSanOnofreUnitsforaspecified setofnozzleloads.2.AloopanalysisofGinnawasconducted byMestinghouse forseismicloading.Actualpumpnozzleloadsareobtainable fromtheanalysis.

Reconmended ActionsRGEshouldhaveVestingfiouse makeacomparison ofGinnavsSanOnofrepumpcasinggeometryandpumpnozzleloadsandscaleresulting stressesfromtheSanOnofrepumpanalysisforGinnanozzleloadingconditions.

Theload/stress comparison andacomparison ofnozzleandcasinggeometryshouldbesufficient todemonstrate seismiccapability oftheGinnaprimarycoolantpumps.

81 p~~r-STRUCTURAL lTIECHAAICS ASSOCIATES ACalif.Corp.5160ErchStreet,NewportBeach,CaN.92BBO(714)833-7552.

SMA12205.20June22,1981Mr.ThomasA.Nelson(L-90)LawrenceLivermore Laboratory NuclearTestEngineering DivisionP.0.Box808Livermore, California 94550

DearTom:

IntheReference 1submittal regarding reviewofopenitemsonRobertE.GinnaNuclearPowerPlant,therewerethree(3)openitemsstillremaining.

1.ControlRodDriveMechanism 2.ReactorCoolantPump3.SteamGenerator TubeSupportsITomChengoftheUSNRCcontacted meonJune16,toinquireifuseofa,sitespecificspectraforGinna,inlieuoftheregulatory guidespectrumanchoredto0.2g,wouldeliminate theoustanding items.'indicated thatfortwoofthethreeoutstanding items,lackofinformation, notmarginalstressconditions, wastheprincipal concern.'s aresultoftheconversa-tion.Ihavereexamined thethreeopenitemsandcaneliminate thesteamgenerator tubesupportoverstress problembyusingLevel0Service(faultedcondition) allowable stressesandscalinguptheoriginalanalysisresultstoapplicable floorspectralaccelerations.

Thecontrolroddriveandprimarycoolantpumpitemsstillremainopenbutshouldbeeasilyresolvedwithsubmittal ofthenecessary information.

Theremaining actionitemsandresolution ofthesteamgenerator tubesupportstressproblemaresum-marizedintheattachment.

Verytrulyyours,STRUCTURAL MECHANICS ASSOCIATES, INC.RobertD.CampbellProjectManagerRDC:lcaAttachment cc:J.Stevenson ATTACHMENT SUMMARYOFOPENITEMSFORROBERTE.GINNANUCLEARPOMRPLANTASPARTOFTHESEPPROGRAMCONTROLRODDRIVEMECHANISM Thesubmittals providedbythelicensee, References 2and3,donotcontainacorrelation betweenloadingusedintheanalysesandaccel-erationsattheRPVsupport.Aconclusion regarding seismicresistance oftheCRDsystemcannotbereachedwithoutsuchacorrelation.

Mesting-houseshouldbeabletosupplythenecessary information toRochester GasandElectric.

REACTORCOOLANTPUMPSTheReactorCoolantPumpStressReportsubmitted forreview,Reference 4,isillegible duetopoorreproduction quality.Alegiblereportneedstobesubmitted forreview.STEAMGENERATOR TUBESUPPORTSSection16ofReference 5documents thestressanalysisofSeries44steamgenerator internals forseismicloading.Aconservative analysisofthetubesupportsresultedinaprimarymembranestressof.7900psiforanequivalent staticcoefficient of0;19ghorizontal.

Thestresswascalculated intheligaments betweenthetubeholesandcircu-latingholesinalocalareaadjacenttoawrapperchannel.Theanalysis'conservatively ignoredaredundant loadpathfromthe.edgeofthetube"supporttothetubeholes.Reference 6,submitted forreview,isanupdateanalysisoftheSeries44steamgenerator butdoesnotaddressthe~1 tubesupportforhorizontal loading.NeitherReference 5or6describethedynamiccharacteristics ofthesteamgenerator anditsinternals noridentifythematerialofconstruction forthetubesupportplates.Reference 7,obtainedduringtheSSMRPprogram,documents agenericdynamicanalysisoftheSeries51steamgenerator forvaryingsupportstiffnesses andlocations.

TheSeries44steamgenerator issimilar,butsmaller.Thefundamental frequency oftheSeries51steamgenerator rangesfrom4.8to9.6Hzdepending uponthesupportstiffness andlocation.

Thefundamental modeispredominantly translation androckingofthesteamgenerator shell.Agenericresponsespectrumthatisflatthroughmostofthisfrequency rangewasusedtocomputeresponseaccelerations andloadfngsinthesteamgenerator.

Areviewwasconducted ofReference 7todetermine thedegreeofcouplingbetweenthesteamgenerator shellandtheinternalstructures andtoestablish thevalidityofconsidering thesteamgenerator asaSDOFsystemforestimating anappropriate equivalent staticcoefficient forevaluation ofthetubesupportplates..Itwasconcluded fromreviewofReference 7thatatthemostcriticalsupportlocation, asdetermined inReference 5,thatthetubesandtubesupportswouldaccelerate asrigidbodieswiththeshell;thus,usingthespectralacceleration fromtheGinnaspectrumforafunda-mentalfrequency ofabout.5Hzisareasonable approximation ofanequiva-lentstaticcoefficient, touseforevaluation ofthetubesupports.

FromtheresponsespectrainReference 8,for7Xdamping,themaximumspectralaccelerations at5Hzare0.58g'sineachoftwoortho-gonaldirections.

Combining thetwodirectional accelerations, theresulting maximumvectoris0.82g.UsingthisvalueandscalingtheligamentstresscomputedinReference 5,theresulting ligamentstressis34,095psi.Reference 7indicates thatSeries51steamgenerator internals areconstructed ofSA285-Grade Ccarbonsteel.Acomparisonoftheallowable stressforthismaterialatthedesign.temperature of556Ftotheallowable stressquotedfortheSeries44steamgenerator tube supportsinReference 5,indicates thattheSeries44tubesupportsarealsoconstructed ofSA285-Grade Corequivalent.

Thismaterialandadesigntemperature of'556Fareusedasabasisforestablishing allowable stressesfortheSafeShutdownEarthquake.

Thetubesupportsareconsidered tobeClass1plateandshelltypecomponent supportsandtheallowable primarymembranestressiscomputedforLevel0ServicefromAppendixFoftheASMECode,Reference 9.Theallowable stressisthegreaterof1.5Smand1.2Sy.ForSA285-Grade Cmaterial, 1.2Sygovernsandtheallowable stressis27840psi.NotethattheoriginaldesigncriterialimitedthetubesupportstresstoS.Comparison ofthecalculated andallowable stressforLevelDService-results ina22%overstress condition.

Ifsitespecificspectraanchoredto0.172gareconsidered inlieuofregulatory guidespectraanchoredto0.2g,thecalculated.

stressdecreases.

Decreasing thecalcu-latedstressbytheratiosofthesitespecificpeakgroundacceleration dividedbythe0.2gpeakgroundacceleration usedtogeneratefloorspectra,theresulting ligamentstressis29,320psi.Thisisstill.about.5'.3XovertheLevelDServiceallowable of27,840psi.Inconsideration oftheconservatism inherentinobtaining thecalculated stress,thecomputed5.3Xoverstress condition isconsidered acceptable forseveralreasons.1.Thesitespecificspectrumenvelopehaslowerspectralaccelerations inthefrequency rangeofthecontainment structure thantheregulatory guidespectrumifbothareanchoredtothesamepeakgroundacceleration.

Conse-quently,thein-structure responsespectrawillbelowerthanthoseinReference 8.'.ThestaticanalysisfromReference 5didnotaccountfor'heredundant loadpathbetweentheoutsidediameterofthetubesupportplateandtheouterrowoftubeholes.'Thedegreeofconservatism couldnotbeevaluated sinceperti-nentdimensions arenotprovidedinReference 5.The'egreeofconservatism iscertainly greaterthan5Xthough.

3.Theevaluation considered theapplicable acceleration tobethevector.sumofthetwoorthogonal directional acceler-ations.Thisassumesthatbothdirectional responses areinphaseandthattheresulting vectorisalignedintheworstdirection.

4-Thein-structure responsespectrawerepeakbroadened

+15%andsmoothedsothattheresulting spectraareessentially flatfrom2-1/2to9Hz,whichcoverstherangeoffunda-mentalfrequency forthesteamgenerators.

Items3and4areconsistent withcurrentregulatory criteriaandareprudentconservatisms tocovermanyoftheuncertainties inthesimplified treatment ofthetubesupport.Recommended acceptance is,therefore, basedontheconservatism ofItems1and2beingsufficient toovercometheestimated 5.3foverstress condition.

Furtheranalysesorsubmittals fromthelicenseearenotconsidered necessary.

REFERENCES 1.SMAletter,R.D.CampbelltoT.R.Nelson,ReviewofOpenItemsResulting fromSeismicReviewoftheRobertE.GinnaNuclearPowerPlantasPartoftheSEPProgram,4May1981.2.HighSpeedControlRodDriveStressAnalysisJune26,1968.3.ControlRodDriveMechanism SeismicFrameCalculations, August13,1968.4.StaticSeismicLoadStressAnalysis, ModelRGEPump(V-11001-BI),

July30,1968.5.VerticalSteamGenerator StressReport,Mestinghouse ElectricCorporation, TampaDivision, April,1969.6.MTO-SM-75-028, 44SeriesSteamGenerator StressReport,ExternalLoadAnalysisUpdate,May,1975.J7.StressReport;51SeriesSteamGenerator, GenericSeismicAnalysis, westinghouse ElectricCorporation, TampaDivision,

December, 1974..8.NUREG/CR-1821, SeismicReviewoftheRobertE.GinnaNuclearPowerPlantasPartoftheSystematic Evaluation Program,15November1980.9.ASMEBoilerandPressureVesselCode,SectionIII,NuclearPowerPlantComponents, Appendices, 1980.

STRUCTURAL ICCHAAICS ASSOCIATES AC~Ilt.Coro.5160BirchStreet,NewportBeach,Calif:82660(714)833-7552SMA12205.20May4,1981Mr.ThomasA.Nelson(L-90)LawrenceLivermore Laboratory NuclearTestEngineering DivisionP.O.Box808Livermore,California 94550

DearTom:

SMAhasreviewedthepackageofdocuments transmitted withyourApril15letteraddressing openitemsonGinna.Ourcommentsandrecommended actionarecontained intheenclosure.

Verytruly'yours, STRUCTURAL MECHANICS ASSOCIATES, INC.piRobertD.CampbellProjectManagerRDC.mwEnclosurecc:J.Stevenson w/encl.

REVIEWOFOPENITEMSRESULTING FROMSEISMICREYIEMOFTHEROSERTE.GINNANUCLEARPOWERPLANTASPARTOFTHESEPPROGRAMReference 1documents areview.conducted bytheLawrenceLivermore Laboratory andtheirconsultants ontheseismicadequacyoftheRobertE.GinnaNuclearPowerPlant.Conclusions oftheadequacyoftheseveralitemsintheNSSSsystemwerebaseduponsummaryinformation provided; however,thesourcesofthesummaryinforma-tionwerenotavailable forindependent review'.ThoseitemslistedinSection5.4ofthereportascomponents forwhichseismicdesignanalyses.

havenotbeenindependently verifiedare:ReactorControlRodDriveReactorVessel'upports SteamGenerator ReactorCoolantPumps-Pressurizer anditsSupportsReferences 2through7wereprovidedtoSMAinresponsetotheaboveidentified openitems.-Thefollowing summaryandconclusions resultedfromSMA'sreview'of thesubmittals-.

1)ControlRodD~iveMechanism Inreference 7,theallowable bending'oment intheCRDMduetoseismicloadsisdeveloped.

Thisreportdoesnot,however,provideacorrelation betweenbendingmomentsandacceleration levels,Thus,noconclusion canbereachedonthebasisofthesubmittal.

Reference 2containsastressanalysisofthecontrolroddrivesupportstructure.

Theanalysisprovidesstressesas'functionofastaticload"P"inpounds.Thereisnocorre-lationbetweenthisstaticloadandacceleration level.Therefore, aconclusion ontheseismiccapability ofthesupportstructure cannotbereachedbaseduponthesubmittal.

ls5~~SMA'sexperience withtheSSMRPreference plant,whichuses106Afull lengthcontrolroddrive.mechanisms, indicated thatalargemarginofsafetyexistsfora1.15gspectralaccel-erationloadingcondition andwewouldnotanticipate aseismicproblemwiththeGinnaCROM.Recomme'rided action-CROMloadsdocumentation applicable toGinnaneedtobesuppliedtoSMAforfinalresolution.

2JReactor'Vessel SuortsDocume'ntation verifying theseismicadequacyofthereactorlvessel;supportswasnotsubmitted.

BaseduponSMA'sSSMRPexperience fornozzlesupported RPY'stheseismicinducedstressesinthenozzlesandadjacentshellareverysmallandthegoverningelementforRPVsupportistheconcrete,

.shieldwall.Theshieldwallwasconsidered inRef.1tobeadequatetowithstand the0.2gSSE.3lSteamGenerator Reference 6containsa1969staticanalysisoftheSeries44steamgeneator.

Themostcriticalareaduetoseismicload-ingidentified inthisreportisthetubesupportbaffleligaments whicharestressedpastyieldfora0.38ghorizon-talstaticload.TheSEPrevisedspectra'resultinanSRSSresponseof0.85ghorizontal whichwillincreasethestressesbyafactorof2.24.Thus,basedonthesubmittal, thetubesupportbafflesareoverstressed forthe0.2gSSE..Thestaticanalysisatthesetubesupportbaffleswas.donequiteconser-vatively, however,andamorerigorous'~nalysis willmostlikelyresultinlowerstateofstress.In1975,anupdateontheseries44steamgenerator wascon-ducted(Reference 4)butthetubesupportareawasonlyevaluated forverticalseismicloadingandnotforthe PI horizontal seismicloading.Thus,theresultsofReference 4cannotbeusedtoupdatetheresultsofReference 6intheareaofconcern.Recommended Action-Documentation evaluating thetubesupportbaffleligaments forhorizontal seismicloadingshouldbesubmitted forreview,4)ReactorCoolantPumsReference 3summarizes thestressesinducedinthereactorcoolantpumpbya0.8g'orizontal anda0.54gverticalload-ing.ThereportedstressesarebelowtheASMECodeallowables, butSMAisunableto'evaluate themodelortheana1ysisduetopoorreproduction qualityofReference 3.Ifthestaticanalysisofthepumpcanbeshowntobevalid,thenthestressesduetotherevisedGinnaspectraloadingwillbelessthanthosecontained withinReference 3,"andthusaccept-able.Recommended Action--AlegiblecopyofReference.3 shouldbeprovidedtoSMAforreview.5).Pressurizer Reference 5containsa1969stressreportofan1800cubicfootpressurizer.

Allpressurizers from800to1800cubic'eetwithcastand'fabricated headsutilizethesamesupportskirts,thusconservative genericanalysiswasconducted for.theheavier1800cubicfootmodels.Basedonthisreport,theloadsresulting fromthenewGinnaspectrawill.causeanover-stressedcondition withinthesupportf1ange,Thisisaveryconservative conclusion, however,sincetheGinnapressurizer isasmaller800cubicfootmodelplus,theflangeanalysisitselfwasveryconservatively conducted usingabeambendingmodelinlieuofamorerigorousfiniteelementmodel.

The1973pressurizer reportreferredtoinReference 1andthepr'essurizer summarystressreport(reference 45ofRefer-ence1)werenotsuppliedtoSMAforreview;thus,theade-quatecapacityfora6.7gequivalent staticloadportrayed inreference 45ofReference 1cannotbeverified.

Lateranalysis(References 8and9)obtainedintheSSHRPprogram,'for theSeries511800cubicfootcastandfabricated headpressurizers showedthesupportstobeacceptable fora0.96g.horizontal and0.64gverticalloadingcondition.

Thislateranalysisutilizedafiniteelementmodeloftheskirtandflangeasopposedtotheconservative beamtheory.usedinreference 5.Recommended Action-Noactionrequired.

Thepressurizer stressanalysiswithinreferences 8and9andthestatements inReferences 5,8and9thatthesupportskirtsareidentical forthe1800cubicfootandthe800cubicfootdesignssub-=stantiate theadequacyoftheGinnapressurizer fora0.2g.SSE.