ML17265A536

Forwards Response to NRC 981203 RAI Re Resolution of Unresolved Safety Issue USI A-46.Util Does Not Agree with NRCs Interpretation.Detailed Bases,Encl

ML17265A536
Person / Time
Site:	Ginna
Issue date:	02/02/1999
From:	MECREDY R C ROCHESTER GAS & ELECTRIC CORP.
To:	VISSING G NRC, NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
REF-GTECI-A-46, REF-GTECI-SC, TASK-A-46, TASK-OR NUDOCS 9902100108
Download: ML17265A536 (31)

Text

CATEGORY1REGULATORY INFORMATION DISTRIBUTION SYSTEM(RIDS)ACCESSXON NBR:9902100108 DOC.DATE:

99/02/02NOTARIZED:

NOFACXL:50-244 RobertEmmetGinnaNuclearPlant,Unit1,Rochester G,,AUTH.NAME,AUTHORAFFILIATION MECREDY,R.C.

Rochester GasSElectricCorp.RECIP.NAME RECIPIENT AFFILIATION VISSING,G.

DOCKET05000244

SUBJECT:

ForwardsresponsetoNRC981203RAIreresolution ofunresolved safetyi'@sueUSXA-46.DISTRIBUTION CODE:A025DCOPIESRECEIVED:LTR ENCLSIZE:TITLE:SeismicQualification ofEquipment inOperating PlantsNOTES:License Expdateinaccordance with10CFR2,2.109(9/19/72)

A-46-GL-87'05000244ERECIPIENT IDCODE/NAME OGC/HDS3VISSING,G.

XNTERNAILECENTER01NREGBNRR/DRCH/HICB NRR/DRPE/PD1-3 EXTERNAL:

NRCPDRCOPIESRECIPIENT LTTRENCQIDCODE/NAME 11XPD1-1PD11MPR/DE11NRR/DE/EMEB 11NRR/DRCH/HOHB 11NRR/DISA/SRXB 11COPIESLTTRENCL111121111NNOTETOALL"RIDS"RECIPIENTS:

PLEASEHELPUSTOREDUCEWASTE.TOHAVEYOURNAMEORORGANIZATION REMOVEDFROMDISTRIBUTION LISTSORREDUCETHENUMBEROFCOPIESRECEIVEDBYYOUORYOURORGANIZATION, CONTACTTHEDOCUMENTCONTROLDESK(DCD)ONEXTENSION 415-2083TOTALNUMBEROFCOPIESREQUXRED:

LTTR13ENCL13 VC4 Ar)rnROCHESTER GASANDElECTRICCORPORATION

~89FASTAVENUE,ROCHESTER N.Y.Id6d9.0001 ROBERTC.MECREDYVicePresident NuclearOperating GroupAREACODE716546-2700February2,1999U.S.NuclearRegulatory Commission DocumentControlDeskATTN:GuyVissingProjectDirectorate I-1Washington, D.C.20555-0001

Subject:

ResponsetoNRC"SecondRequestforAdditional Information" (RAI)ontheresolution ofUnresolved SafetyIssue(USI)A-46.R.E.GinnaNuclearPowerPlantDocketNo.50/244

Reference:

A.LetterfromRobertC.Mecredy(RGEE)toDocumentControlDesk(NRC),datedJanuary31,1997,"Resolution ofGenericLetter87-02,Supplement 1andGenericLetter88-20,Supplements 4and5(SeismicEventsOnly)."B.C.LetterfromGuyS.Vissing(NRC)toDr.RobertC.Mecredy(RGRE),datedApril6,1998,"RequestforAdditional Information ontheresolution ofUnresolved SafetyIssue(USI)A-46."LetterfromRobertC.Mecredy(RG&E)toDocumentControlDesk(NRC),datedMay27,1998,"Response toRAIonUSIA-46."D.

DearMr.Vissing:

LetterfromGuyS.Vissing(NRC)toDr.RobertC.Mecredy,datedDecember3,1998,"SecondRequestforAdditional Information".

1(/QOThisletterprovidesresponses totheNRC's"RequestforAdditional Information" (RAI),datedDecember3,1998(Ref.D).Enclosures 1and2alongwiththeparagraphs belowrespondtoQuestion1partsaandbregarding theuseofGIP"MethodA"atGinnaStation.Responses toquestions 2,3and4regarding specificSQUGscreening methodsandtestingdataareprovidedinEnclosure 3.9902i00108 990202PDRADQCK05000244PPDR Theuseof"GIPMethodA"isdescribed intheGenericImplementation Procedure, Revision2(GIP-2),theSupplemental SafetyEvaluation ReportNo.2(SSERNo.2),andthedocuments referenced inGIP-2uponwhichGIP-2isbased.RG&EusedMethodAtoestimateseismicdemandforcertainequipment within40feetofeffective gradeatGinna.TheNRChasquestioned RG&E'suseofMethodAonthebasisthatMethodAmaybeusedonlyiftheamplification factorbetweenthefree-field groundresponsespectrum(GRS)andthecalculated in-structure responsespectra(ISRS)beingusedbytheplantisnotmorethanabout1.5.TheNRCpositionisbasedontheirinterpretation ofthelanguageonpage4-16oftheGIPwhichsaysthat"theamplification factorbetweenthefree-field responsespectraandthein-structure responsespectrawillnotbemorethanabout1.5...".Rochester GasandElectricdoesnotagreewiththeNRC'sinterpretation.

ItisRG&E'spositionthattheapproachusedforapplyingandimplementing GIPMethodAforestimation oftheseismicdemandonequipment atGinnaforresolution oftheUSIA-46programisappropriate andtechnically justified.

DetailedbasesareprovidedinEnclosure 1.WithrespecttotheNRC'squestionregarding differences betweenthein-structure responsespectraandthe1.5xgroundresponsespectra,RG&Enotesthatthesespectraweregenerated usingconservative methodsandassumptions (typicalofmostnuclearplantresponseanalyses) whichartificially increased theamplifications overthosewhichwouldbeexpectedinanactualearthquake.

Adetailedqualitative assessment oftheseconservatisms areprovidedinEnclosure 2.Basedontheabove,andtheinformation inEnclosures 1and2,webelievethatRG&Ehasproperlyinterpreted theconditions onuseofMethodA,andthattheseconditions appeartohavebeenunderstood andacceptedbytheNRCstaffuntilrecently, afterRG&Ecompleted'the USIA-46reviewsatGinna.Tochangethisinterpretation atthisstageintheprogramforresolution ofA-46wouldbeinconsistent withthespiritandintentofA-46andwouldalsorequirereworkofequipment oradditional analysesandevaluations withoutacommensurate safetybenefit.PleasecontactGeorgeWrobelat(716)771-3535ifyouhaveanyadditional questions.

Verytrulyyours,RobertC.Mecredy Enclosures (3)xc:Mr.GuyS.Vissing(MailStop14B2)ProjectDirectorate I-1DivisionofReactorProjects-I/IIOfficeofNuclearReactorRegulation U.S.NuclearRegulatory Commission Washington, D.C.20555RegionalAdministrator, RegionIU.S.NuclearRegulatory Commission 475Allendale RoadKingofPrussia,PA19406Mr.P.DrysdaleU.S.NRCGinnaSeniorResidentInspector Enclosure 1BasesforInterpretation andImplementation ofGXP-2RulesforMethodAItisRochester GasandElectric's positionthatRochester's GinnaStationhasproperlyinterpreted andimplemented therulesforuseofGIPMethodAaspreviously reviewedandacceptedbytheNRC.Thebasesforthispositionareasfollows:SQUGandRochester GasandElectric's Interpretation oftheGIPThecautiongivenonpage4-16ofGIP-2liststwolimitations onuseofMethodA:Equipment shouldbemountedinthenuclearplantbelowabout40feetabovetheeffective grade,andEquipment shouldhaveafundamental naturalfrequency greaterthanabout8Hz.Theintroductory wordinginGIP-2forthesetwolimitations providesthebasesorpurposesforimposingthem,namely(1)tolimitamplification tonomorethanabout1.5and(2)toavoidthehigh-energy frequency rangeofearthquakes.

Thespecificlimitations whichareintendedbytheSQUG/NRCexpertpanel(SSRAP)andSQUGtosatisfythesebasesareincludedinthetwobulleteditemslistedabove.Thestatement onpage4-16that"theamplification willnotexceedabout1.5"istheexpectedresultofmeetingtheabovelimitations, notathirdcondition.

Atnotimewasacomparison ofMethodAamplification withthatofcalculated ISRSeverintended.

Infact,theentirecontextofthecautiononpage4-16ofGIP-2makesclearthattheadvantage ofMethodAisthatequipment meetingthetwobulletedlimitations above"canbeevaluated withouttheneedforusingin-structure responsespectra..."

2.TheIntentoftheGIPisClearandSSRAPAgreesTheGIP(page4-11)citestheSSRAPreportasthebasisfortheBoundingSpectrumwhichisusedinMethodA,andrequiresuserstoreadandunderstand it.TheSSRAPreportclearlyexplainsthelimitations andconditions whichappearonpage4-16oftheGIP.SSRAP'sreportstates:

"Thus,itisSSRAP'sjudgmentthatamplifications greaterthanafactorof1.5areunlikelyinstiffstructures atelevations lessthan40feetabovegradeexceptpossiblyatthefundamental frequency ofthebuildingwherehigheramplifications occurwhensuchafrequency islessthanabout6Hz.Thus,forequipment withfundamental frequencies greaterthanabout8Hzintheas-anchored condition itwasjudgedthatfloorspectralamplifications within40feetofgradewouldbelessthan1.5whenreasonably computedusingmoremediancenteredapproaches."

[SSRAPReport,page102]TheSSRAPChairmananddeveloper ofMethodA,Dr.RobertKennedy,wascontacted bySQUGandconcurswiththeinterpretation giveninitem1above.TheNRCWasAwareofSQUG'sInterpretation WhenItWasDeveloped TheNRCbackfitanalysisinNUREG-1211, whichjustifies implementation oftheUSIA-46programbyaffectedlicensees, reliesontheconclusions reachedbySSRAPintheirreviewofseismicexperience data.NUREG-1211 statesthefollowing:

"TheNRCstaffhascloselyfollowedtheSSRAPworkandisinbroadagreement withitsconclusions.

Thestaffhasconcluded thatiftheSSRAPspectralconditions aremet,itisgenerally unnecessary toperformexplicitseismicqualification ontheeight(1)classesofequipment studied."

[NUREG-1211, page17](1)Theeightclassesofequipment citedinNUREG-1211 werelaterexpandedto20equipment classes.Notethatthisquotation specifically makesreference totheSSRAP"spectral conditions."

Thespectralconditions aredescribed inSQUG'spositiongivenaboveandwereincludedinGIP-2.TheuseofMethodAwaspreviously reviewedandacceptedbytheNRCandSSRAPrepresentatives duringtwopilotplantreviewsconducted in1987and1988.Thesereviewsaredocumented inGIP-2References 16and25.Thespecificmaterialpresented totheNRCrepresentatives onuseofMethodAisdescribed inthereportoftheBWRpilotreviewasshowninFigure1.Notethattheseismicdemandcriteriadescribed duringthistrialplantreviewarethesameasdescribed initem1above.NRCandSSRAPrepresentatives raisednoobjections totheapproachusedbySQUGinconducting thesetrialplantreviews.Thetopicsdiscussed withandcommentsmadebyNRCandSSRAPrepresentatives duringtheBWRpilot'eview areincludedinFigure2;notethatseismicdemandinformation wasdiscussed insomedetail.

TheRochester Gas&Electric/SQUG interpretation oftherulesforapplyingMethodAisalsoconsistent withtheSQUGtrainingcourseonuseoftheGIPmethods'igure 3isanexcerptfromtheclassnotesusedduringthiscourse.Itshows,inSlide26,severalscreening methodsforcomparing equipment capacitytodemand.Screen52illustrates usesofGIPMethodAasdescribed inItem1above'hat is,ifequipment isbelow40feetandabove8Hz,andtheBoundingSpectrumenvelopes thegroundresponsespectrum, theequipment isacceptable.

ThistrainingmaterialwasusedduringthefirstsessionoftheSQUGtrainingcourseheldduringtheweekofJune22,1993.TwoNRCstaffmembers(P.Y.Chen,MichaelMcBrearty) andaNRCcontractor (KamalBandyopodhyay) attendedthisinitialsessionandlaterprovidedcommentsonthetrainingcourseinaletterdatedAugust28,1992.TheNRCdidnotraiseanyobjections totheapproachtaughtbySQUGinthiscourseforapplyingMethodA.Subsequent tothisinitialsessionofthecourse,11additional NRCstaffmembersandcontractors attendedothersessionsofthiscourse;similarly, noneofthemraisedobjections tohowSQUGwasteachinguseofGIPMethodA.NRCInterpretation RendersMethodANotUsefulThe..NRCinterpretation isthatMethodAcanbeusedonlywhencalculated ISRSarelessthan1.5xGRS.Thisinterpretation negatesthevalueofusingMethodAbecauseitcouldonlybeusedwhenitproduceshigherseismicdemandthanMethodBwherecalculated ISRSareused.Underthisinterpretation, MethodAwouldneverbeused.Thisisinconsistent withMethodA'development anduse,andwasnevertheintent.

FIGUREj.(Figure1containsanexcerptfromGIP-2,Reference 25,whichshowstheseismicdemandcriteriausedduringtheBWRTrialPlantReview.)SEISMICDEMANDCRITERIAAPPLICATION DEMANDCRITERIAEquipment inexperience databaseandlessthan40'bove243',andfundamental frequency greaterthan8Hz.1.ComparegroundSpectrawithboundingspectrum(Figure3.1inSSRAPreport).Equipment inexperience databaseover40'bove243'over281'levation) orfundamental frequency lessthan8Hz.Equipment coveredbyGERS(anyelevation, frequency).

Compareamplified floorresponsespectrawith1.5xboundingspectrum(FigureR1....,Rn, TI,...,Tn).

Compareamplified floorspectra(median-centered) with2/3xSERSforspecificequipment class.4~Anchorage evaluation andequipment-specific stresschecks(excluding valves):Equipment within40'f"grade"(elevation 281'ndbelow)andfundamental frequency lessthan8Hz.Equipment atanyelevation.

Utilizeaccelerations from(1.5xgroundspectra)x1.25.Utilizeaccelerations frommedian-centered amplifiedfloorresponsespectrax1.25.Equivalent staticloadfactorforallequipment (exceptvalves).Staticloadcheckforvalveoperator/yoke checks.Usingappropriate spectrawithmultiplier, use:Peakacceleration forflexibleequipment.

ZPAforrigidequipment.

Acceleration atcalculated fundamental frequency.

3G,Weakdirection.

Note:Ingeneral,forequipment.

withfundamental frequency greaterthan8Hzandwithin40'fgrade.1.5xgroundspectramaybeusedasanestimateofmedian-centered amplified floorspectra.

FIGURE2(Figure2containsanexcerptfromGZP-2,Reference 25,whichsummarizes theSSRAPandNRCcommentsontheBWRTrialPlantReview)Section8SENIORSEISMICREVIEWANDADVISORYPANEL(SSRAP)ANDNUCLEARREGULATORY COMMISSION (NRC)REVIEWSRepresentatives ofSSRAPandtheNRCattendedtheNMP-1walkdownonFebruary1stthrough3rd(Days8through10).OnFebruary1st,following radiation protection traininganddosimetry

issuance, theSSRAPandNRCrepresentatives werebriefedontheorganization andconductoftheNMP-1walkdown.

Theindoctrination andpre-walkdownmaterials coveredbySQUGforthewalkdownparticipants werealsoreviewedwithSSRAPandtheNRC.Theindoctrination/training materials aregiveninAppendixCandincludeinformation ontheorganization andscheduleofthewalkdown, therulesofconductintheplant,plant-specific dataontheseismicdemandlevelsforthewalkdown, andsummaryinformation onGIPrequirements forreviewofseismicdemandversuscapacity, equipment caveats,anchorage evaluation andevaluation ofinteractions.

TheNMP-1seismicdemandinformation usedforthiswalkdownwasdiscussed insomedetail.SQUGrepresentatives explained thattheseismicgroundmotionusedasabasisforthewalkdownisaplant-specific, uniformhazard,ground-motion spectradeveloped byA.CornellandR.McGuixeandisanchoredat0.13G.Thisground-motionspectraenvelopes theNMP-1FSARlicensing basisSSEspectrawhichisanchoredat0.11G.TheNMP-1uniformhazardground-motion spectraisshowninAppendixC.AlsointhisAppendixareamplified floorresponsespectradeveloped forNMP-1usingmodernreactorandturbinebuildingmodelsandthe0.13Guniformhazardground-motion spectra.Mr.Djordjevic (Stevenson sAssociates) reviewedthebasesfortheamplified floorresponsespectraandindicated thattheyarebeingusedasmean-centered, realistic spectra.Dr.Kennedy(SSRAP)expressed theviewthathebelievesthefloorresponsespectraareconservative andgenerally inaccordance withcurrentStandardReviewPlancriteria.

Asaresult,SSRAPconsiders thatitisnotnecessary toutilizetheadditional factorsofsafetyrecommended bySSRAPforusewithmean-centered spectra(1.5foruseofGERSand1.25foranchorage evaluation) inusingtheNMP-1floorresponsespectraduringthiswalkdown.

Asecondareadiscussed regarding theseismicdemandwastheeffective gradelevelatNMP-1.Atthissite,thetuxbinebuildingisfoundedonrockatelevation 243feetabovesealevel.Thereactorbuildingisfoundedonrockat198feet.Gradeelevation is261feet.Intheconstruction ofthebuildings, thesiteswereexcavated tothefoundation level,thebuildings constructed, andtheannularspacebetweenthebuildingandtherockexcavation wasbackfilled withcrushedstoneuptothe251footgradeelevation.

Anelevation viewoftheplantisincludedinAppendixC.SQUGandNMPCrepresentatives explained thatwhiletheybelievelateralsupportisprovidedbythecrushedstonebackfill, ithasbeenconservatively assumedforthepurposeofthiswalkdownthattheeffective gradeelevation isatabout240-243feet.Thiselevation corresponds tothefoundation oftheturbinebuildingandtheelevation inthereactorbuildingwherethestructure changesfromanessentially monolithic concreteblockstructure (including thereactorbasemat)tothatofreinforced concretewallsandfloors.Essentially noamplification isexpectedinthereactorbuildinguptoabout243feet.Onthisbasis,theelevations whichareconsidered tobewithin40feetofeffective grade,arethoseelevations inthereactorandturbinebuildings uptoandincluding the281footelevation.

SSRAPwasingeneralagreement withthisapproach.

PriortowalkdownoftheplantbySSRAPandNRCreviewers, thethreeSRTsdescribed theirprogresstodate,highlighting areastheyparticularly wantedthereviewers toevaluate.

SSRAPandNRCrepresentatives spentmostofFebruary2ndperforming independent walkdowns ofNMP-l.Essentially allsafeshutdownequipment wasseenbythemwiththeexception oftheemergency condensers andrelatedequipment, severalreactorcoolantsysteminstruments, severalreactorcoolantsystemisolation valves,coresprayandcontainment spraypumpsinthebasementcornerroomsandtheequipment inthedrywell,allofwhichwereinaccessible duetotheneedforradiation workpermits(RWPs).Following thiswalkdown, Dr.KennedyprovidedasummaryofSSRAP'sobservations andconclusions:

TheSSRAPwalkdownwasperformed todetermine howtheseismicreviewteams(SRTs)wereoperating, toassesshowtheSRTswereevaluating anddispositioning thesafeshutdownequipment, andtoobtainageneralsenseoftheseismicruggedness ofNMP-1.SSRAPdidnotobservemanyseismicconcernsandnoseriousseismicissues.Theexpectedoutliersidentified bytheSRTswereconsidered bySSRAPtobetypical.Dr.Kennedyremarkedthat,infact,therewerefeweroutliersthanwouldbeexpectedforaplantofthisvintage.Hebelievesthatthisisresultofthenumerousseismicupgradesperformed byNMPCovertheyearswhichwereapparenttoSSRAPduringtheirwalkdown.

ItisSSRAP'sjudgment, basedontheirwalkdown, thattheSRTmembersreceivedadequatetrainingtoperformthewalkdownandthattheyweredoinganadequateandqualified jobofevaluating theseismicadequacyofthesafeshutdownequipment.

SSRAPgenerally expressed theopinionthatwhentheSRTsreacheddifferent conclusions thanSSRAP,theSRTs'onclusions weremoreconservative (i.e.,theSRTsmayhaveidentified moreoutliersthanwouldSSRAP).SSRAPisuncertain iftheutilitySRTsusedduringthetrialplantwalkdownarerepresentative oftheSRTsotherutilities mightusefortheirwalkdowns, sinceSSRAPbelievesthattheutilitySRTmembersatthetrialplantwalkdownhaveconsiderable seismicexperience.

Asaresult,SSRAPcontinues tobelievethatitisessential thattheSRTmembershaveadequatequalifications andexperience inseismicengineering.

Following Dr.Kennedy's summaryreport,NRCrepresentatives presented theirobservations andconclusions.

Dr.T.Y.Chang,USIA-46,Program Manager,reportedthefollowing:

TheNRCgenerally agreeswiththeSSRAPreviewfindings.

TheNRCbelievesthatthewalkdownhasshownthattheuseofutilityengineers isaviableapproachprovidedtheSRTmembershavetheproperlevelofexperience.

TheNRCstillstronglybelievesthatthequalifications oftheSRTmembersareveryimportant, irrespective ofwhetherthemembersareutilityemployees orcontractors.

Further,theNRCbelievesthatthetrainingprogramisnotenoughtomakeanengineeraseismicexpert.TheSRTmembersshouldhavetherequisite seismicexperience priortotheirselection fortrainingandthewalkdowns

.TheconductoftheNMP-1walkdownwasverysmooth.TheNRCcommented thatitisclearthatthelessonslearnedfromtheTrialPlant1walkdownwerefactoredintothiswalkdowninthattherewasaconsiderable amountofpre-walkdownplanningwhichcontributed tothesmoothness ofthewalkdown.

TheNRCwasimpressed withthelayoutofNMP-1.Theplantisopenandhasconsiderable spacewhichcontributes tobothgoodmaintenance andalackofseismicinteraction hazards.TheNRCobservedduringtheirwalkdown(asdidtheSRTsandSSRAP)thatthequalityoftheanchorweldsinsomeelectrical cabinetswasmarginal.

TheNRCnotedthattherelayreviewforNMP-1wasperformed forasampleoftypicalsafeshutdowncircuitsanddidnotcovereverysafeshutdowncircuitandrelayinthisplant.Theynotedthattheremaining circuitsandrelaysneedtobereviewedbeforetheseismicreviewforNMP-1iscomplete.

Therewassomediscussion oftheuniformhazardground-motion spectrausedforthiswalkdown.

Sincethisspectraboundsthelicensing basisground-motion SSEspectraforNMP-1,theNRCconcluded thatthisground-motion spectraisacceptable andmeetstherequirements ofUSIA-46.TheNRCalsonotedthattheyconcurthattheamplified floorspectrausedforthiswalkdownareconservative spectral (Figure3containsanexcerptfromtheSQUGWalkdownTrainingCourseclassnoteswhichshowsthescreening processforcomparing equipment capacitytodemand.)Equipment Capacityvs.DemandScreening ProcessReference Spectrum>IRSScreen1Screen2Below40'AboveSHz88oundingSpectrum>GRSGERS>IRSScreen3Screen4QualiTication Documentation

>IRSOu5iersCapacity>DerrtandResolve-Slide26 ENCLOSURE 2PositionPaerontheUseofMethodAatGinna~PuroseThepurposeofthispositionpaperistoprovidesupporting information forapplication ofMethodAatGinnaasrequested bytheNRCinquestion1ofasecondRAIontheUSIA-46program.Thisenclosure describes manyoftheconservatisms thatexistincomputedin-structure responsespectraandthesafetysignificance ofthedifference betweencomputedandactualbuildingresponse.

1.Conservatism inCalculated ISRSTheprocessofcalculating in-structure responsespectra(ISRS)isacomplicated analytical exerciserequiring asignificant numberofapproximations, modelingassumptions andengineering judgments.

Asaresult,thehistorical development oftheseISRShasincludedatremendous amountofconservatism whichhastypically servedtwopurposes:

Ithasreducedthetechnical debateastothecorrectmodelingofthemanyparameters whichareintrinsic totheISRScalculational methodology, and;2.Ithasreducedthecostsassociated withaverydetailedstate-of-the-art

analysis, (whichwouldattempttotrimoutalltheunnecessary conservatisms)

.AsapartoftheA-46programresolution methodology, theSSRAPdeveloped andSQUGsubsequently endorsedanalternate ISRSestimation technique (referred toasMethodAwithintheGIP)whichwasmuchmoremediancenteredandrealistic thanthetypicaldesignpractice.

RG&E's'position isthattheapplication ofMethodAatGinnawasappropriate andtechnically justified.

ThefactthatdesignISRSmayshowamplifications greaterthan1.5isnotsurprising, nordoesitnegatethevalidityofMethodA.Infact,asnotedintheSSRAPreportitwasevenexpected.

"Secondly, mostunbroadened computedin-structure spectrahaveverynarrow,highlyamplified peaksattheresonantfrequency ofthestructure.

Inmostcasesthesenarrow,highlyamplified peaksareartificially broadened toaccountforuncertainty inthestructure's naturalfrequency.

Thisprocesssimplyincreases theemphasisonthesehighlyamplified peaks.SSRAPisalsooftheopinionthatthesenarrowpeakswillnotbeashighlyamplified inrealstructures athighgroundmotionlevelsasifpredicted bylinearelasticmathematical models,noraresuchnarrowpeakedin-structure spectralikelytobeasdamagingtoequipment asisabroadfrequency inputwhichisrepresented by1.5timestheBoundingSpectrum."

Asdescribed below,threeareasarepresented tosupporttheapplication ofMethodAatU.S.nuclearplantsingeneral,andatGinnainspecific:

A.Measurements ofISRSinActualEarthquakes B.Calculations ofOverallConservatisms inTypicalISRSC.Description oftheConservatisms inISRSinGeneralandGinnaISRSinParticular A.Measurements ofISRSinActualEarthquakes SSRAPdeveloped theMethodAresponseestimation technique basedontheirresearchofbothactualearthquake measurements andonrecent"mediancentered" analysis.

Theyreference (SSRAPreportpage102)themeasuredfloorresponsespectraatelevations lessthan40feetabovethegradeformoderately stiffstructures atthePleasantValleyPumpStation,theHumboltBayNuclearPowerPlant,andtheFukushima NuclearPowerPlantwhereamplifications overthegroundresponsespectradonotexceed1.5forfrequencies aboveabout6Hz.Other,morerecentearthquake datafromtheManzanillo PowerPlantandSicartsaSteelMillinMexico,aswellasseveralfacilities inCalifornia andJapan,hasbeenrecentlyreviewedbySQUG.Thisdataalsoshowsthatstiffbuildings (similartotypicalnuclearstructures) amplifyverylittleatelevations lessthan40feetabovegradeandfrequencies over8Hz.SQUGknowsofnonewmeasureddatathatchallenges GIPMethodA.B.Calculations ofOverallConservatism inTypicalISRSCalculated ISRShaveneverbeenportrayed asrepresenting therealistic expectedresponseduringanactualearthquake.

Aspreviously stated,ISRStypically containmanyconservatisms whichmakethemunrealistically high.Theprimaryreasonforthedevelopment ofMethodAwastoestablish amoremediancenteredmethodofdefiningthestructural responsewithouthavingtoembarkoncostlynewanalysesofallthesitebuildings (Itshouldbenotedthateventhemostmodern,state-of-the-art ISRScontainsignificant conservatisms; eventhoseclassified as"median-centered",

areoftenveryconservative).

ANRCcontractor (LLNL)concluded inastudyfortheNRC(NUREG/CR-1489)thattypicalcalculated ISRScontainfactorsof1.Sto1.8.RecentsurveysbySQUGshowsimilarlevelsofconservatism incalculated ISRS.

Itwasthecontention ofSSRAPthattheISRSfornuclearstructures (considering the40'nd8Hzconditions) wouldbewithinabout1.5timesthegroundresponsespectrum(GRS)iftheplantweresubjected toanactualearthquake.

InderivingtheMethodAcriteriatheyrecognized thatduetothevarietyofgroundmotions,soilcharacteristics andstructure characteristics therecouldbesomepossibility ofexceedances tothe1.5amplification, butstillstronglyjustified MethodA'applicability:

"ItisSSRAP'sfirmopinionthattheissueofpotential amplifications greaterthan1.5aboveabout8Hzforhighfrequency input"isofnoconsequence fortheclassesofequipment considered inthisdocumentexceptpossiblyforrelaychatter'."

[SSRAPReport,Page106]ThebasisSSRAPgavefordrawingthisconclusion wasthathighfrequency groundmotionsdonothavemuchdamagepotential dueto~lowspectraldisplacement, lowenergycontent,andshortduration.

Theyfurthernotedthattheequipment covereddoesnotappeartohaveasignificant sensitivity tohighfrequencies (exceptpossiblyforrelaychatter,whichisaddressed separately intheGIP).C.Description ofConservatisms inISRSinGeneralandGinnaZSRSinParticular Themostsignificant sourcesofconservatism involvedinthedevelopment oftheISRSforGinnaincludethefollowing:

000000000LocationofInputMotion(variation fromthefreefieldinputlocation)

GroundResponseSpectrumShapeSoil-Structure Interaction (SoilDamping,WaveScattering Effects)GroundMotionIncoherence Frequency (Structure Modeling)

Structural DampingTimeHistorySimulation Non-Linear Behavior(e.g.,soilpropertyprofilevariation, concretecracking)

PeakBroadening andEnveloping ClippingofNarrowPeaks'BecauseoftheSSRAPconcernrelatedtopossiblyrelaychatteratfrequencies above8Hz,theSQUGmethodology specifically addresses relaywhicharesensitive tohighfrequency vibration.

SuchrelaysareincludedontheLowRuggedness RelayslistinAppendixEofEPRIReport,NP-7148.

~~Thedegreeofconservatism involvedineachoftheseparameters isspecifictothebuildingbeinganalyzed, tothefloorlevelbeingconsidered, andoften,totheequipment locationwithinthespecified floorlevel.Theseconservatisms typically cannotbeaccurately quantified usingsimplistic calculational techniques sinceeachparameter contributes toanoverallsetofhighlynon-linearresponses.

Thus,itwouldtakeaconsiderable efforttoquantifytheexactexcessconservatisms inherentinthecalculated ISRSatGinna.However,onthequalitative levelpresented below,itiseasytoseetheoriginsandlevelsofthisconservatism.

Thefollowing parameters arethesourceofthemajorportionsoftheexcessconservatism:

LocationofInutMotion-ThedefinedlocationoftheplantSSEistypically partofthedesignbasisdocumentation.

TheSSEshouldtypically bedefinedatthegroundsurfaceinthefreefieldasdefinedinthecurrentStandardReviewPlancriteria.

ThedefinedlocationoftheGinnaSSEisconsidered thegroundsurfaceinthefreefield.Butforpurposedofgenerating ISRS,someplantsconservatively definedtheinput(currently identified asthe"controlpoint"location) atanotherlocation, suchastheembeddeddepthofabuildingbasemat.Thisconservatism canbesignificant depending onthespecificplant/building configuration.

TheGinnaplantsitegeologyconsistsofathinlayerofnaturalorcompacted granularsoil(30to40feetindepth)immediately abovebedrock.Thebedrockisamixtureofsandstone andfissileshalewithshearwavevelocities calculated tobe7000feetpersecondorgreater.Priortoconstruction oftheplant,thesoiloverburden(30to40feetofglacialdrift)wasremoved.AllGinnaStationCategory1buildings, exceptforthecontrolbuildinganddieselgenerator

building, arefoundedonsolidbedrock.Thefoundations ofthecontrolanddieselgenerator buildings wereexcavated tothesurfaceofbedrock.Leanconcreteorcompacted backfillwasplacedontherocksurfacetoadepthwherebytheelevation ofthetopofthefillmaterialwascoincident withtheelevation ofthebottomoftheconcretefoundation ofthatparticular building.

Sections2.5.2.1(Seismicity) and2.5.2.2(MaximumEarthquake Potential) oftheGinnaFSARdescribetheoriginalinvestigation whichwasperformed todevelopestimates ofthemaximumexpected(OBE)andmaximumcredible(SSE)earthquakes forthesite.Itwasjudgedthatthemaximumcredibleearthquake wouldbeoneofRichtermagnitude

6.0 withanepicenter

30milesfromthesiteoroneofmagnitude 7.0ata90-mileepicentral distance.

Aprocedure developed byDames&Moore,usingtheresultsofresearchattheEarthquake Institute ofTokyo,wasusedtoestimategroundmotionatagivenlocationiftheearthquake magnitude, epicentral

distance, andelasticproperties offoundation soilsandrockareknown.TheFSARcontainsthefollowing description ofthelocationofgroundmotion:

"Usingthismethodandtheassumedmaximumcredibleearthquakes discussed above,maximumacceleration onthesitewascalculated tobe8'.ofgravityforsoilsurfaceand7%forbedrocksurface.Plantstructures, systems,andcomponents designated asSeismicCategory1aredesignedtoremainwithinapplicable stresslimitsfortheoperating-basis earthquake (0.08g)andthesafeshutdownearthquake (0.20g)."

Basedontheabovelicensing basisdescriptions, thedesignearthquakes (OBEESSE)wereclearlydefinedatthesoilsurface.SincetheISRSforGinnaweregenerated usingaconservative modeldefiningtheinputmotionatthefoundation level,significant conservatism existsduetothelocationofinputmotion.Thelevelofconservatism involvedinthisassumption isdifficult toestimatewithoutperforming additional

analyses, butpaststudieshaveproveditcanbeconsiderable.

GroundResonseSectrumShae-TheSSEdefinedwithintheplant-licensing basisistheappropriate reviewlevelfortheA-46program.Someutilities utilizedalternative (conservative) spectralshapesfortheearthquake levelsutilizedfortheirA-46resolution (i.e.,submitted aspartoftheir120-dayresponseletters).

Theamountofconservatism isdirectlyrelatedtothedifference betweenthesetwospectralshapesatthefrequencies ofinterestforthestructures beingreviewed.

Thisfactorcanrangefrom1.0toaround2.0depending onthedifferences betweenthespectra.Thelicensing basissafeshutdownearthquake forGinnaischaracterized byasite-specific horizontal groundresponsespectrumanchoredtoaPGAof0.17g.However,ISRSwerenevergenerated intheoriginalseismicdesignofGinnaandthisearthquake wasnotusedfortheUSIA-46program.Amoreconservative earthquake anchoredtoaPGAof0.2gandwithaRegGuide1.60shape(broaderband)wasusedforthegeneration ofISRSintheA-46program.Theuseofthisalternate earthquake inputisconservative for3reasons:The4:dampedspectrawereusedinsteadoftheS:dampedspecificfortheA-46program.Theconservatism istypically quantified bytakingthesquarerootofthedampinglevels,whichwouldresultina1.12(12%)factorofconservatism.

2)TheZPAlevelof02.gis18:higherthanthe0.17gsitespe'cific SSElevelforGinna.3)TheRegGuideShapeandthesitespecificshapearebothbroadbanded,buttheirlevelsofamplification aredifferent andtheirdifferences varyasafunctionoffrequency.

Depending onthebuildinginquestionandthefrequency rangeofinterest, therecanbeadditional conservatisms duetothedifferences inshape.

SoilStructure Interaction SSI-Typicaldesignanalysesdonotaccountproperlyforthephenomena ofSSI,including thedeamplification withdepththatreallyoccursforembeddedstructures andfortheradiation dampingeffectsinherentatsoilsites.Fixed-base analyseshavebeenperformed intypicaldesignanalyses, bothforstructures foundedonrockandforstructures foundedonsoilcolumns.Forrockfoundations, thefixed-base modelhasbeenshowntobeslightlyconservative depending ontherock/structure characteristics.

Forsoilfoundedstructures thisassumption canvarybetweenconservative andveryconservative, depending onthelevelofsophistication ofthemodelingofthesoil-structure system.Thesimplified analysesthatusedthefrequency-independent soilspringsweretypically veryconservative inthatradiation and/ormaterialdampingwereeitherconservatively eliminated orartificially limitedduringtheanalysis.

Soilproperties werealsotypically notadjustedtoreflectanticipated soilstrainlevels.Significant reductions havebeendemonstrated overdesigntypeanalysesusingmoremoderntechniques.

Thesereduction factorsarehighlydependent onthespecificsoilconditions andstructure configurations, butvaluesofaround2to4havebeenseeninpaststudies.TheGinnaanalyseshaveignoredanyreduction infoundation motionduetoembedment effects,wavescattering effectsandradiation ofenergyfromthestructure intothesurrounding media.Theseeffectsarelessforrockfoundedstructures (StandbyAuxiliary Feedwater Buildings andIntermediate Building) thantheyareforthesoillayerfoundedstructures (ControlBuildingandDieselBuilding),

buttheyarenotnegligible.

Thisassumption iscommonlymadeforrocksitesbecauseitgreatlysimplifies theanalysiseventhoughitintroduces conservatism.

TheGinnaanalysesalsoignoredanyconstraint thatsurrounding rockorsoilplacedagainstexteriorsidewallsofembeddedstructures.

Withoutconsidering lateralsupportfromtherockorsoilagainstembeddedstructures, onecomputesstructural responses atgradethataregreaterthanthefreefieldmotion.However,thestructure atgradecouldnotrespondsignificantly greaterthanthefree-field motioniftheembeddedportionofthestructure islaterally supported bythestiffsoilorrock.Aswasthecasefortheveryfirstconservatism described (location ofinputmotion),itwouldrequiresomereanalysis toestimatethedegreeofconservatism involvedintheSSImodelingofGinnastructures.

Itisobvious,however,thatsomenon-trivial degreeofconservatism exits.GroundMotionIncoherence

-Ashasbeendocumented intheEPRIseismicmarginreport(EPRINP6041)therecanbeadeamplification effectonnucleartypestructures duetotheincoherence ofgroundmotionovertherelatively largedimensions oftypicalnuclearstructures.

Conservative reduction factorsasafunctionoffrequency andbuildingfootprint havebeendocumented withinNP6041toaccountforthestatistical incoherence oftheinputwavemotion.These conservative valuesrangefromafactorof1.1toaround1.5.Morerecentstudieshavedocumented evengreaterreduction factors.Thisgroundmotionincoherence isapplicable torocksiteslikeGinna.TimeHistoSimulation

-ISRSatGinnahavebeengenerated usingatimehistorywhichisintendedtoapproximate thedesiredearthquake spectrum(0.20g,Reg.Guide1.60shape).Thisprocessinvolvesthegeneration ofanartificial timehistorywhoseresponsespectraenvelopstheSSE.Theamountofconservatism involvedintheenveloping processhasnotbeenspecifically calculated forGinnabutcanrangeuptoafactorof2ormoreunlesssignificant resources areappliedtominimizethedegreeofenveloping.

CliinofNarro~Peaks-TheSSRAPreportandtheGenericImplementation Procedure (GIP)recommend procedures foradjusting narrowpeakstoreflecttwoareasofconservatism:

Narrowpeaksarenotashighlyamplified inrealstructures asarepredicted bylinearelasticmathematical models,and2.NarrowpeaksinISRSarenotasdamagingtoequipment asarebroadfrequency inputsuchastheReference Spectrum.

TheGIPprocedure recommends anaveraging technique overafrequency rangeof10%ofthepeakfrequency (e.g.,1Hzrangefora10Hzpeakfrequency) usingtheunbroadened ISRS.TheGinnaISRShavenarrowpeaksanddidnotutilizethepeakreduction methodsfromtheGIP.Theconservatism involvedhasbeenshowntobeintherangeof5%to20%fortypicalnarrowpeaksatseveralplants.Weexpecttheconservatism forthepeaksoftheGinnaISRStofallwithinthisrangebasedonasamplingforacoupleofpeaksshowinga10%effect.Thereareseveraladditional sourcesofconservatism (e.g.,structural damping,structural

modeling, structural/soil non-'inearities, peakbroadening andenveloping, etc.)whichaddtotheoverallconservatism inthecalculation ofISRS.Theseadditional conservatisms, coupledwiththosedescribed above,certainly reinforce theoveralllevelsofconservatism inISRSofbetween1.5and8whichwerereferenced bySSRAP(LLNLReportNUREG/CR-1489),

andexplainwhytheconservative GinnaISRSproduceexceedance beyondthe1.5factor.2.NotaSiificantSafetIssueTheexpecteddifferences betweencalculated ISRSandactualbuildingresponsedonotrepresent asignificant safetyquestion.

Thelessonslearnedfromreviewofhundredsofitemsofequipment atvarioussitesthathaveexperienced earthquakes whichweresignificantly largerthanthoseforEasternU.S.nuclearplantsarethatmissinganchorage, seismicinteraction hazards,andcertainequipment-specific weaknesses (incorporated intotheGIPcaveats)weretheseismicvulnerabilities whichcauseequipment damage.Theseareasareconservatively addressed intheGIP.

I~TheNRCstaffacknowledged theseismicruggedness ofnuclearpowerplantequipment inthebackfitanalysisforUSIA-46inwhichtheystatedthefollowing:

"...subject tocertainexceptions andcaveats,thestaffhasconcluded thatequipment installed innuclearpowerplantsisinherently ruggedandnotsusceptible toseismicdamage."[NUREG-1211, page16]MethodAisonlyapplicable tostiffequipment withfundamental frequencies overabout8Hz.AsnotedearlierinSection1ofthispaper,SSRAPandSQUGhaveagreedthatexcitations over8Hzhavelittledamagepotential duetolowspectraldisplacements, lowenergycontentandshortduration.

Thisjudgmentissupported byindustryandNRCguidancefordetermining whetheranoperating basisearthquake (OBE)isexceededfollowing aseismiceventatanuclearpowerplant.EPRIReportNP-5930andNRCRegulatory Guide1.166recognize thatdamagepotential issignificantly reducedforearthquake groundmotionsabove10Hz.Inotherwords,thequestionofwhatistheprecisevalueofbuildingamplification over8Hzhasverylittlesafetysignificance.

3.GinnaBuildinsareicalNuclearStructures Asrequested, RGEEisalsoproviding detaileddescription ofthepowerblock.buildingconstruction.

TheGinnapowerblockstructures aretypicalnuclearpowerplantstructures whichweredesignedtoresistlateralloadswithreinforced concreteshearwallsorbracedstructural steelframesystems.Asummarydescription ofthebuildings andtheirfoundations arecontained intheattachedTable1.4.Determination of"GradeElevation" "GradeElevation" determinations forGinnaStationpowerblockbuildingweredescribed inSection2.3oftheJanuary1997submittal:

"GradeElevation Thepowerblockstructures atGinnaarebuiltonthesideofahi'll.Gradeelevation onthenorth(lakeside)ofthepowerblockis253'.Gradeonthesouthsideofthepowerblockis271'.FortheA-46project,agradeelevation of253'asusedforthestructures onthenorthsideofthepowerblock(DG,IB,SHTB),andagradeelevation of271'asusedforthestructures onthesouthsideofthepowerblock(AB,AF,CB).Thecontainment (RC)isfoundedonrockatelevation 235';235'asusedasthegradeelevation forA-46.ItshouldbenotedthatCB289'sthehighestelevation atwhichseismicSSELequipment arelocated,andthatthegreatmajorityofseismicSSELequipment areatelevation 271'rlower.Therefore, forequipment outsidecontainment, whether253'r271'susedasgradewouldnotimpactthe"withinabout40'fgrade"criterion commonlyusedintheGIP."

ItshouldbenotedthatCB289'sthehighestelevation atwhichseismicSSELequipment arelocated,andthatthegreatmajorityofseismicSSELequipment areatelevation 271'rlower.Therefore, forequipment outsidecontainment, whether253'r271'susedasgradewouldnotimpactthe"withinabout40'fgrade"criterion commonlyusedintheGIP."Inadditiontopreviousdiscussions inEnclosure 2describing thepowerblockstructures andcorresponding "gradeelevations",

ageneralNorth-South sitecrosssectionisprovided.

Conclusions Thediscussion aboveleadstoseveralconclusions:

ClAlloftheGinnastructures arelargereinforced concreteshearwallorbracedsteelframestructures.

Theyaretypicalofthestructures designedfornuclearplantsof.theGinnavintageandare"typicalnuclearstructures".

CITheresultsfromactualmeasuredISRSon"nucleartype"structures supportthe1.5responselevelsadvocated withinMethodA.0Qualitative assessments oftheconservatism inherentwithinthemethodsutilizedtocalculate ISRShavebeenprovidedabove.Theseconservatisms aretypically quitesignificant (ashasbeenindependently verifiedbymedian/modern assessment suchastheLLNLstudy)andcan/willresultinISRSwhichshowamplifications wellbeyondthe1.5factorfromMethodA.RGEEfeelsstronglythatthespecificexceedances notedbytheNRC(beyondthe1.5factor)onGinnaareduetothesehighconservatisms inherentintheISRSmethodsandnotdueto"unusual, plant-specific situations".

Therefore, theapplication=of Method"Atothestructures atGinnaisappropriate andvalid.CIThereislittlesafetysignificance intheexpecteddifferences betweencalculated ISRSandactualbuildingresponse.

Thelargestsafetyimprovements areprovidedbyappropriately reviewing equipment anchorage, seismicinteraction hazards,andcertainequipment-specific weaknesses whereseismicvulnerabilities havecausedequipment damageinrealearthquakes.

ReviewsoftheseareaswereaprimaryfocusoftheSQUGGIPprocess;therefore RG&E'simplementation oftheGIPatGinnaresultedinsignificant seismicsafetyenhancements.

Table1BuildingDGZBDetailedDescription ofBuildingConstruction Thedieseleneratorbuildin(DG)isaone-story reinforced-concrete(Rc)structure thathastwocablevaultsunderneath thefloor.ThebuildingroofconsistsofaRCslabsupported byfourshearwallsthatsitonconcretespreadfootings.

Ztisarelatively stiffstructure typicalofmostdieselbuildings atnuclearplants.Theintermediate buildin(ZB)islocatedonthenorthandwestsidesofthecontainment

building, andisfoundedonrock.Thewestendhasaretaining wallwheretheflooratelevation 253ft6in.issupported.

Thebottomoftheretaining wallfootingisatelevation 233ft6in.Rockelevation inthisareaisatapproximate elevation 239ft0in.Foundations forinteriorcolumnsareonindividual columnfootingsandembeddedaminimumof2ftinsolidrock.SHTBCBThescreenhouse-service water(SH)buildingiscomprised oftwosuperstructures, onefortheservicewater(SW)systemandoneforthecirculating watersystem(thescreenhouseportion).

Theservicewater(SW)portionofthebuilding(bothbelowandabovegrade)isaSeismicCategoryIstructure.

Theservicewater(SW)portionhousesfourSeismicCategoryIservicewater(SW)pumpsandSeismicCategoryZelectricswitchgear.

Thescreenhouse portionhousesthetraveling waterscreensandcirculating waterpumps.Theentirescreenhouse-service water(SH)buildingisfoundedinoronbedrockwiththeexception ofthebasementoftheelectricswitchgear portionwhichisfoundedapproximately 4ftabovebedrock.Sincethebuildingisfoundedinbedrockthebasementwillnotrealizeanyspectralacceleration andtheseismicloadingisequivalent tothegroundmotionof0.08gand0.20g.Thebuildingisconstructed ofRCbelowgradeandhasastructural steelsuperstructure.

Theturbinebuildin(TB)isa257.5-ftby124.5-ftrectangular buildingonthenorthsideofthepowerblock.Zthasaconcretebasementatelevation 253.5ft,twoconcretefloors(amezzanine flooratelevation 271ftandanoperating flooratelevation 289.5ft).Thebuildingisaheavilybracedsteelstructure.

Theauxiliabuildin(M))isathree-story rectangular structure, 70ft9in.by214ft5in.Itislocatedsouthofthecontainment andintermediate buildings andadjacenttotheservicebuilding.

Thestructure hasaconcretebasementfloorthatrestsonasandstone foundation atelevation 235ft8in.,andtwoconcretefloors--an intermediate flooratelevation 253ftandanoperating flooratelevation 271ft.Construction belowgradeis(RC)withastructural steelsuerstructure.

Thestandbauxiliafeedwater buildin(AF)isareinforced-concreteseismiccategoryIstructure withreinforced-concrete walls,roof,andbasemat.Thebuildingissupported by12caissonswhicharesocketedintocompetent rock.Thecontrolbuildin(CB)islocatedad)acenttothesouthsideoftheturbinebuildingandisa41-ft11-3/4in.by54-ft1-3/4inthree-story structure withconcretefoundation matatelevation 253ft.Thefoundation ofthecontrolbuildingissupported onleanconcreteorcompacted backfill.

Therockelevation inthisareaisatapproximate elevation 240ft.0in.Thefoundation ofthecontrolbuildingwasexcavated tothesurfaceofthebedrock.Thefillmaterialwasplacedontherocksurfacetoadepthcoincident withthecontrolbuildingfoundation.

Thebottomelevation ofthedeepestportionofthefoundation matisatelevation 245ft4in.,withastructural slabsupported atelevation 250ft6in.withathickened slabforcolumnfootings.

ThebuildingconsistsofbothRCandstructural steel.RCThecontainment buildin(RC)isaverticalrightcylinderwithaflatbaseandahemispherical dome.Thebuildingis99ft.hightothespringlineofthedomeandhasaninsidediameterof105ft.Thecylindrical concretewall,whichisprestressed vertically andreinforced circumferentially withmildsteeldeformedbars,is3.5-ft.thick.Theconcretedomeisareinforced concreteshell2.5-ft.thick.Theconcretebaseslabis2ftthickwithanadditional thickness ofconcretefillof2ftoverthebottomlinerplate.Thecontainment cylinderisfoundedonrock(sandstone) bymeansofpost-tensioned rockanchorswhichensurethattherockthenactsasanintegralpartofthecontainment structure.

• Buildingdescriptions arefromFSARandUFSAR.

4ScreenhouseDieselgenerator annexTurbinebldg.Intermediate bldg.ServicebuildingReactorcontainment buildingI1I~FacadeControluildin11~SlAuxiliary buildingAux.bldg,additionROCHESTER GASANDELECTRICCORPORATION R.E.GINNANUCLEARPOWERPLANTUPDATEDFINALSAFETYANALYSISREPORTFigure3.7-6Containment BuildingandComplexofInterconnected SeismicCategoryIandNonseismic Structures, FlanView LOWWATERDATUNIEL.243.0'IGH WATEROATUNIEL.247.0'REAKWALL EL.261'ISCHARGE CANALEL.231.5'RADE EL.253'LANTGRADEEL.270'RADEEL.270'NVERT OFDEERCREEKEL.250'CREENHOUSE GUARDHOUSE AlD(II0AItQOICCfl0IIIconnIII0I0CAe00CQIllIllOAZZZrZcnc0Z.g~mrIlltlCI8UnO0Z ENCLOSURE 3SECONDREUESTFORADDITIONAL INFORMATION R.E.GINNANUCLEARPOWERPLANTuestion¹2Inyourresponsetothestaff'sRAIQuestions

¹3and¹4,foranumberofequipment items,theequipment frequencies werestatedtohavebeen-judgedbySRTtobegreaterthan8Hzbyinspection.

ProvidethebasisfortheSRTjudgement regarding equipment naturaIfrequency, especially whentheestimated magnitude fornaturalfrequency isreliedupontodetermine theapplicability fortheuseofGIP-2MethodA.l.Youarerequested toprovidefurtherjustification forthefrequency estimation, orprovideanalytical calculations tojustifysuchestimation, fortheequipment itemsidentified asFT-4084,FT-4085,PSF01AaB,SAFVPCIP,

SAFVPDIP, BVSZ4,DCPDPAB01A&B andDCPDPAB02AaB.

Resonsestouestion¹2ThebasesfortheSRTjudgement was:IIa.TheguidanceprovidedintheEPRIreport,"Guidelines forEstimation orVerification ofEquipment NaturalFrequency",

ResearchProject2925-2,FinalReport,August1992.b.Theexperience oftheSRT.TheSRTwascomposedofRGEEstaffandstafffromanoutsideconsultant, Stevenson 6Associates,(S&A).

TheSaAstaffmembersonthewalkdownswereDr.JohnStevenson, Mr.WalterDjordjevic, andMr.StephenAnagnostis.

Dr.Stevenson has35yearsofexperience inthenuclearpowerindustry; Mr.Djordjevic andMr.Anagnostis eachhave15to20yearsofexperience intheindustry.

Inaddition, Mr.Djordjevic andMr.Anagnostis eachhaveextensive experience performing in-situmodal(frequency response) testsofnuclearpowerplantequipment.

Thistestinginvolvesdozenofpiecesofequipment atmorethantennuclearpowerstations.

Theequipment testedincludescontrolcabinets, motorcontrolcenters,switchgear andinstrument racks.ThistestdataisthebasisofEPRIReportNP-7146,"Guidelines forDevelopment ofIn-Cabinet Amplified ResponseSpectraforElectrical Benchboards andPanels",andtheprocedures forcalculating in-cabinet spectradescribed inGIP-2Section6.4.2(Screening Level3).FT-4084andFT-4085StandbAFWPumFlowTransmitters Theseflowtransmitters areindividually mountedtoasteelbaseplateandanchoredtoareinforced concretewallwithfour(4)3/8"concreteexpansion anchors,asshowninpicturebelow.TheSRTjudgedthisequipment tohaveafundamental frequency greaterthan8Hzbasedonthesmallsize,smallweight,andstiffsupport.4j~~.\4Wl'rQ".rjfo1 TheEPRIguidelines donotdiscussindividually mountedpressureswitches, butdodiscusssteelframeinstrument rackshousinganumberofpressureswitchesandrelatedequipment.

Section3.3oftheguidelines statethat"Often,bracedrackswillhavefrequencies greaterthan8Hz.Thewalkdownteamneedonlybecautiousofverylarge,heavilyweighted, veryweakly-braced racks,orverylowbracedracks".Basedonthisguidance, itisreasonable toconcludethatanindividual transmitter, securelymountedtoareinforced concretewall,willhaveafundamental frequency greaterthan8Hz.PSF01APSF01BStandbAFWPumsCandDTheseare300HPelectric-motor horizontal pumps.Eachpumpismountedonanapproximately 3'-6"wideby10'-6"longsteelskid.Theskidisanchoredtoa27"highcontinuous concretepedestalwithtwelve(12)3/4"cast-in-place bolts.Thepedestaliswellreinforced anddoweledintothefloorslab.Therearenovibration isolators.

Section2.2oftheEPRIguidelines states:"Further, thefollowing classesofmechanical equipment areconsidered tobesufficiently ruggedthatthewalkdownteammayassume,withoutfurtherjustification, thattheirnaturalfrequencies areabove8Hz:PumpsEngineandMotorGenerators AirCompressors FansandAirHandlersChil,lersTestingonshaketablesandinthefieldhasshownthatequipment intheseclasseshavenaturalfrequencies greaterthan8Hzgiventhattheyhavedirectanchorage tothefloorandthatappendages suchasveryflexiblecontrolpanelsarenotpresent.Onepossibleexception isthatdeepwellpumpsmayhaveunsupported cantilever columnsandsuctionbowlswithnaturalfrequency below8Hz(Notethatthisaddressed byaseparateSQUGcaveat).SAFWPCIPSAFWPDIPStandbAFWPumInstrument PanelsThesearewallmountedinstrument panels.Eachpanelis30"widex54"highx32"deepandisweldedtocontinuous anglesrunningalongthetopandbottomofthepanel.Theanglesaresecuredtoareinforced concretewallwith5/8"concreteexpansion anchors(Bothtopanglesandoneofthebottomanglesaresecuredwiththreeanchors;theotherbottomangleissecuredwithtwoanchors.)

TheEPRIguidelines indicatethatthepresenceofwell-engineered topbracingonelectrical equipment issufficient tosupportthejudgement thattheequipment hasafundamental frequency above8Hz.TheSRTexperience isthattypical(30"widex30"deepx90"tall),properlyanchored, floormountedinstrument cabinetshaveafundamental frequency intherangeof10Hz-15Hz.Smaller,wallmountedcabinetsofsimilarconstruction (suchasthesubjectpanel)haveatleastashighafundamental frequency.

BUS14480VSwitchearThisisaWestinghouse TypeDBlowvoltageswitchgear, 160"widex58"deepx76"high.Theswitchgear isanchoredwithten(10)3/4"HiltiKwikbolts.Section3.7oftheEPRIguidelines states:"Aslongasalineuphassixsectionsormore,thenaturalfrequency ofswitchgearunitsmaybeassumedtobeabove8Hz".S&AhastestedasimilarWestinghouse lowvoltageswitchgear attheConnecticut Yankeenuclearstation.Thatunit(designated BUS4atCY)is408"widex54"deepx90"high.Afundamental frequency of9.0Hzwasmeasured.

DCPDPAB01A andBDCPDPAB02A andBAuxBuildinDCPowerDistribution PanelsThesearewallmountedpanelboards (powerdistribution panels).Eachpanelboardis30"widex40"widex12"deep,andwellanchored, nearthetopandbottom,toareinforced concretewallorareinforced concretecolumn.Section2.3oftheEPRIguidelines statesthatpanelboards (wall-mounted distribution panels),ifanchoredtoasubstantial floororwall,willhaveafundamental frequency ofatleast12Hz.uestion3Inyourresponsetothestaff'sRAI,Question¹4,youindicated thattheseismiccapacityvs.demandevaluation fortheUndervoltage RelayCabinetBus24QRAZRC24) wasbasedonshake-table testing.Youarerequested toprovideadetaileddiscussion ofthetestingandtojustifytheadequacyofsuchtesting.Resonsetouestion¹3

References:

AmericanEnvironments Company,Inc.,ReportNo.STR-142280-1, 11/4/80(RG&EProjectEWR-1444)

AmericanEnvironments Company,Inc.,ReportNo.STR-142280-2, 12/31/80(RG&EProjectEWR-1444)

AmericanEnvironments Company,Inc.,ReportNo.STR-142280-3, 1/5/81(RG&EProjectEWR-1444)

Thereareeight(8)relay/control cabinetsofthismakeontheSSEL.Theirequipment designations andlocations arelistedbelow:ARA1CC14ARA1RC14ARA2CC18ARA2RC18ARB1CC16ARB1RC16ARB2CC17ARB2RC17ControlBuilding271.00AuxBuilding271.00Screenhouse 253.00Screenhouse 253.00ControlBuilding271.00AuxBuilding253.00Screenhouse 253.00Screenhouse 253.00Thefourcabinetsoutsidethescreenhouse areeachstand-alone and24"widex24"deepx70"high.Thefourcabinetsinthescreenhouse consistoftwopairsofattachedcabinets; eachpairis48"widex24"deepx70"high.Each24"x24"x72"cabinetisanchoredtoareinforced concretefloorwithfour(4)HiltiKwik-Bolts.

Thecabinetswereshake-table testedasdocumented inReferences 1through3.Theshake-table testswererandom,multi-frequency, andbiaxial(onehorizontal andtheverticaldirection);

thetestswererepeatedwiththespecimenrotated90degreesabouttheverticalaxis.The"RRSvs.FRS"plotbelowshowsthetests'RS(Required ResponseSpectra)comparedtotheenvelopeoftheFRS(FloorResponseSpectra)foralllocations listedabove.TheRRSenvelopes theFRS,'utnotethattheRRSis3%damped,whiletheFRSis4%damped.However,theRRSiswellabovetheFRSinthepeakrange,andtheactualtestresponsespectraexceededtheRRSbyasubstantial amount,particularly forfrequencies abovethepeakrange.Section5.0ofthetestreportsstate:

"Thetestspecimencontinued tofunctionbefore,duringandafterexposuretotheSeismicQualification TestProgram.Therewasnoevidenceofphysicaldamage,orreportedelectrical malfunction observed.asaresultofthestressesofthistestprogram."

10-RequiredResponseSpectrum, 3o/odampingEnvelopeofFloorResponseSpectra,4'4dampingp'L~lprrIrJrrr~l'0.110100Frequency (Hz)uestion44Forthe48DVACMotorControlCenter(MCC)-,youindicated thattheMCCcanwithstand asinglefrequency testconsisting ofa1.35g,5beat,5cycle/beat input,performed atthesignficant structural frequencies.

Itisknownthatsingle-axis, singlefrequency sinebeattests'ostly performed priortotheissuanceofIEEEStandard344-1975, areconsidered inadequate forequipment seismicqualification duetotheirinability toexcitemulti-axis, multi-frequency responses ofequipment (theveryreasonthatplantsareincludedintheUSIA-46program).

Youarerequested tojustifytheseismicadequacyofthismotorcontrolcenter.Resonsetouestion4

Reference:

LetterfromG.R.Geertman(Gilbert/Commonwealth) toC.J.Mambretti (RG&E)datedAu<fust11,1976.ThesubjectMCC(equipment designation MCCL)isa4sectionWestinghouse MCC,66"widex19"deepx90"high.Itisanchoredwitheight(8)1/2"concreteexpansion anchors.TheMCCislocatedonelevation 271'ftheAuxiliary Building.

10Thefollowing plotcontainsseveralresponsespectratoillustrate thediscussion thatfollows.I~I~---r---r--r IJ~I~~II~JI~IIIIJQtI~QI-AB271A46FloorResponseSpectrum(EW/NS

Envelope, Adamping)-Rersponse spectrumtor5cycrbeatwylaeLabTest{4%damping)---Responsespectrumusedbywestinghouse torAnalysis(assdamping)SIIIS~IIII~IcIII~IIIIIIIr1I1r'II1III,I~IIIIIT1TI,III~~IIIIIII~~IIIII0.11~~I~IIIIJJJIIIIII~IIII1~IIII~I~IIIIIrI~IIIIIII10II1TT~IIITIIII~III~~I~II~II~I~IIIISIIISIIIIIIIII~IIIII~I~Frequency (Hs)Thereference containsasummaryreportfromWestinghouse documenting theMCC'sseismi'cqualification.

TheMCCwasoriginally testedin1972atWyleLabsfollowing therequirements ofTEEE344-1971.

Thetest.usedasinglefrequency, 5cycle/beat sinebeatinputdwelledatthesignificant structural frequencies oftheMCC(bytest,thefundamental frequency wasfoundtobe8.5Hz).Themotionwassimultaneously appliedinthe-horizontal direction at1.35gandintheverticaldirection at0.95g.Thecorresponding responsespectrumisshownintheabovefigureasthethinnersolidline.Subsequently, Westinghouse performed amulti-frequency, multi-directional dynamicanalysisusingtheresponsespectrumshownintheabovefigureasadashedline.Westinghouse termedthisresponsespectrumthenGinnaStationSSERequiredResponseSpectrum".

Theanalysisshowedthatthein-structure accelerations inducedbythisresponsespectrumwereabout.1/2thoseinducedbythetest.Basedonthis,Westinghouse concluded thattheMCCwasseismically qualified.

ThebasisofWestinghouse's analysisspectrum(thedashedline)isnotknown.Forcomparison, theA-46floorresponsespectrumforAuxBuilding271'thelocationoftheMCC)isshownintheabovefigureasthethickersolidline.TheA-46spectrumissubstantially belowboththeWestinghouse analysisspectrumandtheWyletestspectrqm)

.particularly inthefundamental frequency rangeat8.5Hz.TheZPAoftheA-46spectrumis0.3g,orlessthan1/4ofthetestinputlevel.Whiletheoriginalsinebeattestwasnotassophisticated ascurrent-day multi-frequency, multi-axis tests,theSRTconcluded thatthehighacceleration levelsusedinthetestweremorethanadequatecompensation.

Notethatthetestdatadocumented afundamental frequency greaterthan8HzandthatAuxBuilding271'swithin40'feffective grade,therefore theMCCmeetsthescreening requirements forGIPMethodA.TheSRTchosetobaseitsacceptance onthetestreport,ratherthanMethodA,becauseitfoundthatthetestreporttobeamorecompelling argument.