ADVISABILITY OF SEISMIC SCRAM 6 E CuMinga J E WallsH E Lambert an G St - PDF document

ADVISABILITY OF SEISMIC SCRAM 6 E CuMinga J E WallsH E Lambert an G St Lager-BarteJun 30 197UC.'U«-5215Prepare fo U.S Energ Researc & DevelopmenAtiminlstratio unde contrac No W-7405-Eng-4LLAWRENCUVERMORLABORATORilU DOCUMEN I UNLIMITE NOTICn amor wa oMarm m M wn^OMBIMO tIMU th IW Raaaarc• Peiiloamm AwtWumnii M M o IMMpWytH W M O UMiriMltlBllQtltllwCOIMIeClOFI0 iRM OMpMyMO Mfke M O O MMM M (Mil MMM O tomoMfMMfo til Metncy coaiplMotM o a* o frow d»;liMd,o | M M woa oo tafrwjl rirjtflNOTICReferenc t • ruaana o pmjhK noJHM o fTOfMMttUo o 1 produc b IMrorafc o CatkW ~ •£ UJ Efaqg Rematc* Dewlonmer. AdjaWMtetk t Ha ma to wKaMoPrinte i th Unite Sla o AmericArailabl froNallcna Technica Informatio Senlc Departmen o Common Por Roya RoaSpringfield V 2216Price Printe Cop J ; Microfich $3.0PaimHMOamaalPag PJaniPrwPai Rant326-35Prl001-02SPai Rant326-35100024-054.0351-37103051-074.5376-4010.7076-iOS.O401-4211.0101-125.5426-SS11.7126-156.0451-47I2.C151-176.7476-5012.5174-207.550I-J212.7201-227.7526-55I3O226-258,0551-5713.5251-27».0576-6013.7276-309.2601-u• 301-32».7"A4 fee eac atftk 10 M tamnem flo (0 u 1.00 MM•4 14.5 fo Mr aaWttom 10 m MGrMM on 1.00 Distributio CategorUC-1iLAWRENC UVERMOR LABORATORUnwstfKorCMbrn Lntrmn.C*ltarmUCRL-5215ADVISABILITY SCRAM G EH G StMS Jun 197f» ..—wotw • — — ' | V mi uf Immwi Wd* it laicf«***. tk«W« M^—•**»•« a .••«lm^»H,lM • '«• «•#*-»** **t M- Hil l* l«*��«Mto«li�Kl«»«n Mf****f*nn� lalr«i• t*f *t fc« ^*nag ^ mrf^DISTSISUIIO O THI DOCUMEN4\ ForeworThi repor wa originall prepare fo th Offic o StandardDevelopment Nuclea KeRulncorCoanisslon-li Content/ib 1 1 Introductio2.Existin Seismi InstrumentatioPropose Seismi Standard1 AreaLikelihoo I'lnn Instrumentatio TrippinHcai-in DurinSpuriou Reacto TripExistin Seismi Tri System HistorVariable Affectin TripCost Tri SystemIncrease Acciden Probabilit t Trip 1Consequence UurinAdvantage a Shutdow Durin 1Gri StabilitHuma JudgemenAftershock 1Desirabilit GeneratinPowe DurinEmergenc ServicePowe Emergenc4 AssessmenFaul Tre Descriptio Tre QuantificatioConclusion Base Outpu5 DiscussioAdvantage a Tri SysteDisadvantage a Tri Systeb7Appendi Interva AnalysiReferenceAppendi fro IMPORTANC Compute Cod 4-iii ADVISABILIT O SEISMI SCRAAbstracTlii stud fo th NucleaRegulator Commissio Assesse thvalu o seismi tri (scrnmsystem oi commercia nuclea poweplants topic wer specificalladdressed I th likelihoo tha instrumentatio woul trith plant " th consequence ospuriou trips 3 th consequenceo trippin durin a earthquake4 th advantage o a controlleshutdown an 5 th desirabilit ocontinuin operation A comparativris assessmen wa mad base oou finding b evaluatio o faultree constructe fo plant wit anwithou seismi tri systems Iwa foun tha seismi tri systemwoul hav a smal an undetermineeffec o ris du t nuclea reactoaccident an ma indee increas thris t societ fro a earthquake1Thi i th fina repor o a stud b Ch Lawrenc Livermor o th advisabilit orequirin a seismi tri (scram ocommercia nuclea powe plantsPreviou studie hav addresse thfeasibilit o providin a seismi system an th presen repor limite t a consideratio o thadvisabilit o suc systemsI assessin th valu o a seismitri system w studie severaspecifi areas• Th likelihoo tha existinplan instrumentatio wil caus a tri durin a earthquake• Th consequence o spurioureacto trip cause b a seismi system• Th consequence o trippindurin a earthquake• Th advantage o a controlleshutdow durin a earthquake• Th desirabilit o allowinnuclea powe plant t continu tgenerat powe durin a earthquakeThroug a generalize faul treanalysi w assesse th ris i eac thes specifi areas W comparefaul tree fo nuclea reactor witan withou seismi tri systems�asir estimate data Th quantifi o th faul tree madpossibl a compariso betwee th o plant wit an withou tri systems an betwee cause ris an otheacciden risks-1 Fro th ris assessmen aninformatio develope i th fivarea o study severa advantagean disadvantage o seismi trisystem ar listed Sinc th degret whic suc system woul protec a nuclea plan ar somewha dependen th amoun o tim betwee trian stron motio arriva afforde bEXISTIN SEISMI INSTRUMENTATIO ANREGULATIONAlthoug ther ar presentl nrequirement fo a seismi tri ocommercia powe reactors certaiseismi instrumentatio ha t b b th operatin organizatiofo sensin an recordin purposesI ma hel pu th seismi triquestio i th prope perspectivt revie th existin requirementan als som currentl proposedSeismi instrumentatio a existin powe plant consist mainl o time-histor accelerographspea accelerographs seismi switchesan response-spectru recorders Thpurpos o thi instrumentatio i t th intensit o a earth on occur an monitorth plant' response Informatiofro th instrumentatio car the buse t hel astablis whethe tha tri system a specia appendi waadde tha estimate fro historicadat ti advanc shutdow timinherentl available Compute outpu i anothe appendi rank thimportanc o th basi event o thfaul Tee t provid a eas meano establishin critica contributort riskoperatin basi earthquak bee exceede an co.apl wit threquiremen tha th plan b shu2 dow i i has Also th instrumentatio help i validatin model an predictio oseismi responseTh response-spectru recorderwit limi switche attache t itvibratin reeds ca b use tindicat a earthquak t Lii cortroroom Othe instrument eithe*Th OB i th mos powerfu earthquak that considerin th regionaan loca geolog an seismolog an characteristic o loca material coul reasonabl b expecte t occu i thvicinit o th plan sit durin thoperatin lif o th plant Thosfeature o th nuclea powe plannecessar fo continue operatiowithou undu ris t th healt ansafet o th publi ar designe twithstan th vibration o thiearthquake2-2 actively or passivel record theilocation A a minimum instrumentatio i positione t rea th frefiel respons an I als locate o containmen foundation containmen walls an sensitiv equipmeninsid th containment Th amoun o depend o thmaximu groun acceleratio nr.'dict.'sfo th saf shutdow earthquake.Ther ar severa system use fodetectin seismi motion Includei all ar switche t detec motiogreate tha som threshold Thsysfm coul b use t initiat oinhibi som actio whe motio velocity o acceleration bu r.nrmall velocit anacceleration exceed som threshold mos commo typ i th oidampe pendulu an variation o itsuc a a meta bal irakin contac adjacen conductors Thityp o switchin i use tinitiate fo example th recordino groun acceleratio wit a stronmotio accelerograph I ca al-3 bTh saf shutdow earthquak ibase o a evaluatio o thmaximu earthquak potential considerin th regiona an loca an seismolog an specifi o loca subsurfacmaterial I i thr earthquaktha produce th m -tmu vibratorgroun Motio fo wl c certai systems an componentar designe t remai functionaluse t inhibi action suc aelevato operatio i tal buildingswher require b buildin codes switche ar basicall aacceleromete wit sensitivit ith bandwidt chose a mor o th earthquakspectru tha o civilization'baekgrojn noise Thes switcheca als b selecte an use fo an — vertica o horizontal a electromechanica device ii a straifihtforwar mitte t adjusmechanica constant' an circui t selec spectru I additio t sensi i a selec bandwidth logi fro separat coul b use t minimizspuri . seismi trip wher spurioutrip woul b costly Thi coinciden logi migh b necessarbecaus a individua senso coulb jarre b som broadban sourc(e.g. a droppe wrench) wit enougenerg i th frequenc o interest excee th se threshold Thiproble ca als b circumventethroug th us o significantlhighe thresholds bu thi ma defeath origina inten o th switchPROPOSE SEISMI STANDARDTh regulation an standard thadictat th presen seismi instru requirement ar Regulator-3 Guid 1.1 an ANS standar N18.5A proppe standard AN 2.10/N643i bein prepare C specif thprope respons th plan operatoshoul mak base o th instrumenreading followin a seismi eventBasicall thi standar relyin ocontrolled manua shutdow proposea tw ste approach I ste 1 i idetermine whethe th OB ha beeexceeded Thi wil b indicate ia minimu numbe o point o threcorde respons spectru havexceede th OB desig responsspectru 0 i ther i a equipmenmalfunction I eithe i th caseth reacto i shu dow i a controlle manne Oa tak fiv tte th planma the requir th concurrenapprova o th Nuclea RegulatorCommissio ..n th plan owner i n clea criterio o whict asses th advisabilit orestartin th plant s onc a plani shu dow i ma b shu dow fod substantia timeI i ste 1 i i determine thath OB wa no exceeded bu thplan operato ha a indicatio tha earthquak di occu the hcontinue operatio an proceed to 2 I ste 2 th recorde dat.i analyze t se i th desig spectru ha beeexceeded If afte analysis iappear tha th floor—responsspectru fo th OB ha beeexceeded th plan I shu downThi analysi ma tak fro on dit on weekNot tha th inten o thipropose standar I t enabl thoperato t determin whethe thplan shoul b shu dow an tshu dow i i appear th OB ha exceeded I judgin this thoperato i onl t tak Intaccoun th conditio o th planan i no t asses th nee foth electri generatin capacit oth plant I.e. t conside mitigat th consequence o th earthquak I th surroundin communityAlso plan shutdow commencesther i n establishe procedur forestart Therefore a i nostands AN 2.1 i adopteo not i th plan I shu dowbecaus o a earthquak i masta ou o servic durin th emergenc perioincludin post-earthquak recovery a seismi tri o lo tri levewer used thi woul assuredl bth cas unles regulation arfurtiie clarified Withou a seismitrip th operato woul hav somdiscretion althoug legall h shu dow a soo a h icjnvince th OB ha beeexceeded-4 3 StudLIKELIHOO O EXISTIN PLANINSTRUMENTATIO TRIPPIN THREACTO DURIN A EARTHQUAKTh sensitivit o th normaplan instrumentatio t th seismimotio portio of th mechanicavibratio an nhoc spectru Ideterminabl bu o questionablvalue detector ar mounte tha the coupl onl t thfre fiel eart motion an signa the ma giv a a direcfunctio o an seismi inpu woulb masked Thi maskin woul bdu t motio attenuatio o amplificatio becaus o plan structuresmountin techniques etcT full understan th responso a instrumentatio channe tsom seismi input tha i whethei wil initiat som safet actionon mus kno it vibratio characteristic durin norma operatio an respons spectru couplin t thdetector Al clas I electrica i suppose - b qualifiet ANS N41. Seismi Qualificatio Clas 1 Electri Egxilpment Thi require tha th equipmen it safet functio durin afte th disturbanc cause ba design-basi earthquak an doeno reall addres possibl vibratioinduce (spurious trips Bvibratio induce trips w mea a tri tha occur whe th sensevariabl doe no change fo tri cause b mechanicashoc o a io chamber A examplo a genuin respons woul bflui leve sloshin i th pressurize o a pressurize wate reacto(PUR) result i a wate levetrip th leve detector arqualifie t ANS Ml.7 the wilsens hig o lo flui leve an accordingly a th variabl a genuin change W wil primaril concerne wit vibratioinduced spuriou instrumen responsi thi sectionTabulatin th "at-power-operationtri channel a typica PW an go throug the t determin th like tha a tri coul b induceseismicall indicate man uncertainties tri channel areN stro flu temperatur (bot absoluttemperatur an temperaturdifferencesCoolan syste pressur wate leveCoolan floCoolan pum statu feedwate floSteam-generato wate leve operationastatu-5 OperatoBot reactor-protectio systelogi train bypassedSkippin ove th individua detector fo eac o thes fo a momenan lookin a th operatio o thtrai o logi (assume her t brela logic connecte t thdetectors som genralizationA simplifie sketc o a portioo th logi trai i show iFig Thi figur indicate thather ar severa relay i seriefo eac channel som requirincoinciden action Appropriat probabilitie ar i conjunctio wit thactua rela qualification an logiimplementatio scheme t mak anquantitativ statemen o potentia responseHowever th relay an circuibreaker probabl hav th samsensitivit t shak an vibrationand sinc the ar al ver similaral thi relayin ca b consideretogether Ther i a basi MIQ specificatio fo relay and shoulth relay an circui breaker bqualifie t an seismi shocapproachin thi specificatio(aroun 1 ar no likelt giv spuriou signalsTh detector likel t b use ith tri channel give abov artabulate i Tabl 1 alon wit somestimat o th detectors individuamotio sensitivities Whil noabsolutel complete th tabl igenera enoug t cove th majorit th detector i th tri channelsA fa a spuriou trip frodetecto respons t a genuin change th mos obviousl detecto i th turbinvibratio monitor Thi monitor'respons spectru overlap (o itsensitiv axis th spectru o mosseismi motion Som fossil-poweunit hav actuall experience a tri durin a earthquake Ga pn Electri Company'Ker Count Uni trippe durin th195 Bakersfiel earthquak becauso a turbin thrus motio sensorI th seismi motio i largenoug th vibratio monito wil an i thi monito i t a automati turbintri tha woul clos th turbinsto valve th reacto woul trifro fas closur o th turbincontro o sto valve I shoul bnote tha eac nuclea plan operato th optio o connectin turbin monitor t th turbintri system an tha severa havchose no to i.e. Pacifi Ga an (a Diabl Canyon anNorther State Powe (aT summarize a generalizatio oth suitabilit o curren plan LosteamgeneratowateleveLofeedwatefloLogi trai 'A• a buReactortriswNotebot bypassetrip reactorVri \ Bypas'BRodrivbuOthesafetfunctionSafetyinjectioblocFeedmasteisolatioTurb ntri (anbac arount reactot»1pSafetyinjectioblocFeedwateisolatioTurbintri (an arount reactotripFig Simplifie logi trai fo stea generato wate leve an feedwate flo portio o plan tri system Tabl 1 Motio sensitivit o typica tri channe detectorsVariablDetectoSensitivit t motioNeutro flu Ionizatio ChambeCoolan temper ThermocoupleaturSom sensitivit t motio givespuriou signals I qualifie tMIL-S-90 an MIL-Std-16 fo structura integrit the woul probablb muc les sensitive Electronicompariso fo tri referenc halo sensitivit t seismi inputLo sensitivit t motion somsensitivit t flexin nea junctionLu probabilit o spuriou outputResistanc thermomete Lo sensitivityCoolan syste Pressur transducerprepsur usingLinea variabldifferentia transforme (LVDTForc BalancStrai gaugMostl lo sensitivit bu a functioo th mas o th sensin elemensuc a a diaphrag o bourdo tubePressur switchesSna actioMercur switcLo sensitivitySensitivit t til fro les tha5 t a muc a 20° usefu ainertia switch Ver sensitiv tacceleratio i somPressurizewate leveConductivit gaugFloa gaugLo sensitivit o senso itselfSom sensitivit a a functio ofloaCoolan floCoolan pumTurbin mas flowmeteOrific differentiapressurClosure an limisensorLo sensitivityLo sensitivit simila t pressurtransducerLo sensitivit o relay an switcheunles ver clos t actuationSteam-generato Simila t pressuriwate leve ze wate leveTurbinestatuThi i par o balanc o planinstrumentatio an a suc i mao ma no hav t mee th Clas I equipmen requirement However idoe ente th scra train Tabl 1 (ContinuedVariablDetectoSensitivit t motioManua turbin Closure an limitri sensorCondensevacuuOverspeeThrus bearinfailurOperatoBot logitrainbypassePressur switchSna actioMercur switcTachometeFrequenc counteVibratio pickuClearanc capacitgaugClosur an manuaCircui breakeand/o rela closureLo sensitivityLo sensitivitySensitiv t til an shakeLo sensitivity sensitivitySom sensitivitySom sensitivitySwitc sensitivit low presenta o variable t th operatorma giv fals cue an a theidiscretio th operator ma trith plantLo sensitivityinstrumentatio fo th secondarpurpos o seismi trippin canno bmade questio require detailestud o th specifi plan aninstrumen installations an thresult woul b applicabl t thaplan only Th instrumen systewil probabl tri th plan durina earthquak mor sever tha thOBE i th plan idesigne fo a larg OBE Howeveri appear t-t b onl marginallprobabl tha th lowe amplitud motio (se VariableAffectin th Frequency unde Spuriou Reacto Trips woul tri ainstrumen syste designe t ANS a tri wouloccu durin stron motio an givn advanc warningShoul trippin th plan bdesire fo seismi disturbancestw alternative presen themselves1 us presen seismi o eartmotio sensor t properl alerth operator an leav th actuatrippin t thei discretion o2 us a wel designed redundanan coinciden triaxia seismitri syste t tri a a chosethreshol tailore t a earthquak"signature. SPURIOU REACTO TRIPExistin Seismi Tri System anThei HistorN commercia powe reactorcurrentl emplo a seismi trisystem However a numbe oresearch test an productio ar no usin o hav usesuc systems an thes ar liste iTabl 2 wit brie note abou thei I general thinstrumentatio use t sens thearthquak i a seismi switcsensitiv t horizonta accelerationan se t tri th reacto a a give pea acceleration I certai logi i use Ith tri circuitr t hel minimizspuriou tripsMos o th reactor equippewit seismi tri system ar iselsmicall activ areas e.gCalifornia Mos o th tri levelar se s th reacto wil shudow i a earthquak o modifiedMercall intensit V (MJ.O g iexperience a th reacto site Aearthquak o thi intensit woulb fel b mos peopl bu no caus damage Som o th reactors wher earthquake o thi intensit ar fairlcommon thei tri level se amodified-Mercall intensit VII(VD. intensit tha shoulheavil damag conventiona structuresbu on tha a wel designe reacto survive A ca b see fro 2 man installation havexperience unwarrante trip frothes systems wit th frequencdependin o tri level circuilogic th seismicit o thsiteVariabl A fe tin Frequenco Seismi TripA unnecessar shutdow cause ba seismi tri syste i no nearlth proble fo th researc reactorliste i Tabl 2 a i woul b foa commercia powe reactor Researcreactor ar designe fo u andow operation whil powe reactorar designe fo stead operationThus seismi tri syste fo a reacto woul hav t b imicmor reliabl tha thos fo researcreactorsUnnecessar trip ca resul frotw causes poo instrumen performanc (nois o o th cri systet low-intensit earthquakes Thfirs caus ca b minimize b uso reliabl instruments redundan logic o discriminatio ith senso t allo respons onl tearthquak spectra Th EBR-I o Experimenta Breede Tabl 2 Seismi tri system employe o reactorsFacilit1 Material TesReactoEngineerin TesReactoLocationzoneIdahzon 3 Tri levelgravitiePerformanc0.0 - 0.0 N spuriou trip2 ProductioReactorWashingtozon 2 0.0 - 0.03 Savanna Rive Pro Sout Carolin 0.0 - 0.0due io Reactor zon 3 N excessiv spurioutrip (use 2/logic)1N problem14 IdahBreede Reacto I zon 3 Sodiu ReactoExperimen anvariou otheAtomi Internationa experimenta reactorCalifornizon 3 0.00 spuriou tripi fiv years'«(no us 2/ logic0.0 - 0.0 Severa spurioutrip16 UCL ResearcReactoCalifornizon 3 7 ResearcReacto3 Valleclto BoilinWate ReactoGenera ElectriTes ReactoCalifornizon 3 Californizon 3 0.00.00 - 0.0Tw trip i fivyears bmino earthquakeSevera trip causeb earthquake an a serie i spurioutrip cause b build settling9 UTR- ResearcReactoCalifornizon 3 0.0 - 0.0N trip ovesevera years10 BoileNeutro Sourc(Lawrenc Livermor Laboratory11 WateBreede Reacto(Shipplngport12 ResearcReactoCalifornizon 3 Pennsylvanizon 1 Californizon 3 0.0 - 0.10.00(2/ logic0.0 - 0.0N trip ove fivvearNo ye installedN spuriou trip ite year•Unifor buildin cod seismi zones-11 Reacto I (se Tabl 2 suggest thtr.'axia seismi trigger an 2/channe logi ca reduc th frequenco spuriou trip t les tha onpe fiv year eve wit a ver lotri leve setting Thi i t bcompare t a expecte spuriou trifrequenc o te pe yea a a moderpowe reactor Th us o a respons recorde shoul improv oth EBR-I experienc eve moreDiscriminatin ou lo intensitearthquake (th secon caus o trips ma b mor o a problem Th firs though woul bt se th tri leve high s thaonl damagin earthquake tri thplant Sinc th plan i designet accep a OB an remai operable ma b th appropriat levelHowever i thi wer don th planwoul no tri o precurso motionan therefor woul shu dow durinstron motion O th othe handsettin th tri leve ver lo woulmaximiz th chanc o shuttin dow th stron motio arrive becaus o aftershocks mighresul i lon period o downtimafte a majo quak o durin a swarm Th Hebg--.i Lak (magnitud 7.2 resultei 2 aftershock durin th simonth followin th mai earthquakean 4 o th aftershock oc-urre ith sam week Danville iliforniarecorde severa hundre measurabl durin a earthquak swari 1970Sinc powe reactor tak manhour t restar afte the artripped th hig frequenc o aftershock immediatel followin a majoearthquak migh resul i long-ter o al nuclea uniti a larg are i th tri levewa se low Thi woul b thcritica perio whe electri powet th surroundin communit i mosneede an whe th les stringentl fossi unit woul probablb ou o servic alsoCertai characteristic o thwave fro a earthquak indicat itsourc distanc fro a instrumentThes wave ar predominantl th P an S waves an the ar usefu i th source Thei differinvelocitie ma giv a tim margini whic som actio beneficia tplan safet migh b taken Thlate arrivin S wav i usuallassociate wit th stron motionan th tim interva betwee th P an S wav arrival coul b enougtha som motio sensitiv equipmencoul b pu int a mod t withstanstron groun motionWit a earthquak sourc largenoug an clos enoug t d damagt a powe plan (i.e. greate thath saf shutdow-12 lim betwee th P an S wav arrivawoul b short Wit propagatio velocitie o th orde o ont thre kilometer pe secon thtim interva i probabl les thaon t tw seconds bu th buildut pea acceleratio (base o previou stron motio records ca b a fe seconds A show i Appendi Ath mea tim t stron motio afteP-wav arriva i thre t fiv secondfro P-wav tremor o 0. g ogreate intensity Thi warnin timi a averag valu an woul holonl fo site wher S-wav motioi significantl large tha anseparabl fro. P-wav motionTher i n genera scal factorelatin th P-wav motio t thS-wav motion I a ha bee determine foon event the othe earthquakefro th sam soui'c are wil havapproximatel th samamplitud relationship Fo earth swarm originatin i th samarea th rati ma sai t b approximatel constant sam ma b sai fo aftershocks magnitude o motio ar frostron motio recordings regardles th spectrum Th frequenc o th earlie P wav migh a stron earthquake s thaa frequenc sensitiv seismi tripeve i se a a lo level woulonl activat fo stron earthquakesInsufficien informatio i availablt determin th feasibilit o thiapproachWit wha i currentl known a lo leve tri (se a les tha0.0 g woul hav t b use t makus o th tim interva betwee thP an S wave (se Appendi"sufficien time interva woulallo activ component o th powe t complet thei shutdowactio befor th stron motio i thre t fiv th plan wil no hav shu down However suc a th contro rods maisteam-isolatio valves an turbin valve woul complet o partiall complet thei shutdow actionSinc th emergenc cor coolin (ECCS i no activate untia acciden occurs thi warnintim woul no affec it operabilityOn o th mai concern abou aearthquak i tha i migh rupturth reacto coolan system leadint a los o coolan accident Ithi wer t occur a tri o threacto thre t fiv second beforth acciden woul allo equalizatioan som remova o th store energ th fue rods resultin i a decreas i th subsequen peatemperatur o th claddingAssumin th ECC works thi ma eliminat an cla perforatio dut th acciden an minimiz th of1sit release However thi i a smal matte whe compare t a cormelt th cor woul mel i wha ECC failure regardles o thwarnin time Not tha accordint Newmark th probabilit o ECCfailur resultin fro groun motioexceedin tha o th saf shutdowearthquak b a facto o 2. ionl 0.02.Cost o Seismi Tri SystemI considerin spuriou tripsther ar thre mai cost t blooke at Th firs i th initiacapita cos o a seismi tri systemA present reacto site ar requiret hav a prescribe amoun o seismi Th increase cos th utilitie o a seismi trisyste wil no� b grea i thsyste incorporate som o thexistin instrumentation Howeverrequirin thi instrumentatio tmee ANS N41. woul incu addecosts secon cos i tha omaintenanc associate wit thseismi tri system Thi cosshoul no significantl increas thnorma maintenanc cos associatewit reacto instrumentationTh thir cos i tha du tspuriou trip o th system Unlikth othe tw costs thi on coulhav a significan monetar effeco th utilities Whil th costassociate wit spuriou trip diffewit locatio an siz o reactorthes cost coul hav a impac o utilit i th spuriou tri ratwa substantial Whe a larg nucleapowe plan suc a on o FloridPowe an Light' 70 MW TurkePoin unit i unabl t operate thextr cos o usin fossi fue treplac th nuclea generate powe upward o a half-millio dollarpe dayAnothe exampl i tha o BrownFerr nuclea powe plant ThBrown Ferr fir i March 1975 shudow bot reactor a th site whethe wer producin 110 M oelectrica power Prio t th fire Brown Ferr reactor wersavin th Tennesse Valle Authoritabou $240,00 pe da i fuecosts powe companie woulhav coi'parabl costsFigur 2 illustrate th effec aincreas i tri frequenc wil hav th plan economics T evaluat dolla amount involved assumi cost $3",3,00 pe da t replacth nuclear-generate powe witfossi fuels an tha i take sihour t brin a reacto bac olin afte a reacto trip Th cospe reacto tri woul the b abou$125,000 a increas i 0 1 1Tri frequenc -1 2trip pe yeaFig 2 Increase cost du tincreas i tri frequency assumin th following outag tim dut tri equal 6 h an cos i$500,000/day-outag o $125,000tripspuriou tri frequenc o on peyea woul ad fiv millio dollar tt plan operatin cost ove a 40yea life Althoug thi i a smalamoun i compariso t totageneratin costs i coul becomsignifican i area wher reactorar subjec t earthquak swarmse.g. California Also w hav no othe economi consequencet th utilit o havin larg poweproducin unit wit increaseprobabilit o sudde shutdown e.g.increase exces capacit requirementTh locatio o a reacto becomecritica whe considerin th monetar effect o seismi trips I a utilit decide t buil foureactor o on site the thautilit woul ru th ris o havinal fou reactor shu dow a ontim i a tri wer cause b a smalearthquake Hig tri leve settingcoul alleviat thi proble but amentione i Sec 2 thi coul th chanc o havin an warnin tim betwee tri arriva o th stron motionIncrease Acciden Probabilit Dut Spuriou TripSpuriou trip durin normaoperatio ma produc a greateacciden ris tha th earthquak Fo example WAS 140 show th greates acciden ris fo a boilin wate reacto (BWR i a transien even causin reactoshutdow followe b los o th hearemova systems WAS 140 alsstate tha th probabilit tha BW remova system wil fai followin a reacto shutdow i 1. x 1 , assumin offsit powe t b availableThi probabilit i muc lowe fo a PU i a stea driven auxiliarfeedwate pum i available Regardless th spuriou scra rat dut a seismi tri syste i 0.2/yri.e. onc ever fiv year (froRef the th probabilit tha a spuriou tri wil caus a fuemeltdow i a BW i (0.2)(1. * 10~ 3. x 10 pe reacto year FroWAS 1400 th probabilit tha aearthquak wil caus a fue meltdowi a reacto withou a seismi trisyste i o th orde 1 . Thus-15 i al thes probabilit value arcorrect a smal increas i spurioutri rat o a BWR' seismi tri woul pos a greate ris o mel tha th earthquak itselfAlso a incrementa increasi tri frequenc wil increas thtota numbe o therma stres cycle plan wil se durin a perio otime increase numbe o cyclecoul possibl hav a effec o thplan life Not tha th currentl tri frequenc o a nuclea reacto wit a seismi trisyste i te pe yearCONSEQUENCE O SCRAMMIN DURINA EARTHQUAKWithou a seismi tri system threacto ma o ma no b shu dowdurin a earthquake dependin oth performanc o existin instrumentatio an th plan operatorsWit a seismi tri system threacto wil b shu dow eithe jusprio t o durin th stron motion th reacto wer t shu dowdurin th stron motion additionastres woul b pu o som Th therma transienma impos therma stresses Alsoactiv component ma bin u o thei movemen durin motion Sinc criticacomponent ar designe t withstanth combine stresse o acciden anearthquak loads an increase stres shutdow du t th thermatransien shoul no caus systePossibl effect o acciv component require t chang stat a earthquak ma b mor Tw classe o activ nee t b examined On i th seismic-category- suc a BW scra valveo relay i th reacto protectivsyste tha chang stat durin a Sinc suc component arspecificall designe t perforsafel durin a saf shutdow the ca b expecte tperfor properly Th othe clas activ component i th noncategory- activ component thamus chang stat durin a scramTypica o thi clas ar thfeedwater an mai steam-isolatiovalve an valv controllers Sincthes item ar no classe acategor I the migh malfunctiodurin a stron earthquak an leat a unpredictabl therma transieni shutdown A detaile stud othes system woul b require tdetermin thisShuttin dow prio t stronmotio woul allo activ componentt chang stat befor earthquakmotio an hel t assur a normatherma transient However sinc th expecte advance warnin timear o th orde o a fe second(se Appendi al systemwil hav complete thei actio bth tim th stron motio arrivesan trippin duri.i a earthquak ma.iffee thei action Som o thsafet systems e.g. contro rodan mai steam-isolatio valvesmigh complet thei actio i thtri leve i se lo enough an tthi exten chance fo a safshutdow wil b enhancedA seismi tri syste woul havlittl effec o th operabilit oth ECC an othe engineere safetsystem calle int pla durin aaccident Sines thes ar initiate respons t a acciden i woul bexpecte tha the woul hav t durin stron motio i tha mot"o cause th accident the ar specificall designet withstan th SS the shoul nob o a muc concer a th nonADVANTAGE O A CONTROLLE SHUTDOWDURIN A EARTHQUAKControlle shutdow durin ofollowin a earthquak offer somadvantage ove a automati seismiscra system Thes advantage pertai t gri stability humajudgements an aftershocks However-17i woul no b possibl t shu dowprio t stron motio i thi strategwa usedGri StabilitI ha bee estimate tha thprobabilit o losin off-sit powedu t a power-gri transiencause b a nuclea plan tri i— 3 11 pe trip Thi estimat imad fo a normall operatin grian singl nuclear-uni trip Sinc-1 i n average th probabilitcoul b highe o lower dependino th sit o th reactor amoun oinstalle capacit an numbe ogri interconnection wit othe powe companies Th probabi o powe failur woul blarge fo a gri tha ha experience a majo earthquak an los o severa nuclea instance fou t sinuclea unit ar planne fo th Canyo sit alon th Californi coast I thes unit weral trippe simultaneousl b seismitri system durin a majo earthquake woul hav a sever impac Pacifi Ga & Electric' gri icentra California Thi coul leat th unavailabilit o powe tth populatio whe it' neede tcop wit th problem tha arisafte a earthquake Huma JudgemenThi subsectio addresse thissu o allowin th operato teffec a controlle shutdow i ihi opinio a damagin o largerthan-operating-basi earthquak haoccurred Thi issu i alsexemplifie b th AN 2.1 standard guidelines Appropriatseismi instrumentatio woul bavailabl t ai th operato i a decision Durin th shutdown operato coul asses an damag t th plan anchec th conditio o th off-sitpowe grid Th controlle shutdow provide tim t chec th providin emergenc powet th sit an brin the intoperatio i cas off-sit powe ilost o thes item plu another uniqu t a soecifi facilityth operato ca decid whethe tresum operatio o ntlnu thshutdownI thi strateg i employedprocec shoul b writte t givth operato a specifi guidelin fomakin hi decision Thi i neede o th concer confusiotha ca b anticipate durin anafte a earthquake Thes procedureshoul als prepar th operato fopossibl malfunction o automatiprotectio devicesAftershockI a reacto containin a seismitri syste i shu down inspectedan brough bac o lin agai aftea earthquake i ma b shu dow mor time becaus o aftershocks th operato tperfor a controlle shutdowminimize th possibilit o repentetrip du t thes aftershock sinch coul ignor the i the ar osufficientl lo level Ther woulthe b les chanc tha electripowe woul b disrupte fo a lon folJowin a earthnuakeDESIRABILIT O ALLOWIN TH REACTOT CONTINU GENERATIN POWE DURINA EARTHQUAKEmergenc ServiceI th utilit gri an thinterti fro th nuclea powe plan th gri� (neit;e o whic arseismic-category- structures iintact th tota ris t thpubli migh b les i th plan t suppl electri powe emergenc service tha i i down Th degre o risk ocourse o ho muc o thpowe o th utilit gri come fronuclea unit an ho man o thesar shu dow b th earthquakeI discussion wit count officeo emergenc service an th utilit-18 jordinato fo th Stat oCaliforni Offic o EmergencServices followin informatiowa obtaine concernin th effec oCaliforni emergenc service iutilit powe i lost• Eac '-'innr offic ha a leastw contro centers Thes controcenter ar use t contro emergencservice durin an emergencsituation Eac contro cente hait ow sourc o powe an caoperat fo severa c ,. withououtsid assistance• Hospital hav a emergencnowe source Thi powe sourc ino designe t completel providr.hu. need o th hospita bui designe t provid enoug powet handl al emergenc situations• Al local stat an federa laenforcemen agencie hav emergencpowe availabl whe utilit powei no available Thes agencieals hav a larg numbe o two-wayself-powere radio availabl t thei emergenc situations• building an largdepartmen store hav emergenclightin fo stairway an othe areas• Telephon companie hav emergenc backu powe i th even utilit powe i lostAl th item mentione s faindicat utilit powe ma no b ovita nee durin a earthquakemergency However th followinpoint wil tempe tha conclusionunde th assumptio tha th earthquak affect a larg area• Wlui hospital d i fachav thei ow sourc o emergencpower o these source appeano t have- al o thei component inte tie seismicall qualifiedTherefore afte a sever earthquakthi powe ma no b available• Vehicl trave become dangerou an slo becaus traffi light inoperativ withou utilit i aggravatedurin th hour o darkness• Electricall powere streelightin wil no b availablbecaus n emergenc sourc wil b available• Departmen stor lightinwoul no b availabl excep foth emergenc stairwa Hghttnmentione earlier• Whil i s home ar ga heatedth thermosta control ar electricall powere an furnace woul becominoperativ withou utilit power• Man alar system requirutilit powe an wil no operatwithou it • +10$ +1+ 1 -+o K--1105.0 ±± 2.5%Compressio ± pr— Extensio "20±10(bFig 3-8 Exampl o o fro a triaxia tes a consolidate specimenTherefore a failur criterio base o axia straiha bee adopted Th numbe oa give strai i define b tw methods• A singl amplitud stiain—fo example a 2.5 i define a th poin wher th doublamplitud strai (peak-to-peak equal 5.0% owher th maximu tensio o compressio excur fro th stati conditio equal 2.5%• Failur strain—foto-pea strai mus b definedA first ste i analyzin th result o testingw hav summarize th por pressur increas an v cycle fo specimen subjecte t th sam confinin pressur (Fig 3-9) Thes datar the replotte indicato cycle require t obtai eithe initia liquefactio o a give leve o strai i term o th stres Th us o thi stres ratiminimize th effect o smal variation i conHnin pressuieSignifican difference exis betwee th strescondition ir th isotropicall consolidatelaborator specimen an in situ specimens.Severa method hav bee propose t accoun fo differences. thes method diffe form the indicat tha th laborator result(fo example th stres rati i Fig 3-10) shoul bmultiplie b a facto rangin fro abou 0. lo 0.fo normall consolidate soils Thes result coulthe t facto mainl applicabl t clea sands Fo soil containin a appreciabl amoun o fines, th ha no ye bee established.Anisotropic situo soi subjecte t initia (static shea stresse i b anisotropicall consolidate triaxia3 O Fossil-fuel-firepowe plano Hydroelectri poweplanFig Theoretica area affecte amodified-Mercall intensit V ogreate b earthquake o varyinRichte magnitud (R-M centere iSa Francisco.Iearthquake an an automati seismitri system woul likel hav a widimpactPowe fo Reacto Emergenc anDeca Hea LoadI a earthquak wer t caus thlos o off-sit power the on-sita powe woul b require t removdeca hea o cop wit a los ocoolan accident I al reacto a th sit hav als beeshu dow b thei seismi tri th diese generatorwoul b th sol sourc o a powerUnde thes condition tw probabilitvalue ar require t determin thoveral probabilit tha th powe o th emergenc o\ decay-hea remova system wil bme followin a earthquake Thfirs i th probabilit o losin electrica gri o interti t thgri i a earthquake Neithe thgri no interti i a seismic system Sinc thprobabilit tha off-sit powe wil los becaus o reacto tri unde-a non-earthquak situatio i 1(Ref w ca us thi a a loweboun fo th earthquak situationTh secon probabilit o conceri th probabilit tha th dieselgenerato set wil fai t provid a sufficien amoun o powe afte aearthquake Al safet relatediesel-generato set a reactofacilitie hav redundan backu anar seismic-category-I i.e. thear designe t withstan th effecto a saf shutdow earthquake Sinc probabilit tha a diesel-generatose wil no provid a sufficien o powe whe neede unde a -21 non-earthquak situatio i- 13 x-1 , thi wil b use a a lowe bound (Th Crook an Vissinrepor state tha a a 50 confidenc level th probabilit osuccessfu diesel-generato se2performanc i les tha 95% whici equivalen t a probabilit o_failur o 5 x 1 o more Thivalu i i roug agreemen wit th3 x 1 value. Combinin th twprobabilitie give a minimuprobabilit fo no meetin powerequirement o th decay-heatremova syste afte a earthquakeA thi point w perfor a comparativ ris assessmen t helestablis whethe seismi tri a nuclea poweplant i desirable W us faultre analysi t conduc a tradeofstud evaluatin th ris o seismi tri system versu ris o no requirin them Wdefin ris a th probabilit tha a los o lif o propert occurs Th tre i Fig 4 consider thris o nuclea powe plan operatiowithou seismi tri system whiltha i Fig 5 consider th riswit seismi tri systems I thisectio w firs qualitativel th faul tree iThi valu i 3 x 1 pe earthquakeI w us th WAS 140 valu o-7 x 1 earthquake pe year thprobabilit o a cor mel fro lac—o emergenc powe i abou 2 x 10pp year Eve thoug thi i a value i i a ver lonumbe whe compare t th normacor mel probabilit o 5 peyear thi ris o seismi tri i probabl no significantI addition bot BWR' an PWR'ar capabl o removin deca heafo severa hours usin stea drivepump independen o a poweran 5 The w quantitativel themFAUL TRE DESCRIPTIOTh faul tree i an 5 ar constructe i a simila manner lef sid o eac faul treconsider acciden sequence tha leat cor mel durin non-earthquakconditions acciden sequence categorize int physicalldistinc processe b whic cormel ca occu a wa don i th Safet Study Th lefsid o Fig 5 illustrate thabecaus o spuriou trips nucleapowe plant wit seismi tri-224 Assessmen •Exclude reade trip causejcccplaM fas', :aus Mrtrit a corLov o heareinova systemcauisei cc mil1 lov or/BHsyste IN•*~*m~mmmm'mm Fig tre plant withou seismi scra system I l^e^w^WPIVftiqinepre safeguars,y tern (prpuen mel£ tSF til beiauiO lOipowe followinPowe fro otheunituiuna lablNuclea pyweplant siubecaus o eartnquakA t'ufflcfmKM Ofluflrani ar\tiU'eartnquak/� iiu a/ ^ tw-caj o «£ A qFig tre plant wit seismi scra system instrumentatio hav increase nonearthquak acciden frequenciesTh righ sid o eac faul treshow acciden sequence du tnuclea powe |/lan operatio whea earthquak ha occurt.'d Thi i furthe divide int th riso plan shutdow o publi emergenc service an th ris o plan operation whicmigh lea t a larg fission-producreleas du t th earthquake'effects referenc t accidentcause b earthquakes th faul treei an 5 diffe i tw• Th faul tre withou seismiscra (Fig 4 include th possibilit tha a destructiv poweexcursio ca occu i th contro ar no full inserte whe aearthquak occur (see Fig 4 even21)• Th faul tre wit seismiscra (Fig 5 include plan whe a stron motio earthquak occur an th plant' inabilitt suppl powe immediatel afteth earthquak (se Fig 5 basievenFAUL TRE QUANTIFICATIOTh faul tree give ian 5 includ th dominan accidensequence leadin t a possibl cormelt sequence wer take-2directl fro th Reacto SafetStudy I addition thes faultree includ earthquak cause eventtha coul lea t cor mel fonuclea powe plant wit an withouseismi tri systems W probabilisticall evaluat bot faul tree establis whethe cor meltcause b earthquake ar significancontributor t th ris o nuclea plan operation 2 determinwha differenc seismi scra maket thi risk an 3 conduc a sensitivit analysi t determinwhic portion o th syste ar oth greates concern Eac probabilit o occurrenc o th basi wil b assigne a mea wit uncertaint factors stha probabilit confidenc intervalca b computedTh firs ste i th quantificatioo faul tree i th assignmen ofailur probabilitie t th basievents estimate o thfailur probabilitie o th basievent wit associate uncertaintfactor i give i Tabl 3 Therro facto expresse th uppe90 limi t th basi failur probabilit b thfollowin relationshipq uppe = q mea +(uncertaintfactor)( q i th^expecte value Tabl even probabilitiesBasiceven Reacto Kxpc.-te rat Uncertaintnumbe Descriptio even tvp laetolos remova system SW 1 • 1 ,' 1reacto tri cause meltdow I. •Los acciden occur durin,- IIKnorma operatio lead meltdow l. •Othe initiatin event leamelDamagin earthquak occursPG (0.1PG • O.if • 10-''/ 1PG " 1.0 •Spuriou tri, cause scra8 Los remo'.a systreact.t tri�im BOthe earthquak .auset'ent9t mel9PG 0.JI0PGllPGOthe earthquak causet mel.'ent12'PGIV PG 0.514I'GEarthquak disable maifeeelw1PG1PG1PG "1.0Earthquak disable auxiRH systemiarIS PGA1PG2PG OU'.itiin•'• (•1.l.'•BI.'110" demanI'W1 .f• deman1 .io-1lirEarthquak prevent controcause destructiv powe excursion2l22a th groun acceleratioAppear tre onl withou seismiC tre onl wit seismiII:-J,-i in-/ev3io/even7 Id"/o',.n) 11 ./even1061 ,/even511 /even2-26 lKarthquak ciusPG 0.2PGJf.'KSK'valvesPG n.2PG 0.5PG •Diesel f 1PG 0,21'G 0.5 l.Olntertle betwee nui-lea plan sitgriPG 0.2jlan atcldento earthquak damag pipesiperat aftePG0 1 •• OLosPGPGe0 0 1 nergenc• 2•• OickupowesystemGriPGPGPGS0 1 fo••• OothereasonsLos powe plant cause gri gdowNuch-a powe plantinstrumenta InnPGPG 0,5PG ].0Nuclea powe plantoperatorsPG 0.2PG I.OSeismi scra syste tripnuclea powe plantexpecte ratUncertainto occurrencfacto5, lO'l 5. 1 , 15 3.2 13 3 10~3 lO^/even3 10 /even'7 3 0.05/deman30.1/deman7 03 0.05/even30.1/even7 0.5/even3 0.01/deman30,05/danan7 0.1/deman1 0.05/even30.1/even7 0.5/even1 0.1/even10.01/even0,2/even0.5/even0.01/even0.I/even1.0/ev^n13 13 2 -27 Th failur probabilitie fobasi event 1 2 3 an 8 appea oth lef sid o eac faul tre anar take directl fro th ReactoSafet Study Fo earthquak causeevents, level o pea grounacceleratio (PGA ar considere ith faul trees1 PGA-0. g [0.1 g PG 0.2 gl2 g [0.2 g PG 0.73 PGA-1. g [PG� Th saf shutdow earthquak i taket giv a PG o 0. gFro Hsieh' work th probabilit pe yea tha a averagreacto sit i th Easter U.Sexperience PG i th intervaspecifie abov i give by16.5 1-2 PG 0.7 4.3 P(0.75g - 1.Newmark' values a cite i thReacto Safet Study fo thprobabilit tha a singl reactosyste wil fai whe subjecte tthes acceleration ar give iTabl 4 Th probabilit o twsystem failin i th squar othes value i the failure arindependent a show i colum 3 oTabl 4 I tw failure ar completel dependent th probabilit ofailur i th sam a fo a singlsystem Th Reacto Safet Studstate tha i i reasonabl tassum tha failur o tw systemi dependent i.e. coupled ThTabl 4 Probabilit o failur o reacto system(s subjecte t varioulevel o groun acceleration.1 3 4 Pea Probabilit o Probabilit o Probabilit ogroun a singl tw independen tw coupleacceleration reacto syste reacto system reacto systemgravitie failin failin failin0.0.0010.0.04 x 101.0.ID"3 x 1-3 x 13 x 1-Uncertaintfacto37 3 *Fo reactor designe t a SS o 0. g tabl take fro Reacto SafetStudy-28 probabilit o th couple failur otw system i give b th relationshiP x P couple * independen dependenti.e. th squar roo o th produco th entrie i column 2 an 3 oTabl 4 Th uncertaint factorresul fro th fac tha '.h inde probabilitie i colum 2 a uppe boun fo failuro tw systems W se tha th o couple damag i Tabl 4 differ b a facto o approximatell' betwee th 0.2- an 0.5- an b a facto o 1 betweeth 0.5- an 1.0- levels Thesspread wer take int accoun whe failur probabilitie fo event 9 throug 4 i Tabl 3 o th failur probabilitie i 3 wer take fro th Reacto Stud o wer estimate base informatio i Sec 3 o thi engineere safet system th auxiliar feedwate syste redundan an seismicall quali Th probabilit o failur othes system whe. subjecte t aearthquak i give b P . , th o failur o tw systemtha ar coupled Th powe conversio system thoug redundant i no qualified an it o failur whe subjectet a earthquak i conservativelassigne th probabilit o failur a singl systemTh earthquak frequencie fobasi event 4 5 an 6 an th failure probabilitie fo 9 throug 4 appl t reactortha ar locate i th easter U.S locate i th wester U.Sexperienc mor frequen earthquakeshoweve the ar designe t a largesaf shutdow earthquake I iassumed her a i th ReactoSafet Study tha thes tw effect an tha th failur dat i 3 ca b applie t alreacto site i th U.SCONCLUSION BASE O COMPUTE OUTPUFro th basi even data wcalculat wit th IMPORTANC2compute code ' th probabilit oth to even fo th faul tree iFigs an 5 I additio wcomput th probabilit o th toeven fo th righ sid o eacfaul tre t establis i earthquakcause event ar significan contributor t risk Th mea anmedia value an 90 confidenclimit o thes probabilitie arshow i Tabl 5 an Fig 6 W als th IMPORTANC compute cod t a sensitivit analysis anth compute outpu an conclusionbase o th outpu ar discusse idetai i Appendi B-29 Tabl 5 Mea value o ris o nuclea powe plan operatio fo variou casesCasAssociate faul tre Mea valu o risI To even probabilitwithou seismi scrasystemII even probabilitwit seismi scrasystemIII cause eventonly withou seismiscra systemIV Earthquak cause eveonly wit seismi st .systemFigFigRigh side Fig 4 Righ side Fig 5 3. x 14. x 1 x 14. x 11010110-- T I = T " 1 -6 I - I ICasIFig Media value o ris a 907. confidenc level fo th casepresente i Tabl 5O th basi o th dat presentei Tabl 5 an Fig 6 w se thaseismi tri system mak littldifferenc i th overal ris onuclea powe plan operation Thris du t earthquak cause eventi eithe cas i o th orde 1 , whic i greate tha reporte i thReacto Safet Study W canno sa thi case however tha ther ia statisticall significan differencbetwee ou result an th resulti th Reacto Safet Stud becauso th followin tw reasons• Ris i thi repor wa definei term o tw basi categories- cor meltdow accidentresultin fro earthquake othe causes-30 - los o emergenc servicedurin a earthquake• Engineerin judgemen wa usei man case t assig probabilitiet earthquak cause event i Tabl Th wid erro band o Fig 6 emphasiz th uncertaintie involvedAnalysi o th result presentei Appendi B sho tha cor meltFro th preceding w matabulat th ke point concernin us o seismi trip o nucleapowe plants Althoug no thi lis cover thmos importan issue tha wcovere hav a bearin o th o unadvisabilit oseismi tripsADVANTAGE O A SEISMI TRI SYSTEProbabl th traditiona reasofo requirin seismi trip onuclea reactor wa t allo movemen o thos component necessarfo shutdown e.g. controbefor the coul b affecte b th motion Presen reacto cor internal ar designe fo severlos o coolan acciden blowdowloads thi concer abou contro woul appea t b no a per today However othe activdown contribut significantl tth ris fro earthquak cause assume th consequenceo bot cor meltdown an los oemergenc service t b th samwhe definin ris i thi analysi(hence the wer o equaprobability the gav th samwhil i fac th consequence arcomponent ar involve i th shutdow o nuclea powe plants anthe als migh benefi i allowe tcomplet thei actio befor stronmotio arrives I appear tha a expensiv earl warninseismi networ i used fro zer t second o advance warnin tim al tha wil b available Thima allo enoug tim fo th controrods sto valves an maisteam-lin isolatio valve toperat bu no enoug fo othe componentsI addition fo sever earthquakewit duration o th orde o 3 sa earl shutdow wil minimiz th o seismi motio cycle thplan experience i th operatin thi way a seismi tri decreas risk Note howevertha nuclea plant ar designe tsafel shu dow durin earthquake5 Discussio-31 o suc severit tha the arexpecte t occu onl onc i manthousand o year (th saf shutdow th decreasi ris i probabl slightI a sever earthquak wer tcaus a accident i woul b aadvantag to hav th plan shu dowwhe th acciden occurred Th mosnotabl exampl woul b a earthquakcause los o coolan accident I plan wer shu dow a fe befor th accident som oth energ store i th fue rodwoul b remove befor blowdow anth resultin los o heat-transfecapability Thi woul reduc thtemperatur transien o th claddinan minimiz th chanc o claddinTh reacto coolan syste idesigne t withstan a ver severearthquak — agai th saf shut1dow earthquake Newmar estimatetha i twic th groun motio oa saf shutdow earthquak i experienced th probabilit o exceedin limit i 0.01 Thusaccordin t hi analysis thprobabilit o a earthquak causina los o coolan acciden i remoteI addition th Reacto Safet Studshowe tha earthquake wer nosignifican contributor t thprobabilit o a cor meltdowaccidentDISADVANTAGE O A SEISMI TRISYSTEA possibl disadvantag o seismitrip i tha the woul causunwante spuriou trips B us ospectrum-sensitiv detector anredundan channe logic electronispuriou trip ca probabl bsatisfactoril minimized Howeveri seismlcall activ areas unwantetrip fro smal earthquake coul ba problem particularl i th trileve i se low Lo tri levelar desirabl sinc the maximizth chanc o trippin a fe secondbefor th arriva o stron motionWha woul b require i a determinatio o a p.ppropriat tri leve foa give reacto site base o it history I thi way ia seismi tri syste wer employed trip coul b minimizewhil th advantage o th systeAnothe disadvantag o a seismitri syste I th effec i woulhav o a utility' electricageneratio an transmissio capabilitfollowin a earthquake Th suddelos o nuclea unit o a utilit migh caus a sever electricatransient resultin i prolongeoutages i th lesstringentl designe fossi unit wer of b th sam earthquakeWhil mos publi emergenc servic-32 coordinator anticipat powe outageo severa hour followin a sever outage o longe duratiocoul severel hampe recover effortsSinc i i no clea ho lon a nuclea plan woul remai dow aftea seismi trip thi trippin actiocoul produc more no les o a ris t th public Thi disadvantag o seismi trip woul requira hig trip-leve settin (t minimiztrip fro aftershocks an weldefine restar criteri (t allorapi restar whe saf t d so)If a seismi tri ideeme necessary existin seismiinstrumentatio coul b upgrade tprovid th tri capabilit withouincurrin a sever cost Instea oa single-frequenc acceleromete sensin th P wav o earlarrivals serie o seismi switcheo varyin frequenc sensitivitcoul b used Thes switche coulb use a selecte point o th powe plant' structur t th desig responsspectru o th structure Shoulth groun motio inpu approac threspons spectru o th structureappropriat actio ma b take awhateve leve deeme proper I ipossibl t us a respons spectrurecorde tha use discret reed foselecte frequencie fo thi purposeShoul thi seismi tri bincorporate int th plan safetnetwork th desig an installatiowoul hav t mee th criteri iBase o th argument developei Sec 3 an th ris assessmen oSec th genera statemen ca bmad tha seismi tri system ocommercia nuclea powe reactor hav a smal an somewha uncertai effec o plan safety Wheviewe fro th standpoin ii rist societ a a whole th seismitri system migh actuall increas ris b denyin a sourc o mucneede electri powe durin thtim o a earthquak emergencyI assessin th advisabilit oseismi tri systems w mus kee imin tha sinc nuclea powe plantar carefull designe t withstana earthquake the woul b expectet affec ris onl durin a versever earthquake Th result o a even woul b seriou i ancas an th nuclea powe plant'respons t th earthquak coul ad t o hel mitigat a seriou situation-336 Summar ANS N42. an b qualifie t2ANS 41.5 Th frequenc ospuriou trip coul b reduce busin 2/ channe logi o somsimila schemeI a seismi tri syste wer tb employed a ke questio woul bth determinatio o th tri levesetting A lo a possibl a settinwoul assur th maximu possibltim betwee tri an stron motioarriva (se Fig A- i Appendi A)I seismicall inactiv areassetting I th migh b t assur this O thothe han thi woul probabl b a lo tri leve i activ areas suc aCalifornia Her a highe tri levewoul b desirabl t minimiz trips Sinc nucleaplant i seismicall activ areaar designe t withstan large suc a Increas itri leve soul no resul ia increase risk Fo an trilevel advance warnin tim oa mos a fe second woul bexpected an suc a shor tim woulno hav a dramati effec o plansafetyAlthoug I ma b likel thaexistin instrumentatio o thoperato wil tri th plan durin damagin earthquake ther i nassuranc o this an i an eventh tti woul occu durin th stronmotio n benefi o advance I seismi tri i deemedesirable a specificall designeinstrumentatio syste fo thipurpos woul b preferable-34 7 Reference1 Tokarz M Zaslawsky an (1 St Leger-Barter of the Jse f .* '." ' * .'-'ran yur'sms for : ov ? p ac' rs, Lawrenc LivermorLaboratory Rept i)CRL-5l'61 (1974)2 :•• r ."•••.v.� .'-' A,� Seis':~'c arid c,- Criteria �;\y ,','. • • . / '.v . .• : ji.r (10CKR100) U.S Atomi Energ Commissio(Kedura KcKl.st.er Washington D.C. Apri 30 1975)3� . ; -r. f- Nuclea Regulator Commission WashingtonU.C. guid 1.1 (1974)4 :••� '•'•'»•-' f r Nuclear !'Mev Plants, AmericaNationa Standard Institute NY Standar N18. (1974)5 ''••!• f :• :•'' '!'•'• Processing, and Evaluation of Records Cl ;'.' " .' '."•,"' instrumentation, America Nuclea Societv proposestandar 2 (1976)6 ,•','• ." ..'ualifi� 'atioi of Class 1 Electric Equipment for A' • ir •7 '••.;• .'•'*-: America Nationa Standard Institute standarN41. (1971J7 Genera Electri Co. Oreg. Trojan Nuclear Plant, Final Safety Analysis Report, Vol. 7, docket-50344-4 (Feb 58 General for Electrical Relays, militar specificatioM1L-R-5757 (Ma 8 1970)9 H Shaw AE Divisio o Reacto Technology privat communicatio tJ Ramey AE Commissione (Jan 1970)10 H Thoma e al. Nuclear Reactors and Earthquakes, AE Div o TechInfo. Rept TJD-702 (1963)11 Curran Argonn Nationa Laboratory privat communicatio (Nov 1975)12 o Nava Reactors U.S Energ Researc an DevelopmenAdministration Preliminary Safety Analysis Report, Light Water Breeder (Shippin Por reactor (1975)13 Powe Co. Charlotte N.C. Preliminary Safety Analysis Report, Oconee Pressurized Water Reactor, docke 50-26 (1968)14 M Newmark Probability of Predicted Seismic Damage in Relation to Nuclear Reactor Facility Design (Draft) Consultin Engineerin ServicesUrbana Illinois prepare fo th Offic o Nuclea Regulator ResearchNR (1975)15 E Siebe "Nuclea Industr - Approac t Performanc Assurance, IProc. Reliability and Maintainability Symposium (IEEE NY. 1976) p 430-35 16 L Scott Nuclear Safety 1717 Safety Study - An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants., U.S Nuclea Regulator CommissionRept 14018 Well an J Savage Fault Tree Analysis of the Emergency Power System of Two Light-Water Nuclear Power Relators, Lawrenc LivermorLaboratory Rept UCID-166619 L Mulloy San Fernando Earthquake of February 9, 1971 - Effects on Power System Operation and Facilities, Dept o Wate an Powe o thCit o Lo Angele20 L Croo an G S Vissing Nuclear Safety 1621 Lambert Fault Trees for Decision Making in Systems Analysis, Lawrenc Livermor Laboratory Rept UCRL-518222 for Protection Systems for Nuclear Power Generating Stations, America Nationa Standard Institute standar N42. (IEE 27923 Class I Electric Equipment for Nuclear Power Generating Stations, America Nationa Standard Institute standar 41. (IEE 323(1971)FW/gw/vt/g-36 Appendi A Tim Interva AnalysiI sufficien tim i availabl betwee a tri signa an th arrivao stron motion th advantag o a seismi tri syste i increased A discussio o "sufficien time i include unde "Variable Affectin thFrequenc o Seismi Trips I thi renort Othe method ca b use tAenhanc thi tim Interval Th objectiv o thi appendi i tdetermine fro Inheren ground-wav characteristics wha th tim intervawil b betwee a tri signa an th arriva o stron motion T accomplisthi task w collecte an analyze 2 accelerograms Thes stron motioaccelerogram ar liste i Tabl A-l ' ' ' Al th accelerogramanalyze wer fro site i th wester U.S Thi i du t th lac oaccelerogram fro th easter portio o th U.S Th 2 accelerogramliste i Tabl A- ar fro 1 differen earthquakes A analysi waperforme t determin i th dat appeare biase becaus o thi fact Thanalysi resul usin onl 1 accelerogra records on fro eac earthquakev th 2 liste i Tabl A- di no diffe significantlyPrio t determinin th tim interva i question w develope a relationshi betwee rang t epicente an tim t stron motion Sinc thdefinitio o stron motio i no fixed th analysi employe severa values th accelerogram analyze i thi stud ar take fro th tim perio193 t 1971 th accelerograph use t recor thes accelerogram wer noal identical an th startin characteristic o eac ar a littl differentHowever i ca b sai that i general thes accelerogram wer triggere approximatel 0.0 g Not als tha th interva betwee th arriva o motio an th tim th accelerograp actuall start i 0. t 0. sFigur A- illustrate th relationshi Vatwee rang t epicente antim t stron motio whe stron motio i define a 0. g Th figurgive th equatio fo a linea fi throug th dat point shown Severa othe tha linea wer tried bu th linea relationshi providea "best fi an seem logica o physica grounds Th correlatiocoefficien associate wit tha regressio i 0.82 Figur A- als illustrate thi relationship bu fo th cas whe stron motio i definea 0.1 gFigure A- an A- sho tha th tim interva o interes increasea th rang increases Thi relationshi reflect th fac tha P an S wave-37 Tabl A-1 Pertinen salien characteristic o accelerogramsAccelerograNoEarthquake datRichteRecordin statio magnitudGrounmotio1 E Centro Ma 18 1942 Northwes Calif. Oct 7 1953 Ker County Jul 21 1954 Ker County Jul 21 1955 Eureka Dec 21 1956 Eureka Dec 21 1957 Sa Jose Sept 4 1958 Sa Francisco Marc 22 1959 Sa Francisco Marc 22 1951 Hollister Apri 8 196 Borreg Mountain Apri 8 1961 Lyth Creek Sept 12 1971 Sa Fernando Feb 9 1971 Lytl Crrek Sept 12 1971 Lytl Creek Sept 12 1971 Parkfield Jun 27 1961 Sa Fernando Feb 9 197 Parkfield Jun 27 196 Lon Beach Marc 10 193E Centro, C7.S W Ferndale6.S44N46Sant Barbara C7.N42Taft7.N21Eureka6.NllN79Ferndale6.N44Sa Jose C5.N59Golde Gat Park S.F5.N10S80S.F Stat bldg. S.F5.N09Hollister C5.S01N89E Centro C6.S W Sa BernardinHal o Records C5.N Or Blvd. L.A6.N W Wrightwood C5.S65Devil Canyon C5.S E Cholame-Shandon C5.N65N25Pacoim Da C6.S74S16Cholame-Shandon C5.N05N85Vernon6.S08N82-38 Tabl A-1 (ContinuedAccelerograNoEarthquake datRichte GrounRecordin statio magnitud motio2 E Centro Dec ^0 1932 Helena Oct 31 1932 Ferndale Sept 11 1932 Olympia Apri 13 1942 Oly.npia Apri 29 196E Centro C67S W Helena6.S W Ferndale5.N45S45Olympia W7.N04Olympia W6.S0476I 5 a 4 E 3 21 0 1 1 1 1 1 Linea regressio equation1 1 -°-l1 -y = -0.2 +— -o --o ^— — O -o ° ^° ^r° ° o°1 1 1 1 1 2 3 4 5 6 7 9 10Rang t epicente - kFig Tim interva v rang t epicente (stron motio = 0. g)-39 7 0 2 3 4 5 6 7 8 9 10RangFig Tim interva v rang t epicente (stron motio = 0.1 g)trave a differen velocities Therefore a Newmar an Rosenbluet havstated th tim betwee th arriva o th P an S wave i variabl a a Afunctio o distanc t epicenter Thi tim interva i use b thseismologis t locat th foc o earthquakes considerin th typ o soian roc th wave mus trave through Th interva help i ou casesinc th slowe S wave usuaii caus th stron motion Th dat scattei an A- ca b attribute t th varie typ o soi througwhic th wave ha t trave t reac th accelerographsTabl A- list th dat point tha wer derive fro th selecte 2accelerograms Th dat point wit stron motio define a 0.25 0.3 an0.3 g ar ver dependen o eac othe becaus the cam fro th samaccelerogramsOn proble tha ha bee discusse elsewher i thi repor i thproble o determinin a wha leve o acceleratio th seismi tri systeshoul b set Whil th fina decisio a t wha acceleratio thi shoulb involve man differen factors Fig A- help t provid a feelin foth relationship involve betwee th leve th syste i se a an th tim-40 9 5 4 4 tha elapse betwee suc acceleratio an stron motion Not tha Fig A-doe no tak int accoun th tim betwee receivin th signa an startinth accelerograph Figur A- als assume tha th 2 accelerogram use ithi analys representativ o wha a typica reacto sit mighexperienceT asses th likelihoo o exceedin a specifie tim interval w neea relationshi betwee distanc an pea groun acceleration Mickey't Acorrelatio an th Esteva-Rosenbluet equatio ar tw sucrelagionshipri Anothe i th relationshi presente b Matthiesen HowardAan Smith Fo th purpos o thi analysis th Esteva-Rosenbluetequatio sh j i Fig A- wil b used Thi equatio is0.77\whera = p i groun acceleratio (g)M = Richte magnitud o th earthquakeR = hypocentra distanc (mi)Stron motioy exp(-24.88xStron motioy exp(-39.66x,o 1 1 l^-t I T I 0.0 0.0 0.0 0.0 0.0Acceleratio tri leveFig Relationshipbetwee tri leve antim t stron motion-42 1.1.21.0.0.0.0.0.1—i—|—i—|—Richtemagnitud^ 8 X _Fig Estimate relationship betweeacceleration distancean magnitude.* Foca i 1 mi2 6 8 10Epicentra distanc12Th hypocentra distanc f relate th foca dept h an th epicentradistanc R b th followin equation«&T (A-2Th foca dept h wa assume t b 1 m i ou analysis Thi i a typicafoca dept o earthquake alon th wes coast Th relationship w neerelatin R t tim interva ar give i an A- fo stron motiodefine a 0.1 g an 0.1 g respectively Th Richte magnitude M waArelate t th numbe o earthquake pe yea pe squar mile N b usin : log = -1.7 - 0.7 (A-3-43 Not tha Eq (A-3 i referre t i Fig A-5 Equation (A-l an(A-2 wer substitute int Eq th numbe o earthquake peyea pe squar mile N-1.7 -N - 1a(R 2250.77(A-4Fo th cas o stron motio define a 0.1 g th followin equation baseo th 2 dat point o Tabl A- ar appropriatea - 0.1 g (A-5T - 1.10.0miwher T i th tim interva i secondsTh numbe o earthquake pe yea pe squar mil wa converte t a relativ probabilit tha ha bee weighte b th circumferenc o a circlwit a radiu equa t th epicentra distance R Sinc R i a functio oth tim interva T a relationshi betwee T an relativ probabilit o T coul b obtained Thi relationshi i illustrate i Fig A-6Afte performin thi analysi o earthquak data w ca mak thfollowin generalization Base o th dat i Tabl A- an th relationshipprovide i an A-2 th expecte interva betwee th startin oth accelerograp recor an th arriva o th stron motio i see t b itn rang o thre t fiv seconds Figur A- show th distributio othes tim intervals-44 " 1 1 i r Lo N = -0.66 -Lo N = -1.7 - 0.7 (1934-1965Lo N = -2.8 -Lo N = -5.0 + 0.0659M(1904-1965a Dat fro 1904-196A fro 1934-1966. 6. 7.Richte magnitudFig Th seismicit o th Wester U.8.A-45 I l.• ' I I I ' I 9.— \ -7.-�5.-I ^V*3.-1i . I i I I I I 1. 3. 4. 5. 6.Tim interva s Fig Relativ probabilit valuefo variou tim interval betweeseismi tri a motio0.1REFERENCE*! G St Evaluation the Use of Seismic Scram Systems for Power Reactors, Lawrenc Livermor LaboratoryReptA2 Dalai Probabilistic Seismic Exposure and Structural Risk Evaluation, Ph.D Stanfor University Pal AltoA3 Earthquake Engineering, (Prentice-Hall Inc. EnglewooCliffs, 83A4 L C A Vo Earthquake History the States, Departmen Environmenta Dat Services Boulder ColoradoPublicatioA5 Motion Earthquake Accelerograms, Californi InstitutPasadena EERA6 Rosenblueth Fundamentals Engineering (Prentice-Hall Inc. Englewoo Cliffs 84A7 G E the Probability Distribution of Peak Ground Acceleration in the U.S. Continent Due to Earthquakes, Universit Lo Rept UCLAA8 C B Seismic Safety PlantsAppendix IV, Public Health Risks of Thermal Power Plants, Universit oCaliforni Lo Rept UCLA-ENG-724-46 Appendi B Outpu an Conclusion fro IMPORTANC Compute CodW use th probabilitie o basi even failure a inpu t thIMPORTANCE compute cod t calculat th probabilit o th to even i thfaul tree i an 5 W als calculate th probabilisti importanco th basi event an o mi cu set i bot faul trees ( cu se i a se o basi event whos occurrenc cause th to even t occur a mi cu i a cu se tha canno b reduce an stil ensur occurrenc o th toevent. I additio w comput th righ han sid o eac faul tre t establis i earthquak cause event contribut significantl t th tota riskProbabilisti importanc fo a basi even i define a th probabilitthj a cu se containin th basi even occurs, conditione o thprobabilit o th to event Th importanc o a mi cu se i define ath probabilit tha a mi cu se occurs conditione o th probabilit oth to event I th referenc give below thes importanc measure arcalle th Fussell-Vesel measure o probabilisti importance.Th IMPORTANC compute cod wa use i computin thes importancmeasures outpu i show i Table B-l B-2 B-3 an B-4 I eactabl th probabilit o th to event basi even importances an cu seimportanc ar shown I th firs colum o eac table th mea probabilit give probabilit whe al basi event ar calculate a theiexpecte additio 98% 90% an 80 confidenc interval argiven colum labele 50 give th media probability A lo normadistributio wa use fo al basi event i expressin th uncertaint ith failur probabilities Examinatio o thes table sho tha fo thprobabilit o th to event th media valu exceed th mea value Th i tru fo th importanc values sinc the ar th ratio o wit th probabilit o th to even appearin i thTh media an 90 confidenc level fo th probabilit o th toeven i Table B-l B-2 B- an B- ar plotte i Fig 6 o Se 4 Widuncertaint band fo eac probabilit ar du t th uncertaintie i thfailure-rat data*See E Lambert Fault Trees for Decision Making in Systems Analysis, Lawrenc Livermor Laboratory Rept UCRL-5182-47 Becaus o th vid overla o thes uncertaint bands w canno sawit a hig degre o statistica confidenc tha seismi scra system makan differenc i th ris o nuclea powe plan operationI Tabl B- (fo plant wit seismi tri se tha thmea probabilit o th to even i 4. x 1 , whic i slightl greatetha th mea probabilit o 3. x 1 give fo plant withou seismi trisystem i Tabl B-3 Thi differenc i du i par t basi even 4(seismi scra syste trip nuclea powe i containe i thfaul tre i Fig 4 bu no i Fig 5 W se i Tabl B- tha basi even4 contribute abou 6% o th tota probability henc w conclud trippinth powe plan a th tim o th earthquak increase no decrease thtota risk I additio w se fro Tabl B- tha mi cu se No 4 basieven 7 an basi even 8 contribute 7% o th tota probability Fro observation w conclud tha spuriou scram du t seismi scra als ten t increas th tota risk However a stateabove difference ma no b o statistica significance Als i wcompar th probabilit o th to even i Tabl B-l 4. x 1 , t tha iTabl B-2 4, x 1 , w se i thi analysi earthquak cause eventcontribut approximatel 10 o th tota risk Th sam conclusio i comparin Table B- an B-4 Th probabilitie o th earthquakcause event ar approximatel a facto o te greate tha earthquakcause event tha lea t cor mel cite i fh Reacto Safet Study A i Sec 4 o thi report thi increas i du t th inclusio o los emergenc service I th definitio o ris i thi report whil i ino include i th Reacto Safet Study-48 E B 7 7E 3 E 6 2 4E 4 E 2 2 7E 5 2B 31 17 16 BJ 27 33 44 45 CB «7 91 62 14 14 63 19 53 23 1' 19 1 3 1I 3B 13 11 78 32 1C 48 29 29 22 95 11 42 24 15 33 97 64 t3 19 30 1 TABL B- (cont'd.RKKEitL-ii. IA. rw tu ri.CU HBB1BB18B2BB3BB4BB4BB3BB4RB4HZB2BBrOB2BB=B4BB4B49 BB4B4BF.'4B4BB4B«BBAS B4BBAS B4BBAS B4BBAS B4BBAS B4BBAS B4BQAS B4BB3B4BB3B4BB3B4B 3B3f-4BB3B4BB3B4BB3B3BB3B3BS3B3BB3BJBB3B4BB3BE 4BB3B4BB3B4BB3B4BBBB-51 6 1 4 3 2 4 3 1 3 6 1 7 4 2 ^ 7 2 7 4 7 4 1 1 3 4 1 1 1 2 3 4 9 2 6 2 6 3 3 4 1 2 3 3 2 4 7 4 7 2 7 3 1 4 4 6 5 3 4 C 4 3 1 3 4 6 4 3 4 3 2 3 1 6 2 3 3 3 9 3 5 7 2 4 6 6 9 4 2 4 3 5 2 8 3 3 4 5 1 8 7 9 1 6 1 9 4 6 1 TABL 9-: (cont'd.r MEA 1 5 10 50 99 99(9B (90 (SO IBO (90 (99I.S40E-0 28SF-0 2 941 0 9 02IE-0 2.033E-0 3.487E-0 6.997E-0 9.190E-01 !9If-0 7 OPOfiO 4 0&3E-0 1 116E-0 3 245E-0 2 173E-0 2.967E-0 4.829E-0rift'.E-O rt a^Dt-O 3 Dizr.-o l 168E-0 2 936E-0 2 293E-0 3.B66E-0 3.203E-07.7-1? O A 0O7E0 6 JP&E0 ?. P3GE-0 2 I41E-J 3 6B2E-0 0.721E-O a 441E-03 rf7«jr-OP ? Tiac-O 1 647E-0 4.089E-0 4 B80E-0S.4701--O 1 B23E-0 893E-0 270E-0 2.17BE-0 4 283E-00.479 ( 1 JJ18E-0 9 73*.E-0 2 91ftE-0 1 076E*O 6.124E-05.4791-0 I 35DL-0 9 3"-2 • 0 3 lOP'i-O 1 112E-0 : 8ZSE-0 3.270E-0 B.704E-04,6Jr. 0 1 51IE-O 8 0O2C-O 2 G21E-0 7 709E-0 2 0O4E-0 3 296E-0 6 966E-03 612"0 I 65GE-0 9.739E-0 3 '.d«- 1 003E-0 8BB9E-0 57X0 1 3e0F.-0 9.690E-O 2 950E"0 8 80B • 0 7 224E-0 3B2C-0 1 377E-0 6.5f!/E-0 2 P05E-0 3 73BE-0 4 963E-0 B.137E-02.382E-C 1 090E-0 7 745E-0 2.443E-0 4.92IE-0 3.496E-0 8.347E-0 2.172E-0 O.iOE-0 6 "58E-0 4 260 J 2 099E-0 4,0376-0 7.843E-02?n7£-r 7 059E-0 B I99E-0 1 853E-0 3.7I6E-0 4 982E-0 8.40BE'0 2 033E-02 M7C-0 I 935E-0 4 898E-0 3 319E-0 9.238E-0 I 90BE-07 6^"3 0 2 9I4E-0 1 608E-0 I 742E-0 3.946E-0 9 B50E•07 f.?3 0 3 474E-0 2,21'E-0 C 300E-0 1 880E-O 1 855E-0 2.999E-0 7.688E-0 0 2 7G3E-0 I 986E-0 3 974.E-0 1 fiSBE-OS 1 5B3E-0 2 701E-0 5.23BE-01.90CE-0 SJ8E-0 5 200E-0 1 672E-0 4 712E-0 4.486E-0 6 99SE-04.764E-0 1 397E-0 2.170E-0 S557E-0 1 003E0 9 6I0E-O 1 433E-0 2 495E-03 2"*BE0 1 31JE-B 8 42PE-0 2 605E-0 6 689E-0 9 743E-0 1 896E-0 fi 226E 03 280E-0 7 2^flE-0 S 72BE-0 1 853E-0 6 043E-0 1 3IOE-0 3.311E-0 B S64E•03 014 0 7 639E-0 3 31SE-0 1 S73E-0 5 6B0E-0 9 648E-0 2 1 B - 0 S.467E-Q3 014E-0 3 B62E-0 4 80BE-C 1 260E-0 6 346E-0 1 430E-0 2 472E-0 fi 228E-03 014C-0 4 933FI0 5 5M Qf 1 554E0 3 .760E-0 1 080E-0 2.418E-O 6.9B7E-0 7.10 0 3 159E-0 3 107E-0 9 9ME0 J 241E-0 1 G13E-0 3.688E-0 1 287E-Cl.C0fct-O 3� ~9E-0 2 J92E 0� P36E-0 3 373E-0 0 834E-0 a G57E•0 9.375E-01 CC 0 T 135 0 Z 547 0 I OP'H •05 3 102E0 7 923C-0 1 e43E-0 1 113E-02 3V)E-tl 1 76U0 2 784C-0 6 919E-0 4 809E-0 1 956t-0 3 937E-0 8.842E-01 11I 0 2 713E-0 ? 3J2F0 G 1GOE 0 2 056E-0 G 201E-0 1 213E-0 5 365E-0A O'SL-0 1 184E-0 9 27OE-0 2 431E-Q 6 079E-0 2 149E-0 5.982E-0 3 466E-03 -SP3E-0 2 ^OOE-0 7 5011-0 3 S59 • 0 2 334E-0 9 B63E-0 3.0B3E-02 t £??E-0 1 0?JE-0 A 334E-0 d.flt.Of-0 3 2J7E-0 7 992E-0 3 JB3E-02.rv.,*.l-n 1 437E-0 3 381E-0 a 412C-0 3 514t-0 1 S05E-0 4 308E-0 1 805E-0 3P3 T 2 D64E-0 1 763E-0 7 3G1E-0 3 981E-0 ? 183 0 7 052E-0 3 9D3E-01 277E-0 6 9-IOE-0 8 2ISE-0 2 2rt4E-0 ? «40E-0 1 74BE-0 3 366E-02^. 3 213E-0 I 890E-0 5 CB3E0 3 372E-0 3 23l£-0 I 103E-0 1l01£-02 723E-0 2 49o - 3 6 75IE-0 A 23SE-0 2 172E-0 3 401E-0 6 39BE-02 1 I66E-0 7 5G4E-0 1 63S - 0 4 036E-0 5 BHE-0 361E-Q9 S9I C P ,3-'SE-0 1 701 0 4 93BC-0 1 4I2E-0 4.414E-0 1 049E-0 « 36BE-0 30^-0 f 3C2E-1 » 059E-0 4 736E-0 1 ISIE-Ot 1 477E-0 A 14BE-0 3.393E*0-53 0 • 2 91 B3 11 99 24 B4 13 23 33 62 14 22 14 14 1J 36 16 66 42 19 3? 39 26 16 5 1 1I 94 12 17 1 E 4 3 e4 24 13 11 •1 •2 12 62 1E 1? I4 94 1 2 11 22 9S 21 2S i9 3i 93 fi.243E-07 3 13 26 22 96 S3 16 22 14 16 16 14 16 33 53 3 REFERENC MI CU SETCU SE NO ORDE BASI EVENTTABL B- (cont'd.SS£ IE 1BIB3BIB3B1B3B1B3«1B3S1B3B2B3B2B3BzB3BB3B2B3B2B3B2B3B2B3B2B3B2B3a2B3B2B3B2B3B2B3B2B3B2B3B2B3B2B3ilda «2 44::E 33 3 4B 4Ei 4 4 )B 3B 3 3»B 4B 4 3 4 4 4B 4• 4B 4 4 4B 4 4 -1 4 4 4B 4 4BF 4B 4B 4 4B 4 43 4 4 4B 4B 4t»& 4-60