ML16340A573

Draft Safety Evaluation Re Amend 52 to License Application. Risk to Public from Seismically Induced Reactor Accident Is Lower than Overall Risk to Licensed Plants Discussed in WASH-1400

ML16340A573
Person / Time
Site:	Diablo Canyon
Issue date:	12/30/1977
From:	NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:
Shared Package
ML16340A571	List: ML16340A570 ML16340A572 ML16340A573 ML16340A574
References
NUDOCS 7909280249
Download: ML16340A573 (24)

Text

t

~

~

)>TUBA)"I'1.2/30/77 I':v;rlua tion o I')ncndrrrcnt

.">2 I I()sjtI'i Scl SrA)C );va1 lillt.'l ()Il IJnit>>

1. and 2, ))i;ibln Crirrynrr !)i tc Voltunc V Tntroduct.i.nn The'isk to t)ic piibli.c from t)ic Diablo Car)yon plant to potential cart)iquakcs i.s c)rarrctcrizcd by t1ie appl.icant:in Airrcrr()rircrrt S2 to thc application in a sct o). complementary crumHativc distribiit::ion.I'iinctions (ccdF)

I'nr radiatinri ()oscs at various locations.

')'Iicsc cc()f'.s;irc generated Crom a T)ostirl:rtc() cc(1)nr seismic accclcration, cvciit/Fault trees to r

I t".>t.'Q

~'I ~

reprcseitt tlic systciu response, conditional probal)il i I'ics nC. cniuponent Failtrre as;i Ciu)etio>> nF scisiuic accc1cration and p'rnhabilistic modeling oC atmos))hc~ ic ()ispcrsion.

)"urthcr inFonrrtion on t)ic ai)alysis pcrlo ~ncd is incorporat:c(l in a lct:ter From tlie a))plic;int: d;itcd October 1)), 1977, which responds to nr;rl qrrcst'ions raised in discussion wit)i t)ic app'licant.

Wc

') (sl

~ a S

t I

S

)rave reviewed t)ic;ipplicant's analysi.s oC t)rc pl;ntt it)i tliout: t)ic ")Iosgri modification";rs

() i scussed below.

Prol>of>H>.ty pl'o i srI>i c Rccol proc ion This review is b;ised o>> the corri))lerner>fary cimrrrl:itive distribution Cru)ctior) Fnr scisiriic:ic(clci;it;ions prcsciitc(1 by t)ic: r)1)licant in Arrrendment No.

52.

'I'hc Pro)):i))i') istic Ar(rlysis StaFF )ias>>nt: cviliiatc(1 t)ic acceptability oI.'liis distril)rrt inri Iiuictioii.

'I'Iic rcsiilts ol')ic 'ii)))licant's;rnalysis indicates tliat t)i(: risk to tlic piib)ic associ itcd t~it)i sci.smic events at the Diablo Canyoi) si tc i.s (lnmi.natcd I)y.>cisrrric accelerations in t)ic range oF 0.8S to,l.S!~ an(l tlic a isk is 1 Iric;rr ry propnit;iori;il to t)ic prnbal)ility assigiicd to t)iat r'rr)!lc.

)'I t)ic (orir!))crircnta).. ciuriul'itive ()isti i))iition Frurction for Y9093803-'f t

scis)ni(: ')ecole)')talons w(.) c to dc(:rcaso i)ioi'c z')I):i(lly as i.Ci)>>etio)z of: acceleratio>>

th;)n pz'cdi.ctcd hy thc ()ppl i.c;int, tlic ovcz';)11 risl;;)ssociatcd witli seis)))ic events

))zd thc dou)inancc ol'. thc 0.85 to '1..5g accclcr'itinn range would also deere isc.

('.O>>vc) scly,,i,l'lic distzil)()tion I'<)>>(..tio>> werc to decrease less rapidly tlic over;)jl risk wo()ld also i>>crease

'zlthoi)gh thc (ln)))i>>;)nce ol'he 0.85g to

1. Sg ')ccc" c)';)t i nn z ange would>>ot ch:)iq!c;)I)prc(zi il)Iy.

L'vent Trees

'1'lic:)pplic;)>>I: s cv<<>>t t')<<cs depict thc sigiii.l I(.;)>>t systc)))-to-system i)ztcractzons pzcdictcd Ior tl)c l)i:)l)lo (.';)>>vo>> p'I;i)it.

'I'lic cv(;)z(

) )(.cs )ised l)y tlic applicant Ior D).;)bio ('.;inyo>> a)'c esse>>ti;)l,ly:idc>>ti.c;)'I to tli()sc d<<ve'loped I;or a diFferent rcilctol'>> (lic 14.')('.tot S')Ict)'t))dy (I)th,)II-'.10(l(I).

)

I l'hoilgll cert 1) ll systc)ll-to-system dope>>dc>><<ics:)t l)i:)l)lo (;;>>zyo)z di.l I'c)

I'ioui those Cou>>d

.i)n tlic 1(c')ctor S)Fcty Study, tlicsc depend<<>>(..ics appc;ir to Ii;ivc b<<<<>>;)etc(Iu;)tcly treated when cou)binii)g tl)( )(.s>>ILS OInuit t)cc;)>>;).I>sci

(

) d((c)')))inc tlic conditio)zal probabil,i,ties ol'Iic accidc)it, sc(li)cn('cs.

DO)1)lllilnt Co)itl' bi)('ol'S to Ih s k Tlie )pplicalzt s ill)ilIyscs ilssi'uilc

( hiit tlic sw i t(. hy;) z'>ds on tho av:)inability oF a.c.

power, thc I;ii'Ii>>c pzoh;)1)i lit> oi

= these systems;il)p) oa(zhcs zuzity Cor acce~eratio>>'bove 0.5g.

IICC))z)sc tlic design incog)oz:)tes;)

stc;u))-tu) bine-dz iven a()xiii )ry fc(

wiiter ~n.u))I) (which docs no(; dcpciid oii a.c.

1)ower), tl)c;)()x.ili;)ry Cccdwatcr systeil) is pz'cd icted to have hi gli 1 ikcl I b.)od oI.

Cx)>>ct lo>> i))i!:it;)ccclcrations ul) to

'O)g') ill)ovc 0

)<~! ('1\\<>i)lit )'il>>J',c ol ac<<lcz")(:I()') (0.85-I.SIl),

))1 I cngi>>cored;I'cat)>>cs arc

)

~

~

P

predictc<l to have;i I'ailiirc rate of )city.

Si>>co tiic co>>di.tio>>al prob;ibi.l.ity of I;iilure is;ilrcady uixity, conunon cause F;)i:I.urcs c;)nnot, increase flic)'c I 0)c>

w<'. hllvc>>otIlo'I'oil!>1)ly'i>>vcs'I'ig(1 t'(i(I I'l(! ()d('(Ill(lc}'I:.t'hc.

applica>>t's trcahnc>>t oF con)Inon cause Cail~nos bctwccii systems.

At lower accolor;itions, tiic 1))c(Iictc<I l.:ikcliiioo<1 of tlic doini>>;iiit accidciit soqucncos

)1 (TIE,B and S Cl}I')

wo()ld I')avc to incroasc by 2 to

) or(lors oC m;ignitudc over that prcdictcd by tlio ))1)pl:icaat: to I)ave a si!~>i Cic'int cf'f'oct.

IVC consider it tv)likely that an error oC this n)ay)it>>dc.is prcsciiI:.

According1y, we believe i.t i.s rc;iso>>abl'c to oxa)oct ti)at thc impact of i>>iy t)>>i<le>>tified common cause(I f':)'i.7(>>cs woiil<1 b<

s)i):i11. o>> tI)c i isk:)sscssni(.))I The conipoiicnt I'ai",lure d;)t() (ised i>> (Iiiiiiti.fyiiig>tlio I-;)t)lt trees is

", derived by <<sstui)in!, tliat t;Iio f:ailure rate bc'low a speci,lied stress or "pivot point" ()'I<.'ld stre...(

I(>>'o!i!i)o>><'>>t's:)>>>I tlic (.o<1<.;)1'iowablo stress for faulted co>>dit.ioiis Ior I)ipi>>I>) is;)ss(unc(I to bo th(

i(>>)<loin.C;iiltire rate lor simil:)r co!npo>>o>>t.s prcscntcd in tho Itoacto) Safety Study (IVA!>!.I-1000).

'I'ho I'a'i 1(iro prob;ibility is asstu)ic<1 b)'hc;ippl icant to increase linearly witii scisn)ic;)c(.< Icr;itio>> above.Iliis pvi>>t, rc;icliing a va1ue of: unity at i st')css oqiiiv(i'le)it t:o tlic <<lt.i)I<)to. t)~ss I'or components c'ln(1 to twico I'))c >." "'. "," 1"'I" >1 ". ': 'r'"" ""r I:!')I tc

>'..Il(litioiis I'or pipi>>g.

'I'Iic stress a>>;ilyscs a)c 1)ascd on Anic>>(hnciit

.">0 t:o tlio applicat"ion, wiiich we I)ave not reviewed.

Assiuning tiicy wcro adcqiiatcly pc) Co)>no<1, in our opinion, the assumnt:ion of':i I i>>c;>>

)'<'l;)tionship:)ppc:).'s to bc conservative.

'1111.11 i'oh:I's t.<<()>>;)(<;i<i('>>t s(>I<<;>>ec i>>i I:i;itc>I I))> a)1 (:arthq()ako-inducod tr;>>isi(.>>L I'ollowc(I by I'.iili>>o <)I tiic p(n, i cu>>version

system,

~ tho )(aH i:)r) f(ic(lwato) systcin i)i(I citlicr tii(. C()iitai)viicnt spr;lv

.injoctio)> systcn), or tho co>>ti)i>>ni(!i)t spr;))'cci)'(.'.<<1;)tion systoin.

S CI)1: rcCors to;i soi.sinic;i'I'ly-i:i(I<<c<)d sin;i'll loss oI cooliint accident

'411owcd by I;iiiuic of'hc ('01)I.',)l>>)lie)ll; spr;ly l>>j((.t'.io>> s}'sto)n thc c)nor!>Cncy co'i'c cool in!> in jccti ))) systcni,;)>>

I tiic co>>t;iii)in('.>>t'))ra) rcci')'cilla L.ioil svs I:cni.

I LI

l'or compone>>t;s whi(.h werc (Iua'I:i.Cicd by seisllli( testing rather t)ran by stress>>1;llys.is, i t w:Is 1sslllllcd that: t)lc I'ai.11>>'e plobabi1ity was unity for acccl.erations greater th;ln the tested v;I.luc.

')'his;lpproach base(1 on testing appears to bc c(lnselv;lt.ivc and, whc>> app.l icd tn tll; diesel. gc>>crators, may result in sollle overestimation of thc signi,fic;I>>ce of'igher levels of accelcl'litic>>.

Bcctluse tile predicted system C Iiltlre rtltcs;Ire dolllinated by the loss o: a.c.

powel, the.C>>sitivity analyses perConllcd lly thc appl.icant indicate C

.that thc calcu'I;ltcd relc;lse c;Itcgory co>>di t'ion;I.I pl oh:lb):).it'.Cs lre rel ltively:insc>>si tive to thc tlssulIIptio>>s Ic)!;Il(li>>I!thc variation of'olllpo>>ent I'll 1111'e I'lte lls 'I ClInctlon 0 I: 'Ie('0 I c1'a t i (>>>.

Ilot(ever,,as noted above, il: thc diesel gc>>er;ltors are clip;Ible of Ctlnctio>>i>>g;Ibove their test accclcr ltion, the system I'ail>>re rates and;1ssoei:lted role;lse category prob;lbi') itics wi I] be 1nwcI'll;>>I those ctllcu'I;It'c,l l>y thc;lpplioant in the accol 01'alt"Io>> rtl>>I!c l)o I'wce>> 0. SJ,';>>ld I'h<.'l('( (. Iilt: I All at wh i ch I';1ill>>e might occur.

I I the;Ictu;I.'I, I'a)1>>rc poi>>t werc;Ibov>> 0.85)!, the;Ippiictlnt's cI]cul;It'c(l pI'ob:lb)1 itics ot'clc;lsc (.;Itegol ies i>> tll;It: r lnl!c ol'ccclcration wollld bc co>>s('I'v:It'ive.

(I>> t'his h;Isis,

.i t:;Ipp(.;II's I'll;lt. the;lssltmptions made i>> thc Inct'ho(IIIlog> llse(I to (lctelallinc t.:lle (:olllpo>>c>>t:lilulc d;lt';I either dn>>ot st I'o>>I!Iy Il I I ( 1'I I h('

(IIII'll> lulls ut I III st Il(l( ul" 'l I Rcl ctlsc

(.II LCI>()I"I ('s

('>>ll'>e I"vcI 'I: I V(,'he;Ippl.l.~

.>>'. '""v.-.".');e:" c'I.<<cnt se;lucnees

.Into release categories.

Tn so (loi>>g, it was assttmc(l that the accident scllul.nces would restllt in:Itnlosphcri(1 rc'le;lscs.i(lcntic;ll to the rc'le;Iscs )>>cdicteI) by the Reactor Sa Coty Stlldy Col. siIII)ltlr,lc(:idc>>t sc(III(.>>ccs.

I'Io i>>dope>>dent

'l 0

I

)

~

~

~ ~

~

~

5

~

~ 5-ca)c<<latin>>s h;ivc I)(.'c>> ma<le l)y (>ithcr thc st'af.'I

<)i t'hc appli(:ant For the" fission prodiict rcmov;)1 and transport: pro(.esses assiimi>>g a core mcl,adown accidc>>t oc(.urs at f)ial)lo (:anyon.

'I')ic ))iahlo C;iiiyo>> containment: dcsigit, equipmcnt layout, an(l 'tlic design oi: acci<)c>>t miLil!a'ting enginccrcd s 8'ety features arc siii)ilar.in m;iny respect:s to those that were oxaii)inc<1 for the pressurized water reactor design analyze(l in thc I',c;ictor S>>Fcty Study.

'Ihus, while (li I'I'crcii<<cs exist in some parameters siicli as cont:>>i>>me>>t i,>>ternal.

volume an() co>>tai>>ment spr;iy Flow rates, t:lic role;isc categories for the T)iablo (.'a>>yon (l<<s.igii shoii)(l l)c reason>>)) I.y corn))al'al)lc to t:liosc ~)ctermined l)y thc

)tea@ t.'or

.'>:i I'(. (:y S t u(iy.

'I1)e fteact'oli.lf<'.ty Stll(ly I'oilll(l t:hat; I')IQ i'Isl; to iiic ))u))lie is largely I

dominat:e>t.c(!Orics 1, 2,

<<>>() 3.

Itc)case

(;;it'c(>oi7 1 is not:

~'

stronf,iy <lcpeii<1c>>t ni) syst'<'.in (lcsig>> since it: i>>v()ives 1'llc occurrcJlcc of.'

stcam <'-xplosio>>

ii) (lie r<<a<<to('essel f'(:I l()ldl>>>> (:('>l"l I>K:I t

(>>i tii (o>>tainmcnt:

)'ailure occur) i>>I>,:i."

t lie i < siii t. of'li>.: st(.'aiii""xf)l()sI>'.. 'I'l>e atmospheric releases associated wi tli Rci<.>>sc ('.at( gori(.s 2;i>>(l 3 coii I<1 va(} as a

Funct;ion OI. plant design.

Ilowcvcr, t:hc i csu1 t:s prc(hctcd:in t:lie i!ca(:f.o)

S>a I cty Staidy in<lieate t))at-thc radioi)(i>. I i(lc>> i>>'hi<<li (loiiiiii;it'<<. th() I)()nc (i>;>>'r<>iv, I'h}'roid a>>d N)io]c body <loses arc io(lii)<<; tcf I<<i ium;)r)() <'<<sium.

'I'hc 14 <<.'t'o) S>>I;cty Stu(ly conc1u<)c<)

L'))ad: c<<<<c)cc::

'c"..Ca::c

.".:.> '(:..i-'",'

'. "" il() l)c rcprcscntc(l as releasing 30'; of')ic tcllin ium a>><l rcspc<;t"ivcly 70". ai)<) 20."'f: thc iodine an<1 S0-o alld 20O of')ic ccsiimi to thc atmnsl)herc.

0))vie>>sly, even H' detailed examination of l)ia))lo (,anyon werc to iii<lic;itc sonic i>><.reise in (lie Fn'action of: a given:isotope rc)case()

t<) 'tlic ci vlro>>mc>>t:, tlic r<<lease f.ract:io>> cou]d

,not excccd 100'..

'Il>>Ls, I'or tlic..imp>)> t:a>>t ra()io>>ucl id< s, errors;in release category assimli)t:)()i)>> could 'iiicrc;isc 'ihc c;i'le>>late(l dos<> pro))ability

,P

4

~ t

-C)-

hy no moro t)in!i a I;ictor of; I;ivc.

('nnsidcriug tlie rcl;itive cotitriluitioii of: tho varioiis niicl ides to tlie prcdictcd dose, i t;ippo;irs very nil.ikoly t)iat tlic c;ilculatod doses coiild l)o in c) ror l)y io()rc tli;ni a Cactol OC tllrco diic to 'i'('1('aso

( l('ego'1'y nsstitlll)t'iotls.

)'<<Ill.L 1 tees

'1')ie systcnt I'niliirc rates wlric)i aro !ised:iii (I~.'i'crmining tlio likoli)iood of: t)ie varioiis;iccidcnt sc()i!cocos werc estrin:ttcd l)y tlic app'1:ic;int by consL'ntctirig I-;iiil,t trees.

'I'lic (atilt trees (tcvol >>).n:!. lo") cal. pattern those combin':.ti~!is. o('o.~ (nant failiircs oi )iiv'ntt c='i-i.i '.ll c.i cot)1<!!lse tho systetti to l':.ll.

I'le )in.)") i.ot. Co.l:l.-;t: ncccs::::.i-to,)c.:fo~". a detailed review of t Fault trees presented in AoicndincnL'52 to tlio;ipp'1 icati.oii since thc Cailure of'.

all <<.c. power:i(re'Iatively 'Iow nccolcr:itic>>s doitiinn(cs tho results oC tlie I:niil.t ti'cc alia 1ys i s bo evaliia(cd tlio sensitivity 'oa t)io risk rcsu'I Ls cal.culntcd by the applicant to vari;itinns iii thc C;nA.t troo results

))y cal.culating tho chango in thc hc(luoncy ol'lolense (.'atogory 2 to i<<creases in Cnult tree res(tits.

'I'l)is I

I.:: -:i;'..I'.".tc". ':,I:"':: (.'".n I" o' is exist in tlio 4'ault trees suc)i t)iat -.i.o i r". t'r.":oi. s';.>~.'.".i tulY ~"i::".'ll:lllcr('~isol)y il Cttcto) of= 1(), tl)(no wo<<l(l l)e ntily;t I ivo-I'old in(;r(:isc it> tt)ic prcdictc(l Crcqucticy of'<c').case (;atcgory 2.i 'llic sc>>siti.vi.ty <<nalyses porter!tie(l by t)lo Bppl.icant 1<< tile ilttnclmicnt. to tlic 0<<tobcr 14, 1977 lottor indicate t)int even a l<<cto) of'() incro:iso.in 1<elcnso ('.ntcgory 2 woiild increase 1

tho cal.culntod pr()1);il)ilit.ics ol: (.",.cccding S1.0 rcm to tlio l)otic ntarrow ol of oxccodinf, a long torm wliolo ))t)i y dose ol 15 r(.m at speci Pic locations by less than:i facto) of'.wo.

T>>(rc:iscs in ))rcdiclcd system failure rates coitld also iticr("tso t)io pro));t))il i.;) ol't l<clcnsc (ntc),o!y 6 release hiit

,P

(

~.

7

~

,\\

<< ~

-'7-this release

.is n relatively smn]l. Co>>L'ributor to (loso.

Bcc<<luso 0I I'hc (10nlln<<lncc 0F thc loss oI'-;I.c, I)>)'Ãci >

lve consi(lcr it.very unlikely that thc Cnult t:rco results N'i.l.l l)c in error by even n Factor ol'0 on thc hit h sido.

'I'hc sensitivity t)1)a'lysos (Iiscussed above indicate th:lt tho res))1 t: 0I'hc a>>alysis;lrc rc'I.;IL'ivoly;i>>.Q>>sitivc t'o an error l

'a of. this ma(!n:i.tu(lc.

(.onsc(jllcncc l'fcth0(1010/!)'

~ ~)S t'onscqllcnco nnct'ho(10'lol.y Iol'ote>>L'i'll ')cci(lent's;>I: thc l)iablo C'lnyon Reactor nro also dcscl'ihc(I i)) Rnen(I)n()nt; S2 to thc;l))pl ic;It'io)).

Ihc consequence calculation

includes the I'01104'I>>f", Io)l'I'<<a'>>CI"II cllIclll<<ltlol)(',Ie)n('Ills'.

JL~

.I<<<<RI)K)l)>>L'l 1;I(lIo;)ct) vc n)'Itc)'I;)J I e'I c;lsc(l 2.

I<<CCQCL's OI; t'hc 1"I(I)o'Ict)vc Ills)tcr3'll oil I:hc ))01)()l'Ition 5.

Ibvc)ncnt OI':hc lnaL'clllll ln 't'ho cllvil'0>>)ncnt 4.

1,0c;ltlo>>

() I t'hc popul;lt"10>>

I"s)eso el()lncnt's:II'(

co>nbi>>c(l Io (lc tellni>>c th(,

('()>>s> (1)I'>>( cs ()I potent:i;11 accidents

<<lt: a nuclcall'0'lx<<CI')Ii))l'(;~

'1'he 1<cnctor S;llctv Stl)rl~ (II<~) norCon~c(l a>> l)n;llvsis to (lot.crmi.no the

$)

risl; to tl)c pl)bl.ic Iron) nuclc'll poNcr at tJ)e

'I irst ll)II rclctors.

'I'h(.

applicant s mothodoloI~ i>>corporntcs nnlch ol thc I!SS:11)pro;Ich to risk assosslnont.

As indicated in R)n('>>(ln)()nt SZ, thc ITS is n I!Cncric I i. I'tu(ly 'nl(1, l)s such, its techniques

)n;ly>>o( l)o diicctly nppl ic;ll)lc to s:i to: l)cc,i.l:ic c;llculntions as perfoiined,in tl)(.

1 i,censing procoss.

Ilowcvcr, tho tcchniquos oF the RSS can ho used to (lotcrmi>>c thc ra>>I!o ol'rob;)1)il.it.ics;lssnci;ltc(l with many ol'. Lhe

~ calculntionnl. rcqllircn)c)lts ('c.(!.,

(I(): e levels) th;It;I)'() pel Conned:in the

~ ~

I

<i 4 ~

~

~

t<

]

'I

~

y\\1

~

~

vl

~

licensin<. nroccss.

)Iorcovcr; those Lech>>iqucs ca>> bc <<scd to present a

l morc realistic <)sscss)nc>>t ol: thc pc>>or>>7. lcvcl ol p<<I)'lie safety;)ssociatcd with a n(lclc;ir pla))l'.

'I'hc apl)ll,ca>>I: 'I>>(',o)7)ola)I't!(I Ils m:>>)y ol I'hc nn)(l(!Is (>>ld tc(.h>>i<I<<cs of IYASII-1400 4

as.it l)C7;ieved w(.l'() appropriate Iol'he'.ir site speci I'I(. calcul;)tion.

In 1)articul;n; they i)so>vers.io>>

I';)cL'0)'s ol I he 15S i.n conj(nlction with the dispersion

)nodcls a>><l <I:)ta which arc speci I.ic t() Ih(. 1)i>>bio ('.)>>yon site.

I'.Cfccts ol'<a(lio:).;I ivi I.v i>>) I'oi)<<latin>>

'I'he lose <<()>>(';rsio>> Iactors of thc IveacLor

!i;(I'(!Iy SI<< ly (sec Appendix VT ol'hc st<<<ly) w()r(

i>>(.o)1)orated:i>>LO the api)1 i(a>>t s (.:)'I(-..<<1:)tions for csti)nati>>I; the <lose,;)))(1 th(.'r()l>y l,hc effc(ts ol

.I:hc r:)(liat ion 0)) the pop<<latio)).

'I'hcse close

"~ ly I

. ~

collvcrsi 0))

5'a('I'.01's (1')'c )live))

I'0') each ol'8,'ll) z (lose p()LI)way ()ll(l isoLopci t)nd i>>elude i))t'0)'>>:)I i <)t()pi(l (I(.<<;)y a>>d do. e (,.0>>I.) ilniI:i<i<<, finn) olhcr organs.

Nc co>>side) the i')

(is(

to I)c;11)p)l)I)rl(iI'c.

I t)v())Rent O'F hbtc') i;CI in Inlvironn)c>>L'

~'I

~

~

<<". i

'fhc appl.ica>>t ca'Ic()lated dispersion 0I: thc )natcri;il in the cnvironmcnt by usi>>p the dispcrsio>>

)no(lc'I Ior continuo<<)

'low lcv( I re'leases as discyisse(i in Beg<<i;)tory f'n)ide 1.4.

'Il)is nindcl ass(uncs;i st).ail,hL l:i>>c trajectory with

<<nl.'Co)ln col)eel)I:7'a I' 0>> over a

22 se( to).

1>> o(lr ol) I >> I ol) tlo s sccl or.-averagi>>g is not appropriat.;e Iol tI)c accide>>t rclcasc cate(.o) ics ol'hc 1)i>>bio CL>>yon analysis l)ccausc ol'heir rc7ativcly short leleasc

<1(>>ation ti)ncs.

'fhe usc of

.,thc sector n)o<lcl ea>> rcd<<cc thc n);mi)mnn clo<<d co>>ccntratto>>s a>>(l <loses by about a I'actor of 5 (Co) sI;il)lc )n<)tcoro'Ini,ic:i'I. (0>>(litin>>s a>>(l I'0) the (list>>>>ccs of conccr)))

as conip;))c(I to the.inst>>).t;)>>eo()s passage dispersion mo<lcl used in thc 8RS.

~

~

~

~

Tn addition, thc applicant's metcorologi(al uloife1 ass(uncs time-invariant win<1 speed:I>>d st:lhil

) ty. -'ea).ly aver;If,c proh;lhil i I'1(-.'ro I'act'ore<1 i>>to tile ca1cu1atlo>>

for tf1('ill(1 speed;l>>d stab'il.ity coulhi>>;ltions at cacll direction.

fhc 16ih )no(1QI <<s(!il;I spc(:L'I'(uu of tlllle dcf)c>>de>>t c()ndi 1 io>>s Ilnd 'ld'fusts tflc probability f'r av(:)')veri wi>><1 <lircctional. I'rcqliencies.

St'))dies have s))own that thc use of thc t'ime-ilaw;)riant model'will I!Qneially l)n<lcrcstiu)atc the complcmenta)y cluu(11;lt ivc <li. t) i.'>l)tion I'lulctio>> (cc(ff) Ior early i'atality consequences hy If)(I>>t:I I';Ictor of'l'h<)>> coiuf):>>(..(I'wi))i thc time-vt)rii>>t 1111C-Vary.A>>

c<llculatio>>s Used,i >> 'I'he liSH

, I s>',

~ I1if)c af)p'I.ic;I>>L'>> mn(lcl (Iocs>>ot i>>c11)(l('. a pl(u)1(: rise Ionuulatio>>.

'fllcrc is s<<hsta>>ti;11 Q>>crpy:Issociatcd witf1 those rclc:Isc categories which dominate tflc

) isl;.

'I'1)is energy litill tc>>d to lift the ra(1ioactivc

)ullter)al. 0 I'; t'e v I'0) I>>(f I thL'. I'<. I 0 )'c I'l('uiI ss I(in o I I'h I:: pl Uluc risc tenn wi'll I,< n(.'I'al fy I'~sllft ',ln Ill)I!letxf)os(>>vs h(.'in/ calclllate<l

'fills conservatism is nof e;lsi fy q(I;I>>tif'.ic<1 bcc>>lsc nl')>>tcr:lotions wit:Il thc otflcr meteorolof.ic;)1 p:II"lul't('1's, hll't, i I: (0111d ()v<'.I'<'.Sf:IIII:Il( lii('. co>>cc>>t)'ations

'tnd (loses t)y I!actor of';11)out: fi) )(Il rll('fist;1>>c(.s col)sider(if il) 11)i'ef)()) t..

'flic appl.ica>>t s;I>>;I'lysis;)iso uses;I silufo'I i I.ied t(.cf)nique to accou>>r. for.

<lownwin<1 decay, p'!(~)le <Icplctio>>,;lnd I'i>>itc c'In)Id Qztcl.n;I] dose corrcctio>>.factors,.

11lcsc

't c'Bus werc',I >>(;0')'f)0)'ll L'Q() '1>>to a co>>s tf)>>t

')")(1 I a'. I()il t')'<Il)sf)ort I'<let'or 'Ior eacil relc;lsc c;It.c!~oiy;)>> I (fist;>>)(:e frou) thc site.

'I'his si)i)pl if ic;)t:io>> is computed on a yearly avcraf c u)<.t<(>>o In!,ical h;Isis Iol the di I fel(>>t (lin:<ltfons ()ho(It t'llc pla)lt.

V fl Thc llsc of th)s I;I fol;I'I,'Iol<ed thc applic;1>>t to <Icco<<pfc tf)c i>>dividllal radio-nucli<lc fanm tile dispersion model.

'Il)ere fo) Q, tf1<

)icf.;)se )II:)I,>>itudcs;)>>d dose

.Convcrsio>>

facto) s (.Ol)'ld he couihinc(f

>>ul s(uiuue(I t<) fol lu dose co>>st;lnts I'or each

<" ~

,re)ease catogo)y.

'Ilrc,deco<<pling of thc iridivid<r:rl..r;rdionuclido from tic <lispcrsion rond<.l.;rs w;is I)orConncd 1)y tlic:ipnl.i'cant can lead to over cstim;rtcs <)'li<. iurxinuun dns<'.s <liiring rrrr C;rvn'r'a))le

<rlru<)sl))r<'.r ic conti.talons (light wirrds and st;rl)le weat)ror):

sigrri I'ic:int rad'io r<..tive docay on route can occur for soirie of'hc import int radionrrclidcs;;md depletion estiuiatcs an<4 the I':in.i.tc cloud sliinc re<hictinii f'actor cou'l<l. 1)c souiewli;it )ii.g)icr than would;ictiially occiir.

)Iec:i<iso of'lrcsc factois, tlio west scvorc air concentrations coul<) bo red<ice<1 wlioii tlioso I-:ictors aro trc:ited properly, i.c., the app1icant overcstinrated tlio severe doses.

'1)io appl.icrnt lias oniitted proc'ipit;ition roiuoval fe.g. rii>>) mccliariisii froin t)ie ana)ysis.

'I'Iiis oinission caii lead to siil>staiitia'I. rrndercstiur:rtcs in doses I'or s)>ociFic sitir<rtinns;rri<l loc:rt'iorr>> an<) can d)';istic;il'ly 'inip'ict t)ic ccdfs for tlic I

1ilgll

< or)so<I<<or)<;o, I<)w Iir<)h;il)i'I i ty 'r",>>)g<i ol';iIi)<'.s.

Ne consi<ler t-I<<dispei si<)ri uio<k.l iiig in)r')'oi I'lr<;rrr;r'lysis to bo quite simpl.if'iod.

b'e li;ivc>>ot prcciscly dotenni>>ed t)ic d<!,r<:<) ol'orrscrv:rtisrrr or non-conscr vati.snr oI tlic:ippl.icant s mode'I )Nit ostiuuri< tliat doses coul.d

'bc al>oiit;i Factoi of;ive )rig)rcr.

Tt:is criic;i;il to re;rl i'.c t)iat tho di CI'<)ror)ces f)r:rI li;ive l>cen noted;ibovc;iro.

strictlv dispcr~inrr u<<)dc'I.i>>!I consider:itions.

'I'hc;r)>1)l ic;>>it s n)odel is not

.vr);recept r))'Ic; 'i.t gorier;ites;i rc;iso>>a))'Ic 1)ml);i))'il.;ity <list'ri))rrti.orr oC concentrations.

Thc only dil.fcrcnci tliat c;in l)o note<) in t)rc meteor'olof,ic;rl

<1;it;i 1)ctwcon ))iab1o Canyon and inost oflier sites is thc douiin;ince oF wi>>ds I'roin:i single direction.

Y)rirty-Civc percent ol t)r<< tiiiio t)rc wi>>d will, corno Cr<uir tlie nortliwest at tlic Di:r))lo Crriyorr si to.

Ilowevei, fhe I)<.) rrI:rr i<>>i >>l)<uil')ial)lo (:;rrryorr is very low lrnd the <)omin;)rice ol. Ilris i'iii<1 <Ii'i'cctioii d.)<is iio(

ii)<:r<'a

<" the risl'o tlic public aroi>>id t)io site.

'I'Iie <I<uiiii;)rice ol s< uie oC thc ot)rer el< iiieiits of')ic conscqiicnco rrrlcirl:rtiorr:is <@pe<..t<d to nv<;riv)rolrr tlio ef;le<its of Ili>> <li.VI'cronccs t)iat )iavo

))<'r ri nntc

<I i i) i ))<

I I "<

~

~

I i

1oc;rt,io>> olnj)rr'I;rr ir)r)

Tl>e I:i>>;i'I clcurcnL'l'ire conscqrrencc c>>lcirl:rtinr) co>>ccrns Lhe ].Ocation OI."

the poprrl;)L'ion.

'I'lre;rppl ictrnt s 'rr)alys;is:in<lie:rt:i.;rh1o (:;rrryo>> ic:iclor sile.is cxtrcucly 'Ioiv.

~ 15 r o>>l'irrrr tliis OI)sc) vatiort and Gnd'only o>>e reactor site with a 1O)ver popul:rtior) <IisLril)irt.ion. 4'hilc J

focusing on tire tj)tee hil!hest popu1ation.'Ioctrtions:rhorrt: Diablo Canyort', the applicant:

1)as perl,o'rrirc<l <1et'ai1c<1 sLrrdics t))at inclrr<lcs potential cv;r<)rr:rt'ii)r) sl',r'rr<.r,i<)s.

'I'lies<.;rrr:r I>'s<)s

< lc:ii ly iii<lic;itc t])at: tl)e public is I

at: co>>si<l(.r;il)'ly low<'r i isl; Izoru scisiuic;il'ly-irrrlrrc<'<I;<<<i<i<'rrt's:rt. thc Diablo C<rnyon )c;rcL'o'r tin<<r I r'<)ru ui;u)y otlret 'rcacLors si>>ce i ts I)r)i)<<L;rLio<<.i.s so sura1'I, when consi<lmir)!! tire rri)<ci.l;rir)t'i<.s <liscrrss<<1;rl)ovc.

'Ihe lorv I)oprrlatio>>.a1)err'iablo Ca>>Yon is exL'i'"lllcI 'lllpol'I':lilt'()I'lic i'I.sl r)ci".I)'<'t,r v<'

~

pl

~.

C~cnsc~ ncncc i'>> icni:ii inn@

OAcc t'lic I ()ill' llllcf.lolul I clclllcllLs 01, tire <rolrs<ulirellcc crl lcillilt'1orls h:iv<: I)<<.'.'ri dcl illcrl, I'lri' rlsl'l c<)rill)iliir)<! I'hcrll:is r'<'Irir'ivol>'t;i';iit!Irtlor'wrlrd.

'I'Iic:rppl'ic;r>>( c:ilc<<lat<d (:(,'l)I's Ior <lose;it'lic Ir)i>> Ii)c;rt.io>>s al)orrt tire site

'I with thc hi<!liest 1)<)i)rrl;rtio>>-dist:rncc r<:I;rtio>>.

(:I;rssic;rl llo>>tc ('allo simulation.

L'cch>>i<Irrcs )vere t;licii (.rliployc<l to couihilic tire

<'.Osc corrstarrt,,

atsuosplrere tz'anspoTt r 'I iict 01' tilt>s.

'I'liis s ii'mil:it'io>> teel>>)'i<lire slrotr id oiily randolll v:)l,rill)les.

I h<.'ose collsL'ilt:, rlt'Iuosl)hcr'3<:.

atlrlospher Ic d i 1r lL' <)ll I 're L'or'lrc i ir I:ict'cl)ciidcrit hc iiscd Ior irrdcpcndcllt 1:i';rllsl)olt I.'rcL'01'lnd v;ii'i:il)lcs.

'1'lic dose col)st"ll)L'lii el) coirrl) i ries L'Ir<. r;i<Ii oiiiicI

.idcs,;is. rvirc<l rio doivruv i rid r;rdioacLivc decay;iiid iio c'lo(id dcplctioii; d<.'c;r>;)rid <lcl)let'ioii:ir'<'..ir)<:l<rdcd;ir)L'hc atrrrosphcric tr;rr)si)r)r't';ict'oi'.

'I'Irc <<L'>>rosI)hcr'ic Lr;rrisI)r)r'L';ict'or;is based iil)on t'lic yc;rl'ly:ivci;il!crrretc<)r'olo!!i<,;il disl)< i'sioii <<)rriliLio>>s (li strrl)i I ity

a A

>.)

r' a

. Ijt Cl

)nd ilie;i>> 4'it)(l sigil); tllese II;>>">>lictcts;ll'0 it)cltXJ iil <Jetttil in tI)e DL'Iitospl)eric dilut:ion I;)cioi (list'r'il>tttio>>s.

Ifcc:<<tse of tl)csc it)t('r;rct:ions, tI)c I

I i)s stir)IJ>t:;1 ol'I 0 I,

.I I l<lei )<.'Ilderlce het )veer)

L'I le dos 0 coils t'a I I ) I, >

t Ile at tltospltc 1'.) c I

tr;tnspot t I;ictoi,;iiid tl)c;tttttospJ)Q) ic <<li l<<tio>> J;;tctoi vat iahtes is invalid.

lllc alsstlt>>pt:loll ol it)d(!peti(l<!'Acc lc<<(ls to cotithinaltioi)s 01'>>) ca iistic

, sit(tantions Nl).icl) ('.;)n bins thc 1csuits.

I'otc>>tirtlly 'I:)1.gc1 (loses cot)1(J bc gcn<:rate<1 t:liai) lian()I(a tltc c;isc.

.>Illlllll~ir 1'Iatsed 0>> I'l(

<<Val I <<rl i Ioil dCSC 1',)hcd a11>OVC >

Ne he I i cvc i'lie tl>>a lySCS PCI'COI'I)lcd hy tl)Q:ti)pi ic:)t)t; to d(tet)nit)Q tl)c coi>>pie)>>eiit;>>y. cul>>t)I;Itive dciisity f(tticuons oC tl)c <lose;it; slicciCic Inc:Itiot)s <<s;i )cstil t: ol';iitl)q<<akes ol'))ious nccclci:itio)is;I'rQ>><>t: (>>>te;tso>>:th'Ic, atss<<ritit)g tl)Q 1>i<>I><<l>i,litics of, tl)c aj SC'I SlillC alCCCJCI'alt l(>i)S all'(.'ot'I Qet:.

Ifowcvet', vc i)ol>> tl);tt tlic c<<rvcs i>>;iy hc loM l>y:) I;tctot'f;il>o<<i: 5 ar)d ti);It'. tl)<<(.i ro) hoi>>ids on il)ci)'nrllyscs a

a

~

'I

'" I

~ a vQg t>>ital)t cztci)d to;ii: Ie;)st';i f;tct'or of'ive oi) pt'ol>;il>i I i ty;t>>d n Irt'cio)'C a

tlt)'ee oil (lose cx(!Ill(li >>g coi)s'i <lct';ttior) of tl)c iiiip;)ct oi) t:l)c cl I'0)'p)'errd 0 (

thc p)cdictcd pi<>l>:)l>iLity ot;i t,ivc>> scis>>iic;iccelei;it'ion 01'l)c possible Vanri.;ttiO>>S in ('Otltp<>>>en' Cspui)SQ to Seistt)i(; (ICCQI(n'It'iO>>S."

TJ)c dct'e)'ri)it)ir)g I;icto)'or)cctning )'isk loi l)i;II>IO t.;tr)yo>> Site is tl);it:

tl)c J>oJ>tilt)i!lot) (I('>>:ai ty i>> tlic v1ci>>ity ol'lic si te,i:

ext) e>>icly 'lou.

Iafctcot'ologic'll ol't,i)el')lodcl1 rig (I i.ff'el'Qllccs a)'('. Qxi>ected to l>Q ovcDYI)QI)l)cd

~

a hy tI)is 1>OI>ttl;it'.io>> (.oi)si(let'Itin>>.

'I'I)et(:I'0)(, N(! I>( I ieve tl)ltt tl)Q risk to

~a'"+

a I

II I

I i

t 1)C p<lit al S<! I Sti) ICal I Iy-i)Id(Ice>

is 30wcr'l):I>> tl)(. Ove)') I'I 1'isk tssoci;)ted wit'll I,icc>>:ed t)(tcJca) 1>ower 1)I;l>>ts sill)i,l;<<'>:Ily"..('d i>> tile Ite;)<.'I'(>I 8;) f(.t'y Sttt(I).