ML20149F777
| ML20149F777 | |
| Person / Time | |
|---|---|
| Issue date: | 01/06/1985 |
| From: | Martin J NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | |
| Shared Package | |
| ML20149B718 | List: |
| References | |
| FOIA-87-743 NUDOCS 8802170360 | |
| Download: ML20149F777 (8) | |
Text
{{#Wiki_filter:. s
" 11/84 PREPR]NT: FOR PROCEEDINGS OF THE ElGHTEENTH MID-YEAR
~ ~ ~ ~ TOPICAL SYMPOSilX (ENVIRONMENTAL RADIATION '85) 0F THE HEALTH PHYSICS SOCIETY, JANUARY 6-10, 1985, COLORADO SPRINGS, CO.
[ OB.ItCT!rt$ of EMCRGENCY RESPomst AND THE POTENTIAL BENEFIT 5 OF
, , , (VACuAT!0m AMD 94ELTER James A. Ma rtin. Jr.
Division of Risk Analysis and Operations Office of nuclear Regulatory Arstarch U.S. Nelear Regulatory Copyission Wa shington. D.C. 20555 AtS T Basic radiatica pectection objectives for the controlled eavi roveat are transferrable vertatm *or use in the uncontrolled or ersergenes situatice. Thes e objectives are: aven nes'-tere injury or fatality and red.ce individsal risks and htal heal th eects te 'evels as los as reasonably achie.atle ( A; ARA). The poteetial bene'sts c' sre' teeing and evacuation in meeting these oefectives daring a response to a seie t ;.1 ace: cent was investigated. It coec be vers difficalt to rTduce total healtS ef'ects because Collective dose (man. red increases mono. tonica117 for scores, pen.aps hunceeds of railes frce a release ocint. In this regard, shelter and ad hot -espiratory trotection appear to te the only rational
. . and feasible near teer erotective actions that sculd te available to most people.
Hoever, the first twe objectives can be eet by early, precautienary evacuation, within two tc three tiles trrestely upon the declaration of a General Emergency (e.g. a core melt accident), and sheltering elsewhert. In the event of an actual major release. later relocation from highly contaminated areas aculd be an integral part of the esergency response. Analyses performed to investigate the various erergency respense options and sciential benefits art described. The perspectives obtainec shes1C be reflected in emergency plans. INTRODoCTI% Radiological emergency plans for fised nuclear released to the atmosphere) (U581; USA 2a). Al though f acilities should be cons tructed to achieve spe. these ' source teres
- are being intensively inves-cific radiation protection objectives in the event tigated (B484) and may be revised in the future, of a future rediological release. Basic radiation the NRC set was used as the best current estimates protectica objectives for controlled enviroernents, of severt accident source terms. The taportant such as a tork place, can be succinctly stated as characteristics of this set are listed in Table 1.
fol b ws: These sculd all be core eelt accidents. The o 40!D serious non stochastic radiation effects estimated probabilities of these accidents occurring (i.e.. near tert. or early, injuries and art very low, twt the estimated release fractions fatalities) and (of the core irventory) to tte atmosphere sculd be large, especially for $5T1. In general terns. o REDurE individual stochastic eists sac total 55T1 would correspond to a coincident early. latent health ef fects to levels as los as *>ssive failure of contairrent, with little or no reasonably achievable. 'crubbing of the release by engineered safety features in a plant. 55T2 would correspond to a inese basic objectives art transferratie ci'ectly coincident major contaiteent failure with degraded to the incontro11ed or accident envirovest, so performance of engineered safety features. 55T3 long as it is recognized that there can tw no could involve a late baseint melt through accident, gua rantees prof fered that all objective s s.on be with efficient scrubbing of particulates and a met in all conceivable circestances in n uncon- smaller release of noble gases, as well. trolled enviroment (U578; U580; In84). For this paper the potential hertfits and practicality of SAct&R004D combinations of evacuation and shelter to achieve these objectives in response to severt light water Clearly, even wit >.out an opergency plan scme nuclear poner plant (LWR) accidents was investigated. off site a ergency res"nse by the Mblic would be espected in the event of these accidents. (narined The LW5 accidents considered was the set her, ses the Westion: Det would be a practical SST1.1572 and 55T3. Sugges ted by the U.S. Nelear Regslatory Comission (hRC) as representative emergency protective action scheme involving groupings of severe accident source terus (frac. evacuation and shelter dien would satisfy ihe basic radiation protection objectivesi It was tions of the cort inventory of radionuclides recognized that individual entrapmer.t sitations 8802170360 880204 PDR FOIA SHOLLY87-743 PDR l /_ f. - - - - - - - - - - -
e t Car be readily visual 12ed; thus, the perspective o In the further event of an actual major was to investigate evacuation and sheltering atmospheric release f rr contawent, tienefits that could acCure for most of the people people in shelten would relocate fry riott of the time given the postulated 4CCidents. highly Contrinated areas lef t in the Certain leading clues were available from wake of the release, previous studies. Three are especially noteworthy. In mungG 0396 (U578), risk vs distance was displayed It f s noted that most core relt accidents in a sinole manner in a fig 7re reproduced here as would not involve early containment f ailure (U575 Figure 1. This figure merely indicates that oose U582. 8a84). Rather.11 is es timated currently vs distance decreases monitonically that only about one out of ten core melt accisents (and free arisk) sour5 point for an atmospheric release, might leaJ to en 55T1 type of release, on the Theoretically, the oetresse varies as r.s.s, a ve rate. For this reason, the phrase 'early, approsisately, where r is the downwind distance, precautier.ary evacuation...in the event of a core Because of oestructions, wind meanoer & wind no t' was used above. shif ts du ring a release period, cose sight more realistically vary as r 8 (inverse souarea). The A version of the CRACP code w+.ich peevides r3 (inverse distance) curve in Figwre 1 was for three rergency response renes was usec fer included as an sic to the reader. Without fuether the caleviations (U583a; USS3t). Peccle in elateration or caveats, the information in Figwee 1 shtiters were provided protection factors o' C.33 can be taken to indicate that protective actions for each of the three major pathways-esternal within a few miles of a release point would be garras frr the plget, inhalation, and 9eocc most beneficial because the risk is clearly greatest c ont a-ins tion. These pectection factors art within this distance. typical for residences with bass ents (0$75). Two otner clues were provided in NUD!G/CR 2239 (U582e). especially in two figures reprocuted here Evacuation mes modeled in the first. or as figwees 2 and 3. Figure 2 displays the probatis- near. 2one by assweing a one have celay af te* ar. ity of exceeding various nebers of early fatalities initial warning (e.g. declaration of a General given an $571 accidental release and various Emergency by the plant). and a 10 pph evacuatier. reegency response assumptions. The botte curve speed radially away fra the plant, theer tHs in the figure clearly illustrates the potential asseption, people kgin to leave one half hour benefits o' a miniral delay before evacuation. af ter the 55T1 release begins. The asseption For the $6rrary evacuation curve; delays of 1. 3 that people leave one half hour af ter a major and 5 hours were astred to occur 0.3. 0.4 and 0.3 re' ease tegins is a generally pessiaistic one. of the time, r?spectively, thfortunately, for the This certainly should not be the planned sequence, botte curve in this figure an evacuation within This has ken noted previously (N77. NBC. N82 gggg), twenty-five riles was assumed, et a speed of 10 miles per hour. This implies that a large area could be cleared in 2.5 hours - hardly a practical The second, or rid rene entended fre the inner zone to 10 riles. People in this area mere assweption for most people ecst ut the time, assrte to relocate af ter four hours esposure to Figure i contains two irportant clues. SA M ground Contamination lef t in the aske of the pJff. The third, or far sone entended fra 10 miles, here is the conditional probability of incurring Here. people are assmed to relocate ef ter eight an early fatality tryond various distances assuming hours exposure to ground contaninetion. These the 55T1 accidental release and various per9ency res pons es . Again, the importance of a sinimal relocation times were estimates of the time it delay before evacuation is clear fra the bottor would tah to locate hot spots, provide notif t-curve which indicates that with a short delay twe cations, and for people to move a short distance the early fatality distance should not exceed two (Ma77. U584) away from the hot spots. ril es . Again, the impractical twenty five rile The New York City meteorological set in the evacuation distance is noted. Further. as indicated CRAC2 data files uns used for the calculations. in this figure, all shtiterees were assmed to This set contains reinfall about eight percent of stay on conta-inated geownd for a full day before relocatter to uncontrinated areas. This is the time. Rain can cause heavier than normal ha-cly a realistic assu ption considerirg that ground contamination. Also, the population distri-dose rates in rany areas could exceed 10 rer per butions in the CRAC2 data sets were used. , hour in the wake of the puf f (U584). Relocation RESULTS ! frr shelter would te espected to occur soon af ter ra:tological monitorin; ters idertified such ' hot Individual Risks s pots" . Principal results of the calculations are CALCULATIONS displayed in Figure 4 and Table 2. As shte in l Following these clues consequence estimates Figure 4. for an 800 megawatt electrical WR at were perforeed using the following protective a coastal site in the U.S. aero early fatalities action assumptions: was calculated for the most severe SST) accident postulated. for en early evacuation distance of o Early, precautionary evacuation within three miles. The uppermost curve in this figure shows that the predoeinantly sheltering protective
- 1. 2 or 3 eiles and inrediate shelter action strategy clearly suffets by carperison to else=*ere in the event of a core melt the early, short range evacuation strategy. Thes e accident (General Entgency). results bear out the intuitive, walitative perspec-tive illustrated in Figure 1.
t
4 l j Identical calculations sert perfomed for These perspectives are illustrated in Table 3 four other LVR sites in the U.S. Principal resu ts i and Figsre 5. Table 3 is a ser.ary of pertinent for these five sites are shown in Table 2. Althcu9h infer,ation in frJRIG 0540 (U577). which illus. the CRA;2 population distributions of these sites trates the pathway and temporal contributions to were used for the calculations, the nsults we" rieriralized to 500 persons per square mile within total calculated latent cancer f atalities for the PWR 1 and pWR 2 accident categories of the Reactor 10 eties. The power levels o' these LWRs ranges Safety Study (U575). These accidents would be of over a factor of two. The results for the lowest the ilk of 55T1. and would include substantial power level are incicative o' what nsults for the releases of long half lived cesive radionuclides, highest powe level would be with a factor of two 11 is readily apparent frF this table that long reduction in the 55T1 source tem. terr esposure path =ays would doeinate the total neber of latent uncers. Osly entensive and Several irteresting aspects of the results expensive decontrination and conde-nation a re sto.e %, or r.ay te inferre: fecr. the inf erre- processes over the long tere would be ef ficacious tion in Table 2. Inece wre ne residents within in rt3ucing total latent uncers. t* *ee rile' cf site nybe' one. arc 2erc early
'stalities mas calculatet. This illustrates the Figure 5 illustrates ho= collective asse an:
potential bere'11 of a very early evacuatter of total latent unter est1 Fates increase with distance. r e a rby areas. i.e. before a release givet. a cere This figure ses tate *. fro- NJRIG/08 2238 (U55"c). melt. Fo' the highes t poner levels. 4 few early Calculated total lateet cancers accrue to large f atalities were ulculatec for 55T1 for site cistances, regardless of the ce positter. and nebe' tw: evea with a three rile early evacuation ra;eitude of the release. Inoted, for the average assu ption. These peo;?e iere uug%t by the front site fully half the latent uncers could accr eu c' the plume. it res t cases. This was a ht9" outside of fif ty rites frcr a release pcint. Depulatior. otesity site. le all cases, several This phenvenor has beet noted previously for tens of persons suf fe rec early injuries (e.g. ro tine atmospheric crissions of purely noble prodrcr.a1 voriting). These ulc614ted injuries gases (Ma74). A core 11ery is that near field occurred at various 42irNths and to distances to erergency protective ressures would provide little 12.5 eiles, tot eostly well within 10 riles. The benefit as regards the objective of reducing collective total neber of early injuries calculated is an dose and total latent health effects. Pt>we v e r . e rtif act of the CRA* cooe. unic*. a$ds up ulcu* for a surely noble gas release the cloud (etternal) lated injuries whereve- they are located. At anJ ga ra dese pathway would doeinste. In this use particular ariewth, no so , than a few earlJ imediate sheltering to large distances would be injuries was ulculated, tich would be the case an ef fective protective action to achieve the for a single suf f. Further. these injuries occurred cbjective of red cing total latent unters ALARA. at los conditior.41 probabilities
- especially where sheltering would not be inconvenient a nyway. This would be little different from an in all cases, peak early fatalities and air pollution alert.
inju cale.'ies af terwre associated transport, or stablewith reinf all. a sudden meteorological The interlay between estir.ates of stochastic-conditions (low wind speed. marrow plumes, nightime latent cancar fatalities tnd costs of condemnation condi tions) . Erergency planners should be espe
- of contaminated property is illustrated in Figurt 6 cially aware of the import of these particularly for the $5T1 accident at the Indian Point site.
adverse, low probability wather conditions, dich The data points for this figJre wre obtained lead to heavy ground contsiination by particulates. f ror. ItJREG/CR-2239 (U582b). These are for various Separate calculations, not shown, indicate that for dose projection criteria for land interdiction the 55T) etlesse, calculated early injuries could be ( conde-nation) . Iersally in CRAC2. people att eliminated by slightly raller particulate source allower tc, rt' ain in contrinated areas where the teres (lower ground contrination). better shieldin9. projectec whole body dose is less than 0.25 Sv (25 f aster relocation or corbinatiorts thereof. rer) in 30 years. As illustrated in Figure 6. for 5571 a very large increment in costs would be Ito early fatalities or injuries wre caleviated incue ed in reducing total latent cancer fatalities l for the 55'? ane 55T3 accident scenarios. for the by teterdicting property at a lower dose projection. noted rergency response asseptions, at all power levels. it should te clear frcr Figure I that Cre accitional perspective is important in early, precautionary evacuation of nearby areas in this rega rd. For a care melt accident with failure the event of a cort telt accident would significantly of contairrent. the constituents of a release may reduce individual latent cancer risks in the event not be known for some time (NB0; %82). Tnus, of a core melt accident. the shelter to large distance (where convenient) option should te predetertined, as well as the Collective Dese evacuation to short distance option. as an irrediate response to the declaration of a General Emergency In contrast with individual riskt of non- (core melt). Indeed. it was in this light that stochastic effects, uh.ch imuld be relatively the protective action assumptions listed under near field or close range effects, estimated total CALCULATIONS above, wre made. l latent cancers would incr1ase monitonically with l distance. Further, for releases which include a (>e caveat is important here. These collective I substantial abuncance of long half lived particu- dose perspectives derive frcr the asseptier of a lates, the collective rist would be associated propertional relationship between risk and ceso with long terr (years) esposure to g oved contrina-tien. Thus, protective actions during the rergency phase sculd prov10e little benefit in reducing total latent health ef fects.
Olstt'5510% The calculations discussed above show that in the event of a core melt accident early evacuation of relatively ens 11 areas near a LWR and sheltering elsewhere newld provi6e significant reductions in individual risks of stochastic and non stochastic health ef fects. The simple emer9ency response schere suggested can be predeteririned for specific in. plant rergency action levels appropriate for . the General frergency class. Because the actions are 50 sie;1e and easily understood, there is an excellent chaate the plan would work. if reed be. The early, irr ediate evacuatior area suggested by these calculations is the size of eeny los population tones arownd LWRs in the thitec States, anc the early evacuation radius (2 3 riles) is less than the distence to the neartst poN1stion center of 25.000 persons, or more, at most LWe sites in the U.S.(U579). Thus, thert should be fe= tenedi-meets to evacuation in these areas most of the tipt. A further conclusion is that redwetio . in collective ese anc latent cancers wowla be very difficult to acnteve in the early (seergency) response phase. Sheltering during the passage o' a prederinantly noble gas re. lease .culd be ef ficacious, tot only if sheltering to long distances were uncertaken. A few caveats to these conclusions are noteworthy: There are large ecertainties in the absolute values of the results of the calculations. hevertheless, the relative potential benefits of v a ri ous ev a c ua ti on/ s hel t e ri ng/re l o ca ti on prot ect i ve action strategies should be clear, especially where large dif ferences in results art obtained. At a few LWR sites in the U.S., and at cany foreign sites, heavily populated areas exist in the near vicinity. iAich could make early. tamediate evacuation difficult. impractical, ce impos sible. For these sites, better shielding protection factors may pertain and smaller early evacuation distances may be justified. Ch the other hand. Some low population renes persist for many siles, and early, istiediate evacuation of $6ch areas say te a reason. able objective for a cert sett accident. It is acknowledged that entrapment situatices can exist for some people or many pople at some tire. Early, imediate evacuation may te physically iepossible or entremely hazardous during a snow or ice storic for example. Special arrangements should be made in mergency plans for identified persons in early evacuation tones who suffer free significent impediments to acbility. In the eveet of a core melt during a highly temiebile situation, e.g., the ice storm resining in shelter and relocation fror hot spots (if a release occurs) as quickly as possible would be the only reasonable and practical alternatives. These are highly unlikely corbinations of unlikely situations and the suggested protective action schase should satisfy the basic radiation protection objectives for most people, most of t%e time. Finally, as corpared to the suggested predeteririned Protective action plans protective action decisions can be ende on an ad hoc basis at any time. The distinction between petTeTervined actions and ad hoc actions is very important for eme rgency pl anners .
4-e atr1Rin:ts last sette11e Colw-tws Laboratoetes.1964. 'tatio. V574. U.S. At les, nep,1 story Comission.1978 nuc196e Release traer LWs Acteent Canettions.' 'Flaaning basis eo ' tre Developent v 5 tate sac 8"b21D4 Vols I.VI. July IM4 (Cc1 veas 0% Local Goverrrect tactological teergency 4espcese 43M1). Plans Plants'. te ksacet o',Lig*.t mater 4 clear PowrDeceeber 1978 (Watrin nWR[G.039( SwS4 tur ke R .P. , Nris tng . C.D. , Alcet ch. C.C. . D.C. 2 lli). IN4. 'le s* ant constoerations for off. site V5't. V 5. an.c t e s. Le ;.* a te > ## s s t er . 19?l. rerpacy respeese te reactor acctete.ts.* Naita Prysics Jr. 46, 763 773. 'Deeog asuc Stat.stics pe r. airing te %cles*
*: w
- Re a ct e 5' tet' . # J8 tG4 344 . Oc t obe ' . 19? 9 la54 Intemat ter.ai Corrisstoa or tactclogica) lmes"egter. C C. 20t!!).
IPCtection. '8cotection e' the Pwllic in the Ivent o' N '*
- Ra e t a t t er Ac ci de nt s : Primcitals for USIl . U.S. A t el a' le b a It o") Coari s s t oa . IM C.
Plammg.' Etocrt for Ccerittee 4. kopted by tse *; rite'* a 'e r sa seto;ics' C.'rece'att e sir Corrissioe. may IM s.1 as/64/*a.5/2. ,5e.q se s p.ors act
. teatwatto*
e esses ses of Prete etaess ir b e:c- c' hsclear Pcier Plaats'. Ma*8 "a *t ta. Jr . . J.a. . 'Calculat t oms o' Dos es . 4.itG4tle. Re 1. %: eatr e 19E* lbese tngtem. D C.
**pwistio' Deses sac Potential nealth t'fects Dwe 2:lllh te A , tseeric teleases o Rag'on c11oes u fect L .5. *a t i e s 8o se r te a c t o.'s tv. ng :
5?t*, . ato*y Corrissier. 1991. Ust;, L'.t. icaigtles, se;lstteattag f issto* Frcestt lasts 'o* Raciatio* tat a sac Rep:*ts. ~34105 319. Jvne 1974. *ie le* sc"i c' D.e t ng '.W:ca cioeets'. mJets.0772, Ju n e Ma* 7 Martin, Jr . , J. A. ,1977. ' Doses dite 195l (has'tagtor. D.C. 20515). traveling u noer w11 estat11thee pienes.' 6 eaith Possics -Jr. . 32. 305 307. USMa. U.S. Acles' Le;.1 story Corrission.1960
' Tee Develo:reat cd 5evere teactor Acc16eet Sovece teres: IW4531*. m;stG.0??3. heveaber 1962 M eS: ka rtir. Jr. J. A. . IM O. ' Perspectives on the Pole of rectological monttoring te an emergency ' (**s*L*f ten. D.C. 2D55 5 ).
Tra*s. k , hgg1, seg,, M , fj}.}lg, U582t. U.S. Aclear te9sistoay Cosef ssion.1962
'Teterical G idsace for 51ttng Cetteria Develop.
aseS* **.a rtin. Jr. , J. A. ,1962. 'LWt a ccietet se*t' . At [G / CE.2229. De ce'r te r 1982, pre pa rte t'y specte.r-release characteristics ses coesessences.. Sanita hattonal 'storatortes (Washlagton. D.C. tr Proceedtags o' toe sortshop om meteorcloeical aspects v perpency response plans for twcTear 20555). Powe plerts. ERIG/CF.D032. %st 1982. U.S. U583a. U.S. Atlear Replatory Comeission.1983, msclear Repslatory Cosristica, hashington. DC 'Calew1stiers o' teatter Accisent Consegseness 20555. Ve rs tor. 2. C3.AC2: Cor p.,t e r Co de 44 e r' s Gui te' . ERIG/Ca.2326. Fabri.e ry 1983, prepared by U175 U.S.le., clear Regulatory Cosef ssion.1975, Lancta mattonal Laboratories (hashington, D.C.
' Reactor Lafet Detober 1975mat (y %1ngton.
Study *.DC WRIG.75/014, 20555). 4pendia y!, 20555). U577. U.S. Aclear heywlatory Casseission.1977 U5834. U.S. Eclear Re>1 story Commission.1963
'Dierview of the Reactor Saf ety Study Consoosence 'CRA* Caleviattoa.s for Accisent Sectierts of
% eel'. ERIG.C340. October 1977 (hashington, feriromentai Staten.emts'. ERIG/CA 2901 Nrch 1983, prepared by Sandia national Labcratories D.C. 10555).
(hashtestos. D.C. 20515). U584. U.S. telear Repistory Courission.1964
'Dese Calcwistio s for severe LWR kc16ent Scenarias.' RVt!G.1D(2. May 1984 (West.ingtor..
D.* 20515). , J
e
.g*
- O istie 1: Che'atte'istics cf Postwisted Severe Act16ent 5cemarios Attiftat Type selease Cases;teristles 3)T 1 35i 2 311 3 barring % le'o'e 0.5 1.0 0.5 setesse (**' .
Release Dwestto' (br) 2 2 f I sectc je,e*gtey fes:*to' s e 'telet te Geen;,,t).ge (th' t*e s tr:u'e e se.a. 13. 1.0 C.9 C . 0M ' 1 21 0.45 0.0C) !!.81 Ct.et 0.1 C.if C .009 16 5)
*e.5t t. 0.64 0.01 !! 5) la.5e 14 C.C? 0.001 l'*H sw 21. C.05 0.002 ff 6!
ta ilf. C.005 0.00 1 1( t) e, as ee' aet ir t>e seattee sa'et) 5twar (U5*!?. t, . fee a 100: we'e' a 8 eme.se'f mow e s'te' sm
- c.* s' en: cf c:ee 1tf e (2 pa's) M.75). Tatie 2: Renw ts assries 5,..... aac 8 tie i
- c. 1 (Br.
- 1 1:ste e IC e t s t rie g'a t t oas t set . octle popwlettoa etstett>witcas.
gas ti vs 1 actgtti eovals 1.0( tilltoa tv etes. C. li l) = 1 s IC* g,,3y ,w , , ,w , , Powe r testwatter hee c' hebe* o' Site tevet aattus ta 17 tacts mo rce* (&e) (rtles) Fataitties' Iq.rtes' 6 1 1100 1.2.63 0 60 2 1100 1 50 20' 2 20 100 3 a 50 3 800 1 40 20C 2 a 9C 3 0 FC 4 450 1 20 130 2 0 70 3 0 60 Table 3: Tasecret eM pathway teatributions to latentCancegfatalttiesforsevere 5 550 1 0 60 source terms PC st%af Perte*t Coat *%tiee t a. harma113eg tg 500 peenoes/sq pile withie 10 elles.
*** N *l N7 t. he resiserts with three eties ewivalent te Este'eM Close evacuation be'eee a release.
t g Inhaistier fror Clowe ta 3 titere.ai Growe s (< 7 seis) 13 le Inte*tal Gma: ( > 7 sais) 42 64
- lehelatt er o8 Resusperdes Cottatir.attom la 2 1*sestion of Conteeteated Foots 5 10 1R TE
- s. Fror t.8tG Cla0 (U!??).
l
- 9. Pea.1 sef Phl.2 are setoft attife't release teteltries TPor the seattt' lafety Study (U171).
The releases are 6f the oPfer o' 5511. 5 _. _ , . - -- ,- .~
/. .
+
o'
- e 1.0 1
I 08 - l R ela t'v e 06 - l Dose Rate l 04 - 1/e 02 - p,7 3/'15 i t i 1 , o 0 Ob 1 7 3 4 b Distaxe (Miled iss rc 3. keraticestik of c:se rate et cistence for
. it,. 1,..I a w ineric re.o .e I 10 0 -
--;- 7,-
3 $$1) 1120 *(t)
- W0 tWERGENCY RES*0N54 1.P. Slit l
. ^ .::_ -
J I w r;T u i i.
^ 3c.d E :
B 3 i a g '
! INRDCLAf C '
1 % : ICMPH I l- 1 { ~. it i SMILTERIN G t u i agt. Iw :c *.1- i .
) k . SHRDELAY 10 MPH 1 .
o . ho erwrgeecy respoese I . o . Swary Evacuation. { ,
. 55*1
= g ggg) g(g) -
le. I he celay withie 1010eiroh. ;
- 25 ml tvec. Radius
. =1thim 25 v1 ga gg, selocation Time
;p:C' I t' itf i t' it' :'
16 0
'3')
1, [Q,;,Y F C (11][$ R.1 ARLY f AT AtlTY DIST ANCE (MI) Fil.ee I. lepast of a range e, eneegemey cesponse atlectio's fig.?e 3. Sentittetts af etely fatsitti cittantes ta on taltvistes early fataltts probattttttet emerpe.ty respcase asseJtioen.
r-e s*. *
.O e
e i M' y
- 851 1 300 M W-e 10 *']
.~
'E w
w ::
- ) -
.00 st w .
5M81 u -
.a - \
go - 2 u
- 1.terty tweevetion 10*e1 3 Rediws(Miles)
M** i e ieiing . . i i i u rg i i
...us 3 i i i iii on i i iiim 10' M' 10 8
M M* M' K. ACUTE FATALITIES fig.re 4 toncitier.a1 teotah ittten et earic.i r #:ers rf etste fetalitie . enhe teg $511 at:ident eer13 c a ... t iu. 61 6841) are a6. and a 61o= relotetter f ra P,ig'Is (9rt *if.eted artes. 1.0 - e e d 0.9 - ggy3 (F"
!! *L
! es - ssT Eg k!] e4 -
- lQ3-1 ., U~ter. . ..
Die tnDvt6.n
- g!
t, , O'
, t , t , t , t ,
ET'. (U) * ! ate.veat+e. 0 40 W M 100 2 10 een.e
.se 3:wre1..t .
DISTANCE (MtLIsl I* 5
~
I I f t.l?
i.'t !
3:'elle
- r (s.'.tv's te s 'e tge t gaeg g, es talt.ggg
- ..:. e:. .... ...,c, .. ,.
,,g.
. . ::m.
't.1 (8.051 e
e 5 i 5
. M =====. m e
. 7m VMMEWLAWBfTteWTW m figem 6 Calsbletes seen teteet eense. fetaittiM eas .3st g trease"s e 185'1 ettle,s**tsewe'e At Iseinen teteMittiee set e s ev e 's opget site),
e _ _ . ~ ._
]6 fl [. -j, t M 4 UNITED STATES yy & NUCLE AR HEGULATORY COMMisslON
!.IN j w.snisofon. o.c. rosss
< \e JAN 15 885 1.[.' . . . . .[
MEMORANDUM FOR: See Attached List FROM: D. F. Ross, Deputy Director Office of Nuclear Regulatory Research
SUBJECT:
EMERGENCY PREPAREDNESS RULE We are preparing an Emergency Preparedness Rule for nuclear power plants based on a graded responr.e philosophy and reflecting the results of the Accident Source Term Program. This work is contingent on a reasonably favorable eval-uation from the APS study group, and will continue only if warranted by their report. The proposed rule will be sent to the Comission at an appropriate time. The rule is being prepared by the Regulatory Analysis and Materials Risk Branch, RES. The project manager for the rule is Mike Jamgochian. The IE Office has assigned Frank Pagano as liaison for this effort. The schedule for continuation of the staffs' development for the rule is proposed as follows, contingent upon completion of the APS review in February: Complete Division Review 12/83 Request Office Review & Concurrence 3/85 Complete Office Review & Concurrence 4/85 Complete ACRS Review 5/85 Complete CRGR Review 6/85 Rule to EDO 7/85 Rule to the Comission 8/85 The technical bases for the graded response approach used in the rule will be contained in NUREG-1082 "Technical Bases For A Graded Response in Emergency Planning and Preparedness". This document is being prepared by Len Soffer, AAB, NRR and will be subjected to a detailed technical review and evaluation by the Reactor Risk Branch, RES under the direction of Gary Burdick. The schedule for completing NUREG-1082 and providing technical support for the rulemaking is as follows: Summary of Preliminary NUREG-1082 Conclusions 3/15/85 Draf t NUREG-1082 for Technical Review 3/15/85 Final Draft NUREG-1082 4/1/85 Principal changes expected in the proposed rule and some potential problems associated with the use of new source term information as a basis for the rule are listed in Table 1. hd d. 2bf D. F. Ross, Deputy Director g . , , 3 A a.2 - '
%q j i.w oc7 Office of Nuclear Regulatory Research
Attachment:
As stated
~ . ..: h lABLE 1 The proposed rule is expected to contain the following major changes:
- 1. Emphasis on a graded response strategy, i.e., emergency planning and protective actions to be taken should be graded or phased within the emer-gency planning zone (EPZ) with respect to both time and distence based on the significant differences in risk that individuals in different locations of the plume exposure pathway face from postulated accidents.
- 2. Exercises limited to graded response areas.
- 3. Graded response capabilities and areas determined using new source tem information.
Using new source tem information as the b$ sis for determining graded response capability requirements and areas in which graded response would be expected raises the following potential problems:
- 1. The new source tems will be plbnt and sequence specific and there are strong indications that the:e may be significant differences in source terms depending principally on containment type. This could lead to a rule which specifies different requirements for different reactor types or which sets out general criteria and requires each t.'.ility to establish l emergency preparedness and response procedures based on a plant specific analysis. Preliminary indications are that the new requirements would not exceed the existing requirements for any plant type.
- 2. The available source term information on which the proposed rule would be based will lack a risk perspective. Source tems and offsi's consequence information will not be available for a suff.cient number of plants and ;
)
,mv , .g.. l} i TABLE 1 2
? sequences to obtain such a perspective. This will probably lead to l requirements based'on accident sequences producing the most serious off-site consequence without regard to the probabilities associated with these sequences. It is expected that information needed to obtain a risk'per-spective will be available in late sumer 1985. However, uncertainties t associated with the information could be large enough to force us to use the-most conservative sequences anyway.
- 3. Even with the new source term infomation, we will not have source terms for externally initiated events. This could open new rulemaking which provides a substantial relief to the criticism that it is based on incom-plete infomation.
!i
n.- 1 g *- , s . L ADDRESSEES FOR MEMORANDUM DATE0: JN 15 E
~
R. Bernero E. Jordan LF. Gillespie T. Speis G. Burdick. D. Matthews F.:Pagano . L. Soffer L. G. Hulman M.'Silberberg J. Taylor H. Denton R. Minogue W. Dircks V. Stello J. Malaro M. Jamgochian e o l l
1 F. . i
.5 )
ANSIENS , International Topical Meeting on Probabilistic Safety Methods
'~
and Applications Volume 2: Sessions 9-16 Volume 2 San Francisco, California February 24-March 1,1985 1 Sponsored by
; American Nuclear Socie'y Nuclear Reactor Safety Division . Northern California Section i
j Ccsponsored by 4 { Atomic Energy Society of Japan
> Canadian Nuclear Society Electric Power Research Institute European Nuclear Society Society for Risk Analysis j
; System Safety Society Taiwan Section ANS in cooperation with
\x International Atomic Energy Agency oECD Nuclear Energy Agency s ,
( u
, 1
,ff,.
k . 1 ,1 ! 4 s ,p 7 l i ! t +
.(
l Paper 128 AN EXAMINATION OF A GRADED RESPONSE STRATEGY IN q EMERGENCY PLANNING AND PREPAREDNESS L. Soffer, J. A. Martin, Jr., and R. P. Grill
.ft -
I U.S. Nuclear Regulatory Comission !
-L .I I,
I 4 . i
-f
..il-d t
f
- r b
1 t i 1 a , 9 i .
. t
; i 4
< f I,
*\ , '
x w m me-
l 1 AN EXAMINATION OF A GRADED RESPONSE STRATEGY IN EMERGENCY PLANNING AND PREPAREDNESS Leonard Soffer, James A. Martin, Jr. and Richard P. Grill U.S. Nuclear Regulatory Commission Washington, D.C. 20555 IMIR000CT10N AND BACXGR0'VND NRC Emergency Planning regulations, which were significantly upgraded in 1980 (1), continue to be a source of some controversy. This may arise in part from the misperception of uniform accident risk within the plume exposure Energency Planning Zone (EPZ). Since promulgation of the regulations, additional risk studies have been performed (2, 3, 4) showinq significant spatial variation in risk over the EPZ, which inplies that phasi, or graded planning and responses, implicitly recognized in the documents (5, 6) which formed the basis for the present regulations, may be a reasonable strategy. The present study was initiated to investigate this. An objective of the present study was to ascertain a protective action strategy capable of dealing with a wide spectrum of accidents. Such a strategy should be flexible, depending on the nature of the accident, and should provide a priority ranking of desired actions, rather than a pre-selected fixed risk objective, or dose criterion, regardless of accident severity. The priorities, in order, should be to avoid early fatalities, raduce early injuries, and reduce other health effects. This study also made use of the severe accident releases referred to as the Siting Source Terms (SST) (3. 4,). Core-melt releases range from the very severe SST) release with an estimated probability of about 1 x 10-5 per reactor-year, to the more moderate SST2 release with an estimated probability of about 2 x 10-5 per reactor-year, and the relatively benign SST3 release with an estimated probability of about 10'4 per reactor-year. These source terms employ the same methodology and are analogous to those used in WASH-1400. Although an intensive research effort is underway to reassess accident source terms NRC efforts are presently incomplete at this time (Noverber 1984). i 128-1 i 4 _ _
TIME FROM INITIATING EVENT TO START OF RELEASE While the spatial variation in dose or risk, showing a sharp initial decrease is well known (5_, 4_), another important factor in emergency planning it the time from initiation of an accident sequence until start of an actual release, since this affects "warning" time. Reference 5 indicated this to be from "0.5 hours to one day," without further elaboration. More recent work (2_).has generally confirmed this range for the most severe releases (SST1), but provides additional insight by indicating that for most releases, time from initiation to start of release is about 2 hours or more. Tables 1 and 2 shows results (2_) for the Grand Gulf and Sequoyah reactors. For the reactors studied, the conditional probability of a severe sequence being a fast-developing one, that is, where the time prior to release is less than 2 hours, ranged from about 3 to 30 percent, with a typical value of 15 percent.
- Hence, an important conclusion srising from this work is that most (about 85 percent) severe accident releases would take longer than about 2 hours from initiation to release.
Table 1 TIMING OF SEVERE RELEASES FOR GRAND GULF PROBABIL!TY OF CORE-MELT SEQUENCES GIVING RISE TO SST1 RELEASES = 3.4 X 10 EVENT TIME OF RELEASE SEQUENCE PROB. (MIN.) PROBABILITY TC-S 5.4 x 10-6 90 min. I 0.16
! TPQl-l 5.3 x 10-6 1190 min. 0.16 TQW- i 1.8 x 10-5 1340 min. 0.54 SI-5 4.6 x 10-6 1400 min. 0.14 l
l i l l l l 128-2
.. ,o Table 2 TIMING OF SEVERE RELEASES FOR SEQUOYAH ,
PROBABILITY OF CORE-MELT SEQUENCES GIVING RISE TO SST1 RELEASES = 3.6 X 10-5 EVENT TIME OF RELEASE SEQUENCE PROB. (MIN.) PROBABILITY V 5 x 10-65 38 min. 0.14 S H-3 2 x 10 6 110 min. 0.55 S HF- 8 5 x 10 6 197 min. 0.14 HF- 8, E, 219 min. 0.085
-6 3 3 xx 10 10-6 238 min. 0.085 TIME VERSUS DISTANCE TO RECEIVE A DOSE AFTER RELEASE Also of interest is the time for an individus1 at a given location to receive a given dose after a release connences. Such a time-dose-distance relationship provides insight as to the relative degree of urgency for response at different distances. This information was developed by generating data (7_) on whole body dose as a function of distance with time after release as a parameter. This was replotted to show dose as a parameter. Examples for SST) and SST2 releases f for adverse weather conditions are shown in Figure 1 and 2. For the most severe release under adverse meteorological conditions, an individual at a distance of 1 mile would receive a potentially life-threatening dose of 200 rem to the whole body about one-half hour after the beginning of the release. In contrast, individuals located 5 and 10 miles away would require exposure times of 3 to 10 hours, respectively, to receive the same dose. For less severe i releases consisting primt.rily of noble gases (SST2) for the sameimeteorological I conditions, doses of 50 7em to the whole body (the threshold for early injury) would be received by an individual at 1 mile about 1 hour after release, while an individual at 4 miles would require 8 to 10 hours to receive the same dose.
For the least severe releases ($$T3) doses (not shown) above the lower level L Protective Action Guide IPAG) value of 1 rem whole body would not be expected l beyond distances of about 2 miles. These results show that individuals at close in distances not only would receive higher doses, but that they would do so within a shorter period of time. j . t l 128-3 I i A
~ - - - - - _ - - _ I 0e I 1 m m I o o m o m I s R R o R
0 0 0 R 0 0 2 1 5 5
= = = =
D D D D I I s P I I I 7 H S f 0 I T A
, L E
l I s R E S C 1 , N
- E A
- L I T
; S f
M , I D
-. 5 EC -
l I I E W S N O h A D p T - E m S I M I r 5 y D T o y i t I t c t i i v l 4 e l i t e b c R a A e t n S amo w k 1 W n "F e M o
- r h o r N Se l
i
- t l 3 u 0 g 2 ar r 1 e :
r s i S 1 D u ht o o - F N . a t D - O se e e c y I T P l a se WFa d e el t g o l M U Re eR n on t B i e 2 S T 1 r
. e i n e l d loh S S H o h A S 2 C S W l 1
- - - - - ~ - - - - - -
3 2 1 0 s S 5 5 4 3 2 3 O 1 1 1 1 m$0I 4m(Um{0 fEow810 a w2P U.a
ll I\I li 31 b
~ *
- I - - - - -
- - - 0 e
m-o I 1 R 0 1 n w = n o I o h h D 1 9 i r t p S
, o cua e:m oi t r r rys e 3
'N ,eaDth5 a. tvd t
o
. O eRs eey s t ci d atc y T ae e o Wibeg I i l
e F Aod I P lWe lo t 1 8 RM Re natapnaB l M si U 20 S T2 r
. tS s" d d rm leo nl m
o S S11 H o~ nF i e h R A S- 2 C~ WShNi o W P I I 5 H 2 i 7 S
=
N O I D T
.. A L
E R I I 6 E C N
- S A E T L
I S I M D
. E I
I 5 E ; C 0 N 0 A E T M
" S I
I T D I I [- I 4 2 e r u g I I 3 i F m o '
.* R -
0 0 . _.~ i 1 1 2
- D m
~ o R
l 0 ' 0 i 1 _ 2
=
D - _ U 2 1 1 1 0 1 S g 7 5 5 4 3 2 1 O sK a g z iu Jn<88 E *O j Ez O a 2oEs y G g?*
3
.. e, .
RISK INSIGHTS ! A major conclusion is that the 10-mile plume EPZ represents a region not only where risk varies significantly in magnitude, but where the need for pr8tective ) actions at close in distances would have a greater degree of urgency, as well, I in the event of an accident, because of timing considerations. In particular, the region within about the first two miles of a reactor and a time frame within about two hours after accident initiation appear to have some significance. A distance of about 2 miles has significance since: o For many core-melt releases (SST3), projected doses would not exceed the Protective Action Guide (PAG) levels beyond this distance; e For most core-melt releases (SST2 and SST3), projected doses would be unlikely to result in early injuriet beyond this distance. It should be noted that an early evacuation within a distance of two miles would not be sufficient to avoid life-threatening doses beyond this distance for the most severe core-melt releases (SST1) under adverse meteorological conditions, and that actions beyond two miles would be necessary in such cases. A time frame of about 2 hours has significance since: i e For SST3 core-melt releases, response times well in excess of this value would be available before projected doses would be expected to exceed EPA
, Protective Action Guide (PAG) levels.
I e For most core-melt releases, an evacuation within this time' would avoid ,; doses capable of producing early in,iuries. e For the worst core-melt releases, warning times (prior to release) of about 2 hours or more are predicted for most (about 85 percent) severe accident sequences. PROPOSED PROTECTIVE ACTION STRATEGY Because of the significant spatial variation of risk as well as timing considerations given above, a protective action strategy taking these into consideration should more nearly meet the desired protective action objective listed earlier. An emergency planning and response strategy intended to emphasizepromptaction;inthehighestriskportionoftheEPZ,whilestiy maintaining planning as well as the flexibility to carry out actions in the remainder of the EPZ, has been given the term "graded response." Since the 128-6
r magnitude of a given release can vary significantly in severity but cannot be readily predicted prior to its actual occurrence, a precautionary evacuation to
, reduce the risk within the highest risk portion of the EPZ shows the greatest urgency and ought to be carried out within a time frame considered likely to avoid non-stochastic health effects. Protective actions (both evacuation and sheltering) may be necessary beyond two miles, but appear to have a somewhat lesser urgency.
'I A graded response strategy based upon these risk insights appears to be a two-step strategy as follows: I In the event of a molten or degraded core condition, or upon declaration of a General Emergency:
- 1. The imediate evacuation by everyone within about 2 miles, to be
- accomplished within about two hours or less as an objective, should i be recomended unless local weather or institutional concerns make evacuation infeasible; everyone within the remainder of the 10-mile EPZ should be advised to seek available shelter and await further instructions.
- 2. Accident assessment should continue, with monitoring of both plant and field conditions, and additional actions, including evacuation or relocation, as necessary, should be recomended for persons in the remainder of the EPZ.
EVAL.UATION OF STRATEGY The graded response strategy outlined above was evaluated by determining the ' dose consequences to an individual from a spectrum of core-melt releases. The probabilistic dose distribution to individuals located 1, 2, 3.g 5, 7 and 10 miles away from the reactor were obtained using one year of meteorological data to display the full variation of meteorology and distance over the EPZ. Results show that for most core-melt releases (SST2 or SST3), an early evacuation out to 2 miles and sheltering elsewhere in the remainder of the EPZ (in a one or two story wood-frame house vithout a basement) with relocation , within 4 hours of ground exposure results in no early fatalities and very low risk of early injuries, as shown in Figure 3. For the most severe releases, shown in Figure 4, early evacuation to 2 miles is insufficient to preclude early fatalities; however evacuation out to about 5 miles (necessary only in the downwind sectors), with sheltering elsewhere and relocation withih 4 hours of ground exposure, would generally do so. - 128-7 i
]
I 1 1 I s 1 I I II I I I I I I IIl I I IIIl 1.0
*1 story wood-frame house, no basement 2 t t I -
c I _ E 10 7 5 3 2 1 Mile g $ oc 10', - in E
~
~
Arrows indicate mean values i e n i ainI e n ii : i e nie 10-2 1.0 10 10 2 103* DOSE Rem Figure 3 Probability of an Individual at a Given Distance Receiving a Whole Body Dose in Excess of the Value Shown, Conditional Upon Occurrence of an SST2 Release and Shelter
- Plus 4 Hrs.
Ground Exposure 1 l
. O f1 e
_ - - - - - - ___- - - - _ i u - 3
~ _ I l
a 3 I V . e er I i hu t I s fp o e I
" I ox sE I sd en cx u I
" Er o -
l e i n.Gs
" i i er I
M sH o D4
"' 1 du ys
' ol
' BP l t fj e*
l r I i m htoe o Wl R ahSe i 2 e
- e . g 5 E nd I
Sinva I. 3 e Oi ee D cs ea Re Y l D ee cR OnaI BtTsS 5 i E iS D t L n Oena H vf io 7 2 WGace s l 0 n [ e i 1 te ar i 0 l u i r l u s 1 a e ac i t v dO uc _ n i i vn i a e io i e - dp nU m i I l e na t a no a e fi i i c oti d yd t t n i n io i l C i s i b , w an b w
. o r oo rh r PS
-_____ A _ 0 1
i 2 _ 0 0 0- 1 4 1 1 e r Gs_m4$mn. i u g F M;,
i CONCLUSIONS It is concluded that the proposed graded response strategy takes suitable account of the spatial and temporal risk variation within the plume EPZ.. It places priority of response on the highest risk portion of the EPZ, while - retaining flexibility in the remainder to accomodate a complete spectrum of accident releases. Although the specific response strategy outlined in this paper is based solely upon WASH-1400 releases, it is clear that the conceptual approach arises from fundamental considerations of risk prioritization and is therefore readily amenable to possible revisions in accident source terms. Hence, the concept of graded response is under serious consideration within NRC as a means by which emergency planning and response requirements may be evaluated, depending upon , the outcome of source term developments. i [ REFERENCES
- 1. Code of Federal Regulations Title 10 Part 50.47, "Domestic Licensing of Production and Utilization Facilities," January 1984.
- 2. Reactor Safety Study Methodology Applic.ations Program (RSSMAP),
NUREG/CR-1659. Vols. 1-4, Sandia National Laboratories, February 1981 - May 1982. I
- 3. R. Blond, et al., The Development of Severe Re'.ctor Accident Source Terms:
1957-1981, NUREG-0773, U.5. Nuclear Regulatory Teietssion, November tysz, l j 4. D. C. Aldrich, et al., Technical Guidance for Siting Criteria Development, g NUREG/CR-2239, Sandia National Laboratories December 195z. I 5. Planning Basis for the Development of State and Local Government Radiological Emergency Response Plans in support of Light Water Nuclear i Power Plants, MUREG-0396 U.S. Nuclear Regulatory Commission, U.S.
. Environmental Protection Agency, December 1978.
- 6. Criteria for Preparation and Evaluation of Radiological Emergency Response Plans and Preparedness in support of Nuclear Power Plant, NUREG-0654, U.S.
Nuclear Regulatory Comission Federal Emergency Management Agency, , November 1980.
- 7. D. Alpert, Private Comunication, Sandia National Laboratories, October 1983.
128-10
~T
- s. e...=
['; , "-;) ' C,.- . , .' dW Commission of the European Communities ; . s *b p, +l. , L Le rad.ia tion protection lg
.k .$sp. -jK Sdf(@
Workshop on methods for assessing AkkU the off site radiological consequences. !
;j of nuclear accidents ..
Proceedings Mid4J b y;\ 15-19 April 1985, Luxembourg @r g; i.fy.
"M n.
" 7 .,5
..~ - "
Edited by: F. Luykx Directorate-General Employment, Social Affairs and Education Health and Safety Directorate (Luxembourg) J. Sinnaeve Directorate General Science, Research and Development Biology, Radiation Protection and Medical Research Directorate (Brussels) E p Directorate General W k@g{fg',$g$ Science, Research and Development - ' " ~ - Directorate-General Empbyment, Sooal Aff airs and Educaton Health and Safety Directorate EUR 10397 EN 1986
'N l
i
.-_ - - -, -.. -_ - . . - . _ _ . - _ . _ _ _ _ - - _ - . . - . . - . ~
i AN ASSESSMENT OF REGULATORY APPLICATION be presented and dis-IN THE U.T.A. OF SOURCE TERM RECEARCH une will not end here, rthodology further on
- rem ins extremely etext. Indeed the ibutions, urban ,
ing within the M W r Europe impose a r L. SOFFER. J.A. MARTIN Jr. and R.P. GRILL vours to contribute national collabor-of the effc,rts of 4 work has teen 'j )odologiesare.m i l A 2 1 U.S. Nuclear Regulatory Cocroission Washington - UNITED STATES OF AMERICA 4 . 1 I 9
)
f d 1 l i [ f. j 53 - ,
~ - _ _ _ _ _ _ _ .
ABSTRACT I i a
.I Ti.e title of this paper is somewhat premature in ti at the assessment of regu-is still ongoing.
latory applications in the U.S. of source tern research However sufficient information has become available to make a tentative iden-tific.ation of sone of the areas affected and to examine a possible approach in one area, emergency planning. e t I i
)
m
.i i
i I 4 e I I t i 55
x *" (w :. _ 1 #' 4 4 BACKGROUND , , .. Source term research results are beginning to become available for review by f. u. .
., s ,
During 'recent months we have seen fg? . studies by the American Nuclear Society ( ANS{g result publication of J-f + interest % , parties. nd the 4 ' BM1 2104 of the peer review by the American Physical Society (APS) 3) , a . The results of . J 4.9 these studies to date, while not at all easy to assess, appear to have [i!l,, . : ;- Somemajor of the '! 3
,x and far reaching implications for off-site consecuence assess.ent. ,75 .;.i4,.
major perspectives emerging froa this research re: ' s
. - Source terms appear to be highly specific to the plant studied and to the f j 1.
accident sequence postulated.
.4 ,
?..l .,
r,z ~. , p' - iQ 4.g -%s l+'
- 2. Many release fractions of specific gioisotopes apoear lower than those predicted by the keactor Study (RS$1 . Q[
m 1 ' ~ :.~ , ' , I{3 3. Some release fractions are predicted to be higher than thote in the R$$. j 9
~
- 4. The release rates of various nuclide groups are more time (and sequence) : "
' dependent than previously predicted. -
e The release of significant amounts of molecular iodine is unlikely. y/ s.
- 6. Containment performance, i.e., the tirting and mode of containment failure g
is a key factor. L. ' ' {F(, ., y -;
.a The NRC is intensively engaged in studying the results of source term research %-Q,
~
to date, and in preparing its assessment of research. This assessment is A 6".'. V' scheduled to be presented to the Comission in May 1985 At the and sameis totime, be published the NRC s: ,e # as draf t NUREG-0956 by about the end of May 1985. staff is assessing the regulatory applications of this research and is develop- 'f
'i .f 4?
-~
ing a plan for the orderly innlementation of source term research regulatory process. into ti,e
- gr g' ;. .
, e
$..e.}.,-{
ll */- l' 1& >.J' ,- ?.. c- K d .<:
& y ?-
- '. #' .f';
.. 'n; $.k- , 73 eq 0
c !' 'g V y 1 , , '[ . r (. : g,
** k. , , {~
a 7
'y ,
4 ' ' rq g i y '.; .
; .6, " .
h I'c. l ;& si * 'Y
.T E.
', , . , _ ,}, ; f.: y p V.,. t ( ' ' -
$. = f
- r. _y1. _ ..} - :;
" : 'sn .( -
. -> , 'L P ),s> : - ..s.
.: -4 <. ' . {. := . ..- -L
- Ju .'. ~..
'* ,.l . . ( ,*%
) l i
a
'; . [ ,', \fp ;,j , f ; ; , * 'fi \' 'f .
l_ .' , ,, ) , .:. ' ., ; ; - f .' % ' .- q; j p ).g; g } h.;:} y , y+ .g,.,., ;-[:s e.R f'. j f .j j,q-
.}. . f
. ; ' L: ;- .
.... t , , . . , . *
,, ;... ? Q4: ' [ [
l" 4
~_, i This plan is also expected tc be presented to the Comission with NUREG-0956.
Some regulatory areas which may be impac,ed by source term research includes the f ollowing: 1 Cortainneet leak tightness i d irtegrity.
- 2. Engineered safety features performance.
- 3. Safety Goal implementation.
4 Equipment Qualification.
- 5. Siting.
- 6. Control rorm nabitabi'ity.
- 7. Emergency plarning.
Although a large n mu ts> r of regulatory areas may be affected, any regulatcry changes arising f rom recent research must anait further study and must be given very careful consideration sirce it is essential that the complex phenomena be well understood and that the uncertainties in our understanding be f actored appropriately into such decisions.
" EMERGENCY PLANNING One arer which has received a great deal ef attention is emergency planning.
the c of "graded response" has received some attention In particular,the recently in U.S. gcpt especially as a possible result of reduced source f . terms. Althcugh it is premature to speculate uoon the outcome of such research 4M upon emergency planning, it is important to note that a "graded response"
-?
h4J .' concept can be developed even for the severe accident source terms of the Reactor Safety Study. NRC g rgency Planning continue to berae source gu l a't iofonsome s , which were significantly controversy upgraded in the United in States. 1980 , This may arise in part f rom the misperceptier of uniform accident risk within the plune exposure Emergency Planning Zone (EPZ). Since g rglga
- g n of the r' showing regulations, additional risk stucies have bee" performed significant spatial variation in risk over the EPZ, which implies that phased or graded planning and responses, implicitly (ff*cg' ized in the documents which formed the basis for the present regulations , may be a reasonable strat-
- h)tip egy. A study was initiated to investigate this.
An objective of the study mas to develcp a protective action strategy capable of dealing witt. a wide spectrum of accidents. Such a strategy should be flexi-
~? ble, depending on the nature of the accident, and should provide a priority ranking of desired actions, rather than a pre-selected fixed risk objective, or dose critetlon, regardless of accident severity. The priorities, in order, should be to: (1) avoid early fatalities, (2) reduce risk of early
., ._ j'.
i 58 i
- e with NUREG 0956. injuries, and (3) reduce risks of other health effectsII3'I#I. Further, the research includes response strategy should recognize the planning that is both necessary and
?
feasible for the three phasg, 4) emergency response, i.e., early, intermediate and late or recovery phases s This study made use of severe co U.S. as Siting Source Terms (SST)g'%t . accident Releases releases ranged f th referred rY 5""'to in the 55T1 release with an estimated probability of about 1 x 10*5 *per 're'a'ctor-year, to ghe more moderate 55T2 release with an estimated probability of about 2 x 10' perreactor-year,agdtherelativelybenign55T3releasewithanestimated f probability of about 10 per reactor-year. TIME FROM IN'T! ATING EVENT TO START OF RELEASE While the spegl yfriation j in dose or risk, showing a sharp initial decrease, another important factor in emergency planning is the time is well known from initiation of an accident sequence until start of an actual release, since this affects ' warning" time. Reference 11 indicated this to g )from *0.5 hours
, any regulatory to one day " without further elaboration. More recent work has generally nd must be given , confirmed this range for the most severe releases (55Tl), but provides addi-lex phenomena be i ticnal insight by indicating that for most releases, time from initiagn to ding be factored start of release is about 2 hours or more. Tables I and 2 ,how results for f the Grand Gulf and Sequoyah reactors. For the reactors studied, the conditional prcbability of a severe accident sequence being a fast-developing one, that is, where the time prior to release is less than 2 hours, ranged from about 3 to 30 t percent, with a typical value of 15 percent. Hence, an important conclusion rgency planning, i arising from this work is that most (about 85 percent) severe accident releases some attention would take longer than about 2 hours from initiation to release, f reduced source of such research Table 1 graded response" ce terms of the TIMING OF SEVERE RELEASES FOR GRAND GULF ly upgraded in United States, dent risk within PROBABILITY OF CORE-MELT SEQUENCES GIVING RISE TO 55T1 RELEASES = 3.4 X 10 Igagn of the showing EVENT TIME OF RELEASE lies that phased 3 SEQUENCE PROB. (MIN.) FROBABILITY documents which #
easonable strat- TC. 90 min. 0.16 5.4 x 106 1190 min. 0.16 TPQl- 5.3 x 10 5 TOW- 1.8 x 10- 1340 min. 0.54 strategy capable SI-6 4.6 x 10-6 1400 min. 0.14 should be flext- . vide a priority l sk objective, or i L; ties, in orcer, J f: risk of early
,t 1
59 - i
__ a t- Of 4s Table 2 4 era ;
.. l - /t rea API -
T I Ml % OF $EYERE RELE ASES FOR SEOUOY AH
, - . '. Ac I
~
[ PROBABILITY OF CORE MELT SEQUENCES GIV NG RISE TO $5T1 RELE ASES 3.( X 10 e EVENT TIME OF RELEASE - PROB. (Mlh.) PROBABillTY R SEQUENCE o 1 . 38 min. O.14 v 5x 10 0.55 5, H- 2s 10 110 min. It _; 5x 10 197 min. 0.14 Si HF - ea 0.085 S' HF- , 3x 10' ' 219 min. 0.Cp5 as Thl- 3 10 238 min. co sa L 5- : 4* h j i - TIME DOSE DISTANCE RELAT]QNSHIPS A k Also of interest is the time for an individual at a given locatior to receive a 0 f-
~
L* glven dose after a release comences. Such a time-dose-distance r(lationship g g provides insight as to the relative degree of urgency for res g eonatwhole different 2 gt distances. This information was developed by generating data body 3 dose as a function of distance with time af ter release as a parameter. This 0 J was replotted to show dose as a parameter. E;amples for SSTI and SST2 releases i f for adverse weather cenditions are shown in Figures 1 and 2. For the Nst T severe release (SSTI) under adverse meteorological conditions, an individual at 0 a distance of 1 mile would receive a potentially life-threatening dose of 200 h
/
( rem to the whole body about one-half hour after the beginning of the release. d y In contrast, individuals located 5 and 10 miles away would require exposure I
? times of 3 to 10 hours , respectively, to receive the same dose. For less d
,1 severe releases (SST-2), for which ncn-nob.e gas radionuclide releases would be [ H significantly reduced, and for the same meteorological cenditions, dcses of 50 '
4 , t rem to the whole body (the thresholb for early injury) would be received by ar b i
' ' individual at 1 mile about I hour after release, while an individual at 4 miles ' -
would require 8 to 10 hours to receive the same dose. For predominantly noble Plume Exposure gas releases (SST3), doses above the lower level U.S.E.F.A. ' l Protective Action Guide (PAG) value of I rem whole body would not be expected ' 9 beyond distances of about 2 miles. These results show that individuals at close in distances not only would receive higher doses, but that they would do ' 7 50 within a shorter period of time, j f J- l ' e ., RISA INSIGHTS "' j j A major conclusion of this study is that the 10-rile plum EFZ represent s a ( -- - region not only where risk varies significaetly in magnitude, but .here the L--
;k L.'
need for protective actions at close in distances would have a greater degree _ jr . h s . '
'{
~
'idRPAh :
z L J T I j J t. E
-- _ )
1 of urgency, as well, in the event of an accident, because of timing consid-
. erations. In particular, the ' region within about the first two miles of a i
reactor and a time frame within about two hours after accident initiation appear to have some sigaificance. 1 A distance of about 2 miles has significance since:
' 3.6 X 10 5 For many core melt releases (SST3), projected doses would not exceed the o
; Protective Action Guide (PAG) levels beyond this distance; LBILITY t t o For most core-melt releases (SST2 and SST3), projected doses would be
).14 unlikely to result in early injuries beyond this distance.
).55
),14 It should be noted that, for the most severe core melt releases (SST)), an 1.085 early evacuation within a distance of two miles would not be sufficient to 1.085 avoid life-threatening doses beyond this distance under adverse meteorological conditions, and that early protective actions beyond two miles would be neces-sary in such cases.
A time frame of about 2 hours has significance since: l ln to receive a G relationship ( o For SST3 core melt releases, response times well in excess of two hours e at different would be availabic before projected doses would be expected to exceed EPA on whole body Protective Action Guide (PAG) levels. Pameter. This lSST2 releases o For most core-melt releases, an evacuation within two hours would avoid For the host doses capable of producing early injuries, individual at For the worst core-melt releases, warning times (prior to release) of g deso c.f 200 o the release, about 2 hours or more are predicted for mo;t (about 85 percent) severe
.uire exposure accident sequences.
se. For less
-ases would be
, doses of 50 P.ROTECTIVE ACTION STRATEGY IMPLICATIONS cQived by an
.al at 4 miles Because of the significant spatial variation of risk as well as timing consid-
=inantly noble erations given above, a protective action strategy taking these into consij-eration should more nearly me't the des. red protective action ob.iective listed lume Exposure t bG 0xpoeted f earlier. An emergency planning and response strategy intended to emphasize ndividuals at i early and prompt actions in the highest risk portion of the EPZ, while still Qhoy could do l maintaining planning as well as the flexibility to carry out actions in the i j remainder of the EP2, has been given the term "graded response.' Since the magnitude of a given release can vary significantly in severity ht cannot be i readily predicted prior to its actual occurrence, a precautionary evacuation in the event of a core melt a:cident to reduce the risk within the highest risk represents a portion of the EPZ shows the greatest urgency and ought to be carried out with-
$ut wtiere the -l in a time frame considered likely to avoid non-stecnastic health effects. Pro-bater tegree tective actions (both evacuation and sheltering) may be necessary beyond two miles, but would have a somewhat lesser urgency.
)
i q f 61 -
A reasonable graded response strategy based upon these risk insights appears, of t to the authors, to be a two-step strategy as follows: ' More-woul In tha event of a molten or degraded core condition, or upon declaration of a ' actu General Emergency: Alth
- 1. The imediate precautionary evacuation by everyone within about 2
- upon miles, to be acconplished within about two hours or less as an ! arts EgEEntgg p objective, should be recomended unless local weather or institutional } read
. . . - concerns make evacuation infeasible; everyone within the remainder of 1 conc F ,. . w *. .f' the 10 mile EPZ should be advised to seek available shelter and await mean
- ! further instructions. These actions for the early, imediate emer-N.
q j% .. . .
.M T l gency response phase should be predetermined and form the basis for determining the actions to be recomended for the following REFE Q f -[g g ," '
i.- . intermediate phase. 1. 'm)G 2. Accident assessment should continue, with monitoring of both plant gg.15 ' and field conditions, and additional actions, including evacuation or 2. A > relocation, as d6emed necessary or desirable at the time, should be %gp" $ 1' . recomended for persons in the remainder of the EPZ. [ . $*' j 3. ih. EVALUATION OF STRATEGY I d., ? The graded response strategy outlined above was evaluated by determining the 4
^x.Ql: 90 dose consequences to an individual from a spectrum of core-melt releases, s , 5 Probabilistic dose distributions to individuals located 1, 2, 3, 5, 7 and 10 5.
miles away from the reactor were ettained using one year of meteorological data @Mf % - lg:^ gg j.5_' to display the full variation of meteorology and distance over the EPZ. 0- Results show that for rest core-melt releases (SST2 or SST3), an early evac-R vation out to 2 miles and sheltering elsewhere in the remainder of the EPZ (in q fM 6 a one or two story wood frame house without a basement) with relocation within 4 hours of ground exposure results in no early fatalities and very low risk of I 6. early injuries, as shown in Figure 3. For the most severe releases (SST)), M{#4sNh' . " shown in Figure 4, early evacuation to 2 miles would be insufficient to pre- 1 d, clude early fatalities; howerver evacuation out to about 5 miles (necessary only 7.I($*4V6.:[I.
~:
in the downwind sectors), with sheltering elsewhere and relocation within 4 4 g;Ik.W"]:{ hours of ground exposure, would generally be so. 7. i V f .k.N 7' i CONCLUS!0NS 8. p pJ ' It is concluded that this graded response strategy takes suitable account of $ 2 g i';'-: the spatial and temporal risk variation within the plume EPZ, the basic ( P% radiation protection criteria, the various phases of emergency response, and j the couplete severe accident spectrum. It places priority of response on the 1 0%, i 9.
%;sj 9 highest risk portion of the EPZ, while retaining flexibility in the remainder 5
., 0.g ; ,~h
.. h
- mv? ,
*, ; -. .* k Mp"._ .
Y 62 - L
l 3 hts appears
- of the EPZ to accomodate a complete spectrum of potential accident releases.
Moreover, it's simplicity should provide a high . level of assurance that it would be easily understood during the planning phase and would work during an )aration of a actual response. Although the specific response strategy outlired in this paper is based solely lhin about 2 upon the WASH-1400 scale releases, it is clear that the conceptual approach less as an . rises from fundamental considerations of risk prioritization and is therefore institutional Jeadily amenable to possible revisions in accident source terms. Hence, the remainder of concept of graded response is under serious consideration by the NRC staff as a e and await means by which emergency planning and response requirements may be evaluated. late emer-basis for REFERENCES following
- 1. Radionclide Release Under Specific LWR Accident Conditions. Columbus,
,! Oh10: Battelle Columbus Laboratories, BMI-2104, 1954
/ both plant ; acuation or ! 2. Report uf the Special Comittee on Source Terms. La Grange Park, ,
, should be
{t Illinols: AmertCen Nuclear 50clety,1954 l
- 3. Report to the American Physical Society of the Study Group on Radionclide Release from severe Accidents at huclear Power Plants (Draf t). Icer1Can Fhysical Society,1955.
raining the 4 Reactor Safety Study. Washington, D.C.: U.S. Nuclear Regulatory Comis-ases* ston, WA5H-1400 (NUREG-74/014), 1975. 7 and 10 I J. A. Martin, Jr., Ob.iectives of Emergency Response and the Potential fo,gicaldata 5. e the Ep2. Benefits of Evacuation and Shelter, in: Proceedings of thi Eighteenth ,early ovoc- Mid-Year Topical Sy@osium of the Health Physics Society, January 6-10, 4 1985. lthe EPZ (in i . tion within l 6. L. Soffer, J. A. Martin, Jr., and R. P. Grill, An Examination of a Graded powriskof Response Strategy in Energency Plann,ing and Preparedness, in: Proceedings l ses (SSTI), of the International Topical Meeting on Probabilistic Safety Methods and >nttopre- . Applications. February 24 Merch 1, 1985, San Francisco, CA, Volume 2, page usary only ! 128-1 to 128-10. bn within 4 ;
- 7. Code of Federal Regulations Title 10 Part 50.47, ' Domestic Licensing of l Production and Utilization facilities," January 1984.
- 8. Reactor Safety Study Methodology Applications Program (RSSMAp),
account of hustfs/CR=lbb9, Volumes 1-4, bandla hattonal Laboratories, tebruary the basic 1931.May 1982. ponse, and
- nsQ on the emainder :
[ 9. R. Blond, et al., The Development of Severe Reactor Accident Source Terms: 1957-1981, NUREG-077a, U.5. Nuclear Regulatcry Comission. Novect>er 19sz. n iE N [
}
t - t i L 5 . i [ - 63 t
9
- 10. D. C. Aldrich, et al., Technical Guidance for Siting Criteria Development.
NUREG/CR 2239, :,andia National Laboratories, December 195z. 4
- 11. Planning Basis for the Development of State and Local Government ;
Radiological Lmergency Response Flens in bupport of Light Water huClear > Fower Plants, NUKEG-0336 U.5. huclear Regulatory Comi s s ion , U.5. $ . Environtrental Protection Agency, December 1978. M,,. . _ _ - _ = Jv
- 12. Criteria for Prenaration and Evaluation of Radiological Emergency Response {
Fians and Freparedness in support of huelear Power Plants, NUKLb-Ubb4, z= U.5. Nuclear Hegulatory comission, Feder41 Emergency Management Agency, & _ E November 1980. 1 i
- 13. International Comission on Radiological Protection, Protection of the Public in the Event of the Major Radiation Accidents: Principles f or Planning, ICRP Fublication 40, Annals of the ICRP, v. i4, n. 2, 1954 ~
N@l (Dec.).
- .' 14 World Health Organizatitn, Nuclear Power: Accidental Releases--Principles of Public Health Action Mio Regional Publications, Luropean 5erles No.
.( .
i lb, Eri0 Regional Of fice for Europe Copenhaven,1984 _ D. Alpert, Private Comunication, Sandia National Laboratories. October ) 3.
& 4A 15.
1983. 'j
- TM .?.? : ;-
e me s snaim _ W .
~
. .e ; ,n g
..l5, o ',i ;
** - h_f__
ff yf - } r
.s
+.)
hy II , ' 4 i -
%;;.>.,Q 6 f >!W i:
-J: :g.5.K .
s E %_.yil , MRIRllE MM
- 1 5_ s .
- r. . c.
h.. . j 's ' oelopment, I I I I I I I I l l
-2 N,.w; 1,1. '.
( w
.TY.s'('
s t Government ] ~ i ter Nuclear 1 sion, u.s. l; _. E. E
-=
3A i 8 1 E e E. E 3 g
- J t.5 se j l g d* 17..
h-cyResponse '
" E e 4 NUREG-0654, e e e .g gif f
- nt Agency, ,
O o o ga '- ' V
< -e y,5 h , .y , . t N,';',; :-/ . .
s ion of the i ,f g [ Ef
'1 iciples fo7 s, n- 2 * ,
.: 4 @ try y.;;
-~ -
.n. [ -Principles 5
-?
I ;- y .o L series No. f , y 2 [g*
> :s, October !
1 . m u is t s.,,
' ai g gg - J, . ,p.i
!_ g y 2 0 f
~
s -e dg $i[hs'f- )cOji, g g .. y e i j .. y E
- r. -
v P t 3 .# - 5 ;- ~'.N G r T
; I .s.. n. .y_
?]*
9 p -
- , . i b CI n.
". j .
8g 4 .: c . % gg - , , t -: ::
..o)xgi m e c. .,:;S. e.-:
=
9 ! ,N/. ..,9 y '. '. e-t t =$ o1> .n
- .-. .px j .e E 2 e g ~n
- a f.
d% e- $ ! 3Ef3 = g
'~
IEEEE 1 [ RDmM i e g i i i
-~
i i i i i , , , , D D " S * * * * . , n , , , , F i g } SWOOH '3Sy313W JO ONINNID33 Wows mg 3
; t - 4 k.
? .
El I I I I I I I I i l l _ g [ I e
~
G a
- f Q -
. j'. g I >>
5 n a El E g ,g j .I _ . i r .I mE.li!F
< no s=*f .
r
- n
- a. g ,
j o P . 5 . u .
- . w .
gg -
%dy' -
la ls: _ c M. h$ ug
/
p.. .t 5 + f
-A -
3 . ij , 3 _ d '
')
f" W ' , ,, T ') _ ,, g:- - " c . i
. 7 7.; .
l,6 g
, f l
h8 $
"'" " - f Ib -
h / m
- ~x :
D D ; 3 a = * = = , a n . . j 4 _. SWOOH '3SV313W JO ONINNIO38 WCW A 3Wt1 $ e 66 -
8 1 2 3 4 5 6 7 s s se DISTANCE, MILES Figure 2. TIME-DOSE-DISTANCE RELATION 5 HIP p*p -M J 5 I 5 5 g 5 II 5 5 5 5 3 5 35g 5 5 5 3 4 E ll] 1.0 _ .
*1 story wood-frame house. no basemant $
+ t i i -
l . - C
=f '
a
<E m 10 7 5 3 2 1 Mile O O 10-1 _
T
. a. -
l
- i Arroe indicate mean values n n i n i sent n
< n i i n' ! ' ' ' ' ' ' '
10-2 1.0 10 10 2 . gg : DOSE Rom fipwre 3 Probability of en Individual at a Given Olstance Asceiving a Whele Sedy Dese in Escess of , sne value sheen. Conditteast upea occurrence of an 55T2 Aslesse end Shelter
- Plus 4 Mrs. !
Crevad caresurs 'I I
{sti i i i i lsi s i i i i .M y k.
- _- 2
~
4 -
"l t
l -
%Il 3 - :35
- a j]t C
'5l
~
E 4 ~ E $ l e 8' 4 $1 _. 3 g ,,
- 1 . e -t e .
- E.js
- p=-
a 3 - cE ,~4, e
'W - r _ g73!
~
e 3;: g j2j O 3. I, .: z B'I i'
,.j e j.c.
~
r 3 s 4> a
% g G*' :4
.J p
.: o k.n,:.. . , , s -
. g .
- ' C- ,
- + E _
3)g hWI.)h'*4 ,. = _ - h - 2 g _ ,
- r j
^- -
+ g ,
.s
=]
. y ;l e 1.:
.?' it i i i i li e s i i i a d i o n
- o e' '
" S S ,g Ailll8V80Wd {*
' ~
i
,e f
r 68 -
r - - x , 4 I d.. ones w ROUTING AND TitANSMITTAL SUP 1/22/85 ,
- l me, emee sym6.#, reem num6er, oste q Tm inR6eis
~; ' , . / Poet)
,i.
,.:. 1.
@?j g G. Burdick
~
f.' g . J. Italaro
- .> 4 11. Ernst
..l, , F. Gillespie Action File Note and Retum ^.'
Approwel For Cleerence Per Conversation As Requested For Correction Propero Repey C6 testate X For Your information See Me Comment lavestigste Signature
- Winetton Justify asmAnns Received this fron H. Spector. It's a section of DOE report to Conoress, due now, that addressed oraded evacuation / shelter response. ?
DO NOT.' i form as a RECORD of approvets. concurrences, disposats. [ chorences. and samilar actions Room No.--81dg. FROM:(We, ces rymbol. Agency / Post) 344 til JMitartin e, n. n.. X37625 - 80u-102 OPDONAL POftti 41 (Rev. 7-76) f,
- cros ties c - asi-us unt>
PP** b *1N v
^e e
\
& /0 ,
y __
' ~
F~b M_J'7d., k 1 y{ Id ,
- 1. INTRODUCTION AND BACXGROUND ir O l Public Law 96-567 "Nuclear Safety Research. Development and N*d
?.
Demonstration Act of 1980." directed the U.S. Department of
$3 Energy (DOE) to undertake an accelerated and coordinated Q% program aimed at developig practical generic improvements to 4 enhance the capability for., safe, reliable, and economical .
4 F.i operation cf light water neclear reactor power stations. The [d,! specific roles of the DOE in this program are to provide the . overall coordination required among the DOE. the Nuclear K.** Regulatory Commission (NRC). the electric utilities, and the g.. nuclear supply and construction indu'stry and to manage and
- direct the department resources for implementing the programs. *
,y DOE efforts under Public Law 96-567 have been organized about *j
,j J i-the conceptual framework shown in Figure 1-1. In this scheme.
the top level goal is to identify practical generic improve- .* ments that would enhance the sconomics, reliability, and j safety of nuclear electric power production, starting from n. this goal, a set of top level functional requirements was defined along with a hierarchy of required supporting func- f'
.., tions. The top level functions are the following: '
- 1. To maintain the normal operating envelope (i.e.. assure
^-
that temperatures, flow rates, pressures, etc., are kept , within the design limits) L
- 2. To protect the core and plant against damage in the event l-P ' that plant processes deviate from the normal operating
- envelope -
. 3. To contain radionuclides released from the reactor coolant .
- r. ' system to assure that releases to the environment are '.
kept within acceptable levels for low-probability events ' when the core and plant protection functions are not t.
~
- maintained [
- i. .
jc 4. To provide adequate energency preparedness to protect the l * ' health and safety of the public, especially for those 1 extremely low-probability events when the radioactivity [ [F containment functlon is not maintained. [ [' These four functions also define the "states" through which v
- ,- an accident aust sequentially progress in order for public c
- health and safety consequences to, result.
Inordertoassureabalancedpershetive,theDOEassembled . industry working groups to address each of these top level g ',i'. functions. The assignment of the working groups was to pro- " Vide recommendations to the DOE on the need for additional I
- l. .
i'. . F 29 ' J
-w= - - ;
._y _
J. .. l,hp light water reactor safety research after careful conside;a-M r.
- sj tion of the following four questions: ,
8 h) . What are the relevant issues and their priorities? f,
}j.9 ]
- i
. - What information is needed to resolve each issue?
fE.'>
- 3 What of the needed information supplied by existing dome'stic U.S. for each issue will be Q'3 light water safety prograns?
[.Q . What information will not be obtained by existing "
.'. programs, and what are the priorities of these remaining information needs?
- S.i
+ _: Those issues that are judged to be of high priority and for '/O 2 which ongoing programs are judged inadequate to provide the ?;y f
needed information can form the basis for future prograa recommendations.
'L .~
This is the report of the working group addressing the C.. emergency preparedness function of the intsgrated approach conceptual framework presented in Figure 1-1. The scope and limitations of the effort described in this report are as follows: The study considers exclusively light water reactors, as directed in Public Law 96-567. The study primarily focuses on issues and information needs for unlikely accidents involving the loss-of-core-cooling sequences that progress to core damage. 1.1 .Purnose and f aDortance of Emeroency Precaredness Function Although the probability of a severe accident at a nuclear
- power plant is very small. and the likelihood of severe health effects is much smaller, it is prudent nonetheless to l
' have plans for protecting the public in the vicinity of operating plants.
l Therefore, the purpose of emergency plan- i ning for nuclear power plants is to reduce consequences ' [. should an accident occur at one of these plants. While some i nuclear accidents can have both onsite and offsite effects, this effort is principally directed at offsite consequences. Li
- Our task has been to determine what'the present research and j.-
- development needs are. if any, to idblement more effective o(; emergency planning.
- b. In order to accomplish this task, it is first necessary to consider the nature of the offsite nuclear
('.' . consequences, the ability of various energency responses to h ,* ' reduce these consequences, and people and their institutions !. required to carry out the responses. l. l , j 31
4 1 j,k 1.1.1 Offsite Consequences 7,9
.i With regard to the nature of offsite nuclear power plant con- ?-f sequences, these can be divided into economic consequences 6.: and health consequences. In most nuclear accidents that have i.1 been analyzed, almost all of the economic consequences would .
44 be onsite. Although thg possibility exists for offsite eco-nomic consequences, to 4 very large degree offsite economic
- f. f.. consequences are independent of emergency responses. For )
r1 l these reasons, offsite emergency planning is directed towards i {'[; reducing health consequences. l It is possible to divide offsite health consequences into two ' broad categories. early health effects and latent health
,. effects. Various studies have shown that, even for severe cc:
hypothetical accidents, nuclear power plant latent health t ,' l' risks are a very small fraction of nonnuclear latent health risks. fj These studies also show that most of these potential radiological latent health effects from severe accidents are due to radiation exposure that occurs long after radioactive f- material was released from an accident. Further, most of this exposure occurs beyond the Energency Planning Zone (EPZ).
' Because of these factors. latent health ef fects are only mar-ginally affected by emergency responses. It is therefore the early health effects that offer the most potential for reduc-tion through energency planning.
Early health effects from potential nuclear accidents include early radiation illness and early fatalities. In the entire L history of nuclear power in the United States, not one offsite early accident. fatality or early injury has occurred from any nuclear f. T, A great deal is known about health effects from acute radia-tion exposure, such as the radiation exposure levels required to produce an early fatality or an early injury. It is also 3 possible to calculate, assuming a certain amount of radio-
- active material has been released to the environment, the l, likelihood of causing an early fatality or injury at a parti-L', cular distance from a nuclear plant. Analyses have shown that the early f atality ris'k. by f ar the more important health con-L
- cern. is limited to a small area near the power plant. For example. 95% of the early-fatality risk from a nuclear power I" <.
plant following an assumed accident at a high-population site was calculated to be within 4 miles of the plant, assuming
' ~ people were exposed to high levels of radiation for 24 hours
% _ - and took no special precautions to protect themselves. Here 1.'.' it was ar,aused that 70% of the reactor's radiolodine. the
' dominant contributor to early fatalities, was released to the
,, environment along with other radionuclides.
3
.~ -
32 ,
,~
.r ~
j; ' il Another characteristic of the early-fatality risk is that it l'.l is very sensitive to the amount of radioactive material (par-Of Liculatty radioiodine) that is released during an accident. ic;( The smaller the amount of radioactive material released. the
<.4
- i. r,c smaller and closer to the site the area over which 95% of the early-fatality risk occurs, stated differently, the number
, '.t of potential early fatali$ies rapidly decreases, as does the '
VS distance over which they gight occur, as the size of the t'. j j radioactive release decreases. Smaller radioactive releases y: also reduce all other offsite economic and health conse-d quences. If only about 1 to 2% of an SST-1= source term is k'T/
./
released during an accident. there will not be any early fatalities even if no emergency responses are taken. (It is
.J usually assumed that almost all of the noble gases escape to ?.O the environment during a severe accident.) Many recent stud- . .J .G; ies and experiments show that almost all potential severe accidents at nuclear power plants lead to very limited radio-C '. iodine and other radionuclide releases, far below the level 9, . required to produce an early fatality. At the Three Mile Island nuclear accident the amount of radioactivity released ;f was very small, far below that required to reach the early , ,
fatality threshold. How much radioactive material is released during an accident? The answer to this question has been the subject of intense study, especially since the TMI accident. Although various experiments and computer program developments are still under- , way to improve the state of the art of what is called source-1113 technology. it is now possible to draw some preliminary conclusions: Previous source-term analyses for the most severe acci-
~
dents overstated the amount of radioactivity released. J.
- For those accident sequences where the radioactive aaterial from a melted-reactor core is kept within the
; containment for several hours. the eventual release is .
quite limited and well below that which would cause an ' early fatality.
.- The second conclusion above has additional implications. It
..' appears that the bulk of the calculated early fatality risk comes from core-melt-accident sequences where there is a l; - short time between core melt and the release of radioactivity L. !" to the environment. Even these rare accidents may not be as severe as previously thought. , i
.n ,
ji N
- *The ssT-1 source term is a large radioactivity release used l ', in various safety studies, see NUREG/CR-2239 for its deft- -
nition. !i P L. I. l 33
. _ _ _ _ _ _ _ . . _ _ -____i------------------
-- - - - ~ ' ' ' - - ' " ' - - ' '
, . i e
?)p A number of experiments have been conducted recently that N .3 simulated important aspects of one of thes) prompt release i 9?.f sequences and in which the amount of radioactive material '
?/} released was considerably below the levels that have been '
q'.) historically assigned to such accidents. It is anticipated *
'd that further ' experiments and analyses will demonstrate that
) aany prompt-release scegarios would not result in early
}'S '
fatalities. The calculated frequency of such prompt releases i at varies from one plant design to another, but they have been
.,19 shown to be very low. In one recent probabilistic risk e.j assessment (PRA), the frequency of prompt releases was calcu-lated to be less than one in a million per year of plant
?n.
operation.
!3 As stated previously, energency planning is also useful in I;j '
reducing early radiation illnesses (early injuries). As in the case of early f atalities, core-melt accidents are most ,y l ,' important in determining the early-injury risk and smaller *
- .,.? source terms mean fewer early injuries.
l:'d 4 1.1.2 Emergency Responses
- ,]
4 Within the EPZ there are two major types of emergency s responses to potential nuclear accidents. (1) evacuation, and (2) sheltering, followed by relocation where necessary.
, Evacuation, if implemented promptly and carried out effi-
.' ciently, can move people out of the zone where the risk is the highest. High evacuation speeds are not required: a few niles per hour (walking speed) is sufficient to eliminate the risk of an early fatality if evacuation starts before the release of radioactivity. The most important consideration in evacuation is timino. To be most effective, evacuatisn
.' should take place prior to the release of radioactive mate-
[- F rial. Evacuation also has certain drawbacks. If there is confusion or delay, there is the possibility that some evacuees will mistakenly-traverse the radioactive plume. thereby increasing their exposure. Since the early-fatality risk decreases
~ rapidly with distance from the plant while the population to be evacuated can. in some cases, increase approximately as the square of the evacuated distance, the complexity (compared E with the benefits) increases dramatically with the distance l- to which people are evacuated. Stated differently, evacua-I, tion near the plant is auch more beneficial and cost-effective
~
than is evacuation several miles from the plant. l Sheltering, followed by relocation where necessary, is also' an effective energency response strategy. Because sheltering
, is far less complex than evacuation. it is less costly to
, , maintain and can be implemented by the public with a minimum
- c. of reliance on energency personnel and systems. There is
.-l i s - also less likelihood for errors. such as inadvertent exposure
= , . n 34
a . - ... . . . l: a ! . i 1 - . a - - a 1% to a radioactive plume. This is because sheltering involves ' JJ simple, familiar actions with ample time to implement them. ; (.V such as remaining in one's basement until the radioactive plume has passed and. upon instructions. relocating to a safe , i area it a particular area is highly contaminated. This relo-j$,9 y! cation should be easily achieved because the contaminated chi area (plume deposition aqea) within the EPZ would be quite EJ. narrow; the distance fro ( the conter of this contaminated d '; area to its edge is likelf to be less than one mile. The r
}l. disadvantage to sheltering is that it entails some minimun l %*.s radiation exposure for the public in the contaminated area '
W { ,' , while sheltering,and while relocating. l sf. Studies have shown that all-evacuation responses (evacuation [7;,, out to the edge of a 10-mile EPE) and all-sheltering responses
~
.. can be about equally effective. It evacuation commences P.. promptly. then the all-evacuation response results in lower
": overall early-fatality risk than all-sheltering. If evacua- ,
Q,: tion is somewhat delayed, an all-sheltering' response could be i
- . more protective of the public. !
The best energency response is a synthesis of these two basic a '. responses. utilizing the better features of each. Consider the area around a nuclear power plant to be divided into ; three concentric planning zones.* The innermost tone can be
, a circle whose radius is the evacuation radius, i.e.. this is r
the area from which evacuation would be utilized. The middle zone is an. annular ring ranging from the evacuation radius to
., the present 10-mile EPZ radius. Sheltering, followed by relo-cation (where appropriate), would be utilized in this annular
,. ring. The cutermost area. beyond 10 miles, would use discre-a tionary measures as is the practice today. ! There are numerous adt 22tages in this graded approach to i minimizing the early ratality and radiation injury risk: ,
- By limiting the size of the area to be evacuated. the full force of the emergency planning system could first be !
l- concentrated on the people in the highest risk area, while
' l preserving the benefits of the simplicity of sheltering in the lower risk area.
2 . Evacuees from the innermost zone would be less likely to
' be delayed on the roads because people in the middle annu- !
lar ring would have already taken shelter, i.e.. stayed lq indoors and off the roads. ,
- r. .
~
k .
,
- Precise geometric zones are not required. Rather, zones 6 l f. should account for geographic features, political boundaries, i~ and the like. as in present emergency planning practice.
L. I 35 lt .,
. ,, a + -
j; l . L p . jf . tip Only a small portion of the middle annular ring would be hj affected by the release of radioactivity hp,; fying the affected area, emergency respon. seAfter forces identi-could concentrate on relocating shelterees out of the area. ?[.'f.d
- jf',
?;'/ Radiation annular zone. ringlevels would bg in the lowercontaminated th4n those in the area of the middle innermost '? : [G protective actions in'the midd1h annular ring before aTherefor
.J' particular radiation level was reachc6.
f.9,d - The cost-benefit ratio for nuclear energency planning f- would be much more societal emergency responses. reasonable and more comparable to other 3 '. . L;e'
,;l The size of the evacuation radius could be adjusted as source-term technology evolves.
i ,1, * [ Site-related refineme'nts could be implemented to optimize 9,', . the size and shape of.the area to be evacuated. It might
- - a site-by-site basis.be valuable to determine the optimum evacuation radius o
! ,~. At sites where there is a high population near the plant but with good sheltering capa- !.. bility (many smaller basements, evacuation area brick night apartment houses). then a (, i all early-fatality risk, result in the lowest over-
- l. sandy area, might have few people and few basements.other Here sites, such a the optimum that evacuation in the previous example. radius could well be larger than h
With also useful regard in to reducing early injuries, this health shelter'ing risk.and evacuation are would be utilized.rosponses dictated by minimizing the early-fatality riskThe energency O p l. and/or evacuation could be used on a discretionary basis.For non-core t discretionary conded because approach of the a to this lesser health risk is recom A Moreover, an emergency uch smaller release of radioactivity. place to respond to the more demanding core-melt. class ofresponse framewo accidents. enorgency planning. Discretionary actions are already part of present j, i boyond the present 10-mile This approach is now assigned to areas viewpoint, EPI.
,' discretionary actions should not have to meetFrom the a regulatory complian rigorous requirements assigned to prescriptive actions, such as a required evacuation capability in the innermost zone.
L Tho various Table 1-1 summarizes postulated n the principal characteristics of j ,, , coorgency responses.uclear accidents and their associated OI i. e 36
- ^ - ' -
.l- .
[q . Ju 3q Table 1-1 JD hh Nuclear Accidents and Their Associated Responses (NhI N. ..h Enercency Response '
?
Accident pner Annular Beyond .
.. Description Zone Region 10-M11e EPZ u
0, l.p.:. . (*{, Core-melt accidents Evacuation Sheltering Discretionary followed by actions relocation i
- 3. Where neces-
.. sary i.,
Non-core-melt Discretionary
- accidents actions Although present Energency Planning Zones extend to 10 miles and smaller EPZs could be justified, no recommendation to i
7 reduce the' size of the EPZ is given at this time. This very conservative size o,f the EPZ should be reviewed at a future date, particularly after source-term technology has evolved further. 1.1.3 People and Their Institutions
- A successful response to an energency requires effective.
', poritive actions by numerous offsite organizations, both gov-
~' ernmental and private. For most accident scenarios, offsite
. ' responses are geared toward bringing offsite assistance to s
, bear at the site (e.g., fire and rescue). However, for those '
care scenarios involving a major radionuclide release, the 4.,
' ' response the plant.
becomes focused on protecting the people living near In such cases, appropriate response by the emer-gency workers (including utility, state, and local government and volunteers), as well as by the individual residents them-selves, becomes important.
,~
As stated earlier, to minimize the early-fatality risk, timing At is the most critical factor in cgreying out a successful eva-cuation. By utilizing a graded approach in which emergency <
.~. forces are concentrated On those t o are in high-risk areas. '
,' a very important step would be taken toward making evacuations c, more timely, if they are needed .
s e o m 37 l _ _ _ _ _ _ ~ , _ _ _ __.__ _ . _ _ _ _. _ _ _ _ __ _ _ _ . . - . _ , _ _ _
pi J. - u _ ~~
. _J
~
,j ,
./ ,
I dj Further steps can also be taken. In an accident at a nuclear y; power plant, it is possible that the focal point of "who is
' 0.
, in charge" will be changing during the first few hours. If 27s
, the public is first warned to take shelter by one authority g'4 and then some time later is directed to evacuate by a higher O4 authority, this could lead to confusion and needless expo-
'9 sure to radiation. Thi concern is particularly valid during
?,^j . severe prompt releases ce the time of radioactivity 49 release would likely tal during the time the center of pfd authority is changing. l Os ! CEl To minialze potential confusion. all officials who address
- s the public should issue consistent guidance, and decision- '
'T making under stressful conditions should be minimized. This ];. calls for emergency plans that are rather prescriptive for ' ,, r core-melt accidents for which public pronouncements would
,p* have been.previously agreed upon. Simply stated. everyone should read from the same script for a given set of accident 7., conditions. ! i; Because time is of the essence for evacuation, the number of I steps and people involved between recognition of a severe-l ', accident condition and the sounding of the General Energency : 1 . alara should be minimized. The process for making and , announcing decisions' should be clear.and simple. t.f L A* key element in ensuring an adequate state of readiness for , l L', an emergency is the conducting of exercises and the training j of energency-response personnel. This would include training ' I for and exercises of a graded energency response for core- (' selt accidents. However, most nuclear accidents would not lead to serious releases. The challenge is to include the . )*', proper degree of real'sa and completeness to ensure readiness j
- - without overburdening the organizations involved. '
n. The current practice of including an evacuation scenario in r - every exercise may lead to preconditioning of emergency
- h. .
workers and the public to expect evacuation in all actual
- - 1 energencies. For example, the ability to make a valid deci '
L'. i' sion ngt to evacuate during an energency could be as important !... to energency preparedness as the ability to evacuate. Drille e J. should be consistent with our state of knowledge about nuclear ,t g risks: most nuclear accidents would require no public action. l.]1, ' some might utilize a sheltering response, and only a very small fraction would require a full-scale graded response. , vi In this vein. Protective Action Guidelines should be recca- ! N[,. sidered utilizing the new guidance provided by the Interna-
- a tional Commission on Radiological Protection (ICRP), along .-
r- with our present understanding of the nature of nuclear !
; risks. for training, exercises, and actual accidents. !
h, Public confidence is most laportant in assuring that the ; individuals take the appropriate protective action (go to
. 1 l -
i, 30 I ! I
----.,,----.m pw-.-- .-w, , ,n , -----r~_onn- - - - - ,---n--n- , , - - - - - , - - - - - - - - - .v-- - - - - - - - - - - - - - - - - ~
their basements when advised to shelter or depart in a quick l' 4 *J and orderly fashion when advised to evacuate). Although no p,S one expects every person to follow instructions (some will '
.. refuse to leave when evacuation is called for, and some will '
.: evacuate if told to shelter), the higher the level of public
,:O confidence, the greater will be the conformance to instruc-
#d tions.
'"" tion afand the ' greater w 11 be the degree of public protec-forded. The bull ng of such confidence begins long i(.' before any emergency arises, through the conduct of public ,
.; awareness programs which introduce the people to what may be
,2 expected of them. to whom to listen, and what to listen for.
.:'d:e 1 In summary, keep the public well informed, keep of f site
/
^
actions by the public as straightforward as possible, agree upon emergency announcements so that the public receives con-sistent statements, simplify the number of steps between 1// recognition of accident conditions and warning the public. reexamine Protective Action Guidelines, and practice drills (;, that are consistent with the frequency and severity of the i nuclear risks involved.
. ,' 1.2 Analyses of Emeraency Responses Analyses have been made to determine the characteristics of a representative graded response to a large, prompt release of H. radioactivity. These analyses show that an evacuation to a distance of two miles, with sheltering (followed by reloca-tion, if needed) from two to ten miles, results in a very low.
f. I. early-fatality risk. 1.2.1 Uniform Energency Responses In Figures 1-2 and 1-3. respectively, we compare evacuations from o to 10 alles against sheltering f rom o to 10 miles. These analyses concentrate on determining the mean probabil-p icy of an early f atality as a function of distance from the
, point of release. Such a determination was done for two source terms. SST-1 and one-third of SST-1 (with no reduc-N tion in the noble gases released). Several conclusions can be drawn from these figures:
The "all-evacuation (o to 10 miles)" response can be better or worse than the "all sheltering (o to 10 miles)"
- r. . response. depending on how long it takes to get people moving.
- 1. .
h.' . The "all-evacuation with a 1-@ur delay time" response Lv and the "shelter for 4 hours. Men relocate" response are W- ; v-l 39
- y"
. 0*1
.n i i
')
i i i 1 : : i g .
..s .
- 337 1 .
M f .' " 9( 11M MWe PWR NYC WEATHER a - i&, %
- d's .
b*= EVACUATION 84 DELAY .
... . \* '
EVACUATION. 3* DELAY .
\
J'. I .
""" SHELTER 64. RELOCATE ,
S*' -
=== SHELTER 44. RELOCATE - -
,,, 3 %
=
""= EVACUATION.14 DELAY
- s. 4 ,
p \
.f 2 0.01 .- .
3 c
~
. W
\ - - . - - . . -
~
- . ko
\ '\ .
. m is -
- . > \\ t
> \- .
3 g .
)
. < \
c o ' '
- e 0,001 a .- .
2 - t -
< ~
\ .
w !c- E . I
, \, .
i
/
\ .
l- 1\\ O.0001
' I ' ' ' ' ' ' ' ' ' '
(' 0 1 2 3 4 5 6 7 T !'. . . , DlJyNCE (ml) i - ' .j , - i .' l .' Figute 1-2. Mekn Probability of Early Fatality vs Distance i (SST-1) , ) I
i , . . t ) r... , 0.1 i i i e i I ' ' ' 9i ,
- l i -
'{"
. l.
I " 1/335T.1 -
,' u
. t-
- j. .I 1120 MWe PWR NYC WEATHER
[,$; *
"} e 3n
; ~
i %- EVACUATION, SW DELAY
, t EVACUATION. 3h DELAY *
, ./", l. .
.} ===== SHELTER 44. RELOCATE "
b ' a - II == =- SHELTER 44. RELOCATE - 4
>- 1 'e. - 4 ..- EVACUATION. th DELAY
- u. 0.01 - i .
' > ~ )3 -
- t. ; lg= -
d 1, w (1 - gu. q 1 i o -
\ -
> In I t *\
- t m
\j .
~ i 4 .
i
.\
m k.) o i I; [ c 0.001 - a - l
. ~
z ~ 1l ! 1 - w - 2 - I 1 _ \ g - 1 L 1i i - I L 1 -
!.1 )l i
O 0001
^
t 1,1 i n f , ! , f , I ' " i . ,' 0 1 2 3 4 5 6 7 DIST*J NCE (ml) , k I
." Figure 1-3. Mean Frobabi1Lty of Early Fatallty vs Distance 1
(1/3 55T-1) 4 +l L
Additionally, the increment in the mean rather similar.' 7,1,, ; early-fatality risk between a 4-hour and an e-hour shel-
< tering response is not very large. ; ~;i'.
At a distance of 2 miles, even for the larger SST-1
;.}A source tera, almost all of the benefits of evacuation * '
3 have already been actieved. At this distance and beyond. ji,, sheltering results 14 very low probability of an individ-ual becoming an early% f atality.
,1;' .
- $; use of a smaller source tera (1/3 SST-1) reduces the mean probability of an individual becoming an early fatality and L
the distance over which this might occur. 1: Figures 1-2 and 1-3 Plot the mean probability of an individ- - f' 7, ual becoming an early fatality versus distance, assuming that
.r.
a severe release has occurred. Since various studies indi-cate that such a release is very unlikely /3 in the range of once per 100.000 years of plant operation down to less than N. once per million years of operation, one can,get an appracia-tion of how small these mean values are b~y multiplying them by their calculated accident frequency. Table 1-2 shows, for two different source teras, at 2 miles. with an aasuaed accident frequency of once in 100.000 years. the mean probability of an early fatality per person per year of plant operation. .l . Table 1-2 Annual Mean Probability That Individual Will Become Early Fatality at 2 Miles from Point of Release . i.; !.. taergency Response SST-1
- h. 1/3 SST-1 h
Evacuation with 1-hour delay 3 x 10-' < 1 x 10-' ' f Shelter for 4 hours / relocate 1 x 10-* t- ,. < 1 x 10-' i / E
?
k
#'
- Delay time is the interval between the time when public .
V warning is given and the time when people begin to evacuate. in all cases. it was assumed that this warning came 0.5 }; hours prior to the release of radioactive material. ; i 42 9
_r
,. . . -l .:
.i ,.,* 3 1.2.2 Graded Faergency Responses
- A .
,,! Figures 1-2 and 1-3 were for uniform energency responses. '1 -
1.e., an all-evacuation or all-sheltering respo,nse from o to
- a. 10 miles. Figures 1-4 and 1-5 depict population risk using k-l graded responses for source terms SST-1 and 1/3 SST-1. respec- .
J '. . tively. In these lattergfigures, the area to be evacuated * - '
- c.- was examined using evacuapion distances of 1 2 and 3 miles .
'D
?' with sheltering for 4 houts (where needed) out to the edge of the 10-mile EPZ. In all cases, it was assumed that evacuation 1 would begin with only a 1-hour delay. In these analyses, a k.9 ?! uniform population density of 100 people per square mile was
',. selected. To adjust the risk numbers for different population
.i t densities. the results can be scaled directly. Thus, for 500 people per square alle. the resultant population risks are
~l. five times larger. Various refinements. such as a site-a specific optimization as discussed in Section 1.1.3 of this 0- report, were not analyzed. The principal input assumptions used in these analyses are given in Table 1-3. --
i . h
;' Table 1-3 Assumptions Employed for Consequence Calculations l
- 1. 1120 MW(e) PWR.
- 2. New York City meteorology. l
- 3. 10 miles per hour evacuation speeds.
- 4. Normal activities assumed before evacuation with the following s,hielding factors: '
Cloudshine factor 0.75 Groundshine factor 0.33
, Inhalation factor 1.00
- 5. Shielding factors for sheltering:
i cloudshine factor 0.50
, Groundshine factor 0.08 Inhalation faq or 0.50
- 6. 100 people / square mile.
i I
)
43 l
.. . ,e .',t , i.'s s . .g, , 5 5 5 y y y V Wg i V V V W Y v vg 4 5 *
- y , y y
" 3 007 1 ===== t as tvAcunftest teesktes etTOese
'. = tta *
'y .
, ,,,,s tswo Pwn,
,ga y,,g ===== M WAcua m eMsL m m ,
. ; : ', No Most*/awa - .. w EvacuAftoes, eseskism ettoreo
'N
- ==== in tet.t. t.ed evacuaftese. tMstfta stvosso .
S t
. *- 88 t,1 - .*.T ; ,.' ', . N.
m, .
==
. d -
t .
- .i t
, , a
. s I
i 4
.. l f
[ t .
, e c 1
. . ...I 10
\i,,,, .it 100
. . . _t > i . .
1000 I. EMt1.Y FATAUT188 4 F* i .- Figure 1-4. Graded Response for Source Term SST-1 1 . e t l
l .; i;: a
.: . ,,.t l s.Y
~ 1< -
A 11 4jt jJ
..g . . . . .
in est.t cancert meets eases:
. tim meepen *
- mec waaruse .
.jl tes8eesta ms .
.o .
+
. . v. J' = g, ,,,
.e .
A,l .
. ,s . .
g . .
. . i t* tvacuartom. sanisa sevome u 0.01 .- -
- a. set - - - I , . . . . ...t i 1 , , , , , , ,
1 10
. too i300 X, EARLY PATALf7188 '
~i t
e' \ a 1 Figure 1-5. Graded Response for source Tera 1/3 SST-1
- (
\
,m. -
" ~ss
, b
.T I
t i
.- _ -. .- . - . -~. -. ..
.
- 9;
. 1 I ' . . (> i Figure 1-4 shows that a graded response is a very effective energency response. For example, with a release frequency of 3t once per 100.000 years of operation, there is but one chance in ten million operating years of exceeding 21 early fatali-
~c9 ties" beyond.
it evacuation is out to 2 miles with 4-hour sheltering pp Ok , O'.9 Figure 1-5 demonstrates he enormous sensitivity of the early-
,6 f atality risk to reduced source terms. The dashed line in , c.?
Figure 1-4 is a replot of the data from Figure 1-5. With such a source ters, a 1-n11e evacuation distance results in (:!9 very low early-fatality eisks. With the 1/3 SST-1 source
?? tera. there is one chance'in ten million operating years of '
i, exceeding 2 early fatalities
- if evacuation is limited to 1 )
.li mile, with 4 hours sheltering beyond to lo miles. With eva-J.:
i-cuation out to 2 miles, no early fatalities are calculated. { 3 With regard to source terms for core-melt release scenarios, ke, .- ' the expected conclusion from various experimental and analy-tical programs now or soon to be completed is that the magni- ' tude of radionuclide releases would be a factor of ten or r Q ' more smaller than the releases used in current analyses, such as SST-1. i b* . 1 Based on the above analyses, and oven utilizing an SST-1 release. a graded response utilizing a 2-mile evacuation ' H distance results in a very low early-fatality risk and makes rapid evacuations more likely. Because of these principal ! conclusions and the realization that even these low-rist fig-ures are quite conservative, the 2-mile graded response is recommended. As source-term technology evolves. the evacua-
- tion distance'and the size of the EPZ should be. recalculated to reflect this knowledge. i 1.3 Descrbation of Eserceney Presareiness Subfunctions V:
h' The Energency reparedness function has been divided into four major sub nettons: , i l ' - 4.1 Plan to Energencies 4.2 Maintain eadiness , 4.3 Implement he Plan During Energere es 4.4 Recover fro the Energency Each of these can be f
, topics. These subfunct ther divided into sub unctional ne and. topics enable he energency activities to take place n an o(ganized, celia le. and effective fashion. Since he pu bose of this e oct is the l
- .' identification of issues. i formation needs. and this particular definition o D&D needs, i
-f heavily influenced by the des the subfunctions an topics was e to provide an ett tive
*At 100 people / square alle. -
to i - !}}