ML20053D562
Text
.
WASHt NGToN 25. D. C.
October 22,1960 is i
!er i
s$
Yh M
h Honorable John A. McCone Chairman U. S. Atomic Energy Cn r-isaion Washington 25, D. C.
I
Subject:
RFJCTCR SIE CRIERIA z
Dear Mr.14cCone:
Ycu have asked that we supply you with criteria which could be
' ~ '
used for judgin6 the adequacy of proposed sites for v. actors.
he Advisory Ccx:cittee on Reactor Safeguartis has devoted con-siderable time to this pmbles. A large part of our delay-in sub=itting site criteria stems from the fact that we believe it is premature to establish quantitative li=its on the variables v'
involved in site evaluations - especiany if such li=its vill
}
appear in Federal regulations, or otherwise be announced as l
Co::cission policy. We recognize that the correctness of the nu=bers which could be celected cov cannot be proved by experi-rental or erpirica* data, and, themfore, these numbers vcrald give a false sense of positivipess which could not be supported upon detailed cerutiny. Du=bers chosen now vin be expected to change as more infomation d,W.s.
For exa.:ple, a quantitative calculation of dosage east include some estirate of the fraction of the total fission pmduct inventory which cay be air-bome.
O his fraction is currently under experire.utal na-htion and the i
estimate may be subject to change.
he Ccc=11ttee believes that the officiauy endorsed nu=bers j
could.itifle pmgress txmtrd a better selection of nu:.bers. h ideas and interpretations from applicants thelves have played a rajor part in the fo: =21ation of the current bases for site evaluation. It would be a significant loss to stop the flow of new it'.eas fmm the applie. ants. he Cc=:ittee also believes that" it is po:sible that the appear:sce of quantitative me.bers in a Fedeml ret:ulntion or policy st:.tement vill reduce the continual S
O 8206070069 820514 PDR MISC PM
i w
80 avnrenes: of the nppliennt that he ha: as:umed a re:pon;ibility to be alert to and to act on unforeceen dicadventaces of a cite even after the site has been approved. The Co=:ittee, therefore, y
[
advices that a quantitative statement of site criteria not be included in Federal regulations, N
j b-3 Rcse co=nents do not mean that the ACMS has no bases' for judgingy;
-- {-
is still largely a r.atter of judgment. Inns =ach as the A RS has 1
4 the adequacy of sites. R ey merely e=phacize that site selection been making cite and reactor evaluations, it may be helpN1 to review the frenevork on which these judpents are being made. It is a prerequisite, of course, that the reactor be carefully and competently designed, constmeted, and opented. It sh reld be 4
inspected during all these stages in a manner to assure preserva-tion of the intended protection of tne public, Also, these factors are applicable only to those reactors on whf ch exp rience has been 9
developed. Reactors which are novel in design, unproven as proto-types, or which do not have adequate theoretical and expe.vimental or pilot plant experience belon6 at isolated sites - the degree of isolation required siepending on the amount of experience which exists.
Our site evaluations stem i'*ua several concepts. Bese are over s/
lapping, but not conflicting:
- 1) Everyone off-site must have a reasonably good chance of not being seriously hurt if an unlikely but c2 edible reactor accident should occur..
- 2) Le exposure of a large'sepent of society in ter=s of integrated man-rema should not be such as to cause a sig-nificant shortening of the average individual lifet'.ne or a significant genetic r'n nge or a nipificant increase in leukemia - should a credible reactor accident occur.
- 3) n ere should be an advantage to society resulting from locating a plant at the proposed site rather than in a more isolated area.
b) Even if the most serious accident poscible (not norrn11v considered credible) should occur, the numbers of people killed should not be catastrophic.
Incidentally, the concept has been proposed by others that the da. age to people from ret.ctor accidents can be accepted if it is no greater than that experienced in other industries. Uc reject this cuCCestion e.s pre-ature, ar.d fellov rather the con-
___ cept.that the consequenrec of reactor a:cidents cust be le:s than this. We reasens for this rejecticn are twofcld: Fir:t,
-2_
t I
l 1
l l.
t.
vc do not have sufficient infomation on the probability of reactor accidents to nnke use of this ccncept in site evalustions. Va do usa, of course, the fact that the probability of a serious accident is very J ov.
Second, ve recognize that the ato=ic power businces has not yet reached the status of supplying an econo:::ic need in a manner sM1a?-
to that of more cature industries; and, therefore, arg==ents for taking comrengonal risks for the greater good of the public are c:cevhat vesk, s
At theyame time, we do not vant to imply that the restrictions placed j.
on sitt locations during the developmental period of atcr:le power vill ta 4
necessidly be carried over to the period of mturity of the atomic power industry.
he reduction of these concepts to a jMc,ent as to the adeqwey of a proposed site requires further legic and the introduction of so:se nu:ncrical estientes. We believe that the searching analysis which 4
is necessary at this stage should be done independently by the owner of the reactor, using the characteristics which are peculiar to his site and to his specific reactor. Bis step, we believe, is essen-tial in developing his contimting alertness to his msponsibility to the cocc:: unity surmunding the site. However, in Ccammittee delibera-tion, we balance his analysis against a genemlized accident which serves as a reference point fd:ss which we can better understand the analysh sNtted by the appl. cant.
Our generalized accident ana4 sis assumes that a serious accident has occurred and predicts in rough tems the consequences of sach an accident. It is obvious that the generalized accident is an l
arbitrary artifact subject to change and has value only so far as it aida judg:sent. As a matter of fact, for certain reactors ani l
conditions juds;nent vill it.dicate that the generalized accident,is too severe. In the genera'ized accident, p m:ust make nuserical assumptions as to the amount, type and rate of radioactivity release (the soune tem), the dispersal of the mdioactivity in the air and in the hydrosphere, and the effect of this radioactivity on people.
Source Tem An arbitrary accident is assu: sed to occur which neults in the release of fission products into the outer:nost building or centain-ment shell. About 100% of the total inventory of noble gases,. 50%
of the halogens, and 1% of the non-volatile products are assu: sed to be so released. It is then assu=ed that this edxture leaks cut of the outemost barrier at a rate defined by the designed and con-fimed leak rate. The reasoning back of this source tem is at:it-tedly loose. It ste=s primarily from a present inability to be com-inced that coolant cannot be lost sc=ehov from the reacter core, either by spontaneous fracture of some element in the pri=ary system l
l -
i 0
~
- .*
- '.1) SS : M E:.
Y
$r j.Krr.nure e.u.ed by --.lo crr. tion (inntrn stal or ;rrun) of the f!
sgn'orel reds.
Shltte.U.;, thic c::uned seur:e term in 1.rce, but yk it' ther@y r.: fords a fr.ctor of cr.fety. In co.e esse: it is justi-i fi. Tale to redice thi: cource toi :. It. is also t:citl: c : red that in this accMent the outemost b:u rier trill not te breeched. T.e lo.ic behind this n:Or.:ption is thst ue require o.11 of the cc: po-nents rest' Ining the pressure of the pri=;j 7:te:: to be ope = tin-I at teurerature: above their nil-ductility tent:rsture.
'Je are,
'herefore, more confident, but not certain, that fcilure vill e: ur b:, tearing rather than by brittle fracture en1 that the probability of c,*. etion of mic lle: which penetrate the outermost barrier is lov. he nece===ry cupportinC structures and chieldirc c1:o p cteet actinct ni :ile dansce.
Dism rs:1 of the F.cdioactivi:y
- 1) l'eteorolory y
~
Ue assu=e a dilution of nir-borne activity usirq ntmospheric diffusion perancters v'..ich reficet poor, rather than averace, acteorological conditions. Choice of specific parc=eter values follows n u a sur-vey of meteorological conditions expected to apply at the site, pri-
=_r11
- vind and stcbility distributions. To analyze the gener ll.'i accicent, we u:e the Standard diffus' ion calculation methodology cut-lined, for exa:Tple, in IJX'U-3066 and t'/OH-740. The ato: phene dif-fusion phenonena 1: the cubject of active research, and new results can be expected to fim up cad i= prove the present methods, cithough we dc not anticipate mjor revisions in this aren.
- 2) Hydrolog Ccnsiderations of hydrology cipbased on chara:teristics of surface and cub-surface flow a: they are related to the pos ible reles:e of conte =inated liquide to the off-site environ =ent.
Tnus, the rate and volu=e of surface flow and the poscib]e presence or cbsen:e of cbcorbin; barrier: of soil betveen the reactor complex ar.d 1. portant underground cquifers r.hould be tahen int: considention. Ec:e facters :ract be fnverable for restrcining the flow of radioactive
-.aterials in case of accident. Design factors, including the capa-bility of pmviding adequate hold-up in the event of adverec hydrol-ocy, are also significant.
Effect of Radioactinty on People Inc upper li-it to the expo ure to a =c :ber of the p0lic in the I ceneralitco. c::identl hould be ne higher than the rr.xi ::s once-in-
'n-lifeti=e c ercency[dese.] Such c level has not been established by IC, 't'e n-e arbitrarily using a fip:re cf about 25 r uhole body
. I,.
- ~
I_4 ?S 7 E '-
crginivalant 1.tc..rcted doce for this level. Lis ficre ic
- Pioned in H:.lboo:: 5') of the *:.tiona.1 Ercau of Steninras,
!3 rafd 0?-70. Jin:r t*te ied.ine do e is often contr:111=;, ve f!
cri!.pntntively e.,nciderin; a thyroid do. e li ita.tien of 200-
~?)
300 rnds. Oc de..,.e so far r.cntioned refer: to 15.!ts to 3
7tople when the people cre concidered =s indepenient in2*.vid-v' 10.
..e helieve that it is c:sential th..t the Atoui: i.nergy Ce: :lec.'cn atLc pt to cc:,nfir:a through its staff or it.s navicors in this field that this cuc.ccated value of 25 r whole bely or 4
e:,uiva3cnt is without si nificant biological effe-t a i the 6
individuals th =icht be subjected to this doce fres be cen-e mliced accident.
'~nen 1rece nr:bers of individuals are expe:ap te sliution.
I another linit clco exists beenu. e of Cenet'e ffc;ts :.nd be-cauce of the statictical natre cf induced leu:.:c..in nr.d the shorteninc cf the life span. *fne licits of c::perra r 2.rce Groups of pecple are better expressed in ter ; of ir.tegrt.ted
- r.r.-rc=s.
We are censiderinc usin., a fip: e ef h z 100. an-re <
for this limit for the people who =icht be exponcJ to radin3 on 1
doses falling between 1 nd 25 rems. ?nis figure of A : 19
==n ens is rou-hly equel to the dose received fre 2 naturc' background by a million people during their reprciuetive lifetirac.
The i= plication of these nu=bers is thi's. About a reactor site, there should be an exclusion radius in which no one resides.
Surrounding this, there should be a region of lov population density, so lov that individuale can be evacuated if the nee 1 arises in a time which vill prevent their receiving r_cre then a dose of 25 r.
Beyond this evacuation artag there :hould be no cities (above 10,000 to 20,000 population / safficiently closea,othattheindividualsinthesecities:. igg.treceive (1) h x 10 rf_n-re::s e: ore than the lover of the folleving:
in the generalized accident, and (2) 200 rems under the ex-tremely imprebable accident in which the outer =oct be_rier fails completely to restrain all of the radiocetivity of the generalized accicent.
The Co=sittee vi:bes to e=phasize again that the nu=bers which have been used in discussion of the generalized accident should not be for:alized into rec 21ations or Co=.iscion policy. The Cc= ittee wishes to acknowledge the help it has received from the Ha:srec D.cluation 3rz.neh in this ratter and succests that these individu11s be en:oraged to present as technical parcro, but not as regulatiens, a cceplete des:ription of their vorhing I
i i
l
. ' 'J S.". ' i.* F.'
4 i
nipro ch to en:: inn j4ments an the ade;ncy of propo:ed reactor l
cites. Such a ye.per, of cource, vould have the etstus of the l
opinion ef cn infor :ed technical individual, but v:rald not i.$y I
Co::ittee approval, nor vould it have the rigidity of a Cc::=ission poli,ty statenent.
l
%i
-)
Sincerely yourn, Las11e 011rernan j
Chnir:nn I
si lI g
h I
f I
9 4
a l
I 1
_