ML20237A311

Policy Statement on Severe Reactor Accidents Re Future Designs & Existing plants,10CFR50

ML20237A311
Person / Time
Site:	Pilgrim
Issue date:	12/31/1985
From:	OFFICE OF THE FEDERAL REGISTER
To:
References
CON-#487-5027 2.206, NUDOCS 8712140397
Download: ML20237A311 (14)

Text

' ' 5017

%Cb204)7A

'W Wl,9e%54 ..

. . . m--me ,7 i POLICY STATEMENTS 00CKETED Ui%C he Commbalon telles upon several there is no genuine controversy. Both of factors in directing the Ucensing Soords Nuclear Reactor Regulation" that was those factors weigh in favor of a finding and, where appropriate the staff to that any deficiencies between present gg pg 16M4). A6 adva{nc(e y notice ofp l conalder carslully the applicability of licensee planning (which complies with l 50 47(c)(1) for the limited pertad the Commission's pre CUAAD proposed rulem ' "

. Severe Accident necessary to finalhe a response to the late relation of 10 CR 50.47(b)(12)) Design C 'ilena." pu bd a @@

recent CUAAD decision. Because b and uture planning in accordance with 2.1980 (45 F1t 65474)is bei withdre Commission has not determitml bow,,or the finalinterpretadon rlplanning by's notice published else e in th s even whether, to defina what constitutes standard (b)(12) as a response to the bsue.

adequate arratigements for offsite CUARD decision. wtll not be safety POR NRNA 8Fo*14AT60*8 Coor7ACT.

Individuals who have been exposed to significant for the brief period in which Miller B. Spangjer. Special Auhtant for dangerous levels of rsdiauon the it takes licensee to implement N hal Policy Development. Division of Commlulon bellewas that untilit standard. S) stems !stegrat on. OUlce of Nuclear provides further guidance on this matter. In addition. as a matter of equity. the Reactor Reguistion. U.S. Nuclear Uceuing Boards (or. in uncontested Commission believes that Ucenalng matim. the stsff) should Orst conside, Regulatory runmineloo. Washington Boarda (and. In uncontested cases. the D.C. 20555. Tele %cne: (301) 492-7305, the applicabibty of to CFR 3a47{c)(1) staff) could teasonably find th,at there before considering whether any an other compelhng reasons to avoid suppt.gssorf Any grown afhis

. addidonal actions an requindte delaying the beensees of those policy statement sets forth the irnplement planning standard (b)(12). applicants who have complied with the Commission's latendom for hg Such ransiderstlon is particularly Commission a pre CUARD section ando h  % wh &

approprists becaun the CUARD Sa47(bX12) requinments.Whers mojy;ng s @ n nlated to nector decialon leaves cS en N posalbulty that applicants beve acted in good faith accidents more seven bn design basb modification or refr.taretation of reliance on the Commisalon's prior accidents. & main focus of this pla nrdng standard (b)(n) could reevlt in interpretation ofits own lation, h statement is oc dedslon procedures a determination that no prior reasonablenen of this faith involving staff approval or. optionally, arrangement need to be made for og. reliance indicates that it wculd be unfair Commission certification of new site individuah for whom b to delay liceulas whDe h Commissico standard designs for nuclear power consequences of a hypothetical acddent comphtes its responsa to b CUARD plants. It also provides guidance on are limited to exposun to redistion - nmand.

~ .

decision and analytical proceduru for in considering the applicability of to Finally,if Ucensing Boards find that CTR $0.4?lc)(1). the Ucensing Boards these factors adequately support the the resolution of severe acddent istues apphcation of to CD for other classes of future plants and for (and. In uncontested cases the staff) eX1h then existing plants (operating reactors and should consider the uncertainty over the those Ucensing Boards d condude continued viability of the current that no hearings would be warranted. plants under construction for which an sneaning of the phrase " contaminated Wnfore, untd & Comission '

operating licenae has been ap I .

f

- injured individuals." Although, that condude ita U rem ad g/

Severe nuclaar accfdents an bJed)W phrtes currendy includu members of Web subhdaldg h h b b the offsite pub!!c exposed to high levels differenti . N Ucensing Boards could g.Ac,-reactor cme whether or not Ibers are -

of radiation. b CUARD court has reasonab y find that any hearing ~. segus g jta ==reences On- -

clearly left the Commf asion the ngarding comphance with to C71 Ocbber 3.1980 the Commission issued discretion to " revisit" that definition in a 60 47(b)(12) shall be limited to issues an advance notics of proposed

~ fashion that could remove exposed whica could have been heard before the avlemaki,ng. Severe Accident Design individuals from the coven e of Court's decision in CUARD v.NRc Criteria that invited public comment planning standard (b)(121. T erefore. M Wm pgents for treating Dated si Washingtons DC this teth day of Ucensing Boards (and. In uncontested stay, t eas. seven accident issus (45 FR 65474). By cases the sieff) may reasonably For the Cornminiest another notice published elsewhere in conclude that no additional actions g,,,g g' g this issue the Commission is should be undertisen new on the withdra

$trength of the present inte pretetion of n eW Cuimis*

this advance notice of that term.

pr_oposed emaking This policy rtatement is a revision of Moreover. the Commission bel.' eves */ So FR 32138 i3 .%

the " Proposed Commission Policy that Ucensing Boards (and. In s*$' Petdeaf SN35 ' 8@ 9 % - Staternent oc Senre Accidents add uncontested casu. the staff) could esasonably find that any defidency (pg gp g r , g 't;.,, Related Views on Nuclear Reactor c

hich may be found in comp'ying with a s'.a

%n 4 ..% Regulation" publiabed for pblic Poecy Stahmerd Sevb RW comment on April 13.1983 (48 FR 18014).

si n a (12 i 218 for the Accidents Regardng Future Designs Twenty six letten of coment oc the pu ses of to CE 50.47(c)(1), De low v-n and Existir>g Piants proposed policy statement were pro abibly of scridents wbcb might

• received. b nudear industry genera!]y cause extensive tadiation exposure - Aeosev: Nuclear Regulatory . supported the proposed policy statement e Commission, j - <. :

ACnosc Polley statemb. ands ested several modifica tions.

I ea e on n po to 4iv}1 3 E ch the citMam of the pW

" E # * #

o d. e oNib y oIeu'*1a ' tussenan:nis statement de~acribes the psNo** t rebd pe a po! Icy the Comrnission intends to use to accident is less than one in a muho.n per yur of operetion). and the slow mooln safety luun mlakd k noctor focused on a qtion of over nif ance on probabilist>c m assesament. .

esclution of adveru reactions to accidents more seve e than design basis espedally when coupled with the overexposure to radiation cre generic accidents. Its main focus is on the Commission's Tafety Coal matters applicable to a'J plants and criteria and procedures the Commission licensing situations and over which Development Program" (43 FR 10771 intends to use to certify new designs for March 24.1983).h policy Statement

, nuclear power plants. %is policy was revised as a result of these statement is a revision of the " Proposed suggestions and critidsms as well as

, A 8 (// -> CorninJssion policy Statement on Severe comments by the Advisory Committee y- Acddents and Related Views on on Reactor Safeguards.

%., Y Y M 4PR-37 December 31,198tureset) 87*2140397 851231 PDR ADOCK 05000293 g p PDR gg

,, J n

POUCY STATEMENTS Many changes have already ben Implemented in existing plants as a needed, to ensun that there la no undue Sutement is reprinted along with other result of the %Q Action Pian NUREG- risk to pubhc health and safety.In information and appendices that provide oseo and NUREG 0737),8infor(mation Implementing such a systematic perepect!ve on the development and resulting from NRC. and industry- approach, plants under construction thei -.

beve not yet received an Opereting le2plementation of this pohey and 10 v It '

sponsored research, and data arising from construction and operating IJcene wiu be treeted euentfaUy the nietes to other features of the Sevu Accident Program. A copy of NUREG-experience.On he beels ofcurready name as the menner by which opereting 1070 will be evahble for laspection et ereitableInformatloa the Commisolen reactors are dealt with.1%et is to asy, e plant opecific review of severe accident the Commiseloe's Pubtle Document concludes that exisdag plants pote as Room.1717 H Street NW., Washinston.

undes rial topublic beelth one vulnerabilities maing thle approach is not and seet mepresent basis for considered to be necessary to determine D.C Copies of NUREC-1070 may be te Purchased by oalhng(202) 27F20eo se action oa' generic rutemaking er esber adequate safety or compliance with NRC safety reguladons ander b (202) 27bt171 or by writing to the regulatory dianges for these plaats % Superintendent of Documents UA because ofsevert Facident risk.The Atomic Energy Act, or to be a necessary Covernment Printina Offles, P.O. Box Comnfesfee ha'i ongoing neclear safety at rootine part of en Opereting !Janse review for this class or plants. staat. Washington, W 300t>7082 or programa thet include: ine resolution of R the decision proceu for the NationalTechnicallnformation new and severalother Unresolved Senice, Departant of Comroercm3285 Safety leaues and Generic Safety Isenes: a new standard plant design- Port Royal Road, Springfield. VA 22101.

b Severe Accident Source Term an approach the Commission strongly Program: the Severe Accident Remarch Peacy statessent encourages for future plante - the Pohey Program operating experience and data Statement affirms the Commlulon's A Indr86CW88 evalueuon regardmg failure of certain belief that a new design for a nuclear Engineered Safety Featurn and s power plant can be abown to be nlated equipment. human errors,afety*

and acceptable for severe accident concerne ththe Room se severe accidentlesees la Butament is prompud by to other scarcos of abnormal events; and ifit meets the fouowing criteria and staFs procedural requirements: class, saf est accidents of this scrutiny by the Of5ce ofInspection and an the embehaud Enforcement to monitor the quality of

• Demonstradon of compliance with .g g ,,gg , gn, plant construction, operetion. and the procedure! requirements and criteria s7ons tukbmajarrisk the pubhc maintenance. Should significant new of the current Commission regulations, safety information become evallatile. including the Three Mile Island d. sad with plaatsedioacgive accidents i

release from whatever sortce, to ques, tion therequirements for new plants as reflected frees

- podear[)eceive 21 of the conclusion of "no undue risk. then the 220ej: in the CP Rule [10 CFR 50.34(f). 47 FR Coaunheic'e s nevere accident poucy is

. technical laenes thus identified would be that es Comunlealoe intends to take di ruolved by the NRC underits beekfit . Demonstnuon of kehnical rusonabh steps to reduce es chances policy and other exhting procedures. twolution of all app!! cable Unresolved d- of a wvem accidet Safety leeues and the medium- and high- involving subesentialdamage to the rule wb fs ti le. f,p#d "g",' n8 reasser anse and es mitigate the One important source of ww

• ' rullability of decay best removal g,nou endian acci h

[,",eg, ".'g*3pgh, spec 15c probabillstic risk assenments.

Each of these analyses, which provide s systems and the rollebility of both AC and DC electncal supply systems; On Apr013,1giB3, the UA Nuclest Regulatory Connaleston issued foe pubbe comuneet e " Proposed

(-)) '

~

• Completion of a Probabuistic Risk detailed awessment of posible accident acenarios, has exposed relatively unique Assessment (PRA) and considendon Accidenta Commhalu Pouc Statement on Seven of the servere accident vulnerabillues bNuclearReactorReguletion" and Re ted Views ce vulnerabibues to avere accidents. PRA exposes along with da 6*ighu 48F3 Generally, the undeelreble risk from 18014) The public comments ave been that it may add to the eseurence of no these unique features has been reduced mvfewed. and, on the baala of furbt to an acceptable level by low. cost undue

, risk to puhuc beslth and safety, e and cono n.

g m abanges in proceduns or minor design p ,

modifications. Accordingly, when NRC

• Completion of a staff review of the H a gul design wtth e conclusion of safety to nguletory dwielon making l and industrylateractions on uvere 88W "'

• 1 accident lasses have progressed acceptability using an approach that 7 ,ung ,nd fu ucl at le%d""Cr'dL"."For",*id'o

, d stresses deter C'M'"ministic ""~'***"'*"'d ens (neering

,,,gn ,g f

rnsulate an inteested systematic approech to sa e;iminatloo of each Cristaen designe that are veristions of the prueent generetice of 1.WRs wtD be

$ro 2 a d"'" $ d 7 @ ~

in has with fts legiale anclear power plant now opereting or mandate to under construction for possibly reviewed in future construction permit ensum het oudear poww pianu should signifleant risk contributors that stight ge plications under the guidelines be plant specific and might be missed entified for a val or certification of e m , q ,,6, ,,,,, ,

standard plant e er===ra air r ps%.ne . win, a obsent a systematic search. Following *ddman * *=e4=ssaara sesser 8===

the development of such an approach. Becanae &ls pobey statementis just or.=m onepart of a ****"' "

en analysis wul be made of any plant progreen. locluding UN.,i 4 -t ev that has not yet andergone an the severs Acci i Rosearch Program, da _ m g. ,,, ,,,4 ,4 appropriate examination and cost- for msolving severe accident lasues, the "'**m' **appamm a i

' e anny w aseded a.susr

- i. . l effective changes will be made. if NRC staffis publishing concurren0y with thh Policy Statement a report on gyp 8 ,' ( " ' M * * . .ency (

'o ew eis rede c d is e Patty sw.mm "NRC Policy em Puture Reeetor Designa: *=ad am,sn pii. m h g .poa ..isron y , j*@'

em sta$able for 8ame =arino er se mere m, ee.

Q~"'"*""hopeco.a et he NRCs P bus Nudear Decialone 8* N Npower onplant Seven

• • • hdara Accident luces in "'dd"'s 64 Regu!adon"

• • d "b * <

i (NUREC.-2WD). In this report the pobey ruisemans ,t*

• D"* Cg"."D' , *n, " 'h**e'

• 8"Nd"8 m maedessi h n., s"'.D"6"um ==uidb , i Deeember 31,1985 (reset)

PS-PR 38

4' POUCY STATEMDWTS

/ pose no undue risk to public beelth and safety, the Commission has esamined commitment (NUREG-coso. Task (l.B.8) (EpRI) on their " LWR standardized on degraded core accidents currendy Future Plant Design Evaluation

, an extenelve range of technicalissues referred to as severe nuclear reactor

. rela Program."

to sevm occident risk that have accidenta:

been ntified since the accident at

• To evold unnecessary delays of It is suumed in this Polley Staternent hm Mile Island.FoHowing plants now under construetlen: that, over the next to to 15 years. utility implementation of numerm and commercialinterest in the Urdted '

r , _x , , ,,; , ,, ,,, r,,,,,,,= 1

[. modifications of lent design and States will focus on advanced light

' yce p n % 3 , m m . olents bose i Mgulato{ rres as dueloped water reactors that involve m eerstion aru andar ens'um Acuon plan (NUREC- improvements but are essentially based l navuumpos e j e&er jgg g.j further Dechts salees on the technology that was i l-c.w 4,- - - , i M"r #4 by soy 4. FNj demonstretad in the destgn.

construction. and operetion of more than i e O, 7 M. ~

-

• To'hchieve taproved stability and 100 of these plants in the Urdted States.

predictabuity of reactor sogaletion in a

{$[MM t s-manner that wpuld emerit impmved puhuc confidenas ir ourregulatory This polley should not be viewed as prejudicial to more extensive changes in reactor designs that might be for rimmedisk

,i-i i. des m.h fuhWrdeg a

decistor snakies. demonstrated during or beyond that rulemaking orobr wy denps k policies arosented ki this time period. Indeed, the Commission for these plants because d oe,m statement wGllead to amendment of encoursps the development and occident risk. Howewr. the occurrence NRC regulations, standard review plans commercialization of any standard of a severe seddent is ma likely at fw Hmasing ha or o&er h der ease plants than at others. At each procefow and artleda es part d NRC's bm$s 6ts, est suchadsht make safey &mugh as em acidpod plant there wiD be systems, componeng greater simplicity; slower dynande er procedures that are the moog %W makes aHowance forSevertre8Ponse Accident Program. Mis to speet conditions irrvolving i

I significant contributors to severe each a the mult of the accident proarso,r events; passive heat accident Ask.& intent of this policy development of new safety information removal for less of coolant accidents:

statement is to provide stilities with of signincanos for design and operating and other charecieristice that promote basis for development of Corsedesio, procedessa.

. smore ofBelent construction, operation.

guidance that wGI aBow ident1Bcadon of in accordance with the activities, and maintenance these contributors and dmlopment of **I*'Y'"D*N'procedurps to enhance

'*d **"*"F-the appropriate course of oction, as views, and poucy developments discussed in this policy Statement the D,.!= gesceptable '

Commlulon believes that it la poulble B. h/icyforNewMont App // cat /one

. ham to complete its ongoing reviews of new ihmduedon y plant designs with an expectation of

__ M"Me,,ftlye"gy noT M. fully snool quesdoes in the severe socident No new commercial nuclear reactors have been ordered in the United States

~jnenageant" includes occident course of the review. einer December 1373. However, the his beliefis predicated on the Comminion has received uveral prevention, socident management to availability of results from the ongoing certail or retard its progression, and applications for reference design NRC. Industry Degraded Core approvals that are currently andar consequence mitigation to forther limit Rulemaking Program (IDCOR). and

.- review. A reference design is one of the its eNects on public health and safety, vendor remich and insights from the options in the Conuaission's .

De Comedesion plans to formalete en Zion. Indian point. Ilmerick, and other standardization policy. When approved approach for a systema 6e safety risk analyses.N review of standard examination of existing plants to by the NRC stag, a reference design dealgas for future Os provideo could be incorporated by reference in a determine whether particular accident incentive to ladustry to addrass sevm vulnerabilities are present end who new Cp app!! cation and. ultimately, in accident phenomena. Indeed, since July an Opera ting IJcense (01.) application.

coes effectin changes are desirable to 1983. the star has completed the During the corresponding Cp and OI.

ensure that there is no undee risk to reviewe and has issued Final Dnign reviews, the NRC star would not public health end safety.in Approvals (FDAs) for two standard imp lementing sed a tyrtematic dupUcate thatportion ofits review designe (General Electric Company's encompamd orits reference design ePyroach, plenu unovr construction that BWR/6 Nuclear Island Design. CESSAR approval. %erefore, even in the absena have not yet received an Opmting II; and Combustion Engineering of new Cp applications,in order to j IJeanse will be treated essentially the Incorporate (s System 30 Design, provide guidelinee for the current same as the manner by wideh opersUng CESSAR). A severs accident review by reference design reviews, the

~

reactors are dealt with, ht is to ny. a the NRC star of the CESSAR D design Commission has recognised the need to plant.spedfbc *sview of uvm accident for forward referenceability is nearly vulnerabilities using this a promptly estabush the criteria by which complete.The reytew included snew desa' ns can be shown to be considered to be ascenar;pproech ' to determine is not asseument of alternatin dulgn acceptable la meeting severe accident odeguate sefety or compliance with changes for severe accident risk NRC safety regulations ender the concerns. De Commisaloa now believes reduction. In addition, the staH has been that there exists an adequate basis from Atomic Energy Ad.or to be e necessary involved with pretendering review of an which to establish an appropriate set of or routine part of an Opersung IJoense appUcation for Westinghouse Electric review for this cleu or plants. criteria.%is beliefis supported by Corporation's advanced presourtsed current opereting reactor experienos, h main ourooses of this Polley water reactor design RESAR-6p/90. In ongoing uvere accident rmarch. and Statemem mow- January 1984, the NRC found the insights from a variety of risk analyses.

• To clartfy the procedures and RESAR-SP/90 appUcation for a De twultant criteria and rocedural requirements for licensing a new nuclear Preliminary Des Approval acceptable requirements m listed be ow.

plant; for docketing in May 1984 the

.lt. Critstia and Procedural Requirements the inst 1 aa at p an PS-PR 33 Decernber 31,1985 (rout)

_ _ _ _ - -- ~

POUCY STATEMENTS web as a propoud custom plant) can 9 requirement may differ considerable Y shown to be acceptable for severe destped to secommodate all of the

- from one review to another. In addition.

acchdent concarus !!!t meets the thelicenue la requind to ensure that host 21e environments resulting from the '-

tonowing criteria and procedural the intent of the ufety requirements la complew spectrum of seven ace dents, regarements: accomplished during procurernent. they can contain a large faction of the

s. Demonstration of complianca construction and operation- radioloskalinventory from e portion of b proceduralrequiremenu am crite a It is rece- that there are a the spectrum of such eevm accidents, of the curnnt Commissie ons, 6 o PRA MWo Nu d For example. large dry containments Inchadag the nr e Isl continue to undergo evolutionary @ g e n, g w M A reguartmenta la new plants as fg development as the results of research b consegeences of a wide epectrum of in the CP Rule [to programs and reliabulty dats from cor,-melt acddents; hence, fur

) mona

• operating reactors become avaQable and rec ^eenents may be annecces$er ary or et resob o dd] bb Unrudna as innovative uses of PRA in safety , dn epgrading current requirements to Safatyissum e aM high- decision conttxts suggest better ways to gain liznited neprevetnents of their priarityGenedeSafety1 achieve the benefits of these methods existing espabiDty may be necessary.

a opedd focus on suuring eugs, dudas while guarding aseinst thdr limitations no Cosedulon expects bt thew ,

nbabW4 o y or impro uses. Wh!!e learning curves i mettm wm continw to be sobbets for bihty o both AC of these study (e.g., in the NRC reseerd) hefectrif[sulpply systems; ds w1U likely continue for a decade or more, it weald nevertheless W in turbe plant-epecfSe sto

c. Completion of a Probabilistic Risk u the Z)co and Inaan Point Apesament (PRA) and consideration of be construction to cocedidak ele expedance at varias stages of PRA probabms6e dak ownsentsh me sevm accident vulnerabilities the development and utilization. At the 1rrtegrated eyetems analysis will be PRA expons doog wie & Indshu present stage of devdopment, a number und to explore whether other 68 "' of posidn uses of PRAs have been contstmnent types exhibit a functional dA oPu b e ty., demonstrated, espedally in identifying- containment capabWty equfvalent to and ht oflarge, dry containment.

(1) nose contributors to servare d.Completloo of a staff review of the acddent dak that are clearty dominant Although containment strength is an design with a conduloc of safety and hence need to be examined for cost- important feature to be considend in

• cceFtabQity using an approach that such an analysis, credits should also be senmees deterministic engineering effectfw those dak wquences eeddee? nductionthatmeasurw are and (2) gins to the inherent anargy and analysis and }udgment complemented dearly insignificant risk contrhters radionuclides absorption capabultles of by PRA~

W m &mfon k @ndy the various designs as weD as other ne fundamatalcdteda listed abon apply to the staff's mim of any am d%wd in between cases are more problematic.

% wm M M wrd combustible gues.

design. In addnasing cdtada N aWeb According! It is dear that core-melt accident pubucation of.within 13 monb of the geshcantfw ce shis uvm accident evaluations and enstainment failure a, [* d '

U ahd! statement, the steff willissue svidence evaluatiou should continue to be '

ddtarnadves and on the forza, purpose and role that PRAs emhatio alternatives to addiwa an to ptsy in severe acddent analysts dperformed operating plants foranda plants repreuntative under -

sample ( '.

the annsolved and generic safety lasues and dedsloo making for both existing constructive and for all future plant -'

and w watch im cost 45 econ and future plant designs and what designs. mee stuees abouldimpron nd sev minimum critada of edequacy PRAs our understanding of the containtoent kee , , should meet. Prom experience to date loading and failure characteristics for

• • 'UO' Q d[ g bough ,

la evident that PRAs could serve as a ,it the various classes oflaculties. %e highly weful toolin assessing the dsk- analysu should be as reahstic as QI

• ogy g" reduction potential and cost-effectiveness of a number ofimaginati" design options for new planu in ponible and shouldinclude, where sppropriata, dpamic and static loechngs fme combusdoo of hydrogen

• • ]s8j8Q

co 8j8 "g,* comparison with desfgn festi;.rea of and othercombustibles, etatic propure 6* Q ""

existing plants. %e PRA guidance will descrh b appropriate combination of and temperature loadings from steam and noo-coodcasibles, basemat social algniacance ed cannot naddy determinfsde and probabilistic penetnetion by coneelt materials, and be quanded in commensurate unna. considerations as a besta for eevere abcu on aerosols on engineered eafety

%e foDowin accident decisions. futures. A dari$catioe of contemnt "C" 0*t*U h"g sec6ans explain in

• • • • *dI*'I* "" I' ne proposed Commission Policy performance expectations will be made appbed to se vadas types dmiews including e decision on whether to est he staRmay amounter. It is Statement oc Severe Accidents tasued utabuA new performance crtteria for on Apr013.1983 recognises the need for u a *b a t n teria striking a balance between accident vention and cons uence mitigation-

{ exploring b needgor eddltional containment eystems and. If so, what these should be.

listad above beforeffad a valor *Ik Commissica .N. neoplses the casadCt is owent' deefgn or opere tfonal features in the importance of such potential that a new design can & rough next generetion of plants to mitigsto the contribwors to uvm acddent risk as

&n==i acs - d = couequences of corecelt accidents. human performance and ubota .%e befm ncetring 61: Baal ap= peal w 6e e-deeion wm ,t,b a wience Issues of both taalder and outs!

cardcados. e. a n

• between socidentprevention sn i aawta,,&,esu a w m duny

• • g ,7 slp consequence mitigetion encompast anal and, to the extent procticeble.

Appegn@A) and 6e a M Mgn actions that improve understanding o will empbesised as spedal A9 roval(TDA).%e unique containment bull failure conMdentions to tbe design and in the circumstances of each dulgn review chmetedsdes and Hign futurn or w ohrating procedures developed for new wil therefora. nquin flexib ty in sedom ht decnue & mm 6 &cdveness d

• O

• gg3[ood of contatnrnent building Am wm W empbstzed

((

t g,' d failures Althoogh not specifically in design and opersting procedure development. A balanced focus w1D be Denmber 31,1985 (rnet)

PS.P440

. POUCY STATEMENTS f peld as te negative impact of human demonstrable progrees la safety he use of PRA in a twcretep review

/ -

./ + on severe aceldent risk as performance. including the reduction in process also raises a number of

' wet as its potenuaDy poeltive frequency of accident precursor events questions. Of particular concern is the someribution to baldne or kmiting the as well as a Al=InMed controversy tindag of the PRA requiressent becaum acasequeman of severe occident among experts as to the ade the completion of a comprehensive and proprussion. Design feMuros abould be nuclear safety technology. quacy of detaued PRA may not be achievable in emptedmed that reduce the risk of earty

• Further progress in severe accident the ebeence of essentiaDy complete and

~=ad====t faAnte, thus providing more risk reduction is a hedge against the tiene Imr he positive contributions of Saal detailed deelen information.

possibility that current risk estimates brefore, to require a menplete PBA at operaturperformance in curtsRing with their broad ranges of uncertainty the PDA stage would not be esahatic.

setuse assident consequences. Also, might unwittingly have been r he '-=iaaton's recent design immeures Aould be given special OPtimleti biased. howmr. Indicates IM a tial seemeen &at serve to decrease the role

• Al the severe accident risk amount of of an detal! thet would gf hamnem error la the sequence af events plant may be perndt llalted. quantitative Iseems to the Initiation nr agravatico acceptable in terms ofits direct oNsite risk analysis exist at the PDA et asse abursdaden. In pardcula,' regional consequences for public health stage. Because the Co==niadaa believes metodo d analysis and associetad data and safMy, the aggregate pmbebillty that risk analysis of this type would be a basse are ander devdopment by es I*ey, over a 30 year period) that one useful design tool. the r ws.wa.

Cassedsska's severe accident emm accident wtB occurin a large expects eat H would be osaplMed as part d 6e PDA

.o,,e,, ,s.g g n@oh,,oe negaove and

.c ons the analyses and populauta and d macte. bold.s addios eiga o . evenia.separetemumpw. m cados prueen. A y.is wo.id noi be a would yield adverse spillover prerogeisite forissuance of a PDA, N2 m ild or consequences for innocent parties in riowever,if this risk analyels is not g, ,g,,g,, other regions (l.a nuclear. oriented performed to the pDA process,it wG1 stilities sad their omstomere), not to neve to be provided as part of any CP R is amend that same d6e severe mandon a changed poudcal eccansat sammarios result in ' 7N applicadon refqthe dedp.

of eGalte consegusaces.

savironment for anclear tion itself If the scope of the rpA reference e5ecting resource costs con *=w ofectiveneescIn programmatic activities.

design applicati.m is limited to an extent his aframtion, there may be ao sleer that would preclude the completice of a hems Imr segulatory acuos bocesse there 8. Application of Criteria for Diferent maaalagful, "we PRA, the is as embetantial efect sa public bulth . Typw of OL and CP Applications reqdrened foricompieb PRA may be or sedety. However, the implementation waived. However, the applicant abould

a. Applications of Certification of stiU perform and submit supplementary of requirumments to control occupational Ae/erenceDesigns with No previous exposure abould be considered along risk analysis, to the extent practical, to FDA. In accordance with the d==aastrate the adequacy of the **

wie sie relatively small e5ects on pubEc heefth and safety for these types Commlasion's standardizauon design. If a n J-- "ve regulations and policy, a new reference

, of areare socidents.The resolution of is not submitted for an FDA, a G/

design can be subadtted for approval OL applicant referencing the approved t

• sost-bens 8t iseuse in severe accident Aret as a preliminary da,lgn and then as design would be toquired to subesit a dedulos saaktag is part of the NltC's Baal design. Ca,~sdq!y, the stag .

plant-specific PRA. For standard destgo Safesy Coal Eralosuon Program. will1aene a Prefirminary Design approvals of restricted scope, additional Alhoid in the licanaln, of axisting Approval and a Final Design Approval. limitations beyond the PRA as may plames thrr ==laalon has determined A PDAis not bowmr, a premiuisig existSu of sud a standard ign by

&M tese anh pow no undw riak 2 for an PDA. An applicant has the option the bonnes applicant may be limited by tb and salsty. ele dodd ad to submit FDA4evel information initiaDy its very nature to a twHtep Bosesing (bac virwed as imply 2g a Comadulon and proceed directly with an PDA procnes, namely, a Construction Pernait pahey that safety improvements in new review.%ese options remain and on Operating Ucense issued pism demens doeld not be asevely anchanged by this Policy Statement. esparately. This would negate some of sought, no Commission fuuy expects After a PDA a Ucation is docketed, the benefits envisioned for en approved est sendore engaged in dulgning new the preliminary bign can be or certified design wherein a previcua}y es.-a.,d (or omstom) plants will achieve referenced in a new Cp application. De approved site could be matched with it corresponding OL sppuestion would in a one-step. combined @/OL process.

a b@ee standard of severe accident safer p=ad-mance than their prior than refmace the approved final design The reference design most settsfy dseigns. This expectanos is based on: (yDA). Of couros. an approved deelsn each of the criteria stated is Section B.2

• The growing volume ofinformation could also be referenced in a new CP before an FDA can be issued.For application. fa. ward referencoobGity of a new Irms industry and goverumment. ne use of an approved standard apa===ed research and opereting standard dalen. the applicant is being design in new CP/OL applications has aHorded in this Policy Statement the reacser emperienos bee improved our received considerable attention under h* L of speci8e severe accident flexibuity of choosing between a the Commnaslan's legislative initiatives Pr=hminary Design Approval (PDA), a valasrobeties and oflowmoet methods on singlHtep licensing It abould be FtnalDesign Approval (FDA), or a for Smir mattigation. Pether learning on noted that a two-step review procnes for safesy vehnerebQldes and innovative Design Certifloation (DCL The a standard design approvalis not. in approvals (La., a IVA or PDA) be meemds is to be'aspected, itself,loconsistent with singlwtep leeued foDowing the completion of the

• The taberent BedbGity of thle IJeonsing.To be tnost effective, single.

pober Statement [that perm!ts risk risk step licensing presumes the existence of stafra review and would be subled to chaDense in individuallimosing tredenSs la eyetems and sub. systems a previously approved design bearings. ne Design Certi6 cation design) encourages thereby innovative essentiaDy an FDA This deelgn could would tie issued by the e -enhaion ways of ashleving an improved overell still be approved in a two-step procou following a rulemaking and systemma sotiabulty et a reasonable cost. as long as both steps were completed in could not be cea!!ensed in ladi e pubac acceptance, and hence advance of the single 4tep licensing bearings. Os or Ota, based on a j invasear acceptancs, of nuclear applica6an, reference deelen that has not been y,s todmakuyis Amp ad at on e's.PR 41 Decemba 31,1985 (rout)

4

. POUCY STATEMENTS 7

r e- - ' through rulemaking, abs 0 be '.

b requot to permit the design to be be a matter of separate considerstiae eM4 met to any deelgo changes arising referenced in new Cp and OL he the rulemaking proceeding in applications for a Exed period of time. spart from this Severs Accident Pobey accordanaa with the t'a==laaion's Statement such as five years, & amended FDA bachat poucy and reguistiosa.The wiu be conditioned as appropriate to d. A New Custom Plant Constructed design cert 1Scadon'would be leeued for permit Applicatiort Itis the ensare that new Cp and OL applications Commisalon's policy to encourope the a longer duration than a dovign referencing the design will utisfy sech appewral & speci8e requirements and use of reimace designs in future CP of the criteria in Section B.2. N seven applicauona. This does not, however.

g % for obtaining design accident review most be completed prior eartfacetions or appvevals wiu be precide b um of a custom deagn.

to the issuance of the new CP or Ola Custom designs shad also be reetewed

==+amakad in a forthcoming mvision to (2) Criterion B.2.c requires the to %r ==laataa's Standardization completion of a comprehensive PRA.!f a against the criteria identified in Sachen Pokey Statement. comprehensive PRA cannot be B.2. As a result of the circumstances and h Spreenlar Certi/icot/on of completed owing the the limited scope timing involved in the ongoing p*=adard taference Designe Previoue/y Granassf design review processes, the of the design, b applicant aball asIn4. In 1983, the NRC staffleemed Commission expects that most. It not all, perform supplementary risk analyses to two Ptnal Dalen Approvals for the extent practicalin support of the ww CP application incorporating a refeream designa. These were approval or rulemaking process. As sference design woold be besed on perm #tned to be incorporated noted above. the !!mited scope of plant essentiaDy final design information. This refarance in OL applications e. the design and PRA analysis would lead to will result in !:nproved safety and ammaponding CP application had a partialloss of benefits in that a two. regulatory precdces, as weU as reduced referenced the PDA. However, the step CP/OL ucensing process would be time to license and construct a socisar desiumis were not approved for required in Doo of a one. step process. powerplant.To obtain as much of this incorporation in new G applica6oes. (3) With regard to completion of a N r% = = tam w now baueves that comprehensive PRA for a mfereues benefit design app setion, beticable for a will the Commission custom these designs are suitable for use in new design the Commission recognizes that require a CP applicedoo for a custom G and OL appbcations under tbs , a PRA would be more meaningfulifit dalen 2 incide design idorma6am maammaa. M below. Amy were based on a substanual portion of that is sufficiency Real and complete to t to these designa, the complete facility des . W refore, permit completion o't an adequate plant.

&an thoes resuldag from the if Justined to the NRC sta . compledon optciBc PRA. It is possible, however.

severe accident review, wiD require the of the PRA by the FDA appucant may be that an desismo to be considad under tbs waived. !f a comprehensive PRA is not aproja plicant or cordAd referencing an da design in Bos provtsions d Sec6cm B.3.a, t.a., as new submitted by the FDA applicant for the cwkm plad wmld have in prospect a y designs. FDA a CP/OL appucant refmacing the sisshdy rdad llun&g Ims since s (1) Each of the two reference desipe daign would be required to submit a staff sffort would not be required-er '

much less would be required-gar a appacants with existing FDAs most plant opeciSc PRA. I . ' ')

regemet that their FDAs be amended to param their designs to be referenced in new O and OL applicadone.N A rererence design app!! cant pavinely granted an mA can pumw the same options of design a design certification as desen valor rmykw g gp d b {d or certi6d detailed changes to socommodate gg

,,, gg,

, ,g, ,p,gj

(-]  !

rogamut must sibe 0) taciude the in the tidormation needed to estisfy each of pmeeding metion fw mfennce des circumstances (e.g.,innovetive the artteria stated in Sec6cn 5.1. er DI) with no pmvins MA. N NA w equipment designs to meet new ASME

,,,,,' [I CPad app 6eas the ta a re e and o or IEEE codes, etc.)

s7bject # C M cfJ 6 3 0iSr @ d ite - -

referendog the deelgri wiD satisfy each n dMd cans of the ar(teria in Section B.2. Requests in (chauen , e, 1.Some General Pnnciples of PoEcy 8 **d *' we imud by Commluton D'Yelopment db doise cadame following a tulemaking proc and h Crmaleslon bas !! censed about to te haMud Review Plan DM could not be challenged in indivi ual 90 cuchar plants and expects to process 8U8CEE hearings. Cps or 014. besed on a app % cations to license approximately 30 s

reference design that has not been l h he Bret osas, the staf wiU amend ad 11tional plants. The Commtamem has rule =akw haD be j the exisdag FDA spa moeipt d es apfed su s

changel arising coasidad at length the questiam of whebt ge: eric rulemaking ebanid be m9 esse to,pennit the daign to be from)ect to any proceeding in new &and OL the rulemaking in undertaken or additional regulations

-apph ootions until the amm accidset accordance with the Commission's should be issued at this time to require mv6ew is completed. N amm backfit polley and regulations.b more capability in opereting plants er mared==d review must be --- m ._ daign certification would be issued for plants under construction to improve 4

ommpleted pr!w to the lesuanos d'"any a loeger durecon than a design sevm a,cedent prevention. l new G or OL whom appucauces approval The spec.ific requirements and conuquence mitigation, or accidmet '

referosos the desLgnD the procedures for obtaening design management that would halt or deley e===='u! completin of the ervem certi$ cations or approvals will be furbr com degradation.

ear 4d==t mytew,the staf will further - establiskd in a forthcoming revision to & TMl accident led to a number of amend the FDA to pwmit the design to b t' Mon's Standardization investigations of the adequacy of design Policy Statement. features, opere ting procedures, and b refersnood in new CP and OL e. A a,,cijver,d Construction Permit appace6ces for a fixed period of time. personnel of nuclear power plants e such as five years. Application. Becean of the many provide assurer.cs of no undue risk be the somed case, the sta5 wiD complex factore involved, the criteria regarding sever $ mector accidents. &

' and procedures for reguletory treatment reporPNRC Action Plan Developed as a amend the existing FDA spon receipt of of reactivated Construction Permits will Result of the TMI-g Accident"(NUREC-December 31,1985 (reset!

_ _ _ _ _ _ -_-________________-____a

POUCY STATEMENTS cosa Wy 1980) describes a i comprehensive and integrated plan reliab!!ity of both AC and DC electrical supply systems;the Severe Accident source, which brings into quution the i ~

involving many actions that serve to Commission's condusion that existing Source Term Program; the Severe increase safety when implemented by plants pose no undue risk. then at that operating plants and plants under Accident Remarch Program; operating time the speciBc technicalissues .

experience and data evaluation construction. The Commisalon approved regarding equipment fauure, human suggesting undue vulnerability wW

- items for implementation and these are undergo dose examinedon and be errors, and other sources of abnormal identi6ed in a nport, ** Clarification of heridled by the NRC under existing i

• TM] Action Plan Requirements" events; and soutiny by the OfBee of Inspection and Enforcement to monitor procedures for laeue resolution including (NUREG&37. November 1980). N the possibWty of generic rulemakin the quality of plant construction, when this is justifiable.However g staffissued further criteria on operation, and maintenance.De emargency operetionalfacilities NRCs amperience sugests that aalsty (NUREC.as7, Rey,1), auxiliary Commission will maintain its vigilance in these programs to offset the issues disevered through operating feedwetar system improvements experience programs, quality assuranos uncertainty of whether significant safety programs or easety analyses often leertved from NURECAse?). and luun remain to be disclosed. Industry instrenestation (Regulatory Guide 1.87 pertain to unique charsetertatics of a Revision 21 research and foreign reactor experience spectSc plant dealso and, therefore, are are also meaningful sources of dealt with through he TMI Action plan led to the t-specine information.

m.J. --ts of over s.400 seperate One important nooros of new modiBcations of to tively modest oost action itaans for opereting reactore and rather than mejorseneric design Bye Near Tern Opereting Ucenses. information is the experience of NRC changes.

About so percent of the action items and the nuclear ladustry with plant. h Seve Accident Research approved for opereting reactors are now speciSc probabWetic risk assessments is progrom as weU as NRC's extensive mesplete and the remainder are that each of these analyses, which severe soddent studies of certala ved to be Snished by the end of provide a more detailed assessment of todividual plants wm aid in determining Sacelyear 1985. There were sat poulble accident scenarios, has te extant to which carefuDy analysad diDerent types of action items approved exposed relatively enique vulnerabilities nfmace plants can appropdataly serve in the Assion pian (an average of so to severe accidents. GeneraDy. the as surrogates for a chas of simGar _

actions plant).Of this total,as unduirable risk from them naique plants as the haals fw any generic .

equipment beckatitema 31 featwee has been reduced to an conclusions. %sse studies wiD also aid.

tavolved procedurel changes, and at acceptable level by low-cost changes in procedures or adnor des' h h % b h M k ecope W rogamed analyses and reports.it is y*P . mach fwloDowsp safety stadue of modifications. A . when NRC v6du plants. Any generic changes impractical to quantify all of the safety **'

kaprovements obtained by these many and industry interactions on severs chang == Neverthelus, the cumulative acddent lasues have progressed pubBeb d an b re sufficiently to denne the methods of trough rulemaking

'efect is undoubtedly a signi5 cant improvement in safety. analysis, the r%a=I==ta= plans to **** e wW be ,e, quired is e 5 formulate an tatsgroted systematic gckB P'U*I' Other information from NRC.and approach to an 9= amination of each .

industry-sponsored research along with ~'2. policy for Operatina Reactors fauure data from construction and nuclear power plant now operating or ting experience have led to under construction for possible in hght of he above pdadples and la existing plants. Also, the significant risk contributors (sometimes condesions, te Commission's policy for NR called " outliers") thet might be plant opereung ructore ladude the has sponsored 11 plant- gegjowtog spedfic and might be missed ebeent a specific pRAs and the industry has systematic search. FoDowing the

z. e oper, guidance' ung ,,cg ear"~'ple'is a

sponsored many more.h evaluation of frequire ao thrtbar 1 severe acddent risk by the interrelated development of such an approach, an story sedo to analysis will be meds of any plant that go ),gth arewe a entluum udeu deterministic and probabilistic methods x signt$ cant new safety information arises has identified many refinements of bas not yet undergone an appropriate examination. The enmination will to quesues wheter &we is a to current design and operating practice that are worthwhus, but has identified include speci6c attention to containment L assurance of no ende risk to .

perfonnance in streing a balance aa " ' # ~event,

% 5= a: "careful ao need for fundamental (or major) between accident prevention and gepl)In

( the(latter changes in design.

. On the basis of currently evallahia consequence mitigation. la assessment shad be made of the severe n5 arm.un. u . r-musion concludes implementing such a systematic acddent velnerebWty posed by the "

let amtine e anta anu ne unha M approach, plans under construction that imue and whethm this vulnesbWtyis to pubbe besin and saferv and aa** an have not yet rewived an Operating ier site ararde or of generic portanna.

, nr. a. k-far wme6ete metion on Ucense will be treated essentially the same as the spanner by which operating e he most cost effective options for seneric rulemmHna or other resul** reducing this valnerebQity shad be

. cbaneen for these plants because of reactors are dealt with.%st is to ,e plart specifle review of severe identiBed and a dedelen shad be " * ' +

sewere ocabdent nn- mw , the reach 6d ocasistant with the onet s 4 - ~ '

w - -"e has ongoing programs vulnerebuttles using this approach is not considered to be necessary to determine effectiveness critada of tbs ~8" *

(described in NUREC-1070 and issued adequate safety or compliance with Cammiaataa's beckfit policy as to which concurrently with this policy Statement) NRC ufety regulations under the optfon er set of options (if any) are theiinchule: the ruolution of futillable and required to be Unresolved Sefety lasues and othat Atomic Energy Act, or to be a necessary implemented.

Ceneric Safety lasues, including a or routine part of an Opereting IJcense ve* la those lastances where the".

special focus on assuring the reliability revlew for this clan of plants.

W.4=e4wentissue beyced currest of decay beat removat systems and the Should signiacant new safety /], tory req information develop, from whatever ta, generlei.y a

! (,, edU be the preferrvd.F v

PS-PR-43 Decernber 31,1985 Ireset!

_ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~ - - - ~ - - ~-

POUCf STATEMENTS pas-sh= In okt cam, b lune should severe accidenta. N Commlulon be Espond of through the convenbecal whether mean or median value, is ao b pesetics of luulng Bulletins and Orders belhves that considerations which so order of 1 chance in 10.000 per reactor ' -'

beyond bt to the poulble need for year. For most plants, only a fracton of or Cameric letters when modiacations safety messurn to control or mitigate an justiSed through backDt policy, or the calculated sevan core damage severe accidents in addition to thou sequences an likely to progress to large throosh plant-epectBc decision maMnf required for conformance with the along b haes of the Integrated Safety Commission's safety regulations or scale core melt Uctil now, few analysts Assenenent Program (ISAF) hava even tried to take that fraction into conformance with the Clarification of wparato consideration, preferring even m6on.* 30 Action plan Requirements,'should to refer to the pnviously calculated i Recopising that plant.epedSc not be addnssed in case related safety PRAs have yielded nbbleimight to hearbss. value as the core melt frequency. Of the sedque plant vainenb(D6a to sevm & Beparate Remarks of Chairman core melt sequences, ty-ically only 1in

,,,* tents feeding to low cost PaDadino and the Dlasentig Views of to, or los, an expectelto yield large

.nemendons, Baneen oloach n! eases of redjoactive material. On Commisaloner Anodatine are attached. virtually every reactor sita in the United opersting reactor wiD be expected to Dated et Wuhington, D.C., this 30th day of States conditions are such that, even per$ urns a hasited-ocope, acddeot safety W tea 5 analysis designed to mecover inetances with a larp release, then la only 1 For the Neelur Regulatory Commiulon. chance in 10 of any early fatauty-and (La. outliers) of partienlar vulnerebulty seemet f sk, so on.nua, the wealth of risk estiinatu to core melt or to enuesaDy poor Secretary ofthe Comaussion containment performance, given core- before us indicate that the risk is quite Remarks by Chairman low-melt accidents. new plant-specioc studies wiD serve to verify that it is often said that one should bewan conclusions developed from latensive of too much trust in the point estimates I bebeve b Commission is on the of probabilistic risk assenments, that severe seddent safety analyses of right course with this decision. &

reference or surrogate plants can be om should consider the uncertainties, severe accident polley statement nia we do. But some then go on to appBed to each of the individual presented bere is based on the demand exact quantitative de5aitions of operating planta. During b next two triuments contained within it, the the uncertainty.nis demand la a form years, b Comminion wiu formulate a additional support of more detailed erstematic approach, inclo the analysis in its companloo document ofPrectee bottom line fallacy' f uncertainty statements o g,,;-43; of guhunes NUREG-1070, the massive support of the come ocly with large amounta af data.

gm L.! criteria, with an expectation inany other related works of this agency At the very low levels of risk with which that auch an approach wfD be and others in this field, and a logical we are dealing the occurrence of actual

. implemented by licemen of b conalstency with other actions of the events is, thankfuDy, very rare indeed.

remaining reactors not yet Commission.

8rstemati y hs, we cannot have exact quantitative inan in simple terms this poucy statement estimates of unce.rtainty. But we can and equivalent or superior mann.e that existing plants pose no undue gto pub!!c health and safety, and that

3. Pobey for Opersting IJoense AppBcetiens for Plants currently Under there is no present basis for regulatory must, continuaUy. explon the of our cetimates and our dedstons to b gape in our knowledge.We have been

'l sensitivity (

Construction changes for these plants due to severe doing that and we will keep at it. -

)

accident risk.nis condusion on reactor in summary, preunt reactors pose no ne same wvere accident plicy safety does not lead us to diamantle our guMawa appua to appucabom for regulatory program; rather we are undue risk to public health and safety.

~

operating Ucenas (Ota) u stated above maintaining a vigorous program of his poucy statement acknowledges surveillance, analysia, and evaluatjon to that and indicates a willingness to fee opendg nuclear power is % permit continued operation of existing with b foDowhg addjbou iteMnis for'*** Posa!ble causes of accidents and reactors as weD as to license new item also apphas to any hearing prevent them. In this perspective, the reactors. This policy statement has been p,,a,,1bes est mi Ccamission has ongoing nuclear safety opmung nector.) ght ads for an studied intensively for over three years,

• Myidad hbg promdbes are 7es:ms bt occident, severe indude:sourceunresolved term a safIt has been reviewed canfuuy and research p endorsed by the Advisory Committee on

• : operating experience

,D7uon a a d data on, and the scrutiny of Reactor Saf ards.! has not been lightly consi$ red nor lightly rt=4A I plant construction, tion and evalostion, t and tigation of "* am confident that the Commluion has r 'informa a enunciated a sound regulatory poucy.

eeddents amore severe than b design from whatever source, to Maunting Views of Commisat--

basas (Class sk %e rew==taalm has enacanced a pouey ngarding Class 9 conclusfon of no undue riskestion th's then the Asselstine scytr=mantal reviews and hearigs in technical issues thus identified would be S its Stateinent oflaterim Poucy on resolved by the NRC under its backfit

• Noc3*ar Power Plant Accident poucy or ohr exhting proceduns. b foremost risk to the pubbe frces Con =Ma stiene Under the National W 1evel of risk found to be b opustion of nucleat nacton dertns Environssental Policy Act of1900"(46 acceptable is wsU documented in b from wre meltdown seddents which FR w1DL. June 13,1980), and expects to basic works of the agency on these can, through b relean of substantial coctinue inis polley. W ecrironmental related sub}ecta. N calculated quantitin of rsdioective materiala, taanse deal essentially with the frequency of severe con damage, result in the injury and death of a utimation and description of the dak of catastrophe number of people.This

's to cra s.wn and "Swement of Pdcy, pohCy statement. whlch $$tablkhes raruwr commwan cwance tar Pwer a.. cia, Commission policies on these severe

.S e 3.noreW Sdey AasenswW Progree Opmdrq tJceno*C 4e rt as2x Decentwr M.

psAP5 sacy u.tn um6 n tasa tsan accident risks, represents one of the most fundamental regulatory decmons December 31,1985 (reset) PS.PR-44

POLICY STATEMENTS

' ever made by this agency, nis the technical support for ht conclusion "6 stat =r==s togehr with thne other and ecceptabiUty of ht risk is based on scientifically accepted si allable.

reissed regulstory decialons. will chart

+ principles and methodology. Second, the poucy statement does not the future coune of this agency and the anctsar industry on nudear safety Absent a detaued discunion of the go far enough in lasisting upon issese ler many years to come. De severe accident rkk posed by existing plants and of the reesoning and redoctions to the seven accident risk of three oeur dedstons are the futwo plant designs. Soch nductions an e scientific besis supporting the much more readily achieveble in new .

's decisico on the

==ys=h8Hty of the severe accident risk Commluton's conclusion on the designs for as yet unbuilt plants than for (

acceptability of that risk, that 1 st tw two opereting Indian point plants, ex] sting plants. Whue b Commisalon's condeston mest be viewed as nothing {

te development of a backfitting rule pobey statement arges reactor designers more than en unsebetantiated essert6on to make safety haprovements to the a substantial safety ^m/ oflittle weight.

for the impoeldon of new the Commiselon's polley destems of future plants, it does nothing wgmireummets together with heavy statement falls to provide any to requin that improvements be mede.

roammas em quantitative coet/ benefit explanatice of the emnmlulon's 3rd, the Coedulon's poucy samtyees, and the dewtopment of a statesset retains the option of treatment of uncertainties in evaluating provisimal, and ultimate a final. the risk of sevm occidents.ne subrtning b Man of construedon d l esamey goal with numeri standards for absence of virtusUy any explanation of futuvo u bend apce only HmitW eruhmsting b acceptability of nudear how uncertainties have been treated in informadon,induding the occadset e---se= risk. Taken together. these four this poucy statement further andermines h'.nl ted dulge informadan which would an actions will set the the nbdity of the Corn =laston's broad be needed to support issuance of a fremmewak for dedding whether the conclusions on the acceptabuity of the pp,u-in.,y design approval pDA PGtC and b industry will pursue risk posed by severe accidents. experienw with nuclear pow t (erplan). Post exandag and fumre significant safety Wrd, the Commissim falls to address du' ocostruction and regulation has i==== whetherfurther vements in in a clear and conslatant manner the h t ne b ' man wEl be pursued for existing need to prevent forht seven reacto' design.as-yoo.befbpitfans approach. W b old and pleu, and how sod accidents. Although b Commlulon a conunohg to eBow b Man d t

~. m Gl be w made. Policy statement pays lip service to this construction with only !!mited sign work umfartmastely, the first two of these dedmans by thec==fanion lead me to h Foure, &

to kcdude se means b r fulon's poh lete, b Commission seems

,,,,Di to repeating the mistakes of the post-mfstates which have led to est we an on the wron c a se statement places undue rebancein ne defernl d etsnIBeant dulgn Isam r__.dh_ .

views opposing the g s a,na.,oht dedsi probabilistic risk aueuments (PRA's) unto the constnedon and pre.opmdon wome est forth in considerable detaff in ven cddent sa, and a need io .mdify wwk q the r==-taelon's written dedston (see qu on r a alind@ propw w ompM

l. 1 CILalMis), and I wiD not rebaarse th" nUence faus to recognha prmat Takaa together, these flaws b the

' views hers.SalSee a to sey that th*7 weakneues in these assenments due to =alon's anm accident pokey

' p Comedesion's unsebetaanated and. > the limited number of PRA's avausble statement cast doubt upon the adequacy

. ,' omty aptinnistic assumpuans en the he far, b variations among b of the Omn=tastop's omall approach to scooptabatty of the spese existing PRA's, the absence of accepted deahng with one soddent risk and stak posed to the psblic guidelines on how to conduct PRA's and undermine the validity of the tese plants how now been to evaluate them in making severe Connataalon's swwplag lodgmats of by hispolley statenset to owera5

• accident risk Judgments, and b the acceptabulty of that risk for existing

- admeag and future noclear WM uncertainties inherent in attempting to and future plants.

la his esentry.In my judgmset b. extrapolate plant specific PRA results to D8 ar-d==

other plants.

rm's action today fads to "r-1 provide even the most rudimentary /btun Monk ' , > Before elaborsthat on the major

==yR==anam of, or justificetlee for, these inBraides d thh poucy statesment, it is

[ eweeping condusions. As a basis for

%e Commission's policy statement is useful to explain what we know about ra equally flewed in its treatment of severe the severe occidset risks to the pubhe, accident risk for future plants. First, the f etimaaldodeloa=aking staasmaat is a comp 1ste faDee.

the--a. dan's severe promises policy statement accident policy ht the ju,y Commission wiu make Sael decisione la Risks am common}y defined as the N N#8 b near term on the acceptabuity of new plant designs for severe accident product of the probability that an event will occur and the consequences of the Isee at least four fundamental flaws purposes. At h same time, the policy e

in te mnminolon's poDey statement as event happening. In regulating the .

j statement acknowledges ht key nuclear industry, b Commlulon makes M appbes to existing plants. Ftret, whue the pou elements in evaluating the ecceptabluty extensive use of a methodology celled i

1 f = 4=.cy statement naches a positive of severe ecc6 dent risk criteria for the probabahtic risk auessment (FRA). In

= ce the seceptabulty of the preparetice and evaluation of PRA's. l seimme amoident risk posed by existing ""d-r'laf a PRA the analyst calculatee '

- containment performance criterie, and the core meltdown probab111ty and. ,

' 'plask B hDs to artienlete winst that -

'* cHieria for evaluating b risk irisk lu; R fads to ldentify the relevant ' sivun a Particular core meltdows ; e contributions due to sabotage and scenado, the analyst the instimatas the

. e 4-s,-a issow evaluated hiamessing human performance-will not be Se b!!!ty of that risk:it faus to conageances to the pubIlc.no .

available for some time. Thus, b explaim those technicalissues war, Commiselce sees the bottomline of e Commission's approach is to egree to these PRA's in dedding whether to r==A-ed and resolved by b make final decisions on uvere accident r==d-daa in naching its positive impeove reactor safety or to relax the

> risk for future plants before the esiety standards even though such W andit fads to demonstrate technical basis for evaluating the nature

n. FRA's do not n=dA=* all contdbute to I

V PS PR 45' DM 31,1985lroset)

J

POUCY STATEMENTS core meltdows risks or quantify aD of N sprud in b estimated core

(' the micertalaties.

A typical result of a PRA which le metadown probebuities for a typical reprewatation of our undentanding of b tune. '

plant range from approximately one used by NRC in maching safety chance in one hamad (1ri per yur A serious consideration of the omre decisions is the estimated core to one chance in one hundred thousand nulldown risks would consider thus fell mahdown probabt!f ty of about one in (10 9 per year, with a unedien value of range of calculated risks and wand tee buand (or 1ri per reactor year. address forthrightly the question of one chance in tan thouand (10-*) per 3 However this probability estimate is year, give or take a few. However, there whether h risk in acceptable or chan based on what is esDed the is no unacceptable, both for the immedhate f that the median of the a fetwo and over the long ter m. '1%e

===tian" value. it is important to , cal led values reflects the actual risk adentand inet wkt the meanlag of *DY more than do the estimates of Ir* th-f aaloe's oaosideration of nesww this bottomline aanbar really is, per year or 1r8 per year. accident risks instead focuses as a Becsees of anajor inadequacies in the median number, ignoring h actual Another typical result of PRA's is the range of of values and the uncertamaties den bem, because of the vest prediction that about 1 out of to core complaxity of nuclear plants, becesse e inherent in es!ng a median number for meltdowns likely will result in lethel trummendous nunber of aneumpoone redietion doses to about 1.000 people.

deciala 1% l must be made in calculating oore Such consequences of core meltdown Since the foremost risk to the pubbe '

mahdown probabulum, and becsese from the commercial nuclear todustry accidents are attributable to degraded laryr scale com meltdown phenomena performance of the containment, which derives from severe accidents, adopting am poorly understood, no e, a policy that seeks to twoolve severe .

can come about in a variety of ways that accident ismee to a definitive ===a=r is catemtetion will yield a remotely are not precisely quantifiable. Because & most baalc duty which can be meaningful probabuity of cataMrophic of theu uncertainties in quantification. undertaken by the Commluloa km

,,, ,,,,,, g ,g,,,g pgg the fraction of core meltdown accidents anting its rupoesibility to decide what

,,3 g 7 ,,g p g ,, ,em,, .a. ,g which would lead to catastrophic constitutm acceptable risk to the Wual MmMas of the one consequences is actua y a range of

& Commissico claims in this Mwn probabih & W values. range could be two or three statement to have examined as M times greater than the above estimate: extensive range of technicalissues Petaas which themselm am not - or it could be two or three times lesa. reladng to severe accident riska to i pmanely known individaa! an=paa*=1 Picking the slaimum factor of 2 sad t reaciu'ng its judynant "that existles folke probabulues, basa error mtes, assuming there are too operating plants do not pose an undue level of risk and boretical models that are thought reactors, the approximate range of

• chances of a catastrophic accident to the public." The Commission's policy to dancrb most of the important statement does not, however, 4

reicalprocesses or engineering between now and the year 2000 would incorporate an axplanatia or for that

, Arry one of thne ladividual be anywhere between tL2 [1 chances le matter even a description, of the annet .

esthmates is as likely to be vahd as the signmeaatissues that have been estunate moulting from any one ofll* tho) u sand). ten and 0. cot tone chance in a ether thousands of calculations. There is Nrefore, b information before the Tuolved and b mannae in which they were resolved.Nor does it include a

(. -I a crectal, but untenable, andartying tw.i.alon indicates that there could description of the snethods of analyses -

assumption that au core meltdown be anywbere between a ao perosat used to resolving the issues or decision aaq===r a have been accounted for in chance and a at t chance of an accident at a critaria that were and for reach the antimatse. N analyst then scans aD at reacter in the next ultimate judgment. It is, therefore.s the 15 yeam that would rwult la lethal lapossible to discern the bases far the of the estimates and picks the probab0f ty value at which half the doses to about 1.000 peopla.The range of Commiulon's decision.

chances could be larger than this if one settamates are above the half are below.

considers eB contributors to b core pgg

%ie sumber is caDad the median. It la, maltdown probabihty and all A paramount concern regarding the according to the N==5asion, the "but uncertainties. Likewise, the amnber of wtimats". When calculated in this way, acceptabuity of the riska to the public deaths could be larger or manaller.

however, one cannot say with any that must be resolved is how to reach a Admittedly, there are many ways of judgment on this issue in the face of ream =nce that this median value is the going about estimating the ranse of tree som meltdown probabuity, enormous uncertainties which are op to riska. However,if them la validated too tisws the audian value used by the Noneth6less, the Co==fa=4a=

i quantitative information on care Commission. Depending on how much airbetrarey choosee this mediaa sunber meltdown risks that is better,it has not

} to one la making its regulatory uncertainties are factored into the d.,6. sana i yet been d==aaatreted. be, becaem decision. judgments could range toe of the amany uncertaintin involved in req substantial e!! orts to reduce calculating both the probabuitia and core tdows rinks to doing nothing sw % ,,,,,,,a ,,,,,,,,,,, the consequences of core maltdowns, about them. Scientifically accepted data a meer meeses.e by e. Aa.tmer commsem en n mm,aar one number dow not give a true pictum and methodology are not available at res entse nepairat som mesame of the actual riak. A range of this thne to red" 4 substan toes wie she cammina m The Acas roommmmend me poealbusties le a more accurete

"",,", QSyg*,gM",,h nacertainties .o eat, as the staff of the NRC has repeatedly told the escammene ens rummh in a substamalaa modoresenate w n. es of uws. m Game PRA andyees hear noir estmense se to CoEunissioit it is " mandatory" to

==a=# reek esametes As hadecetud a s

es now . e-- h i.tes comaldar them la any application of dak madam h that pelat en a spectriam et which half of endereed one of te mesas ,shesL The test uma was whom the Cammhsk s

• dweed WASH leoD assessments. .- , -

m esame tan eh.= and har ein honow.The ga ener s.reer sesdrl a tv s and en me d em. h being informed of b "' ~

> the e.wess valm of as seanner of rinha and to L

== when eer - ap uncertainties in the risk estimates'.'the ah smand he W enks?

cam d amorey ant ila cant tsaa pouceNunarassa sensesar (er d the pewtwood h=idad~ almply ignores theen.The tww im gans to provide any basis y-t Deceanber 31,1985 (reset) PS-pM6

4 POUCY STATEMENTS ,

< for Rs decision to ignore tha*

r uncertannues. Absent some totional estimates" of the core meltdown risks calculeted core meltdown risks by half' without any consideraUon of the effects Unless such a reduction can be i treatment of thm uncertainties or a of the uneartatnues.nis hpproach can canymdng justificadon for why they can " demonstrated", the Commisalon will 1 be ored. the pebbe can han little leed to a decielon to dolns nothing to not consider requiring b change.nle mduce core meltdown risks. Factoring is a much higher barrier to requiring once in ee Commission's into the decision the uncertaintin in improvement in reactor safety than b I conducion that b risha to the pobhc ntimating the level of core meltdown frase a severs accident et a nudeer poucy statement would have a belfeve risks would lead to a decision to search is the Commisalon's policy.

powerplaat are neceptable, ne only for ways to reduce the riska. However.

evnllaoss explanation of the NRCs - Further, the th=wlan's provisional given the current political climata, there safety goal is not intended to regulate on approach to making decisions in the la little sympathy for beckfitting existing the basis of preventing core damage Goes of these ab=h sacertainties is planta. Ms, the Commisalon chooses to accidents, as implied in the above given om pages las through 140 of rely on a faulty number which supports NUREG-tetVNRC pobcy en Fueue purported fundamental objective.

the outcome they prefer and to lenore Rather, b safety goal assumes that the Reactor Deelgne Decislams on Severe Accident leeuse in Nuclear power plant the uncertainties those that are hnown containment is an independent bulwark and quantified and thoos that are not capable oflimiting the external relene Regulation". October tes4. About half of quantinable. of todioactivity to modat amounts for

- the pages am blank and os rand =' most oore meltdown accidents.ha, Whot level of confidence does the are not much better.hls discussica of Commluion have in its hdgment bt according te b Commission, them is no uncertainties is inedequate and fails to core meltdown accidents preunt no need to mgulate on the basis of provide a sufficient basis to justify the undue risks to the public? ne preventing core meltdowns. I am not as r ==tasion's sweeping wacloaloas on Commission nowhere emprases b sanguine as the Commisalon on b the acceptabluty of the severe acddent degree of confidence it seeks to ensure Fieb- acceptability of core maltdown that catastrophic accidents do not accioants.Evenif the containment Another fund ==,-tal ls'ous happen. Yet, ther hmtaalon's chlaf happens to retain moet of the emselutles is the level af risk to 'e safety officat recently wrots "In view'"p[ redjoactive fleefon products in the neut pubtle that r===anably abound be femad the uncertainties ..- severe accident, another accident equal Beyond making a sweeping of secessing severe t' to or more severe than that which uslon bt the severe meddest risk -"rlek,the levelofassarumos (or -

occurnd at %ree MDe faland would be st die existing plaats does est pees an . confidence) of no undee risk to b unseceptable to the pubhc and b undne risk to the public, tbs Commission ' public is regarded as no less important Congress and would be diastrous for anile to addrus this fundamental than the estimated levelof risk itself the nuclear industry and b NRC.

question. la fact, the Commiaalon's (emphasis in the original)." Latter from But more importantly, the f.- technical staff is t now embarking on H.R. Denton. NRR. to A.E. Scherer. Commission's belief that the a program of is that "wiD form Combustion Englosering. be dated containment w!D retain aD but madant part of the baals for a r ammiaa" December 33.1984, subject "SECY.as. amounts of todioactivity during most judgment oc the level of safety presently am Severe Accident policy =4 core meltdowns is not yet espportable adieved by existing plants for sever

• Another problem with the based on scientificeDy accepted occadents."'Since the ra==3=ah is Commlulon's polley statement is that it Principles and methodology.There just begi this program,it cannot clearly contradicts what the simply is no actuarial experience er earve to jus the Commission's direct experimental data ca large scale Commtmalon is dolns in other areas. For

- }edgment on saceptability of the example. in this pobey statement the core meltdown'phaaa-ena or sevwe accident risk. Commlulon states: "A fundamental containment performance i

am its Indiin point dedsloa, the objecuve of the Commission's severe charectaristics given a core nultdown.

r==taalan adopted specific point accident policy is that the Commlulon In the past, estimates of the quantitiu of estuostas of core meltdown risks for the intends to take au reasonable steps to radioactive releases to the environment i Indian Point reactore and found them to reduce the chances of occurrence of a beve been based on not much more than represent an acceptable level of risk. In uvere accident involving substantial interpola tions of extrapolations of the coures of developing this policy damage to the reactor core and to approximations. It la for this reason the statement the Commission expressed mitigate the consequences of such an Commission has an ongoing program, muca laterest in the bottom line results accident should one occur." However, which has cost a quarter of a biluon of al/ completed PRA*a, whether the compare this statement with b dollars in b last few years,in an reported point estimates were the mean Commission's proposed backfittirig attempt to bring some science to or snedian,ne technical staN has standard:"fte Commission shall estimating the core meltdown risks.

repnatedly cautlooed the th=iados require the backf!tting of a facility only However even in this program the data that such bottom line numbers are not when it determines, based on a being generated are from limJted smou credible. What then is the basis for the systematic and documented analysis scale testa.

Comuniasion's position that the level of * *

• that there is a substantialinaease has, a readieg of this po!!cy severe accident risk posed by the in the overaD protection of the public statemenfindicates that the .

existing plants is acceptablet o - bealth and safety * *

• to be derived Commission's. claim that in developing he Commluion's decioloo-making from b *>ackfit andthat the direct and this poucy statament it has examletd an process in developtrqs this poucy indirect cost of implementation for that extenalve range of issues is incorrect. It statement is simply to rely spoe " point facuity are justified in view of this shows tother fbet therammtaston either increased protection."(emphasis added) examined b wrong lasues or gave short

,,,, w,,,g .,me p%, r,u,, The Commluton has already defined a shrift to the fundamental fuues.

p ,o e w p.e.wn a se Acew substantial increau in protection as in failing io define accuretely the level bem 6a Nesient Pow.r Plaat Anguienm6* Oces' meaning a backfit that would at least of severe acejdent risk at the exfsting sent p ar. reduce the " point estimate" of b Pl ants and to addren the need for

) PS-PFI-47 Deosmber 31,1985 (reset)

. ..t

l POUCY STATEMENTS addaonal changa to b plants to make on acceptable wa

& dak accepteW fu b long krm. benent analyses. ys to perfa:m coet. continus to se ca. Wbst actuarial b c-mlaston is speadas pad Further, guidance from b experience we n is sevmly hmited .-

laDune to deal effecunty wie se Comodulon is needed on whether to by ourlack of detailed understanding of

, sever, Ant quadon.& concept d emphasise core meltdown prevention b puformance of b plants, their the reactor containment ortsineDy measurn or con meltdown mitigation daigns, their weak spota. and because '

musum. Of courw. La ordu to dmlop of the wide variations in b designs and evolved as a vweel to contain a fuD core k utWty capabW6u. Furbt, b seeltdown. But in the mid.19eo's, the a pobey on b latter (whether for existing plants or future plants), one usefulnus of actunial experience in reactor de.igners began lacing high drewing broad conclusions about cores into the u ne must first identify the root cauen of of containment.The core meltdown riaka. One must also commercial nuciaar reactors is highly y best of controvelal and fraught wtth those higher powered ooree was so high develop a policy on containroent performance uncertainties.

tiset the acetal ==t vessel could so Bona.

Unfortunatal , the Comm4aion - 7k e-mlaalon argues that credit longer be considand as an effecthe pfuses to can be taken for the improvements nedspendent barrier to the release of the to addren these implemented to addrosa specifle close flesian products evolved during a core issum. Aa eNect!ve guide to regulatory decision-making on the treatment of calis such as the DC accident, the meltdown. At that time, the Ateede Browns Ferry fire and the Rancho Seco severe accident Lesoes requires an Energy t%=miaalon's Advisory trenaient. Each of these were previously th-nittee on ReectorSafeguards understanding of what is expected by way of containment performance, of the unrecognaed (or 41 best inadequately (ACRS) began urging the development appreciated) accident se and implementadon. in about two years. root eenass of core meltdown riska. and also true of, for example,quences. This is of the methods for performing sound the of safety features to protect against a cost benefit analysee. Yet all of these Susquehanna station blackout event loss of coolant accidentin which the elements are missing from the from a single faDure, the Indian Point emergency core cooling system did not Commission's policy statement.%e vulnerabili to a single failure of a week.%e ABCand theindustry battery, the scrcalled interfacing e-6 ton's actual decision making system LOCA's for bolling water bebevt1 that sufBcient data were guidance la this policy statement is eva3able to lus;ify with a high degree of umsted to th.e statement that a new reactors. Noos of these latter events conBdence tb adequacy of the then, requirement might be imposed ifit were identined or high!Ishted through edsung safety standards. Derefort the involves " low-cost changes in PRA's nor were they expected to be.

- AECignored the advice of the ACRS. Procedame or minor design given the level of detail that typicaUy Om the yean, es AEC and the MC anodifications." gas into a PRA and gfun the sub}ective after it have retterated these sweeping Da Commission claims thet PRA's natum of PRA's. Whether tbsee latter and optimisuc statemente on ave identify b plant specific vulnerablutin events should be called close calls is acddent risk. At the same time, tbs that dominate the core meltdown riska. argushle but bit occurrences certainly

- suggat a need to considu the root aumerous techn! cal Bews la the it is true that PRA's can idenufyy some of causes of significant operating events hada='s judgments have become the vulnerabilldes to catastroptue and the conective meaning of those i readily apparent as more in'orma6an accidents. But the Commluton's i.

)

\ ratiooale for retying upon PRA's in events before pening judgment on the \ I and data nearding the Is" *i of Safety of b reactors has become avadable.' assening coe r meltdown risks begs the acceptability of the level of safety l questions: what of the uncertainties in schieved at existing power reactore.

When aD of the avellable data are PRA's? What of oversights in the Common unae also suggests completing t e:.I believe it fair to say that analyust What of b multitude of such an analysh before developing i the eettmated uncertalades in the risk assumodons and approximations in the guidelines for the design of future e.Wla n ons today are as large as they reactors. Yet all of these concerns an were at least tan years agc. Yet, the PRA'ef What of the residual aisks once the specific vulnerability has been swept aalde in the Commission's policy tww.t..lan is once again sweeping statement, fLxed? nose questions are germane to aside thou acertaintias in order to moolving sevm accident inuw. Yet ne DC Action Plan called for a large make b same unsubstantiated and they are not addrewed in b number of modification to the opere ting overty optimistic ganaralizations about Commission's policy statement. plauta,in addition to those the acceptabuity of the current level of Operational experience gives modifications. the Action Plan wvere noddent risk which han bees ,additionalinsight into the level of coaunitted to a rukmaking to consider p,,,em wrong in b past, safety, Actuarial experience with to what extent,if at all, existing nuclear  !

power plants should be required to deal Needed1stpor=anaa'a reactor accidents ladicates that the averses core meltdown inquency is not effectively with damased core and core A disciplined approech to deciding above the opper hmit of the PRA results. meltdown accidents. bre was to be a  ;

whether to requim core meltdown risk demarcation between thoes plants Core meltdown accidents involve alrudy operating er under construction reduction measures should not only multiple feuures and a p lon of specey b Commluion' mots eat make clow and the next tions somewhat Because b y~amiasWa perceind 2nerstion o o identiBable. lf the industr

,,n ad,dma,hg

,,,,,,,e_ c accul ,o,.,,,s,hould c, b core mondown f,og y average cy wm -of 19eo bt there would be a long hiatus in high as 1tr* per reactor year, one would new ,iani orderm,h use eesad io expect more clow calls on con reconsider the General Design Criteria,

'Dr dew carmt(Ma* 6an mte d b daign baus, and the other meltdowns than appear to have h%%4 O,, ne,.

d nguhtlom in Hsht of aD that had been 18,87$a pec ,, ,,,,,, on manaw occurred within the more than 800 reactor years of UA nuclear power lurned through b years of experience in e m_ >desums s Dew okreet Nudase experience. But such actuarial with large powat reaetors, including the s.fgy.o,i rmwrof awarewr DC sccident. From this in-depth g un#nner d Whe== Prem N interences must be made cautiously in part because the operating reactors assenment of the strengths and weakneup of b large power reactor i

December 31,1985 (reset)

PS.PR.48 s

)

POLICY STATEMENTE dulgns and the apprwth taken by A /toriono/Appmoch to S,y, ,

atibtiee loward constructing the plants. condluons are inside the reacter building.

Accident Decisionmaking a dent m*

/, NRC would then be la a stion to ' 't*8' '

articulate safety princip a that it What the Commission should have

,,g,und W

l' ex tad to be inwrporatodinto des (gns done in its policy ststement is to set puts ha controuins the eacsdent into as expertowntal mode lh happeasd durtne b futum appucadona. Mus, the forth precluly and in understandable thret dayL the uncertainty of whether se Cowalasion in W signaled tbm terms what ourpresent estimetion of the would be a significant ste forward in risk of severe accidents is, whether the accident could result in mehr reisenes of edvancing the protection the public, neoecder is too Wdens e e the bmmlulon believes that rid to be enormm dannese to es plant, es expensive

• • d P*UF d Re Ceaunission la this policy statement acceptable or not, what specific takes several steps backwards- technical support can be offered in - ' * *8" ""' P"**

One backward step discussed above support of that judgment, and how the den we's es euver relevant uncertainties have been www aseid hm hs-- " iaccident la the Comuniaslan's decision to accept treated.he Comuniasion thould also had already sons soo'lar se make it esterebie.

thw more meltdown risks as Ibey exist in have come to gripe with a central While throughout h settre document we the current generation of plants without emphastas that fundamental chasons are even addressing some of the most quesUon in our regulatory am:that is, given our present state of owl someary to prevent occidents es serices as fhadamental step is aP issues. Another backward concerning severe accident risks, sho d ThtL we mest not smene that es escident of

_^ of the expressed als or smetw sertonmen omasat happen .

destre for a fresh look at light water we continue to pursue posalble asala, me if te changes a renamend m 1 reactor esisty for future deelps and the improvements in severe accident **

• D ' *d*

prevendon and mitigation?If the insistance on haprovements is tbs lent of severe accident risks for any futum Commission does not believe that the

,,,, % yp,,,"y,,[,ygg, g.w ,'"P,",h,,j'M.a plants. A third backward step is this pruent level of sevm accident rid is potentid impact of sech as acesden on acceptable for the remaining 40.yur life public beelth and ufety, should see occur is policy statement is the return to the the future.

of some exis plants. then the pbliosophy of the 1seo's and tefire that Re Commission uld outline its p construction permits can be lassed for %pM of es prwideM's CommlMoe on based on only partial deelen this lon e Acddent at Aree MoeIsland accepta bounds.g-term risk wi in informstles. Only through such a In order to reduce the severe i ent acc. .14.

For future roector orders, nuclear process can the technical community, de ova Mme k accepublavels,I other public makare and the wWd han mdwiden few spdac utilities vu have expressed a . public undere

, and accept th* inibunk Firet.Iwould hn reqdid a desire for plant designs that are simpler.

safer, and mon forgivin6. Both the Cornmission's judgment on the severe detalld ward for plaW4pdoc Bectric power Research Insutsta (EPRI accident rid quesdon. Unfortuna tely, equipment and design vulnerebHities at such an analysis is nowhere to be found ud Wshe plad 2 b and Edison Electric Institute (12T) have) in de Commission's poll statement. ,,n,ect eau wukneues w% Wd and

/,, - impressed on the Commiselce he need for a fresh look at Based upon the m discussion' constitutes significant omtribotors to s I water reactor I would have rea ed the o

,. technology.%ese n ty sponsored ee rid d a severe soddent. '

conclusion. First, the risk to the p Seed 1w organizations have also todicated that posed by severe occidents at the , lic ,,,,g g,dd han g g, bibW a ,,,j plant construction for new plaats should existin pedem m & m, ,g sd '

O'l'o'o'* "Jm*f"e 2*gni.&e sn"a*g plants is not acceptable special emphasis for the on areas weakness plant. Yet none of these forward A%e*:"Jfn'E2*"

should continue to pursue cost effecuve a g,2;o$;';,74t'i'::=l,-

• thinking requirements are to be found la risk reduction measuru for thue plants. good examples) and on speci8c stilldes

{

the Commiaafon's policy statement. I would ply the as. low as. reasonably.

lasteed, tbs Commission states that it achievab e (Al. ARA) principle to 4 a Wory Mmar@M hm I De W 9.1988 opmhg put M b will be satis 8ed with mere rannaments la the old designs and bt it is willing to 9reducing severe accident DWrisk. h subject er mph a cubuk pmn pardal daigns for

• y*,O[gd re a again demonstrated the dangere ge(,y inherent in the combination of a

""**** **** '"' improvements need not be pursued. I Inarginal plant design and a utility with I cannot leave this latter point without would have slanply acknowledged the marginal opm pedmnance a sad comunentary oc h Comminion's obvious: that the public and the priorities. One lasue la this Congress willnot tolerete. and the hird. I would ve initiated a that compmhensive assenment of the lent commanded great latemet wi ee industry and the NRC cannol allow of safe and the exia lants have Ccammission was kw to circument itsanobr uvm accident as serious as achine ne object of e5 art would suon eat requita a comparison of the %ree Mlle Island accident or worn' be to identify the root ceases of severe a esign to the staff's Standard Review My views in this rian. n is e - w - m u sia by &e go, ,p,mg ard are identical to g,y,,,,y amident risks.his eNort would also ons of one reactor vender. Commission nea,[,, ,, ,sa,, identify a,o.

possible measures which offer ee - mise d eigM8ca m>y redudag the Commissios's efforts to use Whetherin this rticular com we eens severe accident risk by overcoming the this policy statement as a webids to dose a esta ociden ornot,this adverse efects of equipment ,

t the reactor vendor to dreenvent Comunisalon's regulations took bmdh haan umrl precedence over any Commlasine (*"g' e(The should not be allowed w ' defidencies and areas of @[ ,

,,, g, consideration _of such fun <tamantal The acesdent got suf5cian out of hand m despita our best eforts to that thom ettempting to con it wm my Issues as the actuallevelof eenre first two initiatives. Indeed, as the accident risk to the public. the opmting somewhat in the dark. Whue today  !

the cauus are mU undmtood, e months Commission's' chief safety officar noted acceptability of that risk and measures to reduce that risk, potendal eher the eoddent it is still diffloult to know in a June 27.1985 memorendum to the

(.

Executive Director for Operatione: i the precise state of the com and what the '

PS-PR.49 December 31,1985 (reset)

_ - _ - - ___--- )

POUCY STATEMENTS t heheve that'the r$ cent Devis-ilie~es event (STA) poeltion by allowing licensees to alastrates that, in tlw rea! world, erstem and combine one of the required Senior intehded that use of the dedicated STA esoponent rehabulda can degrede below would be an interim measure only until ..

Reactor Operator (SRO) positions with these longer. term goals were achieved.

those we and the todustry routinely assuene in wtimating oon melt freguancsw Our the STA position into a dual role (SRO/ These long. term initia tives

' engulatory abould STA) position. Option 2 provides that a collectively result in an improvement in marsine licensee may continue to use en NRC.

agelset degredenen palso w moset the capabilities and qualificauons of the

&e aceMalnda in nr RA sensates. approved STA program, with certain shift crew and their ability to diagnose modification s while meeting licensed and respond to accidents.Dese Finally, for future plants, I would have opersior stamns requirements.

explicitly required measwes to improve Inidatives include shift atamns arrscTive oats: October 28,1965. increases. training and qualiikation the todgin ci safety agalast severe ' Post runTMan unposessAT>oel CONTACT program improvements, hardware accidents idl future plants and to address modifications, amphasis on human the aldtabe of the past Buch measurve Clare Goodman. Office'of Nuclear Reactor Regulation. U.S. Nuclear factors consideredone. procedural sould lachsde requirements for greater Upgrades. and development of extensive Regulatory Commiselon Washington.

simpucity in plant design. improved DC 20655. Telephone: 301/492-4894. eineqncy wganizauons to maintainability, and a requirement for

• surnasesm&RY pfPORMATIOsc Capabilities during essentially complete klant designs prior M c M -

to the issuance of NRC ppprovalier the Background start of plant construction- Draft Policy Statement Following the accident et Three Mile I believe that these measures would Island in March 1979, a number of On July 25.1963. the Commission be sumcient to bring the risk of severe published in the Federal Register (48 FR studies were conducted to determine accidents within acceptable bounds for why the accident occurred, what factors 337e1) e Draft Policy Statement on the remaining operating hves of the might have contributed to its severity. Enitineering Expertise, on, Shift to exis plants and for the operating ' ' '

hvos any future plants. Moreover. and what the industry and the NRC could do to prevent tne recurrence of the eng eer ng and accident a sst ent sech an approach would de much to expertise must be available to the restore pubuc conndence in nuclear same or a similar accident ihes' studies concluded. emong other thirfgs* ETny'ti 8 t all nucle pow" power and in the effectiveness of the '

that a number of actione should be NRC's regulatoryyrocess. It is -

taken to improve the ability of shift '

unfortunate that me th=laston bas , operating personnel to recognize.

Engineen.sft Poucy ns Expertise on Statement Shift offered on

• chosen another path. However. key licensees of nuclear power plants and diagnose, and effectively deal with plant decisions remain to be made by the applicants for operating licenses two Commiselon in adopting a Baal transients or other abnormal condition.s.

. To address these recommended options for meeting the stamng backfitting rule and a final safety goal. Improvements, the NRC initiated both requirements of to CFR 50 54(m][2] and noes decialons represent a Baal short term and long. term efforts. The the requirement in NUREG-0"37. Ilem opportunity to come to with many short term effort required that as of of the pivotallassesav 1.A.1.1 for a Shift Technical Advisor -

la tble January 1.1980 each nuclear power (STA). Option 2 gave them the f policy statement la that regard, it is plant have on duty a Shift Technical opportunity to co 4

encouraging that there appears to be an emerging aa-asus within the NRC Advisor (STA) whose function was to provide engineering and accident Techru.Opuntors,mbine the licens 6"I" (SRO) and Shift cal Advisors (STA) functions.

senior technical staff and within the assessment advice to the Shift Under Option 1 licensees that did not  ;

ACRS in favoe of safety improvements Supervisor in the event of abnormal or i

want to combine the SRO e,nd STA  !

to reduce severe accident risk both for accident conditions. ne STA was functions could continue with their existing and for future planta, required to have a bachelor's degree in approved STA program in accordance {

engineering or the equivalent and {

with the description in NUREC-0737, '

,, ,, g,p specific training in plant response to " Clarification of TMI Action Plan Puushed totes /es transienta and accidents.The STA Requirements.

Eners>.1easms requirement wu identified to licensees Interested persons, applicants. and via NUREG-0578 (July 1979) ' and licensees were invited to submit wn; ten '

NUREG 0737 (November 1980) and was commen's to the Secretary of the Commission Poney Statement on later mandaled by plant. specific Commission. Fouowing consideration of Engineering Experties m;SW Confirmatory Ordere.

the comments. the Commission Concurrently. the NRC and industry amended the Draft Policy Statement, as Amtwer NuclearRegulatory embarked on a longer. term effort aimed discussed in the following sections.

Comenission. at upgred staffing levels and the

'Y ' " Commenta on the Draft Policy Statement "8 i8 ting stake i r vir Sh na n- A total of 34 responses were received machtne mterface and increasing and evaluated. The pubhc comments euesstAam This Policy Statement capabilities for responding to presents the policy of the Nuclear emergencies. At the time the STA related primarily to the combined SRO/

Regulatory Commission (NRC) with STA position. The following discussion requirement was imposed.it was highlights the major points reised in the respect to ensuring that adequate engineering and socident assessment .

comments and the resolution of those

, expertise is E M by the operstmg ,% ,, comments. A detailed enslysis of all staff at a nuc!(at power plant.This t.r %e. sin m. ,,,,,ci.w,t , % no, public coinments and their resolution Polley Statement offers licensees two er :=erens for e tw mi ow Nac evw onumene was also prepared. (Copies of those Reesu hr H Suse NW. wasMnsm DC. The letters and the detailed analysis of all options for providing engineerirq expertise on shift and meeting licensed C,"'F(Qgggq g rn.aame w t,y .rrmee w cPo. r.o. sei staar the public comments are available for operator staffing requirements. o public inspection and copying for a fee Wee %eien. Dc anmwona. my mer el o tw Option 1 provides for elimination of 8 at the NRC Public Document Room at

• d Imm the NewnelTecWat Inimuan 1717 H Street NW., Washington. DCl the separete Shift Technical Advisor

,T"*,7',jg'"yg*"c*. s2ss p'" Of the 34 letters received.18 included .

December 31,1985 (reset)

PS.PR 50

]