ML19263C958
| ML19263C958 | |
| Person / Time | |
|---|---|
| Issue date: | 10/31/1978 |
| From: | Gossick L, Pederson K NRC COMMISSION (OCM), NRC OFFICE OF POLICY EVALUATIONS (OPE) |
| To: | |
| References | |
| RTR-WASH-1400 NUDOCS 7903200157 | |
| Download: ML19263C958 (6) | |
Text
&
~
/
5 jg af V
-- -x ? 5 l
l
- f..-t (.
bove cer y g/ E Revised: Nov. 2,11 0 h/Vl NRC STATEMENT ON RISK ASSESSMENT AHD
\\.
Y ft f'
4.s s u e 6 i
THE REACTOR SAFETY STUDY REPORT (WASH-1400)
NRC STATEPINT 01 R15K A55ESSHINT AND ig"gp/
TifE RE1.E102 SJJETY STUDY REPORT (L' ASH-1400)
-\\
IN LIGHT OF THE RISK ASSESSMEHf REVIEW GROUP REPORT y
IN tlCJfi 0F THE RISK ASSE55 MINT REVIEV GROUP REPORT g
g
....._... -- _+
-~f--
iritroduct ion Following an interchange of letters with Congressman Morris K. Udall in early 1977 11owing an interchange of letters with Eongressman lio.rris K.
o on the Reactor Safety Study (RSS)* and its Executive Sunsnary, the Risk Assessment Rail in 'early 1577 on the Reector Saf ety'5tudy (RSS)* and its Review Group was forined by the Comnission in mid-1977 with the following charter:
becutive Sorrnary, the P.ist Assessment-Review / Eroup was formed i
"The Review Group wiIl provide advice and information to the f
>y the Eonrr.ission in tid-lf72 with the following charter:
i Coanission regarding the final report of the Reactor Safety Study, "The iteriew Eroup will provide advice and information to WASH-1400, and the peer conunents on the Study, advice and recomnenda.
the Ccunission recerding the final report of the Reactor l
j tions on developments in the field of risk assessment methodology Safety Study, W.Yi-1400, and the peer connents on the j
,i and on future courses of action which should be taken to improve this Study, advice and reco::rnendations on developments ir, the il methodology and its application. This advice and information will field cf risk assessinent methodology and:en future courses i
assist the Connission in establishing policy regarding the use of of action w'iich should be taken to improve this methodology risk assessment in the regulatory process, in improving the base for the and its application. This advice and information will use of such assessments. It will also clarify the achievements and N
assist the Conraission in establishing policy regarding the use of risk assessment in the regulatory process, in ir. proving l
the base for the use of such assessments. It will also The Review Group, under the chainnanship of Professor Harold W. Lewis of the N
O clarify the achievements and lioitations of the Reettor University of California, Santa Barbara, presented its findings and recomnenda-D D
tions to the Ccaulssion on September 7,1978". The report of the Review Group 3
Safe ty Study.=
was issued as an N document NUREG/CR-0400.
- tt F.evin: tro;;r, ur.c'er the chairs.anship cf Professor Harold W. Let:is
- The 65 dRuments an ef fire ~to apply risk assessment techniques to the quantifica-
-' the l'r.iversitf cf talifernia. Sar.:a Sarbara, presented its fincines tion of risk associated with nuclear power reactors. Among the shethodolugles used in the RSS were fault tree analysis, event tree analysis, prnbabilistic modeling, statistical analysis, human factors engineering, and r adiation epidemiology e f!! :e:xer.s ar. ef fer; tc appsy rist assessr.er.t te:hr.icses to
- The other members were Dr. Robert J. Budnitz (Lawrence Derkeley L aboratory.
e cut-tif':e ti:, cf ris k essocia tec: with nuclear po.ar reacters.
University cf California). Dr. Herbert J. C. Kouts (Brookhaven Hattonal t aboratory),
-.c the 'rstbc:!cEins : set in the ESS >gre f ault trees. tver.t Dr. t! alter Loewenstein (Electric power Research Institute), Dr. llilliam D. Rotn:
id, r:Mbilistic r:c'eling, !!atif tical ar.giysis, hu.tr. factors (Environmental Protection Agency), Dr. Frant von Hippel (Princeton University) ye s-9.E. e.c' ra d t-1:n epichiclegy.
and Dr. Fredrik Zachariasen (California Institute of Technology). Or, Budniti is presentely on leave from the University of California and is serving (since Augus t 1970) as Ocputy Director of the NRC's Of firn of Huclear Regulainry I'c.rar ch.
Mhfi.
\\
-l ~ (Ledwupd) s OmTi t_ o j-
\\ i The Cox.ission agrees with the Review Groep that the RSS was a step
, Third, with respect to the continued use of risk assessment in nuclear re.aulation)
~
=
.orward in the efplication of risk assessrent rethodology, including l
the Ch.unission agrees with the Review Group that the RSS was a step forwers' in the the use of fault-trees and event-trees to understand accident sequences
,lncEd5g thkc~omprehensivehse of applicatto of ist assessment methodol and to take a;uantitative estimates of reactor risk. The Comission fault-trees and event-trees to understand accident sequences / make quantitative believes the RSS is a significant contribution to the development and estimates of reactor risk. The Consnission believes the RSS is a contri-rp;'lication of risk assessr.ent methodology to nuclear reculation.
bution to the development..and application of risk assessment methodology to nuclear regulation.
The Review Croup stated that the f ault-tree / event-tree saethodology The Review Group stated that the f ault-tree / event-tree methodology used in the E'SS used in the P.55 "should be aaong the principal ueens used to deal h th ger.eric safety issues, to formulate r.ew regulatory requirenents, to "should be among the principa.1 means used to deal with generic safety issues, to assess and re-validate existing regulatory req::irements, and to formulate new regulatory requirements, to assess and re-validate existing iegulatory evaluate neu desi;ns." The Cor.nission concurs, but with the additionti requirements, and to evaluate new designs". The Conunission concurs, but makes rom.cnt that, schile the staf f does utilire f ault-tree er.d event-tree the additional coninent that while the staff does utilize f ault-tree / event-tree L.alysts to lic h revier:s tf virious issues, further develop.sent analysis to aid in reviews of various issues ecision to use this method as
- f risk as sess.e.! tec'.nolc;y should be persued looking' toetr6s a principal regulatory technique would be a significant departure from current f-rcs e:' r.;rerical rise: evaluation. Accordingly, the Ccmission will practice and should be studied carefully. An evaluation is needed of the role risk s;::cr: tr eg an:ed n:;rra of 1 prove: ents.tnc' extensicas of risk
{
assessment should play in arriving at licensing and regulatory decisions and in I
assas srs.t tee:r.*.,
et-- is, ene ractice u pr:.c te their prt;er and setting research priorities, taking into account the findings and recnninendations of the Review Grcup.( Iurther development of rist assessment tachnology should alsn
( "tctitt Ulf by Eli sep.erats cf the 15C.
i MITill be pursued looking towards improved numerical risk evaluation. Accordingly, the y
g Commission will support an expanded program of improvements and extensinns of 4
. risk assessment theory, methods and practice to promote their proper aivi ef fective DE/IT:. 13-i-78
-B-use by all segments of the NRC.
As the disciplir.e of risk assessment develops, it is expected that As the discipline of risk assessment develops, it is expected that revisions and rrvisions and corrections of findings.will occur. The tonsnission corrections of findings will nrrur.
Ihr feminiss lun secs such adiu.imeni s a. naim al sees Wh f(iusiaenti A5 ntlural and rf regni7c5 iheir ipporttnce in and recognitM their importance in ipining Imptuved underdan finq of rl*l.
t>e InIng ly.e uvri unds st amling ur risk.
e
. (&t e,au d)
~
s
-1 Co raission I.ctions Having completed a prelirainary assessment of the Review Group Report, Cor:rnission Actions the Ccmission t.as deternined that a nu:nter of follow-on actions are Having completed a preliminary assessment of the Review Group Report.. the required.
Cornnission has detennined that a number of follow on actions are required.
f Copies of the R15 Executive fumary will be distributed only when~
Copies of the Risk Assessment Review Group Report (NUi.'*/CR-0400) and of this 1.
ccompented by c: pies of the lieview Group's report and this state:aent.
statement will be sent to all known domestic and international recipients of the Co;,ies of the Risk Assessr.ent Review Group Report (1:URIG/CR-0400) and Separately bound copies of the RSS Executive Suninary will no lonce$ b'e }
of this stater.er.t will be sent to al) known recipients *ad all future distributed;f and copies of the complete RSS will be distributed only when rec.uestcrs cf the F.S$.
accompanied by a copy of the Review Group's report and a copy of this stattwent.
2.
The Cor.ission hereby instructs the staff to ensure that:
risk asses srent rtthoc' ology is used prc.perly as a s ic.portant.
,177 et.s cf caiting the r.> clear replatory process; and k
~. _..
E. T.55 rist es thates anc: codels, ht.en used in the regulah At indicated earlier, per$ ding completion of the short-term tasl.s and policy state. 4 1
rt:ess, t eve thc: cutt. ext.inati:3 ar.( cor. fir. rtiori that ment described below, the Coninission has instructed the staf f not to use RSS ths'.r itses tu t.e cc:. text of tttir a.,;11.atic.s reflect 2
models, risk estimates, or conclusions -- except for the fault-tree / event-tree i
the full ran;e of uncertair. ties and ceficiencies.
methodology -- as the basis for any regalatory decisions, practices or staten.cnts thout the explicit approval of the Executive Director for Operations.
j by&l (w h e y, w e i h sps,j b
iwkh~,
/w/n7 gy /.,
no 7
p i.1.
a w
Iw lf h %.
..m.------
.. c e-
e I
DMIT: 13-1-78.
l The Connission has specified several short-term tasks that are expected to be
- 3. The Cmaission has specified short-tena tasks that are expected Comp eted by the end of 1978. These are described below.
l to be cxrpleted by the end of 1978. These are described below.
A. The Co::r.issien has instructed the staff to review the extenV 1.
The Consission has instructed the staff to review the extent to which past to which past, current, and pending licensing or other reoulatory and pending Ilcensing or other regulatory actions, including Commission, actions have relied or will rely on the risk assessrent modele and ACRS and licensing board actions and statements, hay ( relied on the risk results of-ti.e R55. The Co:;raission will also review its own assessment models and risk estimates of the RSS. The Connission will actions and state.ents' as well as those of the ACRS and licensing examine the results of this review to detennine whether the deijrce of boar d s. The [omisssion will examine the results of these reliance identified was and continues to be justified and to decide whether reviews to determine whether the degree of reliance identified f
regulatory modifications are appropriate.
is justified and to decide whether any regulatory modifications
~~
"~
cousa 2.
The Cocuission has Instructed the staff to develop and present for Conrnission are appropriate.I Such modifications-411-include developmenh I
review and approval procedures for confinning that when RSS models or risk e
s
/
of thatever pidelines say be determined necessary to minimize estimates are proposed for use as part of the regulatory process they are the potu.tial for any future uncritical use of the R55.
technically valid and used in proper context. _
r fW Vhif ff Vg1
&W f YYh f f( y.l nr+
k a cw ny,
i Q9
t a- (wa) b s
- 8. The Cer.r5ssica intends to (t) give vigorous support to ongoin9 I
3.
The Comission intends to continue giving vfgorous support to ongoing es forts
$80ri tired at upcradin; the use of risk assessr.ent ratthoc:. logy in-aimed at ungrading and extending the use of risk assessment methodology the reFrittery prccess and (b) deter T.ine the Ep;ropriate courses in the regulatory proc iere such use is cicarly justified by such factor of actic. for (etiing :ith the issues raised by the Review GrorP.
as data availability and existence of appropriate models. 'Ac'cordingly, the
~
This ove all Ef fert *:ill include, 2 ong other thines, undertakin9 Commission has instructed the NRC staf f to ideEfy in specific tenns those a r! vies c.f 5:tti!Iicti. ethcds and h:.an f actor considerations m
areas where risk assessment methodology analysis is currently being employed Us e; in risk ass ess er.t, as,: ell as C;:,estiens stout earth utkes' and to suggest where, in what manner, and at what cost further extensions
,irts, c r r cause f ailures, and ditt base improve.Ents. The o fthis approach may be warranted. This review is to be comprehensive, with special (c :-issi:- hat ir.structed the staff te address t.+.at chtn;ts should attention to those activities identified by the Review Group as being especially te ;rc;cstc' ir the 2;;rcred TY 7c and ;roposed it E0 research amenable to risk assessment, i.e., dealing with generic safety issues,
- rc;-t-
- actie.e the tiove cbjectives.
formulating new regulatory requirements, assessing and re-validating existing regulatory requirements, evaluating new designs, and formulating reactor safety research and Inspection priorities.
Cm hffd Vf vf, A Tdj If CL
~ - ~ ~ - ~ ~ ~
rP Ga Pedev n d& &.
j 5.
The Comission will undertake to determine the significance of a number of technical issues raised by the Review Group and the appropriate courses of action for dealing with them. These issues include questions about statistical t
methods, data base quality and availability, human factor considerations,
(-
earthquakes, fires, and common cause failures. The Connission will address what changes should be proposed in the appruved FY 79 and proposed TY B0 research program to improve the data base, including that on human behavior.
As an additional action, the staf f has been directed to undertale a review of statistical methods and human f actor considerations used in rist assessment, 3
y s.
E
(
i D'JTT: II-i-7E
- 10 l
4.
The Consulssion will review its current' practices and procedures tri two areas
~
9MsTT 5
of particular concern to the Review Group:
Mhe Ccmission will review its..curyent.ptactices and procedures thehective and inadequate) peer review process accorded the RSS, and (a )
in two grees of concern to the Seview Group -- namely, the peer (b) the coordination among the research and probabilistic analysis staff review process end coordination among the research and probabil-and the 1Icensing and regulatory staff.
1stic enelysis staff and the licensing and reguletory staf f --)with The Comaission will make whatever changes are necessary to assure that ef fective a view toward identifying any needed improvements.
7 peer review and interoffice coordination are integral features of flRC's risk 3
I O
- i Npf ir[/rfy.3/
hpMt 'l Q v
p I'
In early 1979 the Conenission will prepare and make available a policy stat ment on n 1979 the Coar.!ssion will prepare and ratke available additione policy risk assessment which will go beyond this present one by (1) sunanarizing the ceidence which will supple.ent this present statement by (1) sur:narizing resultsoftheshort-termreviewsreferencedabove.(2)providing[ppreexplicit]
I the results of the short-ter:a reviews referenced atove. (2) providing
~. guidelines for the staff application of risk assessment methodology, and (3) any additiona pidelines found to be appropr,iate for_,the, staf f describing an Action Plan together with schedules for svstematically accomplishing application of risk assess:.ent methodology, end (3) describing en Action,
specific, longer-range tasks resulting from the reviews now bef ng undertaken, plan together with schedules f t,r systematically accooplishing specific, longer-range tesis resulting f ro:n the reviews now being undertaken.
The Conunission will keep the Congres ; and ti e p bl aware of its actions on a The Coaatssion vill keep it.e Congress and the public ewere of its continuing basis.
actions,on a cor.tinuing basis.
V s'