ML20135B006
| ML20135B006 | |
| Person / Time | |
|---|---|
| Issue date: | 08/19/1985 |
| From: | Cyr K NRC - ADVISORY COMMITTEE ON ENFORCEMENT POLICY |
| To: | NRC - ADVISORY COMMITTEE ON ENFORCEMENT POLICY |
| References | |
| NACEP, NUDOCS 8509100349 | |
| Download: ML20135B006 (89) | |
Text
,
,- ~Db(L A dviso, k rC. t/
Minutes of the Meeting of the {* "'ne Advisory Committee for Review of the Enforcement Policy w-July 25-26, 1985 @o4c Palo Alto, California 1
Attending: Dennis Wilkinson, Joseph Hendrie, Colin Diver, Howard Parris For the NRC Staff: Karen Cyr, Jane Axelrad The Comittee met from 8:30 a.m. - 5:00 p.m. July 25 and from 8:30 -
12:00 on July 26, 1985, in the offices of the Electric Power Research Institute in Palo Alto, California. Dr. Parris informed the group at the outset that Michael Hasten had called him the previous afternoon and indicated he would be unable to attend the meeting because of last minute business obligati~ons.
. The Comittee discussed the draft sections of the report provided by the members present in turn. (Seeattacheddocuments.) Organizational suggestions, alternate approaches and specific language changes were variously discussed. The Chairman offered to collate all the coments into a second draft of the report and kept notes on agreed upon changes.
~
The Comittee decided that the opening section of Mr. Diver's draft should go up front in the report because it provided a well-stated philosophy of enforcement. The Comittee also suggested that the section discussing the data analysis should emphasize the tentative nature of any conclusions because of the small sample size. The Comittee asked Mr. Diver to provide a description of the analytical e-
.g h f PDR
o methods used to be incorporated as an appendix to the report as an example of what could be done to analyze enforcement actions.
The section on management structure was found to be too unrelated to impacts on the enforcement program and was dropped from the draft report. A change in approach was also suggested for the section on inspector selection and training.
Because Mr. Hasten had been unable to attend or to provide a draft of the section on material false statements, the Committee decided that another meeting would be,,necessary to discuss that section and to develop a final set of recommendations. The next meeting was scheduled for September 23 and 24, 1985.
N l
Karen Cyr NRC Staff Liaison Attachments: 1 - 6 August 19, 1985 L. j
F TABULAR REVIEW 0F COMENTS
- q. -
.g- Effect of Enforcement Policy On: ,
g None or
. Positive Minimal Negative
- 1. Detection a correction of violation 3 11 5
- 2. Reporting of violations 6 7 4
- 3. Prevention of violations e 2 12 3 j
j 4. Safety record of licensee 1 14 -
- 5. Morale of licensee 1 3 14
- 6. Public confidence - 1 14
- 7. Worker's and public's - 6 5 willingness to report violations-l Particular Features of Policy
- 1. Civil penalty levels (A) In General Too high About right Too low 7 4 -
(B) Relative Levels About right Inconsistent (across Supplements) 11 -
- 2. Upward or downward adjustments About right Upward adj. too Downward adj. too Generous Limited Generous Limited (A) As written 5 1 8 (B) As Applied - - -
5 i
- 3. ~ Classification of violations by severity level -
(A) Number of levels -
Too many About right Too few 3 3- -
(B) Assignment of violations to About right Too lenient Too harsh ,
levels ,
1 - .6
- 4. Material false statements policy About right Too lenient Too harsh (A) In General 1 1 15 i ,
l (B) If too harsh, should be restricted to:
- Written statements (not oral) 5 i
- Willful and/or reckless misstatements 10
- Affirmative statements only (not omissions)
- Statements on which NRC did, or have 2 been expected to, rely
- 5. Relative emphasis on negative and About right Insufficient Use of Positive Incentive positive incentives (A) In General 2 17 (B) If Insufficient use of positive incentives - should use:
' Mitigation for past good performance 4 i
W" .
- Public commendation for good performance . 1
- Total mitigation or no citation for corrective action 10
- Reduced inspections 2 e
- 6. Progressive escalation of enfor' cement About right Too harsh or rapid escalation sanctions 8 6
- 7. Procedures (A) Enforcement Conferences About,right Problems t
. 2 5 j If problems:
- NRC not sufficiently communicative 3
- NRC not sufficiently responsive 2
- Too late in the process 1
^
(8) Publicity About right Problems (in general) - -
3 If problems: ,
- Inadequate notice to licensee 5
' Contents too one-sided 7 i
ae t
> Should nok be' issued at all -
- c f .
' Subsequent mitigation not well publicized 1
- 8. Supplement 8 of the Policy Satisfactory Unsatisfactory
. 1 4 Responses to Specific Questions
- 1. Impose sanctions on individuals? Never Willful & serious viol. only Licensed cperators only.
4 11 8
- 2. Effect of regionalization? None or minimal Positive effect Reduced consistency
.j 3 6 12
- ?
'i
- 3. Identification of root causes? Usually Sometimes Rarely t
6 8 5 4
i
- 4. Increased reliance on self- Yes No identification?
13 2 (A) Increased reliance on Yes Yes, INP0-like organ. only No third parties '
5 2 7 o
, Aff. A g O 1
} l l
1 l
advis ay .. , . He e an s.vsy-r .,yim
&c/.a a D
- e. nee r r w - o r d ,, /e 4 3 i % .>r nst ^ es/alh Q ' l no Ab d=* e e-m -s 7Ye e p.t m ee~> osc
.5 Aff's*fL W / ** /r c m .
os & ,,6f n Qey M s du. e e o-- 4 m e. e4 7"w em da c- -expa r' y 4y e-~ -g g an ef< ce . fk pp >p .226t. ~ '22Q
<s- 7L m sac 1 m r gae-2 g ., /, y t> dekm . w . s. % -2 s ed. -f hyv =& pjm .-Q f ace.
c= m <>s m w th m e ~ c 4 7 % -.- , cn cAnp .3 &c h-~ n L u 44 . W .
wng eD~ A .
v.1n n
@w n . , B, pn c - ,:i n, a .m
CL l
l l
l Of & W' njwlMe: > =& 'e A
r -d' pu h. w.0 p heD
@iesf M /S n-2 s st=< p op :cea_
& , AcL7 " W ~.-l h e m 4 3 yq , .zo -,
4-c s --g nD J-e m ui a ,,, , w g NLL AL kN' e ess Nn b ro u e. h of epo95%,
W
/e % -p syt sirir Q ct e D <
i-i m ,
- u.a.4 &,,,
/
f s p //. wny k .&
p Y.5
- l f
c.f v s . Dom l p(gc[4 k V. M+3M ESA.
']a. IdendNe u._ a . 3-s E P O Ih'as*J 4= t- - 4 y + d ete,e
- 7A e % w. r 1
. s e.-wna.am e ns-6.eme-w.-e.2,s .4 - .- - ~ - ,z _,,y-_-p m mam_,,_,,
,_,4 4
O aw w e t/~~y a-,z, :
l i- n n w e + .~-+ p i y -,, g is~~)f M war % mc-- kui.m+ x G l
@l. o && ==-9 m r 4 " " M T** *~ /*< Y' ~ j
&) da W, g-b~~<. viu M ~ 1 <-g g ) u ca.susepy &=,= +7 or &e-D op.
- a. 6x W ow pp, pan. M9 -
e f " spi ~7
7
,NFt. ans,3 .
$ . W Z.A-4. & Asf M'MMT **L-
& a ,4 ~ -/J k a = .;= % cmy ~.s.* 7<g A
fitLf4L M $ of *
, .3 . au. ;z3 rue ofJe n ?' m a m apy a.-yme-ez'-
ogim .A %, a.h %mem 4 ~m my J k 9aw=~.4 ?
A 4. D'As d. s. . s9 S S
- y 4/Q % C4 / J Y ' f' iIg -
l hn ( c!<~4 A oty t<-e ., 4 .1 7 m l-::% n*%
1
[w e l &s h $ *4 SS A.e r A l
[ / [O 4 da h ^j_
& n = .' f <r ss*s M & Wa h-e m 4C- :
l l
4b'~t, ,
1 1
- 'm* w- ,
S I
l -
g
- l *ci of Nh th Eb' &
AnDA m k //-ts,'t&</ , f 3co-3I,IS'tr, y t--a, n rs-, spa s o - rm 7, Str 3 If 1, ten ~
M 117 27 - z s- nts . 9_Jg:=,
A a-n - g -sc n p.c y ~ s m
+Ne L-@ (Evel.svg
3 ff - ,
%
- s t* *7 ""' P" '7 * "
^
N0 $ f% fkl $ fff v t ! fi < > , o% Fe Q c ,:_: m ,,y.< ,ca.,,
/wl,~s/t~y p n i e o. / ss(
- a A42 rect-> m f ,
^
= -
$d A w- Yr e<, ~?STO- o ( g <6C,a;
89 * * * - -Weem wgo a q-- ,, ,
n p <~ ==. r n s enses /2 1
f CPC 5 $%
N' W Y 74 y g de f
'N ;=k h - , Y n art y o y: % r1 S to'~)
- w. /ves % )- d A . % e.u K/r <
.7 h mi /h A 2 t.a
- W
~, -m c- s ,- - s.m
< ~ - uz -
,; _ j . gg , L : = J' M=
= " ,.,,, - -- - . _ 7 - - g & ps - ,-_ .
n ywg c~A wI% >p sy i 4 Deu 7;m L & p.~ 6/ > G.A Qu=> 7:~ .
<< 13 h111ay fm m .s z , yy /um o mea,y -sangg:g,,
4 o 1't <-t d - e 4-y o -n d n .4e e-k
.w-,.. ,=._,--.%.
e O
p g ( N (( w w g (
gyQ^ ;)x 1'A e > ~c A. J. . f m n *1-s% fmp n- 1, As yv n (,< ~ ap q s bbc , % sn d sr x
& owk } c;5= n e A faq m W
, - - > . n s, a ,,9 .
~
. s.g cmn, <
f"mm. /k /cNc t. b A-t- braak 7, k n,arr p ' M n 9 - r o= ck n--. L , t-p,- 4 n - <m , ..r-d, 6h ~.p sawA/ s yys s ~
$N Uf b h & 8 k $
6C he- -ge <wenwn.. -. _m,,. . . , ,
e e
4 skA k nAc,k0 pt c/i &. ~,vc/:,x ?
?-u -a ,9 = - . Q:w'n an 3
- .e.!- A f x :,,r r ~ ~ w y - Q-a.s u e) meLQ rubl~ i~ k ~ & y e h , /g s w ,f n /- - =.'?w s /.e,A P w A 2 ~ 7 }ik t 4
(Am 5- A y m n ,crte- & & V W k sm w~ NeJ tm?S .
so Ac < f,,7/m b n h t<--- e h~m L p iLQ t
9wan4 w < ha s~ s c4 , f,e .+
fiY c-~ g m ?Q % Aniy
&. yy . 4, a e - 5 , , /.,
w Met
- SO (&Ys r 5 ?- . . ,'_ , .* __ u : "-
3- - .-sce n m my,, x f l
-ep. dn Q k c~ /% 6
e ** e A=w Q
e 4
l ob/ 4 % Mm 21. ei as Am gm /Lt4 c k.t p f car <. W Y /5$ W .
h ^+t n b'L'* Q 4> . l'*
- n s,44,49 (e es.se < ) c~m
~ my _a ., e m M -c . Keviuo n h rz;'.,7' u 2 /c < &
n a a ., s - ~
n .zzte 4fi ve nym o ~f esp i w ~ ,p < ~ dw, L /.s . R espann d.s L k -w1 k) pe m.14& A L pp y m se-la n > 4 4.s b.,
c ,u . , , + u . <,
A m e + (,C- 16 1 &s f ys i.s . . '7 1 if
. .~.
3 =eh-miea m 'A,., = ge w e
e 4
h O d !
ILe st w A s fh cammi& .
bn> glaed e
t n s//x*'=4 ie sms . g,= p Pr
-A,
.a,<paitm r s <-mfan Y~en e 7' o~c( '%7-A eniA feel >- MW ~ffeat >3 n pm 7'~-, f un M ,
a
%~m
- .4a <4 mt Jewn p- 14 1
/m > c & 4 T-s, , <2 7 /s .5 < -p p5 &
F -f h d /Lc S(nfp s law /c/
u s'b, f1 ke + g="%1 n4% ng A o m ,s.4 4 ~ ./J 4
. ,9 +e 9 si s oj '7Ze . , ,r ,, 4-e q A 75 daucZ g p,n Q --.- f u e m erc4 Kg
% 6 up s :: t'- % u,c- H e
/~ / Y-ec / ,
e G
1
/d 0f A pu M-w a, J . > cn s > ,. e = = = - : - /.,
c~~P m ed r~ m t-w n =ca< <~ik pm ~4 s u s % .s > up .=4- ok s f~ uy cq ae( m ~ n ka~E 9 no v , .L.f.
.. A11. 3 M -
DRAFT 4. J.M. Hendrie
- 7-21-85
\
fV DDrifTntIDM his section several issues are discussed that may bemg~a ,
procedural in nature > These include the eneral' subject of timeliness of
, enforcement actions, the ature of enforcement conferences, the public information aspect of nf6rcement action and the divis drityJetween headquarters and regions.
A. TIMELINESS In most of the written comments received by the Committee, as well as in remarks of witnesses appearing before the Committee, the issue of timeliness of enforcement action was raised. It was cited as an essential feature of an effective enforcement program by all of those c[mimenting on the issue. It was also cited by almost all as an area in which the' enforcement program as carried out in the past has often been deficient. Enforcement actions involving civil penalties are an especially sore point.
W "
x fivil penalties -tad tYAY=pc:c?
! months after a violation has been v
identified and corrective action taken. If the violation for which the civil penalty is proposed involves Cc-ic !cr rey!e er Office of Investigations or Department of Justice involvement, the time can stretch out to several years.
From the licensee's standpoint, the proposed civil penalty and attendant negative publicity then come long after corrective action has been taken and the licensee feels he has moved on past the violation event. The civil penalty and its announcement seem to come as a gratuitous low blow, largely disconnected from the violation. The effect on plant staff attitude and morale is claimed to be palpable and negative. ,
- . - = - - - - - - - -
/
) %
y A lack of timeliness in enforcement action, especially for escalated hhw%
actions, seems likely to reduce the effectiveness of the action. )dille a tr+-
b
-pesed civil penaltyrever t.~7. impcsed long after a violation has occurg may set an example to be avoided for other licensees, this positive effect is at least wd.'i-^d.
partially offset by the ! rgely negative effects on the licensee}cing long after the initiating violation seem dis-Im ther, givil penalties connected from the violation, and almost as a random afterthought rather than a direct response to the violation.
The importance of timely enforcement action"is well recognized throughout the NRC staff. The Regional Administrators were quite strong on the need for timely action as an essential feature of an effective enforcement program. The staff also recognizes the negative aspects of the cases where escalated enforce-ment action has been substantially delayed. Recent efforts by the enforcement staff to improve the timeliness of escalated actions has resulted in some tighten-ing up. Civil penalties that -dc nct have to go to the C~~"!:sicn and do not hedgh involve the Office of Inspect.cns are now being announced about 10 to 12 weeks erw turny after identification.of the violation. These are typically Severity Level III adkka violations. Information on the timeliness of non-escalated actions, Severity Levels IV and V is not currently collected by Headquarters rtheregions[ tut are m nning; at a guess, betwcen 5 and 0 wc;ks frcm the identifying inspection to -
-; Notice af "iciation.
In April of this year, the staff published a new chapter 0400 of the Inspection and Enforcement Manual, dealing with the enforcement program. This new chapter explains how the regional offices and the Office of Inspection and :
1 Enforcement are to implement the General Statement of Policy and Procedure for NRC Enforcement Actions, 10 CFR Part 2, Appendix C. The avowed aim is"to achieve unifonnity and consistency of enforcement".
i s
The new IE manual chapter sets a fairly brisk pace for enforcement action.
It prescribes an enforcement conference, if one is to be held, before issuance of a Notice of Violation and within 4 weeks after completion of the inspection where 1 the violation was identified or after issuance of an investigation report (if the Office of Investigations is involved). Then, if an escalated enforcement action is . contemplated, the new manual chapter specifies that the action be sent to IE Headquarters for concurrance within 6 weeks of the last date of inspection or of the investigation report and that Headquarters should concur within 2 weeks.
The overall effect is an 8-week schedule f, rom identification of the violation to issuances of the Notice of Violation, with an enforcement conference having been held during that time. If the staff can achieve the 8-week schedule in escalated enforcement actions, that should alleviate most of the problems with untimely enforcement action for those actions that are not delayed by L ...ir ': r
-oe Office of Investigation activiti g et k % d N % d b uEAN N h .
Time 44 ness-problems will-r=in, hcwaver, ir. tne ceses that do go te 1.he j Commission a are subjects of investigations ission and the Office of Inves are w
~
tighten up their parts of the d edule. The C=ittea taae nn raaeen why that c=0t be done by h. tit'.ti.0a1 '
med For3 cases3 ir which the Co-!ss!On =t 50 actified, the-eew manual chapter specifies preparation of an Enforcement Notice which is sent to the Commissioners as soon as the IE Director has concurred in the enforcement action. Commissioners then have 3 days to object: if they do not, the Notice of Violation is issued d daftA -Ka igu ._0_ p o e M . h-ha. cM.aA. u.4aut.-%. au.b. ,es.M,,-Acm.ANR 3
- >- -m +,4niu 4. < ,,m i o ,,.+<~,.m u - - - . . _ - rse -m_ + ,4 e e 4 m,m re Y LA. O., -
e ow . a e n tnet res m tion or t u e 3 + cne eften te cs = te . :t = u:e 4 >00,OCO FNW .a. %
'I.
% M1rpth_ =i C% %n z u'*le cecat a peticy 1 vet! oce_tm. 6e_cLuu. d;.
- e hei-fe i=io g a prect__ cnf = = nt -
OMuu.
- uti- up h orQ;ad. .r. itniaacMo4 s = re_.
y=o3"oid x% bulttd% g.~ bra.M ANL t .
&(bsetifetse M b 4 k_o. so nr 3 A.a.Qt _e == tc eguad.csw.
)=u,mnu4rin4. ey . =3rity cate [ee-bic@wA ti= f dt
$ $ $ c h a A a A.' & b s. s eff$fg tu. Amis %qaqM A$12 0 mg s;~s un L 6 ,
M m5%. W . o n > sP a. p,h L u: a c~ a k n %l x av ' . . am m&o a%&a r n LA % A k ~ Ae.% -
wq ;: a cwm ~
. ,,,, . . . m. c . .',n ~ ,. ,< . _- -
hdhk Expediting the work of the Office of 3 aspect (^^s may be more difficult. y brehRQ q eruiele g it ehnsild he noggible to 35 Sign a high pricrity to the initial inquiru phase n the Office of Investigations is involved in a prospective enfo nt 1 action so tha here can at least be a timely determination as ether the matter requires an in th investigation. Those casa . at do not require full-l scale investigations can then orward i e normal manner and on schedule.
Even where a full-scale investiga i quired, it may be possible to arrange the priorities within th fice of Investigatio move escalated enforcement action cases alon t a brisker oace. As a way of raising visibility and improving ource management of such cases, the Committee suggests t where a ful nyestigation is to be done on an enforcement action, 01 prepare a tar ach;dule for the inve:tigaticn ir consultation with IE, and then stick to it. '
i Cases referred to the Department of Justice are generally the most difficult to deal with from a tineliness standpoint. There have been a number of these J-& m k k% w q<
cases in recent yearsg as alleged-material false nat.ement cases have begun to be-
-routinely-refecced to the-Department of Justice. (The numbers of such csses would- )
QA crihetantially redisced if changes now contomniated in the_defin{-tion-of_.matertgt--
false statements were made, a circumstance that raises some question: about the
' logic cf referring a!! present =aterial fal:0 statement casc: to Jusrtke)v- If an escalated enforcement action is held in abeyance pending completion of Department of Justice proceedings in the matter, the delay may well run to several years.
-Suee it is unlikely that the Commission can do muchsto obtain mere-timely hh 4ecmcc.Murde treatment of its affairs at the Department of Justice).; cther approach is % h ol 9p w; m b a & qV__ that anthemyeoada.eW.
-nccd j. The rnmmitthe holip COm.,,,351gg 3hggid gxggigg yggj gg797gily_
Mether it is necercary in all Oases--to-hold 05 e5Calai.cu enivicement Octica in-ebayance wh!!c the mills at D0J gcind their-inexcreble dy Ter-erd.
1
g l ad +o (. A- l
! i C.ide h w:k Q.Aa ab m M.A. b m .chmt
$4<.a W**A A aAu. % ama.
W e ,&ach %t A c I. u u k s & S d q d p &
L 4 m , h d p .. O J L E [4 Q. @ t ' % 6AD.&M
%;e.
C 4 - S w=vvhh LueRo.d. 6 9h. GI. N i q
'paq .% h k aa Q a. QA JAfsh. k. wa A& w aAQ !
a_ m % m % . % Gmekaa. mob [ D ........ A j h% % a d' hack Ss opt.A q M g
% p h $ m j..- Q m U i p o d . a 4 + Q ,
Q g,. . a . .n ea%. a n .abA6AQ QJLi A & L.0a. 9 ,g wundr kha Qd.. ia D Q,c.cau.u. kuedr% AdAA$&W . ~%m &at i pt
%w a , - c--Q& mQ h
Q 1.E 2.
MQod ,
%(AQca 4 n~. L . eA f 9 s audspa,% 4 Q dn&&e. b - 4%QA 4 ad@ed. . E , % Commaa. -ACh. ht a-d.C. a_ Ga.Q @ % a to ndewQ b@g m b %.
- Is *4, A & %&ubAh
% . d 4 @ i d u.> C & %e4 44 , g & %%..
% C a Ae.a h9.:t ht a. M % aJa saa % % m.
't
, . , , .._,.y .-.
. - . . - . ~ . - . - .
, Recommendations: Timeliness and Enforcement Actions
- 1. The staff should meet the 8-week schedule in the new manual chapter
-for the period from the last inspection on an alleged violation to issuance of the Notice of Violation. .
2 dd2 0n should.'adc#t a pelicy O.
& Mk *%
res01Ying' .=>551cnEi -
sN Y .q&
-cbjectic..The-CO : w to an enforccmcat %wOkQ action by veting >W Ny the actl om W wit iru3 wccks and-wacfdsting ho.a.Rozhh to% Cow Mow.
the majority vote as sett!!n; the matter.
^
- 0. The OfiRe vi Investigations should 5551gn high priority tc in!*-
linquiries on n rcement cases referred to it, in order rmine expeditiously
+
whether a full investiga ort is needed. We s
a full investigation is needed, the Office of Investigation shoul prepate a target schedule for the investigation in consultation with en [tcemen staff and should'walte every attempt to meet it. -
- 4. The N N
should consider the possibility of going fo'rward with an .
escalat enforcement action in a case referred to the Department of Justice Ar6 fore D0J complates its dispositica of the case.
B. ENFORCEMEfR CONFERENCE The enforcement conference is one feature of the enforcement program for which almost everyone had a good word. The regulators all felt that it was one of the more effective and efficient parts of the enforcement program. The' regulated had some bad experiences with enforcement conferences to cite but nevertheless thought it was one of the better elements of the program and thought it could be even better. Clearly, the enforcement conference can be an occasion for effective exchange of information and points of view on the details of the o
particular occurpnce, the violation that might arise from it, the root causes for the event -and short-term and long-term corrective actions that are being taken or should be taken, d daA '
b M.
. .. . we -= **
e b
MA +o ). G
- s. i. aa& ;Aaa , g 4 4.... A g @h 9 y- % ,_nawa. m a m s. ~ k to h c#i u. ehag q n-y.A.
b wMP% mW&N
-pq ce s ua:p vs:apa.
- 4. % . e .._ # , 4 u t w ep u 4 hs , ess , %* & usak pea g 'J&g % k te naiaut. b A L dip,A A d L.x JtQD .
- s. 4 h & a b et 4.4 4 L @ Jdg 6JL h,uAJcA. et a & 4o Abt m aud p d u.4a e tu.% & 6a [A.ug6mmm.Abdl \
& fd@sce uku~ak ko &A %>&
g g T E , L W 1. m 6 4 7 h % ;
m u ot .
.. _ _ _ _ _ . = _
Practice varies between regions as to whether or not the inspection report covering the prospective violation is completed and made available to the licensee before the enforcement conference. Reasonable arguments can be made for either practice. Holding the inspection report open allows the incorporation of any pertinent information that may develop at the conference and allows the possibility of correcting mistaken impressions or information in the draft inspection report. Completing the inspection report in advance of the enforcement conference makes it possible for the licensee to see much more clearly the particular areas of difficulty that are going to be discussed.yrevidiag, cf 4ourse, that the lieeacee get c ccpy Of- the ir.specticr. repcrt in time te "tud-i fore the conference. Having the inspection report complete and file efore the c rence is seen by the staff as helping to avoid any impress that the licensee h een able to exert undue influence in an enforcente t conference and x
cause the staff ton everse itself.
\
The timing of the enforcement o conference has the past been influenced by whether the pertinent inspection eport was be completed before or after the conference. Now, however, the manual pter on the enforcement program requires that an enforcement conference be held More issuance of a Notice of Violation and within 4 we efter completion of the spections pertaining to the prospective vio ion (or after issuance of an investiga on report) The manual does npt specify when the pertinent inspection report shoul 'be completed, leaving at to the regions to determine. Either way, the 4-week requirDen app es and if observed in practice should assure timeliness of enforcement
.cr.fe. eso .
The new manual chapter also requires that the licensee be told that the meeting is an mforcenant conference. That improves upon some past practices where in at least one case tne licensee was told several weeks after the meeting i
l 1
--=a,..--,-.,---h O
k
'k h . G 3
W W u LL p~+ % Em W uka 4,;~s,
& ~ ~ ~ w c. A u .: n a,~.i.. aa w__
Mt% gg,q g 34pj &,y; YS%
a ~
O '0"O
~ -% see-n. _. L g x 4 g Sti% d eur.A w..a n M. gg, A L. g _
S *4 & G k Q -%a am g W Au 4.
a a
% dog Mub %.44 h ggg;%
- *
- 4 +, e ., . * + . , . t.
w wraa Q, .
1 l
l
. _ ~ _ _ _ . _
that he had oeen in an enforcement conference and a Notice of Violation was being issued.
The basic, and unsettled question about the e..forcement conference is what kind of a meeting it ought to be. At present, conferences appear to vary from gcnuine and serious exchanges of information and points of view, through meetings in which the licensee is allowed to present any explanations he may have and outline his corrective programs but in which the staff volunteers essentially no information or point of view, and to occasions when the conference seems only to be an occasion for lecturing the licensee on his sins and info' rming him that a fine is forthcoming. The latter sort of enforcement conference may well be appropriate if it is the second or third time the same subject is being discussed and there has been no improvement to date. But the Committee believes that for most enforcement conferences, a serious exchange of information and points of view would be more productive and more effective %s an enforcement mechanism.
Such a format for the conference requires that reasonable notice be given the licensee so that he can prepare for it. The Committee regards 3 working days as a minimum, and more time will improve the quality of the discussions.
The Cem=!ttee makes nc judgcmcat cr. the relative merit: Of completing and
- !!!ng the inspcctica repcii. beier e er efter the enforcement-conference.
Recommendations:-Enforcement Conference
- 1. Enforcement conferences should generally be serious exchanges between OL staff and licensee of information and points of view about the occurAnce giving rise to the conference. The staff should be prepared to outline its understanding ev of the occurgnce and its position on possible violations. The licensee should be re.
prepared to provide as accurate an account of the occurgnce as he can, the resultsofhisinvestigationoftherootcausesoftheoccurknce,andthecorrect-
-\h d htn. ..y0#:r. Q .
iveactionsthtph: h: :nd will undertchc_
- 2. The licensee should be given as much notice as practicable, with a minimum of 3 working days, that a meeting to which he ,has been called is an L_ L tu GT r a J. ACumJA anddo.
enforcement conferencegefwl the notice should cgtain f c =Q,:cient dc c;-ipticn 8.vN.~ .f_1,____ __ _ _ .mm
~
y
.s,+
..m.
...u.m- . . sc m wu w o. . v. mro. c i vi uic uvii c- c..- ,
C. PRE 55 RELEASES The practice to date has been for the NRC to issue a public announcement of proposed civil penalties. The policy was reconsidered in 1983 by the staff and it was concluded that the practice should be continued. Since the pertinent final documents in an escalated enforcement ac. tion are placed 1,n the Public Document Room, the information becomes public when the Notice of Violation and proposed civil penalty is sent to the licensee whether there is a press announce-ment or not. The press release reaches a substantially broader media audience than is likely to be attendant in the Public Document Room, however, and therein lies the concern of licensees who feel that the' press coverage of a proposed be% -h-&t.
civil penalty 1sfgs damaging to them in a variety of ways apeany dollar amount.
t.h:t might bc !cvied.
The NRC's public information staff believes that it is important to announce proposed civil penalties because they represent the judgement of the staff that there may be a serious breach of the license or of the regulations.
AuA The staff view is that th;'3 :;1 information that should be volunteered to the public and to the Congress at the time the action is proposed. The staff cites the high interest of persons who live near the plants in the operation of the facilities and feels that the staff has an obligation to keep the public informed of significant developments at the time they occur.
The licensee view is that the bad press associated with escalated enforce-ment actions undermines public confidence in the licensees. The newspaper j accounts read by the public with regard to enforcement actions deal almost '
, 9
- exclusively with the " wrongdoing" that resulted in the violation and any sub-sequent positive development rarely comes to the public's attention. A sin;;1e enforcement action can result in negative pressgaccounts two or three times, since the initiating event and the final payment of the penalty may be covered by the press in addition to coverage of the Notice of Violation and announcement of 4
proposed civil penalty. The public may thus be left with the impression that the licensee has been fined more than once. At least one licensee finds the negative press coverage so onorous that they decline to argue for mitigation of any proposed penalty, judging that the potenti,al negative pr5ss coverage of their reply and NRC's final decision would outweigh any possible benefits of partial
- mitigation. They feel that even if complete mitigation were achieved it is unlikely that the public would connect the facts and that it would be difficult, if not impossible, to erase the negative impressions formed months earlier.
It is not likely that any easy resolutfort of these conflicting views is possible. The perceived high " costs" of press announcements by the NRC of proposed civil penalties is part of thg'deterrance factor, just as the civil aA. netd W 6 Aa.e% .
penalties themselves are3 (On the other hand, if the press announcement policy for enforcement actions leads to an unjustified loss of public confidence in the safe operation of nuclear facilities, then it is not serving either the immediate objectives of the enforcement policy nor the broader objectives of the NRC.
It is interesting in this connection to note the language of Order 1000.9D of the Administrator of the Federal Aviation Adninistration (October,1982):
" Publicity. Individual enforcement cases may be publicized only when such publicity will significantly serve the goals of the enforcement program and only in"a manner that will reasonably protect an alleged violator's right to due process and fair treatment." It is hard to argue that the FAA does not regulate matters 1
important to public safety or that the traveling public is not vitally interested in the safety of the aircraft they ride.
The NRC pr:ss releasa practics with regard to escalated enforcement
. actions is a long-standing and deep-seated one and it is clear the agency would have considerable internal struggle in giving it up. Also, since on the licensee's own testimony the public announcement of a proposed civil penalty carries as much or more weight as the dollar amount that may be levied, it is not clear that the agency ought to be asked to forego such an effective enforce-ment weapon. Thus, although a fair case might be made for urging adoption of the FAA publicity policy, the Committee limits its recommendations to some minor modifications in the mechanics of the present practice.
The new manual chapter on the enforcement program prescribes that after an enforcement action has been signed, and the Regional Public Affairs Officer has verified that the subject licensee has been notified and has received a copy of the Notice of Violation, the press release is issued. Licensees continue to report being taken by surprise with announcements of escalated enforcement actions, however. Allowing-one working day's grace between verification that the licensee has received the Notice of Violation and issuance of the press release l would allow some time for the licensee's internal communications to function and for the licensee to prepare to answer press questions about the matter.
A second point has to do with the information in the press release itself.
If the NRC press release is going to command wider media attention and is more likely to be the basis for press accounts than releases from licensees, as almost certainly seems to be the case, the NRC press releases ought, in the interests of fairness and completeness of information to the public, to be reasonably balanced. A press release on an escalated enforcement action will obviously carry that action and the causes as the lead items. But if the licensee has taken corrective action and it is complete or underway, that information ought to be made clear and explicit in the press release. Further, if the licensee has
been operating in reasonably good shape and conformance with regulations for some time, and is again in conformance, then those facts ought also to be noted in the press release in order that the public is not misled by reading solely about the particular violation and proposed civil penalty.
Recommendations:JPress Releases .
- 1. The enforcement program mans.1 chapter section on press releases should be amended to provide for a one working day grace period between verification that the subject licensee has received a copy of the Notice of Violation and issuance of the press release on the propcsed enforcement action'. ,
- 2. Press releases on enforcement actions should include such information as the initiation and progress of licensee corrective programs for the violation in question and plant operating status and recent operating history in order to ,
present a more complete and balanced account to the public.
~
D. REGIONALIZATION There has been some question about the effect of regionalization on the NRC enforcement policy. The obvious concern is that in strengthening the five regional offices of the NRC, appointing their heads Regional Administrators, and granting a certain degree of autonomy to the regional' offices, differences in enforcement programs and practice,s may develop. The better view seems to be that a.mp regionalization has not had : ry g:::h effect on the enforcement program and its implementation y The processing of both routine and escalated enforcement actions is much the same now as before the move toward greater autonomy for the regions.
The only significant change has been the delegation of authority to Regional Administrators to sign escalated enforcement actions which have been reviewed and concurred in by IE Headquarters. Previously all such actions were signed by the IE Director.
- s. .
w
+Mw. e.J m u ad< m a . >
x .
~
/
6
.__.____________..___.___m_____ _ _ _ . _ _ _ _ . _ - _ - _
It is generally felt on both the staff and licensee sides that nation-Y' wide consistency in enforcement is a desirable goal.j The escalated enforcement
, actions, at Severity Levels I,II, and III, are all processed through IE Head-i .
l quarters so that a reasonable degree of consistency is enforced. However, the ,
individual regions handle all aspects of Severity Level IV and V violations and IE Headquarters has not been involved in the past. The Committee notes with approval that a headquarters program of sampling Severity Level IV and V
( violations from each of the regions is being instituted. This should provide t
some basis for adjustment of practice between'the regions to achieve greater consistency for these lower level violations.
. An additional aspect that needs consideration is that corrective actions I
{ proposed by licensees and accepted by regional staffs are not examined to see i
that the treatment and results are consistent across the country. Since it is not at all clear how one might go about examining these matters for consistency, the Committee refrains from making any recommendations in this area.
f .
( M b e A ^ *d M
e to 44 C N 4e '
4 Le. 4a. L m.u% s 'dr g&%. '4eu9. QC ra n & %. Q a. An %
4 g #4 Du .
e e am j m -
I
Diver. Draft 7/16/35 A tt~ 't E*
V.
Penalties and Sanctions b.a.
A. Philosophy The imposition of sanctions can be used to serve any one of three purposes: retribution, incapacitation, and deterrence. Of these, the second and third are relevant to the context of nuclear regulation. Retribution is appro-priate--if at all-- only in the context of crimes against the moral order. Most regulatory violations involve conduct which is prohibited not because it is inherent.ly immoral or evil (malum in se), but because its value is outweighed by the harm it might cause in the particular context.
- 1. Incapacitation has a legitimate role in nuclear regulation, but a limited one. Where a regulated entity shows itself to be fundamentally and chronically incapable of carry-ing out its responsibilities, it should be disqualified by suspension or revocation of its license. One would expect this sanction to be used only very selectively in the nucicar context, however, for two reasons. First, the degree of incapacity necessitating license removal occurs very rarely.
The entities regulated by NRC are typically responsible and com-petent. The corporate entities regulated by NRC--electric utilities, laboratories, hospitals, etc.--have longstanding ties to their communities, employ professional staffs, and guard their reputation. In order to succeed in their lines of business, they must display a generally high level of compe-tence that presumably carries over to their conduct of
_2-
,' activities regulated by NRC. Likewise, the individuals o-licensed by NRC typically have extensive education, train-ing, and responsibility. The licensing process itself, moreover, should screen out those corporations and individ-uals who lack the basic qualifications to conduct regulated .
activities in accordance with NRC requirements. Second, removal of a license can impose a cost on the public by reducing the sources of supply of an essential service. Conse-quently, this sanction should be reserved for , extreme cases.
- 2. Deterre.nce. Sanctions ca'n serve a deterrent function by increasing the perceived cost (as viewed from the perspec-tive of the regulated entity) of engaging in prohibited con-duct. It is the threat of the sanction, not its actual imposition, that achieves the deterrence. The purpose of actually imposing.a sanction when a violation has occurred is to maintain the credibility of that threat.
- a. The Importance of Deterrence The relative importance of the deterrence function of regulatory enforcement depends on two things: the strength of .
existing (extra-regulatory) incentives for compliant behavior and the social consequences of (otherwise undeterred) violations.
In the nuclear area, these two factors often point in opposite l
directions. The existing incentives for compliance are, in j
- o. 1 most contexts, quite numerous and strong. Violating an NRC .
requirement presumably increases the risk of an incident or accident. The consequence of such an incident or accident to the licensee may include injury to its own personnel, damage l
l l
l
,4 W. ,
to its own property, lose of cepacity utilization, liability
.for injury or damage to others, and adverse impact on repu-tation in the financial, labor, service, and political marketplaces. Compared to those existing potential costs, regulatory sanctions are viewed by many observers as having
~
little deterrent impact. While that is perhaps true as a general matter, existing incentives are not equally effective in all contexts. Insurance, for example, may dilute the fear
~ of liability. So may statutory ceilings on liability (as under the Price-Anderson Act). The ability of pome electric utilities to pass increased fuel costs directly through to the consumer may dilute the deterrent-value of a fear of shut-downs. Problems in communication, supervision, and personnel development within large organizations can also weaken the translation of corporate incentives into employee behavior.
A sensible regulahory enforcement program should therefore focus its energies on those areas of regulatsd activity where' non-regulatory incentives are weakest.
Effective regulatory deterrence is also especially important where the harmful consequences of error are greatest.
The less serious the violation, the more uncertainty one can tolerate concerning the effectiveness of nonregulatory .
incentives. Where harmful consequences are catastrophic, i however, those uncertainties should be resolved in favor of e using regulatory penalties as a major supplement to other incentives for compliant conduct.
_4_
\
- b. Implications of the Deterrence Perspective Designing an enforcement policy to achieve effective l deterrence requires the regulator to keep several points in mind.
- 1) Uncontrollabic Conduct. No deterrent function can .
be served by punishing someone for an action or outcome wholly beyond his control. If an agency is perceived as punishing people for freak events, its enforcement program will lose credibility. One must concentrate on preventable occurrences.
Punishing a corporation for its emp"oyees'l behavfbr mus t be i predicated on a plausible belief that the corporation could have controlled its employees' behavior.
- 2) Perception and Reality. What matters for deterrence purposes is the regulated population,'s perception of what the regulator will do ip the event of a violation. A regulator must therefore strive to find out what that perception is, and if appropriate, to change it. Merely changing the reality of enforcement is not enough (and may not even be necessary) to change perception. Communication is also necessary. , j l
- 3) Sanctions. Costs, and Penalties. To deter a particular i
form of behavior, it is necessary to raise its perceived cost to the actor. The penalty imposed is only one part of the en t.
Administrative and litigation costs, damage to reputation, l
exposure to closer future surveillance and increased paperwork (
are additional forms of cost that regulators can and do impose l on violations. In setting the level of penalties, it is important to. keep these other costs in mind. Often they are large enough }y themselves to achieve adequate deterrence.
t V
- e
- , . . . _ . .-_= -.. . .. -- - . . - - . = _ . _ -
~
S
.,- 4) Probability of Detection. The expected cost of i1. -
committing a violation is also a function of.the probability of getting caught. In setting.the penalty level, one must consider whether the conduct involved is considered easy or ;
~
hard'to. conceal from the regulator.
1
- 5) Positive and Negative Incentives.
~
In principle deterrence can be achieved with either positive or negative enducements.
- That is, if one wishes someone not to do X, one can either pay
} him not to do it or punish him for d.oing it. Either way, you eliminate the advantage in doing X. Like the punishment, the -
payment can take many forms (money, serviccs, reduced-administrative i
1' or record-keeping burdens, praise, etc.). Most regulatory agencies
'i j are, however, severely constrained in their use of positive in-centives.- Direct subsidies require explicit statutory authorisation ,
, and appropriations.F' Reduced administrative burdens-typically'mean reduced. regulatory oversight and weakened enforcement. And praise is both suspect and politically dangerous.
i
- B. Overall Evaluation of the NRC Policy 1
- 1. Performance '
The NRCs present enforcement policy has been in effect, subject to minor modifications, since 1980. Has it had a deterrent
, effect?. There is very little solid evidence on that score. What , ,
' evidence we have takes two'formst, statintical.andLimpressionistic. ,
$ Statistics on enforcement are - difficult to assemble for various.
reasons. For one thing, there is no universally accepted measure, a i of performance. The agency gathers various statistics that could be,used as performance measures (e.g. SALP ratings, LERs, shutdowns, !
4 i .
1 4
l *
, downtime violations by severity. level). But they are not gathered centrally nor are they systematically related to enforce-ment effort. Each of the possible performance indicators has .
obvious theoretical problems, and the underlying statistics are frequently incomplete or unreliable. Given its limited resources,'
the Committee could not significant1y' improve upon that state of affairs. It did ask the staff to prepare data on a sample of eleven generating plants chosen to be broadly representative of regions, vintages, and performance levels The" eleven included Arkansas, Browns Ferry, Brunswick, Calvert Cliffs, Kewaunee, Palisades, Pilgrim, Quad Cities, Rancho Seco, St. Lucie, and Surry. For each plant and generating unit data were gathered showing for each year since 1975 the number of inspection hours, violations (by severity level), civil penalties imposed (and dollar amount), orders, LERs, forced and scheduled outages (and hours), and SALP ratings. An analysis of this, data rceealed the following general conclusions:
1 I
l I
i l
l l
1 l
r- _ __ __. _ . _ _
~7-
, -Impressionistic information is not much more helpful in assessing'the performance of the enforcement policy. Most commenters have strong opinions about various aspects of the
. policy. But1few offered any solid evidence concerning its impact.
'Several utility witnesses claimed that they could not identify any particular practice or system that was adopted or modified in response to the policy's adoption in 1980. Many did claim that its excessively." punitive" quality did have an adverse effect on morale. But almost no one was willing ,to ascribe particular. actions (such as employee resignations) to that morale problem. Nor could any witness convincingly separate out the effect of the enforcement policy from the generally heightened visibility and notoriety of the nuclear industry.
On the 'other hand one regiona1 ' administrator stated emphatically that u'tilities go out of their way to avoid civil
~
penalties or orders. The utilities' oft-repeated concern about-the adverse publicity accompanying escalated-enforcement actions lends credence to that assertion. Others conceded that escalated actions do succeed in "getting the attention of management."
The overall impression, then..seems to be that the policy's impact has been modest in magnitude and psychological in nature.
It has made responsible personnel in the industry somewhat more ,
cautious, circumspect, and nervous. . Whether that translates .
into reduced risk of accident is an.open question. Caution can breed care, but irritation can breed error.
2.- Design Lacking solid evidence on the actual impact of the enforce-
ment policy, the Committee has attempted to evaluate its design by applying the principles outlined in the preceding " Philosophy" section. Measured by those criteria, the Committee feels that the policy measures up vell'. The policy claims to base the severity of the sanction on the gravity of the violation. It ,
is very difficult to confirm this claim, since the gravity of a violation cepends on n host of factors peculiar to the parti-cular context that only a trained inspector or engineer can fully assess. But the assignment of violations to severity levels seems on its face to be faithful to'the gravity principle. So does the establishment of relative penalty levels for violations in different_ Supplements, although the margin for error here seems extremely large. Similarly, one senses in reading summaries of recent penalty cases that the size of the penalty generally reflects the, seriousness of the health risk created.
To the extent that this is an accurate reading of the NRC's practice, it sends the correct message to industry about relative costs: subjecting the public to larger risks will cost you more than subjecting it to smaller risks.
I Whether the message is correct with regard to absolute costs is harder to evaluate. Are penalties high enough to provide the incremental deterrence needed? One cannot answer that question I
by looking at the design of the policy. One must look to extrinsic factors such as the strength of nonregulatory incentives and the
^
magnitude of compliance costs.
l The restriction of civil penalties to only the most serious
{
i e D a
W---- -
, violations (Level Is, IIs, and some IIIs) does seem appropriate as a theoretical matter. Level IV and V violations together comprise over 90 percent of all cited violations (an even larger percentage of all detected violations since many are not formally cited).. The. administrative burden (on both NRC and industry) of imposing penalties on these violations would probably outweigh any resulting increment in deterrent impact. The administrative costs entailed in responding to those citations are themselves a form of " sanction" that holds these, violations in check. Also, one must suppose that many level IV and V violations involve no measurable safety hazard. So long as they are corrected, further punishment would be gratuitous.
Secondly, the factors used by the policy to justify escalating or mitigating base penalties seem aphropriate to a sound deterrence strategy. "PreviouI' violation history" is a good indicator of the strength or weaknesses of nonregulatory.fncentives. Where the prior. history is poor, the incentives are weak, and a larger penalty seems needed. For the same reason, " prior notice of the same event" warrants escalation. The " multiple occurrences" factor justifies escalation because it indicates both a higher gravity (the more occurrences, the higher the risk) and a greater propensity to commit violations (weaker nonregulatory incentives). " Prompt ,
identification and reporting" and " unusually prompt corrective ,,
I action" justify mitigation because they signify ef fective self-regulation (strong nonregulatory incentives). Finally, increasing penalties (or refusing to mitigate) for willful or repeated violations j l
makes obvious sense.
-~
I _._
__ i
., While the Committee generally endorses the design of the NRC's enforcement policy, a number of questions have been raised and criticisms made that warrant fuller discussion. The following section addresses those concerns.
C.. Specific Issues
- 1. Positive Incentives. The policy has been extensively criticized as excessively " punitive" and " negative." The Committee is sympathetic with this concern, but doubts that much can be done about it. Truly " positive" incentives like subsidies are unavailable to the NRC. The NRC could bestow public praise on good performances, but there is an obvious danger of its being misinterpreted or distorted. More realistic are suggestions for reduction or mitigation of regulatory burdens. Elsewhere in this report we have endorsed the notion that frequehey of inspection should be based on a licen'see's performance. Since inspection is a form of regulatory burden, reduction in its frequency should be per-ceived as a " positive" incentive for good performance. Mitigation of penalties is another form of reduced burden. It is discussed below.
- 2. Number of Severity Levels. Some commenters propose that there should be fewer severity levels than the existing f,1ve.
Some feel that there is no reason to distinguish between Levels IV ,
and V (since both are treated alike: no penalty). Others com- ,,
bine this proposal with a proposal to eliminate all penalties for '
Level III violations (making it possible to consolidate IIIs, IVs and perhaps Vs).
t
, The Committee does not support these proposals. In its view, the existing structure is ideal. There are enough categories to enable the staff to avoid the need to lump together violations posing wildly divergent levels of risk. Yet the number is small enough to be manageable. Moreover, each of the five levels serves' a distinctive function in the structure of the enforcement program.
(Vs are never treated as escalated enforcement actions; IVs are rarely so treated unless they involve repeated violations or a pattern of similar violations; IIIs are always escalated,but not always penalized; IIs are always subject to a base penalty, but may be mitigated; Is may never be mitigated.) .
- 3. Classification of Violations by Level. Some commenters claim that the policy provides inadequate guidelines on classifying violations by severity level. Our attention was particularly drawn to the fuzzy criter;ia'for making the crucial distinction between a III and a IV. The Committee agrees that many of the published criteria are indeed vague (see, e.g., Supplement I's distinction between a "significant violation" (III) and a "less significant violation" (IV). We urge the NRC to accelerate efforts underway to give greater content to these distinctions, through written guidance, training, and selective headquarters review of regional practices involving Level IV and V violations. But we recognize that given the enormous range of regulated activities and contextual factors, gravity rating of violations necessarily involves a high degree of professional judgment. Bright-line guidelines will never be a substitute for careful selection, training, rotation, and supervision of regional inspectors.
f - --- - . . - - _ .
t 4.
Determination of Relative Base Penalties for the Various Supplements. The Supplements have been criticized for attempting to equate the incommensurable. There is no way, so the argument runs, to correlate, say, reactor operation viclations, with transportation violations. Table 1A of the policy, by assigning relative base penalties, attempts to do just that.
The Committee agrees that ranking violations by relative severity across activities is difficult. But so is ranking violations within activities. It mdy be difficult, but it is not impossible. Indeed it is impossible to avoid making such comparisons implicitly (by the relative penalties in fact assessed).
We commend the NRC for making these comparisons somewhat more explicit than many regulators do. Whether the precise ranking systen adopted by NRC is sensible is a much more difficult question. ,
Several witnesses contended that there are no conceivable violations in, say, the health, physics or transportation area that could approach.in gravity the Level I violations in reactor operations or construction. Consequently, they argue, there should be no level I (or even Level II) violations in the former categories.
The flaw in this argument is that it disregards probabili-ties. It may be true that a worst-case reactor disaster would dwarf in magnitude a worse-case transportation disaster. But the probabi1-ity that such disasters might result from any particular violationI may differ considerably. Since reactor requirements involve multiple redundant-systems, the actual risk that a particular failure will cause an accident is extremely low in most cases. It may, therefore,
make perfectly good sense to equate (in terms of severity). the actual exposure of a person to radiation exceeding 0.5 rems (See Supplemental V, Level I) with, say, a exceeding a reactor Safety Limit (Supplement I, Level I). Whether this is so in fact undoubtedly depends on the circumstances. -
f The relative penalty levels appropriate for violations of different Supplements may also depend on other factors such as the relative strengths of nonregulatory incentives and the relative probabilities that violations will be detected. If, for example, transportation violations were much'less likely to be detected than reactor violations, higher penalties would.be needed in the former area to achieve comparable deterrence.
- 5. Removal of Economic Benefit. In an earlier version of I
the policy, the NRC proclaimed that " noncompliance should be {
more expressive than. compliance." This has been softened to:
" Sanctions should be designed to ensure that a, licensee does not deliberately profit from violations of NRC requirements." The Committee subscribes to the philosophy of the original statement.
That is, effective deterrence requires that the regulated popu-lation believe that noncompliance will cost more than compliance.
This philosophy applies to any preventable violations, not just
" deliberate" violations. The Committec concludes that, while few i
violations are " deliberate," most viola' t ions are " preventable" at least in the sense that their likelihood could have been redttced by the exercise of more effective supervision, auditing, or j 1
personnel development. By failing to install such a procedure, j
l the violator has in effect " profited" by his wrong, and the purpose )
4 i
l i
-1
)
a
~ ~
I of enforcement should be to eliminate that " profit". To the
_ extent that the change in language in the 1985 policy suggests otherwise, the Committee regards it as unfortunate.
On the other hand, the Committee recognizes that actually measuring the " profit" in particular cases would often be .
extremeIly difficult. It is feasible where the violation consists of a failure to install an identified piece of equipment or system.
The Environmental Protection Agency is able to use such a system because each pollution source must obtain a permit that specifies the pollution control equipment of systems raquired. The
" benefit" approach is less workable, however, in dealing with one-shot violation of operating procedures. An attempt to estimate the " profit" in such cases would usually provoke controversy over what alternative procedures might have been used and how effective they would-have b' e e n . A policy of computing the " profit" in all cases would therefore dramatically increase the administrative costs of enforcement and reduce its predictability. Consequently, the Committee endorses the practice of explicitly computing " profit" only in clear cases.
- 6. Ability to Pay. Lurking behind some criticism of the relative penalties assigned to different activities is a concern that penalties more explicitly reflect relative abilities to pay.
A penalty of $40,000 or even $100,000 is small change t'o a large i The utility, but may be devastating to a. materials licensee. {
implication of such comments seems to be either that reactor j penalties should be raised or nonreactor penalties lowered.
But this does not necessarily follow. As indicated above, l
4
- i
.. . -_u u 4
. penalty levels should reflect other factors such as relative gravity of the violation and probabilities of detection. More-over, penalties are only one component of noncompliance cost.
The combined effect on a utility of a penalty, administrative costs, and adverse publicity can be very substantial. Also, since penalties may not be included in a regulated utility's cost of service, it is more likely to get the attention of management than a comparable cost item. Finally, to the extent that the nominal ceiling on penalties ($100,000) is ,t o o " low" for large utilities, the NRC can (and sometimes does) use its "per diem" authority to generate much larger actual penalties for especially serious violations. The Committee urges the.NRC to use this authority liberally to avoid the confinement of an artificial ceiling.
- 7. Nonpenalization of Level IIT Violations. Some commenters recommend that the NRC move further in the direction taken in 1984 by never issuing civil penalties for Level III violations. The.
Committee does not agree. In 1983 only seven violations were classified as Level I or Level II, representing only about one-half of one percent of all cited violations. In 1984 and 1985, the number has been even smaller. We question whether a system that issues so few penalties can maintain sufficient visibility and credibility. At least some Level III violations strike us as 3 serious enough to warrant the imposition of at least a symbolic ..
penalty. The 1984 policy has succeeded in significantly reducing the volume of sma31 penalties (Level III penalties dropped from A8 in 1983 to 23 in 1984). A further reduction seems counterproductive.
E
,. 8. Complete Mitigation for Self-Identified and-Corrected Violations. Related to the previous proposal is a proposal from several utilities that civil fines never be imposed for a violation identified and promptly corrected by the licensee. Existing policy I ,
allows 'a maximum 50 percent mitigation for each of these two factors.
(This policy does not apply in the case of Level I violations where no mitigation is permitted. We will discuss the wisdom of this 1
, ~1 imitation in the next section.) *
{ The Committee prefers.the existiat polict to the proposed change. Self-identification and correction are merely two factors among several that should go into the determination of a penalty.
They should not, for example, totally _ excuse a poor prior history, j prior notice of a similar event, or wi11 fulness. A universal policy of giving licensees a free bite of the apple (so long as
, they identify and co"rrect the violation) may create an excessive i
bias in favor of cure at the' expense of preven' tion.- "
- 9. Nonmitigation of Level I Violations. Th'e NRC's refusal to permit mitigation of Level I violations has been criticized as_
i.
illogical. The Committee agrees. The theoretical and practical f
justifications for allowing mitigation apply with equal force to all violations, even the most. severe. Since mitigation is based j on a percentage of the base penalty, Level I' penalties (even if I mitigatable) will still be considerably higher than_. comparable -
Level.II penalties, reflecting the former's: greater severity.-
Allowing mitigation of-Level-Izviolations will also eliminate any incentive ~ to'misclassifyla-I as a II in order to give recognition to good perfqrmance. .
.r 4
mm _______.__- _.__ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ - . , _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _
. 10. Penalizing Individuals. The Committee heard very little support for the idea of imposing sanctions on individual employees or contractors of corporate licensees. Some commentators took the position that civil penalties should never be imposed on individuals; others,- that they should be imposed only on licensed operators for particularly egregious violations.
The Committee subscribes to the more extreme view: civil penalties'should never be imposed on individuals. Since most NRC requirements run to the corporate. licensee, it is inappropriate to penalize individuals for contributing to violations of those requirements. It is up to the licensee to decide how best to meet its commitments. This includes deciding how to allocate responsi-bility and duties with the organization, how to select, train, and develop personnel, how to monitor their conduct and reward or punish Imposing NRC penalties directly on individuals inevitably
~
them.
interferes with that discretion. The NRC becomes drawn into managerial prerogatives, and employees are subj ec ted to the demoralizing conundrum of having to serve two masters. Furthermore, the attempt to assign individual blame will often require prolonged inquiry.
This does not'mean that the NRC is placed in the position of " condoning" -individual wrongdoing. The corporate licensee is ,
expected to take such remedial action as it feels appropriate, a,n d that expectation is backed by the NRC's requirement that all violations be " corrected." Correction means taking action to prevent the violation's recurrence. If the licensee fails to discipline, retrain, or reassign an errant employee, the NRC can
. ~ . . . . . - . - _ . ,
- i
., refuse to mitigate the penalty or can even escalate the penalty for failure to take " prompt and extensive corrective action."
I Licensed operators present a somewhat different case, j since they have already been singled out for attention from the NRC. Arguably they owe individual obligations to the NRC which l it can enforce directly. Since operators are licensed only to a particular facility, however, we feel that their primary super-vision should come from management. Only if they exhibit a serious incapacity to discharge their licensed functions should they be subject to direct sanctions from the NRC. In that case, the appropriate sanction should be suspension or revocation of the license. Like a civil penalty, a license action imposes a direct financial cost on the operator. But it serves the added function of removing the licensee from regulated operations and thereby protecting the publid,
- 11. Orders vs. Fines. Some witnesses recommend that the NRC make greater use of orders (and less use of penalties). Orders are perceived as more " remedial" and less " punitive." In fact there has been a modest increase in the use of orders of late.
(There were 4 in 1982, 8 in 1983, 10 in 1984.) Still, they are !
I comparatively infrequent. .
The Committee takes no firm position on this issue. It la difficult to assess the value of orders in'the abstract, since .,
their value depends on their content and enforcibility. Orders can serve an educative function and, since their use imposes administrat$ve (and perhaps publicity) costs on utilities, they l
can also serve a deterrent function. But orders require l
,. i specification of requirements to a degree that is not always feasible or wise. We have heard extensive criticism of technical specifications as excessively detailed and inflexible. We fear that widespread use of orders would compound these problems. .
~
The NRC ought generally to be moving in the direction of sub-stituting perfornance (output) standards for engineering (input) standards. The former allow industry more room to " breathe" (to innovate and develop efficient s ol u t io ns ) . Furthermore, orders like other requirements must be enfbreed. Eventually, NRC must be able to make a credible threat to impose costly sanctions on violators.
- 12. SALP Ratings. Like most things, SALP is good in theory but imperfect in execution. The main criticism seems to be based on the subjectivity of the ratings.
SALP ratings are heavily influenced by resident inspectors who may have a limited or distorted perspective.
Some subjectivity is unavoidable in any c2 ting system.
-The Committee agrees, however, that steps must be taken to reduce the impact of interpersonal subjectivity. All SALP ratings ought ideally to be done by the same group of people. Whatever their biases they will hopefully be applied consistently across .
I the board. We therefore urge the NRC to reduce the influence of individual residents in the SALP process by making greater use of SALP rating teams that will rate several plants.
l
_ .1 ~ ' ~~
f . =.
4
. D. Data Collection The Committee's difficulties in evaluating the performance of NRC's enforcement policy dramatizes the im-portance of establishing an ongoing system of collecting and evaluating data. The Committee does not mean to diminish either the difficulty of that endeavor nor the NRC's considerable accomplishments in this area to date. But more can and should be done.
As indicated earlier, NRC gathers extensive data on licensee performance'that could conceivably be used as a basis for evaluating the success of enforcement. Most of.
these statistics suffer from the obvious drawback that they are only weakly responsive to variations in enforcement level or weakly re?.ated to the ultimate eriterion of regulatory success --
safety: On. balance, we feel-that the best test of enforcement effectiveness is the extent of licensees' -
compliance with their substantive obligations. " Extent of compliance" must, of course, reflect both the quantitative I and qualitative dimensions of compliance (how often did violations occur and how serious were they). The best ex-isting measure of the latter is the severity levels assigned, to violations by NRC inspectors. -Using the severity levels one can compute a " noncompliance index" for any particular ,
facility. The procedure would be to weight each violation by a percentage figure corresponding to its severity level (derived from- Table 1B of the enforcement policy), and add the
'f ,
i resulting values for all violations in the sample. Index figures could be computed for an entire facility (e.g.,
generating unit) or for each of several activities conducted by a unit (corresponding to the Supplements). Then by tracking the movement of these index figures from year to year, one can roughly gauge the relative impact of enforce-ment efforts.
The Committee recognizes the shortcomings of this approach.
First, thu averity level classification system,is only a very rough and imperfect measure of relative safety-significance 1 of violations. But it is the best measure available. Obviously, i e NRC should continue to refine and improve it, whenever possible. Second, the violations detected by NRC (including those reported to it) do not comprise all nor necessarily a representative sample of the violations that actually exist.
For this reason, NRC may wish to adjust each licensee's index figure by a factor representing the relative intensity of the'NRC's regulatory oversight or the relative reliability of the licensee's self-detection and reporting program. Third, not.all variations in noncompliance levels can be attributed to variations in enforcement. For that reason, NRC should count only those violations that are in principle preventable. But, i as explained earlier, it should be doing that in any event. -
It is true that the level of compliance may vary because of
' independent changes in the strength of nonregulatory incentives
_(new state laws, changes in the insurance market, new manage-ment). But those are precisely the sort of events that NRC needs to know about. If compliance falls because those
, incentives weaken, NRC must toughen its enforcement program accordingly. One may also object that no single measure of compliance (or noncompliance) can do justice to the wide variety of requirements (especially tech specs) applicable to different licensees. For that very reason we do not
- recommend using the " noncompliance index" to compare licensees (or plants or even units). Rather, it should be used to track the performance of any one licensee (plant or unit),
over time. -
This system requires that all preventable safety-significant violation;be reported centrally. That may not presently be the case since an unknown number of Levd IVs and Vs are never cited (as an inducement to self-regulation).
Since the criteria for noncitation. include self-detection,
^
reporting and correction, omission of those violations is probably not a sericas loss. Incorporating these' violations s
in a performance index might well reduce the incentive for effective self-regulation. Another obstacle is the fact that Level IV and V violations are not routinely reported to headquarters. But that is not a serious problem, since data on those violations are captured in the 766 system. .
Presumably the system can be easily modified to report and ,
present the data in an appropriate' format'and to the appropriate.
l officials in IE HQ.
l
. e t
l I
et ,,-
F, . '
A G . 6'~~
- \
l
- l l
INSPECTOR SELECTION, TRAINING AND EVALUATION NRC has established a progam for the selection, training and evaluation of inspectors to meet two basic and essential objectives:
e To ensure that inspections meet minimum knowledge and
-qualification standards, and e To maintain inspector qualifications and training in general conformance to standards for comparable job activities in the nuclear industry.
SELECTION Hiring of inspectors in general
- follows the standard process utilized by other federal agencies. Career opportunity announcements are listed in NRC Form 114 with'a description of duties. The form also lists qualifications required by the job and the rating factors that are utilized t-. evaluate the candidate. Interested candidates apply through the submission of a personal qualifications statement, SF 171, and a vacancy application status notice, NRC Form 115.
Basically, three types of appointments are made:
i Resident Inspector, Senior Resident Inspector and Regional Inspector. The first two types of inspectors must have been an NRC employee for six months and two years, respectively. This is advantageous in that appointments can take into account actual performance on the job. Regional inspectors can be a new employee.
TRAINING Newly hired personnel seldom possess all of the required qualifications. Therefore, both formal class room training, as stipulated in an Inspection and Employment Manual, (Chapter 1231) and on-the-job training are utilized to ensure that the necessary knowledge is acquired.
Each inspector must attend the required training or verify through successful completion of.a written equivalency examination the desired level of knowledge in the inspector's
. specialty, such as Radiological Safety Inspection.
- Certain positions require that the candidate already be an NRC employee.
{
^
3 o
The passing grade for each course, series of courses, or equivalency examination is 70%. It is puzzling to note that the passing grade for licensing of operators in 85% - a case where the pass-fail criterion for the inspectors is less than those of -
tha inspected, i.e. operators. <
Other requirements, such as oral boards or regional training, may be used to supplement or enhance training. For example, there is a training manual, "IE Training and Qualifications Journal" that goes with IE 1231. Upon completion of the training journal, the region.may convene a " Board" to give the candidate an oral examination, bpon the Board's recommendation, the inspector is certified.
Neither the composition of the Board nor its process is standardized. Two regions utilize formal boards, two regions do not utilize boards at all, and a fifth region utilizes an informal process wherein one regional representative conducts a walk th% ugh evaluation. Little is known about the reliability and validity of_either the formal or informal board evaluation process.
All newly-hired personnel and assignees are required to complete the training activities or take and pass equivalency examinations within the first 24 months of being assigned.
In general, only those inspectors who have successfully completed the required training will be allowed to perform inspections. However, if regional management evaluates the background and performance of an individual inspector and concludes that_the inspector has demonstrated an ability to perform inspections in specific areas, even though the required training has not been completed, the Regional Administration can authorize the individual to perform inspections in those areas.
EVALUATION Ferformance evsluations are accomplished by standarized I annual appraisals. This evaluation puts major emphasis on objectivity as described 3y a chapter in the IE Manual. The emphasis upon objectivity leaves the impression that deficiencies !,
in other' areas; such as job knowledge, work habits, communi-cations $3Blens skills and accuracy of_ findings, can be forgiven C as long as objectivity is maintained.
In summary,_NRC has put in place a program for hiring, training and evaluating inspectors so as to achieve policy 1 objectivesofassuringthecompetencenecessaryforinspect$e$Is to discharge pcsition requirements in an effective manner. l 1
}
l l
l
)
/
Oveldll, possibly be the madeprogram is apparently even better. working Accordingly, well it is but could1 (
recommended that NRC carefully evaluate three aspects of the current process \
with the possibility of revision:
- Paise passing scores for training courses and'l_ _
equivalency examinations to the same passing score:
required of operators.
e D'isallow individuals to conduct inspections as the]
primary inspector who have not passed the required training or equivalency examination. (This would notl_
preclude an individual assisting in an inspection underl
..the direct supervision of a qualified inspector).
e The rationale of utilizing or not utilizing boardsl shouldbethoroughlyreviewedwiththeobjectiveoh
/estabhshrns standarhai.iun. If boards re to %
utilized, a;samplin of board acti ns shou d be(
car'gf ully _ ex mined t assure at t e origi al in en and process a e being stringen y fu filled and t at a
( loose ess has ot infi trated t
\ proessthtwoud(]
precl(ube the a hieveme t of vali ity d reliabilil y
'" td 0 ,a d &
- c' A t ,$41( -'
L p us f
/
HLP:ba i WP:P-02
f 3
II.
GENERAL COMMENT
S l
A. A Brief Description of Current Program The NRC enforcement program is intended to " promote and protect the radiological health and safety of the public, including employee's health and safety, the common defense and security, and the environment." With vetur-- a exceptions, whenever noncompliance with NRC requirements is identified,' enforcement action is taken. Principal tools / in the enforcement actions are hordersandfines. Violations are categorized in terms of five levels of severity to reflect their relative importance.
br:r:11y pendent upon the seriousness of the violation, a civil penalty in the form of a monetary fine evied. The intent of the fine is to assure prompt corrective action and to deter future noncompliance. The severity of fines was signifi-cantly increased in 198 by Congress upon the recommendation of the NRC. The rationale for the increase was that the heavier 3, fines would O2 u suu ur-fewer violations. In actuality, there was
-,[(f[ .
an increase in the number of violations in both 1982 and 1983,as contrasted with prior years, which cannot be explained merely by the larger number of plants becoming operational.
V B. Perceptual Differences between the Regulators and the Regulated.
Unfortunately, existing regulatory practice has resulted in almost a completely diametrical viewpoint between the regulators 1
-- . - . . , - - . - ~ , , . ,,. .,, , ,e, , , , _ , , -m,r.---
, 'e
- d t.
and. utility sources (per, for example, the Federal Register Vol. 47, No. 46, March 9,1982) . The regulators believe their
^
actions are effectively serving the public interest. The utilities, while accepting the fact that they are working in a .
regulated environment, contend that there has been some 6
unintended consequences of the regulatory process which are
- ~ ;
counter productive to the intended objective and asscrt that a system which is supposed to be corrective has, in. fact, become ,
. \
- frequently punitive and adversarial in nature. The* intensity of
- utilities reaction to the current system is indicative of the dissonant and discordant impact of the current system as
! perceived from their vantage point.
' s a
Both regulators and owners share the common goal of plant and public safety; the problem is how best to achieve it. While i
continued fine tuning in the program will be required to establish a truly effective team effort, this paper addresses key issues contributing to problems in the current program and offers l
recommendations for the consideration of the NRC which the Study '
j Committee believes would~ contribute to more optimum results.
l l
C. Summary of Public Comments
^
A terse summary of salient utility-comments is given below.
Topics covered by these comments will be treated later on in this report in greater detail.,
I i
2
t o Material False Statement l
The current policy on material false statement has had a chilling effect on timely and accurate communications -
l 4 which inhibits a free and open flow of communications so important to an effective program. It is perceived by most utilities as being extremely sensitive, counter productive and either the first or second priority issue
~
requiring resolution.
o Enforcement Philosophy Enforcement philosophy is based almost exclusively on a punitive system and too little recognition is given to good performance and effective licensee audit programs.
o Inadequate Distinction Between Severity Levels There is a lack of clarity in the example of violations
, in the supplements and there are inadequate distinctions between severity levels assigned to violations in the various activity areas. Utilities feel they are in a no-win posture and that penalties are inevitable regardless of diligence because NRC seeks perfection rather than excellence.and fails to consider extenuating l
circumstances or exemplary past performance. l I
l 3
l
In addition, NRC frequently.is seen as guilty of exercising arbitrary powers unchecked by an effective oversight. Management conferences, which could be a
,very effective tool if conducted in an open and free -
exchangeprocess,aretopoftencharacterizedbyan
. arbitrary and hostile attitude being displayed by the NRC representatives.
o Mitigation Factors More recently NRC has seen fit to increase the frequency of appropriate mitigation actions, i.e., 50% for prompt identification and reporting and 50% for prompt correc-tive action. The utilities have lauded this action as a step in the right direction but complain about the
.~
exclusion of severity Level 1 cases for mitigation
. consideration.
o Consistency Between Regions Inconsistency between regions is perceived as a major problem and there is general agreement within the NRC that consistency would be desirable. Regions have authority to initiate _ routine enforcement actions but, due to-differences in philosophy and the lack of standard operating procedures, an event judged.to be a )
i violation in one region may not be evaluated as a l i
4
-. . . . ~ ~ . ~ . _ _ - - - - - - - - - - - - - - - -
l
.. -, i violation in another region. The-turn-over-ofmbe,rt of-the NRC. Commission-and-theis differences-regarding regulatory-phi-losophy is parceived as m tributing factor-iTrthe lad vf a standar(rplicy uwoughout all . l v i
% OAS, i
o Enforcement Publicity Current enforcement publicity emphasizes the negative and sends a false message to the public at large that utility managers are inept or dishonest (or both) and impairs confidence in the industry. The negative i
publicity has a deleterious impact upon morale and, as currently conducted, can result in a single enforcement being given a negative _ press two or three times.
-Utilities appear to be as much or more concerned about negative-publicity than fines per se.
o Tech Spec Problems 4
There has been a proliferation of Tech Spec 4since the
- a(
infamous TMI 1N(incident. Utilities claim that some of ;
I the tech specs are not' safety significant and others are 1 I
almost unintelligible._ NRC agrees that tech specs need- j cleaning up but, due to constrained resources, estimates that it would require ten years. -Violation of a tech spec can result in a violation and-fine.- Utilities ,
l 5
claim that it is inevitable that some tech spec will be violated and particularly when a violation is a matter of broad interpretation.
U o Imposition of Sanctions Directly on Individuals This issue is unresolved. Utilities desire mapades responsibility to manage their own personnel and oppose the imposition of sanctions directly on inBividuals.
Since causal factors of violations are sometimes a s't multiple in nature, it is difficult to responsibility to just one person. If done at all, it should be for a malicious or willful act for which the utility has been exonerated.
'~
o Morale of Licensees and Their Employers Negative publicity is not counterbalanced with recog-nition about all the things that are done right. Too little recognition and reward are given for good per-formance. Rather, there are thousands of opportunities but made one mistake and you are nailed. One very objectionable characteristic of the present system is untimely action; a problem is identified and corrected and a fine, with all the attendant negative publicity, comes along several months later which is particularly deleterious to morale in local communities. The 6
l S l frustrations from regulators pressure has contributed to some degree to employee attrition, o Collection, Analysis, Reporting of Data -
The current prugram is perceived as being largely reactive to isolated individual cases which do not lead to long-term,-broad-scale corrective actions. More emphasis should be placed on' identifying rbot causes and a pragmatic approach with emphasis upon prevention.
III. DISCUSSION OF SPECIFIC ISSUES A. Technical Specifications Until 1974, all nuclear power plants had " custom" technical specifications *, there was no consistent standOand format. In 1974, the NRC implemented Standard Technical Specifications (STS). Since 1974, and particularly since TMI, several sets of changes to the technical specifications e to (1) implement new requirements, (2) keep the technical specifications up-to-date with plant modifications, or (3) to t
make improvemenpf b ased on requests by the licensee.
The growth ~of technical specification requirements over the years has brought problems with it. Some of the tech specs are almost unintelligible and are subject to broad interpretation which can 7
k -
- . . _ 1 l
l vary significantly as a function of the training and experience of the individual making the interpretation. In addition, utilities contend that many specs are not safety significant.
ed.' ,L &
Furthermore, some utilities accepf poor tech specs so as not to -
prejudice their chance for license. The NRC agrees that the tech specs need to be cleaned up but, either out of inertia or inadequate resources (or both), have taken no tangible steps to introduce reform. An NRC estimate to effect the necessary reform N .
is ten years .
P9rt the roblem apparent is th lack of explic uidance and direction from the co mission. The e has bee talk abo t t ch spec reform r
ttle if anything h be done about i .
./ In the meantime, both the utilities and NR must live with tech ec deficiencies.
d @t ch 7 &~
B. Root Causes - Effectiveness of Dissemination c-C'xe > w
_A-prin' pal.objectiveoftheNRC.g .d cu, P gn i to identify why violations occur and to disseminate guidance and information so as to preclude their reoccurrence. Unfortunately, this objective is not being met to the degree of efficiency that it should. In many cases, the root causes of violations are very complex and may constitute multiple variables; design, communications, training, procedures, human error,-etc. The enforcement policy supposedly focuses on the fundamental problems and root causes of violations which in turn necessitates the acquisition of a lot of 8
facts whenever an individual item of noncompliance is identified--the who, what, why, where, and when regarding the violation. While there is guidance to the inspectors in a manual on how to nforcement actions and how to write an -
inspectionreportj)thereisnotastandardformdesignedto capture all the necessary information, nor is there any coding in inspection reports at all. Citations are tracked in a computer system but the total informaticn needed for analysis to determine root causes may be on different documents in differ %nt places in lieu of a standardized format for input into a computer-based data system amenable to effective analysis. NRC makes extensive use of a computer-based data system on inspection reports called the 766 Form. This form reflects the readout of the licensee's
, enforcement history, the text of the violation, the requirement that was violated, and will categorize what the severity level s
was. It does not, however, address root causes. NRC is currently structured, the major source for root cause analysis appears to be the office for Analysis and Evaluation of Operational Data (AEOD). The objective of the office is to collect, screen, analyze and feed back operational experience or data to the NRC,.the industry, and to the public. The AEOD has access to a large number of reports from the industry and foreign event reports based upon bilateral agreements. A principal source of information is the Licensee Event Report (LER)--a report which documents reportable events occurring at the operating facilities. Over 2000 LER's were received in 1984.
However, the looseness of the reporting format was such that the i
O LER's were practically worthless in determining root causes before 1984. In recognition of the dominance of the human element in many of the reported events, the LER system was modified in January 1984 to get better information on human factors variables contributing to the event. Even though the NRC does not get good information with every LER, the system has improved and the AEOD is generally successful in deriving root c}
causes @.c4 &
v;cict:cas involving equipment and hardware but are
~ MN___cas, less successful for the people oriented vac_ i.e., human error. Currently, the analysts in AEOD represent a partial multidisciplinary capability but they have only a limited capability in human factors. When queried if it would not be advantageous to bolster the capability.by adding human factors specialists, the answer was rather surprising in view of the large number of human factors specialists available. "We'd very
)
much like to. Those people are very hard to find." That some-l thing needs to be done is evidanced by the fact that a study of significant events in 1983 by AEOD and INPO concluded that the incidence of human error was around 40% and was particularly prevalent in the test and surveillance area.
I In conclusion, it is the belief of this committee that the NRC's I ""Far [ .nt P Mr N2sy would be more effective if it concentrated more on the identification of root causes and co$aective actions end its: On punitive ections includiu3 civil pencitle:. Root cause identification could be enhanced by promoting and encour-aging a cooperative atmosphere between the NRC and the utilities 4 e
10
which would be characterized through day to day interface, better/more frequent use of management conferences and enforce-ment conferences with free and open exchange between the NRC and the licensees. -
A large number of reports are disseminated by the NRC, such as notices, bulletins, circulars, generic letters, and a series of
( summaries, for example, of abnormal operational events and the
~
information notice process. Th'c apparent prcblems with these releases, as reflected by the recipient utflities, is the frequent lack of timeliness and in many cases, the lack of iden-tification of root causes and appropriate explicit solutions.
Industry interaction in other ways (sugh as pertinent news releases and shared information between utilities).and the issuance of good practices by INPO are helpful and constructive and these efforts should be accelerated along with a parallel effort by the NRC to issue more timely releases with explicit recommended corrective actions.
C. Inspection Policy & Staffing, Training and Rotation The purpose of the inspection program is to verify that licensees are meeting their commitments and regulatory programs and approximately one-third of NRC's resources are directed to the inspection program. This. includes the actual inspection, the development of programs and procedures to carry out the programs, and support to the actual on-scene inspection. The inspection 11
program is managed by the Office of Inspection and Enforcement and involves the five regions as well as some direct inspections via special teams out of the Office of Inspection and Enforce-ment. However, the bulk of the inspections are carried out by -
regionally based inspectors and resident personnel.
Another way in which violations are identified is when the licensee himself identifies the problem and reports to the NRC.
To the surprise of this committee, the NRC staff coIild not identify the percentage of_ violations identified by the NRC inspectorsandthoseidentifiedbythelicensehthemselves.
Obviously,theinspectkprogramforoptimumsuccessdependson thecontributionofboththeNRCandthelicensefs. One highly 1
placed NRC official' reported that it was his belief that the majority of violations reports were reported by the licensees.
Finally, the last way that violations are identified is through the process of allegations and subsequent follow-up of those allegations by either the investigative staff or the inspection h
staff. It is NRC policy neither to encourage or discourage allagation reporting but, upon receipt, allegations are systematically evaluated, f l &
In ' concern -for ~~a _-potential loss of obje ctivity, the NRC had a policy originally of~ frequent rotation but the.NRC has now satisfied itself that loss of objectivity is not a problem and
~
most inspectors now stay on their assignments for at least five H
l l l
l years. It is difficult to say whether this policy is predicated on the cost of relocation or inspector efficiency. The NRC staff contends that while there are no standards or quotas, they can
" smell" it when there has been a lapse of objectivity or effi- -
ciency--aninterestingclaimnotfoundedinscientifickigor.
There seems to be a problem of burnout and the fee. ling of some inspectors that confinement to one plant does not give them the proper frame of reference for comparative purposes wheti they are only familiar with one plant. Another complaint is'that the volume of paper work reduces the time available for actual inspection activities and that there is in effect a " quasi" quota system which is contradictory to official policy, i.e., in the absence of reported violations the resident inspector is likely to receive qu7stioning phone calls.
A major problem seems to exist in the area of materials licensees, i.e., fuel facilities, uranium mills, new fuel produc-tion plants, hospitals, and other industrial users of uranium products. There are roughly 8000 licensees of this type licensed by the NRC throughout the United States. The NRC tries to look at those with the most significant' quantities of radioactive material and those that represent potentially the greatest threat to public health and safety. Those licensees in these areas are inspected usually about once a year, but in the case of other materials licensees it may be years before they are inspected by the NRC. This program, like the reactors, also depends upon voluntary reporting.
13
}
.. \
Another complication in the materials licensee category is that, in addition to the approximate 8000 cases where the license is issued by NRC, there are that many again where the licenses are issued by agreement states.
By'NRC's admiss oni they have,,not
/
looked at this total population
/
of users of~ regulated nu, clear
~,
j materials to derive an'evaluative conclusion of the quality of
/
conduct with regard to the agreement/state licensee and the NRC l
l licensees.
/ .
hey do, however, coordinate their ac ivities with agreement states.
Pttethermera, it is a requirement that the NRC judge the compara-bility of agreement states programs to,the NRC programs. /This would apparently mean that NRC should have_in hand-a' sufficient <
body of information, not now available, that can be evaluated so OL- ';
as to able to form judgments about the quality and comparability l of the agreement states programs, i'.'e., number of enforl cement :
/
actions, and the magnitude and sco/
/ pe of problems which are surfaced. - - - - - -
~ - ~ ' ~ - - '
In addition to the inspections discussed above, NRC also makes use of SALP (Systematic Assessment of Licensee Performance) y wctep J ratings as a generic evaluation of the licensee in areas such as
/7 operations, radiological controls, quality assurance, and safety. 3 ALP ratings, while not necessarily a violation, are .
perceived and used as a vehicle for effective dialogue with the
. licensee about his problems.
14
The inspection system of operational plants generates thousands of notices of violation annually which are candidates for fines. The largest number of notices of violation involve ,
severity level 4 and 5 violations which are issued routinely by .]
the regional inspe tion process. These violations are not d M Ctt d reviewed by the N C aff before they are issued. It is estimated that there are 1000 to 2000 notices issued annually not previouslyseenbytheofficefofIandE. The I aiid E staff has, in recent months, been getting copies of the notices to do some kind of auditing as to what is going on but, because of the large number of notices and a small staff, it is difficult for I & E to get a handle on it. The problem of inconsistencies on the overall program between regions is therefore complicated by the fact that I & E does not have a good fix on the direction and consistency of the regions in regards to severity level 4 and 5 cases. NRC " thinks" they have reasonable consistency on 4 & 5, but no steps have been taken to assure that there is a relatively even-handed perception and applications of the 4 & 5's among the regions. I & E justifies this situation by the fact that they have a' limited number of personnel and cannot 1cok at every-thing. However, it is significant to note that when an initial ;
effort was made to try to smooth out inconsistencies, or at least reduce strong differences, the effort was discontinued with the encouragement of the commission.who wanted it to be essentially a regional program.
I 15
Adequacy of Resources On the surface, at least, a case can be made that more resources could be advantageously utilized in the enforcement program. .
Four problem areas discussed above continue without resolution wherein more resources, carefully selected and effectively utilized, could have a positive impact. These problem areas are: the achievement of consistency between the regions in the perception and interpretation of requiatory policy,* cleaning up a voluminous and frequently unintelligible tech spec file, a more effective and multidisciplinary analysis of violations so as to determine root causes, and a more frequent and effective evalu-ation of the quality of inspection programs in the material licensee category--both those licensed by the NRC and those issued by agreement states. There is no question that the man-datory activities in the material licenses area must continue and require NRC responses apart from responsibilities for operating plants.
A fundamental question regarding more resources is the impact of these additional resources en improvement in safety.
Unfortunately, there does not seem to a direct answer to this question since there are no quantifiable performance criteria to j determine the efficacy of the current program and if current resources are being optimally utilized. Valuable resources, both NRC and the utilities, are being dissipated by undue emphasis on problems of minor significance. Level V violations are very 16
.____________i_____._____________._._._________.___._.____.
. .. m -
numerous and normally receive no attention except by the licensee cited. However, they still require experienced manpower to support and respond to the activity. Calling these events \
deviations, rather than violations, and resolving at the local .)
levels without volumes of formal paperwork could represent a potential for significant savings of resources. In other words, )
f
, if a licensee has an acceptable program for identifying and evaluating deviations and taking corrective actions, these category V events would be treated as' nonconforming-deviations and resolved locally without all the attendant paper work associated with a violation.
Reducing the frequency of inspections at plants with good performance records and increasing the frequency at plants with poor records could be a more effective use of resources while
^
serving as a important motivator to plant managements. In addition, it would concentrate attention on licensees who are in need of increased guidance and assistance. The emphasis would I l
also shift appropriately to prevention rather than after-the-fact I l
, fixes. To negate any possible lethargy and to maintain the viability of this approach as an incentive, a reasonable minimum of frequencies should apply to all licensees.
hfundamentalquestionshouldconcentrateonthefocusofthe ,
_A '
k_ inspection process and are manpower resources being effectively 9 i (utilized? 'For example, is there a lot of inspection activity goina on in areas that do not accrue major benefits in regard to 17
~-
c safety? Are there existing regulatory requirements which can be eliminated that have a marginal importance to safety? Would it be effective to divert more manpower to the establishment of a viable preventive enforcement program rather than the- expenditure g_
4 -
'-O-ef these resources in after-the-fact investigations? (Very little is being done in preventive enforcement at the present time due to limited resources.)
In summary there seems to be little doubt that the Enforcement program could utilize more resources. However, probing cost-benefit analyses should be conducted to assess the priorities for generic safety issues, the need for new regulations and the deletions of marginal requirements, and the overall effective utilization of manpower resources. Questions other than those posed in the prior paragraph will undoubtedly surface as a result of a thorough study of the resources issue.
Management Structure Since the 1979 TMI accident, a number of reports and studies have questioned or been critical of the NRC management structure and organization of the NRC. One report stated the NRC "is not so badly managed as it is not managed at all."
A recently completed study, the-Bevill Report, concluded that the organization and management of NRC is a blueprint for regulatory paralysis and that the collegial decision making of the 5 n ?fm 18
o commission is a fiction. Instead, " personal hidden agendas" held by individual commissions thwart collegiality, sow distrust at h
the highest agency' levels, and sends mixed and confused messages to the major staffs and employees. Finally, there appears to be -
a lack of a clear understanding, voiced by many former and present commissione as to their roles, responsibilities, and authority. Certainly, while the chairman of the Commission is perceived as an official spokesperson for the agency, there are few commissioner decisions, opinions,"or letters sent to Congress ;
which do not include minority or dissenting opinions from indi-vidual commissioners. A case in point is the single adminis- l trator concept as an alternate to the current 5-man commission.
A second major finding of the Bevill Report is that there exists a costly lack of discipline. The agency staff, composed of four major program offices, is almost a separate entity within the NRC and lends credence to the allegation that the NRC is managed bottom-up. An example of this is the regional approach wherein each region is managed like a fiefdom, which represents the personal philosophy and perceptions of the five regional administrations.
)
1 A clear case of lack of discipline and control and consistent policy is represented by backfitting , Backfitting, an over-reaction to TMI II, has run into the hundreds of millions of dollars for some plants. For example, Beaver Valley faces increased costs of over $200 million for 15 backfits. Unfor-19 w e
~ ~ ~ -~ ~~ ~ ~
c tunately, however, there did not exist immediately after TMI II a master, well-coordinated, carefully formulated scheme for the imposition of backfits. Instead, they were levied by many different levels of the NRC:
on-site NRC inspectors, regional -
inspectors, and management officials, and by staff reviewers.
Collectively, these levies have represented personal inter-pretations of licensing requirements and their own criteria for operations. -
The.NRC,'at least in some-instances, also overreacted and over-prescribed licensing requirements as a consequence of TMI II.
Two examples support this contention: (a) The TMI accident had 3 or 4 known causes and yet NRC prescribgd 180 new requirements and (b) in accordance with 1972 technical specification requirements, a licensee performed 390 surveillance _ tests per-year; in 1984,
- s the number had mushroor-?d to 14,000 surveillance tests per year.
The NRC, to their credit, has recognized these problems and now has a regulatory review process based on cost-benefit analyses to judge whether a new requirement should be imposed or not.
In the case of backfits, two actions have been taken (or proposed) to correct admitted deficiencies. First, the respon-sibility for backfitting activities has been placed in the-hands of the Executive Director for Operatior.s; something that has not
. existed before. Secondly, NRC_has proposed and received comments on a new backfitting rule, which should significantly improve 20
2 s.
I backfitting procedure by requiring cost-benefit analysis unless l the Director of Muclear Reactor Regulation feels that there is an urgent need to take care of an important safety problem at a l
1 particular plant.
NRC " hopes" to implement this rule in the near -
)
future after all the comments have been received and digested.
Testimony from key NRC personnel also reveals a good appreciation of the fact that the system must strive for excellence rather than perfection,A*A esk that a proliferation of small civil penalties may be counterproductive by diverting the skills and talents of key personnel, in preparing and responding to the. paper work, away from more important safety matters.
Since many of the fundamental deficiencies of the NRC as an agency are traceable directly or indirectly to the Commission, it is appropriate to look at some of the circumstances under which _
they must work. The commissioners themselves are highly motivated and possess eminent professional qualifications in one or more technical areas. However, their actions and effective-ness are significantly influenced by laws which, while well intended, have had a counterproductive effect on the collegial process.
The first of these is the Sunshine Act which, as currently inter-preted, means that the Commission cannot meet as a body to hold evaluative discussions and resolve differences except in public sessions. Sinck,the resolution of differences is a give-and-take I
l l
21
i' 1
0 trade-off process, holding the sessions in public has a constraining influence and hampers t'he effectiveness of the collegial process. ,
i The second major constraint is the ex parte considerations which.
in' practice, does not allow the commissioners to make fully effective use of the abundant technical talent within the NRC in directly advising the commissioners on specific cases. Thestafp is a technical resource that should be available to'the commissioners without overly restrictive legal constraints.
l e The relaxation of the restrictions should have a positive impact
- upon the achievement of the rationale ,for the collegial process. This committee did not have the time nor resources to examine the pros and cons of all alternative approaches to the 4
present commission structure. These alternatives include:
single administrator, 3-man commission, and possibly one commis-sioner with a technical advisory staff analogous to an industry j l
- CEO with a staff or a Board of Directors. A specific study of
1 this issue may be necessary to spell out all options and the pros and conc of each. Less elusive, however, is the need for con-solidation of NRC facilities. The NRC headquarters staff is l spread out in eleven different buildings thr:oughout the 4
Washington, D.C. metropolitan area. This involves a valuable
! waste of resources in the expensive and time-consuming process of
. traveling between these locations. Both the Kemeny report and l the Grace Commiselon have urged geographic consolidation, and 1
4 22
c this committee report strongly endorses those recommendations.
, \
In summary, NRC has recognized and is taking action about some problem areas (such as a more deliberate approach to backfittin' '.g
\
and new requirements), have recognized but are dependent upon !
Congress or other agencies for resolution of some problems (Sunshine Act, Ex Parte, facility consolidation). The NRC should, however, initiate more careful analysis and study so as to maximize total manpower efficiency. Particular attention should be given to the priority significance to safety as regards the focus of inspections. Guidance and training programs should l' be initiated to achieve more consistency between regions.
Overall Impression of Effectiveness of Policy The Bevill report contains the following statement, "The concern
/
is not only with waste of the taxpayers' money on a poorly functioning agency but, mor impey/tantly, the increased cost to
, the ratepayers for the ene y theystequire, the demise or at least postponement,,o nuclear optio'nband most importantly, the possibility'that NRC actions are making plants less safe."
1
-UnFnr*nn=tel"-ssertionsabouttheeffectivenessoftheenforce-
# I obt.
ment policy and its impact on safety bq largely subjective in the absence of quantifiable performance criteria. With the exception of exposure to radiation by personnel who work in the material licensing area, where there it. documentation that exposure has 23
4 been reduced, other conclusions concerning safety are largely subjective in nature. It is the opinion of this committee that the lack of an attempt to derive quantifiable performance criteria, by which the overall effectiveness of the program can be judged, represents a serious omission. Admittedly, this is a difficult task but, with advanced psychometric techniques and modern computers, it is believed to be amenable to solution. Four parameters are tracked by all French nuclear plants: o Equivalent availability , o Forced outage rate o Number of planned outage days o Load rejection rate (load rejects without s~ rams) The French parameters are obviously more related to plant effec-tiveness than safety. Other candidates, more specifically related to' safety, which could and should be evaluated for feasibility, might include: o Unplanned automatic scrams 24
4 o Workers exceeding radiation exposure limits (5 rems /yr) o Average total radiation dose per plants o Lost time accident rate o Incident of human errors A concentrated effort in this area would undoubtedly unearth other candidates. In the absence of currently existing quantifiable performance criteria, subjective judgments must be, exercised to assess the current effectiveness of the system. Following this approaq,h, a e list of unweighted positives and negatives regarding the enyorpe-ment program can be derived. Positive There is no question that an enforcement policy is required since there are a percentage of plants (probably 10 to 15 __) that must be forced to discharge their responsibilities and by the admission of at least one utility who was heavily fined, "The fine really got our attention." A major deficiency identified by the Kemeny Commission, inadequate training, has been vigorously addressed by both the 25
4 NRC and INPO to the degree that it can be confidently concluded that training has been strengthened and improved. For example, with the effective use of task analysis, explicit training requirements have been identified so that training is not only - more thorough and efficient but, by being specifically related to performance requirements, is also more cost-effective. The licensing process has been evaluated and impro.vements have been made and work continues apace to'effect furthet improvements. Procedures, and particularly emergency procedures, have been improved. . , Maintenance has finally been recognized for its importance and is now receiving accelerated and overdue attention. The industry has established a high level NUMARC 4 Committee to work coopera-l tively with the NRC to further address and resolve problems in l this area. For example, more emphasis is being given to preven-tative maintenance. New, improved operators' aids, such as the SPDS, have been derived and are being implemented. These aids help to reduce the stress and workload of the operator, and thereby make the operator more effective in both normal and emer-gency operations. Some of the mots flagrant human engineering design deficiencies identified by the Kemeny Commission and o Rog(vin reports have and are being corrected, i.e., improved labeling, color, and shape coding of instruments and controls; 26
4 improved environmental conditions; improved layout of panels and consoles; improved communications, etc. Research continues to effect further improvements in annuciator alarm systems, the use of CRTs, and computer-generated display systems. . The NRC has recently seen fit to reward good performance (prompt identification and correction of problems) by more extensive use of mitigation of fines with the exception of Categ.ory I fines which, for some inexplicable reason, l's still disallowed. Negatives Despite the fact that the overall efficiency of the program depends upon the best efforts of both the NRC and the licensees, the NRC cannot answer the question as to the number of violations that were identified by the licensees. 1 No attempt was made to study the enforcement policies of foreign countries and attempt to adapt some of their best features. At the request of this committee, a quick survey, via use of questionnaire, was made of the major countries with operating reactors with which signed information exchange agreements exists. Timely responses were received from France, Federal Republic of Germany, the United Kingdom, Sweden, Switzerland, Belgium, The Netherlands, Japan, and Korea; Taiwan was expected to reply at a later date. Results from this survey reflect the fact that the respondents find the " adversary" approach in the i 27 l
- _ _ - - _ _ _ _ _ _ _ _ _ _ _ - - _ . - . - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ - - _ _ . _ _ _ _ - _ _ - _ _1,
J
-e United States difficult, if not impossible, to understand. The foreign regulatory agencies primarily view themselves with pride l as working as part of a total nuclear team working in the best interests of their country. In this context, the regulatory -
agency works hand in hand with industry rather than distancing itself from it. The current enforcement policy does not go far enough to use the regulatory process as a means of constructive interface between the NRC and the licenses to maximize the safety and reliability of the plant, and seems to ignore that the greatest incentive of all for the detection and correction of potential violations is the owner's concern for its substantial capital investment in plant and equipment and the protection of its own employees and reputation. Despite the continuing responsiveness of the utilities to queries from the NRC regarding the enforcement policy (starting in 1982), the NRC has, in the main, chosen to ignore these learned and thoughtful inputs and perceptions. A case in point is the very j sensitive issue of material false statement wherein the utilities have consistently pointed out that it has had a chilling effect on the timely inputs so vital to the identification and resolu-tion of safety-oriented problems. Despite the intensity of these utility inputs, no positive corrective action has been taken by the NRC. 28
- - = = = -
o The current policy has helped promote a public perception that nuclear power is unsafe and that plants are not operated any safer today than at the time of the Three Mile Accident. Rather than a policy to educate the public and political bodies of the .
- country of the safety of nuclear power, so essential to the future viability of the industry, the current policy is too punitive in nature. The reward for prompt reporting is the probability that these reports will be converted into written violations, escalated enforcement actions, poor SALP reviews, fines, and negative press releases. By contrast, it is likely that a policy with a focus on root cause, identification, and positive incentives will improve overall enforcement and minimize the need for NRC sanctions except in isolated cases of non-
, responsible, noncooperative licensees. Even in these cases, orders are perceived as being more effective than fines. A number of deficiencies or problems have been previously iden-tified and reported elsewhere in this report. In terse summary, the most salient ones are: ( Lack of effective discipline and central direction and
/\
Control. o A very sensitive Material False Statement issue has not 1 been resolved. i o Lack of quantifiable performance indicators. 3 29
. .= _
l e o Root cause identification has not been addressed on an 1 i extensive, systematic basis.
~
qcd k ! o emphasis on punitive measures has created an A adversarial relationship between the NRC and licensees. I b $ o A recognized problem with Tech Specs'iq not receiving
~
i appropriate attention.
, o No systematic effort has been made to study foreign enforcement program and adapt their better features.
Too little emphasis placed on preventive enforcement. Lack of clarity in the example of violations in the e i supplements. 1
' Inadequate distinctions between severity levels.
K V 400--l-14t-le-and-ineffactua1--use-of--managenent , \ conferences. o Negative publicity releases have'sent false messages to the public and impaired public-confidence in the. industry, and has had a deleterious impact upon morale , of the managers and technical personnel who must operate 30
- . . n=.-- w s
e and maintain the plants. Too much emphasis upon insignificant and trivial events. 4 Within constraints of time and resources, this report does not attempt to restructure or rewrite a revised enforcement policy in explicit detail. However, it is believed that a restructure of the policy to establish an improved environmental. climate that is both equitable and acceptable to the iicensees woul5 improve overall effectiveness. Since overall effectiveness is also inextricably related to safety, it follows that safety should also be enhanced. The framework for such a system would have the i following characteristics: i o Establish quantifiable performance indicators. o Establish and maintain timely and open communications channels. 1
)b s Differentiate and concentrate on real safety concerns with deemphasis upon trivial and insignificant events.
J o Reward. timely identification and correction of problems for all severity levels. I o Reduce regional inconsistencies through' improved training and the formulation of standard operating j 31
e policies and procedures. yq' Recognize utilities innate interests, investment, and responsibilities and afford them more flexibility to , conduct their programs. o[s Put more emphasis on a more pragmatic approach with more attention to prevention and less emphasis,upon after-the-fact actions, which are punitive in nature and frequently untimely. o Make expanded use of management conferences.
% For the small percentage of hard-core problem utilities, make more frequent and effective use of orders.
) Replace the adversarial and punitive approach with a cooperative " team" approach.
o Restructure enforcement publicity-to reduce the negative and " false message" connotations. o Initiate a priority and accelerated program to clean up Tech Specs. o NRC should increase the timeliness of their enforcement decisions. 32 1
. . -- . ::=
4 o
\ <
Exercise centralized and disciplined control of backfits and new licensing requirements. o Initiate an accurate reporting system which would establish the source of violation identification broken down by severity level. Derive an improved data base'so the quality of the agreement states programs can be better judged.
+
l O 33}}