ML20147F807
| ML20147F807 | |
| Person / Time | |
|---|---|
| Issue date: | 11/30/1978 |
| From: | Rubinstein D NRC OFFICE OF MANAGEMENT AND PROGRAM ANALYSIS (MPA) |
| To: | Hendrie J NRC COMMISSION (OCM) |
| Shared Package | |
| ML20147F798 | List: |
| References | |
| RTR-WASH-1400 NUDOCS 7812260020 | |
| Download: ML20147F807 (2) | |
Text
- . - . . -
g - '
if* *% .-> -
umTED STATES
'/ \ ' NUCLEAR REGULATORY COMMISSION f hy
= ~
WASHINGTON, D. C. 20555 }
a , t k +n..
,/ .flosamber 30, 1978 h.
5
, MEMORANDUM FOR: rman Hendrie
. Commissioner Gilinsky '
' 'J LM Commissioner. Kennedy Commissioner Bradford :*
Commissioner Ahearne i J U 10: 0 ;
FROM: David Rubinstein, Applied Statistics Branch .'
0ffice of Management and Program Analysis '
t SUBdECT: .PERS0flAL' IMPRESSIONS ON THE' EVALUATION OF THE LEW,'So ,.
1' COMMITTEE REPORT AND ON RISK ASSESSMENT ,
In line with the -open door policy, I am writing to'youL about issues .
that have been nagging nie since shortly after my employment. in_ the ,
Applied Statistics Branch three years ago. These issues have come to the' forefront since the. Lewis Committoe Repcrt and its still evolving aftermath. While my preoccupation is with the nuts and bolts of statistical analysis and methodology I am not oblivious to '
the broader issues . implicit in risk assessment as a regulatory tool. .q First, I should like to state briefly two subjective impressions:
-(1) The Lewis Committee Report reinforces my very negative opinion of WASH-1400.
(2) The Commission, at least initially, seemed to try to resolve the WASH-1400 problem too quickly.
The problem that does not seem to. be urgently addressed is: To what extent is WASH-1400 symptomatic of. the working of NRC? To raise some pertinent questions based on the Lewis Committee report: ,
(1) To what extent are conclusions reached on the basis of inadequate data? ,
A (2) To what extent are inappropriate statistical methods used in NRC analyses?
(3)' To what' extent are NRC documents inscrutable or insufficiently reviewed by peers?
(4) To what extent do NRC summaries lead to false impressions?- l
\
I 781226oGO
/
. ... .w- .. . . - - .-.---.-_...--.-.a.-.. . . . . , - - . . . . - -.,,.,a . - . , . _ , , . , , -
q- . . . ..
The Commissioners - 2.- flovember 30, 1978 .
d In an ideal world one should provide explicit answers to these ,
questions. I am not prepared to provide them categorically. The questiens derive from the ratner intimate relationship of WASH-1400 to-fiRC in general. Specifically:
(1) WASH-1400 is the product of AEC/f1RC.
(2) WASH-1400 was widely accepted in f1RC; in fact by part or even perhaps by most of the Commission it was treated like a sacred cow. We have presently strong defense of WASH-1400 from some quarters. The question to ask is, why did we need the Lewis Committee to tell us what the trouble is?
(3) Risk assessment is the basis of fiRC regulation.
The last point deserves some elaboration. I presume that most if not all our regulations are in some way influenced by the perceived seriousness of a threat or consequence. This implies formal, informal, or intuitive risk assessment.
The p'oint in examining questions such as those above, is not to find scapegoats but to improve our regulation. In fact we might even go a step further, that is, to examine thoroughly the inherent limitatiens-and profound difficulties of risk assessment in the nuclear field. Then with more explicit cognizance of the limitations we can improve the formulation of sensible policies and principles for regulation.
I do not claim extensive study of WASH-1400 or of the broader issues of f1RC performance. I am not even sure whether these issues can be satisfactorily resolved or whether as a matter of practical policy they better be left submerged. I do feel, however, that even partial resolution will enhance flRC performance and that adverse effects are likely to be temporary and moderate. Therefore, I urge consideration of these issues by people with better and broader perspective of flRC than I can bring to bear. If these issues have already been considered at highar policy levels then I apologize for taking up your time unnecessarily.
14&N Adk:At David Rubinstein Applied Statistics Branch, MPA cc: L. V. Gossick fl. M. Haller Dw E 5 . i 11 R. Moore ts ., m, , 2, ' l
- .m.e b d o W 'A "~' ' [ ' '
.% c
, s ,. ,.s - m re ma n J n%
w g ., . - - - , , , __. m ,