ML19340E333
| ML19340E333 | |
| Person / Time | |
|---|---|
| Issue date: | 12/04/1980 |
| From: | Labarkin M Advisory Committee on Reactor Safeguards |
| To: | Sege G NRC OFFICE OF POLICY EVALUATIONS (OPE) |
| References | |
| FRN-45FR71023, RULE-PR-50 NUDOCS 8101140066 | |
| Download: ML19340E333 (3) | |
Text
D b.
/p* *%
UNITED STATES Y
(45 M 7/023
{
g NUCLEAR REGULATORY COMMISSION
~
c ADVISORY COMMITTEE ON REACTOR SAFEGUARDS o,
f WASHINGTON, D. C. 20555 s, -
f 4
9 December 4, 1980 9
O DOCKETED
\\
G. Sege, OPE I. Roo, DEC 8 Sf6 >
J kin, Assistant Executive DirectIr 0
FR0ti:
M.
b 6
-M, for Project Review H
O G
SUBJECT:
DRAFT " APPROACHES TO SAFETY GOAL FORMULATION:
A WORKING PAPER ON ALTERNATIVE FRAMEWORKS" The outline proposed appears to suffer from a certain amount of definitional confusion. Since the paper is intended to outline alter-native approaches to safety goal formulation, the items considered should be limited to safety goals. The categorizations attributed to Slovic, et al in the section headed " Alternative methods of approach" deal with decision techniques rather than safety goals. To be sure, safety goals and the tecnniques used to determine the adequacy of a given endeavor must be compatible, but safety goals and decision tech-niques are not interchangeable.
Methods which fall into the categories described by Slovic will yield results which must be tested against established goals before a conclusion
-~
as to acceptability is possible. See, for example, the discussion begin-ning on p.168 in Slovic which makes clear that a number of different criteria (" safety goals") are possible within the context of a fomal i
(cost-benefit) analysis. These include "... the greatest good for the greatest number..." and the "Pareto optimality criterion..."
I suggest that the outline be restructured so as to focus entirely on safety goals, against which the results of analysis techniques can be checked. The first two sections should be titled (something like):
"Non-quantitative approaches" and " Quantitative approaches." A decision as to which approaches, of either kind, should be included in the pre-liminary paper cannot be made until the matter of the criteria against which the sagety goal will be tested has been settled (presumably, some-time after 12/4/80') Meanwhile, I offer the following additional sources for suggestions of appropriate safety goals which can be tested against the criteria finally decided on:
Non-Quantitative Safety Goals W.W. Lowrance "Of Acceptable Risk, Science and the Determination of Safety." A section on " Guides to acceptability" includes the following, with some discussion of each:
810114 0 og
.. _. g.
7
- Reasonableness Custom of usage Prevailing Professional Practice Best Available Practice Degree of necessity or benefit No detectable adverse effect Toxicologically insignificant levels (although the above are discussed in the context of environmental pol-lutants and food additives. I believe that in most cases the concepts could be recast in a more generally applicable form)
" Issues and Problems in Inferring a Level of Acceptable Risk,"
Salem, Soloman, Yesley, R-2561-DOE.
Includes, in an array of Risk Reduction Goals: " Minimization of Socially Perceived Risk".
In an earlier paper, Slovic, et al, discussed the concept of
" Expressed Public Preferences" as contrasted with " Revealed Public Preferences".
Quantitative Approaches R-2561-DOE, cited above, includes, in an array of Risk Reduction Goals :
Minimization of maximum accident consequences liinimization of probability of most probable accident Minimization of both total accident and normal operational risk flinimization of only accident risk Zero total risk As low as reasonably achievable risk Equitable share of risks and benefits Minimization of peak risk General and occupational population share equally in risk In summary, the suggested goal approaches you have included, plus those noted above, plus any others which come to light should be screened by testing for conformance with the criteria (or sets of criteria) finally decided on. The result should be a set of options for approaches to safety goals, each of which satisfies, in some measure, a stated set of criteria.
I believe that, in most cases, the nature of the safety goals will determine at least the form or type of the techniques which are to be used to check the degree to which any individual technological application satisfies the
i t
fr.=)'
. =.
. goals. The outline should, therefore, include a section dealing with proposed technique options, including uncertainty, which you have, of course, already included.
cc: E. Hanarahan, OPE R. Bernero, RES K. Goller, SD M. Ernst NRR G. Page, NMSS D. Thompson, I&E f
e l
l l
l l
I L