ML20010A760
| ML20010A760 | |
| Person / Time | |
|---|---|
| Issue date: | 07/29/1981 |
| From: | Sege G NRC OFFICE OF POLICY EVALUATIONS (OPE) |
| To: | |
| References | |
| FRN-45FR71023 45FR71023, NUDOCS 8108120188 | |
| Download: ML20010A760 (11) | |
Text
hCh bk Y0kjGh (45 V 8 c41023 DEVELOPMENT OF A SAFETY G0AL:
E Overview of the Issues
<f N
and Highlights of the Wrap-up Session of the pf Doegg7g0 4
Second Safety Goal Workshop usuae 5 198;, _3' 2
AUG T
George Sege 61 Office of Policy Eva!Jation ofth 9
0' Presentati'.n Before the Subcommittee on Reliability and Probabilistic Asses
,n y
Advisory Committee on Reactor Safeguards c
~
s x '
Los Angeles, California
/9 Ig k
July 29, 1981 4.2 W e oZ
,~
D,l,l
<=
OVERVIEW 0F THE ISSUES I am going to present a very brief overview of some salient dy they fared in the reference safety-goal statement before you. The overview will suggest some questions for your consideration.
The Discussion Paper reflects a proposed settlement of a number of issues, pretty much in line with prevailing views that emerged from the firit workshop and other discussions.
Also addressed in the paper, sometimes incompletely, are issues that the first workshop and other discussions left very much open. There may well be room for further discussion about both groups of issues, but the Subcommittee's views are particularly sought about the latter group.
The reference safety goal includes the following elements that appeared to elicit wide agreement:
1.
Qualitative goals are formulated and used as a basis for quantitative goals.
8108120xes 810729 PDR REVCP NRCISCDS PDR
-p-2.
Quantitative goals are supplements to qualitative principles, and do not supplant them.
3.
The implementation process is an integral part of the goal specification.
4.
The safety goals are envisaged as dynamic, with special provisions for existing plants, transition, and future evolution.
C 5.
Costs, benefits, and practicalities are considered in both the basis and the implementation of the goals.
6.
Protection of the individual and society are both-provided for.
9 7.
The basic scope of coverage at this time is reactor accidents.
The reference goal statement also addresses the following issues about which no wide consensus had emerged.
On some of these issues there was 4
controversy at the first workshop.
On others, the discussion did not reach the issue or did not oroceed to'a conclusive result.
I'll skim the surface of 10 substantive issues in this category.
I 1.
What should be the qualitative goals? The goals stated, in Section VI of the Discussion Paper, should be read in conjunction with the principles guiding development of a safety goal, in Section IV. The qualitative goals themselves are stated in terms of limiting risk to any one person and in keeping aggregated social risk adequately low, in relation to
. other risks.
Such qualitative goals tend to have something very general and commonplace in their articulation -- perhaps necessarily'so, perhaps properly so. We would, however, welcome improvements that you may suggest in the draft qualitative goals, to help make them as sound and as useful as possible.
2.
What should be the extent of quantification? Our draft's approach is simple, limiting itself to three elements.
Is more needed? Can more be managed in a practical schene of implementation? Should some of the greater elaboration of the ACRS proposal be adopted?
3.
What quantitative elements should be specified? At what values? The reference goal specifies an individual exposure limit, a social impact per unit of energy production, and a core-melt probability. Should other parameters be specified, in. addition to or instead of these, for example, from the ACRS proposal? What are Subcomittee members' views about the numerica1 values for the quantitative goals?
4.
What qualitative elements should be specified? The reference goal statement includes rudimentary specifications for individual and social protection and discusses some qualitative principles that have guided development of the goals.
What should be the implementation process? How should the goals be used? How should uncertainties be dealt with? How should any burden of proef be allocated? The proposal in the Discussion Paper would
i.
. c implement the safety goals with a very light hand. The goals would influence rulemaking and other standard-setting with respect to major safety issues.
They would also influence decisions on issues not covered by regulations -- backfitting of existing plants, exemptions from rules, enforcement actions.
But the key is analysis and disclosure
-- disclosure of the bases and uncertainties as well as results of analysis -- and consideration of this information as one f actor among others in the major safety decisions involved. The manner and extent of the.use of that information would be determined by the nature of the issue.
Uncertainties would be respected; prescription and proof would generally not be involved.
6.
Should the scope of the safety goals be extended to aspects other than accidents and to facilities other than nuclear power plants? 'The Discussion Paper suggests deferment of any such extension.
7.
How should equities be taken into account, particularly (a) equities in the distribution of risks and benefits and (b) genetic risks? The Discussion Paper acknowledges that it is not possible to devise a system of regulation whereby the distribution of risks and benefits is always equitable to each individual.
However, if the risks are small enough, there should be correspondingly " educed concern by individuals regarding the balance of risks and benefits.
Genetic risk is not specified as suun.
Rather the reference goal structure depends for validity on the thesis that regulatory actions with respect to reactor accidents would not be sensitive to inclusion or exclusion of a genetic risk specification.
. 8.
Should there be safety goals beyond a specified level of minimum adequacy?
The Discussion Paper endorses a concept of 'as low as reasonably a~chievable,"
without specifying a numerical formula for cost-based cutoff in cost-risk trade-offs.
In a broader sense, the issue of safety improvements beyond minimum adequacy is left somewhat open by the non-prescriptive nature of the implementation process.
9.
Should there be a special emphasis on high-consequence accidents even at low probability of occurrence? What form should that emphasis take?
The Discussion Paper expresses interest in considering incorporation of some special approach. The first workshop produced criticism of the concept of an exponent on the consequence factor (at 1.2 or any other value), but arrived at no widely supported specific alternative. We would welcome suggestions. The paper notes remote siting and measures to mitigate accident consequences as specific actions to reduce catastrophic potential.
- 10. What should be the role of economic considerations? The Discussion Paper provides proposed guidance on recognition of economic factors in ALARA (the "as low as reasonably achievable" concept), in decisions concerr.ing possible backfitting of existing plants to new requirements, and in timing of corrective actions when required in an operating plant and the severity of the problem is not such as to demand immediate action.
. I hope that Subcommittee members will give us the benefit of their views on a number of these issues. An at least equally important desired result of this review from our standpoint is identification and guidance on any additional issues that should be considered further at this time.
v 4
~
e w
d'
SECOND SAFETY G0AL WORKSHOP: HIGHLIGHTS OF WRAP-UP SESSION I shall now turn to the Second Safety Goal Workshop, particularly its final, wrap-up session. The results of that session. offer perspectives on the issues that may be of interest to the Subcommittee.
The wrap-up session was devoted to participants' statements of what they viewed as the most significant judgments concerning the OPE Discussion Paper and changes needed in its reference safety goal statement.
I shall present some highlights of the discussion, noting particularly views that appecred tc have vide support.
Complete consensus on the issues involved was, however, not sought--or found.
Summary A widely-shared general theme appeared to emerge from various statements of views by participants:
The highest safety goals should.be qualitative goals, devised as aspirations a.
to be aimed at, rather than as requirements geared to what is currently achievable.
b.
Quantitative goals should be stated as standards against which performance could somehow be me:sured.
I c.
Safety goals should include distinct elements expressed in language and specificity appropriate to the various users and uses (articulation of i
public concerns and values, cuidance and decision rules for staff).
d.
Introduction of safety goals should be a gradual'and circumspect phase-in, i
rather thar an abrupt' displacement of current practices.
'btably, quanti-I tative goals ~should not be allowed to displace the defense-in-depth concept.
t
-B-The safety goals should address avoidance and mitigation of catastrophic e.
accidents (i.e., high-consequence, low-probability accidents).
1.
Qualitative Goals as Aspirations A number of participants advocated making a clear distinction between (a) broad and symbolic goals that set the aim and thrust of safety regulation and (b) narrower, operational standards that codify specific decision rules.
The general goals should be qualitat.ive and need not be--in the view of some, should not be--constrained by what can currently be demonstrated or even by what is believed to be currently attainable. They should be. aiming points, perhaps distant, but ser b sly pursued.
Specific cafety policies and standards should be related to these goals and adjusted from time to time, en the basis of periodic review of performance in relation to such l
primary goals.
l The qualitative goals as stated in the Discussion Paper were viewed as 1
needing considerable expansion.
l 2.
Quantitative Goals as Standards l
For reactor licensing by NRC, goals are needed that are stated as standards, in specific technical, including quantitative, terms. These should be consistent; be coherent with the primary. qualitative goals; admit com-pliance. verification, at least in retrospect; and reflect balance of a trio of components--accident probabilities, accident consequences, and social values.
s.
.g-The quantitative goals should be stated in.a manner that is integral with the implementation process, i.e., the relationship between goals and processes to achieve them should be stated.
They should be related to the qualitative goals by a clear rationale. The regulatory decisions to which the quantitative standards are to be applied should be identified. Decision rules should be spelled out. These should include procedures for use of probabilistic risk assessment (PRA).
Reservations were expressed about quantitative safety goals in several respects:
(a) in view of problems and uncertainties in PRA, there may be too much room for error and abuse in implementation; (b) quantitative standards tend to drive out qualitative ones; (c) quantitative safety goals may weaken the defense-in-depth concept; and (d) we may be far away from having proposed the right quantitative goals.
One participant noted that use of PRA does not require probabilistic goals.
3.
Gradual Introduction, Dynamic Evolution Several participants stressed the need for cautious, gradual, tentative introduction of safety goals.
After initial experience with the goals applied on a trial basis, performance against the goals and utility of the goals should be evaluated. Adjustments in the goals, perhaps especially in the quantitative goals, and in the method of implementation should be made as warranted.
l
The ct.utious attitudes about introduction of safety goals stemmed from several considerations: uncertainties in technical knowledge underlying PRA, problems in the quality of PRA calculations, social and equity issues not taken into account or only addressed in a tentative or controversial manner, uncertainty as to precise effect of particular goal elements on regulatory actions, possibilities of counterproductive effect through excessive displacement of established safety practices, and a general need to learn from new operational data and from experience with the safety goals.
There should be.a conscious effort to keep what is valuable in current practices.
Safety should be enhanced over time.
PRA practices should be improved over time.
/
4.
Problems of Comparison Some participants cautioned against simplistic comparison of technologies in seeking a suitable goal for nuclear power plants. The risks are quali-tatively different.
Uncertainties are involved in estimating the risks of both nuclear and alternative technologies. The risks of old coal-fired electric generating plants should not be assumed to be acceptable today, nor should today's risks be assumed to be acceptable in the future.
A suggestion was made that the relevant comparison is with existing nuclear plants: whether the safety goals and standards should be set as equal to, higher than, or lower than achieved by existing nuclear plants.
r,
~ 5.
Eauity and Social Values as Matters of Policy The view was 1xpressed that the safety goals should articulate public views and concerns, rather than take an apparently patronizing posture of " con-tributing to better understanding by the public...".
Policy principles should be stated as agency decisions ("We come down here, for these reasons"), rather that asserted as "self-evident," without stated basis.
Where social or equity issues on which cpinion is sharply divided are _
avoided, or addressed incompletely, there should be acknowledgment of the incompleteness and provision to revisit the issues. The issues of equity of distribution of risks and benefits and of special consideration of high-conseqtrace accidents were mentioned as examples.
4 I
e 9
l l
i y
c..
.,m.,,,
m.,
, -