ML20003F824
| ML20003F824 | |
| Person / Time | |
|---|---|
| Issue date: | 04/14/1981 |
| From: | Sege G NRC OFFICE OF POLICY EVALUATIONS (OPE) |
| To: | |
| Shared Package | |
| ML20003F822 | List: |
| References | |
| FRN-45FR71023, RULE-PR-50 NUDOCS 8104230658 | |
| Download: ML20003F824 (9) | |
Text
EL 0 06 bC tCf O
gs me 9/o25 DEVELOPING Ait NRC POLICY STATEMENT ON SAFETY GOALS Geerge Sege Office of Pclicy Evaluation p
e U.S. Nuclear Regulatory Comission
+
Presented at cecem g
The Atomic Industrial Forum's estme APR M
. :.,16198t' i,r Licensing Workshop
-1 New Orleans, Louisiana tv
,. Se tenfmy
\\
---%r;&g" p April 14, 1981 Eg.g s\\d I am pleased to have the opportunity to make a progress report on the Nuclear Regulatory Comission's safety-goal project to the AIF Licensing Workshop.
Last October the Comission instituted a project to define more clearly the '.evel of protection of the public health and safety that it believes is adequate.
In accordance with the project plan, in late December 1980, the NRC Office of Policy Evaluation submitted for the Comission's consideration a pre-liminary policy paper, along with a proposed preliminary Comission state-mant, and a companion paper, providing a more detailed supporting discussion.
The Comission has recently completed consideration and revision of these materials. On March 26 the Comission published in the Federal Register, for public coment, a statement of preliminary policy consioerations in developing a safety goal, and authorized publication of the revised sup-porting document, NUREG-0764, "Toward a Safety Goal:
Discussion of Preliminary Policy Considerations."
l 810423 0 ~6,Q I
2-The coments received will constitute a part of the considerations leading to preparation of the eventual safety goal.
In addition to consideration-of the written coments, workshops and public meetings are being held. A subsequent paper will be preoared by the Office of Policy Evaluation and is scheduled to be submitted to the Comission for its consideration in August 1981. After consideration of that paper, the Commission intends to seek additional public coment on the proposed policy statement. Those coments will also be considered in the preparation of the final safety goal itself.
The purpose of the Commission's preliminary statement is to elicit coments by indicating the kinds of considerations which may enter into an articulation of the Agency's safety goal.
The Advisory Comittee on Reactor Safeguards has prepared "An Approach to Quantitative Safety Goals for Nuclear Power Plants" (NUREG-0739). Comenters are urged to use that document as one example of a concrete application I
l of the concepts discussed in the Comission's statement. The Comission l
l has at this time not formed any views as to the merits of the approach described by the ACRS or any other single policy. The ACRS has characterized its proposal as intended to " serve as one focus for discussion."
It is the only specific proposal that has so f ar been formally submitted to the l
Comission, and is detailed and accompanied by background information.
The scope of early policy development efforts has focused primarily on radiological accidents associated with power reactors.
It is, however,
3-intended to include considerations of sabotage and other-external events and to seek adaptation of the safety goal to other facilities _and operations where the need is evident.
For NRC, the safety goal should provide a means of testing safety regu-lations for necessity and adequacy.
It should provide a systematic basis for judging whether a nuclear power plant is safe enough, perhaps indirectly, via regulations.
For industry, it could guide design and operational practices.
For the public, it should provide a criterion for judging-whether NRC is doing its job.
The first of the two workshops planned as part of the safety-goal project was held the week before last in Palo Alto, California.
I'll tell you about it -- in terms of a few highlights. My remarks will be tentative and preliminary.
Brookhaven National Laboratory is preparing a report on the workshop, which should be available within two months.
j Our Session Chairman, Ed O'Donnell participated in the workshop.
He may wish to add some observations of his own when we get to the time allotted for discussion following my remarks.
ae g, - - -.
9y-m.------,,,-
y
-,-.,.---e sa y
eue-- - --
y w--
-is
t 4-5 The workshop was de, signed to illuminate the important issues of safety goal formulation, including both quantitative and qualitative elements and economic, ethical, social, and political issues as well as technical considerations.
It was a discussion workshop, involving invited know-ledgeable persons representing a broad range of viewpoints, drawn from -
technical and social disciplines, from industry, public interest groups, universities, and elsewhere. During most of the meeting, the participants were divided into three separate discussion panels: Panel A, Quantitative Safety Goal; B, Qualitative Safety Gol; and C, Economic, Ethical and Sociopolitical Issues. At the plenary sessions, at midenurse and during the last half day, interim and final reports of the Panels became the subject of general discussion.
Some of the views developed at the workshop commanded a wide measure of consensus; other elicited vigorous debate.
Here are some views that, by and large, appeared to be widely shared:
. There should be improved quantitative safety gorls, in order to enhance the protection of the public -- not directly, but U. rough making the regulatory process less capricious and more objective.
Goals must be clearly stated so as to permit a reasor.ed political test.
Improved quantitative goals can force quantitative analysis of system and subsystem standards. They can provide a basis for deciding what measures are important for safety and what measures are not so significant.
4-M o
-e-am e r---rrg--r-
,,-p-trw--
ew
+7 c-w--r---*
T--
e=w--
-v-
5-The revised goals should be constructed around a qualitative statement, related to safety of the people.
This should be supplemented by quantitative requirements for achieving the qualitative goals.
Some-where, but not necessarily as part of the goals themselves, there must be highly specific instructions on where and how to use these quantitative limits in a reasonably unambiguous way.
The revised goals should also include quantitative components related to individuals more highly at risk and to tfie aggregate risk to society.
l They should also include a component related to financial impact on society.
For the present, the licensing process should continue to be deterministic with the detenninistic requirements justified through demonstration that they assure meeting the safety goals. Both subsystem and whole plant analyses can contribute to this, but it is recognized that, in the present state of the art, a large element of judgment will still be involved, as well as recourse to operating experience. The one exception to the deterministic rule should be that an applicant for a license should be free to propose a new system or subsystem, and to prove, by analysis, that it is a better way to achieve the goal.
The quantitative aspects of new goals will require political consensus, i
but development of the techniques for calculation requires much more technical work.
In particular, there may be some subgoals fcr which the Calculation is now beyond the state of the art.
e=
h u-6 Weh' w-,-
y
- -rw-w we-yrw v w-
,r we-
==ar--yr-,g--yg---e&
+w wy-y,_er- + - - - - -er--
yo--
-se---
-*w,*--%v-w gr
-+-+-----------=--,em
-w-~a r
TT*-
4 There is a conflict for resolution here between desirability and complete feasibility.
Of course, it would be better if compliance with all safety goals and objectives could be determined with a high degree of assurance as to validity of the conclusion. The need for an element of judgment is not unique to this human endeavor, though it is as acceptable here as it usually is.
. The new goals should be dynamic, to respond to progress in technology, but "grandfathering" plants already approved should be normal policy, in the absence of overriding safety considerations to the contrary.
~
The dynamic character is especially needed because improved quantitative goals will probably find their principal application to plants that will not come into existence and operation for scre than a decade.
They must be adaptable to conditions that may be important then. Political consensus, and public acceptance, are essential for the end product, but the responsibility of the NRC is to protect the public and not to satisfy it.
These are not always compatible, nor always conflicting.
Satisfying the public is a job for other elements of government whose function is meant to be more political than the NRC should be.
. The question of the mode of use of goals is very imocrtant, perhaps at least as important as the goals themselves. The = ode of use deserves considerable further study.
. While there are numerous, important advantages of quantifying safety go'als, the NRC must recognize that there are imcortant disadvantages.
h
_,.-..-~.-
l 7-Virtually all of these disadvantages can be overcoxe by recognizing -
them and giving them careful attention.
. Accident probabilities such as 10-7 are statements on the border of what is knowable. Current estimates must be qualified carefully as theoretical with careful detailing of the data, model, and methods of arrivng at them. The principal method for reducing confidence intervals about these estimates is a central collection and analysis of data about individual " micro" component failures and other difficulties. An explicit feedback process is necessary, taking current experience, analyzing it, improving models, and changing operating instructions and design.
. Qualitative and quantitative safety goals are complements, not substitutes.
Congress inevitably stipulates qualitative goals. The NRC must specify qualitative, quantitative and process goals. These will serve to enhance safety by providing a coherent framework for regulation.
. Implementation of qualitative safety goals requires coordinating quanti-tative, qualitative, and process goals at various levels. The goals make no sense without procedures for verification and for ensuring that the system operates as modeled.
In particular, the NRC must oversee detailed specification of " good practice" that translates safety goals to the operatier.i level.
y.
w y
,--.-w 7
,+-e---
m -
m--
, -. For the NRC to establish acceptable safety goals, it must identify and
^
involve all those who have a stake in the outcome. This is difficult and even cumbersome but necessary. The NRC must search for a " felicitous" process for gaining information and arriving at decisions-without paralyzing action or absorbing inordinate resources.
Information given to these groups must be arranged to facilitate their understanding, but it must be complete, not patronizing, and not sugar coated.
The workshop also called attention to some further special problems that will need to be taken into account and some issues on which contrasting views were expressed.
. There was a division of views on whether safety goals should be stricter for nuclear power than for other power sources. Related to this was some controversy on whether nuclear or competing technologies involved greater uncertainties as to risks.
. There was controversy about whether a safety goal should be limited to l
establishing an adequate level of safety or call, in addition, for improvement beyond that level when justified on a benefit-cost basis, such as the as-low-as-reasonably-achievable (ALARA) concept.
l
. There was call for explicit treatment of genetic risk.
It was argued that early and delayed death are not good surrogates for genetic effects.
l l
l
_g.
. Questions were raised about whether institutional arrangements can maintain standards oemanded of people as the scale of the industry becomes larger; also about the relation of safety goals to an increased dependence of the country on the nuclear part of electric generation and increasing vulnerability to shutdowns as an increasing fraction of.
the electric power demand is met by nuclear power plants.
Altogether, the workshop was very helpful in illuminating the chief issues and pointing to directions for narrowing the safety-goal options for further consideration. We value that help.
Thank you.
1
- -.. -,