ML25168A025

01-10-83 ACRS Report on SECY-82-1B

ML25168A025
Person / Time
Issue date:	01/10/1983
From:	Ray J Advisory Committee on Reactor Safeguards
To:	Palladino N NRC/Chairman
References
SECY-82-1B
Download: ML25168A025 (1)

Text

UNITED STATES NUCLEAR REGULATORY COMMISSION ADVISORY COMMITTEE ON REACTOR SAFEGUARDS WASHINGTON. D. C. 20555 January l 0, 1983 Honorable Nunzio J. Palladino Chairman U.S. Nuclear Regulatory Commission Washington, D.C.

20555

SUBJECT:

ACRS REPORT ON SECY-82-1B: PROPOSED COMMISSION POLICY STATEMENT ON SEVERE ACCIDENTS AND RELATED VIEWS ON NUCLEAR REACTOR REGULATION During its 273rd ACRS meeting, January 6-8, 1983, the Advisory Committee on Reactor Safeguards discussed SECY-82-1B, "Proposed Commission Policy State-ment on Severe Accidents and Related Views on Nuclear Reactor Regulation,"

dated November 24, 1982.

We also considered the memorandum of October 25, 1982, Samuel J. Chilk, SECY, to William J. Dircks, EDO, "Staff Requirements

- Discussion of Severe Accidents - Policy Statement and Research Plan.... 11 In our review, we had the benefit of a Subcommittee meeting held on Decem-ber 21, 1982.

The Committee has commented on earlier drafts of this SECY paper in reports dated February 8, 1982 and September 14, 1982.

As a result of the October 25, 1982 memorandum, s. J. Chilk to W. J. Dircks, the ACRS arranged a series of three Subcommittee meetings to discuss the proposed NRC research program in support of a regulatory approach for deal-

ng with severe accidents as described in "Nuclear Plant Severe Accident Research Plan," NUREG-0900.

The first of these meetings was held on Decem-ber 21, 1982.

You may recall that in our report of August 18, 1982 on NUREG-0900 and in our report of September 14, 1982 on SECY-82-lA, "Proposed Commission Policy Statement on Severe Accidents and Related Views on Nuclear Reactor Regulation, 11 dated July 16, 1982, we expressed a number of concerns about what we considered to be the lack of a coherent and workable approach to dealing with severe accidents in the licensing of new plants and in the regulation of existing plants.

We concluded that we could not judge the appropriateness or the adequacy of the research program without having examined one or several feasible approaches to which a research program could be related.

With these comments in mind, we requested the NRC Staff to present, during the Subcommittee meeting of December 21, 1982, whatever additional informa-tion had been developed on approaches to deal ~,ith severe accidents.

We were surprised when we were informed that SECY-82-1B was, in the Staff's view, what the Commission is likely to adopt as its policy.

The substance of SECY-82-1B is, so far as we can see, little different from that of SECY-82-lA.

1961

Honorable N. January 10, 1983 In its statement of purpose, the policy statement is said to have been "revised to reflect Commission and ACRS comments." In our opinion, the policy statement of SECY-82-1B does not reflect the comments we have made in previous reports to the Commission.

Our convnents on the various drafts of SECY-82-1 are summarized below to-gether with some further recommendations.

Additional details can be found in the attached excerpts from several previous ACRS reports.

1.

As we understand the proposed policy, judgment as to whether an ap-plicant for a license has dealt appropriately with severe accidents will depend heavily on the results of probabilistic risk assessment (PRA).

Decisions will involve comparison of the results of the PRA with the numerical guidelines suggested in Revision l of "Safety Goals for Nuclear Power Pl ants, 11 NUREG-0880.

There is general agreement that large uncertainties exist in our ability to predict both tne proba-bilities and the consequences of severe accidents.

Furthermore, there is no generally agreed upon method for comparing the results of a PRA with the guidelines given in NUREG-0880.

2.

For existing plants, it appears that some as yet undefined set of plant specific and generic PRAs will be used to draw generic conclu-sions about groups of plants.

An effort will then be made to draw conclusions about specific plants.

The process to be used is not yet defined, nor is it clear what methods will be used to define it.

In-deed, we observe that experience gained with PRAs suggests that it may be inappropriate to use generic results in the evaluation of individual plants.

3.

No specific guidance is given as to an appropriate balance between prevention and mitigation of severe accidents.

Except for some rather general comments about the need to explore the behavior of containment systems, and some equally general comments about filtered vented con-tainment systems and core retention devices, mitigation is largely ignored.

It appears that in principle, under the proposed policy, only an appeal to prudent engineering practice or the use of ALARA in risk reduction could be used to generate containment specifications, for example, and requirements for other mitigation systems important to public health and safety.

4.

We have in several reports expressed reservations about a strong depend-ence on PRA alone in decisions dealing with severe accidents.

We note, however, that the Commission policy as expressed in SECY-82-1B would use PRA as a principal criterion in detecting and correcting weaknesses in design.

We recommend that before issuing a policy statement on severe accidents, the Commission give consideration to tne possibility of in-cluding more specific directions for systems or approaches for dealing with severe accidents.

As examples we suggest:

1962

Honorable N. January 10, 1983 (a)

A statement that effort will be made to specify the performance of containment systems including subsystems for heat removal.

It may not be feasible to do this at present, but an effort to do so can guide research that may be needed to detenni ne if it is possible.

(b) Specifying improved performance for decay heat removal systems.

(c}

Giving direction to a licensee that a plant design must include specific consideration of features to decrease the probability of damage from sabotage.

It appears to us that because of the close relationship that must exist among a safety goal, a policy on severe accidents, and a siting policy, a much more integrated approach is needed.

We recognize the considerable effort that has gone into the various drafts of SECY-82-1 and associated documents.

We understand that the task is difficult.

We nevertheless consider SECY-82-18 to be seriously flawed.

Sincerely,

Attachment:

List of Relevant Comments from Previous ACRS Reports References:

1.

SECY-82-18 from W. J. Dircks, Executive Director for Operations, to NRC Commissioners,

Subject:

Proposed Commission Policy Statement on Severe Accidents and Related Views on Nuclear Reactor Regulation, dated November 24, 1982.

2.

SECY-82-203A, from W. J. Dircks, Executive Director for Operations, to NRC Commissioners,

Subject:

Revisions to Nuclear Plant Severe Accident Research Plan, NUREG-0900 (Draft}, dated August 30, 1982.

1963

Attachment to January 10, 1983 ACRS Report on SECY-82-1B - List of Relevant Comments from Previous ACRS Reports

• "We believe that, before embarking on the course proposed for future CPs in SECY 82-lA, a concerted effort should be made by the NRC Staff and the ACRS to develop policy guidance on as many of the relevant safety issues as are tractable, and to propose an alternate approach to the Commission in which such policy guidance is provided to appli-cants for future standard plant designs." {Ref. 5, p. 2)

• "With regard to existing plants, we believe it would be productive for the NRG Staff to draft alternate positions on the most significant safety issues and to establish what would be needed in order to eval-uate the alternatives."

(Ref. 5, p. 3)

• "Neither the original nor the revised version of NUREG-0900 contains a delineation of an approach for dealing with severe accidents. This is needed to judge the appropriateness of the proposed research program.

We continue to urge that the work necessary to pro vi de one or more approaches be carried out.

We 1 ook for requirements that might be placed on components or systems required to deal with severe acci-dents, description of what is now known about these, specifications of what, if any, information is required to describe system performance with the necessary accuracy, some indication of whether the informa-tion can be obtained from research in the time and with the resources available, and what research is planned to obtain the needed informa-t i on. 11 (Ref. 3, p. 1 )

11As an example, we note that, in the draft Implementation Plan for Safety Goals (July 16, 1982) provided to us, the NRC Staff concludes that it is not now feasible to specify the performance of containment systems.

The NRG Staff further expressed an opinion that the informa-tion and approach needed for such a specification should be developed.

We, therefore, looked at NUREG-0900 for a description of what new information is needed to specify performance of the various kinds of containments and containment systems now in use or proposed. Although there a re elements of the program that could certainly contribute to more accurate specification of containment performance, we find no systematic descriptions of what information is needed or what part of the proposed program is designed to provide the information." (Ref. 3,

p. 2) 11 We recommend that alternate containment performance criteria be developed and evaluated for existing nuclear power plants as part of the trial implementation program.

A separate set of alternate trial containment performance criteria should be developed and evaluated during the trial period for plants yet to be designed."

(Ref. 4,

p. 2) 1964

Attachment * "With regard to future plants, we believe that the NRC should examine and evaluate the safety-related changes now proposed or underway for LWRs in countries like France, the Federal Republic of Germany, Japan, Sweden, and the United Kingdom before arriving at its own judgment on what is appropriate for the U.S.

For existing nuclear power plants, it is premature to assume that the available PRAs provide a generic basis for decision-making. On the contrary, despite their uncertainties, the PRAs indicate the existence of important plant-specific differences which need to be factored into the formu-lation of policy.

Again, the specific backfitting approaches cur-rently underway or contemplated for LWRs in other countries should be examined and evaluated for their relevance to U.S. policy." (Ref. 5,

p. 3)

• "In our recent reports specific attention was called to the need for organizing the research under this Decision Unit to answer questions likely to arise in connection with the ColJlllission's stated intention to modify the licensing process to take specific account of acci-dents more serious than those generally identified as Design Basis Accidents." (Ref. 2, p. 9)

"However, there is still a lack of definition of even one approach to deal with the severe accident issue.

Considering the difficulty of the problem, effort should probably be made to define several alternatives."

(Ref. 2, p. 9)

• "We find that the NRC program, as proposed, is not responsive to

[previous] recommendations [that funding t>e reallocated to provide the information needed for the severe accident rulemaking].

The programs *** should be restructured so that the primary priority is to provide the information needed for decision-making concerning features to mitigate the consequences of accidents involving severe core damage or core melt, for reactors in operation and under con-struction and for reactors yet to be designed.

This would allow the elimination of a substantial portion of the longer-term experimental and code development work." (Ref. 1, p. 9)

• "A focused priority effort is needed with respect to risk contrib-utors such as seismic events, design errors, operator errors of commission, sabotage, and systems interactions to provide a metho-dology suitable for incorporation into PRAs on a trial basis or to identify and evaluate sources of uncertainty which make this im-practical and to suggest regulatory approaches in light of these uncertainties." (Ref. 2, pp. 8-9)

• "Insofar as feasible, all accident initiators and risk contributors (other than sabotage) should be included in PRAs and in benefit/cost analysis.

If the uncertainties are su~h as to make a meaningful 1965

Attachment quantification for some initiator or contributor impossible, this should be documented in sufficient detail and an allocation of risk to this contributor justified." (Ref. 4, p. 3)

• "We believe that, in view of the continuing uncertainties to be expected in the art of PRA and a continuing inability to satis-factorily treat all initiators and other contributors to core melt frequency, and in view of the potentially very large differences in release magnitudes among different core melt accidents, containment performance design objectives are needed and should be developed expeditiously." (Ref. 6, p. 5)

Related ACRS Reports:

1.

"Review and Evaluation of the Nuclear Regulatory Commisson, Safety Research Program for Fiscal Year 1983, 11 NUREG-0864, dated February 1982

2.

11Convnents on the NRC Safety Research Program Budget for Fiscal Years

, 1984 and 1985, 11 NUREG-0875, dated July 1982

3.

11ACRS Comments on Nuclear Plant Severe Accident Research Plan," NUREG-0900 (Draft), dated August 18, 1982

4.

11ACRS Report on the Draft Action Plan for Implementing the Commission's Proposed Safety Goals for Nuclear Power Plants," dated September 15, 1982

5.

11ACRS Report on SECY 82-lA:

Proposed Commission Policy Statement on Severe Accidents and Related Views on Nuclear Reactor Regulation," dated September 14, 1982

6.

11ACRS Comments on the NRC Staff Questions to the Commission Concerning the Policy Statement on Safety Goals for Nuclear Power Plants," dated September 15, 1982 1966