ML20058L971

Summarizes & Discusses Public Comments Re Proposed Rule 57FR44513 on Alwr Severe Accident Performance.Recommends Commission Endorse Plans to Continue Discussions W/Acrs Re Generic Rulemaking

ML20058L971
Person / Time
Issue date:	08/18/1993
From:	Taylor J NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO)
To:
References
FRN-57FR44513 SECY-93-226, NUDOCS 9309030352
Download: ML20058L971 (87)

Text

_ _ _ _ _ _ _. _ _ _ _ _ _

---

g g o T g s s,,,,,,,,,,,,,,,,

5 l

RELEASEDTOTHE PDR

[

/b f** "4 A ese ini 9

f: gf ;

..........'....... ps :

s(v /

t

• sse*

POLICY ISSUE August 18, 1993 SECY-93-226 (Notation Vote)

For:

The Commissioners From:

James M. Taylor Executive Director for Operations Sub.iect:

PUBLIC COMMENTS ON 57 FR 44513 - PROPOSED RULE ON ALWR SEVERE ACCIDENT PERFORMANCE

Purpose:

To present a summary and discussion of public coments received regarding the proposed rule on advanced light water reactor (ALWR) severe accident performance and to provide staff's recomendations regarding the policy issues raised in these comments.

Sumary:

In a Federal Register Notice (Vol. 57, No.188) dated September 28, 1992, the NRC issued an advance notice of proposed rulemaking (ANPRM) on the subject of l

ALWR severe accident design criteria. Comments as well as responses to fifteen specific questior.s regarding various aspects of a possible generic rulemaking were invited. Eleven parties responded, and this paper summarizes those comments.

Of the eleven responses, three were in favor of generic rulemaking and provided suggestions on how a rulemaking would best be accomplished.

Seven responses were not in favor of generic rulemaking. One individual did not comment either way in support of or opposition to generic rulemaking.

(A table is provided in Enclosure 1 to give a summary of the responses to those questions that could be answered by "yes" or "no".)

NOTE:

TO BE MADE PUBLICLY AVAILABLE WHEN THE FINAL SRM IS MADE AVAILABLE 300075 1

l

Contact:

Brad Hardin

{

492-3733 l

3/cA e j, y

,:p.

. )-

v

+

m-

,t b

E The Commissioners 2

It is the staff recommendation to continue to develop a draft generic rule but to defer a decision to issue the rule until after the Final Safety Evaluation Reports are issued for the GE ABWR and CE System 80+ designs.

Backaround:

The advanced notice of proposed rulemaking is a reflection of the extensive research over the past two decades regardi~

" vere accidents and methods for preventing and mitigating them. The ANPRM addresses licensing requirements for protection against severe accidents in advanced light water reactors (ALWRs).

It is not intended to apply to existing plants.

This rulemaking would codify the already existing Commission guidance on severe accident issues that has resulted from reviews of the EPRI ALWR Requirements Document, the GE ABWR and the CE System 80+ designs.

This guidance is presently documented in the SRM based on SECY-90-016 and in SECY-93-087.

It is expected that severe accident licensing issues will primarily be resolved for the ABWR and System 80+ designs through the individual design certification rulemakings for these two evolutionary designs.

However, the staff is considering a procedure wherein if generic rules are put in place sufficiently early to facilitate (through reference) the design certification process for reactor designs licensed after the evolutionary designs, such generic rules or parts of the rules, could possibly be utilized. The potential use of this approach is noted in SECY-91-262, " Resolution of Selected Technical and Severe Accident Issues for Evolutionary Light Water Reactor (LWR) Designs," August 16, 1991.

The ANPRM identified the major severe accident issues that would be addressed in such a rule, described three alternative ways that the issues could be addressed and provided fifteen questions for which feedback was requested.

The ANPRM was discussed in a Commission paper (SECY-92-292, " Advance Notice of Proposed Rulemaking On Severe Accident Plant Performance Criteria For Future LWRs," August 21, 1992).

Listed below are the organizations that responded to the ANPPM Federal Reaister notice with written comments:

1.

Department of Energy (DOE), E.C. Brolin, Acting Deputy Secretary 2.

Tennessee Valley Authority (TVA), Mark J. Burzynski 3.

Florida Power Corporation (FPC), Paul V. Fleming 4.

Entergy Operations, James J. Fisicaro, Director 5.

Individual, John S. Fuoto, PE 6.

Ohio Citizens for Responsible Energy (OCRE), Susan L. Hiatt, Director 7.

AECL Technologies, A.D. Hink, Vice President 8.

ALWR Steering Committee, E.E. Kintner, Chairman 9.

Westinghouse Electri: Company, N.J. Liparulo, Manager i

i

+

The Commissioners 3

10. NUMARC, William H. Raisin, Vice President

11. Winston & Strawn (representing Niagara Mohawk, et. al), Mark J.

Wetterhahn, Esq.

Discussion Of the eleven responses, three were in favor of a generic rulemaking and offered suggestions for such a rulemaking.

Seven responses did not favor a generic rulemaking, and one commentor did not state either his support or opposition. The negative responses mainly included detailed arguments for their position and did not offer responses to the individual questions. to this paper gives the major points raised in the comments for each question and includes a summary table. includes copies of the submitted comments, and Enclosure 4 is the Federal Reaister notice for the ANPRM.

Arauments for Generic Rulemakina Main arguments given by those in favor of the generic rulemaking included:

(a) there would be value in providing a consistent regulatory basis for certification and licensing of future plants; (b) policy statements do not provide a substantial basis for licensing and, therefore, rules are needed; (c) generic rules that apply to all advanced designs should be promulgated to resolve severe accident issues prior to commencement of design certification proceedings to the extent practicable (and regulatory guides should be developed to provide more complete guidance on acceptable rule implementation for specific reactor designs); and (d) the importance of severe accidents to public risk requires the addition of severe accident regulatory requirements if public acceptance of future plants in the United States is to be assured.

Arouments Aaainst Generic Rulemakina Main arguments given by those opposed to generic rulemaking included:

(a) generic rulemaking may be appropriate at some time in the future but could impact negatively on the design certificatior Ethedules for the four evolutionary and passive reactor designs if a generic rulemaking is ongoing concurrent with the design certification rulemakings; (b) generic rulemaking is not necessary in light of the established design certification process provided by 10 CFR 52; (c) generic rulemaking could, in fact, deter from regulatory stability as opposed to its intent; and (d) resolution of severe accident issues is best carried out through the staff's review of the EPRI ALWR Requirements Document.

Evaluation of the Comments This discussion will be presented in two parts: (1) the value perceived for a generic rulemaking and its timing, and (2) if generic rulemaking was to be implemented, which alternatives should be pursued?

The Commissioners 4

In the responses to the ANPRM, the opinions regarding the value df a generic rulemaking on severe accidents expressed three different viewpoints: (1) those in favor of rulemaking now, (2) those possibly supporting rulemaking but not until after design certification is completed for the two evolutionary and two passive reactor designs, and (3) those who do not appear to favor generic rulemaking at any time.

The staff agrees with many of the arguments presented by those who are not in favor of proceeding to implement a generic rule at this time. However, the staff believes that there are overriding arguments for continuing with the development of a draft rule as will be discussed further below.

It is agreed that it is not necessary to have a generic rule to license the designs currently under review. The design certification process provided under 10 CFR 52 is adequate. Severe accident concerns may be dealt with individually for each design during the respective rulemakings. Any severe accident requirements that the Commission determines are important for those designs could be implemented through the individual rulemakings as well as by a generic rulemaking.

Furthermore, the efficiency in licensing and time saved by having a generic rule would not appear to be particularly significant given current circumstances. The number of individual designs to be processed through design certification in the near future are not foreseen to be large in number (four designs as stated).

If an applicable generic rule was available in time for reference in the passive reactor design applications, some duplication of licensing steps could be avoided.

But it is likely that the documentation from the earlier design certification of the evolutionary designs would also provide some referenceable technical bases for severe accident issues for the passive design certifications.

A concern expressed by those not in favor of generic rulemaking is that a protracted rulemaking could delay the design certification schedules for any designs in the process of being licensed. Since the NRC has not yet completed licensing a design under 10 CFR 52, the potential impact of a generic rulemaking conducted in parallel with an individual design certification rulemaking is not clear. An important point to be made, however, is that the critical technical content of the two different rule approaches regarding acceptable means for preventing and mitigating severe accidents would be essentially the same. Critical technical content, as used here, is that content that would be used in determining acceptability of a design. The issues to be addressed in any rule have been known, evaluated and discussed at great length over the past ten years. Acceptable approaches for dealing with these issues were identified in SECY-90-016 and expanded on in SECY-93-087, and it is unlikely that new issues would arise in a generic rulemaking that were not includc1 in an earlier design certification rulemaking.

For this reason, the staff believes that a generic rulemaking would in effect be a codification of acceptable approaches for addressing severe accident issues where those approaches would be the very same ones to be included in the i

individual design certification rulemakings.

For this reason, the staff believes that the finalization of a proposed rule should proceed following the i

I l

i

Te

,ammissioners 5

completion of the Final Safety Evaluation Reports for the ABWR and the CE System 80+.

What is left to consider is the potential value of a generic rule An issue not address' by any commentors is that of the state of the regulations.

The current regulations do not completely reflect the present available knowledge regarding the important regulatory issues regarding severe accidents, namely the risk posed to the public and effective design approaches for their prevention and mitigation in future reactor designs.

10 CFR 50.34 (i) does address certain of the issues, and this regulation is referenced in 10 CFR 52.47 where it is stated that the technically relevant portions of 10 CFR i

50.34 (f) must be addressed in future LWR applications.

However, the staff t

believes that the criteria contained in 10 CFR 50.34 (f) should be revi ed to reflect current information on severe accident issues.

Based on the above considerations, the staff believes that there could be regulatory value in pursuing development of a generic rule. However, it is not expected that a generic rule could be implemented in time to be of use during the design certifications of the two evolutionary designs.

As indicated above, the second area to be discussed is the approach to be taken for structuring a generic rule. Since most commentors did not support a generic rulemaking, there was not enough input to ascertain any public consensus regarding which of the three alternatives discussed in the ANPRM would be most favored. Comments that were received comparing the alternatives were useful, however, and should be considered if it is decided to proceed.

The staff believes that the best approach to be adopted for a generic rule would have to be determined after a more detailed evaluation of sample draft rules and discussion of the z.vailable options with the ACRS.

It is clear from the responses to the ANPRM, there presently is no consensus within the public regarding the best structure for a generic rute-if a rule was to be pursued.

However, certain advantages and disadvantages have been identified for each of the alternatives based on past discussions. These advantages and disadvantages have been described in the ANPRM. As indicated in the ANPRM, for a generic rule to be most effective, it would have to strike a balance between containing sufficient detail and prescriptiveness to be interpreted without ambiguity in a regulatory framework yet not be so prescriptive as to rule out alternative and possibly superior design approaches.

Because of uncertainties in predictions of severe accident behavior, the staff does not believe that it would be practical to develop officially sanctioned computer codes with controlled models and data bases such as was done for ECCS evaluations. Therefore, it is not proposed to include the subject of l

requirements for analytical codes in considering alternatives for a generic rule.

l Although there are uncertainties in the initiators and the course of severe accidents, the major challenges to the reactor and its containment that contribute significantly to public risk are well established and agreed upon by both industry and the NRC. There is also general agreement on the most

s 4

' The Commissioners 6

effective means for addressing these challenges through special design features and procedures aimed at preventing and mitigating these accidents.

These approaches have been previously documented in SECY-90-016 and its SRM and in SECY-93-087. They are also included in the EPRI ALWR Utility Requirements Document (URD). The staff believes that the implementation of the approaches described in these documents as is planned during the design certification proceedings will help assure that the risk to the public from severe accidents is maintained at a very low level. As noted above, it is unlikely that new phenomena will arise or that new approaches for addressing the challenges will be developed in the near time frame. The staff is continuing to examine the alternative approaches for codifying this information in a generic rule and intends to discuss them with ACRS prior to recommending a final rule structure.

On June 10, 1993, the staff met with the ACRS to discuss the subject of this paper. As a result of the meeting, the ACRS provided a letter supporting the staff's proposed approach for proceeding with the drafting of a generic rule.

(Letter from J. Erne;t Wilkins, Jr. to James M. Taylor, June 18, 1993, ).

In its letter, the ACRS also provided three specific recommendations for this activity. The staff will consider the ACRS' recommendations during the development of the draft rule.

OGC has reviewed this paper and has no legal objection.

Recommendation:

The staff recommends that the Commission:

(1) Endorse the staff's plans to continue discussions with ACRS concerning generic rulemaking to address severe accident licensing issues for future LWRs. A final decision to issue a rule and the approach to be used would not be determined until the Final Safety Evaluatier. Reports for the ABWR and the CE System 80+ designs are issued.

(2) Note that, pending approval of the Commission, the staff involved in drafting the generic rule will closely follow the reviews of the two evolutionary and two passive designs, completing some work in parallel. This approach should allow completion of a draft generic rule soon after the issuance of the Final Safety Evaluation Reports for the ABWR and CE System 80+

designs.

Also, to ensure consistency, the staff developing the draft generic rule will coordinate the development of this rule with the development of the design certification rules for the ABWR and CE System 804 designs.

Digoing revisions to 10 CFR 100 on source term and siting issues are expected to be completed in the first quarter of CY 1994.

Because of the related but different nature of the issues on severe accidents, a rulemaking for these issues in Part 50 would follow the completion of the revisions to 10 CFR 100 and 10 CFR 50 which address siting and source terms.

i i

4 The Commissioners 7

(3) Note that the staff would apply the results of such a generic rulemaking only to the licensing of future reactor designs - not to existing plants.

..e a s H. Tay r Exe,cutive Director for Operations

Enclosures:

As stated Commissioners' comments or consent should be provided directly to the Office of the Secretary by COB Thursday, September 2, 1993.

Commission Staff Office comments, if any, should be submitted to the Commissioners NLT Thursday, August 26, 1993, with an information copy to the Office of the Secretary.

If the paper is of such a nature that it requires additional review and comment, the Commissioners and the Secretariat should be apprised of when comments may be expected.

DISTRIBUTION:

Commissioners OGC OCAA OIG OPA OCA OPP EDO ACRS SECY l

INDIVIDUAL QUESTION RESPONSES The questions included in the Federal Register notice are paraphrased below.

After each question, a discussion is orovided on the range of comments provided by the public.

After the question discussion, a table is provided to summarize a selected set of the question responses. The table gives a quick assessment of the agreement and disagreement among the commentors on those questions that could be answered with a "yes" or "no."

In some cases, the responses were qualified, and notes are indicated to provide further explanation.

Question 1: (a) Is a rulemaking addressing si '.e accidents desirable?

If so, why? If not, why not?

(b) Would a rule provide better coherence and predictability for future reactor design licensing, or is individual design certification rulemaking sufficient?

Responses to Question 1: Three commentors expressed support for generic rulemaking; seven were not in favor, and one response did not explicitly state either response or opposition. The main reasons for these positions were provided above under the Discussion section of this paper.

Question 2 : Would a new severe accident rule in Part 50 provide a basis for revising the existing Emergency Planning Zones (EPZ) for future LWRs?

Responses to Question 2: FPC states that a basis for revision is provided if a rulemaking results in source terms that are significantly adjusted such that the resulting dose rate projections support EPZ revisions. OCRE argues that, due to the importance of emergency planning as a hedge against uncertainty, a basis for revising the EPZ would not be provided if the result was a weakening or elimination of present emergency planning requirements. AECL states that since the rule alternatives discussed in the ANPRM do not address offsite consequences, a basis for revising the EPZ does not exist.

It is further recommended that offsite consequences should be included in a generic rule and that EPZ revisions would then be supported.

NUMARC states that while the nuclear industry generally believes that a generic rulemaking for severe accidents is not necessary, a generic rule may be necessary to accomplish ALWR emergency planning and that the industry is continuing to evaluate this.

Question #3: (a) Is the proposed 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> containment performance criterion a suitable substitute for conditional containment failure probability of one in ten?

(b) Is the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> an appropriate time?

(c) Is the degree of conservatism appropriate considering uncertainties? What other criteria (probabilistic or deterministic) might be considered?

w Responses to Question 3: Fuoto suggests that there are weaknesses in both the I

proposed.1 conditional containment failure probability and the deterministic (24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> leak-tightness based) criteria and offers instead a multi-part criteria. His suggested criteria includes a probabilistic criteria for a prompt containment function failure combined with deterministic criteria for i

fission product retention and off-site dose. See Enclosure 3 for specifics.

OCRE suggests that the proposed 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> criteria is insufficient and that one week should be considered instead.

AECL suggests that specific criteria for containment performance should not be implemented due to the wide range of containment challenges.

It is suggested instead that general criteria concerning event frequency and radiological consequences should be in a rule with design specific details in regulatory guides.

NUMARC states that the industry is in favor of deterministic containment performance criteria as a means for providing defense in depth and that this is a suitable substitute for a probabilistic criterion. NUMARC stated that industry analyses indicate that a 16 hour1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> criteria would be adequate and that while a 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> criteria is probably conservative, the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> criteria is a part of the ALWR utility requirements document (URD). As planned, the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> criteria will be codified as part of the design certification rulemakings for the ALWRs.

It was also stated, however, that the generic codification of the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> criterion proposed in the ANPRM would be inconsistent with existing Commission guidance on safety goal policy because the use of a 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> criterion would, in effect, impose a de facto higher safety goal than is represented by the Commission's quantitative health objectives.

Entergy suggests that an additional criteria that should at least be encouraged is that of containment venting capability due to its importance as an accident strategy.

Question 4: (a) Are there analytical tools that are sufficiently developed to implement the phenomena based rule of Alternative 27 Should any such codes be approved by the NRC? (b) Should the codes and input parameters be approved by l

NRC? Should acceptance criteria be codified in a rule or put into a regulatory guide?

Responses to Question 4: FPC suggests that it would be difficult to get agreement on standards by which to judge acceptability of severe accident analysis codes and which codes should be used as the official approved codes.

NUMARC states that the URD will include appropriate analytical approaches and that this will again be codified during the design specific ALWR rulemakings.

Question 5: Should future designs for reactor containments include features beyond those described for Alternative I?

Responses to Question 5: OCRE suggests that additional features and requirements should be included and offers a number of suggestions aimed at ensuring a containment capable of withstanding any accident.

Details are l

given in Enclosure 3.

On the other hand, AECL states that the features indicated in Alternative 1 are too restrictive and are an impediment to innovative design solutions.

NUMARC states that the features given in Alternative 1 of the ANPRM are already included in the URD.

2

9 -

Question 6: What criteria should be used to determine what severe accident challenges should be considered?

(a) Should the challenges be specified in more detail?

Responses to Question 6: FPC suggests that only high-level requirements should be considered.

OCRE suggests that challenges should be specified in more detail and cites the ACRS May 17, 1991 letter on this subject. AECL suggests that there should be general requirements for all reactors in the first part of a rule and design-specific requirements in a second part. NUMARC states the URD and the design certification process will adequately address this issue.

Question 7: Are there any reasons why the criteria proposed in the three e

alternatives would not apply to the passive reactor designs?

Responses to Question 7: FPC states that criteria for passive designs should be the same.

Fuoto suggests that risk-significance should be considered to determine applicability. OCRE believes the same criteria should apply.

NUMARC states that the passive plant version of the URD will assure coherent application of criteria.

l Ouestion 8: What features could an advanced LWR design include to prevent or mitigate fuel-coolant interactions?

Responses to Question 8: Fuoto suggests that while uncertainties make it difficult to specify design and procedural criteria to prevent fuel-coolant i

interactions, the capability for the addition of coolant during an accident should always be considered beneficial.

NUMARC cites design features being provided for the various ALWR designs to protect against fuel-coolant interactions and cites the URD as assuring adequate protection.

Ouestion 9: If a design includes a rapid RCS depressurization system to mitigate high pressure melt ejection (HPME), should that design be required to also provide a reactor cavity and vessel support system design to mitigate HPME?

Responses to Question 9: FPC responds that the likelihood of a HPME is too low i

to merit consideration. OCRE responds that a reactor design and vessel i

support system design should be included in the event of failure of the i

depressurization system or an act of sabotage. AECL responds that the need for such additional features should be determined based on event frequency l

considerations and that it is better to design to prevent this event rather than to mitigate it.

NUMARC states that the URD provides for both approaches although it is believed that the reliability of the depressurization systems specified in the URD precludes the need for the additional design features.

Question 10: (a) Should future LWR designs include an on-line instrumentation system to monitor containment atmosphere for containment bypass? (b) Would such a system allow modification of leak testing requirements in Appendix J?

Responses to Question 10: FPC responds that such a system is not necessary, 1

that its safety value would be minimal considering the cost required to 3

i

O develop such a system.

Fuoto responds that such a system would likely be inadequate in most cases.

OCRE responds that there should be on-line containment monitoring but that it would not adequstely substitute for Appendix J requirements.

AECL indicates that such a system is provided for in the CANDU design and that it should provide a basis for modifying Appendix J requirements.

NUMARC responds that the URD provides for a periodic check of containment status to ensure that penetrations are not inadvertently left open after maintenance operations but that such a system cannot be considered a replacement for Appendix J.

Question 11: What design criteria could be used to establish containment integrity during shutdown conditions?

Responses to Question 11: OCRE cites agreement with the ACRS recommendation for containment protection regarding providing for ease of emergency closure during shutdown conditions.

Fuoto responds that improved technical specifications on available power sources and makeup systems and containment configuration should be adequate to provide protection during shutdown conditions.

NUMARC responds that managing risk during these relatively brief and infrequent periods is best accomplished by plant system configuration control and capability for mitigating any loss of coolant inventory.

Duestion 12: Should plant equipment provided for severe accident prevention or mitigation meet the same qualification standards as design basis equipment?

Responses to Question 12: FPC responds that severe accident mitigation requirements should be less than design basis requirements.

Fuoto responds that risk importance should be a determining factor and that the Station Blackout Rule provides a good model for accident equipment requirements.

OCRE responds that the same qualification standards should be imposed.

AECL

. responds that the requirements for specific items of equipment should be

' tailored to their specific expected accident environments and the need for tFa e quipment including their reliability goals.

NUMARC responds that the e

4

' standards for severe accident related equipment should be lower in accordance I

"*vith their lower likelihood of being required for use during an accident.

P Ouestion 13: (a) Are the proposed Service Level C stress limits for steel containments appropriate? (b) Could these same limits be used for missile loads? What equivalent limits could be used for concrete?

Responses to Question 13: FPC responds that existing design requirements with IPE related enhancements should be appropriate.

Fuoto responds that Service Level C is probably too limiting considering the accepted standard of applying Service Level D to LOCA.

OCRE states agreement with the use of Service Level C. NUMARC indicates support for the use of Service Level C but states that less conservative limits should be used for missiles.

NUMARC also recommends that for concrete containments, the unity factored load combinations of Subsection CC of ASME Section III would be appropriate.

Question 14: What information is available regarding the costs of the design features that would be required under these alternatives?

4

' Responses to Question 14: None of the commentors provided any information in response to this question. NUMARC stated that it had not evaluated these relative costs.

'Ouestion 15: If a rule such as those alternatives considered in the ANPRM was implemented, should that rule allow the elimination ~of the consideration of SAMDAs under 10 CFR 51?

Responses to Ouestion 15: NUMARC responds that while it is not in favor of a generic rulemaking, it concurs in the underlying thrust of Question 15 to avoid imposing additional requirements on the ALWRS (i.e., if a generic rule were to be implemented, there would be no need to consider SAMDAs further.).

FPC responds similarly.

SUMMARY

OF RESPONSES TO SELECTED ANPRM QUESTIONS (questions one through six)

M E

M M

M M

2 DOE Not now - Suggests postponing decision TVA No - Cites agreement with Niagara FPC No Yes Yes No No Gen'l Regt's*

ENTERGY N/A' N/A N/A Vent'g' N/A FUOTO Yes Yes Yes N/A No No OCRE Yes No No No Yes Yes AECL Yes Yes No No No No ALWR No - Cites agreement with NUMARC response 5

WESTINGHOUSE No - Cites agreement with NUMARC response NUMARC No maybe Yes N/A N/A N/A WINSTON &

No N/A N/A N/A N/A 64/A STRAWN Notes for Table 1:

1. It is recommended (DOE) tho perhaps the central issue is the timing for a possible rulemaktag.

Due to the advanced state of the designs planned for design certification, a generic rulemaking initiated at this time would have the potential for delaying the schedules for these designs.

DOE suggests that i

after the design certifications of the evolutionary and passive designs are completed, the value of generic rulemaking could be revisited.

2.

It is recommended (FPC) that if severe accident design criteria are to be codified, only high-level general requirements should be specified.

3.

In this table, "N/A" indicates that the question was not addressed by the commentor because he or she chose not to or that an earlier position obviated the need to address the particular

question, e.g.,

for those commentors opposed to a generie rulemaking, questions regarding the manner in which a rulemaking would be carried out were generally not addressed.

4. It is recommended (Entergy) that controlled venting capability should be encouraged due to its importance as a strategy for severe accidents.

6

Enclocuro 2

/

'o UNITED STATES y'

',g NUCLEAR REGULATORY COMMISSION

{

r ADVISORY COMMITTEE ON REACTOR SAFEGUARDS o,

g WASHINGTON, D. C. 205$5

%, +..../

June 18,1993 Mr. James M. Taylor Executive Director for Operations U.S.

Nuclear Regulatory Commission Washington, D.C.

20555

SUBJECT:

PUBLIC COMMENTS ON PROPOSED RULE ON ALWR SEVERE ACCIDENT PERFORMANCE During the 398th meeting of the Advisory Committee on Reactor Safeguards, June 10-11, 1993, we discussed with members of the staff public comments received on the Advance Notice of Proposed Rulemaking (ANPR) on ALWR Severe Accident Performance.

We had the benefit of the documents referenced.

It is our understanding that the staff's proposed approach for proceeding with rulemaking involves the following four elements:

1.

Continuing discussions with ACRS concerning a potential generic rule, 2.

Delaying a final decision on implementation of the rule until after final safety evaluation reports are issued for the Advanced Boiling Water Reactor (ABWR) and the CE System 80+,

3.

Coordinating the efforts of drafting a generic rule and the design certification rules for the ABWR and the CE System 80+

to ensure consistency, and Following the reviews of the evolutionary and passive reactor 4.

designs to ensure consistency of the draft rule with these reviews.

We agree with this approach.

In our reports on this subject dated May 17, 1991 and May 14, 1992, we developed and subsequently endorsed what is designated as Alternative 3

in the ANPR.

We continue to recommend this alternative.

i

Mr. James M. Taylor 2

June 18, 1993 4

For your further consideration, we recommend that your approach accommodate the following:

1.

The amended regulations should not be so prescriptive as to preclude the use of a design feature which substantially reduces the challenge (s) to the containment.

For example, the approach should not require accommodation of large amounts of hydrogen generation if a design change (such as different core materials) precludes significant hydrogen generation, 2.

The recognition of passive design features to cope with some phenomena, e.g., a large volume-high strength containment, and 3.

Consideration for dealing with combinations of containment loads from severe accident phenomena, e.g.,

steam explosions and hydrogen combustion / detonation.

We expect to have further discussions with the staff on this matter.

Sincerely, J.

Ernest Wilkins, Jr.

Chairman

References:

1.

Memorandum dated May 14, 1993, from Warren Minners, Office of Nuclear Regulatory Research, for John T.

Larkins, Advisory Committee on Reactor Safeguards,

Subject:

Summary of Public Comments on Proposed Rule on ALWR Severe Accident Performance

- 57 FR 44513 (Predecisional Draft Commission Paper Attached) 2.

Report dated May 17, 1991, from David A. Ward, Chairman, ACRS, to Kenneth M. Carr, Chairman, NRC,

Subject:

Proposed Criteria to Accommodate Severe Accidents in Containment Design 3.

Report dated May 14, 1992, from David A. Ward, Chairman, ACRS, to James M. Taylor, Executive Director for Operations, NRC, subject:

Advance Notice of Proposed Rulemaking on Severe Accident Plant Performance Criteria for Future LWRS

I s.

e f

9 9

(

i

-I COPIES OF COMMENTS RECEIVED ON ANPRM 6

i

't s

_s s o s J'

... 'T 5 0

\\

Department of Energy (5 7 F R W si3) i g

ag

..f...

[

i e

4 Washington, DC 205S5 4

s Decembe r 24, 1992

~o2 DE 3 g.p Mr. Samuel Chilk Secretary of the Comission U.S. Nuclear Regulatory Comission Washington, D.C.

20555

Dear Mr. Chilk:

This letter provides Department of Energy (00E) coments on the Nuclear Regulatory Comission's (NRC) advance notice of proposed rulemaking entitled,

" Acceptability of Plant Performance for Severe Accidents; Scope and Consideration in Safety Regulations." The proposed rulemaking has the potential for impact on the review schedules for design certification applications currently before the Comission, which DOE is sponsoring in cooperation with the nuclear industry.

We believe that the primary issue is whether generic rulemaking should be pursued at this time or whether the design-specific rulemakings that are currently scheduled should first be completed. We recomend that the difficult issues related to severe accidents first be resolved in the context of specific designs through certification rulemaking.

If necessary, the certification rulemaking could be followed by a generic rulemaking.

The advanced status of the designs submitted for certification obviates one of the major benefits of generic rules, which is to provide uniform and consistent guidance to designers by identifying the requirements that must be met during the safety review. The passive Advanced Light Water Reactors (ALWRs) have been designed and Standard Safety Analysis Reports have already been submitted to the Comission for review and approval.

Further, these designs are in accordance with the generic severe accident approaches incorporated in the ALWR requirements document. Since there are no other U.S. ALWR designs anticipated for some time, the generic rulemaking can be conducted at a later date and still achieve this benefit of generic rulemaking for any future designs. The lessons learned from the design approval process for current ALWRs can be incorporated in a future generic rulemaking.

In addition to being not necessary to the review of the two passive ALWR designs currently before the NRC, it is probable that the generic rulemaking would adversely impact the certification schedules for these plants.

Based on prior rulemaking experience, it is likely that the schedule for a generic rulemaking, which needs to address issues and situations that would be applicable for all future ALWR applications, would exceed the current plant-specific certification schedules. Thus, the likely outcome of pursuing a generic rulemaking at this time will be to delay the time when certified passive plants are available as an option in the United States.

L

~.I 193 4Ci.WC

• i

2 For these reasons, we request the Conrnission delay proceeding with generic rulemaking at this time and consider the need for a generic rule after the experience gained through the passive plant reviews is obtained.

Sincerely, E. C. Brolin, Acting Principal Deputy Assistant Secretary for Nuclear Energy r

6 5

h b

C ! K. e i i 1% !

l 10

~....

A

'93 JAN -4 P 3 :05 Tennessee valw.e Av; horny 1101 Market Street Chatta%OQa Ie9Tssee 37402

~

December 29, 1992

+.

Mr. Samuel J. Chilk Secretary of the Commission ATTN: Docketing and Service Branch U.S. Nuclear Regulatory Commission Washington, D.C. 20555

Dear Mr. Chilk:

NUCLEAR REGULATORY COMMISSION (NRC) - REQUEST FOR COMMENT ON ADVANCED NOTICE OF PROPOSED RULEMAKING; ACCEPTABILITY OF PLANT PERFORMANCE FOR SEVERE ACCIDENTS: SCOPE OF CONSIDERATION IN SAFETY REGULATIONS The Tennessee Valley Authority (TVA) has reviewed the subject advanced notice of proposed rulemaking, which was noticed in the September 28, 1992 Federal Register (57 FR 44513-44518), and is pleased to provide the following com: rent.

TVA supports those comments submitted by Winston & Strawn on this proposed rulemaking.

Sincerely, i

Mark J. Burzynski Manager Nuclear Licensing and Regulatory Affairs cc:

U.S. Nuclear Regulatory Commission Region II 101 Marietta Street, NW, Suite 2900 Atlanta, Georgia 30323 Mr. Thomas King Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, D.C. 20555

AN _

593

.cyr. "i%C Di ~"! C'b

~~ -~ """"*""*"

Lf " ~ N g

~ SO II (57 F2 Rsh)

~

January 21, 1993

~ Nif The Secretary of the Comission,

'93 Jm 27 P3 :13 U.S. Nuclear Regulatory Commission Washington, DC 20555 ATTN:

Docketing and Service Branch c

Subject:

Comments on Proposed Rulemaking -

" Acceptability of Plant Performance for Severe Accidents; Scope of Consideration in Safety Regulations"

GENERAL COMMENT

S:

The NRC is contemplating revisions to 10 CFR Part 50 to include design considerations for challenges from severe accident (SA) phenomena for future light water reactors (LWRs).

As stated in the proposed regulation existing nuclear plants do not pose undue risk to public health and safety because present design and regulation is sufficient, although very complex and burdensome. The NRC states that it recognizes the need to "... strike a balance between accident prevention and consequence mitigation in exploring the need for additional design features in the next generation of plants." However it is not clear how the NRC demonstrates that recognition in light of the proposed regulations.

In proposing the regulation the NRC identifies two fundamental areas of concern:

1) SA mitigation through improved design based on insights from IPE and resolution of several issues such as station blackout, anticipated transient without scram, hydrogen generation and control, and 2) Containment performance; establish new performance criteria.

The former concern simply is not be expressed in terms that designers are able to use without skyrocketing the cost of building a nuclear plant. The latter concern attempts to encapsulate the SA in an impenetrable (energy / mechanical) device.

To aid in the determination and applicability of this proposed regulation three alternatives are provided. The first of these (Alternative 1) attempts to define hardware needs in terms of six (6)

" risk significant" SA phenomena.

Unfortunately the features that would be defined in the rule were not available for review now.

Although a statement is made to address "... these new regulations would not be considered to be traditional design basis requirements",

it does not differentiate what that means with respect to operability type evaluations.

It is pointed out that a regulatory guide would be issued to "...

provide additional guidance on such design details such as redundancy, diversity, system capacity, power supply, equipment survivability, and analytical assumptions." However it is not clear why this much detail is proposed for such a regulation if an understanding of the relationship between prevention and mitigation is really appreciated.

Finally, this alternative closes with a statement that would allow the licensee to justify design without performing

" extensive" analysis to show compliance.

The problem is NRC still wants justification based on likely SA scenarios combined with deterministic analysis.

This is quite costly and may not be able to prove anything.

9Q i

g FLORIDA POWER CORPORATION Power eUCLEAR OPERATIONS N

O 8Ox 2ie CRYSTAL RIVER. FLORIDA 32623 0219-s

Nuclear Regulatory Commission Page 2 of 4 Alternative 2 attempts to define containment performance requirements in terms of analytical methods, assumptions, acceptance criteria and guidance on design criteria for SA hardware. The NRC recognizes that this approach would probably

~

i (necessarily) limit the types of analytical tools and code to those specified by the staff due to the diversity, complexity, and uncertainty of available or yet to be developed codes.

It would be very difficult, if not impossible, to agree on code selection and analytical methodologies.

Alternative 3 uses the General Design Criteria (GDC) approach. Because SA design basis would (probably) be different from traditional design basis, regulations would necessarily be more detailed and complex. This approach would rely on many of the methods used in Alternative 2 to define related basis which may or may not be acceptable to the NRC.

COMMENTS TO SPECIFIC CONSIDERATIONS:

1.

A rulemaking for severe accident space is not good business.

Rules and regulations should more closely correspond to issues that really challenge the health and safety of the public.

Integrating IPE information into design criteria for future reactors may be applicable for cases where significant safety improvement is derived.

2.

If source terms are significantly adjusted (with or without SA design modifications / codification) such that dose rate projections and significant benefits are derived from revising Emergency Planning Zone requirements then a basis for revision is indicated.

3.

Under worst case conditions containment pressure is projected to reach as high as 200 psig.

To prevent challenging the containment venting is indicated. It is not reasonable to design a containment to withstand such pressure and codify it. Although UPC is relatively high for many types of containment designs venting is a reasonable strategy to prevent exceeding it thus reducing the probability of an uncontrolled release.

If a hardness factor based on realistic failure probability can be applied to design criteria established for " traditional" design basis then ti.ere may be some safety benefit derived.

4.

This relates directly to earlier comments made to Alternative 2 & 3.

Some codes and analytical methods may need to be developed.

It's not clear what standard (s) (if any) would be used to develop and evaluate such codes. With the wide variety of assumptions and existing codes, including revisions / versions it would be very difficult to get global buy in on any particular one, unless of course it is specified in regulations which may or may not be analytically correct.

5.

No.

Nuclear Regulatory Commission Page 3 of 4 6.

If SA design criter Jst" be codified then the only realistic approach would be high levei, sral statements. Unfortunately this approach does lend itself to a host of other problems which typically begins with differences of interpretation leading to varied levels of compliance.

7.

SA space is independent of reactor design in that SAs begin with a degraded core which is "beyond" design basis (or should be).

Therefore applicability to passive designs should be the same.

8.

This is an open ended question and not really appropriate here.

9.

No.

There is no evidence that a HPME is even a reasonably possible (probably) event.

Even if it were the consequences of such an event are not worthy of design considerations in terms of safety benefits.

10.

Although this is an interesting concept surveillance activities and containment controls are sufficient.

Should a bypassed or impaired containment develop during a SA the usefulness of such instrumentation, provided it would be available, is minimal.

The cost of development to some arbitrary design criteria, installation, calibration, maintenance, t

and actions to take upon degraded performance would be very difficult to demonstrate safety benefits, 11.

This is a separate issue.

12.

If equipment is targeted for SA mitigation and "must" be regulated for that function then those regulations need to reflect a significantly lesser set of requirements.

What those would be and what they would be based on is not readily definable at this time.

13.

Existing design requirements with enhancements as identified by IPE and adjusted per coment 3 may be an appropriate means of establishing limits.

14.

The answer to this question should be directed to and obtained froe NUMARC.

15.

If containment performance criteria includes SA phenomenology then it would bound all other regulatory concerns.

t

1 Nuclear Regulatory Commission Page 4 of 4 CONCLUSIONS:

1 Although the concept of defense in depth may be applied to SA mitigation strategies as a part of overall accident management, it is not reasonable to generally include those in design criteria regardless of separation from

" traditional" design criteria. If significant vulnerabilities exist which pose a threat the health and safety of the public (in SA space) then we are obliged to revise our design criteria (for the future) and operating practices.

Based on opening NRC comments, existing nuclear power plants do not pose such a threat.

Therefore, consideration of regulatory change to include SA mitigation design criteria should be limited to reflect significant insights from IPE and clearly identified design feature improvements where safety benefits are meaningful and quatifiable.

/

,,c2-f aul V. Flemin 1

Senior Nuclear Licensing Engineer l

/

m

~~.3

• ( (I

[3

~

===== Entergy

~ ~" oa ? ?"

'"c-

~

(5 7 F R WS 13)

Operations 3g=

j

'92 DEC 21 P4 :~4 i

December 16,1992

R OCAN129203 Secretary of the Commission Attn: Docketing and Service Branch U. S. Nuclear Regulatory Commission Washington, DC 20555

Subject:

Proposed Rulemaking on Acceptability of Plant Performance for Severe Accidents Gentlemen:

On September 28,1992, the NRC published an advance notice of proposed rulemaking in the Federal Register (50FR57188) concerning acceptability of plant performance for severe accidents for future light water reactors. The NRC requested comments on the proposed regulations.

Please find our following two comments:

Specific Considerations, Item 5: Asks about future LWR design features beyond those described in Alternative 1. One of the aspects of Alternative 1 deals with controlled elevated venting (if provided in the design) after the initial 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period.

Comment:

While there is no requiremgfor Containment Venting capability in.this proposed rule, this feature should.be encouraged due to its importance as a strategy in the severe accident environment.

jal.

_ 153 Acirc..2 ;s 7 = ::...........

-=

U.S.NRC 7

December 16,1992 OCAN129203 Page 2 l

Specific Considerations, Item 7: Asks for what reason would any of the criteria proposed in the

,three alternatives not be fully applicable.

Comment:

i RCS depressurization capability. While the capability to rapidly'depressurize the ECS may be applicable in some cases, it may not be the best option for all scenarios. MAAP runs have shown that RCS depressurization can cause earlier containment failure and at the same time not provide any benefit to accident mitigation.

Very truly yours, ym James J. Fisicaro Director, Licensing 1

JJF/NBM/sjf l

i cc:

U. S Nuclear Regulatory Commission 4

Document Control Desk i

Mail Station PI-137 l

Washington, DC 20555 i

f i

a p

n

~

[ l'M COC '.ET NUMBER g.','05EDRULEqT. Ep j

dG7 F A W5/3)1825 Prelude Drive mm U5NHL Vienna, VA 22182-3345 November 2, L 32 92 W --9 P 3 3 The Secretary of the Commission U. S. Nuclear Regulatory Commission 2;

Wasnington, DC 20555 ATTENTION: Docketing and Service Branch 57 FR 44513 - PROPOSED RULE ON ALWR SEVERE ACCIDENT PERFORMANCE 1.

Desirability of Severe Accident Performance Rulemakino A rulemaking on severe accident performance is appropriate and is highly desirable.

First, it will address for ALWRs many of the considerations that went into 10CFR50.44 that addresses a number of plants that will not be built. Second, a rulemaking will provide a consistent regulatory basis for certification and licensing of future plants.

Policy staten'ents have their uses, but they do not provide a substantial basis for licensing and regulatory actions that may be required.

2.

Would a New Rule Provide a Basis for Revisino EPZ Criteria for Future LWRs?

A properly crafted rule could provide a basis for revising EPZ criteria for future LWRs.

First, plants designed to explicitly meet severe accident performance criteria could have lower off-site dose consequences from core damage sequences. This can provide a technical basis for reducing the EPZ from a ten mile radius to something smaller.

Second, the prospect of reduction in the EPZ can likely give substantia 1 incentive to even more robust containment designs, since there can be a measurable reduction in public risk.

Third, from a public acceptance point of view, how can one claim one has developed a safer reactor design when there is no change in the EPZ? Thus, coherent regulations in this area can demonstrate in a tangible way that the advanced and evolutionary LWR designs indeed do result in lower risk to the public.

3.

Containment Performance Criterion Ootions Obviously, the selection of a one-in-ten conditional probability of containment failure is based first on determining a suitably low core damage frequency probability. The use of a leak-tightness criterion option is based on the assumption that leak-tightness equals radionuclide retention, which is only strictly true in the case of noble gases. Existing NRC-sponsored studies show that substantial increases in containment leakage rates l

following severe accidents lead to only small increases in public risk. I believe that the suggested surrogates miss the point. It is how well the containment performs to reduce risk to the public that is the issue. Thus, neither of these options alone is sufficient to address the issue of minimizing public risk in a cost-effective way.

.N 21 19 kchrn...c ;:.y

.........93

U. S. Nuclear Regulatory Commission November 2,1992 Page two I believe that a more appropriate containment performance criterion can be crafted. I would suggest multi-part criteria along the following lines:

a.

That the conditional probability of prompt (within about one hour after initiation of core damage) containment function failure be less than some appropriate va!ue. This probability would likely be less than 10% and probably on the order of not more than 1%, and b.

That the containment function be capable of retaining at least 95% to 99%

of radiciodines and particulate radionuclides following failure. This appears readily achievable based on NUREG-1465 and would give full credit to radionuclide retention features such as sprays, chemistry controls, controlled filtered venting, and in-plant radionuclide plate-out, and c.

That off-site doses due to noble gas release be reduced on the order of at least 50% to 75% below that which would occur if the noble gases were released as soon as they were generated and transported to the containment breach (es). This is consistent with criterion a. above.

4.

Analysis Reovired for Alternative 2 I believe that Altemative 3, Severe Accident General Design Criteria, is the best choice Thus, I have no further comment on this point.

5.

Inclusion of Additional Features Bevond Those Addressed in Alternative 1 I believe that Alternative 3, Severe Accident General Design Criteria, is the best choice However, the severe accident phenomena and concems addressed in Alternative 1 summarize the scientific, regulatory, and utility consensus on potential mechanisms for containment challenge. Based on my review of severe accident and PRA literature, I do not believe that any other mechanisms need to be considered.

6.

Phenomenolooical Severe Accident Challenoes - Level of Soecificity The phenomena to be considered should not be specified with some arbitrary level of challenge such as 100% clad-water reaction, etc. First, such specificity will not provce incentive for designers to avoid or minimize challenges. If a designer can reduce hydrogen generation, surely there should be positive regulatory incentive to do so.

l l

i U. S. Nuclear Regulatory Commission November 2,1992 I

Page three Second, arbitrarily determining level of challenge a priori will not promote cost-benefit tradeoffs to be made among protection features. Clearly, more weight should be given to prompt containment failure sequences than those that may occur later and to more t

probable failure mechanisms to less probable ones.

Third, a review should be included to assure that design features and procedure-related strategies do not introduce new failure mechanisms. Such a review does not have to be overly complex and should reflect best judgement. This would be consistent with the intent of 10CFR50.59, and would be a more complete treatment of severe accident i

containment challenges than would be likely under detailed rule-specified level of challenge.

7.

Basis for Non-Aoolicabifrty to Passive Desianed LWRs Given that the fundamental basis for passive designs is to rely on methods that will "always" work, determination of applicability based on risk-significance is most appropriate. This would give full credit for a passive design's ability to eliminate certain containment failure sequences or to reduce the public health risk consequences.

8.

Prevention or Mitigation of Fuel-Coolant Interaction Obviously, any feature which minimizes the probability of sustained core uncovery and prolonged inadequate core cooling goes a long way towards preventing fuel-coolant interaction. Given the phenomenological uncertainties, it is difficult to determine precisely what design and procedure strategies should be followed to minimize fuel-coolant interaction. However, I strongly recommend that the regulations presume that without clear evidence to the contrary, addition of coolant is always beneficial, no matter where in the core damage sequence it is finally (re) introduced.

9.

Mitication Features for Mitication/ Prevention of High Pressure Mett Election This question concerns level of specificity required for containment feature performance assuming an Attemative 1 type rule. Please refer to my comments 3. and 6. above.

10.

Use of On-Line Instrumentation System to Determine Gross Containment Leakace See comment 3. above. Unless the designer chose a double-walled containment design.

or a secondary containment completely enclosing the primary containment and all of its penetrations, such instrumentation would likely be inadequate when it would really be needed during containment bypass sequences and unnecessary for all other sequences.

l U. S. Nuclear Regulatory Commission November 2,1992 Page four 11.

Containment intecrity Design Criteria Aoolving to Shutdown Conditions During shutdown conditions, sensible and decay heat loads are far less than at power.

This substantially reduces the rate and the magnitude of radionuclides that can be generated. Additionally, the same features that are added to ALWRs to increase the ECCS reliability will be available to keep the core covered during shutdown event

+

sequences.

As far as off-site public risk is concerned, shutdown events are a tempest in a teapot.

l I have reviewed calculations on an existing PWR that would indicate that a core melt sequence starting from cold shutdown with the containment hatchway open would be required in order to have off-site releases corresponding to a General Emergency.

Improved Technical Specification requirements pertaining to available power sources, available makeup systems, and containment configuration should more than cover shutdown events.

12.

Standards Reovired for Severe Accident Eouioment There appears to be little difference in the performance of well-designed and maintained

" safety-related" and "nonsafety-related" equipment. Given the low probability of severe accident, " gold-plated" OA requirements are clearly inappropriate and will result in unnecessary cost with little or no risk benefit. The standards to be used should be performance-based, reflect the existence of the maintenance rule, and reflect importance as determined by probabilistic safety assessment.

The process for determining appropriate standards reflected in the Station Blackout Rule is probably a good model for severe accident equipment, since SBO is typically a significant contributor to core damage.

13.

Mse of ASME Service Level C Umits 1

As stated above, I believe that Afternative 3, Severe Accident General Design Criteria, is the best choice. However, use of level C stress limits is probably too limiting. Given that Level D stresd firnits are used for once in a plant lifetime design basis events such as LOCA, it certainly would be inconsistent to use level C stress limits for events that are supposed to be beyond the design basis and of even lower probability. Given that the real function of the containment is to mitigate off-site dose consequences of severe accidents, significant structural plastic deformation could occur without affecting the ability to retain radionuclides.

)

l i

3 U. S. Nuclear Regulatory Commission November 2,1992 i

Page five l

i

~

i 14.-

Information Available on Costs

(

I am not qualified to comment on this point.

lf l

15.

Containment Performance Objective versus 10CFR51 SAMDAs l

See my comment 3. above. Use of multi-part criteria suggested above is consistent with

[

the Safety Goal Policy Statement and can form a reasonable outer bound beyond which

" remote and speculative" sequences can be removed from consideration based on explicit consideration of the end goal - reasonable measures to reduce public risk.

i By using surrogate conditions that may be more conservative than required in most l

cases, and that may not adequately address prompt failures against which no public l

protective actions can be effectively implemented, then the goal of coherent regulatory i

policy is missed.

i By focusing severe accident mitigation features on reduction of public risk, rather than some intermediate expedient surrogate, the technical and safety basis is clearly defined, -

need for future backfits can be more readily determined, and most importantly, a determined intervenor cannot use apparent inconsistencies among regulations to use a j

non-technical court system to set aside regulations. This has been done before, successfully, and to allow history to repeat itself would be unconscionable.

l Sincerely, 8

John S. Fuoto, PE l

l 1

+

1 i

l 4

m s

@\\M

,. [ '...-

~

_ _ '.j [ h,,

[G2 Fg wg13 y=

p 3L9 December 27, 1992

~92 OEC Z A10:03 COMMENTS OF OHIO CITIZENS FOR RESPONSIBLE ENERGY, INC.

(." 0C R E." )

ON ADVANCE NOTICE OF PROPOSED RULEMAKING. " ACCEPTABILITY OF PLANT PERFORMANCE FOR SEVERE ACCIDENTS; SCOPE OF CONSIDERATION IN SAFETY REGULATIONS,"

57 FED. REG. 44513 (SEPTEMBER 28, 1992)

In this notice the NRC proposes regulations which would require future nuclear power plants to be able to withstand severe acci-dent phenomena which present a challenge to maintaining contain-ment integrity.

OCRE generally supports this advance notice of proposed rulemaking.

The major flaw in the NBC's safety regalations for nuclear power plants is the exclusion of severe accidents from the design basis.

Instead of being designed to withstand all possible accidents, the current generation of nuclear reactors is designed to withstand only a limited set of events which are assumed to be mitigated by the operation of the ECCS, with minor radiological consequences to the public health and safety.

Severe accidents were deemed to be " incredible," at least until the TMI-2 accident proved otherwise. Due to this regulatory gap, severe accidents pose the greatest risk to the public health and safety from nuclear reactor operation, since current plants have limited capabilities to cope with severe accident phenomena.

E. g.,

NUREG-1150 found that even the strongest of current containments would not withstand severe accident loads.

It is OCRE's position that severe accident phenomena must be made part of the design basis for future nuclear power plants.

This advance notice of proposed rulemaking would accomplish precisely that.

Hence, OCRE supports this effort.

In addition to enhancing safety, this effort will also benefit the nuclear industry as it seeks a revi~al.

The public is un-likely to accept a new generation of reactors that cannot with-stand severe accident phenomena.

Of the three alternatives presented in the notice, OCRE prefers a combination of Alternative 1, the hardware-oriented

rule, and Alternative 3,

GDC-oriented rule.

Due to its prescriptive nature, Alternative 1 minimizes reliance on the uncertain, imper-feet analytical methodologies which are required in the other two alternatives.

However, Alternative 1 does not make severe acci-dents part of the traditional design basis, which Alternative 3

does.

The disadvantage noted for Alternative 1, that it could 1

JAM l'.53 AC/EW4CCCC l,' TO ------ -

discourage the development of other design approaches, does not appear to be that much of a problem; the requirements are not so restrictive as to preclude innovation, and in the event that they

are, the NRC could grant an exemption or modify the rule if necessary.

Innovative designs could also be evaluated in the SAMDA review process.

OCRE would suggest the following improvements to Alternative 1:

A.

The containment performance objective states:

"The design shall include a

containment system that provides a

barrier against the release of radioactive material for a

period of approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the onset of core damage under (emphasis likelv severe accident challenges.

the more added).

Any insertion of the words "likely" or " likelihood" into the rule will defeat its purpose by allowing licensees and plant designers to claim that they need not comply with its -provisions because severe accident challenges are not likely.

This asser-tion will of course be buttressed by a PRA to which fudge factors will undoubtedly have been applied.

Even without the fudge

factors, FRA has enough limitations and uncertainties that it should not be used to evade compliance with the rule.

This is a

retreat to the " accidents can't happen" mentality of the past.

The present wording of this Alternative creates a

substantial loophole which must be closed if the rule is to have any force.

B.

The 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> timeframe for maintaining containment integrity is insufficient.- OCRE prefers the application of the present GDC i

16, which requires the containment to provide "an essentially

.. for as long as postulated accident leak-tight barrier conditione require."

This is the approach suggested in the ACRS letter oi May 17, 1991.

If containment venting is not to be prohibited, it should not be utilised until at least one week after the onset of core damage.

A one week period would provide more time for fission product removal mechanisms to work, would increase the likelihood that the accident would be recovered or stabilised, and would allow more time for offsite protective ac-tions.

The use of diverse containment heat removal systems should be required.

t C.

The description of the severe accident phenomena should be more detailed and comprehensive.

The description of the phenome-na in the ACRS letter of May 17, 1991 is far more appropriate than the cursory descriptions in the Federal Register notice For example, the ACBS description of the hydrogen sources to be considered is: "in-vessel and ex-vessel oxidation of core mater:-

L als, including (1) core degradation from overheating and melting.

(2) steam explosions or high pressure melt ejection in the l

2 P

't

l a

presence of water, and (3) interaction between molten core mate-rials and concrete."

This is an accurate description of the actual sources of hydrogen gas.

The description in Alternative 1,

"a 100-percent metal-water reaction of the active fuel clad-ding,"

is not complete and accurate.

This description ignores many sources of hydrogen, such as the oxidation of metal other than active fuel cladding in the core, the reaction of boron carbide control rods with steam, and core-concrete interactions.

For another example, consider the ACRS-proposed criterion on core-concrete interaction:

The containment system would have the capacity to accommodate the following challenges resulting from the thermal decomposition of concrete by molten corium: (1) the degradation of containment cooling and of cleanup capability due to aerosol formation, (2) slow overpressurization resulting from the evolution of noncondensible gases, (3) functional degradation of structural concrete by erosion, including basemat penetration, and (1) combustion of carbon monoxide.

Dllenges to the containment should not be sufficient to render Anoperable that equipment required for containment cooling or atmospheric cleanup, nor to cause leakage in excess of the rate specified in Criterion 16 or to allow any release through the basemat within an appropriate time of the onset of the corium-concrete interaction sufficient to cause significant contamination of the groundwater.

This is a far more complete and accurate description and require-ment than is contained in Alternative 1.

i OCRE recommends the incorporation of the description of severe i

accident phenomena and acceptance criteria from the ACRS letter into Alternative 1.

D.

The use of deliberate ignition to control combustible gases should be prohibited.

Reliance on ignition to control hydrogen requires that the igniters have a power supply, which is suscep-tible to failure.

If the power supply to the igniters fails and is restored at some later time, after the hydrogen has built up to higher concentrations, a severe deflagration or detonation could result.

The sudden condensation of steam in a

steam-i inerted containment could likewise result in high hydrogen con-i centrations which would then be ignited, with severe results.

Ignition of hydrogen adds the heat of combustion to the contain-ment atmosphere, can create pressure pulses which can damage the i

containment or equipment within it or resuspend fission products i

3 i

which had plated out on structural surfaces, and creates the potential for accelerated flames or deflagration-to-detonation transition.

The NRC should require containment atmospheres to be l

inert.

This may be accomplished through full pre-accident inert-

ing, or through partial pre-inerting or post-accident

inerting, or some combination of these two approaches, provided that the inerting mechanism is reliable. The potential for deinerting during the course of the accident must be eliminated.

E.

The calculation of containment stresses must be based on the as-built state of the containment.

The as-built containment will undoubtedly contain deviations from the perfect structure usually i

considered in such analyses.

For

example, the containment structure may be out-of-tolerance, out-of-round, or contain weld

(

joints with indications dispositioned "use-as-is".

The effects i

of these imperfections, which are probably acceptable for the loads resulting from the design basis LOCA, need to be evaluated for the higher pressures resulting from severe accidents.

The NRC should consider the potential for other failure mechanisms of steel containments, such as creep and low-cycle fatigue from multiple steam explosions or hydrogen deflagrations, should the NRC decide to allow deliberate ignition.

j d

i

Also, for containments using suppression pools, the NRC should l

require evaluation of the effects of hydrodynamic loads induced by steam explosions and'other rapid pressurization

events, in-i cluding hydrogen combustion pressure pulses, if the NRC decides l

to permit ignition.

I F.

The NRC should require all future containments to have a

i minimum design pressure of 45 psig and a minimum net free volume of one million cubic feet.

l The NRC should prohibit the use of pressure-unseating equipment hatches and personnel airlocks in containments.

A pressure-unseating equipment hatch, with the hatch cover mounted on the outside of the containment, has such a large surface area that the closure bolt preload is overcome at relatively low pressures.

Containment integrity is then maintained by elastomeric

seals, which may be degraded by the high temperature severe accident environment.

l The NRC should establish standards for containment penetrations (including but not limited to electrical, mechanical, hatches.and personnel locks, and purge / vent valves) which would require the use of penetrations types and waterials which have been demonstrated to be the most resistant to severe accident condi-4 1

)

tions.

RESPONSES TO QUESTIONS POSED IN THE NOTICE OCRE is responding to selected questions posed in the Federal Register notice.

The numbering corresponds to the numbers of the questions in the notice.

1.

As stated above, OCRE believes that this rulemaking is necessary and desirable.

A rule will provide better coherence and predictability to the design review and certification proc-esses than an ad hoc approach.

A rule will also enhance the NRC 's authority by requiring compliance with a rule rather than with non-enforceable guidance.

1 The ACRS, in its May 17, 1991 letter, provided a cogent explana-tion for why new containment criteria are needed.

The ACRS ad-vanced three reasons for the new requirements: to reduce risk and uncertainty, to clarify what is expected of applicants and to bring greater coherence to the regulatory process, and to in-crease the

" robustness" of containments.

The ACRS explained

' robustness" as follows:

"A containment cleverly and narrowly designed to mitigate a set of accidents that has been precisely identified may not be able to cope with the unexpected.

A truly

~robust' containment would have improved capability to deal with the unexpected.

A containment that has been designed with ex-plicit consideration of a more extensive set of challenges is likely to be more robust than one designed with consideration of only a limited set."

OCRE agrees with this assessment complete-ly.

2.

The rule proposed herein does not provide a basis for revis-ing emergency planning standards.

OCRE presumes that the NRC is contemplating the elimination of offsite emergency planning,

.or shrinking the plume EPZ to the site boundary, or some equivalent

concept, as has been advanced by some in the nuclear industry.

The present emergency planning requirements should be retained and strengthened.

Emergency planning provides an-important hedge against uncertainty.

Even with the additional requirements being

proposed, it is always possible that the containment will still fail.

We must have the emergency planning infrastructure in place if that happens. - Emergency planning also provides the last layer of defense in the event that plant safety features are deliberately defeated in acts of radiological sabotage or terror-ist attacks.

In addition, emergency planning yields significant collateral benefits to the community in that it is useful for

.a spectrum of natural and manmade disasters,

e. g. tornadoes, hurri-

canes, floods, chemical spills, nuclear attack, etc.

In

fact, 5

i h

many utilities use this as a selling point for emergency plan-ning.

Emergency planning ought to be in place for every communi-l ty, regardless of whether it is host to a nuclear power plant.

3.

As noted above, OCRE believes that the containment should be t

designed to remain leak tight for the duration of the accident.

In the event that the NRC chooses to permit containment

venting, it should only be permitted after one week from the onset of core i

damage.

As explained above, this is to allow sufficient time for fission product removal mechanisms to work, for recovery of the accident, and for offsite protective measures.

4.

The difficulties mentioned in this question are among the reasons OCRE does not favor Alternative 2.

.In the event that the r

NRC adopts Alternative 2, OCRE recommends that it include a more i

detailed description of the phenomena and that acceptance crite-ria be codified.

5.

OCRE believes that future containments should have additional I

featur'es and requirements.

Some of these have been addressed

above,

e. g., minimum design pressure and internal volume, prohi-bition of pressure-unseating hatches and airlocks, and require-ments for penetrations.

OCRE would also support the "supercontainment" concept.

That is, the containment must be designed to withstand any

accident, without consideration of probabilities.

If an accident or-phe-nomenon is possible, the containment is designed to withstand it.

Germany is currently investigating supercontainments for future reactors.

Initial cost estimates suggest small impacts (less than 5%) on total reactor costs.

This is partially due to the fact that current German containments are conservatively designed to withstand extreme aircraft accidents.

See Forsberg and Reich,

" Worldwide Advanced Nuclear Power Reactors with Passive and Inherent Safety:

What, Why, How, and Who," Oak Ridge National Laboratory, ORNL/TM-11907, September 1991.

l OCRE believes that U.S.

reactors should provide a level of pro-tection at least equivalent to that required in Germany and other countries.

The United States should be the world leader-in protecting the public from reactor safety hazards, not the lag-gard.

OCRE supports the ACRS proposal that containments should protect against aircraft crashes, explosions, and other threats external to the plant.

i L

The NRC should also require the use of a core catcher.

The 6

i I.

l I

l 1

following characteristics appear promising (taken from Forsberg i

and Reich, supra, p.

59):

I Under the reactor vessel, a portion of the concrete mat has 1

a specially controlled chemical composition.

The concra+.e con-tains a

mixture of different aggregates.

The aggregates are chosen so that when the various aggregates -cement, steel

rebar, and core materi als-melt, a waste glass that incorporates the core materials is created.

The glass contains one or more aggre-gates containing neutron poisons to prevent any possibility of a

criticality accident.

The glass chemical composition is chosen

)

to have a very high affinity for volatile fission products.

The aggregates are chosen to minimize gas generation upon melting

)

and, hence, minimize aerosol formation.

The glass also has a

high surface tension to minimize aerosol generation.

The depth and width of the concrete mat with the special i

concrete aggregate is chosen to contain the reactor core.

A heat l

balrace exists between radioactive decay heat and (1) heat needed

]

t' melt the concrete, and (2) heat conducted out or removed by ather mechanisms from the molten core / concrete matrix.

Eventual-ly, heat conduction out of the waste matrix will exceed heat generation and the molten core / concrete matrix will begin to solidify.

The special aggregate concrete mat is sized to exceed the maximum volume of the molten core / concrete matrix, and the area is chosen to maximize cooling.

In particular, the top surface area is large enough to radiate sufficient decay heat so that it will cool and solidify the waste matrix over time, with-out meltthrough of the reactor basemat.

The concrete aggregate is a relatively low-melting aggregate (400 to 900 deg-C).

Low melting points are desirable for the following reasons:

(1).

A low melting waste matrix will.quickly spread the molten core / concrete matrix over a wide area under the reactor.

This improves heat transfer and cools the matrix to quickly form a

solid.

(2).

A low melting waste matrix minimizes gas and aerosol gener-ation by two mechanisms.

First, the rate of release of semivola-tile radioactive gases is temperature dependent.

Lower tempera-tures imply less gas.

Second, the rate of release of semivola-tile radioactive gases is dependent on the concentration of thosa materials in the waste matrix.

Diluting the core material re-duces the fractional releases of radioactive materials.

[

7

i i

t 6.

OCRE does not believe that the likelihood of accident scena-rios or phenomena should play a role in determining.whether the containment should be designed to withstand them.

Our notions of

" likelihood" are too uncertain, incomplete, and subject to manip-ulation to be sufficiently accurate for regulatory purposes.

OCRE agrees with the ACRS ' assessment: "It is because quantits-tive risk estimates are not perfect that defense in depth is a

useful philosophy, and that separate containment performance guidelines make sense."

As stated above, OCRE believes that the challenges do need to be specified in more detail and with greater accuracy and complete-ness in the rule.

OCRE supports the incorporation of the de-scriptions of challenges and severe accident phenomena contained in the May 17, 1991 ACRS letter.

7.

OCRE sees no reason why the criteria proposed in the rule t

should not be fully applicable to passive LWRs.

9.

A reactor design should be required to have a reactor cavity design and/or a

reactor vessel support _ structure capable of mitigating and accommodating a high pressure melt ejection even if the design includes the capability to rapidly depressurize the primary system.

Why?

Because the depressurization system may fail or be deliberately defeated in an act of radiological-sabo-tage.

10.

OCRE agrees with the ACRS that there should be a provision for on-line monitoring of containment isolation status.

However.

if such a system is only capable of detecting

" gross leakage' then it is not a sufficient replacement for the leak rate testing requirements of Appendix J.

11.

OCRE agrees with the ACRS that containments should be de-signed to provide for ease of emergency closure during shutdown operation including station blackout conditions.

f 12.

Equipment provided only for severe accident prevention or mitigation should be subject to the same requirements as design basis equipment.

It hardly makes sense to require such equipment only to have it fail when needed because it was not subject to a

nuclear quality assurance program or was not environmentally l

qualified to withstand the very conditions in which it is re-quired to function.

Since severe accidents pose a greater rick to the public than design basis accidents, it makes little sense to have more stringent requirements for those accidents posing the lesser risk and less stringent requirements for those acci-dents posing the greater risk.

F 8

i i

13.

OCRE believes that the ASME service level C stress limits for steel containments are appropriate for severe accident condi-tions.

Use of the level C limits is necessary to provide suffi-cient conservatism to account for all the uncertainties inherent in calculating such stresses.

For an excellent discussion of these uncertainties, see "On the Uncertainties Associated with Containment Analysis," Griemann and Fanous, NUREG/CP-0056, pp. 229-244.

15.

The codification of a containment performance objective does not provide a basis for the elimination of further review of SAMDAs for future LWRs under Part 51.

Consideration of SAMDAs is necessary because our knowledge of severe accident risks and I

phenomena will undoubtedly increase with the passage of time, and design innovations will undoubtedly be developed.

For

example, just 10 years shutdown risk was thought to be virtually nonexistent.

The SAMDA review process is essential for the rational evaluation of risk and mitigation measures that may be unique to a plant design.

Respectfully submitted, jnd i

/

Susan L. Hiatt Director, OCRE 8275 Munson Road Mentor, OH 44060-2406 (216) 255-3158 9

h7 Y *.kh.$bbb D %':L3 b T 0 - -

. g.

Cr n R w s)

AECL AECLTechnologies

'92 DE 24 W. :06 9210 corporate Bouievara Suite 4i0 Rockvdle Maryland 20850 USA 1-800-USA. AECL (301) 417 0047 Fax (3011417-0746 i

Telex 403-442 i

File: 33-0002-122 December 21,1992 l

Mr. Samuel J. Chilk Secretary of the Commission U.S. Nuclear Regulatory Commission Washington, D.C. 20555 Attention: Docketing and Service Branch Re:

Advanced Notice of Proposed Rulemaking: Acceptability of Plant Performance for Severe Accidents; Scope of Consideration in Safety Regulations [57 Fed. Reg. 44513, (September 22, 1992)]

Dear Mr. Chilk:

AECL Technologies has reviewed the subject advance notice of proposed rulemaking. We appreciate the opportunity to review this document and to contribute our views for Staff and Commission consideration. AECL Technologies supports initiation of rulemaking to resolve severe accident issues generally prior to commencement of design certification proceedings to the extent practicable.

Sincerely, I

A. D. Hink rice President / General Manager AECL Technologies

"*""'5" JAN 21 :93 Mmvw ; 2: by ca.t.......................

A Deseen c4 AE CL tag

A 6

= *.-

• d--

u--

E-3

. bec:

J. Kennedy, NRC M. Muntzing, Newman & Holtzinger M. Bonechi, SP-2

- R Durante FI, Feinroth L. Rib i

R. Ferguson R. Curtis M. Fletcher

. i 9

6 a

t p

4 4

I 8

?

i

. I 1.

i i

h i

l

,e

.a, e

December 21,-1992 l

4 i

AECL TECHNOLOGIES (AECLT) 1 1

COMMENTS ON THE NUCLEAR REGULATORY COMMISSION'S (NRC) i ADVANCE NOTICE OF PROPOSED RULEMAKING (ANPR) t CONCERNING ACCEPTABILITY OF PLANT PERFORMANCE FOR SEVERE i

ACCIDENTS; SCOPE OF CONSIDERATION IN SAFETY REGULATIONS i

[57 FED. REG. 44,513 - (SEPTEMBER 28, 1992)]

t I.

INTRODUCTION 1

The ANFR states that one purpose of a severe accident rule would be to " Provide assurance that the performance of future LWRs under severe accident conditions is consistent with assumptions about severe accident performance used. in i

developing new source term information. "

57 Fed. Reg. 44,514.

.)

To achieve this purpose, AECLT believes that the rule 7should j

establish comprehensive requirements applicable to all reactors and that Regulatory Guides should be developed which provhagfjc_ guidance concerning ways of meeting the requirements for specific reactor systems or reactor $ types.

These guides should become available in draf t form at the time a proposed rule is issued.

Provided herein are AECLT's comments on the ANPR, as well as responses to those questions in the ANPR applicable to Pressurized Heavy Water Reactors (PHWR) such as the CANDU 3 reactor.

Since questions 8, 11, 13 and 14 are not applicable to PHWRs, no responses to these questions are provided.

I.

GENERAL COMMENT

S:

1.

The overall criteria for protecting the public frem severe accidents should be the same for all water-cooled reactors.

A severe accident rule should specify these overall criteria.

AECLT. believes, the_ format of such a rulo should.-be similar to. themfomac described as i

Alterw'isthe ANPR. Adoption of this format wou'd i

encourage-designer flexibility and inventiveness in :te incorporation of severe accident prevention and mitigation features to reduce the-frequency.

and consequences of such accidents.

2.

The ANPR indicates that the criteria discussed in tnis ANPR would codify much of the Commission's guidance f:r general application to all future LWRs.

AECLT believas that this guidance would also be applicable to PHWRs j

Presently, the Imc is conducting a preapplication rev:-w of the CANDU 3 design.

In conjunction with NRC's rev:aw of CANDU 3 severe accident prevention and.mitigat: n design features, AECLT has prepared at NRC's request.

e 1

l n

December 21, 1992 comparison of the CANDU 3 features to the NRC Staff's recommended criteria in SECY-90-016, as modified by Commission

guidance, concerning severe accident prevention and mitigation in LWR designs.

Based on this comparison, AECLT concludes that the CANDU 3 design will conform with the SECY-90-016 recommendations and guidance.

P 3.

The ANPR discusses three potential alternatives for design requirements related to prevention and mitigation of severe accidents.

Alternative 1 would prescribe hardware requirements to address risk-significant phenomena.

Alternative 2 would require designers to address risk-significant phenomena in the design, but would not prescribe specific hardware requirements.

Alternative 3 would specify General Design Criteria to describe the nature of the severe accident challenges as well as associated success criteria.

From the description of each alternative in the ANPR, AECLT cannot tell whether the alternatives are intended to beiequally comprehensive in scope.

AECLT believes that, regardless of the format adopted for the severe accident rule, the rule and accompanying guidance concerning implementation of the rule should be comprehensive in scope and should address the following matters:

(a) criteria for establishing event sequence frequencies; (b) radiological consequence limits; (c) capacity and reliability of the design feature; and (d) criteria to establish load combinations and environmental conditions.

Additionally, implementing regulatory guidance should address redundancy, diversity, power supply, equipment survivability, analytical methods, and acceptance criteria.

4.

Specifically, in the rule and implementing guidance the following matters should be addressed:

A.

iBelection Process for Severe Event Secuences

. Considered in the Desion. The selection process should be based on event frequency.

The process would establish the frequency limits to: (1) define the events requiring design changes to reduce their frequency, (2) define the events that require features to mitigate the event's consequences and (3) define events that need not be considered in the design.

B.

Consecuence Limits: For each event sequence defined by A (1) and A(2) above (e.g. reactivity events, loss of heat sink at High/ Low Pressure),

acceptable 2

December 21, 1992 t

consequences for the event frequency should be defined on an overall basis (e.g. containment stress and leakage, radiological consequence limits).

In addition, a phenomenon. acceptance criterion should define the acceptable consequences for each individual phenomenon (e.g. hydrogen, molten fuel, non-condensable gas) associated with the event consistent with the overall acceptance criteria and the design features that produce the phenomenon.

C.

Phenomenon Accentance Criteria: For each phencmenon acceptance criterion, systems / features should be identified which provide the me ns to mitigate the consequences of the phenomenon.

D.

System / Feature Desian Criteria:

For each system / feature, design criteria should be established for

capacity, load combinations,

~

environmental conditions vs time, and reliability.

The reliability criteria should include: redundancy, diversity, power supply, separation (from each other and from systems / features whose failures are involved in the severe accident event sequences),

and environmental qualifications.

E.

System / Feature Demonstration Recuirements: For each system / feature, the demonstration analysis / test requirements should be defined.

These should include assumptions, acceptance criteria, analytical methods, and test requirements.

5.

For a criteria-oriented rule, similar to Alternative 3, I

which AECLT favors, items A and B above should be included in the rule; items C, D,

and E above should be included in a Regulatory Guide.

6.

Because each reactor type may have some unique requirements, AECLT suggests that the rule be structured in two parts. The first part should present the overall requirements to be met by all reactors.

The second part I

would have separate sections for each reactor type (i.e.,

)

Pressurized Water, Boiling Water, Pressurized Heavy Water) PLWR, BLWR, PHWR (i.e.

CANDU) ).

Other reactor types (e.g.

sodium and gas cooled reactors) could be added at a later date.

7.

A severe accident rule ideally should be of sufficiently comprehensive scope to permit severe accident closure I

determinations to be made for new designs.

AECLT believes that a rule and implementing guidance of the 3

i December 21, 1992 scope described above in paragraph 4 would permit these determinations-to be made.

Issues identified in 4 which are not addressed in a severe accident rule will have to be addressed in individual Standard Design Certifi:ation rulemakings or in COL proceedings.

8.

As discussed in 3 and 4 above, a severe accident rule i

should specify a cut-off event frequency such that events below this frequency need not be considered in the design and for which further analysis is not required.

i NUREG/CR-5368,

" Reactivity Accidents"

- reported the results of analyses of light water reactor reactivity events performed by Brookhaven National Laboratory.

For i

that

effort, Brookhaven categorized potential event-sequences as being worthy of further analysis, or not.

One of the screening criteria used to determine the importance of a sequence for further analysis was whether the sequence required too many low probability events to occur in combination.

Brookhaven established a screening l

methodology with which low probability events could be-eliminated from further consideration.

Event sequences with a treguency of less than 1E-7 per reactor year were consicered

" incredible" and nct recommended for further study.

AECLT believes that the generic severe accident ru.-

should codify similar screening criteria.

4

~ December 21, 1992 II.

ANSWERS TO NRC'S SPECIFIC QUESTIONS Ouestion 1 Is a rulemaking addressing severe accident plant performance criteria desirable?

If so, why?

If not, why not? Would a rule provide better coherence and predictability to the design review and certification processes for future reactor designs or is rulemaking on these issues via individual design certification sufficient?

Response 1 t

AECLT believes a rule establishing generic severe accident criteria and plant performance criteria to prevent-and to mitigate severe accidents is desirable.

AECLT prefers that the ru3 e be in the i

format described by Alternative 3 in the ANPR.

As discussed in comment #4 above, the rule should establish the event frequency bounds, design criteria and radiological consequences associated with severe accidents.

These criteria should be independent of reactor type.

Such a rule would provide predictability in the certification process.

It would provide assurance to the publ.ic that the individual design certifications would be consistent with respect to degree of protection provided from stvere accidents.

t The rule should be supplemented by Regulatory Guide (s) which 1

identify the phenomena identified to

data, the acceptable systems / features to cope with phenomena, and the acceptance criteria

.l for such systems.

The supporting Regulatory Guide (s) should be issued at the same time

-l as the rule.

Question 2 l

Would a new rule in 10 CFR part 50, concerning plant performance for severe accidents, as discussed in the three alternatives, provide a basis for revising the requirements on Emergency Planning Zones for future LWRs?

If so, why?

If not, why not?

Response 2 I

i The three alternatives discussed in the ANPR do not address the

}

offsite radiological consequences of a severe accident; therefore, they do not provide an adequate basis for revising the requirements on Emergency Planning Zones.

As discussed in our Answer Oc Question 1, AECLT believes that the severe accident rule should address such consequences.

If the rule does so, the rule would provide a basis for EPZ simplification for all future reactors (born LWRs and PHWRs) encompassed by the rule's scope.

5 2

l l

i

December 21, 1992 I

Ouestion 3 One option for an overall containment performance criterion that has been considered is that the conditional f ailure probability.of the containment shculd be less than approximately one in ten.

Two of i

the alternatives use a deterministic surrogate that states that the containments should remain leak tight for a period of-approximately l

24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the onset of core damage and after that time remain a barrier against the uncontrolled release of radioactivity when f aced with challenges from the more likely severe accident phenomena.

Is this criterion a suitable substitute for the conditional containment failure probability of one in the ten?

If so, explain why.

If not, explain why not.

Is a period of approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> an appropriate time frame? Is its degree of conservatism appropriate considering uncertainties and defense-in-depth?

If not, what alternative would be appropriate?

What other criteria (probabilistic or deterministic) might be considered?

Response 3 i

Because of the wide range of the types of challenges to containment that may result from severe accident events, the NRC should not i

specify any specific criteria for evaluating these challenges in the rule.

Instead, general criteria concerning event frequency and 3

radiological consequences should be in the rule.

Design specific l

criteria should be in the Regulatory Guides.

The applicant should l

provide the traditional justification for the analysis of containment performance during severe accident events.

l Ouestion 4 Alternative 2 would require extensive reliance on analytical tools i

that calculate the affects of severe accident phenomena.

Are there t

I analytical tools that are suf ficiently developed and adequate to l

allow effective implementation of such a phenomena-based rtGe?

If so, what are they, and for what phenomena could they be used?

How would alternative 2 be implemented? For example, should the codes and input parameters be approved by NRC? Should acceptance criteria be codified or put in a regulatory guide?

l Response 4 l

Alternative 2 may dampen innovative approaches to the prevention and mitigation of severe accidents.

Alternative 3 would not be so dependent on the state of technology and so difficult to change ::

incorporate the results of ongoing research programs.

6 I

'l l

December 21, 1992 Ouestion 5 Should future LWR containment designs include features beyond those described in alternative 1 to prevent / mitigate severe accidents?

If so, what are they?

i i

Resnonse 5 AECLT believes that Alternative. is unnecessarily restrictive. By I

codifying specific design requirements based on current knowledge, Alternative 1 does not allow for alternative designs.

This is-an l

impediment to innovation based u1on increased understanding of alternative technologies. Alterna'.ve 3 would codify the acceptance criteria and permit innovative / signs to meet those criteria.

i Ouestion 6 Alternatives 2 and 3 specify phenomenological severe accident challenges that should be considered in the design. Alternative.1 l

is based upon the same phenomena / challenges. Are there other severe accident phenomena / challenges that should be considered?

What i

should the criteria for. deciding whether a severe accident phenomena l

or challenge is likely and should be considered?

Should the j

challenges be specified in more detail (for example, specifying the

.1 amount of hydrogen generation) or is a general statemenc of the j

challenge more desirable?

.j l

Resoonse 6 As discussed above in General Comment 4, the criteria for deciding whether a phenomena or challenge should be considered in the design should be based on the event sequences to be considered in the design and the phenomena they produce.

This requires a systematic j

review of the plant for potential events and an analysis of their event frequency and their phenomena. As discussed above in General l

Comment 8,

AECLT believes that ' phenomena associated with event l

frequencies less that 1E-7 should not have to be considered.

Qyestion 7 i

For what reason (e.g. not a risk significant phenomena, not a cost i

effective solution) would any of the criteria proposed in the three alternatives not be fully applicable to passive designed LWRs?

l Response 7

}

i Alternative 1 may be design-dependent, as may Alternative 2.

Alternative 3 would be independent of specific reactor designs and, therefore, would be applicable to passive designs.

i 7

I i

December 21, 1992 Ouestion 9 If a design includes the capability to rapidly depressurize the primary system, should it also be required to have a reactor cavit /

design and/cr a reactor vessel support structure capable cf mitigating and accommodating a high pressure melt ejection?

Response 9 The need for either preventing or accommodating a high pressure melt ejection should be established on the frequency limit for the events, that should be considered in design.

If the f requency 1: mit can be met or exceeded with system (s) that prevent this event, it should not be necessary to accommodate the event.

It is more prudent to design to prevent this event rather tIan design to accommodate this event. The design of preventive syste is, (i.e. depressurization systems, power supply, feedwater, etc.) is straightforward. The design of accommodation systems is speculat; ve because the conditions of the high pressure melt ejection tre uncertain.

Question 10 Should future LWR designs include an on-line instrumentation syst em that monitors containment atmosphere for gross leakage to reduce t ae risk from an inadvertent bypass of containment function?

WotLd application of this system be sufficient basis to modify leak ra:e testing requirements under 10 CFR part 50, Appendix J,

" Primary Reactor Containment Leakage Testing for Water-Cooled Powtr Reactors."

Response 10 In the CANDU 3 design, containment air pressure and temperature along with other data, are monitored while the plant is operating to provide a timely indication of any gross breach of containment.

The provision of a gross leakage monitoring system should be a sufficient basis to modify the requirements of Appendix J.

Question 12 Should equipment provided only for severe accident prevention or mitigation be subject to (a) the same requirements as design basis equipment (e.g. redundancy / diversity, power supply, environmental qualification, inclusion in plant Technical Specifications, 8

December 21, 1992 maintenance priority, quality assurance); or (b) lesser standards

( e.. g., reduced design margins or the regulatory guidance found in-appendices A and B of Regulatory Guide l'.155, " Station Blackout?").

If lesser standards, what standard would be appropriate?

Response 12 The question appears to suggest only two alternatives for requirements; however, there is a third alternative that considers the nature of the design feature, its safety function and the conditions under which it should operate.

The requirements for severe accident prevention or mitigation equipment / features should be appropriate for the specific feature, the time-history of the conditions associated with the-event, and the desired reliability goal for the equipment / feature.

For example, the hydrogen igniter system, depressurization systems and heat removal systems would have different requirements from the reactor cavity and basemat.

Ouestion 15 The containment performance objective discussed in Alternatives ;

and 2 (i.e. containment shall provide a barrier against the release of radioactive material for a period of approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the onset of core damage) represents a level of safety for a 3800 Mwt plant cited in accordance with 10 CFR part 10:

approximately three orders of magnitude below the Commission s-Safety Goal Policy Statement.

It could be argued that a future LWF design meeting this objective through analyses and the incorporati:n of design features need not consider the addition of other features.

since-these other_ features would be directed at even more high;.

unlikely severe accident phenomena and sequences which could :-

considered " remote and speculative" under the National Environment a.

Policy Act (NEPA) and 10 CFR part 51.

Therefore, would tna codification and compliance with such a contain.nent performanc-objective be sufficient to also define a point of truncation an:

serve as the basis for an amendment to 10 CFR part 51 eliminatin:

the need for further review of SAMDA's for future LWRs under 10 CFL part 51?

Response 15 Regardless of the rule format (Alternative 1,

2-or 3), the ru -

should be sufficiently definitive to eliminate the need for furtn-:

review of.SAMDAs-for future LWRs and PHWRs under 10 CFR Part The approach described in General Comment 4 is of sufficient sc ;-

to permit severe accident closure under NEPA for designs meeting * *. -

requirements.

The rule should include a determination to tnr 9

.l December 21, 1992 J

[

effect so that the issue cannot be raised successfully in a-design certification or COL proceeding.

\\ar.rp. tf

.l i

5 I

i h

f F

?

l L

I i

?

i 9

i i

i I

.i

.?

l r

'?

?

?

I i

i i

10

?

?

I

A& qr (5 7 F/2 WS /3)

~T 5

r,. r.n. n ALWR n-: "

• 3 L4.hs.

d ADVANCEO t ICHT W ATTR RF ACTOR 23 December 1992 Mr. Samuel J. Chilk, Secretary Office of the Secretary of the Commission U.S. Nuclear Regulatory Commission Mail Stop 16 GIS Washington, DC 20555

Subject:

Advanced Notice of Proposed Rulemaking (ANPR),

" Acceptability of Plant Performance for Severe Accidents"; Scope of Consideration in Safety Regulations" (57 Federal Register 44513 of September 28,1992)

Dear Mr. Chilk:

The ALWR Utility Steering Committee assisted by the Electric Power Research Institute (EPRI) has reviewed the ANPR, " Acceptability of Plant Performance for Severe Accidents; Scope of Consideration in Safety Regulations," and has provided its input to NUMARC for incorporation in their letter forwarding overall industry comments. We agree with and support the NUMARC letter.

In addition, we wish to identify specific concerns regarding the ANPR that relate to the role of the ALWR Utility Requirements Document.

EPRI and NUMARC both commented in 9 January 1989 letters to the NRC on staff proposals to initiate severe accident rulemaking for advanced reactors.

The Utility Steering Committee (USC) reiterated our concerns in a letter to the Chairman on 4 May 1990. Key points included the following:

Rulemaking is neither necessary nor desirable to implement the NRC's Severe Accident Policy Statement.

The ALWR program is committed to meeting the letter and intent of the Severe Accident Policy.

0 33 AcknCW 3d;60 by CUTd.JAN.

-.. --.- - ~..

EPfU ALWR Utility Steering Committee 3412 Hillview Avenue, Palo Alto, California 94304

• Telefax: (415) 812-2874

~

Mr. Samuel J. Chilk

~

21 December 1992 Page 2 The ALWR Utility Requirements Document (URD) is the appropriate s

mechanism for translating industry and NRC policy into a specific set of baseline design requirements for ALWRs.

The ALWR URD with its attendant NRC Safety Evaluation Report followed by design certification.will achieve the necessary resolution and codification of severe accident issues for future designs.

The steps taken in the Requirements Document toward reducing the frequency and severity of core damaging accidents will reduce the probability of accident measures being needed. Nevertheless, all the severe accident issues and approaches to resolution were already identified and addressed by the Evolutionary Plant URD at the time generic rulemaking was first proposed.

No significant new severe accident issues have emerged, and no major changes have occurred in our utility requirements on how these severe accident issues should be addressed in the designs.

From what we have i

learned to date, no significant changes in the design of either the ABWR or t

System 80+ have been made and we do not expect any significant changes in Passive Plant requirements or designs.

Specifically, the " Purpose of the Rule" as stated in the ANFR can be achieved without an intervening generic Rule as follows:

" Codify the Commission's guidance on severe accident and containment issues that resulted from the review of advanced light water reactors" (achieved via the Design Certification rulemaking]

" Provide assurance that the performance of future LWRs under severe accident conditions is consistent with assumptions about severe accident conditions performance used in developing new source term information" (achieved via approval of the URD and the FDA process}

" Provide guidance to future LWR designers and potential applicants" (this is done via NRC input to and industry conformance to the URD)

" Add consistency and standardization to the resolution of severe accident issues based on the current technical information" (this is done via the conformance to the URD) i

" Facilitate design certification rulemakings" (URD approval already accomplishes this; generic rulemaking will conclude too late to affect}

Intervening severe accident rulemaking cannot facilitate DC rulemaking for evolutionary plants because DC rulemaking will likely be completed well

l l

Mr. Samuel J. Chilk 21 December 1992 i

Page 3 1

before the generic rulemaking process could conclude.

Intervening rulemaking for passive plants will not be needed once agreement and approval i

of the Passive Plant URD are obtained.

Intervening severe accident l

rulemaking (already in development for over five years) will become a major source of unnecessary schedule delay in the ALWR design and approval j

process. This continuing regulatory instability would further erode investor confidence in the nuclear option and place a large drain on the already strained NRC resources.

We understand that the Commission gave tentative approval for proceeding with generic rulemaking "where appropriate" in its SRM on SECY-92-262. We believe the policy and schedule implications of the current path are so severe j

that a Commission review of the appropriateness of this path is needed.

i i

Sincerely, E. E. Kintner, Chairman i

ALWR Utility Steering Committee j

576L/TUM/cdl 1

c Chairman Ivan Selin i

Commissioner Kenneth C. Rogers Commissioner James R. Curtiss Commissioner Forrest J. Remick Commissioner E. Gail de Planque j

James M. Taylor, Executive Director of Operations, NRC j

Thomas E. Murley, Director, Nuclear Reactor Regulation, NRC Docketing and Service Branch, NRC John J. Taylor, Vice President, EPRI Joe F. Colvin, President and CEO, NUMARC i

1

-n,..

,~

nn

~,,,,,

l. ' M ' l C. '

V W l'"

,,,,.S.

1 ?.UlC w

._..- _ k [ 0 W

!.* T

~

m.

~

(51 FR M5 0) 3 -

3, Westinghouse Energy Systems W DEC 28 P4 :16

"" 355 Electric Ccrporation Anmv PevsvNania 1523C C355

~

.f ET-NRC-92 3788 NSRA-APSL.92-0269-Decemb< r 22,1992 "Mr. Samuel J. Chilk The Secretary of the Commission U.S. Nuclear Regulatory Commission Washington, D.C. 20555 A'ITENTION:

DOCKETING AND SERVICE BRANCH

SUBJECT:

" Acceptability of Plant Performance for Severe Accidents; Scope of Consideration in Safety Regulations" (57 Fed. Reg. 44513, September 28,1992) - Advanced Notice of Proposed Rulemaking (ANPR)

Dear Mr. Chilk:

The purpose of this letter is to provide Westinghouse Electric Corporation comments on the subject proposed rulemaking.

On June 26,1992, Westinghouse submitted to the NRC an application for final design approval (FDA under Appendix 0 to 10 CFR Part 52 and a Standard Design Cenification (DC) under 10 CFR Pan 52 for the AP600 plant design. As one of the three vendors with Advanced Light Water Reactor (ALWR) designs currently under review by the NRC, we have carefully reviewed the subject ANPR with particular focus on the impacts the proposed rulemaking would have on the pursuit of the FDA and DC for the AP600 design.

Westinghouse believes that generic rulemaking to address severe accident issues for future plants is neither necessary nor desirable. Contrary to a stated purpose of the proposed rule, we believe that the proposed rulemaking will not " facilitate design certification rulemaking" but would instead increase uncertainty with little or no added value. To pursue generic rulemaking for severe accidents at this point in the AP600 Design Cenification effon would at best result in a duplication of effon, as well a a duplication of regulation and guidance. The intrusion of such rulemaking at this time introduces substantial uncenainty and confusion into the Pan 52 licensing process, and could easily jeopardiec design certification for the AP600 as well as other advanced plants.

ALWR severe accident issues are presently being addressed by the industry and NRC through the ALWR Utility Requirements Document (URD) and design cenification interactions. The utility requirements developed for both evolutionary and passive plant designs address a wide spectrum of severe accident phenomena and containment challenges. The Westinghouse AP600 is designed to meet the utility requirements for passive ALWR designs. As such, we believe that NRC review ot ihc JM,

300'1

~

.,.,.: ww===

~

onu

ET NRC-92-3788 December 22,1992 NSRA-APSL-92-0269

~

URD, as well as the design certification review of the AP600 design, provides the preferable method for resolution of severe accident technical issues. This process allows for generic resolution of severe accident issues to the extent possible via the ALWR URD and review and codification of design specific implementation as part of the AP600 design certification process.

We are particularly concerned that a generic rulemaking for severe accidents separate from the design certification process will significantly disrupt the design certification pmcess being pursued for the AP600 with the potential for causing costly and unwarranted delays. As noted in the ANPR, *this rule would be generally applicable to passive LWR designs. However, as detailed design information becomes available and review of the passive systems is completed, further rulemaking may be necessary." The implication is that the net result of this potential series of generic rulemakings is a rule which would reflect the information obtained through the specific passive plant design certifications. Thus there is no need for a separate rulemaking at this time. We expect the Design Certification application for the AP600 to be docketed within the next month (a docket number has already been assigned). The review of the AP600 design by the NRC staff is already in pmcess and we expect the AP600 FDA to be issued in mid to late 1994. This time frame does not lend itself to the process envisioned in the ANPR.

Westinghouse has participated in the preparation of industry comments on the proposed rulemaking and we support the comments being offered on this subject by NUMARC.

We appreciate this opportunity to comment on the proposed rulemaking and urge that the Commission not proceed with the proposed generic rulemaking for severe accidents. As discussed above, we believe the NRC staff should continue to work toward resolution of severe accident issues for advanced plants via the design certification process.

Very truly yours,

/Y g

N. J. Liparuto, Manager Nuclear Safety and Regulatory Activities

/nja ec:

The Honorable Ivan Selin, Qiairman Commissioner Kenneth C. Rogers Commissioner James R. Curtiss Commissioner Forrest J. Remick Commissioner E. Gail de Planque Mr. J. Taylor EDO em

i q-

.'Og,,,

KM @M-. j O

w e,c= =%

.q, _

?j'{TF (57 PR '-i'is 1)]

l

??'3E!ED NUCLEAR MANAGEMENT AND RESOURCES COUNCIL

1. 'C I

1776 Eve Stee? N W e Suce 300 e was%ngm CC 2000624% g y,, 3

,,,,3 (202)372 '280 Wittiam H. Rosin vce Lescem? 6 Lrec'cr

'9CPCO nion

~

December 22,1992 i

Mr. Samuel J. Chilk Secretary, Office of the Secretary of the Commission U.S. Nuclear Regulatory Commission i

Mail Stop 16 G15 Washington, DC 20555

SUBJECT:

Advance Notice of Proposed Rulemaking (ANPR), " Acceptability of Plant Performance for Severe Accidents; Scope of Consideration in Safety Regulations" (57 Fedeml Register 44513 of September 28,1992)

Dear Mr. Chilk:

I The Nuclear Management and Resources Council (NUMARC)', on behalf of the l

nuclear power industry, has reviewed the ANPR, " Acceptability of Plant Performance for Severe Accidents; Scope of Consideration in Safety Regulations," and offers the following i

comments for consideration.

The industry has long been aware of NRC consideration of the need for generic '

rulemaking for advanced light water reactors (ALWRs) to supplement specific design l

certification rulemakings. In a January 9,1989, letter to NRC Executive Director for Operations, Victor Stello, NUMARC provided its perspective, which we have maintained

{

consistently since, that generic Part 50 miemaking to address severe accidents, such as

[

that contemplated by the subject ANPR, is not necessary and,in fact, may be counter-productive. Instead, the industry has supported generic resolution of severe accident I

2NUMARC is the organization of the nuclear power industry that coordinates the i

combined efforts of all utilities licensed by the NRC to construct or operate a nuclear i

power plant, and of other nuclear industry organizations, in all matters involving generic j

regulatory policy and on the regulatory aspects of generic operational and technical issues that affect the nuclear power industry. Every utility responsible for constructing and operating a commercial nuclear facility is a member of NUMARC. In addition, i

NUMARC's members include major architect-engineering firms and all the major steam supply vendors, i

J M 2 ; 333 Ackn:wiecgsa by card..,,,,,,,,,,,,,,,,,,,,,,,,,

9

1 Mr. Samuel J. Chilk December 22,1992 Page 2 issues to the extent possible via the ALWR Utility Requirements Document (URD),

followed by review and codification of the design speci6c implementation as part of the design certification process. This position was explained to the Comnussion and NRC staff on several occasions, including our January 27,1992, letter to the Office of the Secretary commenting on SECY-91-262, Resolution of Selected Technical and Severe Acciden: Issues for Evolutionary Light Water Reactor Designs. We continue to feel strongly in this regard, and this letter and the enclosed responses to the fifteen ANPR questions reiterate our position and provide the basis for our conclusion.

We believe the URD/ design certification process is the superior approach for the resolution of severe accident issues for ALWRs, both evolutionary and passive. A full discussion of this position is delineated in response to ANPR Ouestion Number 1 contained in the enclosure, the key points of which are as follows:

I severe accident issue resolution via the URD and design certifications is consistent with the approach embodied by the Commission',s Severe Accident Policy Statement and NRC Generic Letter 88-20, Individual Plant Eraminations for Severe Accident Vulnerabilities; the URD, together with design certification reviews and rulemakings, provide the optimal approach for comprehensive, imegrated and design specific resolution of all safety issues, including severe accidents, associated with individual standard plant designs; severe accident resolution accomplished through the URD and design certification will avoid uncertainty associated with the impact of a

" competing" generic rulemaking, thus promoting predictability and stability in the Part 52 licensing process; and there is minimal value added by one or more generic rulemakings that produces severe accident resolution essentially the same" as that which will be accomplished via design certi5 cation.

1 A separate generic rulemaking on technical issues associated with ALWR designs, such as severe accidents, is not warranted. Rather than facilitating the design certification process, as indicated in the ANPR, generic severe accident rulemaking would more likely complicate the process and unnecessarily consume signi5 cant NRC and industry resources. Further, this approach would be inconsistent with the initiatives of the NRC and the industry to identify unnecessary, contradictory and/or duplicative regulations and guidance.

f i

Mr. Samuel J. Chilk December 22,1992 Page 3 We appreciate your careful consideration of these comments and urge the Commission to direct that severe accident issues be addressed via the URD/ design certification process rather than move ahead with a proposed severe accident rule for future plants.

Sincerely, py. -

t

....?

s.

=

m William H. Rasin WHRhRJB\\ljw Enclosure cc:

Chairman Ivan Selin Commissioner Kenneth C. Rogers Conunissioner James R. Cuniss Comnussioner Forrest J. Remick Commissioner E. Gail de Planque James M. Taylor, Executive Director of Operations, NRC Eric S. Beckjord, Director, Office of Nuclear Reactor Research Thomas E. Murley, Director, Nuclear Reactor Regulation, NRC Docketing and Service Branch l

__a--.-_--

--.---a-_-------------------a-----.-----

Responses to ANPR Ouestions i

ANPR Ouestion Number 1 Is miemaking addressing severe accident plant performance criteria desirable? If not, why? Would such a rule provide better coherence and predictability to the design cenification review and certification processes for futare reactor designs or is rulemaking on these issues via individual design certification sufficient?

Industry Resoonse We believe that a generic Part 50 rulemaking to address ALWR technical and severe accident issues would not be appropriate and is neither necessary nor desirable.

A substantial amount of severe accident research has been accomplished by the NRC as well as the industry since the Three Mile Island Unit 2 (TMI-2) accident in 1979. Indeed, the industry has incorporated into the ALWR development the lessons learned since TMI-2 through development by EPRI of the ALWR Utility Requirements Document (URD) and vendor development of individual ALWR designs, including design specific PRAs. ALWR severe accident issues are presently being addressed by the industry and the NRC staff via complementary URD and design certification interactions. Utility requirements have been developed for both evolutionary and pass:w ALWRs, including explicit consideration of a wide spectrum of severe accident phenomena and potential challenges to containment [1,2]. These challenges encompass containment bypass scenarios, random system and equipment failures which could lead to breach of the containment boundary independent of any severe accident conditions, and potential phenomena that could challenge the structural integrity of the containment as a result of a core damage accident. The challenges considered include not only those referenced in the first two alternatives of the ANPR, but also the eight groups of challenges proposed by the ACRS for inclusion in the General Design Criteria [3).

References 1 at.J 2 identify the extensive list of challenges considered in the evolutionan and passive ALWRs and contain a summary of the design features available in each plant to limit the potential for or accommodate each of these challenges. Design speafic PRAs are being performed which evaluate the potential for significant challenges given j

these design features and address plant response to those challenges that dominate risk i

Individual ALWR designs that implement the udlity requirements and reflect the insights i

of a design specific PRA are being reviewed by the NRC staff and will become codified i

via design certification rulemakings. In sum, the industry and the NRC staff have

)

pursued a course emphasizing generic resolution of severe accident issues to the extem possible via the ALWR URD followed by review and codification of the design speafic implementation as part of the design certification process.

1

i Consistent with this approach, the NRC staff argued persuasively in SECY-91-262 that individual design certification rulemakings provide the most efficient and effective mechanism for codifying the resolution of technical and severe accident issues. As noted in SECY-91-262, the design certification process provides for a thorough NRC staff.

ACRS and Commission review of ALWR design requirements and criteria leading to meir codification via the design certification rulemakings. Moreover, addressing severe udent issues in the context of the URD as well as within the overall design t

certification process provides greater conSdence that these complex and interdependent technical issues will be coherently resolved. We concur in the essential thmst of SECY-91-262 that, considering the advanced state of the ALWR designs and associated reviews, the diversity among the standard plant designs, and the potential for delaying the design certification process, generic rulemaking to address ALWR technical and severe accident issues is not necessary or desirable.

Not only is resolution of severe accident issues via design certification amenable to evolutionary designs, as recommended in SECY-91-262, we believe severe accident resolution via design certification is the appropriate, superior and most technically coherent course for passive designs as well. Indeed, an approach for passive plants that instead relies upon generic miemaking would be contrary to the rationale behind the NRC staff issuance of Generic Letter 88-20, Individual Plant Examinations for Severe Accident Vulnerabilities, which recognized that severe cccident challenges and resolutions are somewhat design specific. A generic rulemaking on severe accidents, especially one based principally upon evolutionary plant designs as suggested by the ANPR, may not address the particular performance characteristics of passive plant designs or allow for the realization of the full regulatory benefit in their regard. As a result, a generic rulemaking would likely not obviate the need to scrutinize severe accident issues in detail as part of a passive plant design certification. Therefore, we conclude that generic technical issues, such as severe accidents, can be most efficiently and coherently resolved for all ALWRs, including passive designs, via the URD and design certification reviews and rulemakings.

Furthermore, we are particularly concerned that promulgating a rule based upon this ANPR would likely generate uncertainty and could significantly disrupt the design certification process for both evolutionary and passive designs. Because evolutionary plant resolution of severe accident issues reflected in Part 52 design certifications and the rule envisioned by this ANPR are expected to be " essentially the same," there would seem to be minimM value added by such a generic rule for evolutionary designs. Indeed.

we are concerned that issues relating to the consistency and applicability of the generic severe accident rule may be raised at a critical time in design certification proceedings with the potential for causing costly and unwarranted disruption of the certification i

process.

2 i

1

4 Regarding passive designs, parallel design certification reviews and severe accident generic rulemaking would add another dimension of uncertainty to an already complex undertaking. It is not clear what benefit a generic severe accident rulemaking based on evolutionary designs could add to the comprehensive and integrated resolution of all technical and severe accident issaes via individual design certification rulemakings.

The ANPR essentially acknowledges this by noting that, "as detailed design information becomes available and review of the passive systems is completed, further rulemaking may be necessary" Presumably, Ae net result of such a series of generic rulemakings would be that generic resolution of severe accident issues for passive de-igns would be

" essentially the same" as that reflected in passive plant design certifications. As in the case of the evolutionary designs, there would seem to be no value added by a largely duplicative generic rulemaking. It is our view that the potential benefit of such generic i

rulemaking(s) to a passive plant certification proceeding is outweighed by the greater i

likelihood that a rule to resolve expressed concerns would not be timely and could generate uncertainty, rather than clarity, in the design certification process.

In the deliberations leading to publication of the Severe Accident Policy Statement, the Commission considered whether it was necessary to amend the NRC regulations. The Commission concluded that a statement of policy was more appropriate as both NRC and industry studies demonstrated that operating plants posed no undue risk to public safety and health. ALWR designs satisfy the Commission's stated expectation that future plants achieve a higher standard of severe accident safety performance. This enhanced performance is being achieved, in part, by integrating severe accident considerations throughout the utility requirements and design processes, thus providing for enhanced severe accident prevention and mitigation. While addressing severe accidents for future plants is appropriate and consistent with the Commission's Severe Accident Policy, inclusion of criteria for future plants related to severe accidents in the Commission's generic Part 50 regulations would be inconsistent with the Commission's previous conclusion that ;eneric rulemaking was not necessary because current operating plants do not pose an undue level of risk to public health and safety. Because advanced plants achieve a higher standard of severe accident safety than the current operating plants, we would expect the Commission's previous conclusion regarding generic rulemaking to continue to be appropriate. Analogous to the plant specific approach to sevece accident issue resolution for operating plants via Individual Plant &aminations, an integrated, design specific treatment of severe accident issues within design certification reviews and rulemakings is the best approach for future plants.

Finally, the industry and NRC staff are presently proceeding in efforts to take a liard look at eliminating or revising unnecessary, duplicative, or potentially confusing regulations and regulatory practices. The generic rulemaking envisioned by this ANPR would be inconsistent with that process in that it would promulgate duplicative severe accident regulation on top of comprehensive design certification rulemakings, potentially i

introducing technical inconsistencies, schedular disruptions and licensing uncertainty. In addition to being the most practical and efficient approach to ALWR severe accident 3

I

issue resolution, the URD/ design certification process will provide for severe accident issue resolution that is clear, tailored and codified for each standard design, and reilective of integrated ar.d coherent consideration by plant designers and NRC staff reviewers alike, of the full spectmm of safety issues associated with ALWR designs. _

{

[1]

" Matrix Approach to Evolutionary Plant Contamment Performance,"

E.E. Kintner letter to T.E. Murley, May 9,1991.

[2]

DOE /ID-10291, " Passive ALWR Requirements to Prevent Contninment Failure," December 1991.

[3]

" Proposed Criteria to Accommodate Severe Accidents in Containment Design," ACRS letter dated May 17,1991.

ANIQ Ouestion Number 2:

Would a new rule in 10 CFR Part 50 concerning plant performance for severe accidents, as discussed in the three alternatives, provide a basis for revising the requirements on Emergency Planning Zones for future LWRs? If so, why? If not, why not?

Industry Response:

Plant design to enhance performance under potential accident conditions and the improved state of knowledge regarding accident consequences provides an appropriate basis for reevaluating the requirements on Emergency Planning Zones and offsite emergency response programs. The traditional requirements should be reconsidered for ALWRs due to the enhanced plant design, including containment performance and other relevant technical criteria in the Utility Requirements Document. Reconsideration of offsite emergency planning requirements is also appropriate due to developments regarding improved understanding of, and anticipated changes to, the source term for ALWR designs. As emphasized throughout the industry response to this ANPR, the industry does not believe a generic severe accident rulemaking is necessary. However, a generic rulemaking may be necessary to accomplish ALWR emergency planning and is being considered by the industry subject to finalizing the technical basis through review and issuance of the FSER on the passive plant URD.

ANPR Ouestion Number 3:

One option for an overall containment performance criterion that has been considered is that the conditional failure probability of the containment should be less than approximately one in ten. Two of the alternatives use a deterministic surrogate that states that the containment should remain leak tight for a period of approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the onset of core damage and after that time, remain a barrier against 4

the uncontrolled release of radioactivity when faced with challenges from the more likely severe accident phenomena. Is this criterion a suitable substitute for the conditional containment failure probability of one in ten? If so, explain why. If not, explain why not. Is a period of approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> an appropriate time frame? Is its degree of-conservatism appropriate considering uncertainties and defense-in-depth? If not, what alternative would be appropriate? What other criteria [probabilistic or deterministic]

might be considered?

Industry Response:

Deterministic contninment performance criteria are a suitable substitute to a quantitative conditional containment failure probability for demonstrating containment performance for ALWR designs.

The underlying purpose of containment performance criteria is to provide defense-in-depth in the design of containment and containment systems by assuring a balance between severe accident prevention and mitigation. For a plant with highly reliable core damage prevention systems, it may be possible to show consistency with the safety goals simply by assuring an extremely low potential for core damage. The quantitative health objective, the large release guideline, and the mean core damage frequency guideline could all be achieved without the need for a containment. The intent of a containment performance criterion, either probabilistic or deterministic, is to assure that no matter how low the potential for core damage, there still will be a containment function that is available under postulated severe accident conditions. As provided for by the Commission SRM in response to SECY-89-102, the industry has pursued a deterministic approach to demonstrating ALWR containment performance.

Indeed, the URD establishes design features and criteria which accomplish the containment performance objective of this ANPR, as well as that suggested by the NRC staff in SECY-90-016, thereby assuring adequate severe accident containment response for ALWRs. ALWR plant designs codified at the conclusion of the URD/ design certification process will have reliable containments and containment systems capable of preventing or accommodating a wide spectrum of severe accident challenges as measured against the deterministic containment performance criteria of the URD. This deterministic approach to evaluating containment performance meets the intent of a quantitative conditional containment failure probability for demonstrating ALWR containment performance.

The time frame for containment remaining leak tight after the onset of core damage should be long enough to: (1) allow for fission product radioactive decay and aerosol settling so that if a release were to occur thereafter, the site boundary dose would be below the threshold for acute health effects when analyzed realistically; and (2) allow time for protective action as part of ALWR emergency planning. The URD establishes design features and containment performance criteria based on a 24-hour leak tight period to provide reasonable assurance that each of these objectives is 5

1 l

accomplished with significant margin. With these design features and using the best estimate methodology of the URD, ALWR containments can be shown to remain leak tight for at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following core damnge. In actuality, a period of 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> or even less, depending upon the speci5c plant design and site, provides ample time for meeting both of the above objectives based on the half life of noble gases, the effectiveness of fission product mitigation systems and natural removal processes, and historical emergency evacuation experience, including ad hoc evacuations. Thus, the approximately 24-hour deterministic containment performance criterion proposed by the NRC staff in SECY-90-016 and this ANPR, in the absence of a common understand.ing regarding realistic evaluation methods, may be overly conservative.

This conservatism is acknowledged in Question 15 of this ANPR which notes that the 24-hour contninment barrier criterion provides a level of safety that is three orders of magnitude more consetrative than the quantitative health objective of the Safety Goal Policy statement. While 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> is an appropriately challenging URU requirement to ensure that ALWR designs contain features that provide a high level of severe accident protection, codifying this determmistic criterion would be inappropriate because doing so would establish a de facto higher safety goal. That result would be contrary to Commission guidance provided in response to SECY-89-102 concerning NRC staff imposition of industry design objectives as regulatory requirements.

Based on the preceding discussion, the criterion for protection against uncontrolled fission product release after 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> is similarly conservative and is inappropriate to codify. As noted above, it is estimated that a controlled release that would pose no undue risk to the public could occur significantly prior to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following core damage. Controlled release from the suppression pool airspace in the form of an overpressure protection device is being considered for the advanced BWRs.

Overpressure protection may also be provided by demonstrating that the size and strength of the containment allows meeting appropriate ASME limits for approximately

,t.hree days, which provides adequate time for actions to bring the accident under control.

'ThElatfer'nethod of protection from uncontrolled release is being implemented for large PWR containments.

In summary, design features for overpressure protection are specified by the URD and will be codified for individual ALWRs via the design certification rulemakings.

These features provide the basis for assuring substantial time for fission product removal and emergency response prior to any significant release, and their codification via design certification eliminates the need for generic severe accident rulemaking. The industry concurs in the appropriateness of deterministic performance criteria, and as discussed above, severe accident resolution for ALWRs achieved via the URD and individual design certification reviews and rulemakings will reflect use of these criteria for demonstrating ALWR containment performance.

6

ANPR Ouestion Number 4:

Alternative 2 would require extensive reliance on analytical tools that calculate the effects of severe accident phenomena. Are there analytical tools that are sufficiently developed and adequate to allow effective implementation of such a phenomena-based rule? If so, what are they, and for what phenomena could they be used? How would Alternative 2 be implemented? For example, should the codes and input parameters be i

approved by NRC? Should acceptance criteria be codified or put in a regulatory guide?

Industry Response:

Specific response is not provided to this question relative to the implementation of Alternative 2 on the basis of the industry's overall position embodied by the response to ANPR Question 1 that generic severe accident rulemaking for ALWRs is neither necessary nor appropriate.

With respect to the URD/ design certification process for implementing ALWR severe accident resolution, a combination of quantitative analyses and qualitative considerations has been used. The appropriateness of the implementation and criteria for a given ALWR will be indicated in the FSER for the design and codified via the certification rulemaking.

ANPR Ouestion Number 5:

I Should future LWR containment designs include features beyond those described in Alternative 1 to prevent / mitigate severe accidents? If so, what are they?

Industry Response:

Specific response is not provided to this question relative to the implementation of Alternative 1 on the basis of the industry's overall position embodied by the response to ANPR Question 1 that generic severe accident rulemaking for ALWRs is neither necessary nor appropriate.

As discussed in response to Question 1, ALWR designs contain features required by the URD intended to limit the potential or accommodate a wide spectrum of severe accident phenomena and containment challenges. These features include those identified in Alternative I of the ANPR and, together with specific insights from the design PRA, provide the basis for severe accident issue resolution for ALWRs via the URD/ design certification process.

7 l

ANPR Ouestion Number 6:

Alternatives 2 and 3 specify phenomenological severe accident challenges that should be considered in the design. Alternative 1 is based upon the same phenomena / challenges. Are there other severe accident phenomena / challenges that should be considered? Should the challenges be specified in more detail (for example, specifying the amount of hydrogen generation) or is a general statement of the challenge more desirable?

Industry Response:

Specific response is not provided to this question relative to the severe accident phenomena / challenges considered by Alternatives 1,2 and 3 on the basis of the industry's overall position embodied by the response to ANPR Question 1 that generic severe accident rulemaking for ALWRs is neither necessary nor appropriate.

As discussed in response to Question 1, the URD identifies an extensive list of ALWR severe accident phenomena / containment challenges. These include containment bypass scenarios, random system and equipment failures, and phenomena with the potential to challenge containment integrity. For each ALWR design, features are provided to limit the potential or accommodate each of the phenomena / challenges.

These features, together with specific insights from the design PRA, provide the basis for severe accident issue resolution for ALWRs via the URD/ design certification process.

ANPR Ouestion Number 7:

For what reason (e.g., not a risk significant phenomena, not a cost effective solution) would any of the criteria proposed in the three alternatives not be fully applicable to passive designed LWRs?

i Industry Response:

l

[

Specific response is not provided to this question relative to the implementation l

of Alternatives 1,2 and 3 for passive plants on the basis of the industry's overall position I

embodied by the response to ANPR Question 1 that generic severe accident rulemaking fo ALWRs is neither necessary nor appropriate.

As discussed in response to ANPR Question 1, severe accident issue resolution for passive plants is being coherently addressed via the passive plant URD and passive plant design certifications. This approach thus avoids the potential problem of incongruities between passive designs and a generic severe accident rule based on evolutionary designs.

8

l ANPR Onestion Number 8:

What features could an advanced LWR design include that would prevent or nitigate fuel-coolant interactions?

Industry Resoonse:

Fuel-coolant interactions (FCI) can be postulated to occur either in-vessel or ex-vessel. The potential for in-vessel FCIs inducing containment failure (alpha-mode failure mechanism) was identified in the WASH-1400 Reactor Safety Study. ALWRs have depressurization system capabilities which could result in low enough reactor coolant system pressures such that energetic FCIs could not be excluded based on system pressure. The reactor pressure vessel, and internals, head closure, and missile barriers included in the ALWR design are similar to those employed in existing LWR plants.

Thus, the same mechanisms are operable in ALWR designs that led to past conclusions that the occurrence of an energetic FCI which could lead to alpha-mode containment failure is very remote in probability.

In the unlikely event of reactor pressure vessel lower head failure, ex-vessel FCIs can result in both significant steam generation rates and shock waves induced by ex-vessel explosive interactions. In the PWR designs, any shock waves which occur will be within the thick walled reactor cavity, and the containment boundary integrity would not be challenged. Additionally, the thick walled reactor cavity provides a barrier against the generation of missiles which could impact the containment boundary and challenge its integrity. The lower drywell boundary provides the same type of barrier in evolutionary BWR designs.

In the SBWR, the lower drywell walls form the containment boundary. However.

)

a thick walled shield is being provided to protect the lower drywell portion of the J

containment boundary from the effects of severe accidents. Thus, containment boundarv integrity would not be challenged by shock waves or missiles.

For the AP600, the capability has been provided to externally flood the reactor pressure vessel and attached piping such that ex-vessel heat removal can be established.

This feature could protect the vessel lower head and prevent its failure. Maintenance of vessel integrity serves as a preventive means of precluding ex-vessel FCIs since it avoids j

the discharge of molten core debris.

1 The above ALWR capabilities and features, which are being incorporated in the j

advanced LWR designs via the URD and individual design certification reviews and rulemakings, provide adequate severe accident protection relative to both prevention and

)

mitigation of FCIs.

9

)

i I

i l

ANPR Ouestion Number 9:

If a design includes the capability to rapidly depressurize the primary system, should it also be required to have a reactor cavity design and/or a reactor vessel support structure capable of mitigating and accommodating a high pressure melt ejection?

Jndustry Response:

The capability to reliably depressurize the primary system provides sufficient means for limiting the potential for high pressure melt ejection (HPME) events.

ALWR designs include reliable means of primary system depressurization such that additional mechanisms and features, to address HPME are not required.

However, it should be noted that reactor cavity / support design has been extensively considered relative to assuring ex-vessel fuel / debris cooling under low pressure core melt scenarios. The resulting design of these features is also expected to limit the transport of fragmented core debris outside the reactor cavity region in the unlikely event of HPME.

ANPR Ouestion Number 10:

Should future LWR designs include an on-line instrumentation system that monitors containment atmosphere for gross leakage to reduce the risk from an inadvertent bypass of containment function? Would application of this system be sufficient basis to modify leak rate testing requirements under 10 CFR 50, Appendix J?

Industry Response; The Utility Requirements Document for advanced LWRs for both evolutionary l

and passive designs requires that a means be provided to enable the operator to perform a periodic check for gross leakage of containment atmosphere during normal operation.

The intent is to provide the owner / operator of the plant with added assurance that penetrations connected to both the containment atmosphere and the environment are not inadvertently left open during normal operation. It was not the intent that such a system be maintained on-line continuously; instead, it would be used as a periodic check.

e.g., following maintenance operations which involve opening of containment penetrations.

It is not practical to use such a system to modify the leakage rate testing requirements in 10 CFR 50, Appendix J. Testing conducted at the containment pressure which exists during normal operation would likely not be sufficiently accurate for detection of design basis leakage rates such as are detected by testing in accordance with Appendix J. Also, testing during normal operation would identify leakage in penetrations connecting the containment atmosphere to the environment, but not in t

10

other penetrations which are covered by Appendix J testing (e.g., penetrations of lines connected to the reactor coolant system). Thus, a system for periodic checking of gross containment leakage during normal operation is a requirement of the URD. However, it is not puctical or necessary to use this system to modify Appendix J leak rate testing requirements.

ANPR Ouestion Number 11:

What design criteria should be developed that provide assurance that the containment's integrity cculd easily be established during certain shutdown conditions?

Industry Resnonse:

Like severe accident issues associated with power operation, shutdown risk issues are being appropriately addressed for ALWRs via the URD/ design certification process described in response to Question Number 1.

The risk associated with shutdown operations is dominated by a few key plant configurations and operator actions initiating and responding to events occurring in the plant during shutdown conditions. Rese conclusions are variously reflected in NUREG-1449 [1], NUREG-1410 [2], the IAEA conference report on shutdown accident sequence modeling [3] and industry initiated studies associated with shutdown risk [4,5,6].

ALWR designs reflect past operating experience to minimize the risks associated with the events identified in the referenced documents. He need to isolate containment during an event which might occur during shutdown conditions is dependent on the risk associated with the postulated event.

For example, for tr:.nsient initiators which occur with the refueling cavity full, many hours are available prior to the beginning of bulk boiling even at the beginning of an outage, and days are required in order to boil off the inventory to the point that fuel damage would occur. His provides substantial time for operator actions to recover lost systems or mitigate the depletion by providing makeup from external sources. If the containment is open, the only risk during this period is that associated with the releases from the coolant activity. Such releases would be minimal (a small fraction of 10 CFR 100 limits), and therefore offsite consequences from this type of an event are expected to be very low. He URD specifies that the reactor designer demonstrate that the consequences of pool boiling with containment open are acceptable.

For those configurations when reactor inventory is low, such as head removal, the time available for recovery is short. Managing risk during these relatively brief and infrequent periods is best accomplished by plant system configuration control and by providing the operator with the capability to mitigate any loss of inventory. Again, the l

11 i

I

URD specifies that sufficient water inventory be available to provide makeup for losses of reactor coolant system inventory.

Because the risk associated with shutdown operations is very low and because residual risk is best addressed through configuration control and procedures, design features intended to limit releases by isolating containment are of limited incremental benefit and have not been established as design criteria. For purposes of establishing additional margin, the ALWR URD specifies that, where practical, capability shall be provided to restore containment closure through simple manual actions.

l

[1]

NUREG-1449,"NRC Staff Evaluation of Shutdown and I.ow Power Operation," February 1992.

[2]

NUREG-1410,"Ioss of Vital AC Power and the RHR System During Midloop Operations at Vogtle Unit 1 on March 20,1990," June 1990.

[3]

IAEA, "Modeling of Accident Sequences During Shutdown and Iow Power Conditions," November 1991.

[4]

"Seabrook Station Probabilistic Safety Study, Shutdown (Modes 4,5 and 6)," May 1988.

[5]

" Shutdown Risk, Prairie Island Dual Unit Shutdown," Northern States Power Co., ANS Executive Conference on IPEs, October 1992.

[6]

"The Diablo Canyon Shutdown Safety Assessment," Pacific Gas and Electric, ANS Executive Conference on IPEs, October 1992.

ANPR Ouestion Number 12:

Should equipment provided only for severe accident prevention or mitigation be l

subject to (a) the same requirements as design basis equipment (e.g.,

redundancy / diversity, power supply, environmental qualification, inclusion in plant Technical Specifications, maintenance priority, quality assurance; or (b) lesser standards (e.g., reduced design margins or the regulatory guidance found in appendices A and B of Regulatory Guide 1.155, " Station Blackout?"). If lesser standards, what standards would be appropriate?

Industry Response:

Equipment provided only for severe accident prevention or mitigation should not be subject to the same requirements as design basis equipment. This position regarding equipment survivability is consistent with that expressed by the NRC staff in SECY,

12

4 l

i

\\

l l

016 and subsequently accepted by the Commission via their SRM dated June 26,1990.

The NRC staff recently reiterated this view in SECY-92-070, which observed that applying the requirements developed for safety-related equipment to equipment prosided for severe accident would amount to inclusion of severe accidents in the design basis for the plant design. The industry concurs in the NRC staff position indicated by SECY 070 that, TE]xisting requirements and the high degree of pedigree associated with them provide the design basis of the plant."

As stated in SECY-90-016, equipment provided only for severe accident prevention or mitigation should be designed so that there is reasonable assurance that it performs the function for which it is intended over its mission time. However, as indicated in the SECY, the equipment pedigree requirements need not be as stringent as those for equipment credited in design basis accident analyses. Specifically, features provided for severe accident protection only, need not not be subject to: (a) 10 CFR 50.49 equipment qualification requirements, (b) all aspects of 10 CFR Part 50, Appendix B quality assurance requirements nor (c) 10 CFR Part 50, Appendix A redundancy / diversity requirements. Likewise, severe accident features would be included in Technical Specifications only to the extent that a design basis accident function is served by the same equipment.

The URD specifies that the design of equipment identified as useful for severe accident mitigation shall provide reasonable assurance that the equipment can perform its identified function during severe accident conditions. Design considerations for equipment under severe accident conditions include the circumstances of applicable initiating events and the environment (e.g., pressure, temperature, radiation) in which the equipment is expected to function. In addition, such equipment will be located and arranged, to the extent practical, to enhance its usefulness under severe accident j

conditions. Safety-related equipment that also serves for severe accident mitigation will be qualified and provided with quality assurance consistent with design basis requirements.

ANPR Ouestion Number 13:

Alternative 1 discusses not exceeding ASME service level C stress limits for steel containments under certain severe accident conditions. Are these limits appropriate for severe accident conditions? If not, what limits would be appropriate? Could these same stress limits also be used for loads generated by missiles? If not, what limits would be appropriate? What equivalent limits would be appropriate for concrete containments?

Industry Response:

Adherence to the ASME service level C stress limits for steel containments under severe accident conditions provides reasonable assurance that the containment will perform its intended function as an additional barrier against potential fission product 13

release. The contamment will perform satisfactorily if its integrity is not compromised and no unacceptable leakage develops d. mg and after the severe accident. By limiting the stresses in the shell to the code yielt. < service level C limits assure that the contamment remams in the elastic domain oy a margin of at least 20% Within the stress range, the containment integrity is only minimally challenged since significant margin exists between service level C and ultimate failure. Tests and analytical modeling have shown the margin to be of the order of two or greater [1,2,3,4].

Limits less conservative than the service level C stress limits should be used for loads generated by missiles. Missiles have a local impact and do not threaten the overall integrity of the containment. Criteria based on strain limits are appropriate for this type of loading. The energy is absorbed through plastic deformations limited by functionality requirements (i.e., no leakage and no zipper effect that would compromise the overall integrity of the structure).

For concrete containments, the unity factored load combinations of Subsection CC of ASME Section III are adequate requirements for severe accidents. They are based on the same philosophy as service level C stress limits applied to Class MC components.

The limits placed on stresses and strains in the reinforced concrete load resisting elements, strain in the metallic liner plate and liner plate anchorage systems and other liner design details, provide significant margin against catastrophic failure and containment pressure boundary leakage.

+

The containment boundary strain levels associated with service level C will be within clastic limits such that any leakage that does occur can be expected to be via penetrations through the primary steel membrane. Tests have been performed on individual penetration types including airlocks, hatches, electrical penetration assemblies, and mechanical penetrations at pressures comparable to containment ultimate capacities.

Such tests have demonstrated that typical penetration designs can withstand pressures substantially greater than equivalent service level C containment pressure without loss of structural integrity or any significant degradation to their leakage barrier function.

The containment loading criteria discussed above are being incorporated into ALWR designs, as appropriate, based on utility requirements and will be codified via the design certifications. Therefore, as discussed in response to Question 1, generic rulemaking such as that envisioned by this ANPR is unnecessary.

[1]

Keck, J., F. Thome, hak Behavior Through EPAs Under Severe Accidem Conditions," Proceeding of the Third Workshop on Containment Integntv.

NUREG/CP-0076.

[2]

Julien, J. T., S. W. Peters, hak Rate Test of a Containment Personnel Airlock," Fourth Workshop on Containment Integrity, NUREG/CP-0095.

l 14

i s

[3]

Brinson, D. A., G. H. Graves, " Evaluation of Seals for Mechanical Penetrations of Containment Buildings," NUREG/CR-5096, August 1988.

[4]

Crapo, H. S., R. Steele, Jr., " Containment Penetration System (CPS) Tests Under Accident Conditions," NUREG/CR 5043, August 1988.

ANPR Ouestion Number 14:

What information is available regarding the costs (capital and operational /

maintenance) of design features that would be required under these alternatives?

Industry Resoonse:

Based on the industry position strongly preferring the URD/ design certification process for accomplishing ALWR severe accident issue resolution / codification, the industry does not have the appropriate data and has not addressed the comparative costs of implementing the three alternatives outlined in this ANPR.

ANPR Ouestion Number 15:

The containment performance objective discussed in Alternatives 1 and 2 (i.e.,

containment shall provide a barrier against the release of radioactive material for a period of approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the onset of core damage) represents a level of safety for a 3800 MWt plant sited in accordance with 10 CFR Part 100 approximately three orders of magnitude below the Commission's quantitative health objective for prompt fatalities, as defined in the Commission's Safety Goal Policy Statement. It could be argued that a future LWR design meeting this objective through analyses and the incorporation of design features need not consider the addition of other features, since these other features would be directed at even more highly unlikely severe accident j

phenomena and sequences which could be considered " remote and speculative" under the National Environmental Policy Act (NEPA) and 10 CFR Par: 51. Therefore, would 1

the codification and compliance with such a containment performance objective be sufficient to also define a point of truncation and serve as the basis for an amendment to 10 CFR Part 51 eliminating the need for further review of SAMDAs for future LWRs under 10 CFR Part 51?

Industiv Response:

The industry concurs in the underlying thrust of ANPR Ouestion number 15, consistent with the intent of design certification under Part 52, that additional requirements should not be considered for ALWR designs whkh will have a level of safety substantially exceeding that established by the quantitative health objectives of the Safety Goal Policy Statement. In keeping with the intent to resolve all design related issues at certification, the industry and NRC are addressing the requirements under 15

NEPA relative to consideration of severe accident mitigation design alternatives (SAMDAs) as part of the design certification process. The staff recommendation in this regard, later concurred in by the Commission in an SRM dated October 25,1991, was contained in SECY-91-229.

While the concept of generically elimmating NEPA/SAMDA requirements for future plant designs demonstrating a high level of safety is sound, this benefit does not outweigh the industry's substantial concerns regarding the implications of conducting a generic severe accident rulemaking that would be necessary to provide the basis for such a change. Further, as a practical matter, a generic rulemaking would likely not be timely to eliminate the NEPA/SAMDA review requirements for evolutionary plant design certifications.

16 v

--m

\\.k.I I,.14 5,b.'d.5.bb d 6 0 PRC?Ci-D F.UO 7

WIN (47 FR '/'/5/3)fPJVN STON & S FREDERCK H wfNSTON 08531686) 1400 L STAEET N W.

cmcaoo cerca S!LAS H E'nAwN l'esi446, WASH NGTON O C 20005 3502 as west wAc=Em Dawn

'92 DEC 30 P 3 R2= "~o's-(202) 371 5700 NEW v0Aw OFFICE FACSMLE (2C2) 3715950 ns warEn sTnEET amerta s pimECT W AL huwSE P NEW YORM Ny soomget (212) 26F2900 December 28, 1992 Mr. Samuel J.

Chilk, Secretary U.S. Nuclear Regulatory Commission Washington, D.C.

20555 Attention:

Docketing and Service Branch RE:

Response to Advance Notice of Proposed Rulemaking Acceptability of Plant Performance for Severs Accidents; Scope of Consideration in Safety Regulations 57 Fed. Rec. 44,513 (September 28. 1992)

The Nuclear Regulatory Commission

("NRC")

recently published an advanced notice of proposed rulemaking ("ANPR") that would, for the first time, require new nuclear power facilities to be designed to withstand the effects of a severe accident.

57 Fed.

Reg. 44,513 (September 28, 1992).

On behalf of the licensees identified below,l' we are filing the following comments in response to two aspects of the ANPR: (1) the propriety of a rule amending 10 C.F.R. Part 50, as opposed to utilization of the existing design certification process codified in 10 C.F.R. Part 52; and (2) reliance on severe accident mitigation design 1

requirements as the basis for amending 10 C.F.R. Part 51 to preclude consideration of Severe" %ccident ~ Mitigation Design Alternatives

(*SXMDKd*T.

These comments are timely filed in accordance with the September 28 Federal Reaister notice.

1.

The NRC: Staff ~~Bhould Refrain from Proposing Regulations that Sper 4Meally: Address Severe Accident Mitigation, and Instead Pursin'Buch'Instres In the Part 52 Desian Certification Process The ANPR evidences a fundamental shift in the NRC approach to severe accident risk.

In the past, the Staff chose to limit consideration of issues involving severe accidents to l'

These comments are submitted on behalf of the following licensees of existing nuclear power facilities:

Niagara Mohawk Power Corp.,

Northeast Nuclear.. Energy Co.,

TU Electric, Tennessee Valley Authority, and Washington Public Power Supply System.

N~;Q cC C i/2tcc: t':..:,,,

Q U.S. Nuclear Regulatory Commission December 28, 1992 Page 2 environmental evaluations and general safety goals, articulating this position in two separate policy statements.F Even when the Staf f recommended in a generic industry communication that existing nuclear facilities with Mark I containments consider beyond-design basis over-pressurization events, the Commission recognized that, absent voluntary commitment, the limitations of the backfitting rule applied.21 In the ANPR, however, the Staff proposes to require future facilities to specifically design against a severe accident, i.e.,

an accident beyond current design bases, in accordance with discrete design requirements to be codified in 10 C.F.R. Part 50.

Not.only is the Staff now focusing on an entirely new and previously undefined realm of design requirements (as opposed to the consideration of alternatives in an environmental review) associated with severe accident mitigation, but also, at the same time, the Staff is proposing to implement this new philosophy in l

binding regulatory requirements to be codified in 10 C.F.R. Part 50, thereby casting doubt as to the appropriateness of the design i

basis for existing facilities.

This approach, if adopted, should not be implamanted via a rulemaking that would amend 10 C.F.R. Part 50.

Rather, the NRC should implement design criteria of the posited nature by means of the design certification rulemaking process codified in 10 C.F.R. Part 52.

Because the ANPR would require new design criteria only in connection with the licensing of future light water reactors, and because, as a practical matter, the design of such plants will be certified by rulemaking pursuant to 10 C. F. R.

Part 52, design criteria governing the performance of future LWRs under severe accident conditions should be romulgated in the context of a l

design certification rulemaking F

Sag 50_ Fed. Reg. 32,133 (Aug.

8, 1985), " Policy Statement on Sever 6 ReactorrAccidents Regarding Future Designs and Existing l

Plantg# 51 Fed. Reg. 30,028 (Aug. 21, 1986), " Safety Goals for the 0peratiions of Nuclear Power Plants; Policy Statement. "

2' Egg Generic Letter 89-16, " Installation of a Hardened Wetwell Vent," September 1, 1989.

s' Indeed, the Staff has proposed the form and content for a design certification rule. Sag SECY-92-287, " Form and Content For a Design Certification Rule," August 18, 1992.

In pertinent

part, the proposed rule would require the consolidation of all design-related information into a single, stand-alone document called the Design Control Document i

("DCD").

e U.S. Nuclear Regulatory Commission December 28, 1992 Page 3 2.

NRC Should Not Revise The Part 51 " Remote and Speculative" Finding On the Basis of The Proposed Severe Accident Design Recuirements The Staff specifically asked for comments on the advisability of using a

new containment performance design l

requirement as the basis for revising 10 C.F.R. Part 51 to " define a point of truncation" and to eliminate "the need for further review" of SAMDAs.I' Presumably, the Staff would cite compliance with the proposed Part 50 severe accident design requirements as l

the basis for concluding that severe accident risk, as a result of i

implementation of the proposed severe accident design criteria, is

" remote and speculative" within the context of the National Environmental Policy Act ("NEPA").

42 U.S.C.

4331 21 a.cq.

As a result, the benefit to be realized from the installation of SAMDAs would not have to be cost-justified under' NEPA.

Unuer such a scheme, the proposed Part 50 severe accident design remiirements would become a " floor" below which environmental alternatives regarding severe accident need not be considered further.

To proceed with such a plan would inappropriately precondition a Part 51 finding of " remote and speculative" risk on

(

the inclusion of SAMDAs in plant design.

To date, the NRC has dismissed SAMDAs in its NEPA analyses without requiring that facilities specifically be designed to withstand the challenges of severe accidents.

For example, in an operating license proceeding that focused, in part, on the need to consider the mitigating effects of SAMDAs in an accompanying NEPA analysis (where the plant was not designed to withstand severe accidents),

the Staff

"' discovered no substantial changes in the proposed action as previously evaluated in the FES [ Final Environmental Statement]

that are relevant to environmental concerns nor significant new circumstances or information relevant to environmental concerns and bearing onwthawl:1,cansing of Limerick Generating Station, Unita 1 and 2.'*8' Moreover, the United States Court of Appeals for the Third Circuit has aptly observed that the Commission itself has j

noted that,the.impaccaof SAMDAs on the environment will dif fer with a particular plant's design, construction, and location.

Limer m l'

57 Ped. Reg. at 44,518 (Question 15).

I' Supplement to the Environmental Impact Statement, Limerict Generating

Station, Units 1

and 2,

at 1,

cuoted in Philadelchia Electric Co. (Limerick Generating Station, Units 1 and 2), LBP-89-24, 30 N.R.C. 152, 153 (1989).

l

9 U.S. Nuclear Regulatory Commission December 28, 1992 Page 4 Ecoloev Action, Inc.

v.

United States Nuclear Reculatory Comm'n, 869 F.2d 719, 738 (3d Cir. 1989).

Thus, the Court expressed doubt as to the feasibility and validity of according SAMDAs generic treatment under NEPA.

Id. ("it is unlikely that severe accident mitigation can be treated as a generic issue.")

Even if feasible, generic treatment of SAMDAs, as a practical matter, may deter the development and installation of cost-beneficial design changes at individual facilities.

In sum, we reiterate that Part 50 should not be amended because there is no need to require design e hanges to acccamodate remote and speculative accidents.

In addition, in response to the question posed concerning Part 51, we believe that even if the Staf f should proceed with rule changes to Part 50 to address severe accident risk, those changes should not be used as the basis for revision of Part 51.

Sincerely, i

Mark J. Wetterhahn Kathryn M. Kalowsky WINSTON & STRAWN

ine

.i W-i i

~

l e

1

-1 i

I FEDERAL REGISTER ANNOUNCEMENT ON ANPRM b

s

-l r

f r

?

i r

l 4-

)

i I

f Monday e-2s-e2 g

Vol. 57 No.1se September 28, 1992 Pages 44481-44650 i

)

E R

R E

E w

M1 t

e.->

e u

j r,-r-

+

+, _

4 i

e Federal Register / Vol. 57. No.168 / Monday. September 2& 1992 / Proposed Rules 44513 TAett 1..--ESTruAtto tuPacis or 1993 ARP OPTIONS Options l~

1 2

3 4

5 AAP (%) _

..~...

oo 25 5o 75 to o Particutspn (%)..

94 94 94 S4 94 Plamed Aces 0000 Ac )

3003 2920 2835 2751 2667 P'o@ct oa 4W cwt).

165 6 1612 156 8 1524 147 9

~

Domestac/Ress%st Use (W cwt)...-.... _... -

s76 972 a9 8 BG 4 89 o Ecorts (W cwt)..

76 o 74 o 77.8 7o o 6e o Erong snacas tnui cwt)...

32 9 31.o 28 9 2tLO 243 54acas/use Raco (%) Praar 3 veers...._..

20 5 19 3 18.o 16 e 15.5 Season Avg encs 41/ cwt) 6.7o 845 7 oo 720 7.35 Net our.rys (W S)..

736 707 665 811 572 Accordmgly, comments are requested NUCLEAR REGULATORY Regulatory Commission, Washington,.

whether to establish an ARP for the 1993 COMMISSION DC 20555, telephone (301) 492-3980.

rice crop, and if so, at what reduction suppt.turwTAny swronuATiosc percentage.

10 CFR Part 60 The final determination of this RAN 3150-AE34 percentage will be set forth at 7 CFR ne Nuclear Regulatory Commission part 1413.64(a)(4).

AcceptabWty of Plant Perfonnance for (NRC)is considering developing Ilst of Subjects in 7 CDt Part 1413 Severe Acckients; Scope of regulations under to CFR part 50 for Considervtion in Safety Reguistions future LWRs to address the ability of the Cotton. Feed grain, Related programs, plant to withstand challenges from Nuclear Regulat,m7 phenomena associated with severe core Rice, Wheat.

Aotwcy:

Commission.

Accordingly,it is proposed that 7 CFR damage accidents. Severe core damage part 1413 be amended as follows:

Acmc Advance notice of proposed accidents are low probability events rthmWg-beyond the design basis established in PART 1413-FEED GRAIN, RICE.

~

UPLAND AND EXTRA LONG STAPLE soeeuAsry:Thi Nuclear Regulatory to CFR part 50 that can ! cad to significant core damage and radioactive COTTON, WHEAT AND RELATED Conssion (NRC)(s cesidermg an material release from the reactor fuel PROGRAMS amendment to its regulations for future light water reactors (LWRs).De pins %e NRC believes that research 2 iWFi

1. He authonty citation for 7 CFR amendment would add provisions for

,,,refcd pheno na, ev nt part 1413 contmues to read as follows:

the design of the plant structures i sequences, and cost effective methods to Authoriry: 7 U.S C.130tL 1306a.130E 1441-withstand certain challenges fro 2.1444-2.1444 f.1445b-3a.1481-1467.15 phenornens associated with severe core mitigaR them, coupled with its U.S C. 714b and 714c.

damage accidents beyond the current understanding of the details of future design basis accidents." De NRC :s plant designs, have sufficiently matured j

to all w the development og p ant

2. In i 1413.54. paragraphs (a)(4) (ii) issuing this notice to invite advice and and (iii) and (d)(3) are revised to read as recommendations from interested puf rmance entuia for futum 1.WRt follows:

his advance notice of proposed parties on fhe proper scope 'and method rulemaking (ANPRM) requests from i 1413 54 Acree9e reouction m to incorporate thge provisions into provmaona.

safety regulations.

interested parties advice and p)...

DATas: Comment period expires

. mm ndh a 2e propa swpe and method to incorporate these (4)(i)...

Decembu 2& 1992. DektC will considerations into the NRC's c ns ments meeMahn,%.

regulations.

(ii) * *

• 1992 rice.0 percert.

• ' but (iii)1993 rice, within the ranFe of 0 to his ANPRM reflects consideration of ba 8

• 2 ere ociden area et rrr r oun db th no paid land diverwon.

. I8 M .

NRC has already taken various actions "C'

Anoetsses: Send wdtten commants.to:- in response to severe accident concerns.

De Secretary of the Commission, U. S-

- On October 2.1980, the f%mmission Id)...

Nuclear Regulatory Commission, issued (45 FR 65474) an ANPRM that (3) Shall not be rnade available to Washington DC 20555, Attentiork _.,

invited advice and ecommendations on producers of the 1993 crops of wheat.

Docketing and Service Branch. Deliver the consideration of degraded or mohen feed grams and rice, as determined and comments to:11555 Rockville Pike.

cores in safety regulation. Based on announced by CCC.

Rockville. MD. between 7A5 am and 4:15. recommendations received from that pm on Federal workdays. Copies of ANPRM. the Commission developed a Segned on September 2.2.1992, et comments received may be examined at policy statement that addressed severe Washington, DC.

the NRC Public Document Room.2120 L accident consideranons and withdrew Ksith D. Bjerke.

Street NW (Lower Level), Washington.

the ANPRM (August & 1985: 50 FR Esecutive %ce President. Commodity Credir DC.

32151). In its ~ Policy Statement on Co'P"Doa-Fon rusm4tn euro8tMATION CONTACT:

Severe Reactor Accidents Regardmg IFR Doc. s2-23385 Fded 9-22-412:4.19 pm)

. Thomas King. Office of Nuclear Future Designs and Existing P! ants" auma conc *****ws Regulatory >Research. U.S. Nuclear published August & 1965 (50 FR 33331.

j

\\

7 9

)

1 F.

h4*ral hneter 7 Mal. SL MO. OE t/ fMonditS.: September 28, Mr '/ Prnposed, Rules u tg a

smsdits and the Commission's Safety the Commission stated its intentions far tsoetht attdmrcezardGamrniesmn Onal Dnbry.( Au pt.4. MC1.ER rulemakangs andnther regulatory Emdmce nn the kJ.phcatmnmf thew actions for resolvingiaevere acodent proposed severe accident and 28044) In addition this rule could safety issues. Tor e s.inting planta, the contamment cnteria to the esolutionary complement and support the review of Comminuortconcluded that thase plants DVRdessns now under revie4 Were Accident De+itmMugetion Alternatives ISAMDAs) on future LWTu posed no undue risk to pubhc health and Guidance from the Comtrussion was provided in a *5tafrRequirements as.part of the environmcr.ta! review saTeW. Therefore,it did not mee a need Memorandum. S. Chili to I.. Taylor.

4arned out under MCFR Per:Frt for immediate acrinn on Fenerm daed.}une 26.1990. The ontene Ba*,s.for theJune rukmaking Tur thewylante tecauseof

,dryctrryed m-this PtNPRM would codify the low severe accident rakJhe much of the1 Commission's guidanceTor Thu, rule would reflect theWRC%

Commission bas continued to take all Seneral REphcation to All. future.LWRs.

current tmdmtendmg of severr reasonable steps to'further reducelhe Additionally. the :.'RC plans t accident issues from its research.

rek frorn+evere accidents et existing mtprove itstegulatremr fer-futme plants experience with hght water reactors plants through its regsflatory programs.

by separating 1decouphng)'the now in operation. end reuew of future For exarnple. the Cornmission completed acceptance entene for a reactor site designs. Accordingly. this rule would rulemakmgs on several key issues from the acceptanae-cnnene lar.the g

related tasevese accliente(i.e.estation design of vanous engineeredaefety only, but could rentbly p-ovide blacknat.untmapated tnmaients wittnut features (ESF) via rulemaking changes guidenecim estabhshmg errteria-for scram. hydragan,ganamuan sad to to CFR Parts 50 and 100..TW other reactor types.The develepment of control). has tmplemented a phase of this decouphng of sitmg cnteria gg g,

.y containment performance improve ment from design enteria focusesvrrupdatmg &

dh prograntbased apart insigh te. rgf.ardmg and.reywrm erecritana. TL-d Since the accident at Three+1ile containment perioncane nnder. severe phase di trus,processmouDLfocus on IslarJ mnsmscderabbenearchan accident condatiaascandlas. initiated,a updating.10 CTR part3cTor'Etm EWRs severe acadents.has been perfonned.

prog um for todwiduahplanj to:

TL M1ayexplowd the exarrunatma(IEE) foraesere amident (1)!mplement, mew 1WRe w

pnenomena associated within. vessel vulactabihhas.

informa tion.

ahwmebem cadent (2) apec44 edomancea ntene for pmcessee: byttregen generation and dta 1 shed th c te plant design features based on.impraved Pob control; the. form.4pantity andummgcf andprocedural supaunderufsch a.m kaledge.of the hof sedioectwe radwachmded re e*&

design for a cursentr. power phmt emid containnarnt; dhdllegge.s.tn.centramant be.engplaldt"for tueettcg amtere sourcedennLand integnty: and'the consequence.s.to.the (alatpeedy critenmfstr plant pubbc. Utis research has led to the pedomance umseressimr.smcident development.cf. data andeanakucal th w

annr==. Ehe-enteraad. a r +. tochanabememN b a ccident preventeers a nd cenammaenne con MN arew--

ath a portami ad h m the, t

b8""*P

""' N '" N severeJLccident.riskand e. valuate generation of4 plants.,Alsocthe NRC atmve-potentisl risk-reductiortimpravements in expeated that theemew plantwruld aciunve e hrgherstnadardsf severe Purposa ciftbeTdle design and operation.

accidentrealety performansethan prior Tb Ebbw&WW @

gbmmmMe s[

to specifpssaptatAef astysrfermam" has occurredin rrnmy arces.One that.e l

rn in wp nac t saveseasa dents would comprehensive applicatien has been the cj, g

performancemill he snade.mcf adaga a c mphsh.the.fnHawuls develqpment.olWUREG4uo '. "Se vere decision en whetheNn,estabbetraciw

1. Codify the Commissank.gmdaase Accident Risk: An Assessment forrFive pedormanoorniena for tantamment on severe nezuderti and -tainment U.S. Nuclear Power Plants." NUREG-systems and..lf so. what thane should issues that_tradited1rsm.the necew.nT 1150 used pmbabilistic techmques to adva nurikght. water.reacicts.

analyr.e five operating plants from a The NRC staff has been revtwm8 2.?tovide assurance.1 hat.the severe accrdentrsk penrpertrve. This propoacd entenecfor-hrturr3 LWRs performance orfatuseIWRsmnd" analysie provnied4heNRC4tstff with subtmited by ElectnraPower Besammh severe acrIdentrandihaJa -defent be see irrrgh% into the>important event instittne (EPRI) and-savemakwwLWR with assumptiana.about.savere aocident sequencesatt st-cen4 ecd t& severe designs widt respect-tathe perfortnance used in.Beveloping new accidents and the mechamrms that can Comnrssion a 4evereatmde at palmy sm w.. durmahon.

lead to a loss of containment function and the-deeigntstufzcanon aspetssihD 3.:Prorrdepreenxtttfvturr' LWR during severe accidents. These tasrc CFR partT. In.perfuruung these designers and potentral e-M!s.

neighne identified challerges,to reviews.the.NRCrataff bestprmposal

4. Add'etmsrstency sud containmentantggrityshatamn.be entena to4ddressaevere;eenranttand standardsatien ttrtherrschttrorver dwided. antr >sworgraups:4aergauc or contamment msues that zicpart'fmr, the severe accident issues beard rm-the rapid energy releases. and i, lower, e Alstmg regLhttmDEaFor the cu rent tedhnmalirffermrtion.

evolutionary 1WPATm.anany of 5 Feelliterte ties gn-cett!!!catitm

• cm= of swasm acwimw these ptcposed cdica.m -- a Jned.m ndem & M Supenmendent ot Documents ULGmemament a papcT:prDvidEthtatine*NRC

'Fhre rule wou!d then$ help-essure-Set Pnn: ins crr.o P.o. som s'on w..Lasion Dc am2 oc cop,e. m si.e em: we Iroe>e Cornmanienera mCzrnun:y.n.mm.

thembto the puMrifmmse rere Nan =t.w n sus a~smswt SECY 2%DtAWadenxxury*IM*R accidents ta futun LWRs is maintained-Certifccaban,In.wesm:dIFbs:r at verpowdaneirin.ercevtieweewftb Q

j 7,Qc Relsuensht;Mo CmrentBrektmy e xpe n e ncea from,exsetmgs plansa, cerrvest puu,c poc.,.etoon run, sont. uw sto er RequirrtnerrvaNrheMRCasafLhan instghtsdrtmvr.sketud,es4md nc wreh Lew4 %eshwon. DC

s (e

Federtl Register / Vol 57. No.108 / Monday. September 28. 1992 / Npd Rules 44515 gradually evolvirrg releases to the cioced System 80+). It to hkely that the phenomenon. this alternative would corttainment rystem. Ltamples of resolution of severe acc> dent issues for require that the applicant pmvide an containtnent loadmgs in the first gmup these designs will occur via the evaluation of the phenomenon with incjode hrgh pressure core melt crectmn individual design certifications before respect to the overall containment wath direct contamment heating.

completion of this rulemaking. De NRC performance objectrve specified in the hydmgen cx>mbust2on. and the amtial expects the resolution of severe accident mie. His alternative is derned from the release of stored energy from the reactor issues for the evolutionary LWRs and containment perfonnance criteria coolant system. Decay heat and the results of this rule to be essentially developed as part of the Commission's noncondeosible San Feneration from the same.

advanced reactor reviews and core.concete mteractions typdy the The NRC staff is also reviewmg future essentially codines those criteria.

g.oup of slow energy rekases to the LWR designs that use a passive design In this approach, those features of the contamment. Further maights from this concept. In contrast to the evolutinary design needed for severe accident analysis idenched maior contnbutors to designs, the NRC staff has reviewed prevention and mitigation would be nsk to the poblac and potential design only conceptual design infortnation imm specified directly in the rule.These solutrons.

the passtre plant vendors.This requirements would be an ** overlay" en Also, the NRC has frequently prehminary review has not identiSed the e.xisting design basis requirements in interacted over the past several years any unigee features that would prevet to CFR part 50 for nuclear power plants..

with EPRI and vanous mactor deargners the evaluation of these desrgns under ne requirements would be considered concenung reguletory mtena for the the rule disenssed in this ANPRM-and justiSed on an enhanced safety future evolut2onary and passive LWRa.

Herefore, this tsle wcmid be generally basis (i.e. using safety goal cost.beoeht Some of these interecnons have applicable to passive LWR designs analysis and other appropriate addnseed both probabikstic and Hewever, as detailed design information conuderations such as defense-in-depth dstermmistic mteria awociated with becomes available and review cf the and uncertain 6es) and wodd plant performance under sesere passive systems is completed. further com'plement the e.xisting design basts to accident conditions.nese potental rulemahng may be necessary

,h, h levd d adety. However.

cnteria have been diset.s.ed in vanous kcause d b bw thhhood d senre a

wrrespondence with EPRI and the acadents. these new requtrements reactor des:gners, and are documented I3iscussed below are three potential wodd e Wdend w W in dreh safety evaluation rworts on the alternatives for incorporating plant traditid de'p ksis nq-a EPRI Advanced Light Water Reactoe performance criterie for severe For empla.% btuns W Requrements Document.

accidenta into the regulations.

mly b severe dded Mgsben A.

he I

Yro

,,,,,,,,,,,,, ga,g,,,, g,,M,g g,g, : woeld mot be subiect to the same i

NRC provided by an independent study of his atternetive (as are the other c nservative analysis and design requiremmts that an necessary b containsnent design critena made by the altermstrves discussed in this ANPRM)

Advisory Onmmittee on Reactor is based opun ensuring that the risk symems W to cope wd dense basis accidents. A regulatory gmde Safeguards (ACRS). In a leuer to the signtficant severe acedent phenomena.

Commission on May 17.1991. Ibe ACRS which may cause a lose of contarnment would proesde addihanal guidance os such % detans as p*mdancy proposed a set of cnteria addressing the function 4e an LWR. are considered in specinc challenges posed by severe future LWR designs. Bened cpon diversity, system capaaty, power accidents to the containment design for currently available infortsation, supply. equipment survivability amd future light water reactor nucleae power including the results of risk studies and

""*U*iC*I ****'"PU'*-

plants. In SECY-42-070, dated February severe ecc> dent reeeerch progran,a, An ameple of tMs af ternatrve I*II"

28.1992 the NRC staff analyzed these these risk significant severe eccident ACRS cnteria with respect to EPRI phenomena are-50XX Prevention and untigatsos of design requtrements for evolutionary

1. Hydrogen generwtion and trentport.

,ere - evt-m 1

and passive LWRs. evolutionary vender includmg bundng endfor detonation.

designs passive vendor designa, and the resulting frein teetal-wa ter and core (s) Applicchility.%e criterse of Game 1

existmg Commission guidance for the concrete reactxme; section apph to the design d hght mess {

)

NRC risff's review of severe accident

2. Mg6 pressure efection of meken nuclear po==er noctors king &

issues b evolutionary LWRs. nis core material from & noctor veseel for a constroh pendt er ego %

ANPRM reflects cxmsideistion of the -

3. Intersc5one between snotten co,,

licensa mk WR part 50 m results of this analysis and the ACRS debris and reecser beeemet meterial.

apphcatums usader ID CFR part 52 en er proposed entena.

containment wuD and etnictural after the effective date of this rde he j

material:

criterna of this sectroo also may prwaer i

Appbcalmiity of tb Rule

4. Containment overprepare and guidance in estabbinng the The NRC has accumulated an overtemperstwe frois decay best. non, i@as for other types of reer*=r understandmg of the evolutionary light condensible ges geneset on, metal-wster de*i D8-S weter reactor designs to complement tts reactions:

(b) Contrineerit M-ca understandmg of severe accident issoes 5 Steam exp&ossene fran %0 coolant O&fecerve. he design sbaH inclode e frorn operatmg reactors. Based on this interscrious: and containment eyetem that provides e understandmg. it is expected that the

6. ContainmeM bypees.

barrier against the release of red 6ancm.

attena developed in tins rule would be Ahernattve 14 oald specify material for a period of spproximess+r cons stent and compatilde with the reasos=ble design feetares or ett:ibeles - 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> foGewtag the onset of corv cntens being de reloped and sppt6ed in of design festures direcsed toward demege tmder the mere Uke!y revw, evolutionary LWR reviews. However.

prevenoon or r:dtigatice of the ebove accident cireBenges.FoBawing thes as due to theadvanced stage of the phenomena. Where design festarws hour pened, the conte-r n ent sho=6d reviews ni the current evohstionary cannot be predsely specified to pw-on contmtne te maintain a barner er'***

designs (CE ABWR and ADB/CE or1suttigate a severe moe=d:M the uncentre! Led release of large

7m 1

7 a

1 Federal Register / Vol 57. No.188 / Monday. September 28.1W2 / Proposed Rules 44516 diverse containment heat removai i

quantitles of fission products.This shall located outside containment and be accomplimbed by:

connected to the RCS to an ultimate systems or rely on the restoration of,

(1) including plant design features rupture strength at least equsi to the full normal containtnent heat removal that RCS pressure, capability if enough time is avadable for (il Provide the reactor coolant system W Not creditmg use of containment maior recovery actions.

t The advantages of this approach (RCS) with the capability to rapidly and ventmg dunng the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period.

melude presenbmg those design 7

reliably reduce RCS pressure.

followmg the onset of core dama6e m (ii) Provide a reactor cavity design evaluating the design for compliance features to reduce the nsk from severe that restncts as much as practical the with the contamment performance accidents.thus promoting a more amount of ejected core debns that objecove in paragraph (b) above.

standardaed resolution to severe re&ches the upper containment or (c) Equipment Surm abihty. Features accident issues. In effect. this impinges directly on the containment provided for severe accident presention alternative is also prestnptive regardmg wall.%e cavity design. as a mitigating or mitigation shall be designed to the severe accident phenomena that a feature, should not unduly interfere with operate for the ttrne penod needed m the future LWR design must address. since operations including refueling.

environment {e.g., pressum. temperature, the design features specified are a direct i

maintenance. or surveillance activities.

radiation) tn which the equipment is result of the phenomena considered. The I

(iii) Provide a reactor vessel support retted upon to function, including presenptive nature of this alternatne structm e sufficient to retam the reactor consideration of the circumstances of will also tend to facihtate the NRC vesselin place under the loads apphcable initiatmg events (e.g.,

review and design certification process generated by a high pressure core-melt transients, loss of AC power. loss of by focusmg the review on the severe ejection.

coolant accidents).

accident p' enomena which rnust be (iv) Provide for containmenWnde Matntaining contamment intepty for considered and the basic features which bydmgen contml (e.g. igniters. large a penod of approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> the design must incorporate to address yolume), that accommodates the fouowing the onset of cort damage those phenomena. thus enhancing hydrogen reeuttice from a 100-percent provides time for the remaining airborne regulatory efficiency. In addition. this i'

metal-water reaction of the active fuel activity m the containment (principally approach essentially codifies cladding. and limits containment noble gases and iodinal to decay to a Commission guidance on severe hydrogen concentration to no greater level that. when analy2.ed realisucally.

accident and containment issues from I

than to percent or provides that the would be unkkely to cause proc:pt the advanced LWR reviews.This post-socident atmosphere will not health effects if containment failure or.

approach does not regmre the appbcant support hytt n combustion controlled venting were to occur afte: -

m peria esive severe accident (v) Reduce potentral for and effect. that time. In addition. si represents a of interactions with molten core deMs - level *of-safeiy'eigulficantly belowthe

".,"j'f[e)p c]nfco d pnm n y rely by:

quantitative health objectwe for' prompt upon design features which. through

+

>or iataIines defined in the Cornmission s previous analyses and research. has e

( A) Pmviding reaetor cavrty ik.

space <o promote core debris spreading Safety Goal Policp However..

been shown to be effective in reducmg and coolabihty;

- considering the uncertainties involved in the risk fmm the more likely sesere

[B) Prmiding a means to flood the analycing the severe accident accident scenarios. coupled with reactor cavity to assist in the coehng pbecomena and progression and deterministic analysts to confirm that process and scrubbing of fission

- emphasis ou defense-in<!epth. it is not the contamment performance obiecta e products-imreasonable to include some (C) Pmtectmg the contsiament liner - conservatism in the criteria. This time The disadvantage to this option is that and other structural members from-period would also enhance the time.

it could discourage designers of future direct contact by molten core debns; available for offsite protectiv$ actions, LWRs from developing other design (D) Employmg besemat materials To the extentbractical A.nng this which reduw the production of non-period the passive capabihty of the approaches that might be more cost-effective innovative, or safer.

condeceible gases when in contact with, containment a.nd any; elated design

~

molten core debris: and features (e.g, suppression pool) sho !d A/temative rfhenomeno Onensed flu /e (E) Ensunng thet c.ontamment

- provide for contaimnent integrity.~

"Ris alternative is a modified servon temperature and pressure increases or Following this geriod.Jhs sontainment of the first altemative. It.like the firs the generation nf missiles resulting from.should continue to provide a bamer

. alternative, states an overall decay heat. fuei-coolant interactions.

against the uncontrolled release of, conteinment perfortnance goal and as I

combustible gas genemtion and control. firsion pmducts. However, in Leepmg.

based upon prev,enting or rmtigating the and core-basemat matenalinteractions, with the concept of allowmg for involving o range of event sequencea intervention in coping with long-term or same severe accident phenomena as t

descnbed in Alternative 2. However which release core debris intothe gradual energy. release, controRed.

containment do not cause containment elevated ventmg (if provided in the instead of specifying hardware stresses to exceed ASME service level C design) may be given credit in the design requirements in the rule to meet the limits for steel containmersts, or analysis after the initial 29 hour3.356481e-4 days <br />0.00806 hours <br />4.794974e-5 weeks <br />1.10345e-5 months <br /> period goal this alternative specifies the sewere equivalent for concrete containments. or to reduce the chance of containment accident phenomena that need to be significant degradation of the failure.The intent of epecafying in the addressed in the design. Based upon containment design leak rate.

design analysis no reliance ett analysis of these severe accident (vi) Reduce the possibihty of-containment venting durfug the initial 24 phenomena, the designer would develop containment bypass and a loss of hour period is to achieve tha design -

and propose the actuaJ design fes tures coolant accident outside containment by, objective of high containment integrity necessary to meet the goal. Regula tory designing. to the extent practical all.

but not to constrsin use of venting guides would address items such as elements of systems and subsystems dunng operation if for some reason analytical methods, assumptions.

(cat. p9 ng. Instrument lines, pump

. venting were the desired course of o acceptance aiteria and guidance on 1

seals. heet eu+ager tubes, and valves) action. Altematively, a design may sae design criteria for severe accrdent s

.e, o

Federal Registar / Vol. 57. No.188 / Monday. September "3.1972 / Proposed Rdes M517 hardware. An cumple of this design. nis alternauwe could be made nr alternative would differ from the alternative follows:

more prescnptive by specifying. for other alternatives in that the eusteg to 50 XX Prevention and mitsation of example. the amount of hydrogen or CFR part 50 design basis would be molten core debns which must be modified to include severe accidents.

severe accidents considered but. cevertheless. would Accordingly. the design requirements for (a) Apphcobsry. Le cntena of this provide designers with considerably the severe accident equipment (e g.,

sect;on apply to the design of light water more desap flexibdity to address sesere quality assurance. equipment nuclear power reactors bems considered accident issues than Alternative 1.

s urviva bih ty. redundancy / di ver sit s )

for a construction permit or operauna Apphcants would be required to provide would need to be determmed m rdataon bcense under 10 CFR part 50 or analyses showing that their proposed to those for traditional desi n basis F

appbcat2ons under 10 CFR part 52 on or design meets the containment eqmpment, Different design I

after the effectite date of this rule.ne performance objective. However. this requrernents may be appropriate for cntena of this section also may provide attemative would plcce a heavy rehence severe accident equipment because of guidana in establishing the on analytical codes to predict the the low probability associated with requirements fer other types of reactor likehhood of severe accidents and their sesere accident scenanos. nis designs.

behavior accurately. Imrutanons of these (b) Containment Performon&

analytical codes and gaps an knowledge attemative would give the designer Ob/ective. ne design shall include a of the phenomenological progression of flexibility in devising proposed solutions l

severe acodents may make such a to revere accident phenomena. Like contamment system that prov> des a attemative 2. this altemative would barner apsmst the release of radioactive beavy reliance unacceptable, &

material for a penod of approx 2mately boundmg parame1ers are used. Uke req; ire applicants to submit analyses 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> followmg the onset of core Alternative 1. this attemative would showing that the entena are met. Uke damage under the more hkely severe facihtate the NRC review and design alternative 2 this would place a heavy accident challenges. Followsng this 24 certificstion process by focusing the reliance cn analytical codes to predict hour penod. the containment should review and hmitin6 itrgation on the severe accident behavior accurately and l

contmue to maintam a barner against severe occident phenomena which mut would leave open to review and the uncontro!Ied release oflarge be considered-however. It wedd leave litigation in a licensing beanns or design quantit es of fisuon products. This shall open to review and htigation whether certification rulernaking whether the be accomplished by; the designer has adequately addreened designer has adequately addressed the (1) Mimmzmg the likehhood or effects the severe accident phenomena.

severe accident pbecomena.

on containment integnty of the followina Accordmgly. this attemative could Plans and Sckdules severe accident phenomena:

potentially require considerable NRC (i) Uncontrolled hydrogen buming and revsew effort pnor to ecceptmg en The plant performance regturements detonatierr applicant's analytical results. SimBar to desenbed in this ANpRM are part cf the (ii) Interactions betweca. molten core Alternative 1. this attemative would be second phase of a program to decouple debne and the reactor basemat cnaterial.

in the form of an overisy on the existing siting and destgn critena. In this phase, reactor vessel support structure-design basis specified in to CFR part 50 plant performance requirernents for containrsent wall, and other strucrural and testified on an enha: teed safe +y severe archts,in combination with matenals; ba sis.

(ui) High pressure meh e4ectim.

other necessary changes to 10 CFR part

50. will result in a rule that would liv) Containment bypass and loss of Ahernative 37 Ceaera/ Dengenh interfacmg :ystem integnty ICDC)OrsentedMe complete the decoopims of siting and tin 6 e M ssm n p k ns (v) Steam explosmas due to fuel-In this alternative, the NRC would coolant interaction; and to publish the proposed rde for develop e eet of new (:esign comment in aud.1993 and to pubbsb the (2) Not crediting use of containment requirements that would include final nde in mid-19H.

senting during the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period defimtion of specific challenges posed followmg the onset of core damage in by sevece accedents and issue them as SPaeh Considerations evaluating the design for compliance changes or additions to appendix A.

with the containment performance De NRC invites comments and

" General Design Cnteria" (GDC). to to objecute in paragraph (b) above.

recommendations from interested CFR put 50. Fach new GDC would (c) Equipment Survivability. Features desenM the nature of the severe persons on g g ggm g g provided for severe accident preventim accident challenge or containment load I"P"*d and mitiganon shall be designed to altunatives,if desm Furdem &

operate for the time period needed in the as well es a ruecess enterion. Ususity, success would be defined eimply as

~

environment {e g.. pressure, ternperature-maintenance of the containment kgal and Weal ennam n 6 tadianon)in which the equipment is function for an appropriate period I U "I"8 relied upon to function. inchtding fMiowmg th partieder chaBenge.

1. Is a ruleavAing addressing sevete considerat2on of the circumstances of applicable initiating events (e g, Regulatery Godes would be developed acadcot plant perfonnance entens to provide additional gddance on items deszrabler if so. why? If not. why nott trar.sients. loss of AC power. loss of coolant accidents)-

such as analysis metbeds ar'd WW a nde pnmde bener derm The approach in this phenomena' assumptions.The ACRS orthned tbfe end pW% b ee des # remw onented alternatrve would be similar to approach in more detail in a letter to and certiLeatman procesees for feture the hardware-onented alternative in NRC Charrman K. Carr. datedMey 27..

We deswas arls rWemabag on gpgg,e these Laanes via todrvidual derrgn that it is prescriptive regard.ng the certification auf5cient?

severe accident phenomena which snust l

be addressed in the desig;r however,it g

the toc vst.rit Docauneet Rom. 2120 L Sheet. FN Reghwy Rmard M Ma&$m l

does provide flexibility for the designer g w i,,,g w,

,% oc5.,%,p com-w e emn. oc musarw m.

e to propose solutions specific for the

...Jn f=n w Tbmas s;ms ONr of fudear M *s u mi l

s

]

f.

Federal Register /;Vol. 57. No.188 / Monday, September 28. 1992 / Proposed Rules 44318

=

a

2. Would a new rule in to CFR part 50 generation 1 or is a general stateinenrof-- and 2 (Le., containment sh barrier against the release of radioactive the chauenge more desirable? 5 concerrung plant perforTnance for severe

7. For what reason le g., not a mk -

material for a period of apprournately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the onset of core accidents. as discussed in the thee attematives, provide a basis for reviring significant phenomena. not a cost damage) represents a level of safety for effective solution) would any of the the requirements on Emervency cntena proposed in the three a 3&x)Mwt plant sited in accordance Plannmg Zones for future LWRs7 ff so.

attematives not be fully apphcable to with 10 CFR part 100 approximately utiy? If not why not7 passive designed LWRa?

three orders of magnitude below the

3. One option for an overal EL What features could an advanced Commission's quantitative health contMnment performance cntenon that LWR desqrn include thetwould prevent obtective for prompt fatahties. as has been considered is that the or mingate fuel-coolant interections7 defined in the Commission's Safety Goal conditional failure probability of the

. 9.lf a design mcludes the capabihty t Policy Ststement. It could be argued that cootsmment should be less than rapidly depressurue the pnmary system. a future LWR design meeung this approximately one in ten. Two of the should it also be requtred to have a objective through analyses and the alternatives use a detenntnestic reactor cavity design and/or a reactor incorporation of design features need surrogate that states that the vessel support structure capable of not consider the addition of other containment should re. main leak tight for mitigating and acco ating a high a penod of appecximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> wou d be directed at even more highly foUowmg the onset of core damage and.

Pg$S opfu W ndude unlikel

• "em addat phmomma i

WR des' Y

after that time. remain a barner agamst an on4me instmmmtauon system that and sequences which could be, _

the uncontrolled release of radioactmty monitors containment atmosphere for considered remote and speculative when faced with challenges from the grosa leakage to reduce the nsk from an under the National Environmental Pohey more likely severe accident phenomers. inadvertent bypass of containment Act (NEPA) and 20 CFR part $1.

Is this criterien a suitable substitute for functiopf Would appbcation of this.

hnfore, would the codification and the conditional containment failure system be sdscient basis to moQ Icak probability of one in ten? If so. explain rate testing requirements cdct 10 CFR compliance with such a containment wby. H not. explain any not. Is a penod part 5a appeda L **Pmnan Reactor performance objective be sufficient to of epproximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> an Catammmt 14akage Tesung for also define a point of truncation and appropriate tune frame? Is its degree of Water-Cooled Power Reactors.

serve as the basis for an amendment to 10 CFR part 51 eliminating the need for conservatism appropriate considermg

11. Wl af design critena should be.

6tncertainties and defenscan-depth? If developed that provida assurance that further review of SAMDAs for future not, what alternat2ve would be the containment s integnty could easily. LWRs under 10 CFR part 51?

appropriate? %bt other cr.tema be established denng ceytato shutdown.

.IVe rehmina'Y views expreued m (probabtitstic or deterministic) might be conditions?

this ANFRM nay change after consideredT

12. Should equipment provided only-considering the comments received. In

^Itemathe 2 wouM remu.re for severe accident prevention or any case. the NRC will provide an extensive reliance on analytical tools miti ation be subject to (a} the same opportunity for additional public 6

that calculate the effects of severe requirements as design basis equipment comment on any proposed rule accident pher omena. Are them te g redundancy / diversity, power developed as a result of this notice.

ana ytical toow that are sufficiently supply, environmental qualifica tion, developed and adequate to aUow inclusion in plant Technical Ilst of Subjects in 10 CFR Part 50 effect2ve unplementation of such a Specifications, maintenance priortry.

phenomena.bned rule? If so what are -

quauty assuranceh or (b) lesser Antitrust Classifiedinformation.

they, and for what phenomena could 8 * "*

• 8 Cnminal penalties. Fire protection.

they be used? How would attemative 2

'h*

[ig' d

inc rp ration by reference.

o se implemented? For example, should appendice. A and B of Regulatory Guide Intergovernmental relations Nuclear the codes and input parameters be 1.155. 'Sta tion Blackout?"). If lesser Power plants and reactors. Radiat2on approved by NRC7 Stould acceptanm standards, whM standards wodd k protection. Reactor siting entena.

[

entene be codified or put in a regulatory appropriate?

Reporting and recordkeeping guide?

1.3. Attemative 1 discunes not.

requirements.

5 Should future LWR containment exceedmg ASME service level C stress designs include features beyond those hmits for steel containmente under desenbed in altemative 1 to prevent /

certain severe accident cond>tions. Are d cument is: Sec.181. Pub.'L aM03. 68 mitfpate severe accidents? If so, what these limits appropriate for severe Stat. 948. as amended [42 U.S.C. 2201L accident conditions? If not, what limits Sec. 201. Pub. L 93-438 88 Stat.1:42 a s are they?

would be appropnate? Could these same amended (42 U.S.C. 5841).

6. Altematives 2 and 3 specify stress limats also be used for loads

,f phenomenological severe accident generated by missiles? If not, what Dated at Roc.kville. Maryland. tius 22d de y challenges that should be considered in Irmits would be appropriate? Wht I

the design. Alternative 1 is based upon equivalent limits would be appropriate l

the same phenomene/challerwes. Are for concrete contalnmentst

%g p

there other severe accident phenomenaf

14. Wht information Is availab'le.

i challenges that abould be considered?

regarding the costs (capital add -

'"^

' operatiorfal/ maintenance)of derign

>=t 1 04k.

j

%ht should be the entena for deciding

~

t whether a severe accident phenorcens features that would be required under Secrerary of the Comuniumn.

I or challenge is likely and should be considered? Should the challenges be these altematives?

p Doc med e ao aN 15.The containment performance epecified in more detaillior ex. ample, objective dacussed in Altematives 1 earme coat nme-a epecifying the amount of hydrogen e

e

_