ML13330B194
| ML13330B194 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 04/10/1987 |
| From: | Medford M SOUTHERN CALIFORNIA EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 8704160265 | |
| Download: ML13330B194 (15) | |
Text
Southern California Edison Company P. 0. BOX 800 2244 WALNUT GROVE AVENUE ROSEMEAD. CALIFORNIA 91770 M.O.MEDFORD April 10, 1987 TELEPHONE MANAGER OF NUCLEAR ENGINEERING (818) 302-1749 AND LICENSING U. S. Nuclear Regulatory Commission Attention:
Document Control Desk Washington, D.C. 20555 Gentlemen:
Subject:
Comments on Proposed Revision to Appendix 3 of 10 CFR 50, "Containment Leak Rate Testing" San Onofre Nuclear Generating Station Units 1, 2 and 3
Reference:
Federal Register Notice Volume 51, No. 209; pp. 39538-39541 In the above referenced Federal Register, the Nuclear Regulatory Commission proposed a revision of 10 CFR 50, Appendix 3, "Leakage Tests for Containments of Light-Water-Cooled Nuclear Power Plants" and invited comments to be submitted by January 26, 1987. This comment period was subsequently extended to April 24, 1987 by Federal Register Notice dated January 16, 1987.
Southern California Edison Company (SCE) appreciates the opportunity to comment on the proposed revision to 10 CFR 50, Appendix J. The significance of the proposed revision to the nuclear industry and SCE is substantial and, contrary to the Federal Register Notice, does contain new containment leak rate testing criteria.
Provided as Attachment I to this letter are the responses to the 15 specific questions put forth in the Federal Register Notice. Provided as Attachment II are SCE's comments regarding proposed Regulatory Guide MS 021-5, "Containment System Leakage Testing."
This proposed Regulatory Guide would become the basis for acceptance of Containment Integrated Leak-Rate Testing and as such, must be evaluated in the context of the proposed revision to Appendix 3 to 10 CFR 50.
Even though SCE has provided a response to the specific questions listed in the Federal Register Notice, SCE does not consider that the proposed revision to 10 CFR Appendix 3 should be promulgated either in its entirety or in part. The basis for this is the backfit analysis of the proposed revision which states:
"The analysis does not conclude that there is a substantial increase in the overall protection of the public health and safety or the common defense to be derived from the backfit."
8704160265 870410 0
PDR ADOCK 05000206 P
-PDR
Document Control Desk April 10, 1987 Given this conclusion and the NRC's intention to revise Appendix J again in a year or two (Question #8), SCE requests that the proposed revision to Appendix 3 to 10 CFR 50 be withdrawn in its entirety.
Very truly yours, CEW:7940F Attachments
0 ATTACHMENT I
RESPONSE TO FEDERAL REGISTER QUESTIONS REGARDING PROPOSED REVISION TO 10 CFR 50, APPENDIX J "LEAKAGE RATE TESTING OF CONTAINMENTS OF LIGHT-WATER-COOLED NUCLEAR POWER PLANTS"
- 1. The extent to which these positions in the proposed rule are already in use.
Response
The majority of the "revisions" provided in the proposed Appendix 3 reflect current interpretations of the regulations and industry practice. However, the remainder of the changes are not currently in use because they do not represent desirable practices nor would they provide increased assurance of containment integrity or public safety.
A containment that would pass the existing Appendix J test acceptance criteria would also pass the proposed Appendix 3 test acceptance criteria but would require a longer test duration to show acceptable results.
The existing containment penetration and isolation valves will also be able to provide acceptable test results when evaluated by the proposed Appendix 3 acceptance criteria. The difference between the current and proposed Appendix 3 is that the proposed Appendix J will increase the difficulty of demonstrating acceptable test results with the consequence of increasing test duration and critical path outage time. Therefore, because the existing containments and containment penetration and isolation valves will continue to be in use by licensees without modification, the proposed Appendix 3 will not provide increased assurance of containment integrity or public safety.
- 2. The extent to which those (positions) in use, and those not in use but proposed, are desirable.
Response
The proposed revision to 10 CFR 50, Appendix 3, has some desirable elements. The proposed revision is less confusing to interpret, provides for the relaxation of unnecessarily stringent requirements (airlock testing), and provides for the performance of a containment integrated leak rate test (CILRT) of less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. This last item, the possibility of a reduction in test duration is of great importance in that the performance of a CILRT is always a critical path plant outage item. Any reduction in test duration would thus decrease outage time by a corresponding amount.
However, the above advantages of the proposed rule are outweighed by the following:
- 1. Reference to Draft Regulatory Guide MS 021-5 as the CILRT acceptance criteria (see Attachment II of this letter),
-2
- 2. NRC planned comprehensive revision to Appendix 3 within the next "year or two" (see response to question 8),
- 3. Increased reporting requirements (see General Comment 2),
- 4. Required Technical Specification changes (see General Comment 2), and
- 5. Proposed revision to Appendix J is considered to be an improper Backfit (see response to question 7).
The current rule, and the practice in use, allows for the use of ANSI N45.4-1972, "Leakage Rate Testing of Containment Structures for Nuclear Reactors" which is the appropriate and sufficiently conservative standard for the conduct of CILRT's.
- 3. Whether there continues to be a need for this regulation.
Response
Yes.
The containment structure is essentially a very large pressure vessel and all safety related pressure vessels require periodic testing.
- 4. Estimates of the cost and benefits of this proposed revision, as a whole and of its separate provisions.
Response
Based on the fact that the proposed rule is a clarification of the existing Appendix J requirements and test acceptance criteria, the impact as a whole of the revision is extensive. However, most of the separate provisions of the proposed rule will have a minor impact. The proposed revision, by referencing Draft Regulatory Guide MS 021-5 rather than the existing ANSI N45.4-1972, would significantly impact the cost (direct test expense and additional critical path outage time) without increasing the benefit of existing programs or increasing public health and safety.
- 5. Whether present operating plants or plants under review should be given the opportunity to meet the current Appendix 3 provisions if the proposed rule (reflecting consideration of public comments) becomes effective.
Response
It is considered necessary that licensees be allowed to continue to meet the provisions of the existing Appendix 3. As indicated in the preamble of the Federal Register notice of this proposed regulation, the existing Appendix 3 has been in effect since 1973. Therefore, a considerable portion of licensees have implemented leak rate programs which satisfy the provision of the existing regulation.
0 0
-3
- 6. If the existing rule or its proposed revision were completely voluntary, how many licensees would adopt either version in its entirety and why.
Response
It is expected that licensee's with programs presently complying with the existing Appendix 3 would continue to retain their programs. The proposed rule would require CILRT program changes but would not reduce cost or increase safety. An undesirable aspect of the proposed rule is referencing the proposed Draft Regulatory Guide as acceptable guidance in satisfying the requirements of the proposed rule. The regulations are not considered to be the place for providing guidance in satisfying the rules. The Technical Specifications or the FSAR is considered the appropriate forum for identifying acceptable methodologies to be used in complying with the regulations.
The referenced draft regulatory guide identifies a new methodology for demonstrating compliance with the requirements of the proposed rule. As indicated in the preamble of the notice, this is not the intent of this proposed rule. The scope of the revision is to provide corrections and clarifications. Identifying a new methodology within the proposed rule is not considered by SCE to be a clarification.
As described in the response to question (4), the methodology described in the draft regulatory guide is not considered to be the appropriate CILRT acceptance criteria for satisfying the requirements of Appendix 3.
SCE's concerns with the draft regulatory guide are taken up in separate comments provided in Attachment II of this letter.
- 7. Whether (a) all or part of the proposed Appendix 3 revisions would constitute a "backfit" under the definition of that term in the Commissions Backfit Rule, and (b) there are parts of the rule which do not constitute backfits but which would aid the staff, licensees or both.
Response
Under the provisions of 10 CFR 50.109, "Backfitting," a backfit is defined as:
the modification of or addition to systems, structures, components or design of a facility; or the design approval or manufacturing license for a facility; or the procedures or organization required to design, construct or operate a facility; any of which may result from a new or amended provision in the Commissions rules or the imposition of a regulatory staff position interpreting the Commissions rules that is either new or different from a previously applicable staff position."
The proposed revision is intended to provide clarification, provide corrections and exclude new criteria. However, the proposed revision would change the CILRT acceptance criteria from ANSI N45.4-1972 to Draft
-4 Regulatory Guide MS 021-5. Therefore, as the proposed rule would be "an amended provision in the Commissions rules" and does not provide "a substantial increase in the overall protection of the public health and safety," this change to CILRT acceptance criteria would be an unwarranted backfit. The remainder of the changes proposed are not an unwarranted backfit.
- 8. Since the NRC is planning a broader, more comprehensive review of containment functional testing requirements in the next year or two, whether it is then still worthwhile to go forward with this proposed revision as an interim updating of the existing regulation.
Response
If the NRC is conducting studies on containment requirements, this proposed rulemaking is premature. Resolution of all the issues that could influence this particular regulation should be completed prior to initiating a proposed rule. Such action results in licensees expending resources on resolving concerns with the proposed rule which in the broad scope may have been unwarranted and undesired. These efforts would have to be reinitiated when the NRC's long term studies result in a final rule for containment testing.
- 9. The advisability of referencing the testing standard (ANSI/ANS 56.8) in the regulatory guide (MS 021-5) instead of in Appendix J.
Response
The intent of regulations (10 CFR) is to provide the requirements for operating nuclear power plants. The regulations are not a place to provide guidance or direction on how to comply with the regulations. It is appropriate to reference guidance documents or provide methodologies in regulatory guides, NUREG's or other similar formats.
Including guidance in the regulations prevents flexibility in compliance necessitated by plant specific considerations, for example, and prevents timely updating of acceptance criteria.
- 10.
The value of collecting data from the "as found" condition of valves and seals and the need for acceptance criteria for this condition.
Response
SCE believes that there may be some benefit in collecting "as found" leakage data for the detection of possible common mode valve failures between CILRT's. However, the existing criteria regarding "as found" valve leakage has proved to be sufficient for this purpose. Therefore no additional "as found" acceptance criteria are required to demonstrate containment isolation and penetration valve integrity.
-5
- 11.
Whether the technical specification limits on allowable containment leakage should be relaxed and if so, to what extent and why, or if not, why not.
Response
In order to assess the need to relax the containment leak rate limit, resolution of the long term broader issues should be completed. As indicated in the preamble of the Federal Register notice, the NRC is investigating the containment function and testing requirements. Also, the work being performed on the source term issue will have some impact on acceptable containment leak rate. Only after these issues have been thoroughly investigated and resolved can any assessment be made on relaxing the containment leak rate limits.
- 12.
What risk-important factors influence containment performance under severe accident conditions, to what degree these factors are considered in the current containment testing requirements, and what approaches should be considered in addressing factors not presently covered.
Response
Factors which would influence containment integrity would be isolation valve, penetration, and airlock leak tightness. These factors are currently addressed by the existing Appendix 3 requirements and plant specific programs. No additional factors are considered to be significant contributors to the overall containment leak rate.
Therefore, additional factors need hot be included in Appendix 3 or the plant specific CILRT programs.
- 13.
What other approaches to validating containment integrity could be used that might provide detection of leakage paths as soon as they occur, whether they would result in any adjustments to Appendix 3 test program and why.
Response
SCE is not currently prepared to identify other possible approaches to validating containment integrity. Other approaches to validate containment integrity should be the subject of extensive studies. An important part of these studies would be to determine the feasibility of implementing and performing any new or alternate approach.
- 14.
What effect "leak-before-break" assumption could have on the leakage rate test program. Current accident assumptions use instantaneous complete breaks in piping systems, resulting in a test program based on pneumatic testing of vented, drained lines.
"Leak-before-break" assumptions presume that pipes will fail more gradually, leaking rather than instantly emptying.
-6
Response
The leak-before-break issue is another area that needs to be included in the overall investigation of containment integrity. This can be used in relaxing the allowable containment leak rate. Again it is emphasized that the broad scope of this issue needs to be resolved before a proposed rule is approved. The resolution may result in an overall relaxation of the regulatory and technical specification requirements relating to containment integrity.
- 15.
How to effectively adjust Type A test results to reflect individual Type B and C test results obtained from inspections, repairs, adjustments or replacements of penetrations and valves in the years in between Type A tests. Such an additional criterion, currently outside the scope of this proposed revision, would provide a more meaningful tracking of overall containment leak tightness on a more continuous basis than once every several years. The only existing or proposed criterion for Type B and C tests performed outside the outage in which a Type A test is performed is that the sum of Type B and C tests must not exceed 60% of the allowable containment leakage. Currently being discussed by the NRC staff are:
- a. All Type B and C tests performed during the same outage as a Type A test, or performed during a specified time period (nominally 12 months) prior to a Type A test, be factored into the determination of a Type A test "as found" condition.
- b. If a particular penetration or valve fails two consecutive Type B or C tests, the frequency of testing that penetration must be increased until two satisfactory B or C tests are obtained at the nominal test frequency. Concurrently, existing requirements to increase the frequency of Type A tests due to consecutive "as found" failures are already being relaxed in the proposed revision of Appendix 3.
Instead, attention would be focused on correcting component degradation, no matter when tested, and the "as found" Type A test would reflect the actual condition of the overall containment boundary.
- c. Increases or decreases in Type B or C "as found" test results (over the previous "as left" Type B or C test results) shall be added to or subtracted from the previous "as left" Type A test result.
If this sum exceeds 0.75La but is less than 1.0 La, measures shall be taken to reduce the sum to no more than 0.75 La.
This will not be considered a reportable condition.
If this sum exceeds 1.0 La, measures shall be taken to reduce the sum to no more than 0.75 La. This will be considered a reportable condition.
The existing requirements that the sum of all Type B and C tests be no greater than 0.60 La shall also remain in effect.
-7
Response
(a) This proposal is not adequate since it could possibly require the performance of Type B and C tests during plant operation. Test performance in these plant modes is significantly more difficult and hazardous because of increased radiation levels, containment access restrictions, different system pressures and temperatures, and plant systems equipment availability. The current criteria regarding the "as found" leakage condition are adequate.
(b) This proposal could possibly require Type B and C testing during plant operation and is subject to the same concerns identified in response (a) above. Type B and C test results do not directly correlate to Type A test results. With this concept in mind, it is not clear how performing Type B and C tests during plant operation could significantly increase the confidence of containment integrity.
(c) This proposal would directly link Type A test results to Type B and C tests on a continuing basis. This concept is not valid since every Type B test has at least one pressure boundary tested in the "reverse direction" and almost every Type C test is conducted on at least two valves, even though only one is assumed to leak. At the very least this proposal would require that maintenance test boundary isolation valves be maintained in a condition approaching that of a containment isolation valve. This would expend considerable licensee resources with no significant increase in the confidence of containment integrity.
General Comments
- 1. The proposed revision to 10CFR50, Appendix 3, is intended to clarify the current regulations. However, in Section IIIA(6) of the proposed Appendix 3, the acceptance criteria are not clear. It is suggested that the last sentence of this section be revised to include the formula in ANSI N56.8 paragraph 3.2.6(B)(1), if the proposed revision is promulgated.
- 2. The Federal Register Notice (Reference) states that:
"The scope of this revision to Appendix J is limited to corrections and clarifications, and excludes new criteria."
However, the proposed revision to Appendix J includes several new reporting requirements and required Technical Specification changes.
Specifically, the following sections of the proposed Appendix 3 would require increased reporting:
(a)Section III A.(4) - Proposed change would require the licensee to include new information in the report to the Commission on valve leakage requiring corrective action before, during or after the test specified in Section VI of Appendix 3.
-8 (b)Section III A.(8)(b) -
Proposed change would add the requirement for the licensee to submit, for NRC review and approval, a corrective action plan when any Type A CILRT fails the acceptance criteria.
(c)Section V.B - Proposed change would add the requirement for Type B and C tests performed during the same outage as a Type A test to be documented and reported for both "as found" and "as left" conditions.
(d)Section VI.A(2) - Proposed change would add the requirement that intermediate Type B and C test failures be reported within 30 days of test failure.
(e)Section VII.B - Proposed change would require licensees to submit an implementation schedule for the revised Appendix 3 180 days after the rule is effective.
All of the above items would increase the burden on licensee resources by increasing reporting requirements. The following-sections of the proposed Appendix 3 would require Technical Specification changes:
(a)Section V.A. - Proposed change would add the requirement for licensees to include leak test methods, procedures, and analyses for a steel, concrete or combination steel and concrete containment and its penetration and isolation valves in the Technical Specifications.
(b)Section III.B(4)(d) - Proposed change would add the requirement for licensees to include the individual acceptance criteria for all airlock testing in the Technical Specifications.
These Technical Specification changes identified above are contrary to the Commissions policy statement on Technical Specifications dated February 6, 1987.
Superficially, Criterion 2 of the policy statement appears to include containment leakage as a process variable as an initial condition of a Design Basis Accident. However, upon examination of this criterion, as provided in the policy statement discussion, it is clear that Criterion 2 is intended to apply only to "those process variables that have initial values assumed in the Design Basis Accident and Transient Analyses, and which are monitored and controlled during power operations."
Therefore, as containment leak rate test methods and acceptance criteria are not process variables controlled by the operator, they do not belong in the Technical Specifications.
- 3. In addition to the above items, as discussed in the response to Federal Register questions (4), (7) and (9), SCE considers the Draft Regulatory Guide MS 021-5, as the new NRC CILRT acceptance criteria, to also be a new requirement of the proposed revision to Appendix 3. It is therefore recommended that footnote (1) in Section I, "Introduction" be removed if the Commission proceeds with promulgation of the proposed revision.
-9 Response to Request for Comments on the Application of the Backfit Rule The Federal Register Notice asks whether the Commission should continue to apply the Backfit Rule to all rulemaking or whether the Backfit Rule should be revoked as it applies to rulemaking. Further, it is asked whether the Commission should amend the Backfit Rule to waive the "substantial increase" provision and to indicate explicitly that the non-monetary benefits may be weighed by the Commission in the cost-benefit balance, when such considerations are found by the Commission to be in the public interest.
Response
In response to the first portion of the above question, SCE considers that the Commission should continue to apply the Backfit Rule to proposed rulemaking.
The imposition of a backfit by the Commission has the same impact on the licensee whether the backfit is required by a plant specific order or by rulemaking. Not requiring a Backfit Analysis for proposed rulemaking would strip the Backfit Rule of its original intent by creating a regulatory path around th e Rule. Additionally, imposing a new rule that requires the expenditure of licensee resources but does not "increase the overall protection of the public health and safety or the common defense and security" is what the Backfit Rule was designed to prevent. Therefore, the Commission should continue to require a documented Backfit Analysis regardless of the source of the backfit.
It is also difficult to understand how the application of the backfitting rule would require the NRC to expend "resources wholly disproportionate to any conceivable benefit to the public."
The Commission has previously acknowledged that the systematic and documented analysis required by the Backfitting Rule represents essentially the same type of analysis performed by the Commission in the past in considering proposed requirements. Thus, the Commission has noted that the backfitting analysis has "precedent in existing NRC practices as seen in the Regulatory Analysis Guidelines of the U.S.
Nuclear Regulatory Commission, NUREG/BR-0058, the approved CRGR Charter and the Commission's approved plan for the management of plant-specific backfitting..
." (50 Fed. Reg. at 38103, col. 3).
The continuation of this type of prudent analysis cannot be considered a new burden on NRC resources in excess of the public benefit.
Responding to the second portion of the above question, SCE considers that the Commission should not waive the "substantial increase" provision of the Backfit Rule under any circumstances. Again, the basis for this position is that the ability to waive this provision would create a regulatory path around the Backfit Rule thereby preventing the Rule from performing its intended function.
CEW:7945F
ATTACHMENT II
COMMENTS ON DRAFT REGULATORY GUIDE AND VALUE/IMPACT STATEMENT, TASK MS 021-5, "CONTAINMENT SYSTEM LEAKAGE TESTING" As indicated in SCE's comments on the proposed revision to Appendix 3, the proposed draft regulatory guide MS 021-5 is not considered to be a desirable replacement to ANSI N45.4-1972 as the appropriate CILRT acceptance criteria.
Provided below are specific comments on this draft regulatory guide.
Section B. DISCUSSION Item 12 and 13 Because CILRT's must be started once the containment air mass has reached temperature stabilization, most CILRT's have an air mass versus time plot that appears as a decreasing parabolic function which reaches an asymptotic level.
(At the start of data recording, the indicated mass leakage rate may be three to four times greater than the allowable leakage rate.)
The result is that a test that would show acceptable results with the current methodology would show a test failure with the proposed methodology.
This is due to the introduction of a "limit of curvature" criterion in the proposed methodology. Without the current 24-hour test methodology and acceptance criteria contained in ANSI N45.4-1972, the CILRT would initially fail based on the "limit of curvature" criterion (After test failure, the "failed data" would be thrown out and not considered in the next CILRT attempt). The CILRT would be restarted and continue to fail until the parabolic appearance of the curve was reduced enough to pass the proposed new "limit of curvature" criteria. At this point, a final test could be started that would show acceptable results.
This iterative process has the result of needlessly increasing containment leak rate test duration and thus increased plant outage time, but does not increase confidence in containment integrity.
The limit of curvature criterion is too restrictive, and is based on false assumptions regarding containment air mass temperature stabilization.
Additionally, as most licensees have a CILRT program based on the existing acceptance criteria, the option of completing a test utilizing the criteria of ANSI N45.4-1972 should be allowed.
Section C. REGULATORY POSITION 6.(5)
The basis for starting the verification test "as soon as possible" following the Type A test is not clear. SCE suggests that the start time should be left to the discretion of the test director to prevent verification test failure due to possible conditions unrelated to containment integrity.
-2 12.2 The second stabilization criterion needs to be clarified further as to basis and method of calculation. An example of the calculation similar to that provided in ANSI 56.8 for the first criteria should be included.
13.1 After temperature stabilization of the containment air mass, the start time of the test should be left to the discretion of the test director alone.
13.3 It is not clear why a "limit of curvature" criterion must be applied to the Mass-Plot method of ANSI N56.8 if the curve is smooth and commences at a value greater than the allowable leakage rate. Currently, all the Mass-Plot data from test start to finish is evaluated by "least squares" linear regression. Thus, the slope of the "least squares" line (which is the containment leak rate) includes the initial test data that appears parabolic in nature. As the initial data points indicate a containment leak rate that exceeds the acceptable leak rate (from historical CILRT data), the slope of the "least squares" line is always skewed to indicate a higher rate than actually exists. The proposed acceptance criterion is less conservative for the following reason. As discussed in the comments on items 12 and 13 above, if the CILRT data fails the "limit of curvature" criterion, the data is thrown out and the test restarted. When the data passes the "limit of curvature" criterion and the "least squares" line computed, it will not include the initial higher leak rate test data and will for the same CILRT, show a smaller leak rate than the current method. Therefore, as proposed, the acceptance criteria are less conservative yet will require a longer test duration. It is suggested that the CILRT acceptance criteria of the proposed regulatory guide be revised to prevent unnecessary test failures. Also it is suggested that efforts be initiated to extensively verify the proposed new criteria against actual historical CILRT test data to validate its basis and viability.
APPENDIX Define each term used in an equation explicitly in this appendix. Provide examples for use of each of the equations like those of ANSI N56.8. These actions will clarify the correct implementation of this Reg. Guide and provide for more consistent results.
CEW:7953F