ML20244E526
| ML20244E526 | |
| Person / Time | |
|---|---|
| Issue date: | 12/18/1986 |
| From: | Ballard R Office of Nuclear Reactor Regulation |
| To: | Hulman L Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20244D710 | List:
|
| References | |
| NUDOCS 8612240163 | |
| Download: ML20244E526 (5) | |
Text
_ _ _ _ - - _ - _ _ _ - _ _ _ _ _ _ _ _ _ _
Ref. 10 i .
DEC 18 386 H
MEMORANDUM FOR: .L. Hulman, Chief, PSB, BWR L. B. Marsh, Chief. EB, DPLB W. Minners, Chief, RSIB, DSRO W. Shields, OGC FROM:
Ronald L. Ballard, Chief, EB, DPLA GENERIC LETTER, CONTAINMENT INTEGRATED
SUBJECT:
LEAK RATE TESTS 26, 1986, r? quested concurrence or comments
' Tom Novak's memorandum of November from NRR Division Directors on a proposed gen (ric letter to all light water reactor applicants or licensees regarding integrated leak rate tests'(CILRT).
The generic letter was to be transmitted to J. Sniezek for information prior to its general release.
November 26 memorandum (copies attached), suggesting alternatives or needed improvements to the generic letter.
The November 26 version of the generic letter has been redrafted, incorporating a number of suggestions that have been received. Included in the new draf t are references to-the Appendix J revision that has been released for connent, a
" grandfather clause" describing the letter's applicability In addition, only to afuture CILTsT we are preparing
. activities, and technical clarifications.
- brief Regulatory Analysis that will accompany the generic letter to J. Sniezek's office.
The alternative of issuing a limited scope rule change rather than a generic letter was considered at length on several occasion and has attractive features. However, our present objective Thus, your is to proceed with the preparation suggestions are solicited on the of an acceptable generic letter. In the interest revised letter, a copy of which is attached for your review.
of expediency, handwritten marginal notes or editorial revisions will be suitable. Comments are needed by COB on Friday, December 19, in order to meet Give me (X27433) or Jim Pulsipt,er (X27793) a call our established schedules.
if additional information is needed.
Origina! Signed By Ronald L. Ballard, Chief Engineering Branch Division of PWR Licensing.A DISTRIBUTION:
Attachment:
As stated Central Files JPulsipher PAEB Reading File cc: See next page RBallard GBagchi G 4 : : :
- PAEB o . :
- PAEB OFC : PAE6 d : : : : :
er:ws GBa : RLBal NAME :JPuls' DATE :1 /86 12/ /86 : 12//4/86 :
OFFICIAL RECORD COPY b_ .
h
4 Those Listed DEC 18 B86 cc: R. Bernero F. Miraglia .
T. Novak
/
C. E. Rossi
DRAFT T0i All Light Water Reactor Applicants and Licensees
- f. Sf!BJECT: CONTAINMENT INTEGRATED LEAKAGE RATE TESTS Gentlemen:
A recent NRC review of the different types of test computations used to determine leakage rate for containment integrated leakage rate tests (CILPTs) performed throughout the industry has indicated that applicants and licensees are frequently usino a test computation method that is not in compliarce with the explicit regulatory requirements of 10 CFR 50, Appendix J. The method being used for computation is the mass point calculation described in ANSI /ANS-56.8-1981, " Containment System Leakaae Testing Requirements."
In order to meet the regulatory requirements of Appendix J, applicants and licensees must measure the integrated leakage rate using one of the CILRT methods described in ANSI N45.4-1972, " Leakage-Rate Testing of Containment Structures for Nuclear Reactors."Section III.A.3.fal of Appendix J requires that all CILRTs shall be conducted in accordance with the provisions of ANSI N45.4.Section III.A.3.(c) of Appendix J also requires that test leakage rates be calculated usina absolute values corrected for instrument error. Section 5.1 of ANSI N45.4 permits two methods of performino CILRTs, the absolute method and the reference-vessel method. In addition, Section 7.9 of the standard describes the types of leakage rate computations permitted for the P4-hour period absolute method CILRTs. The only types of leakage rate computations specified in this standard are the point-to-point and the total time calculations. Hence, compliance with the regulatory requirements of Aopendix J for absolute method CILRTs can only be achieved through the use of the point-to-point or total time leakage rate computational methods, corrected for instrument error.
It should be noted that the total time test methodology presented in the Bechtel Topical Report BN-TOP-1, Revision I, " Testing Criteria for Integrated Leakage Rate Testing of Primary Containment Structures for Nuclear Power Plants" has been determined to be in accordance with ANSI N45.4, with a test duration as defined in BN-TOP-1.
i i
r _ - .. __ _ -. _ __ _ -- . _ - _ - _
4 DF. AFT ANSI N45.4 also requires that a " statistically averaged" leakage rate be
- determined (Paragraph 7.9). . However, ANSI N45.4 does not provide a specific method for accounting for random and instrument error during a CILRT, as required by Appendix J. The NRC staff believes that the calculated leakage rate as determined by using the mass point method described by ANSI /ANS-56.8-1981- has technical merit. Although the NRC staff has in some instances accepted test results calculated by this method, recent review has determined that compliance with the regulatory requirements requires a CILRT leakage rate calculated by total time or point-to-point, with a test duration of at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or as provided for in BN-TOP-1. Random and instrument error is.. typically accounted for by calculation of a statistically-determined
.95% confidence level. For a test with a duration of at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, the 95%
confidence level calculation will be acceptable if it is derived from either:
(1) the total time or point-to-point method; or (2) the mass point method, as provided in ANSI /ANS-56.8-1981. For a test with a duration of less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, the leakage rate and the 95% confidence level must be calculated in accordance with BN-TOP-1. The calculated leakage rate, with the 95% confidence level amount added to it, must then meet the acceptance criterion of Appendix J
'. (i.e., not exceed 0.75 La or 0.75 Lt, as appropriate).
The supplemental, or verification, test required by Section III.A.3.(b) of Appendix J must also have its leakage rate calculated by the same method (i.e.,
total . time or point-to-point) as is the CILRT.
The regulatory requirements concerning acceptable calculational methods, described above, are applicable only to future CILRTs, and will not be applied to the calculational method of any test previously determined by the staff to l
, be acceptable. It is suggested that applicants and licensees review their CILRT programs with respect to the above to assure future consistency with regulatory reovirements.
In a related action, the NRC has released for public comment a general revision to Appendix J to 10 CFR 50 (51 FR 39538, dated October 29,1986). This I revision would, among other changes, relax existing constraints on CILRT l-~ - _ _ _ - - - - _ _ _ _ _ _ _ _ . _ - _ - - - - _ _ __ _ --_- _ - _ - _ --_--_ _ __ _ ------_ _ - - _ - - _ - _ - _ - - - - - _ _
,s -a l
i DRAFT .
calculational methodologies and would pemit the use of the mass point method for future CILPT determinations. A proposed Regulatory Guide (MS 021-5, October, 1986) has also been published for comment as a companion document to,
, and referenced by, the Appendix J revision. The draft Regulatory Guide identifies the mass point method, with certain constraints, as an acceptable method for future CILRT determinations, in accordance with the revised Appendix J. However, until a revision is complete and in effect, the current i Appendix J requirements apply.
No written response or specific action is required by this letter.
Therefore, no clearance from the Office of Management and Budget is required.
. If you have any questions on this matter, please contact your project manaaer. ,
Sincerely, Harold R. Denton, Director Office of Nuclear Reactor Regulation L __ _..m.____ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -