ML20207L918
| ML20207L918 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 01/07/1987 |
| From: | Bloch P, Jordan W, Mccollom K Atomic Safety and Licensing Board Panel |
| To: | |
| References | |
| CON-#187-2114 79-430-06-OL, 79-430-6-OL, LBP-87-01, LBP-87-1, OL, NUDOCS 8701130035 | |
| Download: ML20207L918 (4) | |
Text
4.
2f/ f D0LKETE0
'J 5tli' C UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION
'87 JAN -7 A10 :46 Before Administrative Judges:
Peter B. Bloch, Chairman d im. ~ ' I
- Dr. Kenneth A. McCollom Dr. Walter H. Jordan SERVED JAN 0 71987 In the Matter of Docket Nos. 50-445-OL
)
50-446-OL TEXAS UTILITIES ELECTRIC COMPANY, et al.
ASLBP No. 79-430-06 OL (Comanche Peak Steam Electric Station, Units 1 and 2)
)
)
January 7, 1987 MEMORANDUM (Adequacy of Record: Errors and Sampling)
As the case has progressed, the Board has continued to consider the relationship between the requirements of Appendix B and the acceptabili-ty of the Comanche Peak Response Team's (CPRT's) alternative approach, consisting of inspection of samples of hardware.
In this memorandum, we set forth concerns that we have determined ought to be addressed in the interest of an adequate record.
We understand that the sampling process employed by the CPRT is to provide a screen for detection of the existence of deficiencies within a given population.
However, we have not seen an adequate justification for a 95/5 (or 95/95) sampling program as the screen for all systems.
Althcugh Texas Utilities Electric Company, et al. (Applicants) has said that it does not rely entirely on the inspection of samples, we do not
$0 oy4, go,,
enn
\\
Adequacy:
2 I
understand how the CPRT's other work improves the level of assurance of the program of reinspection by sampling.
For example there may be areas of the plant or types of hardware for which Applicants will rely en-tirely on reinspection through sampling.2 This reliance on the reinspection of samples may affect one of the post TMI-2 requirements: the "[p]erformance of a simplified AFWS reli-ability analysis that uses event-tree and fault-tree logic techniques to determine the potential for AFWS failure under various loss-of-main-feedwater-transient requirements."3 It appears to us that reliance on the CPRT reinspection program could require a revision to the required-analyses.
I See Applicants' Memorandum in
Response
to Board Memorandum ITtatistical Inferences from CPRT Sampling), January 31, 1986 at p.
17.
We do not find precisely the same kind of statement in Applicants'
Response
to Board Concerns
[ Concerns Response],
December 1,1986, so we are not sure whether Applicants continue to rely on the same reasoning concerning not relying entirely on a sampling program. Furthermore, we note that the use of two.95/.05 screens may have some effect on the level of confidence or level of error of the sampling program; and Applicants may wish to discuss the effect of double sampling to assist us to understand this research tool better. Concerns Response at 6-7.
2 To the extent that the CPRT credibly confirms the reliability of all or part of the QA/QC program, this would of course enhance the credibility of findings that portions of the plant are adequate for safety because they pass the CPRT's sampling screen.
The difficulty we forsee, concerning the need to rely on the sampling program, is most relevant should the CPRT discredit substantial portions of QA/QC or should the evidence about the QA/QC program be equivocal.
3 Safety Evaluation Report, NUREG-0797, July 1981, at 22-38.
]
v Adequacy:
3 We are interested in the relationship between the 95/5 sampling program and the level of safety achieved at nuclear plants by properly certified, trained, and supervised craft personnel and by an appropriate quality assurance / quality control program (QA/QC), which would catch and reduce errors made by craft.4 We recognize that errors are expected. But what level of errors is expected? What level do codes anticipate? What effect does the error level have on probability risk assessments?
We are not setting any particular time deadline for a response, but we are interested in reading the response we will receive.
Intervenors may respond. within a month of Applicants' filing.
Staff may respond with all deliberate speed.
4 We do not entirely accept the argument that, "If a population passes the 95/95 sample screen, the populat'nn of items is free from programatic deficiencies."
Applicants Memorandum in Response to Board's Memorandum, January 31, 1986 at 9.
All we would conclude from a population passing the 95/95 sample screen is that it is probably free from programatic deficiencies that result in an error rate of five percent or more. The question this raises is whether that is adequate assurance of safety.
a 1
Adequacy:
4 FOR THE ATOMIC SAFETY AND LICENSING BOARD Peter B. Bloch, Chairman ADMINISTRATIVE JUDGE b' alter 4. Jfrc.an u
ADMINISTRATIVE JUDGE c
Kenneth A. McCollom ADMINISTRATIVE JUDGE Bethesda, Maryland