ML20212R588
| ML20212R588 | |
| Person / Time | |
|---|---|
| Issue date: | 11/04/1986 |
| From: | Bender M Advisory Committee on Reactor Safeguards, QUERYTECH ASSOCIATES |
| To: | Igne E Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1868, NUDOCS 8702020724 | |
| Download: ML20212R588 (2) | |
Text
'
R$5
[ '~ r ~ - sarety $5E ron= h ms9 aeon g5P 1 - sn * ,y a r..e , u,gg _. -
OUERYTECH A550CIRN5 Engineering and Technical Advisors 535 d7L//47 ESP :
9040 6scutlw Park Drhe, Suite 217 MQ '
j, c, !
Hnoxvtlle, Tennessee 37923 Phone:(615) 690-2728
, L;r d ^[%; j/ 1
'l November 4, 1986 -
Mr. E. P. Igne, ACRS Staff Member Advisory Committee on Reactor Safeguards U. S . Nuclear Regulatory Commission 1717 H Street, 10th Floor Washington, D. C. 20555
Subject:
Review of Regulatory Guide, " Format and Content of Plant Specific Pressurized Thermal Shock (PTS) Safety Analysis Reports for Pressurized Water Reactors" for implementation of PTS Rule 10 CFR 50.61,
Dear El,
My review of the subject document and comments led me to conclude that the basic concepts set forth in this regulatory guide are responsive to the PTS 'rulemaking and should be supported, however, I think that, in total, the procedure is too complicated and puts too much emphasis on showing a probabilistic treatment of the PTS induced vessel failure risk.
The computed risk values, of themselves, have little meaning.
As Dr. Catton showed, by selecting from a menu of combined probabilities one can establish risk values that may be greater than the 10-5 PTS probability goal. I was able by somewhat arbitrary logic to generate probabilities as high as 10-2 as well.
as values as low as 10-6 for the same combination of events.
Everything hinges on operational assumptions. I tend to agree with the Staff argument that the event probability combinations I will normally be overly conservative in the probability estimates i
because too many things will be assumed to go wrong at once, but the Chernobyl potential cannot be disregarded. Hence, I think the guide needs to put most of its emphasis on procedural matters rather than on the analytical procedures for establishing the action threshold. The paper generated by this guide for each nuclear unit could overwhelm the review process.
I offer the following recommendations concerning the guide:
- 1. Require a reference risk analysis for each type of nuclear unit.
hO 4 861104 DESIGNAUCD ORIGINAL Certified By [/ -
L
- 2. Require each licensed reactor unit to perform a compara-tive analysis relating the unit to the reference system and provide a list of control measures to assure that the event probabilities remain within in the desired risk limits.
- 3. Make sure that each reactor owner has a complete record of the materials and fabrication history of his reactor vessels and a carefully reviewed basis for establishing initial RTNDT values for critical high fluence areas of the vessel. cumulative fluence should be tracked by combining cumulative fuel burnup and neutron flux distribution ,over the operating plant history so that changes in toughness as defined by shift in RTNDT will be known in time for needed corrective measures should the occasion arise.
The above list would be a manageable set of responses to the rule that would be understandable to the licensee and the regulatory staff. The content of the proposed regulatory guide could then be used as a preamble to the guide to provide background and to elaborate on the technical issues to be considered.
Sincerely, M. Bender l
-- - --