ML20056C535
| ML20056C535 | |
| Person / Time | |
|---|---|
| Issue date: | 04/10/1993 |
| From: | Ward D Advisory Committee on Reactor Safeguards |
| To: | Kress T Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-2068, NUDOCS 9306240436 | |
| Download: ML20056C535 (1) | |
Text
cc Dean Houston April 10,1993 To:
Tom Kress, ACRS db From: Dave Ward,
[PM hd'[N ACRS consultant
Subject:
Comments on March 11 paper, Minners to Larkins, "ACRS REVIEW OF DRAFT COMMISSION PAPER ON LARGE RELEASE DETERMINATION" I see no reason to disagree with the Staff proposal to discontinue work on development of a large release definition pending public comments on the revised regulatory analysis guidelines. This does not have to mean that the subject is dead. I dont believe it should be dropped forever.
The Staff seems to have done what we asked them to do a couple of years ago, namely perform a series cf sensitrvrty analyses around the parameters important to the concept. The draft paper and its attachments present some of the results of
- heir analyses, but not enough, it would be helpful if they could provide a more comprehensive set. For one thing, the analyses presented all seem to be centered on the idea that a large release is equivalent to a single prompt fatality. That overty limits the ranges of their sensitively analyses.
The memo from Forrest Remick attached to the paper Dean Houston gave me for 1
comn mr.'is interesting. I agree with him that some of the basic definitions should be reconsidered now that v o have had time and more informabon for reflection.
In particular, I now believe the ACRS hierarchy should be somewhat revised. Namely, the 10E-6 for a large release,10E-4 for CDF, and 10E-1 for CCFP should all be at the same levelin a hierarchy. The ACRS proposal had the latter two as subsidiary to the l
10E-6. The reason for a change is this:
The 10E-6 is an adequate surrogate for the OHOs, all by itself, it is probably too conservative, but it is simple and understandable. Except for the part about what is a large release. Given what I can see of the Staff"s sensitivrty analyses, it seems to ine that a proposal for a qualitative defirvbon like, loss ofsigni6 cant containment function due to system failure or bypass, accompanied by signi6 cant coCe damage, will be most serviceable.
The other two values 10E-4 for CDF and 10E-1 for CCFP are needed to put i
flesh on the general words about defense in depth. That is, even though the large release value is met, the Safety Goal Policy insists that plants will also meet the other two values. A plant cant get away with a great core but a lousy containment or vice versa.
This is essentially what I was proposing a year ago when you and Hal and I, along with
%an and Steve, were meeting in strange places to hammer out a SG position. But, it got lost in other debates. I propose it again.]
l also note that Remick would associate the 10E-4 probability with " core on the floor". I agree with that. Hal complained for years that we were not being expl' cit enough about what we meant by ' core damage". He was right. See the recent GE discussion of severe accidents in ABWR where they defined core damage as a PCT > 2200 F. It is a long way from this latter to core on the floor.
q 93o624o436 93o41o 3 31CNATED ORIGIEAL mu =X Y
SW#e""
Poa g