ML19344B477
| ML19344B477 | |
| Person / Time | |
|---|---|
| Issue date: | 07/09/1980 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Quittschreiber Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1268, NUDOCS 8010210161 | |
| Download: ML19344B477 (2) | |
Text
~
< e / -/Jll,2 r
03 1
V9 D 4 C4
%c,g,@gp July 9, 1980 MEMORANDUM FOR: 3.R.Quittschre)er
- 1. Catton 6h FROM:
RELIABILITY AND PROBABILISTIC ASSESSMENT
SUBJECT:
QUANTITATIVE G0ALS FOR SAFETY ASSESSMENT 1 JULY 1980 l
Probabilistic methods have the ability to impose order by organizing failure causes into comprehensive logical schemes that can be treated The problem is that the assessment of failure probability is difficult because frequently there are uncertainties in the data ba matically.
t where final results may be misleading.
to decompose a comples problem into simpler components and the data base is weak.
Some maintain that methods to point out important phenomena needing study.
is weak invalidates the use of the probabilistic approach.
i certain physical phenomena means that the development of techniques lts bability is unrealistic for a real LWR core melt accident if quantitative re are desired.
Certain aspects of a core melt accident lend themselv f
components of engineered safety systems can be done with a high d analys~is.
done confidence as it is not random. probability potential, separates the p The latter part could well and those requiring a great deal of speculation.The results would encourage us to be carried out using a worst case approach. understand the pny ity gets to the public.
I personally am uncomfortable when probabilistic m to ~
Insurance is more comforting.
ii effort'needs to be spent on increasing one's confidence if probabil st cm assure me that I am safe.
methods are to be acceptable to the lay publicSuch efforts,.i mentioned by Dr. Zebroski of EPRI.
ffD-l y-I. Ce%
/
8010210 161
- ',# 9 -
'.f It was noted by one of the ACRS consultants that NRC has the power to make a decision as to what is acceptable risk.
If this is the case, then they should do so.
I would favor hearings being initiated with rulemaking as the goal as soon as possible. Delay does not make a problem go away.
9 4
-. _ _..