ML19352A395
| ML19352A395 | |
| Person / Time | |
|---|---|
| Site: | Farley  |
| Issue date: | 12/20/1977 |
| From: | Mattson R Office of Nuclear Reactor Regulation |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML18025B195 | List:
|
| References | |
| FOIA-80-587 NUDOCS 8104170028 | |
| Download: ML19352A395 (2) | |
Text
.
r
'[
b'g UNITED STATES g4g y *.,,, 7 NUCLEAR REGULATORY COMMISSION
^
O:k
.. g WASHINGTON, D. C. 20555
's i
\\~
p 2 0 1977 DEC Note to: Harold Denton, Director, Div. of Site Safety & Environmental Analysis From:
Roger Mattson, Director, Division of Systems Safety
Subject:
SAFETY OBJECTIVES FOR ACCIDENT ANALYSIS
Reference:
Appendix C to SER Supplement on FNP: " Evaluation of Potential Accidents in the Vicinity of a Nuclear Power Plant" I believe there are several topics in the reference that are suitable for DSE-DSS discussion, especially in light of ATWS objectives. We should meet soon.
Some thoughts as preparatory for this discussion are:
1.
Relation of traffic accidents and industrial accidents to unacceptable offsite exposures seems indirect.
Since such events have the potential for structural damace, it seems that conventional criteria such as yield strength or missile penetration resistance, would be more direct.
2.
I do not believe we fully use the safety objective that "the risk to the public from all reactor accidents should be very small compared to other risks of life".
First, the ACRS quote is dated, and in context does not apply to present systems.
Also, the RSS study could be interpreted to show that comparatively the risk is not "very small", depending on what other societal risks you are considerino. Finally, and most important, our present review methods (ATWS excluded) do not start with any probabilistic safety objective. Rather it is to:
i) reduce probability of accident by high-quality design, construction, inspection, operation, maintenance, and, ii) reduce consequences by mitigating systems.
Does the RSS show that per plant there is 10-6 chance per 3.
year of exceeding part 100 auidelines? (In this context do we consider realistic releases from successfully - terminated
)
LOCA?)
4.
I do not see why (in your paragraph 4) there is consideration of the "16" plant accidents in terms of a probabilistic safety objective.
Our consideration of LOCA is, for example, deterministic and our review concludes'.0CAis(andmustbe)successfullyterminated.
- Further, i
the LOCA we review (i.e., one that meets 10 CFR 50.46 criteria) results in no core melt and little or no release of fission products.
81041706 %
~._
es 3
is 10~7 probability of an accident such as a transportation event If the 5.
and you believe it should be a design basis, how would I argue that pressure vessel failure (with perhaps comparable probability) should not be?
6.
How well can one demonstrate identification and correction of events with probabilities in the range of 10-7 per reactor year.
~
It is doubtful if the list of 16 on P, C-3 is complete.
For example, transients followed by several different potential failures may yield unacceptable consequences. Would each sequence have to meet 10-7 criterion or the combined sequences have to meet 10-7 criterion (i.e., if ten sequences, then each has to be 10-8 probable?).
I suggest the following participants for this meeting:
~s S. Hanauer R. Mattson (Ross, Novak, Thadani, Minners etc.)
H. Denton (Bunch etc.)
R. Boyd S. Levine V. Stello 03/1 c s, oger Matt on, Director Division of Systhms Safety cc:
S. Hanauer E. Case S. Levine V. Stello R. Boyd D. Ross D. Bunch T. Novak-W. Minners A. Thadani
.