ML20043D758
| ML20043D758 | |
| Person / Time | |
|---|---|
| Site: | Millstone, Sequoyah, Seabrook, Surry  |
| Issue date: | 03/25/1990 |
| From: | Davis P Advisory Committee on Reactor Safeguards |
| To: | Kerr B Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1963, NUDOCS 9006110145 | |
| Download: ML20043D758 (2) | |
Text
_
i oh
'TO. Bill Kerr [M e
p From. Pete DaVI6.pW
%M' M'//9/
SUEUECT: Comments from Sub-Committee Meeg@ns uSREd4ql March 20 & 21,1990 ggty: utEx
Dear Bill,
b
%......... mv r. ' -
A few comments, from my perspective, ba$ed on information presented by the NRC staf f and contractors during the subject meeting.
The Staf f seems to have concluded that sevore accident research into phenomena which occurs during and af ter vessel breach is important. The primary objective of such research appea4 to be to gain Information which can be used to reduce uncertaintles in estimated risks. I have a few concerns with this conclusion, as follows.
A. I have examinu, essentially all PRA results currently available, and for those PRAS which estimate risks, none have shown risks in excess of the NRC Saf ety Goal even taking into account upper bounds on the uncertainty results. However, several PRAs are showing core damoge frecuency results in excess of the of ten used guideline (and used in early versions of the safety goal) of IE-4/yr. In f act, the average CDF for published PRAs a in the neighborhood of IE-4/yr. which suggests that a core damage event may occur with a probability of 20% over the next i
twenty years. It also seems that such an event would be devastating to the nuclear power Industry, and terd to destabilize our electrical power suppply. Thus. It has been my conviction that we should concentrate our limited resources on research to prevent core damage events, rather than attempt to better understand their progression af ter they occur. Current PRA results suggest that we have successfully designed plants to I
adequately mitigate severe accident consequences, even with current l
uncertaintles in severe accident progression. However, it seems ar(iuable, based on these same PRA results, that current designs have adequately provided protection against the occurrance of core damage accidents.
L B. For PWRs (representing about 70% the U.S. population), most PRAs estimate that off-site r1sks are domin3ted by interfacing LOCA accidents (V-Sequence). This is the case for Millstone-3, Seabrook, Surry, Sequoyah, etc. Thus, phenomena oc curing af ter vessel melt-through would be of little significance in reducing risk uncertainties since, for these plants, the t
PRAs Indicate that the most important radionculide release would have lJ already occurred prior to vessel'#each.
M l
DESIGNATED ORIGINAL, 61 g 5 9003 Cortified By_ -- //sh PDC CT-1963 b
L, y
..a gu.
Io, C. The staf f Indicated that they believe uncertiantles in the vessel'~ "',
- e;
. f ailure scenarlo, and subsequent events, are the major contriteors to risk uncertaintles based on PRA results. I find this questionable. I would
+'
expect that uncertainties in the external event risk contributors l' particularly seismic, would be more significant. Also, uncertainties in ks{ 'C human error and common cause modeling as well as consequence calculations are known to be significant contributors to uncertainty.
[f Thus, it is not at all clear that significant uncertainty reduction can be anticipated upon resolution of the severe accident issues being addressed by the NRC programs. Certainly for plants (see item B above) whose risk is dominated by the V-sequence, significant reduction in uncertainty would not be expected from this program.-
D. I rmain unconvinced that the BWR Mark i liner fallure is a significant severe accident issue. Studies show that the consequences of limer f ailure are not likely to be significant because of the tortuous path still remalning ior radionuclide transport to the environment.
Furthermore, it !s my belief that for most (if not all), severe accident secuences which lead to potential liner failure, most of the radionuclides would have already been released from the Superheated corluto upon vessel breach. Most of these radionuclides will be trapped in the suppression pool, or released from a f ailed or deliberately vented containment by the time the cooler corium reaches the liner area. Thus, the added source term from liner integrity loss may be insignificant.
Please call if you have any Questions on these comments.
cc: Dean Houston, ACRS
.a
?
~
i.
J
_..,..... _ _ _ _