ML20245A442
| ML20245A442 | |
| Person / Time | |
|---|---|
| Issue date: | 03/30/1987 |
| From: | Congel F Office of Nuclear Reactor Regulation |
| To: | Minners W Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20236K304 | List:
|
| References | |
| FOIA-87-380, TASK-082, TASK-82, TASK-OR NUDOCS 8704010581 | |
| Download: ML20245A442 (5) | |
Text
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _. _
[f A
'o UNITED STATES
,' g NUCLEAR REGULATORY COMMISSION e,.
g.
a W ASHING TON, D. C. 2M55 i
',E s
/
3 0 VAR BM I
MEMORANDUM FOR: Warren Minners, Chief Reactor Safety Issues Branch Division of Safety Review and Oversight 1
FROM:
Frank J. Congel, Chief Reliability and Risk Assessment Branch Division of Safety Review and Oversight
SUBJECT:
SUPPLEMENTAL COMMENT, GI-B2, BNL DRAFT REPORT "BEYOND DESIGN BASIS ACCIDENTS IN SPENT FUEL STORAGE POOLS" We provided our review coments on the subject draft report by a memorandum dated March 20, 1987. Since then, based on discussion we had at meetings of March 24 and 25, we have developed supplemental coments to further elaborate on our previous remarks on the treatment of seismice.11y induced spent fuel pool accidents.
These comments are included in the enclosure.
Please contact N. Chokshi (x28347) of my staff if you need additional infomation or clarification.
Frank J. Co gel, Chief Reliability and Risk Assessment Branch Division of Safety Review and Oversight Encisoure:
As stated cc:
T. Speis B. Sheron I
H. VanderMolen E. Throm
{
L. Reiter
(\\ (\\
7.).
A. El-Bassioni S. Acharya Pf
. (3 L7
)..
,(
\\
t
ENCLOSURE t
SUPPLEMENTAL COMMENTS ON THE BNL DRAFT REPORT "BEYOND DESIGN BASIS ACCIDENTS IN SPENT FUEL STORAGE POOLS" The probabilities for beyond design basis accidents in the spent fuel pool due to complete loss of water inventory as a result of a seismic event are estimated to be 2.3 x 10 6 for Millstone 1 and 1.6 x 10 5 for the Ginna plant in the draft BNL report.
The loss of inventory results from the seismically induced i
structural failure of the pool structure.
In order to estimate these probabilities, a number of assumptions were made regarding the seismic hazard estimates, and structural fragilities.
Also, a weighting scheme was used to estimate the likelihood of various hazard and fragility curves. While the adequacy of these assumptions and their effects on the estimated probabilities are discussed later, the estimate of frequency of the spent fuel pool accident in the range of 2 x 10 5/RY raises the following immediate question.
The seismic event, more than any other event, will affect the entire plant and not only fuel pool structure.
In past PRAs, seismically induced core damage frequencies were generally estimated in the range 10 ' to 10 8/RY.
These frequencies are comparable to frequencies of spent fuel pool failure estimated in this study.
As discussed later, the seismic capacities of the spent fuel pool structures (which are seismic Category I Structures) are expected to be higher for most plants than the components which would govern core damage
j sequences.
Therefore, examining spent fuel pool accidents alone may lead to a distorted view of their importance.
This is not to say that tne seismic initiators should be disregarded for the spent fuel pool accident. but if the outcome is controlled by major structural catastrophes (at earthquake levels significantly higher than the design basis levels) rather than failures associated with low level earthquakes, the risk from fuel pool failure must be examined in the context of the possibility of the concurrent core melt accidents.
The seismic hazard estimates used in the study are primarily based on the results of a study conducted by the Lawrence Livermore National Laboratory (LLNL).
These results for the Millstone site are typical of the high range of hazard for that site. Alternate estimates by the utility and EPRI would be approximately an order of magnitude less.
The assumption of an upper bound cut-off to acceleration in the hazard curve needs further examination and sensitivity studies. The fragility estimates are based on the fragilities of the Oyster Creek 8eactor Building and a shear wall in the Zion auxiliary building.
The median capacities for the two structures are 0.75g and 1.lg, respectively.
For a number of reasons, these estimates are judged to be conservative by themselver and more so as a representation of a typical shear wall of the pool structure.
For example, data from PRAs of ten plants indicate that the median capacities of shear walls in seismic Category I structures (other than containment) range from 0.9g to 8.2g.
The expert panel on the seismic margin program recommends that up to the review level earthquake of 0.5g, because of their high capacities, the shear walls can be screened out from consideration in order to estimate the high confidence of low probability
l l
)
of failure level for the plant.
For the most plants, the pool structures are 1
located at lower elevations near grade level'and also have heavier walls than stegory I structures such as auxiliary or diesel generator buildings.
Therefore, it is 4ery likely that a specific analysis will improve these fragility estimates.
In any event, the fuel pool walls will experience loads such as hydrodynamic loads, impact loads' from the up-lif t of the racks, and algnificant non-linear interaction between racks and structures.
Tnese pieenomenon and loads, in general, are absent for shear walls of other category I structures and extrapolation of fragilities of these walls to the fuel pool structure should be checked.
Failure modes for the pool structures may be governed by items such as excessive strains in the pool liner which might i
I indure tears and the excessive and sufficient cracking of walls such that inventory depletes rapidly.
Because of the assumed high hazard and high fragilities (low capacities), the estimate of 2 x 10.s/yr is judged to be a conservative estimate.
=
While the study has focused on the loss of inventory because of the structural failure, the effects.of other seismic induced interactions which may not
.necessarily lead to the loss of inventory are not discussed.
Specific examples of these interactions would be the coliapse of block wall surrounding the fuel i
pool or failure of the overhead crane.
These interactions are likely tu be very plant-specific.
Such phenomena as severe distortion of racks or esccatites should also be addressed.
As discussed earlier, a number of sensitivity studies are required to examine effects of various assumptions made in this study.
Until these studies are completed and some improvements are made to be more p? ant specific, results of
_ _ _ ~ - _ _ _ _. _ _ _ _ _ _ _ _
.3 this study must be viened as preliminary.
Particularly, the frequency estimates based on the seismically induced f ailure of the pool structure are judged to be conservative an: neither generic nor plant-specific.
These estimates indicate the need for further refinement of the analysis.
C
- _ - - _ _ _ _ _ - -..