ML20245A438
| ML20245A438 | |
| Person / Time | |
|---|---|
| Issue date: | 03/27/1987 |
| From: | Miraglia F Office of Nuclear Reactor Regulation |
| To: | Speis T Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20236K304 | List:
|
| References | |
| FOIA-87-380, RTR-NUREG-0612, RTR-NUREG-612, TASK-082, TASK-82, TASK-A-36, TASK-OR GL-85-11, TAC-M53347, NUDOCS 8704010234 | |
| Download: ML20245A438 (5) | |
Text
n-_-_-______
\\
,hS 58 co
- q'e, UNITED STATES
[
[}
NUCLEAR REGULATORY COMMISSION 1
t WASHINGTO N. D. C, 20555 8
March 27,1987
% ='... /
"-":T :" FOP :
Themis P. Speis, Director i
Division of Safety Review and Oversight l
FROM:
Frank J. Miraglia. Director Division of PWR Licensing-B
SUBJECT:
GENERIC ISSUE 82, "BEYOND DESIGN BASIS ACCIDENTS IN SPENT j
FUEL STORAGE POOLS" (TAC M53347)
\\
In response to your request, the Engineering, Peactor Systems and Plant, Elec-trical, Instrumentation and Control Systems Branches of the Division of PWF.
Licensing-B have reviewed the subject report.
PEICSB Eranch comments are as follows:
]
1.
The report on beyond design basis spent fuel pool accidents does not clearly state the criteria or basis for recommendations to resolve the 1
issue. The report's only recommendation (Section 6) states that plant specific evaluations should be performed, but no criteria for determining resolution are established.
Further, the report summary does not include this recommendation. Additional discussion of these issues should be provided.
2.
The beyond design basis seismic event considered in this report would affect the entire plant.
If this event is of concern, its effect on the core mitigating systems, non-safety systems, systens interaction ard personnel should be evaluated as well as its effect on the spent fuel pool.
3.
The report does not consider the improvements implemented in all plants as a result of the resolution of Task A-36 which is reported in NUREG-0612. " Control of Heavy Loads at Nuclear Power Plants," July 1980.
Further, the staff determined on.the basis of a cost / benefit analysis as documented in Generic Letter 85-11 that the risk of unacceptable offsite radiological releases due to heavy loads handling concerns has been reduced to low enough levels by actions already taken in this area. The I
report does not document a refutation of this conclusion.
4.
The considerations given to risk reduction measures are superficial compared to the analysis requirements of paragraph (c) of 10 CFR 50.109, "Backfitting."
For example:
The proposal to limit the quantity of stored spent fuel and its asso-ciated radioactivity is not established as being either significant l
or insignificant to reducing the risk from beyond design basis spent
[fd bd.
l fuel pool accid _ents.
~i CONTACT:
L]b'. Lf R. Ferguson, PEICSB/NRR 49-27065
t 2
The characteristics of new " acceptable" storage configurations for discharged fuel including their seismic design requirements and cost effectiveness are not established. The basis for the' risk reduction factor of 5 to 10 assigned to such proposed configurations is not apparent.
The analysis procr. dure and criteria for the recommended plant specific evaluations to be performed are not established.
The characterir, tics of proposed post-accident spray systems over the spent fuel pool as a means of reducine spent fuel poul accident risks are not established.
The Reactor Systems Franch had no comments. The Engineering Branch comments are in Enclosure 1.
l 4ynjoyH F. J. Miraglia, Director Division of PWR Licensing-B
Enclosure:
({
As stated cc:
T. Nnvak R. Bernero C.E. Rossi G. Lainas W. Minners P..
Bosnak E. Throm C. Thomas L. Marsh m_m___.m__ _ _ _ _ _ _ _ _ - _... _.
~
ENCLOSURE 1 l
COMMENTS ON THE GENERIC ISSUE 82 "BEYOND DESIGN BASIS ACCIDENTS IN SPENT FUEL STORAGE POOLS
1.0 INTRODUCTION
1 i
i The Engineering Branch has reviewed Section 2 of the generic issue entitled "Beyond Design Basis Accidents in Spent Fuel Storage Pools" as requested.
In the document two plants Millstone 1, a BWR, and G1nna, a PWR, were selected for probabilistic analyses.
As a result i
of these analyses, Chapter 2 concludes that the failure of spent fuel pools due to shipping cask drop resulting from human error and earthquakes are the two dominant events among all the possible acci-dents studied.
The conclusion was based on the fact that the calcu-lated probabilities for these two events were higher than that of other events.
For example, the probability of the pool structural failure from cask drop was reported being 3.1 x 10 5/RY for both Millstone I and Ginna, from seismic being 2.2 x 10.s/RY for Millstone 1 and 1.6 x 10 5/RY for Ginna, from airplane crash being 1 x 10 20/RY for both Millstone I and Ginna, and from tornado missiles being 1 x 10.s/RY for both Millstone I and Ginna.
Since the conclu-sion was based solely on the comparisons of calculated probabilities, the validity of the conclusion depends on the degree of adequacy of these calculations.
(1 The text was fairly clear on how the probabilities were derived, as well as the assumptions used during the process of derivations. How-ever, justifications were not provided for some assumptions, and other assumptions of fundamental importance appear to be unsubstantiated.
These will be listed and discussed below.
- 2. 0 JUSTIFICATIONS THAT NEED TO BE PROVIDED Justifications for the assigned weighting factors for the seismic a.
i.azard curves and fragility curves on Table 2.5 were not provided and need to be provided, b.
It is stated on page 2-17 that a conditional probability of 10%
has been arbitrarily selected for the hazard calculations. Justi-fications were not provided for the selection of 10% and need to be provided.
The distribution of failure frequency of human errors in heavy c.
crane operations, as listed in Table 2.7, was stated as being estimated by the BNL staff and not supported by actual data.
There was no discussion on the significance of the shape of this distribution curve to the structural failure probability of 3.1 x 10 5/RY of the spent fuel pools.
Justifications are needed unless it can be shown that the influence of this distribution shape is insignificant to the final value of probabilities.
~,
. d.
The assumption.for each of the ten items listed in Table 2.8 was reported as being used in the hazard calculations.
The assumptions for three items were borrowed from a reference book.
They are the probabilities of having mechanical failure of a crane being 3 x 10 5/ operating hour, electrical failure of a crane being 3.x 10 8/ operating hour, and human errors being 6 x 10 4/11ft.
On the subject of human errors, the text states on page 2-16 that the applicability of this specific value from the reference to crane operations is not obvious.
The only reason for using the value of 6 x 10 4/11ft was that the value is consistent with the data listed in the NRC handbook on human reliability analysis.
This value was responsible for the final conclusion that structural fcilure of the pool due to the shipping cask drop resulting from human error being one of the two dominant events in the hazard analyses. A better justification should be given to the selection of this important value of 6 x 10 4/11ft.
If a better justification cannot be given due to the lack of data, the author, BNL, should make its positions clearer as to whether it agrees with the number listed in NRC handbook or whether it has expertise to make such a judgment so that the reader can have a better understanding on the relia-bility of the number and its subsequent conclusion.
No justifica-tion or reference was given to the other seven assumptions listed in Table 2.8 and justification or references should be provided.
e.
It is stated on page 2-18 that the probability of structural failure due to cask drop on pool edge caused by human error is 3.1 x 10 5/RY.
(}f No discussion on how this value was calculated.
Justifications or derivations are needed.
3.0 ASSUMPTIONS THAT APPEAR TO BE UNSUBSTANTIATED The text states on page 2-9 that fragility curves specifically for a.
spent fuel pools have never been developed and, therefore, the fragility curves developed for the Oyster Creek reactor building and Zion auxiliary building were substituted for the spent fuel pools at Millstone 1 and Ginna.
Only the fragility curves for the Oyster Creek reactor building were attached to the text and are reproduced in Enclosure 1.
The fragility curves shown in Enclosure 1 define the failure frequencies of the Oyster Creek reactor building versus peak ground accelerations during an earthquake.
The fragility curve so defined is structure dependent. Two different structures located at the same place and subjected to the same peak ground acceleration may have different fragility values.
This is because the two dif-ferent structures possess different strength and ductility and, therefore, will fail at different levels of peak ground accelerations.
The spent fuel pool structure is different in shape from a reactor or auxiliary building even though it may be located in a reactor or auxiliary building.
Thus, the fragility curves should all be different for these different types of structures.
Even among the
w s-spent fuel pool str uctui es the fragility curves may also be dif-ferent for the pool structures designed for different levels of earthquake motions.
In the absence of proper justification, sub-stituting the fragility curves developed for the Oyster Creek reactor building and Zion auxiliary building for the spent fuel pool structures at Millstone 1 and Ginna is questf.onable at best.
1 b.
The text did not address the validity of the fragility curves of I
the Oyster Creek reactor building and Zion auxiliary building which had been used as the basis of this analysis.
Since the fragility curves have significant impact on the final probabilistic values of pool structural failure due to earthquakes, their validity must be examined before use.
The fragility curves in Enclosure 1 appear to be questionable with respect to actual structural behaviors for two reasons.
4.0 CONCLUSION
The text was written clearly; however, some assumptions of fundamental importance appear to be incorrect and some other assumptions need to be justified.
O w
_-