ML20126H642
| ML20126H642 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 06/12/1985 |
| From: | Fiedler P GENERAL PUBLIC UTILITIES CORP. |
| To: | Zwolinski J Office of Nuclear Reactor Regulation |
| References | |
| GL-83-28, NUDOCS 8506180408 | |
| Download: ML20126H642 (2) | |
Text
GPU Nuclear Corporation NUOIME 100 Interpace Parkway Parsippany. New Jersey o7054-1149 (201)263-6500 TELEX 136-482 Writer's Direct Dial Number:
June 12, 1985 Mr. John A. Zwolinski, Chief Operating Reactors Branch No. 5 Division of Licensing U.S. Nuclear Regulatory Commission Washington, D. C.
20555
Dear Mr. Zwolinski:
Subject:
Oyster Creek Nuclear Generating Station Docket No. 50-219 Generic Letter 83-28 Reference (1) NRC Letter dated April 5, 1985, John A. Zwolinski to P.
B. Fiedler, " Request For Additional Information Follcwing Preliminary Staff Review of Licensee Responses to Generic Letter 83-28" (2) NEDC-30844 dated January 1985, "BWR Owners' Group Response to NRC Generic Letter 83-28, Item 4.5.3" (3) G. Apostolakis et al, " Assessment of the Frequency of Failure to Scram in Light-Water Reactors", Nuclear Safety 20(6): 690-705 (Nov - Dec 1979)
The following information is provided in response to your April 5, 1985 letter concerning Item 4.5.3 of Generic Letter 83-28. General Electric Company, on behalf of the BWR Owners Group has performed an analysis which addresses the reliability of the Reactor Protection System for Boiling Water Reactors (NEDC-30844). This analysis has evaluated the availability of the RPS for the existing intervals for on-line functional testing taking into consideration the uncertainties due to component failure rates, common cause failures, operator errors, component "wearout" and reduced redundancy due to testing.
The results of the analysis, performed using a generic availability model representative of the General Electric BWRs, show that the RPS failure frequency is low and that the existing on-line functional testing for the RPS supports this. Additional analyses performed by GE address the differences between specific plants and the generic model. The significant differences between the Oyster Creek design and the generic design analyzed by GE are the type of relays used in the sensor logic channels and in the redundancy of scram contactors.
/
B506180408 850612 PDR ADOCK 05000219 l
1 0 P
PDR s
GPU Nuclear Corporation is a subsidiary of General Pubhc Utihties Corporahon
1 Incorporating these differences in the analysis, the results s RPSfailurefrequencyisstilllowandapproximately1.5x10gowthatthe per year.
This result is comparable to results of other studies referenced in this analysis which shoy point estima rangeof1.9x10-0to9.0x10gesoftheRPSfailurefrequencyinthe per year.
Additionally the analysis (NEDC-30844) addresses two other areas to which the RPS failure frequency is sensitive; these are common cause failure and human error. At the extreme, the upper bound error factors for common cause failure and human error produce estimates of RPS f OysterCreektypeplantofapproximately6.1x10gilurefrequencgforthe and 1.4 x 10- per year respectively. These values are at the upper bound of the studies referenced in the analysis.
Whentheupperboundofbothcommoncauseandhumanerroraregakentogether and translated to " failure en demand", the result is 1.8 x 10. This is comparabletotheupperboundoftheRPSfajlurefrequencydistribution provided in Reference 3, which is 1.2 x 10- per demand.
Based on the foregoing discussion, GPUN endorses NEDC-30844 to tne extent that it substantiates that the Oyster Creek Plant has a low RPS failure frequency and that the existing on-line functional testing supports this when the differences and sensitivities cited above are taken into consideration.
Very truly yours, Vice President and Director Oyster Creek 1r/0487t cc: Administrator Region I U.S. Nuclear Regulatory Commission 631 Park Avenue King of Prussia, Pa.
19406 NRC Resident Inspector Oyster Creek Nuclear Generating Station Forked River, N. J.
08731