ML19257B040
| ML19257B040 | |
| Person / Time | |
|---|---|
| Issue date: | 01/24/1979 |
| From: | Baranowsky P NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Edison G NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| REF-GTECI-A-30, REF-GTECI-EL, TASK-A-30, TASK-OR NUDOCS 8001150139 | |
| Download: ML19257B040 (5) | |
Text
e be
,8 UNITED STATES s
g' NUCLEAR REGULATOF:Y COMMisslON b*b WASHINGTON, D. C. 20555
/
M8D 1., k '
f g,
v January 24, 1979 NOTE T0: Gordon E. Edison, Head Systems Engineering Section, PAS FROM:
Patrick W. Baranowsky Systems Engineering Section, PAS
SUBJECT:
SUMMARY
OF MEETING WITH EVALUATION ASSOCIATES ON DC POWER SYSTEMS STUDY A meeting was held with members of the NRC, Evaluation Associates (EA) and Sandia Laboratories in Bethesda, Maryland, on January 16, 1979, to discuss the progress of the DC Power System Study.
The list of meeting participants and meeting agenda is attached.
EA provided a handout with revised estimates of probabilities associated with the loss of offsite power (LOP) events, a minimum DC power system diagram, and PWR and BWR fault trees for core meltdown due to DC power systems failure.
These items were discussed in detail. A summary of the major points of the meeting is attached. The next meeting is scheduled for February 20, 1979.
/N $fkI.>
- . xt -
Patrick W.~ T3aranowsky e Systems Engineering Section, PAS Attachments:
As Stated cc:
A. Buhl P. ficGrath J. Pittman S.HanaueX y
F. Rosa G. Lainas E. Wenzinger Meeting Attendees 1750 092 8001150 /3 9
e Summary of Meeting with Evaluation Associates on DC Power _ Study - 1/16/79.
1.
EA has begun to review the detailed LLR's on LOP provided by a 00R survey.
2.
The probability of a LOP following plant trip will be reassessed usihy u.: detailed LER's and compared with the current assessment of 0.0027.
3.
EA estimates the probability of occurrence of a single DC power failure which would lead to reactor trip at 2.4x10-4 per reactor year by fault tree analysis.
Previously reviewed data did not include a DC power failure leading to reactor trip.
EA will search LER's on LOP for this type of failure.
For zero failures leading to trip a statistical upper bound will be estimated for comparison with the synthesized fault tree evaluation.
4.
As agreed earlier with NRR. Power Systems Branch, all DC power analyses are being performed on the basis of the minimum acceptable system defined by NRR (i.e., the Hadam Neck Plant).
For analyses which require decay heat removal system evaluations, the minimum DC system will be coupled with the Surry Unit 1 plant for PWR analyses and with the Peach Bottom Unit 2 plant for BWR analyses.
EA will review the applicability of the minimum DC power system to these plants.
5.
Common cause failures of the DC system and their probability will be evaluated.
EA will synthesize a CCF rate by reviewing failure data for patential common causes and by a review of more detailed design information, e.g., system layout drawings, testing, maintenance, and tech specs.
The NRC will assist in obtaining detailed design information as specified by EA.
In addition, EA will look at bounding the CCF probability for comparison with synthesized values.
6.
EA and NRC will review the AFWS CCF in the RSS to check the AC power loss contribution and possible double accounting in the DC power system failure fault trees.
It was noted that most diesil generators are load sequenced and that trip due to no load sequencing was considered a major CCF contributor in the RSS.
7 EA will include in its draft report a discussion of error bounds and confidence limits.
1750 093
8.
The reactor coolant system pump seal failure problem appears to have low impact on the outcome of the DC power failure analysis.
This is due to the relatively low leat rate anticipated from scal failure and the fact that the probability of the initiating event (loss of all DC and AC) is apparently much lower than the probability of an S2 LOCA.
The need for further work on this matter will be determined later in the program.
9.
The next meeting is scheduled for February 20, 1979. A draft report will be provided by EA prior to the necting for discussion at the meeting.
In addition to the draf t report, specific discussion of the P(LOP / trip) review, DC power dependencies in decay heat removal systems, and common cause failure probability will be included in the next meeting.
10.
The DC power systems study by EA was originally scheduled for completion by Febraury 15, 1979.
Dif ficulties with' the collection and evaluation of failure data and the common cause failure assessment are impacting the schedule o.f this program.
It is estimat.'d by Sandia that funding for this program ($30K) will be exhausted F
.he end of February or early fiarch of this year. flRC will evaluate the need for added funding, availability of funds, and impact of an extended schedule.
1750 094
DC Power Systems Study Meeting Attendees - January 16, 1979 f1RC G. E. Edison P. W. Baranowsky R. Fitzpatrick J. Olshinski Ejl M. Fedele M. Wilson Sandia J. Hickman 1750 095
AGEt;DA FOR DC POWER ftEETitiG JANUARY 16, 1979 8:30 A.M.
PAS C0f!FERENCE ROOM, MNDB 1.
DC Dependencies in Decay Heat Removal Systems 2.
Portion of DC (Single Division) Failures Resulting from Human Error 3.
Common Mode Failure Frequency 4.
Fault Tree Analysis PWR Fault Tree BWR Fault Tree CCF Consideration RCP Seal Failure Treatment 5.
Data Discussion Component Failure Definition P(LOP / Trip) Data LERs indicating Trip Data Discrepancies 6.
Minimum DC Power System & Decay Heat Removal System (s)
PWR System Description BWR System Description 7.
Direction & Schedule of Future Ef forts 1750 096
_