ML19291C062
| ML19291C062 | |
| Person / Time | |
|---|---|
| Issue date: | 09/21/1979 |
| From: | Angelo J Office of Nuclear Reactor Regulation |
| To: | Hanauer S NRC - TMI-2 UNRESOLVED SAFETY ISSUES TASK FORCE |
| References | |
| REF-GTECI-A-17, REF-GTECI-SY, TASK-A-17, TASK-OR NUDOCS 8001110327 | |
| Download: ML19291C062 (4) | |
Text
-
'v
[
UNITED STATES 3N i1 NUCLEAR REGULATORY COMMISSION 3 ' m[I g
VVASHINGTON, D. C. 20555 N '[p/
SEP 2. 749
%..J Generic Task No: A-17 MEMORANDUM FOR: Stephen H. Hanauer, Director, Unresolved Safety Issues Program FROM:
John Angelo, Task Manager, Task A-17, Systems Interaction in in Nuclear Power Plants
SUBJECT:
SUMMARY
OF MEETING WITH SANDIA LABORATORIES ON SEPTEMBER 11-12 1979 TO DISCUSS PROGRESS OF CONTRACT ASSISTANCE ON TASK A-17 On September 11-12, 1979 members of the NRC staff met with representatives of Sandia Laboratories in Albuquerque, New Mexico to discuss progress of the work or. Task A-17, Systems Interaction in Nuclear Power Plants.
In particular, we discussed the analysis of the fault trees on the Reactor Coolant Pressure Boundary (RCPB) safety function using the set Equation Transfomation System (SETS) computer code. We also discussed the general structure of the fault trees on the Decay Heat Removal (DHR) safety function which we m completed during the first week of September 1979. The specific items of s1.nificance that were discussed are summarized in the following paragraphs of this summary report. A list of ;:ersons who attended the meeting is included as Enclosure 1 to this repo rt.
1.
Status of the Task At the time of this meeting (September 11-12,1979) the fault trees on the reactor coolant pressure boundary (RCPB) safety function, and the analysis of those fault trees using the SETS Code, have been completed and the results are reported in the Third Interim Report by Sandia Laboratories dated August 31, 1979. Also, at the time of the meeting the comparison of RCPB interaction potentials were compared to the Standard Review Plan and reported in the Third Interim Report.
In addition, the fault trees for the decay heat removal (DHR) safety function were completed and a limited number of copies were made available at this meeting for the purpose of discussing the general structure of the DHR safety function fault trees.
Sandia Laboratories expects to complete the fault trees on the reactor subtriticality (RS) safety function by mid-October 1979, and complete the analysis of the DHR fault trees by mid-November. Sandia Laboratories expects to complete the draft of the final report by the end of December 1979, but this may be a slightly optimistic estimate.
- 2. ' Analysis Techniques and Results The techniques used by Sandia Laboratories to perfom the analysis of the RCPB fault trees were discussed in considerable detail. These techniques are described in :he Third Interim Report by Sandia Laboratories dated A gust 31, 1979. These discussions augment the description of the analysis tecnniques that are presented in the report. We recommended that, for the final report, Sandia Laboratories should add further descriptive infomation to the text on analysis of fault trees, particularly with regard to the extent to which the 8001110.32 7
Stephen H. Hansuer SEP 211979 investigation was conducted for actuation power, both control and motive power, for pumps and valves. Also, further descriptive infomation woulo be useful to convey the extent to which location was investigated as an inter-active characteristic.
We discussed the fact that the analysis by the use of the SETS code is carried out for each of the three safety functions separately. The computer does not have the capacity to reduce all of he Boolian algebra for all three safety functions as input to an OR p;ce. What this means is that a particular basic cvent (or fault) may app.rar in more than one safety function cut set, but.will not be identified specifically as having occurred in more than one safety function analysis. However, we are confident that this particular technique will not obscure potential systems interactions.
We discussed the specific problem associated with basic events or faults that are much more likely to be the result of an operator error or lack of action on the part of the operator. The most comon example is the strictly manual valve which can be positioned only by the action of an operator from outside the control room.
In these cases the event will be discriminated by input of either "zero" or " unity" according to whether the event is assumed to occur with probability of one or probability of zero. The cut sets that are retained from the analysis are the sets with three or less independent events.
(If several basic events are linked by interaction, then that group of events is treated as one independent event).
It is important that the analysis report makes clear that the cut sets that are retained can be biased by the particular plant arrangement.
For example, consider a plant with three code safety valves and one or more power-operated relief valves on the pressurizer. Failure to relieve pressure could result if all of the valves fail.
If there is no interactive linking of these valves then the cut set would include four or more independent events.
(Each code safety valve failure would be treated as an independent event, and the relief valve failures would be independent events).
Since we limit the retained cut sets to three events, this particular cut set would drop out. However, if a olant has only two code safety valves and one relief valve the cut set would be retained. Therefore, it is important to make this distinction when generic analyses are to be used. Sandia Laboratories has stated that the report would contain enough details to make this matter clear to the users or reviewers.
We infomed Sandia Laboratories ttfat the techniques used and the method of presenting results appears to be acceptable and that they should plan on preparing the draft of the final report in the same manner as the interim report, except for our recomendation that additional descriptive text should be added to convey a better understanding of '.he extent of. the analysis of fault trees.
1734 098
Stephen H. Hanauer SEP 21 W9 3.
Review of the DHR Fault Trees Sandia Laboratories has completed the DHR fault trees and presented an overview of the general structure of the trees to assist us in our review of the details.
John Angelo, Task Manager Generic Task No. A-17
Enclosure:
As Stated cc:
J. Hickman/W. Cramond Division 4412 Sandia Laboratories P. O. Box 5800 Albuquerque, New Mexico 87185 1734 099
t ENCLOSURE ATTENDANCE LIST MEETING WITH SANDIA LABORATORIES ON SEPTEMBER 11-12, 1979 IN REGARD TO GENERIC TASK A-17, SYSTEMS INTERACTION IN NUCLEAR POWER PLANTS Sandia Laboratories Brookhaven National Laboratory W. R. Cramond A. Buslik G. J. Boyd R. A. Bari A. M. Kolaczkowski NRC
- 5. Asselin J. Zwolinski H. George D. McCloskey D. C. Fischer M. Srinivasan J. A. Murphy M. B. Aycock J. Angelo 1734 100 I
Distribution:
SEP 211973 Central File LWR #1 File k
D. Fischer M. Srinivasan J. Zwolinski S. Varga J. Angelo M. Chiramal H. George L. Soffer F. Rowsome M. Taylor J. Murphy' W. Vesely W. E. Morrison J. Norberg T. Scarbrough R. Major NRC PDR )
.