ML20126G118
| ML20126G118 | |
| Person / Time | |
|---|---|
| Issue date: | 12/23/1992 |
| From: | Taylor J NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO) |
| To: | |
| References | |
| SECY-92-428, NUDOCS 9301040070 | |
| Download: ML20126G118 (5) | |
Text
i e
.....................oo.
j
?,
RELEASED TO THE PDR 3
/$pg\\
//v/93 4
a g
o
! kW L 1 dater iniuvs 8
- $,g-g.' f.'s annosooooooooeeeodoooooo
[
POLICY ISSUE 1.
December 23, 1992 SECY-92-428 d
(Information) i_
for:
The Commissioners From:
James M. Taylor i
Executive Director for Operations g
Subiect:
STAFF USE OF PROBABILISTIC RISK ANALYSIS
?
Purople:
To provide a second status report on work by the staff to 3
evaluate the manner in which it currently uses probabilistic si risk analysis (PRA).
Backaround:
In a July 1991 letter, the Advisory Committee on Reactor 1
Safeguards identified a number of problems with the staff's current uses of PRA.
In response, I formed the "PRA Working 5
Group" with the following objectives:
o To develop guidance on consistent and appropriate uses h
of PRA within the NRC;
'M o
To identify skills and experience necessary for the appropriate execution of each category of staff use; and O
o To identify improvements in PRA techniques and associated data necessary for each category of staff j
use.
9
_a A
In a Staff Requirements Memorandum dated January 24, 1992, i
the Commission requested a status report on the staff's i
eialuation of PRA uses. A first status report describing a
tht: plans of the Working Group was transmitted to the a
Commission in SECY-92-273, dated August 6, 1992.
This paper provides a second status report and summarizes the Working Group's progress to date.
j
=
Contact:
M. A. Cunningham, RES 492-3965 TO BE MADE PUBLICLY AVAILABLE IN 10 WORKING DAYS FRO.gHhl) ATE SECY NOTE:
C-
~
T-d osma gygy g(g 'y j
OF THIS PAPER.
1 I
The Commissioners Discussion:
The Working Group has completed two of its five tasks, the identification of present staff PRA uses and a assessment of limitations in the uses.
A key product from these two tasks is the definition of some important characteristics of and limitations in the uses of PRA by the staff, based on a survey made by the Group of staff users of PRA.
Some of the key results of the survey were:
Experienc_e and Trainino o
Almost all of the staff surveyed had taken one or more of the NRC training courses; however, Many of the staff surveyed had limited (less than one year) experience and familiarity with PRA techniques; and The staff surveyed also indicated limited experience and familiarity with PRA-related technical skills such as statistics.
o Most of the staff surveyed cited substantial experience with topics not directly related to, but useful to understanding PRAs, such as reactor systems, instrumentation and control, and reactor operations, o
A small percentage of the staff surveyed had a formal education in PRA-related topics (e.g., courses in statistics, reliability, etc.).
Guidance o
In many cases, there was limited or no formal guidance on how to perform or use PRA methods and results in a particular regulatory activity.
[It should be noted that, in parallel with the Working Group's activities, guidance is under development in some of these areas.]
o No staff member surveyed identified a use of formal decision analysis methods in their work.
Methods and Scope o
Almost all of the staff surveyed stated that they used best estimate calculations in the PRA. However, conservatisms in PRA models and assumptions were cited, mostly in the prioritization use category.
o The majority of PRA applications and studies were level 1 PRAs (i.e., with the product being a core
l The Commissioners damaae frequency or change in core damage frequency),
o The majority of these applications by the staff relied on adaptation of existing PRAs.
o No preference for any specific PRA method or PRA was reported.
Past PRAs such as the Reactor Safety Study, NUREG-1150, and industry-sponsored PRAs were equally cited.
Given these results, the Working Group is proceeding with its remaining three tasks:
developing guidance for PRA uses, assessing training and staffing needs, and assessing PRA methods development needs.
With respect to the first task, the Working Group's principal focus is now on two types of general guidance:
o Guidance on PRA uses Guidance is being developed by the Working Group on the scope, products, decision criteria, and quality assurance for two general types of staff PRA uses:
screening and prioritizing issues or events, and performing more detailed analyses of specific issues or events (e.g., operational events, generic safety issues).
These two types were chosen because they encompassed a number of staff PRA uses and there was essentially no formal guidance on what scope of PRA to perform in these uses.
(It should be noted that, in parallel with the Working Group's activities, present guidance on decision criteria used in regulatory analyses (the " Regulatory Analysis Guidelines") and generic safety issue prioritizations is being updated.
The Working Group will interact with the groups performing this work to maintain consistency in the guidance developed.]
o Guidance on PRA terms and methods Recognizing the limited experience of many staff users of.PRA (identified in the Group's survey), guidance is also being developed by the Working Group on basic terms and methods in technical areas important to appropriate uses of PRA by the staff.
This guidance provides definitions of terms used in PRA and related skills (with the goal of agency-wide adoption of these definitions), a description of methods commonly applied in the agency's business, including descriptions of the strengths and limitations of each, and a summary of references for. obtaining more detailed information.
In addition to this general guidance development, the Working Group plans to review and comment on other guidance being developed (e.g., the draft Regulatory Analysis-t
l t
The Commissioners Guidelines).
The Group also expects to make recommendations, where appropriate, for the development of more detailed guidance for specific staff PRA uses, which would be then reviewed by the Group.
The Working Group is still in the midst of its discussions with respect to the final two Working Group tasks (assessing training and staffing needs and PRA methods developments needs).
As such, the Group has no specific perspectives to provide at this time.
The Working Group met with the ACRS for a second time on October 9, 1992.
In that meeting, the Group described the results of its first two tasks and summarized the guidance it intended to develop, i.e., the results and guidance noted above.
The meeting was intended as a status report; as such, no written comments were requested or obtained from the Committee.
As described in SECY-92-273, external reviewers with expertise in risk analysis, statistics, decision theory, safety analysis, and NRC's regulatory process have been asked to review and comment on the Group's work.
The reviewers (Dr. Herbert J. C. Kouts, Defense Nuclear Facilities Safety Board; Dr. B. John Garrick, President, PLG, Inc.; Dr. Bernard Harris, Professor, Department of Statistics, University of Wisconsin; and Dr. Ralph Keeney, University of Southern California) were briefed by the Working Group on October 20, 1992. The briefing included discussions of the history of the Group, the results of the survey of present staff PRA uses, and the two types of guidance being developed.
The Working Group will continue to. periodically interact with the external reviewers and the ACRS.
By present-plans, the Group will provide a draft report to and meet neith the external reviewers in January 1993.
After resolution of their comments, the Group will transmit a revised draft report to the ACRS for comment; this is planned to occur in February 1993.
The Group's report is planned to be completed in April 1993.
After some experience with the guidance, revision of the guidance will be considered.
/
9
/
M. T y or E cutive irector for Operations cc:
SECY OGC
- - - -.. - -. -... - - -.. -. - - _ ~
k' i
i i
I s.
s DISTRIBUTION:
Commissioners OGC 3
5
{
CAA i
i 1
l j
1 i
e l
. -. - -.....,, -. _.. _.... -. _. -. _.