ML20052F417
| ML20052F417 | |
| Person / Time | |
|---|---|
| Issue date: | 05/24/1979 |
| From: | Arnold J Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-CT-1117A, NUDOCS 8205120440 | |
| Download: ML20052F417 (3) | |
Text
__ ____
67-MM PD& O k i
May 24, 1979 j
J. H. ARNOLD 9,
O
.p o a THE LER REPORTING SYSTEM g
y E
~
i ;g g u The LER Reporting System clearly has the potential to serve a number of important needs including (1) the collection and storage of information on safety significant events, safety component / system failure rates and safety significant human errors; (2) the Staff monitoring (by exception reporting, i.e., LERs) of the control of the safety features and safeguards of individual plants, within the operating bounds anticipated by the individual plant operating licenses; (3) prompt disemination to each nuclear plant operator, of infomation on safety problems being experienced industry wide; and (4) publication of infomation on all safety significant events of interest to the public.
One of the more apparent needs of the LER system is the correction of the substantial imbalance between the number of LERs and the resources available for optimum use of the information they generate.
It may well prove beneficial to strike a balance, reducing the number, improving the efficiency of handling, and increasing the resources.
Care should be taken in reducing the numbers, to preserve precent functions of the system, particularly the procurement of failure rate data and the monitoring of the control of the safe operation of the plants.
The information input to the LER system furnished by industry immediately following an event, is most beneficial as a form of instant alert fer both Staff and industry, to plant problems..The use of this information for l
expansion of failure cause and frequently data bases and other safety related assessments, should be guarded and important data points should be confinned following i inore complete analysis of the event. Also a general evaluation of the quality of " instant" information should be made and realistic improvements in quality should be sought. Best results will p,qbably be achieveable only with enthusiastic..sunoort.ef the,areoram by industry managers And supervisors.
l L
ORIGINAL h 5 g 40 790524 Ceitiriol By
[d_
CT-1117A PDR f
- p r
)
LER Reporting System - J. H. Arnold Page 2 One of the more elusive of the latent benefits of an LER reporting system is the ese of reports having little or no direct safety significance, as a source of teachings useful for the enhancement of public safety.
I believe it would be worthwhile to make a control search of a reasonable number of such LERs based upon limited objectives, for example a search for (1) inaccurate concepts which have lead to events that were not understood at the time and to disorganized human response, (2) u nusually high failure rates of common components in the particular duty compared to similar components in general, (3) unusual coupling of events, failures, and human errors, (.4) types of failures in areas of little safety significance, wnich could be serious had they occurred in areas of greater sensitivity, etc. The brevity of the infonnation available from most LERs now on file might defeat the purpose of such a search, but the making of the search might lead to beneficial additions to the reporting requirements for future LERs.
Investigation of Special Situations Among the needs for new technology crucial to the full realization of the potential of the LER system, are techniques for assessing by how much margin a particular event, missed developing into an event which could have brought harm to the public or generated great public concern.
Consideration should be given to approaches based on engineering judgments and on probabilistic analysis methods.
Either approach will undoubtedly require a more substantial data base than now exists, and the efforts of more than one group over an extended period of time, before results useful in any quantitative sense could be achieved.
However, the attempt produce quantitative results at this time will almost certainly lead to important new engineering insights. The application of probabilistic assessment technology will be particularly difficult because the import' ant question will not be close the miss would have been had the plant in question possessed the reliability of an average plant, but just where in the map of reliability did the plant in question lie before the event occurred. An even more.irnportant question is what would the margin have been had the plant in qu::stion been among the few of least reliability for the game has changed from the risk assessment of the average plant of WASH-1400 to the protection of every plant.
~,%-
- i 7
l j LER Reporting System - J. H. Arnold Page 3 The goals, organization, procedures and industry participation for the LER Reporting System appear to be worthy of more detailed evaluation. There may be merit in having the evaluation conducted by NRC and industry representatives not previously involved in the current system. Matters to be reexamined includes the broadening of the window of the system to plant events and operating data; the processing of the information in manners which lead to more prompt use for improving plant hardware and practices; the development of greater common understanding and support for the separate requirements and responsibilities of the NRC and industry in matters of plant reliability needs.
In this latter regard, the NRC is required to limit its actions to matters of public safety. The utility on the other had is required to manage an extremely large number of reliability needs, one of which is public safety.
l Many of the other reliability needs of industry are inextractably intertwined with reliability for public safety. Reliability for public safety, is of course, a paramount concern of both the NRC Staff and industry.
Reliability for the production of power economically and dependably as well as safety, and ir.co reliability for avoidance of nuclear events which result in extended plant outages and high repair costs, are also a paramount concern of industry. The need for comnon understanding and cooperation between NRC and industry in the simultaneous achievement of reliability for public safety, power production and. plant safety is pbvious.,
4
-6 D
4 i