ML20044A736
| ML20044A736 | |
| Person / Time | |
|---|---|
| Issue date: | 06/14/1990 |
| From: | Carr K NRC COMMISSION (OCM) |
| To: | Babb A NATIONAL RESEARCH COUNCIL |
| Shared Package | |
| ML20044A737 | List: |
| References | |
| NUDOCS 9007020156 | |
| Download: ML20044A736 (3) | |
Text
-
M O-
' p aso
+
o UNITED STATES
!1
- Ig NUCLE AR REGULATORY COMMISSION n
h{
E WASHINGTON, D. C. 20566 -
June 14, 1990 CHAIRMAN Dr. Albert L. Babb, Chairman Comittee on Future Nuclear Power Development Comission on Engineering and Technical Systems National Research Council 2101 Constitution ~ Avenue, NW Washington, D.C.
20418-
Dear Dr. Babb:
I I am responding to your letter of May 11, 1990,-in which you stated that the.
Committee on Future Nuclear Power Development is considering making a quanti-tative stateme.it on the likelihood of a severe core melt accident in the next one to two decades and requested the Nuclear Regulatory Comission's (NRC's) view on whether estimates submitted by the Comission in 1985 to the House Subcommittee on Energy and the Environment provide a suitable basis for determining the probability of a severe core melt accident in U.S. reactors in the next 20 years. At the time that this estimate was made, the Commission recognized that-such simple estimates are subject to much uncertainty inherent in projecting core damage probabilities; these averages are driven.by plants that may have much higher core damage frequency than the majority and for this and other reasons, are subject to potential misuse.- Accordingly, considerable caution should be exercised in reliance on use of these averages.
There are-large uncertaintics associated.with quantitative estimates.of core damage frequency. These uncertainties result from lack of data to fully quantify the potential for multiple failures from common causes and, more importantly, the difficulty in quantifying human perfomance (errors that may aggravate conditions or actions to recover from a potentially serious con-dition). Further, there is no technical basis for. assuming that the severe core damage frequency of an individual plant will remain constant over a-prolonged period of time.
The dynamic processes associated.with1the aging of plant components, the normal replacement of plant equipment, redesigns responding to safety concerns and operating experience', changes in'organi-zational attitudes and operating procedures, turnover in operating personnel, and improved diagnostic aids made possible by technical innovations make such an assumption highly questionable.
t Perhaps of greater value than reliance _on-individual quantitative estimates of core damage frequency is the encouraging trend in the operation of nuclear power plants, which suggests that the actual collective risk is most likely decreasing. Since 1984, when we first began nuclear plant trend analysis, c
there has been a clear and significant improvement in the key operational safety parameters for operating reactors.
In particular, the number of I
N 9007020156 900614 l0
(
PDR COMMS NRCC s
~CORRES,PONDENCE PDC
\\
Dr. Albert L. Babb -
significant operating events h6s decreased, the unplanned automatic plant shutdown or scram rate has shown steady improvement, the number of safety system actuations has decreased, the average radiation exposure to plant 1
workers has decreased, and the number of significant precursor events (which 1
could lead to more serious conditions if additional failures were to occur) has declined. Continued attention to operational safety is being maintained to ensure that this trend is not reversed. A detailed examination of five operating U.S. nuclear power plants, presented in NUREG-1150, " Severe Accident Risks: An Assessment for Five U.S. Nuclear Power Plants", which has received extensive peer review and is scheduled for publication in final form later this year, includes findings that the NRC Safety Goals are shown to be met for all five plants studied, including the Quantitative Health Objectives for both fatality risk and latent cancer fatality risk.
In our view, the real value of integrated risk analyses is in their potential to identify modifications in plant design or operational priorities that would lead to a reduction in severe core damage frequency. Additionally, the performance of integrated risk analyses provides an opportunity to learn more about plant design strengths and weaknesses and how systems interact.
- Indeed, the experience to date indicates that virtually every probabilistic risk assessment (PRA) performed has led to some modifications in plant design or operational practices that would reduce the estimated severe core damage frequency. The Contrission has requested each licensee to perform an Individual Plant Evaluation (IPE), a probabilistically-based study, to search for any vulnerabilities that might cause a plant to depart significantly from the level of safety associated with an " average" plant. We anticipate that performance of the IPE studies will lead to modifications in plant design or operational practices to correct these vulnerabilities.
l We would emphasize that the PRAs that have been perfonned to date indicate that the results are highly plant specific, a finding that has been confirmed by the recently published second draf t of NUREG-1150, completed subsequent to the estimates to which you referred. This is so because of the high variability from plant to plant in the design of supporting systems (e.g., ac and de power, auxiliary cooling water systems, etc.). The results of one study, or even a group of studies, may not represent the industry as a whole and hence are of limited value in reaching industry-wide conclusions.
For these reasons, we have not attempted to formulate an " average" severe core i
damage frequency and likewise would strongly encourage your consnittee not to use any number based on assuming an average severe core damage frequency a
applied to the nuclear industry as a whole to estimate nuclear power plant risks. We would suggest instead that your committee consider stating that.
Dr. Albert L. Babb
-3 based u)on state-of-the-art probabilistic' risk assessment methodology, there is reasona'le evidence that the ensemble of operatir9 U.S. nuclear power plants a
meets the NRC Safety Goals and that there is reasonable assurance that the health and safety of the public are adequately protected.
Sincerely, Kenneth M. Carr
,