Probabilistic risk assessment practices in the USA for nuclear power plants
Introduction
It is the purpose of this paper to highlight activities associated with the probabilistic risk assessment (PRA) of nuclear power plants in the United States of America (USA). Topics to be covered include a historical summary of PRA related activities, the motivation for risk assessment, the language and methods of analysis employed, the form of the results, and public perception and acceptance of PRA. The paper concludes with brief comments on the impact of a generic standard and a few remarks about the response to a presentation of the paper at the European Commission-Joint Research Centre International Workshop on “Promotion of Technical Harmonization on Risk-Based Decision Making.”
Section snippets
History of nuclear power plant risk and safety assessment activities
Early practices (1950s and 1960s) in the USA of managing the risk and safety of commercial nuclear power plants were a carryover from the military uses of nuclear energy in the development of nuclear submarines and ships. The first commercial nuclear power plants were small in size (generally less than 100 mW electric) and to a large measure were extrapolations of submarine designs. Secondary containment systems also had their origin with the submarine program as a result of one of the
Purpose of nuclear plant risk assessments
As noted above, a quantitative, i.e. probabilistic risk assessment, was first performed in the Reactor Safety Study (WASH-1400) for the US Atomic Energy Commission starting in 1972. The Reactor Safety Study was published in 1975. This evaluation was not required by the regulations. While the details are not completely clear, it is generally believed that WASH-1400 was triggered by a letter from a United States Senator, Senator Pastore, to James Schlesinger, the then Chairman of the US Atomic
Nuclear plant risk assessment terminology and scope
In general, a standard appears to be emerging in the USA as to just what is meant by “risk.” That standard is the triplet definition of risk (Kaplan and Garrick, 1981). This definition says that in order to answer the question, “What is the risk?” it is necessary to answer three subsidiary questions: What can go wrong? How likely is it? What are the consequences? Many industries, segments of industry, government agencies, and academic institutions in the USA have adopted this definition.
One of
Risk assessment methods for nuclear power plants
Probabilistic risk assessments for commercial nuclear power plants in the USA are in various stages of development and maturity. Since the RSS in the mid-1970s, the data and models have evolved to a level of maturity that is seen in few other industries. The approach currently used by the commercial nuclear power plants generally involves five major steps. The steps are (1) develop and screen scenarios, (2) develop models, (3) estimate parameter ranges and uncertainties, (4) perform
Completeness of PRAs, regulatory authority, and existing standards
The scope of probabilistic risk assessments in the USA has been varied. In the USA, there is no existing regulatory requirement to update the PRAs performed in response to Generic Letter 88-20 for existing nuclear power plants. Most utilities have chosen to maintain some PRA capability where it can be shown that such capability will be cost beneficial, but there have been a few utilities that decreased their PRA staffs following the publication of the response to Generic Letter 88-20. A few of
Characterization of nuclear power plant risk and risk comparisons
The most common form for presenting risk results for commercial nuclear power plants in the USA is the core damage frequency. In most cases, this is done as a single point estimate. In some cases, a large early release frequency is also calculated, but the value is not widely used, understood, or referenced. As noted above, in some PRAs, a complementary cumulative distribution function is defined for public health effects. The CCDF is also known as the “risk curve” and was introduced to the
The impact of a generic standard
Standards that would enhance the decision-making process in risk management would be the most useful. Standards on the definition of terms, methodology, and data requirements are considered less important. Emphasis should be on such issues as health effects and environmental impacts that are scientifically based and permit comparisons across different industries and applications. Standard setting should take a systems approach and consider the costs, risks, and benefits of the full cycle of
Public perception of nuclear power plant risk
It has been the historical record of almost all new technologies that there is initially a period of distrust and fear as the new technology is introduced. For example, approximately 100 years ago, most states required that every automobile had to be preceded by a person carrying a flag to warn people that an automobile was approaching. However, within 10–20 years of the introduction of the automobile, such restrictions were no longer in place.
There is a high amount of distrust and fear among
Background
During the last 4 years there has been a number of initiatives in the USA to define and implement what has come to be called “risk-informed, performance-based regulations.” The overall purpose of these initiatives is to make the regulations of the US Nuclear Regulatory Commission more effective and efficient. By effective is meant that the regulations will address the most important contributors to risk. By efficient is meant that the regulations will address the most important contributors to
Other initiatives
The initiatives summarized above are just some of the activities that are underway in the USA to change the existing regulations regarding nuclear power plants. Other efforts that are related to risk-informed, performance-based regulation but not as directly tied to PRA as the initiatives above include:
Revised Reactor Oversight Program (RROP) [replaced the Systematic Assessment of Licensee Performance (SALP)]. This revised program has a framework that is tied to PRA, but at present the
Summary
In some sense, what has been occurring in the USA in the last few years can be considered the initial stages of a revolution in the licensing process of commercial nuclear power plants. Formal recognition has been made by the USNRC staff that the overall public risk and radiological consequences from reactor accidents is dominated by the more severe core damage accidents that involve containment failure or bypass. This fundamental concept, which was recognized and acknowledged by the PRA
Post-workshop impressions
Since this paper was originally presented at the EC-JRC International Workshop referenced earlier, a few remarks growing out of that presentation are appropriate.
The workshop was a meeting of international experts organized by the European Commission to discuss the merits of developing an internationally accepted generic “standard” for risk-based decision making. Nuclear power was one of nine general topics covered by the workshop.
References (23)
- Farmer, F. R. 1964 (November). The Growth of Reactor Safety Criteria in the United Kingdom, Anglo Spanish Nuclear Power...
- Federal Register. 2000 (12 January). Proposed Rules, Vol. 65, No. 8, pp. 1829 and...
- Fullwood, R.R., 2000. Probabilistic Safety Assessment in the Chemical and Nuclear Industries. Butterworth-Heinemann,...
- Garrick, B.J., Gekler, W.C., 1967 (May). Reliability Analysis of Nuclear Power Plant Protective Systems. US Atomic...
- Garrick, B.J., 1968. Unified systems safety analysis for nuclear power plants. PhD thesis, University of California,...
- Garrick, B.J., et al., 1979 (August). OPSA, Oyster Creek Probabilistic Safety Analysis. Draft. Prepared for Jersey...
- Kaplan, S., Garrick, B.J., 1981. On the quantitative definition of risk. Risk Analysis, 1...
- Kemeny, J.G., 1979 (October). Report of the President's Commission on the Accident at Three Mile...
- Pickard, Lowe and Garrick, Inc., Westinghouse Electric Corporation, and Fauske and Associates, Inc. 1981 (September)....
- Pickard, Lowe and Garrick, Inc., Westinghouse Electric Corporation, and Fauske and Associates, Inc. 1982 (March)....