A review of the fundamentals and methods of accident analysis using the best estimate plus uncertainty approach in nuclear power plants

Abstract

In 1989, the United States Nuclear Regulatory Commission (USNRC) revised the licensing rules for nuclear power plants (NPPs), allowing the conservative approach to be replaced by the Best Estimate Plus Uncertainty (BEPU) methodology. This replacement had two consequences: First, it became possible to use Best Estimate codes such as RELAP5 in licensing and design processes (previously, conservative codes were used). Second, methods were developed to study and quantify the uncertainty of these codes’ results and implement the BEPU approach according to the developer’s goal. The primary goal of the BEPU approach is to organize safety and thermal-hydraulics (TH) concerns for NPP licensing processes and frameworks. This approach is not uniform and various methods are used to implement it, but some concepts and basics are common. Some methods are pioneering and have initiatives within their structure, such as CSAU, GRS, and UMAE. Others improve the weaknesses of the pioneer methods (such as IMTHUA). This study aims to clarify the fundamentals, review the implementation methods, and assess the progress of studies based on the BEPU approach.