A Study of Safety Importance Screening Criteria in the NPP Risk-Informed Safety Classification Process

Abstract Long-term practical operational experience and analysis of nuclear power plants have shown that component classification based on determinism can be optimized. Some safety-related components are not important from a risk perspective, while some non-safety-related components may have a large impact on the safety of the plant. With the development of probabilistic safety analysis (PSA) and the application of risk-informed technology, a risk-informed classification method combined deterministic and probabilistic evaluation has been developed. Risk-informed safety classifications (RISC) process can help nuclear power plants to focus their resources on high safety-significance component, thereby improving plant safety. By relaxing the regulatory requirements for component of low safety-significance, the operating cost can be reduced to a certain extent and the economic benefits of the power plant can be enhanced. The first step of RISC process is risk characterization of SSCs, using the sum of Fussell-Vesely (F-V) importance and Risk Achievement Worth (RAW) values as measurement parameters. F-V importance and RAW examine the impact of events on nuclear power plant risk from different perspectives and are commonly used risk importance parameters in risk quantification work. However, in Chinese engineering practice, Risk Spectrum (RS) software is often used to calculate the risk importance of basic events, and by creating component groups, it can directly give the risk importance of component. In this paper, we will analyze the calculation of the component risk importance provided in the RS software, using the PSA models of a generation II reactors and a generation III reactor as example, compare it with the measurement parameters provided by the NEI 00-04. In addition, by collating the risk importance distribution of nuclear power plant component, this paper makes some suggestions for screening metrics, which provide a reference for the subsequent pilot work of risk-informed component categorization in operating power plants.