Dynamic scenario quantification for level 2 PRA of sodium-cooled fast reactor based on continuous Markov chain and Monte Carlo method coupled with meta-model of thermal–hydraulic analysis

Abstract

ABSTRACT In probabilistic risk assessment (PRA), an event tree (ET) methodology is widely used to quantify accident scenarios which result in core damage and fission products release. However, the current approach using the ET methodology is not applicable to evaluate dynamic characteristics of accident progression, when the accident progression is time-dependent and headings in the ET have inter-dependency between events. Thus, a dynamic approach of accident scenario quantification is necessary to evaluate more realistic PRA. This research addressed this need by developing a dynamic scenario quantification method for the level 2 PRA by coupling of Continuous Markov chain and Monte Carlo (CMMC) method and a plant thermal–hydraulic analysis code for a sodium-cooled fast reactor (SFR). The CMMC method is applied to protected loss of heat sink (PLOHS) accident of the SFR to analyze dynamic scenario quantifications. The coupling method requires heavy computational cost and it makes difficult to quantify the whole accident scenarios by comparing the results from existing plant state analysis codes. Thus, a meta-analysis coupling method is proposed to obtain dynamic scenario quantifications with reasonable computational cost. Also, a categorizing method is used to depict analytical results in a transparent manner.