Modeling Safety-II based on unexpected reactor trips

Abstract

Safety-I is defined as a state where as few things as possible go wrong. Until now, the methods for analyzing the safety in nuclear power plants (NPPs), i.e., Probabilistic Safety Assessment and Deterministic Safety Analysis, have been developed from the perspective of Safety-I. However, focusing solely on Safety-I may miss opportunities to 1) learn from successes, and 2) observe how human adaptation contributes to successful outcomes, despite novel circumstances and resource limitations. In this light, a paradigm shift from ensuring that "as few things as possible go wrong (Safety-I)" to ensuring that "as many things as possible go right (Safety-II)" has been suggested. This study aimed to develop a model of Safety-II for unexpected situations in NPPs. Safety-II concentrates on the conditions in which as many things as possible go right. First, this study suggested and characterized a qualitative Safety-II model, which was modified from a resilience model developed by the Electricite de France (EDF). Second, event reports from unplanned reactor trips at Korean NPPs were analyzed based on the elements of this characterized Safety-II model as well as event severity. Third, a quantitative network model of Safety-II was developed by performing a correlation analysis. Finally, a feasibility analysis of Safety-I and Safety-II concepts for explaining event severity was performed. The result of this research suggests a new methodology for the safety assessment of unexpected reactor trips in NPPs, which could complement the conventional probabilistic safety assessments and deterministic safety analyses. (C) 2018 Elsevier Ltd. All rights reserved.