Study on the method of collecting and processing initiating event data in probabilistic safety assessment

Probabilistic safety assessment (PSA) is a structured methodology for evaluating the safety of nuclear power plants, developed since the 1970s based on probability theory. The analysis of initiating events and their associated frequencies serves as the foundation of PSA and represents one of its principal technical components. An initiating event refers to an internal or external occurrence that disrupts the plant’s steady-state operation and results in abnormal conditions, such as transients or loss-of-coolant accidents. Such events necessitate responses from both the plant’s mitigation systems and operating personnel. Inadequate or failed responses may result in severe outcomes, including core damage. This study explores the procedures for initiating event (IE) definition, screening, and classification, as well as statistical criteria and parameter estimation approaches. Techniques for IE data acquisition and processing are examined and synthesized. Challenges related to data collection are identified and discussed, offering valuable insights for application in nuclear power plant safety assessments. Suggestions are provided to enhance future data collection methodologies.