A Fault Diagnosis Support System for Identifying Abnormal Operating Procedures in a Nuclear Power Plant

Abstract

In order to prevent safety hazards that can result from inappropriate decisions made by the operators of a nuclear power plant (NPP), this study was undertaken to develop a fault diagnosis support system to reduce the complexity of the decision-making process by aiding operators’ cognitive activities, integrating unusual symptoms, and identifying the most suitable abnormal operating procedure (AOP) for operators. The study was conducted from the perspective of human factors engineering in order to compare the process that operators originally used to diagnose potential and actual faults with a process that included a support system for diagnosing faults. First of all, based on interview, procedures studies and observations, the different kinds of symptom source related to the AOPs were discussed. Next, the unusual symptoms were collected and the rule of identification was constructed, which formed the fault diagnosis support system. Afterwards, an experiment was conducted to verify system effectiveness and reduction of workload by computing decision time, the number of errors and NASA TLX task load index. The results of the study indicated that the existence of a support system for fault diagnosis makes the task of fault diagnosis easier and reduces errors by quickly suggesting likely AOPs. With such a support system in place, there were clear improvements in human performance, i.e., decision-making time decreased by about 25%, and the accuracy of the operators’ decisions, judged by the successful resolution of specific problems, increased by about 18%. In addition, there were fewer erroneous solutions implemented, and the mental workload was reduced. Hence, it is recommended that the fault diagnosis support system be applied in identifying the AOPs in the main control room (MCR).