HuREX – A framework of HRA data collection from simulators in nuclear power plants

Abstract The fundamental issue in reference to human reliability analysis (HRA) in a nuclear power plant is the lack of empirical data for human error probability (HEP) estimation, and lower level information of human performance that can be used to estimate HEPs. In an effort to resolve this issue, the Korea Atomic Energy Research Institute (KAERI) developed a framework, known as human reliability data extraction (HuREX), for data collection and analyses from a simulator to generate HRA data, such as HEPs, or for correlations between performance shaping factors (PSFs) and the associated HEPs. HuREX provides guidance on the identification of unsafe act (UA) and the processing of collected data. In addition, it allows us to analyze collected data based on the associated forms and taxonomy on generic task types and error modes. An application study was carried out using two sets of full-scope training simulator records to confirm the suitability of HuREX and to generate the HEPs of generic task types with respect to reference plants. As a result, 37 HEPs were successfully quantified for 21 generic task types.