Human Reliability Analysis (HRA) methodologies have been developed to understand, manage, and mitigate the potential for human error, while addressing the inherent limitations of expert judgment and the discrepancies between actual operational conditions and analytical assumptions. The introduction of advanced digital technologies, such as fully digital panels and remote control, is being explored. Some HRA methods have proposed using simulators and actual Nuclear Power Plant data to evaluate potential improvements; however, there is no generally accepted criterion for determining the amount of HRA data required to support such improvements, creating uncertainty regarding the data volume needed to bound analytical uncertainty to an acceptable level.　Moreover, the simulators available for collecting such data are limited in both number and scope, resulting in constrained HRA datasets. Consequently, the limited availability of data and the associated statistical processing may introduce greater uncertainty compared to HRA results for existing nuclear power plants.
The objective of this research is to estimate the HRA data required to update HRA elements for fully digital control rooms. As an example of state-of-the-art HRA methods, this research focuses on IDHEAS-ECA, which was selected because its methodological foundations and supporting HRA data are explicitly documented.
This research first identifies and organizes the improvement items required to adapt IDHEAS-ECA to fully digital control room environments. This work examines the potential impact of improving a specific CFM and PIF on HEPs and their uncertainties by applying statistical analyses to the outputs generated by IDHEAS ECA, informed by commonly used data sources and statistical methods. This includes an examination of the IDHEAS ECA data sources and their corresponding evaluation elements. The outcomes of this analysis inform potential HRA enhancement options that could support more robust and context appropriate HEP estimation for fully digital control rooms.
Additionally, this research investigates the required quantity of HRA data to reduce uncertainty due to the adjustment of the HRA elements for automated environments. The outcomes of this analysis will inform the estimation of the minimum number of HRA data points required to support potential IDHEAS-ECA improvements in digital environments, while keeping the uncertainty of HEP within the range that is commonly accepted in HRA studies, such as existing HRA benchmark research. The findings of this study are expected to contribute to the development of a framework for HRA data gathering in fully digital environments and to be useful for other HRA methods currently under development.