Speaker
Description
Over the years, the nuclear fleet has relied on labor-intensive, time-consuming preventive maintenance (PM) programs, driving up operation and maintenance (O&M) costs to achieve high capacity factors. The primary objective of the research presented in this paper is to develop scalable technologies deployable across plant assets and the nuclear fleet in order to achieve risk-informed predictive maintenance (PdM) strategies at commercial nuclear power plants (NPPs). A well-constructed risk-informed PdM approach for an identified plant asset was developed in this research, taking advantage of advancements in data analytics, machine learning (ML), artificial intelligence (AI), risk model, and visualization. These technologies would allow commercial NPPs to reliably transition from current labor-intensive PM programs to a technology-driven PdM program, eliminating unnecessary O&M costs.
The research and development approach (see Figure 1) presented in the paper is being developed as part of a collaborative research effort between Idaho National Laboratory and Public Service Enterprise Group (PSEG) Nuclear LLC. This paper presents the results of analyzing the heterogeneous data associated with the circulating water system (CWS) from both the Salem and Hope Creek NPP sites. Fault modes present in the data were identified based on logs and correlated with data to develop salient fault signatures associated with each fault mode. The fault signatures are used to develop diagnostic models using scalable predictive analytics and integrated with plant-level risk and economic models. The paper also outlines the development of a user-centric visualization application to ensure the right information is available to the right person, in the right format, and at the right time.
The research outcomes presented in this paper lay the foundation and provide a much-needed technical basis to start focusing on additional needs such as explainability and trustworthiness of ML- and AI-based technologies as part of future research.