Physics-Based Prognostics for Optimizing Plant Operations

摘要

Scientists at Pacific Northwest National Laboratory (PNNL) have optimized energy plant operation using DSOM (Decision Support Operation and Maintenance) which has been deployed at several sites. This approach has been expanded to include a prognostics component and tested on a pilot-scale service water system modeled on the design employed in a nuclear power plant. A key element in plant optimization is understanding and controlling the aging process of safety-specific nuclear plant components. This article covers the development and demonstration of a physics-based approach to prognostic analysis that combines distributed computing, radio frequency data links, measurement of aging precursor metrics and their correlation with degradation rate and projected machine failure.