Implementation of remaining useful lifetime transformer models in the fleet-wide prognostics and health management suite

Research and development efforts are required to address aging and reliability concerns of the existing fleet of nuclear power plants. As most plants continue to operate beyond their original 40 year license life, plant components are more likely to incur age-related degradation mechanisms. To assess and manage the health of aging plant assets across the nuclear industry, the Electric Power Research Institute has developed the Fleet-Wide Prognostic and Health Management (FW-PHM) Suite, a set of web-based diagnostic and prognostic tools and databases comprised of the Diagnostic Advisor, the Asset Fault Signature Database, the Remaining Useful Life Advisor, and the Remaining Useful Life Database, that serves as an integrated health monitoring architecture. The main focus of this paper is the implementation of prognostic models for generator step-up transformers in the FW-PHM Suite. One prognostic model discussed is based on the functional relationship between degree of polymerization, the most commonly used metric to assess the health of the winding insulation in a transformer, and furfural concentration in the insulating oil. The other model is based on thermally induced degradation of the transformer insulation. By utilizing transformer loading information, established thermal models are used to estimate the hot spot temperature inside the transformer winding. Both models are implemented in the Remaining Useful Life Database of the FW-PHM Suite. The Remaining Useful Life Advisor utilizes the implemented prognostic models to estimate the remaining useful life of the paper winding insulation in the transformer based on actual oil testing and operational data.