The Influence of Modeling Transformer Age Related Failures on System Reliability

Abstract

The paper investigates the effect of age related failure of power transformers on the identification of most critical transformer sites for system reliability. The end-of-life failure model of power transformers is modified first to integrate loading conditions effect. The adopted Arrhenius-Weibull probability distribution, which represents the effect of thermal stress on the transformer's end-of-life failure, was compared with the commonly used Gaussian probability distribution model. The sensitivity of results to the uncertainty in model parameters is thoroughly assessed, and acceptable level of uncertainty is determined. The results demonstrated the importance of integration of loading conditions into the failure model. The sensitivity analysis revealed that the identification of critical transformer sites is not significantly affected by the uncertainty in the failure model parameters and that approximate ranges of parameters can be used instead of accurate values without significant, if any, loss in accuracy. The case studies were performed on a realistic transmission test system with 154 power transformers.