Evaluation aging failure probability of generating units using data analytic method

Abstract

This paper presents a method to predict failure probability related to the aging of generating units in power systems. To calculate aging failure probability, Weibull distribution is usually used due to age-related reliability. The shape and scale parameters of Weibull distribution can be estimated by using hybrid method using both normal and Weibull distributions. In this paper, data analytic method applied to type II censoring is proposed which is relatively simpler and faster than the hybrid method for the estimation of parameters. The real historical data of combustion turbine generating units in Korean power systems are used in the case study to show how to calculate the mean life and its standard deviation by the proposed method. Copyright © 2008 John Wiley & Sons, Ltd.