On the use of statistical methods and models for predicting the end of life of electric power equipment

This paper is devoted to the problem of estimating the end of life of electric power equipment. Several statistical methods and models commonly used in reliability engineering are discussed, focusing on (a) the technical merit of each described method, and (b) the suitability of each method for determining the end of life of power equipment. Description of statistical methods considers both censored and uncensored observations. The problem of model selection for analysis of censored data is given special consideration: a novel approach is proposed for this. The points made in the text are illustrated by several numerical examples pertaining to the lifetime of transformers, however the methods can be applied for other types of electric or electronic equipment (e.g., electronic relays).