Spare Assessment of Distribution Power Transformers using Two Markov Models

In earlier spare assessment studies of distribution power transformers at Hydro One, the issues of the full utilization of mobile unit substations (MUSs) and their reliabilities were not fully addressed and therefore, the results of spare studies may have been underestimated. This paper describes a study that has been performed recently to address the two mentioned issues. The study used a simple and flexible probabilistic approach that shows how the two issues can be properly addressed and helps explain the results of earlier spare studies. The proposed assessment approach uses two Markov models: one representing minor transformer failures and one representing major transformer failures for a group of similar distribution power transformers. The MUS utilization factor introduced in this study is incorporated into each failure model in order to obtain the group availability as a function of the number of spare units. The results of a sample distribution system show that the two issues can have significant impacts on the spare assessment results. The purpose of this paper is to describe the study and its findings and to compare its results with the earlier spare methods of assessment.