One Markov Model for Spare Analysis of Distribution Power Transformers

This paper describes a probabilistic method for evaluating the number of spare units, regular spare transformers (RSTs), and mobile unit substations (MUSs) required for a group of similar distribution power transformers used at distribution stations in order to meet a pre-determined level of the group availability. The method uses a single Markov model for representing Class I and Class II failures of distribution transformers and a two-state model for representing the mobile unit substation (MUS). The two models are then convolved to obtain the required number of spare units. The proposed method is illustrated using a Hydro One sample of the distribution transformers and the results are compared with the previously published method.