Failure analysis and replacement strategies of distribution transformers using proportional hazard modeling

The paper presents a graphical method for plotting mean cumulative repair function (MCRF) for different capacities or types of distribution transformers. The plots provide information about their failure behavior with time. The intensity function graphs obtained using parameters of Weibull process proportional hazard model confirms the results obtained from nonparametric MCRF graphs. Proportional hazard modeling (PHM) technique is quite helpful to consider the effect of covariates on the failure performance of different transformers with time. The property of proportionality is validated with the help of graphical and analytical methods. The higher values of intensity function for 100 kVA transformers and rural environment transformers show their poor performance compared to less than 100 kVA transformers and urban environment transformers respectively. This information is quite helpful in evaluating maintenance/replacement policies. It is considered that normal life of distribution transformer is about 25 years, for a capacity of 100 kVA or less. The present study indicates that in case of 100 kVA transformers and rural transformers an early replacement is required.