Statistical analysis of non-homogeneous life-data of 50 kV tap changers for residual life assessment purposes

Statistical tools are useful for the residual life assessment of the population of power system component. More specific, by fitting a statistical distribution to the life-data of an investigated population its life-stage can be recognized. However, there is a number of requirement with respect to the data quality that is supplied for the analysis. The homogeneity, independence and randomness should be recognizable in the sufficient amount of data to enable distribution fitting and assure the representative and exact results. However, due to characteristics of the power network, the life-data representing populations of high-voltage components is not always homogeneous with respect to the operating regime. In addition, the amount of failure data available for the analysis is often low in comparison to units remaining in service. In our paper, we will present the analysis of the Weibull distribution applicability for the population of 50 kV tap changers, knowing that the past loading profile varies between individual units.