Estimation of Remaining Useful Life and Failure Probability of an Electrical Machine - Case Study

The main purpose of this paper is to provide a methodology for a condition assessment of an electrical machine as a significant asset that requires maintenance. The remaining useful life (RUL) and the failure probability of an alternator with salient poles rotor are determined for various temperature profiles in order to demonstrate how RUL and failure probability change under different levels of main stress factor which is a temperature in this case. As a nominal lifetime changes non-linearly in the relation to the stress factor, the consumed RUL and reliability change as well. The most stressed location was identified and the empirical power-law aging model for a particular thermal class of electrical insulation system (EIS) was selected for the RUL estimation. Subsequently, the exponential and Weibull distribution incorporating parameters obtained through literature research were used to evaluate the failure probability. The numerical and graphical results were obtained via the custom script which analyzed a huge bulk of data recorded during one year of continuous device’s operation.