Bayes inference for reliability of HV insulation systems in the presence of switching voltage surges using a Weibull stress-strength model

This work shows a Bayesian approach to evaluate the effect of switching overvoltages on HV insulation reliability. Indeed, HV insulation is often designed to withstand rated temperature and nominal (AC or DC) voltage but not voltage surges, whose amplitude can largely exceed rated voltage even in the case of switching overvoltages. Since both voltage surges and insulation electric strength (affected by aging) have stochastic nature, reliability evaluation involves the use of the relevant probability distributions, whose parameters are generally known with large uncertainty. Thus, the methodology for reliability assessment here presented, based on Bayesian estimation, is a useful tool in the case of HV insulation subjected to switching voltage surges. Under proper assumptions, among which a particular Weibull stress-strength model - leading to a loglogistic distribution for lifetime the proposed method enables the analytical determination of prior and posterior distributions of reliability and other related parameters, such as given percentiles of service life, and their Bayes point and interval estimates. The analytical feasibility and simplicity of the procedure are illustrated in the paper by means of a numerical application, and the efficiency of estimates is shown through extensive Monte Carlo simulations. The method is applied to HVAC cable insulation.