Statistical failure reliability analysis on edible and non edible natural esters based liquid insulation for the applications in high voltage transformers

Abstract

In the recent decades, the natural esters have gained more attention as an alternate liquid insulation instead of the petroleum based mineral oil due to the development on power system networks and their environmental friendly nature. The failure free functioning of high voltage transformer mainly depends on the reliability of insulation used inside it. In this work, edible and non edible natural esters are studied and analyzed by using statistical distribution of the AC breakdown voltage in order to ensure their reliability. In this investigation work, sunflower oil (SFO), palm oil (PO) and sesame oil (SO) are studied under the category of edible natural esters, while honge oil (HO), neem oil (NEO) and punna oil (PAO) are analyzed under the non edible natural esters category. The applicability of investigated esters as liquid insulation is analyzed with distribution models like normal, Weibull and generalized extreme value (GEV) distributions. The statistical failure reliability analysis is carried out with the distribution functions such as survival rate, hazard rate and failure rate from the 100 distributed values of AC breakdown voltages of the investigating natural esters. The statistical analysis reveals the better characteristics for natural esters (edible and non edible) such as superior survival rate at higher value of electric stress, possibility of failure occurrence only at the wear out range and elevated statistical withstand voltage at various failure rates. Comparing both the category of natural esters, non edible natural esters have shown the superior characteristics than the edible natural esters. Consequently, it is evident that the natural esters are reliable and potential surrogate to the conventional liquid insulation for the application in high voltage transformers.