Evaluating the accuracy of different DGA techniques for improving the transformer oil quality interpretation

Abstract

In each power utility, a numbers of different tests are conducted for evaluating the quality of transformer oil. The information provided by each measurement and analysis techniques has varying degree of significance towards the identification of quality of oil. Consequently, it is important to understand the accuracy of different interpretations. In this paper, Dissolved Gas in oil Analysis (DGA), one of the most important tools for transformer fault diagnosis is investigated. In a case study different DGA diagnosis methods are compared at different rated voltage levels and based on the obtained results accuracy of each method is calculated. The examined dataset consists of 120 different fault cases from a power utility.