Identification of Transformer Defects via Analyzing Gases Dissolved in Oil

Abstract

A description of almost all the current widely used methods used for the interpretation of the results obtained in analyzing gases dissolved in oil (AGDO) is presented. In addition, a new method for identifying defect types, an algorithm for its application, and the advantages thereof over other methods are described. A technique for testing the reliability of the AGDO interpretation methods and results obtained upon testing these methods are presented. The quality of the detection provided by the considered methods has been compared with respect to various types of defects. In the course of the testing, special attention was paid to examining the issue that different methods of AGDO interpretation can identify different number of defect types. The evaluation of the quality of the methods and the testing of their reliability have been performed for a test sample containing 134 cases of transformer damage. For each case, there is a detailed description of the defect development process (the results of analyzing and testing of a transformer for the preceding period and the records in the operational documentation), as well as the results of transformer investigation, in the course of which the cause of damage done to it has been established. © 2019, Allerton Press, Inc.