Performance Assessment of IEEE/IEC Method and Duval Triangle technique for Transformer Incipient Fault Diagnosis

Abstract

Power transformers are among the most important equipment in a power system. The assessment of the transformer’s condition which is connected to a system is needed to avoid disastrous failures and the subsequent power outages. Although transformers are highly dependable and efficient, but are always under the influence of thermal, electrical stresses which are caused due to the deterioration of transformer insulation. Internal faults in a transformer can be classified as incipient faults and internal short circuit faults. The former type of faults develops over an extended period of time and lead to the degradation of insulation. Incipient faults are generally the result of deterioration of the dielectric strength of transformer mineral oil due to hydrolysis, pyrolysis, and oxidation, which result in the formation of gases that are collected in the gap between the oil and tank and partially get dissolved in it. If incipient conditions are not corrected timely, they may finally lead to collapse of the system where such a transformer is installed. Dissolved Gas Analysis (DGA) methods are extensively accepted by utilities for detection of developing faults. In this work, the brief study of two of the prevalent DGA methods for identification of incipient faults has been carried out. The methods that have been studied are IEEE/IEC Ratio Method and Duval’s Triangle Method. The published DGA data where the actual fault is known is used for identification of the incipient faults and the results have been compared.