Failure Analysis of Power Transformers

This paper deals with three types of power transformer failure analysis tests which can help in monitoring the transformer condition. The three methods are Conventional oil test, Furan derivatives test and Markov models criteria. Conventional oil testing procedure includes determination of gas concentrations in the transformer oil sample and analyzing the fault by considering ratio of two suitable gases. Furan derivative analysis includes determination of Furan derivatives present in the transformer oil sample and helps in deciding the transformer insulation condition. A computer program was developed for the Hidden Markov model, which helps in finding out the fault probability of the power transformer. Here, we demonstrate the application of these three methods on a 132/33kV, 15MVA power transformer from Port substation, Andhra Pradesh, India, under three different conditions namely healthy, moderately deteriorated and extensively deteriorated conditions and results obtained. Keywords—Transformer oil testing, dissolved gas analysis, furan derivatives analysis, hidden markov model, fault diagnosis, MATLAB