DGA, Wavelet & Ann Techniques for Failure Analysis of Power Transformers

In the proposed paper, we have an evident comparative study of three types of power transformer failure analysis tests. The methods are Dissolved Gas Analysis, Wavelet technology-applied temperature sensor concept and Artificial Neural Networks approach. In Dissolved Gas Analysis, the gas concentrations are determined in the transformer oil sample for diagnosing the fault from the ratio of two suitable gases. Computing method was developed for Wavelet technology application through temperature sensor to find out the fault intensity. Artificial Neural Networks criteria was utilized for considering the key gas concentration ratios to analyze corresponding faults. Here, an attempt has been done to demonstrate the application of these three concepts on a 132/33kV, 15MVA power transformer from Port substation in Andhra Pradesh, India, under three different conditions i.e., healthy, moderately deteriorated and extensively deteriorated conditions and results obtained.