Feasibility Study on the Detection of Turn-to-Turn Fault Severity in the Transformer Winding by Frequency Response Analysis and Numerical Indices

Abstract

Frequency response analysis (FRA) provides valuable information on internal electrical and mechanical faults in a transformer. This information can also be used to diagnose the turn-to-turn fault (TTF) in a transformer winding. This paper concentrates on using numerical indices for the evaluation of FRA to diagnose the TTF severity of the transformer winding. This evaluation is essential for expanding the knowledge toward the interpretation of FRA results. In this paper, the FRA traces of a winding are obtained in the laboratory for a healthy and faulty transformer with various TTF severities. The test results are analyzed using the indices related to all points of FRA characteristic and the indices related to the resonance and anti-resonance points. In this analysis, the possibility of diagnosing the TTF severity is addressed. To confirm the experimental results, simulation models of different windings are implemented to repeat the evaluation. By choosing an appropriate frequency range for the evaluation, the numerical indices demonstrate a monotonic performance versus the fault severity, i.e., the TTF severity can be determined in a transformer winding by assessing the FRA results with numerical indices.