Fault classification in power transformer using polarization depolarization current analysis

The decomposition of insulation structure due to ageing and faults in transformer will changes its material properties, i.e. material conductivity and its charging and discharging current behavior. The primary aim of this work is to investigate effect of faults in transformer on polarization and depolarization current spectra. This paper also utilized Artificial Neutral Network (ANN) to identify each spectrum characteristic for faults identification purposed. The insulation oil from in-service power transformers with normal and different fault conditions were sampled and tested for Dissolved Gases Analysis (DGA) and Polarization and Depolarization Current (PDC) Analysis. The result shows that transformers with normal, partial discharge, overheating and arcing fault can be identified based on its polarization and depolarization current pattern. The ANN study had found that the faults identification is more accurate when using depolarization current compared with polarization current. A detailed analysis has demonstrated that depolarization current can provide more detailed information on the condition of transformer.