Condition diagnosis of transformer oil-paper insulation using dielectric response fingerprint characteristics

Abstract

The life expectancy of a transformer is closely related to the condition of its oil-paper insulation. In the present study, oil impregnated pressboard samples with different aging states and water contents were prepared under controlled laboratory conditions to analyze the oil-paper insulation condition of the power transformer by the dielectric response method. Dielectric response measurements, i.e., polarization and depolarization currents (PDC) and frequency domain spectroscopy (FDS), were conducted on the pressboard samples at different temperatures. Results show that the dielectric responses of oil impregnated pressboards are highly sensitive to aging, water, and temperature, but the influence of the temperature on FDS can be effectively eliminated by using the master curve technique. Meanwhile, dielectric characteristic parameters, which are sensitive to aging state and water content, were extracted to form a fingerprint database to indicate the various states of the oil impregnated pressboard. A grey correlation diagnosis model was first introduced to assess the aging state and water content of oil-paper insulation. Finally, FDS measurements were performed on a field transformer to verify the effectiveness of the new evaluation technique proposed in this paper for the condition assessment of the transformer oil-paper insulation.