Practical issues of online partial discharge measurement for in-service substation transformer

Abstract

Partial discharge (PD) measurement provides a means for online monitoring and assessing the insulation condition of a substation transformer. However, the characteristics of PD pulses may vary depending on various factors such as geometry of insulation system, types of PD sources, properties of PD sensors and measurement systems, and sampling rate. Among these factors, sampling rate of PD measurement system has significant effect on the acquisition of PD waveform and the extraction of PD signals. This paper evaluated different sampling rates for online PD measurement of in-service substation transformers. The results of different sampling rates are compared in terms of the waveforms of the acquired PD signals, properties of the extracted PD signals, formation of PD patterns, and dataset sizes for proper PD evaluation, which will be presented in the paper. Other important issues such as online PD measurement system setup, filtration of power frequency harmonics, and PD signal extraction using wavelet transform are also discussed in this paper.