A practical PD online monitoring system used for transformers

In this paper, we develop a partial discharge (PD) online monitoring system with frequency band from 50 kHz to 1 megahertz used for transformers. The monitoring system mainly consists of wide-band signal detectors for multi-terminal measurements, signal pre-processing circuit, high speed A/D converter and cascaded 2nd order IIR lattice digital notch filter. The major measures to suppress narrowband interference are the signal pre-processing device and the digital notch filter. With analysis to field monitored signal from a transformer at a 220 kV substation, it is shown that the monitoring system can detect and extract the discharge pulse signal effectively. With PD pattern recognition method, the monitoring system can get more stable and reliable detection results.