On line Partial Discharge Measurements by means of signal processing tools integrated in micro processing hardware

Summary

The evaluation of the insulation condition of high power transformers, cable systems, etc. by means of on line partial discharge measurements have been traditionally an important challenge due to the difficulty of removing the background noise existing in the substations. The discrimination of PD pulses in the power transformer from other external PD pulses coming outdoor equipment, or even PD pulses coming from the bushing or from the cable termination has been also an important challenge. Power signal processing numerical tools used for filtering and for clustering have been integrated in a new micro processing hardware to improve the capabilities of on line partial discharges diagnosis. In this paper a new on line PD monitoring system is introduced that uses non-invasive sensors placed in each input phase of the power transformer, and an additional sensor at the tank earth connection. The signal acquisition of each sensor is synchronized to discriminate whether the eventual PD source found is inside or outside the transformer or along a cable system. The paper presents the selection and the architecture of the PD sensors used when the power transformer is connected by bushing or when it is connected by cable. The signal processing tools used for noise suppression, PD clustering, PD location and PD phase resolved pattern recognition are also presented. The result of the signal processing analysis permits to identify whether the PD source is in the power transformer, in the bushing, in the cable termination or along the cable system.

Additional informations

Authors

Gal, Zhao Jingxuan, Chen Peixing, Li Huan, Zhou Fusheng