Investigation of online detected partial discharges in power transformer

In this paper some advantage of online monitoring in power transformer and the sensors that can be used for partial discharge detection are described. Several detected PD in one power transformer and one distribution transformer are depicted. Matlabs wavelet package is employed to de-noise the recorded signals. As it appear from the captured and the de-noised signals, PD signals in an aged power transformer are often enough large that can be easily detected through the noisy measured signals. By examination of the frequency spectrum of several PD signal in this paper it is obvious that this signal contain two major frequency band, one under 1MHz and the other one between 7 to 9 MHz.