Novel Method for Discrimination of Transformers Faults from Magnetizing Inrush Currents Using Wavelet Transform

Discrimination of power transformer inrush from fault current is most important to proper operation of differential relays. Most differential relays discriminate the inrush current from short circuit fault according to high content of second harmonic of inrush. Unfortunately, in some power system operation, it has been seen that this algorithm is not accurate enough due to the presence of some kinds of loads. On the other hand, current transformer (CT) saturation due to large inrush or fault current also can cause the discrimination algorithms mal-operation. This paper presents a novel algorithm for power transformer differential protection which differentiates internal faults from those of magnetizing inrush currents in the presence of CT saturation. In order to eliminate the impact of CT saturation on the performance of the proposed technique, at first the currents are compensated by a CT saturation compensation method. Furthermore, a technique based on wavelet transform is used for discrimination of transformer internal fault from magnetizing inrush currents. Using discrete wavelet transform (DWT) features of primary and secondary differential currents and those of flux measured at transformer primary winding, an algorithm was proposed. Discrimination criterion is based on synchronization of maximum or minimum points between wavelet coefficient of differential current and flux of the transformer. The efficiency of the algorithm was evaluated using different scenarios of inrush and internal fault current of a power transformer simulated in MATLAB/Simulink software environment.