Transformer Inrush Currents and Internal Faults Identification in Power Transformers Using Wavelet Energy Gradient

The main problem with differential relays performance in power transformer protection is a correct distinction between internal faults and inrush currents. A new methodology to differentiate transformer inrush currents from internal faults using the root-mean-square of the wavelet energy gradient is presented in this paper. The technique is based on calculating the root-mean- square of wavelet energy gradient using the differential current sliding data window due to inrush currents and internal faults over a specified frequency band. The method effectiveness has been verified in several computational simulation cases using EMTP/ATP and MATLAB®, analyzing inrush currents and internal faults situations, including cases of sympathetic inrush in a power transformer. The proposed method main advantage is to eliminate the need for setting thresholds, since the criterion used is based on comparing two internal indices. This work demonstrated that the proposed method is able to perform a perfect separation between internal fault currents and inrush and sympathetic inrush currents. © 2016, Brazilian Society for Automatics--SBA.