Differential Protection of Three-Phase Transformers Using Wavelet Transforms

This paper proposes a novel formulation for differential protection of three-phase transformers using Wavelet Transforms (WTs). The new proposed methodology implements the WTs to extract predominant transient signals originated by transformer internal faults and captured from the current transformers. The Wavelet Transform is an efficient sign processing tool used to study non stationary signals with fast transition (high frequency components), mapping the signal in time-frequency representation. The three phase differential currents are the input signals used on-line to detect internal faults. The performance of this algorithm is demonstrated through simulation of different internal faults and switching conditions on a power transformer using ATP/EMTP software. The analyzed data is obtained from simulation of different normal and faulty operating conditions such as internal faults phase/phase, phase/ground-, magnetizing inrush and external faults. The case study shows that the new algorithm is highly accurate and effective.