Power Transformer Differential Protection Using the Boundary Discrete Wavelet Transform

Currently, the differential function is the most used for the power transformer protection, leading to a reliable discrimination between internal faults and other events. However, the conventional phasor-based differential protection function presents difficulties in the detection of some internal faults and its performance depends on the harmonic restraint and blocking functions in order to avoid relay misoperation during inrush currents. However, internal faults and other disturbances present transient, which can be properly detected by using the wavelet transform. This paper recreates the phase current and the negative-sequence current differential elements by means of the boundary wavelet coefficient energy. The proposed method was evaluated by using representative simulations of internal faults, external faults, and transformer energizations in two different power systems. By using the boundary wavelet coefficient energy instead of phasor estimation, the proposed method was quite fast, accurate, was not affected by inrush currents in transformer energizations and fault clearance, and could be used in a real-time application with low computational burden. In addition, the proposed method presented no failure in fault with overdamped transients, was scarcely affected by the choice of the mother wavelet, presented no time delay associated with the wavelet filtering, and was not affected by typical noise.

I. Introduction

Power transformers are devices that require continuous monitoring and fast protection because they are essential to the electrical power systems [1]. About 10% of faults take place into power transformers, in which 70% of these faults are caused by short-circuits in its windings [2]. In addition, faults in transformers are generally considered one of the most severe disturbances in transmission networks [3]. Therefore, a fast power transformer protection is of utmost importance.