Differential protection of power transformers using the wavelet transform

Several disturbances involving power transformers commonly occur in the Electric Power System. External faults near the transformer, internal faults, as well as transformer energizing and overexcitation are the main events evaluated in the protection of the power transformers. All of the aforementioned events present transients, which can be properly analyzed by the wavelet transform. The Maximal Overlap Discrete Wavelet Transform (MODWT) can be used to detect transients in some power system disturbances. This paper presents an algorithm based on MODWT for detection and identification of external faults near or far the transformer, internal faults, and transformer energization by using differential currents in the wavelet domain. A system of 230 kV was modeled using ATP/EMTP software. The results reveal the advantages of using this type of modeling, since It provides reliability and speed in detection of these events.