Experimental detection and localization of fault in the winding of 220kv generator transformer using gabor wavelet

Standard method adopted for testing of a power transformer under impulse is by comparison of neutral current at reduced and full voltages. Partial or complete failure of transformer is reflected in the change of voltage or neutral current transient characteristics. The changes in characteristics depend primarily on the extent of failure. For example, a disc to disc failure in the transformer winding may result in an instantaneous dip in voltage and corresponding change in neutral current whereas a winding failure may result in sudden collapse of voltage and abrupt rise in current. In order to detect such faults, the Transfer Function Technique (FFT) is used widely by several testing agencies. This technique has been applied for evaluating the transient admittance of the winding. These transient admittances are deconvoluted into frequency domain using Fast Fourier Transform (1, 2, 3) method. Though, the Transfer Function technique has proved quite effective in diagnosing the fault in the transformer, some of the lower range frequencies is difficult to detect with clarity. A new technique based on Mexican and Morlet wavelets 94, 5) is reported in literature which can be effectively used for detection and localization of fault in disc/turn of transformer winding. This technique is an advancement over transfer function, since signal can be analyzed with small as well as large windows. Hence, even micro details associated with signal can be absorbed. Wavelet analysis is a relatively new method and is a powerful tool in the analysis of power transformers transient phenomena because of its ability in both the time and frequency domain. Hence, Wavelet Transform (WT) has been used to determine the time and frequency characteristics in case of any failure in the transformer winding. The present article deals with analysis of neutral current based on Gabor wavelet technique. The measured neutral current is subjected to FFT for determination of dominant frequencies. These neutral currents are also analyzed using the above wavelet. The concept of wavelet transform, the experimental studies and result of application of wavelet are described.