Techniques for the diagnosis of oscillatory transients resulting from capacitor bank switching in medium voltage distribution systems

Abstract The oscillatory transient has well-dened characteristics such as short duration, considerable variations in the voltage’s amplitude, and frequency, which can be manifested in any part of an Electric Power System and mainly caused by the capacitor bank switching. Knowing the effects caused, for possible mitigation of these, it is initially necessary to recognize their origin, since, for example, the possible disturbance recorders installed or other measuring equipment available on medium voltage systems do not point out the causes. The methodology presented in this research comprehends a complete study since the detection of an oscillatory transient until the extraction of the main characteristics of the signal in analysis, resulting from the capacitor bank switching. Initially, a comparative analysis of the processing time and the mean error in the detection was carried out by using Notch Filter, Wavelet Transform, and Mathematical Morphology techniques to identify the beginning of oscillatory transients in synthetic signals, with and without the presence of noise. The results obtained pointed out that the Mathematical Morphology identified with greater speed and less computational effort the presence of an oscillatory transient in comparison to the others. Later, applying the Matching Pursuit Algorithm with a structured data dictionary, it was possible to estimate the frequency (f), damping (d), and phase angle (a) parameters characteristic of the analyzed oscillatory transients.