The statistical analysis of harmonic influenced electrical treeing partial discharge data using the Weibull distribution

This paper highlights the statistical analysis of electrical treeing (ET) partial discharge (PD) activity affected by varying power quality. The harmonic influenced electrical treeing PD data captured from 42 laboratory experiments is statistically processed applying the two parameter Weibull function. The excitation waveforms employed are defined by varied harmonic orders and magnitudes quantified by the total harmonic distortion (THD) index ≤ 40% and waveshape factor (K_s) ≤ 1.6. To facilitate an almost effortless statistical analysis, a user friendly MATLAB based software tool was developed to automatically process the PD data, displaying the statistical results. Electrical treeing PD activity, captured on application of waveforms incorporating the 7^th harmonic, exhibited larger scale (α) values, relative to waveforms with an equivalent THD level-suggesting increased magnitude of PD activity and potentially accelerated electrical tree growth. For composite waveforms incorporating the 1^st, 5^th, 7^th, 11^th, 13^th, 23^rd and 25^th harmonics, there was a direct correlation to the magnitude of THD with α and shape (β) values.