Quantifying the Probability of Partial Discharge in VFD Fed Electric Motors under Voltage Harmonics Concentration

Partial discharges (PD) diagnostics are considered as reliable techniques for the insulation health assessment of various electrical components of the power system. Generally, PD diagnostic tests are performed in laboratory by ignoring the effect of harmonic pollution in test voltage waveform. However, the operation of electric motors (EMs) fed by variable frequency drives (VFD) produces harmonic distortion in the sinusoidal test voltage waveform. During low speed operation of VFD, the influence of harmonic distortion in test voltage waveform is increased. The additional distortion parameters in the test voltage waveform can significantly increase the PD activity in the EM insulation. Therefore, for correct understanding of PD data, it is essential to consider the actual harmonic spectrum of the test voltage waveform during online PD measurement. This paper presents a methodology for the quantification of the probability of PD in VFD-fed EMs under different concentration of voltage harmonics. For the experimental work, PD diagnostic tests have been performed under variable operating conditions of eight EMs and PD severity is investigated by calculating the PD characteristic parameters at different harmonic distortion levels. Also, the probability density functions and cumulative distribution functions for several probability distributions were evaluated to statistically process the PD data at different harmonic levels. Finally, the distribution fitting tools have been implemented to find an appropriate distribution that can precisely characterize the probability of PD under different concentration of voltage harmonics. The proposed method may be utilized for studying the effect of harmonic distortion on PD activity and estimating the lifetime of stator insulation in EMs.