On data analysis and processing of partial discharge pulses using software orientated techniques

The detection of partial discharge activity from the solid insulating systems of high voltage plant is well-recognised as a measurement whose interpretation can prove valuable in the assessment of insulation integrity. Partial discharge detectors have traditionally been configured to display the pulses from the partial discharges superimposed on the AC power cycle. Knowledge of these pulse amplitudes, their phase distribution, polarity and overall number per cycle can be used, often in conjunction with the knowledge of how these parameters vary with applied voltage and, at a fixed applied voltage, with time, to provide a valuable insight into the nature, extent and location of degradation sites within the solid insulating system. Given the plethora of parameters to be measured and presumed necessity for accuracy and objectivity, it is not surprising that digital technology has been utilised to this end and many such systems are now being reported in the literature. Indeed, the data manipulating power of such systems is now producing results which indicate that more complex pattern changes can be analysed to provide increasing sophisticated assessments relating to insulation integrity