Partial discharge magnitude distribution analysis in characterising ageing phenomena in high voltage rotating machine insulation system

Over the years, many diagnostic/condition monitoring tests are in vogue for evaluating the health of HV machine insulation. While it has been possible to assess the general state and condition of the stator insulation, experience gained so far has shown that these diagnostic tests are not sensitive to detect or predict the breakdown or failure of the insulation, which is a microscopic phenomenon. In other words, the conventional diagnostic tests such as dissipation factor, capacitance, polarisation index, DC leakage current measurements are average/integrated effects and are not capable of reflecting the microscopic phenomenon such as an insulation breakdown. In view of these facts, there is a need to device a test method or to find a more searching and sensitive parameter capable of predicting the insulation breakdown. In this context, partial discharge measurements assume greater importance as it is believed that partial discharges (PD) are capable of revealing the incipient faults. The paper presents the partial discharge degradation studies carried out on the stator bars of 1.1 kV. 135 MW hydrogenerator. The generator bars have been subjected to accelerated electrical ageing and the partial discharges have been monitored periodically using Multichannel Analyser (MCA) in order to characterise the ageing phenomena. The MCA provides the distributions of partial discharge pulse magnitudes against their repetition rates. The distribution profiles have been examined in detail in order to characterise the internal discharges, slot discharges and endwinding discharges that occur in the stator winding.