Memory propagation of negative point-to-plane corona under AC voltage

Partial discharge (PD) detection is an important means of evaluating the condition and quality of an insulation system. It has long been regarded that discharge distribution patterns may contain deterministic parameters relating to the insulation condition. With the advancement of electrical circuitry and computer systems, it is feasible to interpret measurable PD distribution patterns with a discharge physical mechanism. Earlier research work has found that the parameters of an arbitrary discharge event in a series are closely related to the time separation from the preceding events. However, under AC criteria, these parameters are difficult to calculate due to the add-on impact of periodically changing fields. Obviously, the characteristics of pulse-to-pulse memory propagation of Trichel-pulse corona are superimposed by the influence of the applied AC voltage. To evaluate the superimposed impact on discharge distributions, both corona amplitude and energy are employed to study the properties of negative Trichel-pulse corona. Relevant statistics of unconditional and conditional distributions are calculated for comparison and analysis. It is found that a critical time interval t_c exists between consecutive corona pulses, under which the AC voltage impact may be minimised. With the knowledge of t_c, it is possible to compute the discharge parameters relating to characteristics of the discharge mechanism under AC voltage. The results obtained from a number of conditional distributions are discussed to interpret the behaviours of negative Trichel pulses in distribution profiles