Detection of corona with RF methods and spectra analysis

High voltage power electronic systems are increasingly being used in applications such as renewable energy generation and in transportation systems. Such high voltage systems are often placed in compact packages to minimize weight and volume. In minimizing package size, there are consequences on the design of the high voltage insulation system which is likely to be operated closer to design limits. This can lead to an increasing number of failures due to tracking, partial discharges or disruptive discharge. An opportunity therefore exists to develop a monitoring system that can detect a range of electrical discharges and which is able to work in environments polluted by electromagnetic noise generated by rapid switching of power electronics under any load conditions. Of the methods that have been developed in recent years (optical, ultrasonic, RF etc.) the ones based on RF techniques are the most promising. To enable an easy differentiation between the RF signals produced by a PD and the random environmental noise it is crucial to understand the type of emissions that can be generated by different kind of discharges. In this experimental work carried out in a controlled electromagnetic noise-free environment, arbitrary corona discharges (of known magnitude and polarity) have been detected with RF techniques and the signals have been analyzed. Results relative to positive and negative corona generated with blade-plane gap specimen at varying voltages (1-4kV) and at differing pressures (11-101.3 kPa) are reported. This is the first part of a study targeted to classify the spectral emissions associated with different PDs; future work will analyze emissions due to voids, floating particles, tracking etc. The final aim of this project is the development of a system capable of a monitoring a range of discharge types in high voltage power converters including those operated at altitude in aerospace applications.