Radiometric partial discharge characterisation
The process of generation, transmission and distribution of electric energy appears to be very simple. However, the continuous flow of power through electric networks is affected by several factors, normally of two kinds: failures caused by ageing of insulating materials and failures caused by defective equipment or design. The energy supplied to small and big consumers can be interrupted for long periods causing disastrous economic losses. The origin of most failures normally is found in electrical insulation, which is an essential component in all electrical systems especially in power equipment such as transformers and circuit breakers, the cost of which is very high. Before the insulation fails, there are manifestations of this condition, called partial discharges (PD) and because of this, PD have been subject of an extensive study and investigation. PD identification relies on the intrinsic characteristics of the generated discharge. This thesis shows the behaviour of radiated discharge through the analysis of the characteristic frequency spectra and pulse distribution to create pattern classification in the laboratory. This research is based on a novel technique that uses wide band, high resolution equipment, composed of an antenna array, that receives the radiated discharge, and a high resolution oscilloscope that accurately measures and samples the propagated signal. The second part of the laboratory setup comprises the characterization of partial discharges using a high voltage transformer and measurement devices that accurately apply a range of voltages to specimens containing insulation defects. The advantages of having accurate PD sources are evident: availability, free of ambiance noise, accurate AC reference and PD location. Results show that pattern recognition to identify PD'S using a non invasive system and point on wave information is possible and reliable.