With the rapid development of HVDC technology and the issue of smart grid, it is a new challenge to monitor and diagnose performance of the related power facilities. This thesis dealt with the statistical characteristics of partial discharge(PD) pulse in SF6 gas under HVDC in terms of discharge inception voltage(DIV) and discharge extinction voltage(DEV), discharge magnitude, and pulse count as well as the statistical characteristics extracted from the discharge distribution and density function.
To simulate the typical insulation defects in gas insulated switchgear(GIS), electrode systems such as a protrusion on conductor(POC), a protrusion on enclosure(POE), a free particle(FP), a void inside spacer(Void), and a crack inside spacer(Crack) were fabricated. All of them were filled with SF6 gas in ranges from 0.1MPa to 0.5MPa. A HVDC source was generated by a rectifying circuit which is composed of a 100kV diode, and a 0.5μF capacitor. PD signal produced from the electrode systems was detected through a 50Ω non-inductive resistor and was analyzed by a digital storage oscilloscope(DSO) with a sampling rate of 5GS/s and a DAQ system based on LabVIEW program.
The DIV and DEV in POC, POE, and Crack increased with the gas pressure. The gas pressure did not strongly affect the DIV and DEV in FP. The DIV and DEV in Void were almost similar to the increase of the SF6 gas pressure. The maximum discharge magnitude Qmax and pulse count in 5 seconds of each electrode system increased as the applied voltage was raised. However, the mean discharge magnitude Qmean did not change significantly. For each electrode system, the discharge distribution and density function presented distinguishable patterns. Therefore, it is possible to identify the type of defects in gas insulated equipment operated under HVDC by analysing statistical characteristics extracted from the discharge distribution and density function.