Study On The Gas Insulated Substation Partial Discharge Diagnosis Based On Target Decomposition Gases And Its Application

Considered its outstanding insulation, compact design, Gas Insulated Substation (GIS) is widely adopted recently. Even though GIS had been thought to be free of maintenance, some faults still occurred because of latent faults, such as protruding, metallic particles attached to epoxy insulators, which were introduced into GIS during manufacturing and assembling process. Such latent faults, which is very difficult to be detected during normal check, might lead to serious casualties or property loss. In2011, National Grid recorded2026GIS compartments with faults, while Southern Grid reached recorded0.2GIS compartments with faults in every100compartments.Partial discharge diagnosis is an important method to make early warning to the faults of GIS. A lot of researches have been conducted on the area, among which are impulse current method, photoelectric method, ultrasonic method, UHF method, Chemical analysis method and so on. According to the operating experience of Guangdong Power Grid, few of the method mentioned above is effective enough to detect the types and severity of latent faults.In the first phase of the research, a small experiment apparatus was designed, in which different faults can be simulated. Gas decomposition was detected and analyzed. CS2, a first introduced gas relating to epoxy insulators damage in GIS, was discovered. Besides, the relationship between the amounts of SO2and SO2F2showed different features in different type of discharge. In order to validate the existence of CS2under epoxy insulator discharge, the relationship between the amounts of SO2and SO2F2under different discharges, and accordance of the experiments to field running, experiments were carried in110kV GIS in the second phase. The results showed accordance to those obtained in the small experiment apparatus.Inspired by the dissolved gas analysis of power transformer, more and more researcher begin to focus on the gas decomposition analysis of GIS, trying to explore an effective method of GIS partial discharge by adopting target gases, that CS2would be detected under spark discharge on epoxy insulator and the generation speed of SO2is lower than that of SO2F2under corona discharge, which reversed to that under spark discharge. In order to back up the discoveries from the experiments, Quantum Chemistry was adopted to explore the generation mechanisms of different decompositions under discharges. As the diagnosis method is based on gas analysis, a new GIS gas decomposition analysis technique also gets discovered. Based on the experiments results and Quantum Chemistry analysis, a GIS latent fault diagnosis method was raised, which adopted CS2, CF4, SO2and SO2F2as target gases. Once SO2or SO2F2is detected, discharges might happen. The generation of CS2or CF4could help prove the epoxy insulators are under deteriorating. Otherwise, discharges might happen in other parts in GIS instead on the surface of epoxy insulator. Then higher generation of SO2than that of SO2F2shows that spark discharge happened otherwise corona discharge is happening.The method has been applied to the field, where it has been validated.