Research On The Decomposition Characteristics Of SF₆ Of Solid Insulation Defect In PD Of GIS

The advantages of GIS (gas insulated combination switch) is obvious and has been widely used. But there are some insulation defects which will lead to the GIS insulation degradation that induced by partial discharge (PD) and affect the running safety. The main methods to detect the partial discharge of GIS equipment are: pulse current method, the method of ultra-high-frequency (UHF), ultrasonic method and decomposition gas test, etc.By comparison, although the sensitivity of decomposition gas test is low, but it means that the defects are more serious defects once detected, and the decomposition gas test has the advantages of strong ability of resistance to electromagnetic interference and has become a research hotspot now. Based on decomposition gas test, selecting the the more common solid insulation defects as the research object and setting up electrode materials for iron and copper solid insulation defect model respectively on the real 110 kV GIS and setting up on the high voltage conductor spike as control group at the same time to study the trends of SF6 gas decomposition product which caused by solid insulation defect in GIS bureau down to provide the reference for gas test type of insulation defect diagnosis technology.The general bureau of peatlands unchanged during the process of test and apply a 240 h of continuous pressure in the insulation defect model, and at regular intervals during the testing process by taking port within the GIS model to monitor the SF6 gas composition to observe the change trend of component concentration. Is obtained by experiment:for solid insulation defects, sensitive gas was mainly CO2, CO, SOF2, the extension of its gas production has significant growth as the discharge time, and gas production of CO and CO2 and gas rate are higher than that of high pressure spikes conductor defects. The iron electrode of solid insulation defects will produce a large number of CF4 and a small amount of CS2 and is different from the characteristics of the copper electrode gas; Regardless of the electrode material which is iron or copper, solid insulation defects can produce COS and is used as the difference between the characteristics of the gas in high pressure spikes conductor defects. Finally in this paper, the method of gas concentration ratio of solid insulation defect type diagnosis method and the fuzzy clustering method have proved its diagnostic accuracy which is higher than using a single sensitive gas in diagnosis, and construct the corresponding decision tree ID3 algorithm.