Research And Application On The Fault Diagnosis System Of SF₆Electrical Equipment Through SO₂Detection

SF6electrical equipment has been widely used in high-voltage transmission network. The event of failure of these devices will be a serious threat to the security and stability of the electricity network. Therefore, the detection technology of SF6electrical equipment’s operating status is a hot issue in today’s researchers.The excellent performance of SF6electrical equipment depends on its SF6gas, which has been used as its insulating medium. It is possible to judge the fault behavior by analyzing the decomposition product of SF6as different internal insulation faults of electrical equipment will result in different decomposition product of SF6. Up to now, the using detection methods of decomposition product are called’periodic sampling-offline censorship’, which are time-consuming and not suitable for the detection of equipment failure in time.In this dissertation, ultraviolet spectroscopy techniques were used to detect SF6discharge decomposition derivatives and a method based on SO2was proposed to diagnose SF6electrical equipment fault. A monitoring system was finally designed, which could be used on-line and on-site.The main research of this dissertation is as following:SF6and its derivatives ultraviolet spectrum characteristic are researched. For decomposition mechanism of SF6in discharge, predecessors have done much research, they are consolidated in this decomposition and finally SO2is selected as the characteristics of the derivative in the partial discharge equipment failure. SF6and its derivatives ultraviolet spectrum are measured in this decomposition. It finds that CS2, H2S and SO2ultraviolet spectra are recognizable, however SF6absorption spectrum in ultraviolet region do not have significant statistical law. Absorption spectrum of SO2has special fluctuations in the structure of the valley at295nm-305nm and209nm-219nm, is that absorbance peak concentration is positively correlated with the trend. Based on the achievements, a new method to monitor partial discharge of electrical equipment online is advanced by detecting ultraviolet spectroscopy of SO2.The influence of environmental factors on the low concentration of SO2is researched. Environmental factors affecting the disturbance cannot be ignored online. Temperature, barometric pressure, optical path and other environmental factors are analyzed to test the influence of SO2ultraviolet detecting with a series of experiments in this decomposition. By analyzing the experiment data, the influence of field temperature to SO2absorbance can be ignored. When the total pressure and SO2 partial pressure changes simultaneously, the absorbance change and the pressure difference is approximately linear law. When the total pressure changes and SO2partial pressure is unchanged, meaning that the concentration keeps a certain and background pressure changes, the absorbance change and the pressure difference is approximately nonlinear law. The change of optical path affects only the SO2absorption peak amplitude, and attenuation coefficient waves in optical path reduction ratio of neighborhood. Based on the achievements, the spectral compensation scheme of ultraviolet detection system that when barometric pressure and optical path changes is advanced to eliminate the adverse impact of environmental change on the detection system.In view of the SO2content by the discharge is usually low in the electric equipment, a detecting algorithm based on frequency domain analysis of SO2in ultraviolet spectrum of weak signal is put forward in this dissertation,use FFT spectral analysis, autocorrelation detection, power spectral density analysis method to trace the characteristic spectrum signal extraction of SO2in SF6. On this basis, the dissertation puts forward the binary state criterion, according to the results of the three judgments, using the voting method to estimate SO2information, and then estimate the partial discharge fault. Finally, using the standard gas experiment to determine a criterion for judging apriori threshold, and obtained quantitative calibration curve of SO2in (295nm-305nm) and (209nm-219nm) the concentration of the two characteristics band. The method has solved the detection problem of SO2in SF6electrical equipment, the minimum limit of detection is1ppm.According to the statistical properties of SO2in the ultraviolet spectral region, On-line monitoring system for SF6electrical equipment has been developed. This dissertation has detail description and design of the system, such as ultraviolet spectrometer, light source, MCU and other major components. Through a number of field case detection, the internal discharge fault of the SF6current transformer is founded, and a large number of field measured data is acquired. Compared with the traditional electrochemical sensors, the system have the advantage of low cost, simple structure, no frequent calibration and adapt to the environment changes, and has favorable application prospects.