Study On Raman Spectroscopy Properties Of Dissolved Gases In Transformer Oil And Their Detection

Dissolved gas-in-oil analysis is one of the most convenient and effective method todetermine early latent failures of oil-immersed power transformers. Research oninsulation online monitoring methods of power transformer based on dissolved gas-in-oilanalysis is important for improving the operation and maintenance level of powertransformer and protecting the safe and stable operation of power system. Current onlinemonitoring methods detecting dissolved gas-in-oil of power transformer mainly are GasChromatography, Fourier Transform Infrared Spectroscopy, sensor array method,Photoacoustic Spectroscopy. Raman spectroscopy, based on Raman Effect, is a molecularidentification and analysis technology. Raman spectroscopy gas analysis technique is ableto directly analyze the components of mixed gases without gas separation, which isespecially suitable for gas-in-oil online monitoring. Studying the application of Ramanspectroscopy in online monitoring of dissolved gases in power transformer oil hasimportant significance for the continuous improvement of insulation monitoringtechnology of power transformer. The main work of this thesis is as follows.①According to the research progress of detecting dissolved gas in transformer oilthrough Raman spectroscopy, the author studied the electromagnetic models and thequantum models starting from the basic principles of Raman scattering. By analyzing thebasic structure of confocal microscopy Raman techniques and comparing the principlesand applications of a few common Raman enhancement means, a method for detectingdissolved gas in transformer oil using Raman spectroscopy was established. The thesisintroduced the basic principles of density functional theory calculations. And the Ramanspectrum, Raman frequencies and Raman activities of the7fault characteristic gases (H2,CH4, C2H2, C2H4, C2H6, CO, CO2) of power transformer were simulated by this method.②A Raman spectroscopy test platform for detecting dissolved gas in transformer oilbased on confocal microscopy Raman system and optical fiber enhancement method wasdesigned. And the structural frame and equipment selection of this platform wasdetermined. Utilizing a similar Raman detection platform of Chinese Academy ofSciences, the Raman spectrums of the7fault characteristic gases were detected. A varietyof signal processing methods, such as Savitzky-Golay polynomial smoothing, wereapplied to improve the Signal to Noise Ratio of the original signal Raman data obtainedby experiments. Test results verified the feasibility of Raman spectroscopy applied to dissolved gas-in-oil analysis③According to simulation analysis and experimental results, Raman characteristicspectrums of the7fault characteristic gases were elected,which realized the qualitativeanalysis of the7fault characteristic gases based on Raman spectroscopy. And the Ramancharacteristic spectrum of H2, CH4, C2H2, C2H4, C2H6, CO and CO2were4157cm-1,2918cm-1,1973cm-1,1344cm-1,2954cm-1,2142cm-1,1388cm-1, respectively. Throughsimulation analysis, the attributions of these characteristic peaks were identified. Usingthe external standard method on the base of principle of least squares, a Ramanspectroscopy quantitative analysis model of dissolved gas in transformer oil wasestablished.The research work of the thesis laid the foundation for Raman spectroscopy applied toonline monitoring of dissolved gas in transformer oil.