Power Transformer Fault Diagnosis Based-on Dissolved Gas Analysis

The typical oil-filled electrical equipments include transformers, reactors and oil-filled mutual inductor etc.It's widely applied in power system.As the hub of the power transmission equipment,the operation will directly affect the safety of power systems and engendered a significant economic impact and damage with a fault occurred.Therefore,the research on insulation condition of the transformer,detect the potential transformer failure early to ensure their safe operation and improve the reliability of power supply is a long-standing important issue to the electricity sector.Based on insulating oil analysis,the research of the transformer fault diagnosis technology have great significance to improve the level of transformer operation and maintenance.DGA is an effective method in transformer fault diagnosis, and how to improve the method accuracy and reliability of fault diagnosis is one of the focuses. In the article, the relation between the characteristic gas and the faults was analyzed with the mechanism of generating dissolved gas in transformer oil. And the fault can be judged according to the gas content threshold and the gas production rate, the fault nature can be judged according to the gas characteristic content and the total hydrocarbon, the degree of insulation aging can be judged according to the characteristic quantity of CO and CO₂. Then, the method of diagnosing fault with dissolved gas ratio in oil was discussed. In allusion to the limitations of the three-ratio method, a transformer comprehensive fault diagnosis was realized with the improved three-ratio coding method, dimensional graphic method and MATLAB as a platform. And the practical example was presented.In addition, this article studied the transformer fault diagnosis based on gray relation analysis, the dissolved gases, such as H₂, CH₄, C₂H₆, C₂H₄, C₂H₂ ,are selected as the characteristic parameters to constitute the vector of Standard state model, the type of fault is determined by analyzing the grey relational grade between the diagnostic fault mode and the all standard modes. However, the condition of solid insulating material is not fully taken into account in the diagnosis. Therefore, the associated growthsituation of CO with the main characteristic gases is considered as theassistant for diagnosis, solid insulating fault is discussed by analyzing thecorrelation coefficient of CO and H₂ or CH ₄.Thus the cause of the fault canbe identified in specific details, and a large number of diagnosed examplesare taken to prove the effectiveness of this improved method.