Research On The Discharge Characteristics Of Typical Defect Models In Insulating Oil And The Law Of Free Gas Production

The power transformers are important equipment for transmitting electrical energy in the power system.When there is a failure,it may evolve into an explosion or fire,which will inevitably affect social production activities.And it will also affect social production and human life.Although the accuracy of transformer protection has been improved to a certain extent by continuous optimization of protection algorithms and the upgrade of related protection equipment,there are still problems such as low sensitivity of internal minor fault recognition and mis-operation of gas protection caused by external factors.The reason is the insufficient realization of the internal physical changes in the process of transformer faults which make people unable to propose new protection methods for specific stages of internal fault development.The existing electrical protection methods are ineffective for minor internal faults,non-electrical protection methods have high rate of mis-operation,and although the online monitoring method can sensitively respond to minor internal faults,its application is limited and prone to misjudgment.These reasons have still caused the transformer deflagration accidents from time to time in the world.This thesis relies on the Smart Grid Joint Fund Project"Fundamental Research on Multivariable-Based Adaptive Protection and Safe Operation for Power Transformer".This project builds a multi-variable adaptive protection through the study of various physical quantities such as electricity,magnetism,pressure,flow rate,and free gas under transformer internal faults.On the basis of summarizing previous researches,free gas has been chosen as the focus of this thesis,aiming to obtain the characteristics of free gas generation when a typical discharge fault occurs in the insulating oil,so as to provide a free gas characteristic criterion for multi-parameter transformer protection.Firstly,the gas protection of power transformers,the detection methods of internal discharge faults in transformers,and the previous researches of insulating oil discharge and gas production are summarized in this thesis.Then,an experimental device for conducting electrode discharge experiments in insulating oil is manufactured and three electrode models including corona discharge in oil,creeping discharge in oil and gap discharge in oil are also designed to simulating typical discharge fault in transformers.After that,the step-up method and the constant voltage method are used to conduct discharge experiments on the three discharge models.The high-frequency pulse current is also used to quantify the discharge characteristic information during the development of the fault.From the perspective of discharge characteristic parameters such as the average apparent charge,the characteristics of fault development under different discharge models are analyzed qualitatively and quantitatively.At the same time,combined with the gas component content information in the free gas collected,the characteristics of free gas production at different stages of different discharge fault types are analyzed.The innovative research findings in this thesis are:in corona discharge in oil,creeping discharge in oil,and gap discharge in oil,when insulation breakdown discharge occurs,the average apparent charge rises to more than 1×10⁸ p C/s,and when no insulation breakdown discharge occurs,the average apparent charge is obviously less than this value,and only when the discharge per unit time exceeds this value,free gas will be produced.At the same time,whether there is an insulating paperboard in the discharge model,H₂ and C₂H₂ in the free gas generated by the discharge experiment are always the main components.These experimental conclusions put forward in this thesis have certain reference significance for the improvement of transformer gas protection and the understanding of the change law of free gas composition under transformer internal discharge fault.