The Pattern Recognition Of Partial Discharges In Transformer Based On DGA Energy State Analysis

As one of the most pivotal facilities in electrical power system, power transformer has played a crucial part in the secure and reliable operation of power grid. It is universally believed that discharge defects are the primary reasons giving rise to the insulation failure in power transformers. Therefore, prompt discharge-defect diagnosis and correct discharge-pattern recognition are of utmost importance to the stability of the power system. The severity of insulation damage caused by the partial discharge defects inside transformers is strongly related to the discharge energy. Currently, though dissolved gas analysis (DGA), the most widespread-used and reliable method for years of incipient defect diagnosis in oil-filled transformers, could distinguish the types of the incipient defects, it does not involve the information of discharge energy and development pattern of various defects. For this purpose, this article established a partial discharge pattern recognition of different discharge defects though energy analysis on the basis of DGA method. Firstly, we calculated formation energy for each DGA gas based on enthalpy theory; secondly, the correlation between the formation energy of DGA gases and the discharge energy was terrified through multi-model discharge experiments based on the practical operation condition of transformers; further, a partial discharge pattern recognition was proposed to identify the different types of transformer discharge defects. The specific research works are as follows:1.Based on enthalpy change of thermodynamics theory, this article investigated the dissolved gases in the perspective of formation energy and constructed the gas-generating model according to the material composition of oil-paper insulation. By employing enthalpy change analysis on this model, we proposed the calculation method to obtain the formation energy of DGA key gases.2.Combined with the insulation structure and faulty location of transformers, the experimental platform of typical discharge models was set up, including ball-plate, column-plate, and needle-plate, in order to study the discharge development tendency in different discharge models. The correlation analysis was studied between the discharge energy, traditional electrical characteristics, DGA gas-formation energy and combustible gases content of different discharge models in different time period. It turned out that DGA gas-formation energy showed good correlation with discharge energy and the discharge development was divided into three stages as prophase, metaphase and anaphase according to the DGA gas-formation energy.3.Based on the development discipline of DGA gas-formation energy, this article analyzed the discharge development tendency of three typical discharge patterns. The results indicated that the DGA gas-formation energy of the column-plate discharge pattern takes the S-type on the growth trend while that of ball-plate and needle-plate shared the same growth trend. Furthermore, by proposing the energy density ratio R= c(C2H2)/c(CO), we discovered that the R value would keep a sustainable growth in needle-plate discharge pattern while the R value of needle-plate discharge pattern would increase in the prophase, then stagnate in the metaphase and increase again in the anaphase.