The Analysis On Thermal Field In Winding Of Large Oil-immersed Forced-air Cooling Power Transformer

Oil-immersed power transformers as one important equipment in powersystem, with the increasing capacity of power transformer, which led to thethermal issues become more prominent. Winding as the major loss source intransformer, if local overheating will cause the insulation aged，thereby it willaffecting the life of insulating materials. Therefore, to study the temperaturedistribution of the transformer winding, reduce the temperature rise has greatsignificance.This paper based on fundamentals of electromagnetic theory, establish two-dimensional solution mode accord to the actual transformer winding structure,using field-circuit coupled method for getting the distribution of leakagemagnetic at the high voltage-medium voltage winding working at rated situation,and analysis the distribution of magnetic flux leakage at the inside and outside ofthe winding. And then, get the magnetic flux leakage in the radial and axial ineach unit. Derived the formula of the eddy current loss, calculate the eddycurrent loss for each winding unit, and get the sum eddy current loss and thedistribution in winding，and analyzed the factors of eddy current.On the basis of obtained the winding sum loss, according to heat transferand fluid mechanics theory, then use the field-structure interaction method tocalculate and analysis the temperature field and fluid field of winding, the majorresearch is about the affect of different number of oil board and different locationof oil board, and also consider the affect of vertical oil channel at the middle ofradial winding for the temperature field of winding. Analysis affect of oilchannel structure for oil flow.Finally, use the thermoelectricity analogy method, derived the formulawhich consider the oil viscosity changes with the top oil temperature and the formula which consider the load loss varies with temperature. Obtained thetransformer winding hot spot temperature at the condition of which the variousload with time changes, and compared with the methods which discussed in theIEEE C57.91, prove the accuracy of the proposed method, it has a certainreference value in forecast winding hot spot temperature rise.