Extracting Frequency Spectroscopy of Oil-Immersed Paper Based on Havriliak–Negami Model Without Known Insulation Structure of Transformer

Abstract

When using the Frequency Dielectric Spectroscopy (FDS) method to evaluate the insulation condition of a transformer, it is essential to calculate the frequency spectroscopy of the oil-immersed paper from the major insulation of transformer with the structural parameters. However, in on-site measurement, the structure of the transformer is often unknown, which greatly limits the application of the FDS method. This paper proposed a method for extracting the frequency spectroscopy of oil-immersed paper from the oil-paper insulation without the known insulation structure. Firstly, the frequency spectroscopy of the oil-immersed paper with different moisture content, aging states, and temperatures was measured and analyzed. The results showed that the Havriliak-Negami (H-N) model with hopping conductivity is the most suitable frequency spectroscopy model for the oil-immersed paper in different conditions. Secondly, the transformer insulation model was constructed, and the frequency spectroscopy of the corresponding oil-immersed paper was calculated. The calculation showed that the error of the structural parameters influences the calculated frequency spectroscopy of oil-immersed paper, especially in the low moisture condition, which leads to the increase of the fitting residual of the H-N model. According to the smallest fitting residual, the structural parameters can be analyzed, and the frequency spectroscopy of oil-immersed paper can be extracted accurately from the oil-paper insulation without the known insulation structure.