A comparison of Return voltage measurement and frequency domain spectroscopy test on high voltage insulation

Return voltage measurement (RVM) and low frequency domain spectroscopy (FDS) have been successfully used to assess moisture content of oil/paper insulation. This paper is focused on modeling the dielectric response of transformer insulation with RVM data and FDS data respectively to get a relationship between these two methods and to facilitate a better interpretation of test results. The model is based on linear dielectric response theory. A nonlinear least squares optimization algorithm is used to estimate the parameters in the model. Experiments have been performed on two transformers in the Pacific Power Advanced Technology Center (ATC). When the transformer is relatively "dry" good agreement was obtained between these two methods, but some difference was found when moisture content of oil/paper insulation in transformer increases.