Parameters Extraction and Mechanism Analysis of the Moisture in Oil–Paper Based on the Spectral Decomposition of Dielectric Response

Frequency-domain dielectric spectroscopy (FDS) is widely used in the insulation assessment of oil-immersed power equipment, and the moisture in insulating paper could be qualitatively predicted by mathematical approximation. However, due to the still lots unexplored underlying mechanisms, until now, there is still no convincing and physical method for the diagnosis of moisture content. With the aid of spectral decomposition methods, this article aims to explore the effects and mechanism of moisture on the independent dielectric response spectra. First, the measured dielectric response curve is unfolded and reconstructed by derivation of the real part curve. Second, the independent polarization and conductance spectra of damp samples are extracted, and then, the relationships between moisture content in insulating paper and spectral parameters are established quantitatively. The results show that moisture introduces a new polarization in the midfrequency region at room temperature, and both polarization and conductance increase with the moisture content. The multiple adsorption model and the proton exchange model of moisture are employed to illustrate the changes of the polarization and the conductance spectra, respectively. This research provides a methodological reference for the moisture assessment of oil–paper insulation.