Dielectric Mechanism and Spectrum Prediction Model for Insulating Paper Based on 3-D Frequency-Domain Spectroscopy

Moisture is a serious concern for transformers, as it will degrade insulation properties and accelerate aging. The dielectric mechanisms of oil-immersed paper (OIP) and the influence of moisture still remains contentious in existing studies, which prevents the application of Frequency-Domain Spectroscopy (FDS) technology in moisture assessment of transformers. OIP is a biphasic composite material composed of oil and paper, and the existence of oil would complicate the understanding of the dielectric mechanisms in solid insulation. Therefore, this study focuses on investigating the dielectric properties of non-immersed insulating paper using FDS. Firstly, the spectra of paper with different moisture content were analyzed, and the relationship between dielectric parameters and moisture was quantified through the proposed dielectric model of insulating paper. Then, a three-dimensional dielectric model for spectrum prediction was established, and the feasibility and universality of model was verified. Results show three distinct dielectric response processes in insulating papers: the hopping conduction (RI), the low-frequency relaxation polarization (RII), and the high-frequency relaxation polarization (RIII). With increasing moisture content, RI and RII processes shift synchronously and exponentially towards higher frequencies with the dielectric curve shape maintained, while RIII process is unaffected. Moreover, the proposed spectrum prediction model can accurately predict the spectrum of insulating paper with varying moisture content. The dielectric response mechanism and three-dimensional dielectric model of insulating paper identified in this study provide the theoretical basis and data support for subsequent studies of the dielectric mechanism and moisture assessment of oil-immersed paper.