Hump Phenomenon-Based Aging Estimation of Liquid Insulation Used in HV Transformer

In this study, aging impact analysis on dielectric behaviour of mineral oil, natural and synthetic ester is performed and their aging status is estimated using the hump phenomenon obtained from dielectric modulus spectrum. For this purpose, equivalent aging of 5, 10, 15 and 20 years of each oil are emulated by mixing formic, acetic, and levulinic acid with mineral and ester oils. Thereafter, Fourier Transform Infrared Spectroscopy and Frequency Domain Spectroscopy (FDS) are conducted on the prepared aged samples to assess the impact of aging on physicochemical alteration and low-frequency interfacial polarization characteristics at different temperatures, respectively. Based on the FDS results, dielectric modulus spectrum is obtained from which hump phenomenon is identified. Besides, another important factor i.e. conduction dominance factor (CDF) is introduced to quantitatively investigate the dominance of conduction mechanism over relaxation polarization. Based on the experimental results, a correlation with the aging status of liquid insulation with the three aging sensitive parameters (hump frequency, hump peak and CDF) is established which can reliably be used for estimation of their aging status.