Influence of charging voltage magnitude on time domain dielectric response of oil–paper insulation

• Published in: IET science, measurement & technology Vol. 13; no. 6; pp. 874 - 882
• Main Authors: Banerjee, Chandra Madhab, Dutta, Saurabh, Baral, Arijit, Chakravorti, Sivaji
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 01.08.2019
• Subjects: charging voltage; different charging voltages; insulation conductivity; insulation response; method performance parameters; neural nets; nonlinearity; oil–paper insulation; performance parameter calculation; polarisation; polarisation current profile; polarisation–depolarisation current; power transformer insulation; Research Article; space charge; specialised equipment; time domain dielectric response; trained artificial neural network; transformer oil; voltage magnitude; performance parameter calculation; voltage magnitude; polarisation–depolarisation current; nonlinearity; polarisation; insulation conductivity; method performance parameters; time domain dielectric response; charging voltage; insulation response; polarisation current profile; trained artificial neural network; specialised equipment; oil–paper insulation; paper; space charge; transformer oil; different charging voltages; power transformer insulation; neural nets
• ISSN: 1751-8822; 1751-8830; 1751-8830
• DOI: 10.1049/iet-smt.2018.5276

Polarisation current profile varies non-linearly with a magnitude of charging voltage. This implies performance parameters obtained from polarisation–depolarisation current and those measured using specialised equipment cannot always be compared. Results reported in this study suggested that the non-linearity in insulation response can be explained in terms of the interaction between space charge and insulation conductivity. The influence of using different charging voltages and its effect on performance parameter calculation is investigated in this study with the help of laboratory samples. It is shown in this study that with the help of a trained artificial neural network, it is possible to identify the charging voltage at which insulation response exhibits minimum non-linearity. To test the accuracy of the proposed method performance parameters measured using commercial equipment are compared with those calculated from the polarisation current profile.