Condition Assessment of Dielectric Liquid Used in Large Power Equipment Using Time-Varying Model

This paper presents a non-invasive approach for assessing the condition of dielectric liquids, particularly mineral oils, used in power transformers by employing a Time-Varying Model (TVM) to analyze polarization current profiles. Ageing in dielectric liquids can compromise transformer reliability. Hence, effective monitoring is essential. In this study, mineral oil samples are subjected to accelerated ageing, with polarization current measurements analyzed via TVM. Performance parameters, including conductivity, dissipation factor, and relative permittivity, are evaluated using variance analysis to reveal more distinct ageing trends compared to traditional standard deviation methods. Findings suggest that variance analysis provides a more accurate depiction of oil degradation, supporting condition-based maintenance and reducing transformer failure risks.