Non-destructive and in-situ analysis of insulating materials in high-voltage power transformers

Abstract

Rapid in-situ analysis of the condition of power transformer winding insulation kraft paper, to provide information on its aged condition, is an imperative in the power industry. The ability to achieve this with portable instrumentation would be very beneficial. The insulation paper degrades over time, and the degree of polymerisation (DP) of the paper affects its structural integrity. At low DP levels the risk of electrical breakdown increases. Currently used methods of estimating DP are only approximate or destructive. The use of spectroscopy together with multivariate statistical analysis (MVSA) provides a powerful non-destructive evaluation of the condition of paper. From initial feasibility studies, we have developed a simple, portable system (TRANSPEC) using fibre-optics and broad-band spectroscopy that can measure the degree of polymerisation of various aged transformer papers to a precision of approximately 50 DP units with a spatial resolution of 25 mm/sup 2/, and also measure the chemical composition and condition of the insulating mineral oil. MVSA regression models were constructed from library spectral data, and these models are used to predict the DP of other papers with parameters that fall within the range spanned by the set of calibration samples. Separating oil and moisture information from wetted paper is possible and water content can be determined to an accuracy of approximately 0.1%. Very low to high concentrations of water have been determined using this method. We have also confirmed that both the absorption rate and water capacity of the paper are strongly affected by the degree of degradation. With a single TRANSPEC system, in-situ analysis of the degree of polymerisation of insulating paper, the water content and, if required, a chemical analysis of the insulating oil and its water content is possible providing a rapid cost-effective method for transformer insulation condition assessment and monitoring, which correlates well with current established methods.