Condition monitoring of biodegradable oil filled transformer

Abstract

Natural and synthetic ester insulating oils have higher fire points and excellent biodegradable characteristics. Therefore, in order to reduce the adverse environmental impact and to improve the fire safety of transformers, there is an increasing demand for natural and synthetic ester insulating liquids as a transformer insulating oil. However, present understanding on ageing behaviour of ester oil-paper composite insulation system and knowledge on application of existing condition monitoring tools for ester based insulation systems are inadequate. This impedes the cost effective and reliable field applications of ester insulating oils, particularly application of natural esters. To reduce this knowledge gap, series of controlled ageing experiments are performed in this research project to provide a better and comprehensive understanding on ageing behaviour of ester oil-paper insulation systems. Furthermore, applicability of existing chemical and electrical based condition monitoring techniques for ester oil-paper insulation systems is systematically investigated in this research project. In this thesis, ageing behaviour of dry pressboard insulation in mineral and three different ester insulating oils under simulated transformer operating environment is investigated. Moreover, the ageing behaviour of natural ester-pressboard composite insulation in moisture rich environment is also compared with that of mineral oil-pressboard system. Degree of polymerisation of pressboard samples measured at different ageing interval is used in this research to determine the ageing condition of pressboard. Moreover, applicability of oil related diagnostic parameters such as concentration of dissolved furanic compounds, acidity value, Dielectric Dissipation Factor (DDF), viscosity and colour to assess the degree of degradation of both ester and mineral insulating oils is thoroughly investigated in this research project. The potential of FTIR (Fourier Transform Infrared spectroscopy) techniques for characterising the degree of degradation of paper insulation is also discussed in this thesis. In addition, this thesis concentrates on characterisation of the charge dynamic in insulating oil through modelling of their dielectric responses. The comparison of gassing behaviour of ester and mineral insulating oils under two different low temperature faults and low energy electric discharge condition is also presented in this thesis. Dissolved Gas Analysis (DGA) results presented in this thesis depicts that faults gases detected in ester and mineral oil samples subjected to a similar fault are akin in type but quantitatively different. The quantity of fault gases produced in two different natural ester oils is also dissimilar. For example, soy-based natural ester produces a large quantity of ethane (C2H6) under low temperature overheating condition than sunflower oil based natural ester. Therefore, this research project investigates the applicability of well-established DGA interpretation schemes namely Duval II | P a g e triangle, IEC gas ratio and IEEE key gas method on ester based insulation systems in order to identify the possible faults based on DGA data. Investigation of impact of moisture, temperature and ageing on dielectric response of ester impregnated pressboard insulation by Frequency Domain Spectroscopy (FDS) and Polarisation Depolarisation Current (PDC) is a major contribution in this thesis. In order to study that, dielectric responses of well-controlled pressboard insulation samples which are impregnated with four different insulating liquids have been critically investigated. These measurements have been performed with two natural esters, one synthetic ester and one type of mineral oil under varying moisture (0.3%-8.7 %) and different temperature levels (35, 55, 75oC). In addition, the effect of ageing on dielectric response of natural ester impregnated pressboard insulation has been compared to that of mineral oil impregnated pressboard insulation. In this thesis, low frequency conductivity based model is proposed to determine the moisture content in transformer solid insulation based on FDS data of unaged pressboard insulation. Furthermore, the effect of ageing and temperature on the proposed model is quantitatively analysed. In order to interpret the FDS data of pressboard insulation in frequency range 10-4-10-3 Hz, a hierarchical equivalent circuit is derived based on Dissado-Hill’s cluster theory. The equivalent circuit model parameters are then used to explain the influence of diverse oil properties and moisture on microscopic level charge transport and polarisation phenomena in oil impregnated pressboard insulation. Furthermore, this research has used activation energy to characterise the temperature dependent dielectric response of pressboard insulation impregnated in mineral and ester insulating oils. The applicability of commercially available FDS based moisture diagnostic tool on ester based insulation system is also systematically investigated in this research project.