| Summary: | Conventional Current Transformers (CTs) provide the input signal required by protection relays, control equipment and energy metering in power networks. Catastrophic failures of CTs may lead to injury of personnel within the substation, interruption of power supply and damage to adjacent high voltage equipment. One of the causes of CT failures is high values of insulation power factor. Research indicates that with increasing primary insulation power factor values, partial discharges develop between the paper insulation leading to its exponential decay with the end result of an ultimate CT failure. Reports in Eskom Distribution indicated that a number of CTs from one manufacturer were exceeding the specified insulation power factor value. This research was initiated to investigate the impact of high power factor on the premature failure of CTs. This study serves to analyse the significance of power factor on paper-oil insulation within a hair-pin type CT. The internal primary insulation of a hair-pin type CT used in Eskom Distribution is reviewed in terms of its design, construction and relevant tests. Sample CTs rated at 132kV and manufactured in the year 2007 to 2009 from the specific manufacturer were selected for insulation power factor testing. The Doble M4100 diagnostic test system was used to perform the testing that also assisted in providing a comparison between the units with high insulation power factor values and that which were within the specified limits. The results show that the high values of insulation power factor give a direct indication of the dielectric losses (I2R) within that CT, which inadvertently indicates the shortened serviceable life of that CT. High moisture content within the primary insulation, low quality insulating oil and inadequate quality assurance were identified as some of the contributory factors in the CTs non compliance. The effects of high values of insulation power factor are the primary factor for continuous on-line condition monitoring techniques that enable data trending and provide for early warning of an imminent CT failure. The testing of the sample CTs provided a more dynamic approach for recommendations to prevent the installation of such units into the power network. === Thesis (M.Sc.)-University of KwaZulu-Natal, Durban, 2010.
|
|---|
