Optimal Time Selection for the Polarisation and Depolarisation Current Measurement for Power Transformer Insulation Diagnosis

Non-destructive dielectrometry measurement techniques such as polarisation and depolarisation current (PDC) measurement is gaining popularity in power industry to evaluate the insulation condition of transformers. Due to nature of the polarisation species involved in transformer insulation, a perfect PDC measurement would take a very long time, to allow all the interfacial polarisation species to be relaxed. Test duration of 10,000s charging and 10,000s discharging has been considered a reasonable compromise in the context of power transformer insulation. However, in a practical field situation power utilities usually cannot afford such a long measurement time due to economical factors. The authors of this paper explore the possibility of further reducing the PDC measurement time to meet the preference of power industries. Since polarisation and depolarisation are reverse process to each other; it would be possible to estimate oil and paper conductivity from short duration measurements. An experimental investigation has been performed on a distribution transformer and on a model transformer to estimate and compare the accuracy of conductivity computation with different charging and discharging times. These test data helps to decide on the optimal PDC test duration. To help extrapolation of PDC data from short duration to longer duration (i.e. 10,000s), both the current difference and charge difference methods have been adopted. The results from these methods have been critically analysed in this paper.