Compensating the effect of residual dipole energy on dielectric response for effective diagnosis of power transformer insulation

Abstract

Analysis of relaxation current is a widely accepted method for diagnosis of power transformer insulation. The accuracy of such diagnostic tool is dependent on insulation model parameters which are formulated using relaxation current. This implies that the accuracy and hence the reliability of existing insulation diagnosis methods indirectly depends on the accuracy of the recorded polarisation depolarisation current. Sometimes during field measurement relaxation current measurement equipment fails to record proper current, even after application of dc charging voltage. As per utilities, this primarily happens due to improper/loose connections (this cannot be avoided entirely due to the involvement of human factors) and such situation is usually followed by checking and rectifying improper connection. The analysis presented in this study shows that the polarisation current recorded immediately after rectifying the correction is inaccurate and leads to the erroneous diagnosis. Furthermore, it is observed that in these cases, the measured and calculated (using insulation model) values of performance parameters like dissipation factor, polarisation index, and paper-moisture differ by a large extent. This work is aimed at removing the effect of this residual dipole energy introduced during the improper connection phase.