Investigation on digital algorithm for on-line monitoring and diagnostics of metal oxide surge arrester based on an accurate model

The digital algorithms presently available for on-line monitoring of metal oxide surge arrester (MOA) are normally based on a simplified equivalent model consisting of a capacitor and a nonlinear resistor in parallel, which however in applications poses some deficiency compared with experimental results, leading to a certain phase discrepancy between the line voltage and the resistive branch current, due to a capacitive effect in the branch. This paper develops a new digital algorithm based on an accurate model of metal oxide surge arrester, in which base-frequency approximating method and iterating scheme are combined in use. Computer simulations are carried out to validate the new approach, showing the digital algorithm gives an application accuracy of less than 1% even with harmonised power source and different levels of background noises while within 10% of the measured sum leakage current.