Modelling of Acoustic Wave Propagation Due to Partial Discharge and Its Detection and Localization in an Oil-Filled Distribution Transformer

Abstract Partial discharge (PD) is the main cause of the insulation decay and hence periodical testing of the insulation condition of a distribution transformer is necessary. This paper presents a model of PD acoustic wave propagation, detection, and localization in an oil-filled distribution transformer using finite element method supported by COMSOL Multiphysics software. Using an acoustic module and AC/DC module of COMSOL Multiphysics software, oil filled distribution transformer, and the acoustic piezoelectric sensor are simulated to analyze and detect the PD inside the transformer. PD is numerically simulated in the transformer windings, core, and oil ducts that produce acoustic wave signal. The distribution of the acoustic pressure wave inside the model transformer is analyzed first. Next, the acoustic piezoelectric sensors are modelled at four different locations of the model transformer to detect the pressure acoustic wave signal induced due to PD in the transformer. Finally, using an artificial neural network (ANN), the localization, and identification of PD in various parts of the transformer have been analyzed. The results obtained for location and detection are quite encouraging.