Approaches for Numerical Simulation of Partial Discharges

Abstract

Abstract-Insulation degradation is the primary reason for aging and eventual failure of electrical equipment. Any measurement that forewarns an impending breakdown of the equipment can be a boon to asset managers. Partial discharge (PD) is one such phenomenon that can be monitored for assessing the quality of insulation. However, the phenomenon of PD is quite intricate and requires an understanding of various concurrent processes. In this paper, some of the methods used to simulate this complex event have been reviewed. At the fundamental level, PD is a localized breakdown that occurs without complete bridging of the insulation. These are feeble and extremely fast, nanosecond discharges. They can be thought of as motion of charges under the influence of an external electric field, which has exceeded the breakdown strength in a small region of the insulation. When such discharges occur in voids in solid insulation, they generate signals in the electrical, chemical and acoustic domains. Also, PD causes physical damage due to bombardment of these charged particles on the walls of the void. The charges are generated by ionization of the gas filled in the void. Once ionized, the motion of these particles is governed not only by the external field, but also by the space charge so created. There are various methods that can be used to simulate the motion of these charges. Of these, particle-in-cell (PIC) method can be applied in a Lagrangian frame of reference, where the particles retain their identity, and their trajectories are monitored by calculating the force on them. This force is determined by the field at their location. This method is fairly intuitive. The other class, Eulerian, treats the charges as a density distribution and then determines its motion. In this, the flux-corrected transport (FCT) scheme basically considers the motion as governed by the charge continuity equations coupled with the Poisson's equation. The continuity equations also account for the generation of the charges, due to ionization, and their possible subsequent recombination or attachment. However, these events, as recorded, are also statistical in nature, and the effect of previous discharges on the subsequent ones can hardly be ignored. The method of including these in the simulation has also been reviewed.