Analysis of actual multiconductor transmission lines: effects of non-uniformities and nonlinearities

The analysis of transients in electrical networks has been a topic largely investigated in the literature for a long time; however, researchers are still focusing on this topic to progressively improve the available models of devices operating in the electrical system, and to increase the network reliability in case of undesired events (e.g., faults, direct or indirect lightning, etc.) or switching operations. Adopting a systemic approach, at first different aspects involving transmission line modelling are analysed; the most relevant ones are identified, and chosen as the main focus of the presented research activity. This work aims at analysing transient events occurring along overhead multiconductor transmission lines, evaluating the impact of non-uniformities and nonlinearities, as lightning events, protective devices (surge arresters), and corona effect, on the magnitude of the predicted overvoltages. Although electromagnetic transients programs are available to perform these studies, equivalent circuits and models of electrical devices are frequently embedded and not accessible to the final user. To this aim, an implicit Finite-Difference Time-Domain algorithm with second order accuracy has been implemented in Fortran for the solution of the coupled telegrapher’s equations. The main code has been integrated with additional routines, which have been developed to account for the nonlinear behaviour of surge arresters (modelled through lumped equivalent circuits), for the coupling with external electromagnetic fields produced by return stroke currents, and for the corona effect (as a nonlinear distributed physical phenomenon). Further analysis has been performed in the frequency domain to assess the effect of additional ground wires, installed below the phase conductors, in mitigating the overvoltages across the insulators in lightning studies. In depth considerations are devoted to the validity of the approximation introduced when the tower grounding impedance is modelled by means of a resistive device. Both studies, primarily performed in the frequency domain, are suitable for time domain simulations, when included by means of inverse Fourier transforms and time domain convolutions (as to the transient impedance of grounding systems).