Emanuel Petrache
Lightning-induced voltages in power and communication systems are nowadays one of the main causes of power quality and electromagnetic compatibility. In recent years, due to the increasing demand by customers for good quality in the power supply along with the widespread use of sensitive devices connected to distribution lines, the protection against lightning-induced disturbances became of primary importance. As a consequence, for a correct protection and insulation coordination, the accurate estimation of lightning-induced overvoltages is essential. In addition, sensitive electronic components used in power and communication systems may suffer logic upset or damage at significantly lower levels of induced electromagnetic interferences. As a result, the evaluation of lightning-induced disturbances in underground cables has recently gained more interest compared to the past. The major aim of the present thesis is the development of models and computer codes, allowing the computation of the voltages and currents induced by an external transient electromagnetic excitation, especially due to lightning discharge, along realistic -and hence complex- transmission line networks and in buried cables. After a brief introduction on lightning phenomenology, Chapter 2 presents an overview of lightning return stroke modeling and methods for the calculations of lightning-generated electromagnetic fields above and inside the ground. We show in this chapter, that the simplified expression recently proposed by Cooray for the calculation of electric field penetrating the ground yields very good approximations to the exact numerical solutions, for distances as close as 100 m. The main original contributions of this thesis are presented in Chapters 3 through 5. They consist of theoretical and experimental work as follows. In Chapter 3, based on the Agrawal et al. coupling model, extended to the case of a multiconductor line and interfaced with EMTP program, a new software tool called LIOV-EMTP96, developed in the framework of a collaboration with the University of Bologna, is described. This software tool is able to analyze the response of complex distribution systems to indirect lightning. We carried out several experimental campaigns to test and validate the LIOV-EMTP96 code. First of all, we used the NEMP simulator of the Swiss Federal Institute of Technology (SEMIRAMIS) to illuminate a reduced-scale model of a multiconductor line. Then, we used a more complex 27-line reduced-scale network illuminated by the VERIFY NEMP simulator belonging to the Swiss Defence Procurement Agency (Spiez). And, finally, we used experimental data obtained by means of artificially-initiated lightning in 2002 and 2003 on an experimental distribution line at the International Center for Lightning Research and Testing, Camp Blanding, Florida. Calculations performed with the developed program have been tested versus the obtained experimental results, and a very good agreement between the simulations and the measurements is found. It is concluded that LIOV-EMTP96 is adequate for the analysis of the response of complex distribution systems to indirect lightning and for their protection in view of optimal insulation coordination and power quality achievement. Chapter 4 presents efficient calculation methods to estimate lightning-induced disturbances in buried cables in both time- and frequency-domain. Concerning the parameters of the underground cable, different expressions for the ground impedance are analyzed and compared. We propose a logarithmic approximation for the ground impedance of a buried cable that is shown to be very accurate. In addition, unlike most of the considered approximations, the proposed expression has an asymptotic behavior at high frequencies. As a result, the corresponding transient ground resistance in time-domain has no singularity at t = 0, and therefore, does not require any special treatment in a direct time domain solution. The time-domain solution of electromagnetic field-to-buried cable coupling equations is also investigated using the point-centered FDTD (Finite Difference Time Domain) method. The coupling model includes the effect of ground admittance, which appears in terms of an additional convolution integral. An analytical expression for the ground transient resistance in the time-domain is also proposed which is shown to be accurate and non-singular. For the case of coaxial buried cables, the coupling with the inner conductor is also investigated using a frequency domain approach based on Green's functions. Chapter 4 provides, in addition, recommendations concerning the modeling and computation of lightning-induced disturbances on buried cable. Chapter 5 presents extensive experimental results obtained at the International Center for Lightning Research and Testing (ICLRT) at Camp Blanding, Florida, in 2002 and 2003. We measured currents induced by triggered and natural lightning events at the ends of a shielded buried cable, both in the cable shield and in the inner conductor. The horizontal magnetic field was also measured. Additionally, a close natural lightning event has also been recorded during the 2003 summer campaign. For the natural lightning events recorded in 2002, the waveforms are correlated with the data obtained by the U.S. National Lightning Detection Network (NLDN). The obtained experimental data were used to test the theoretical models and the developed computer codes for the analysis of lightning-induced disturbances in buried cables. In general, reasonably good agreement was found between numerical simulations and experimentally-recorded waveforms. It was also found that the ground conductivity affects in a significant way lightning-induced currents in buried cables. Therefore, an accurate representation of the ground electrical parameters is a key issue in the determination of such disturbances.
EPFL2004
