Conference Agenda

Dimensioning, testing and maintenance of the lightning protection system of wind turbines can be improved using the local distributions of lightning current parameters instead of those on which lightning protection standards rely. This paper presents a prototype system for measuring lightning currents’ waveforms on wind turbines. The prototype is being developed at the Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia. The system’s fundamental components are two Rogowski coils and their corresponding integrators. One coil is optimized for lower frequencies and amplitudes, while the other is for higher ones. The cRIO real-time controller is used as an acquisition system, with acquisition logic developed using LabVIEW. The controller has one digitizer for each coil and a GPS synchronization module. The system was tested in the High Voltage Laboratory at University. It was recently installed on an actual wind turbine in Croatia, located in an area characterized by high winter lightning activity. Simultaneously with the prototype measurement system, the lightning activity in the wind turbine micro-location will be monitored by three other systems: the lightning location system, additional lightning monitoring sensors installed in the blades of the same wind turbine and a high-speed camera installed at the wind turbine location.

This paper presents a new aspect of shield wire (SW) modeling for assessing the indirect lightning performance of overhead medium-voltage distribution lines: a flashover (FO) between an SW and the reinforcing bars of a distribution pole or a crossarm (SWFO). In general, an SW is periodically grounded approximately every four to ten poles and not grounded at the other poles. However, owing to lightning-induced overvoltages, the voltage difference between the SW and the distribution pole may even exceed 100 kV, and the SWFO can occur. In this paper, we evaluate the effect of the SWFO on the number of FO occurrences of phase-conductor insulators by the Monte Carlo method using a 2D finite-difference time-domain-based indirect lightning surge analysis program. The effect of the SWFO is more significant in lines with high soil resistivity (a soil resistivity of 1000 m was assumed) regardless of the installation of surge arresters: the total number of FO occurrences markedly differs by up to 50% between cases in which the SWFO is considered and not considered. The analysis presented in this paper will assist the formulation of lightning protection measures particularly, in regions with high-resistivity soils.

This paper investigates the effect of the impulse corona inside the cross-arm on the lightning performance of a Y-shaped composite pylon with a cable as an internal down-lead through the hollow cross-arm. First, the electromagnetic transient model for the down-lead system of the composite pylon is built. The simplified steps for calculating the surge impedance of the cable down-lead are given. In addition, the mutual coupling between two down-leads is considered. Through a laboratory test on coaxial cylinders resembling the structure of the cable and cross-arm, the dynamic capacitance of the corona on the surface of the cable is obtained and included in the electromagnetic transient model. Then, the effect of the corona on the traveling waveform and mutual capacitance is discussed. Furthermore, the influences of the ground electrode length and the phase voltage on lightning performance are also studied. Finally, the backflash rates of the composite pylon with cable as down-lead are calculated. The results show that the impulse corona has a limited impact on the critical current, and the composite pylon with cable as down-lead shows a promising lightning performance.

In this work, the authors model the number of lightning flashes in transmission structures with Ground Flash Density, height tower, and elevation above sea level as variables. 333transmission structures were located in the northeast of Colombia, where an influence radius of 150 meters was established for each one. Cloud-to-ground lightning strokes detected by the LINET network between 2014 and 2021 in the studied area were grouped into flashes and counted for each radius. Likewise, each structure was characterized by its height, elevation above sea level, and GFD of the area where it is located. In this way, a model that describes 75.66 % of flashes in transmission structures with a standard deviation of 0.32 flashes was built. It is concluded that there is a linear relationship between the dependent variable and all the independent variables.

The paper focuses on evaluating the stress on distribution class surge arresters (SAs) caused by lightning strikes. It proposes a procedure for estimating the statistical distribution of energy absorbed by SAs due to both indirect and direct lightning strikes, which is a crucial step for assessing the probability of SA failure. Two different SA representations are considered, namely, a static nonlinear resistance and a dynamic, frequency dependent model. After analyzing the overvoltage and current waveforms caused by lightning strikes and considering the effect of flashover occurrence, the paper assesses the effect of several factors on the current and energy absorption, namely the presence of a periodically grounded shield wire, of the grounding resistance value, and of the distance between subsequent SAs. The analysis shows that the static model can be considered accurate enough for evaluating the stress originated by direct and indirect lightnings on distribution class SAs.

The differences on the lightning response of transmission lines, whose towers are subjected to the impression of pulses of current of first negative return strokes and of positive flashes, are determined by computational simulation. It is shown that the amplitude of the resulting transient overvoltages across insulators per unit of impressed current is much larger for negative return strokes. As most of the traditional procedures for assessing lightning performance of transmission lines ignore these differences, their results can exhibit significant errors.