Instrumentation of the Säntis Tower for Lightning Current Measurement

In this thesis, a new experimental station was designed and built for the measurement of lightning currents. The station (Säntis Tower) was instrumented using advanced and modern equipment with remote monitoring for an accurate measurement of lightning current parameters. In Chapter 2, we present the phenomenology and classification of lightning discharges. An overview of different types of discharges is given and their typical signatures in terms of current waveform were described. Lightning current measurements obtained using different techniques are also summarized. It is emphasized that, despite the important effort in obtaining experimental data on lightning currents using different techniques, the number of available data is still limited and more data are needed to better understand the physics of lightning and to better characterize the lightning current parameters associated with different types of cloud-to-ground lightning discharges. Chapter 3 presents the characteristics of a system suitably developed to measure lightning current waveforms on the Säntis Tower in Switzerland. The system was designed considering the EMC constraints, using fiber optics as backbone for transmission of the measured signals and paying special attention to the design of the cabling, measurement boxes, protection systems and shielding. In addition, due to the harsh conditions inside the tower, where the temperature ranges from -15°C to +35°C, a system comprising a heater, a ventilator, moisture exhaust holes, and a thermal insulation material was designed to keep the temperature within acceptable limits for the components and to avoid condensation inside the PCB cards. In order to overcome the limited high frequency response of the Rogowski coils, we propose to combine them with an improved design of a B-dot sensor which does not have the shortcomings of conventional loops. The designed sensor is characterized by an upper frequency cutoff of 20 MHz and a 50 Ohm matched termination. Laboratory tests carried out in the high voltage laboratory of the EPFL show the effectiveness of the joint use of Rogowski coils and B-Dot sensors for the measurement of lightning currents. The measurement system allows an over-the-Internet remote maintenance, monitoring and control overall system. In particular, the status of each pair of sensors is monitored and controlled by means of a system designed, and built, using National Instruments Compact- RIO modules linked via 100Base-FX Ethernet, which uses fiber optics as a transmission medium. A data analysis software (SENDIS) was developed to analyze the obtained data, to extract statistical parameters and to allow the remote monitoring, control and programming of the different components. Algorithms implemented in SENDIS allow the detection of current components (return stroke, M-component, continuous and continuing currents) within each flash and the automatic extraction of statistical parameters of the current. In Chapter 4, we present statistical distributions of the lightning current parameters based on the lightning current and current-derivative waveforms measured at the Säntis Tower site in 2010 and 2011. The total number of flashes analyzed in this study was 167 which include nearly 2000 pulses, constituting the largest worldwide dataset. The statistical distributions are associated with upward negative flashes. The obtained data reveal that many flashes are characterized by a large amount of charge transfers (in excess of 50 C). The median value of the flash multiplicity or the number of pulses per flash was found to be 8, a value much larger than those associated with downward flashes. Interestingly, 6 flashes exhibited a number of pulses in excess of 40. Chapter 5 presents an analysis of measured current waveforms associated with positive and bipolar flashes recorded on the Säntis tower from May 2010 till January 2012. The overall number of recorded positive flashes in the considered period was 30, while 3 flashes were classified as bipolar. The percentage of positive flashes (15%) was found to be considerably larger than the values observed in other studies in summer months (3% to 6.5%). The time- derivatives of the current pulses associated with upward stepped leaders are found to be much larger than those of the main pulse. Our recorded data constitute the first directly-measured evidence of M-components of both polarities during a continuing current lowering positive charge to ground. The observed positive flashes are characterized by a median peak current of 11.8 kA, and a median flash duration of 80 ms. These values are consistent with those associated with the data recorded at the Gaisberg Tower in Austria. On the other hand, the amount of transferred charge was substantially larger in our dataset, with a median value of 169 C (6 times as large as the values obtained in Monte San Salvatore and in Japan, and 3 times as large as the value obtained in Austria). Eight flashes out of 30 transported positive charge to the ground in excess of 500 C. The obtained results confirm also the findings of Saba and co-workers according to which positive lightning flashes may combine high peak currents with high charge transfers (or flash durations). In the same chapter, we analyze three bipolar flashes recorded during the considered time period which occurred during one storm on August 27, 2011. In Chapter 6, we present an evaluation of the performance of the European Lightning Location Network (EUCLID) using the obtained data on lightning current measured on the Säntis Tower. The flash detection efficiency was estimated to be 93% and the median location error was 123 m. The EUCLID peak current estimates were on average larger than the measured current with a median peak current estimation error of about 60% for strokes over 10 kA. Measurements included five typical positive flashes successfully detected by EUCLID. The location errors for the positive events ranged from 1 to 3 km, with a median of 960 m. Finally, in the Appendix, we present a thorough analysis of the noise source and characteristics affecting the measured signals at Säntis. We also describe appropriate signal processing methods, which are adopted and implemented to post-process the measured signals. The Appendix contains also a description of the data analysis software which was developed in the framework of this thesis to analyze the obtained data and to extract statistical parameters.