Vernon Cooray

Engineering return stroke models can be categorized either as current generation (traveling current source type) models or current propagation (transmission line type) models. The current generation models are described among other parameters by a corona current distributed along the channel. Recent studies show that there is equivalence between the models of current generation and current propagation types. Due to this equivalence, any engineering return stroke model of current propagation type can be described in terms of an equivalent corona current per unit channel length. The measurements conducted within 10-500 m from triggered lightning flashes show that the electric field of subsequent return strokes at these distances flattens within 15 mus or so. In this paper, the constraints imposed by this feature on the temporal and spatial variation of the equivalent corona current are investigated. The results show that in order for the close fields to flatten within 15 mus or so, the equivalent corona current, should be bipolar and the corona current wave shape at late times should be identical to that of the longitudinal current time derivative. This is in contrast to most of the engineering models of current generation type, in which the corona current is assumed to be unipolar.

A measurement of the positive streamer charge has been performed at atmospheric conditions in a quasi-uniform electric field where the discharge was electrically triggered. By applying a square impulse voltage to the trigger electrode, which was close to, but electrically isolated from, the anode, a positive streamer discharge with one discharge event was produced. The voltage, current and luminosity associated with the streamer discharge were measured simultaneously. By placing a photographic film at the cathode, the number of individual streamers hitting the cathode was estimated within one discharge event. The amount of net charges in streamer discharges has been evaluated for different background electric fields. From the results, it has been deduced that each individual streamer channel contains a charge represented by an excess of 1.1-2.0×10¹¹ positive ions m^-1. The lower value corresponds to a background field of 400 kV m^-1 and the higher one to 600 kV m^-1. The streamer has also been simulated using a simplified streamer model. The results of the measurement and the calculation are compared and discussed.

In most of the studies dealing with electromagnetic field interaction between nearby lightning flashes and electrical installations, the cloud flashes are assumed to play an insignificant role. However, our studies show that cloud flashes can generate induced voltages with amplitudes comparable to those of ground flashes when the electrical installation has a limited size or length and the lightning flash is a few tens of kilometers away. Furthermore, the voltages generated by cloud flashes have much shorter risetimes and pulse widths than those of ground flashes. This paper presents the results of a study conducted to investigate the coupling mechanism between lightning electromagnetic fields from both cloud and ground flashes and an artificial low-voltage power installation (LVPI). In the experiment, a Y-shaped artificial LVPI was constructed and exposed to lightning-electromagnetic fields. In the experiment, both the vertical electric field and the induced voltage in the LVPI were measured simultaneously. The analysis was conducted in the time domain by using the Alternative Transient Program-Electromagnetic Transient Program (ATP-EMTP) and the horizontal electric field at the installation was calculated by applying the concept of wavetilt to the measured vertical electric field. A reasonable agreement between the measured and simulated induced voltages was found. However, the results given in this study might not be true for longer structures such as power distribution lines and close lightning flashes.