The effect of transient voltages on dielectrics — III (An investigation with very high lightning voltages)

Abstract

During a thunder storm lightning voltages that reach the transmission line appear across insulators, transformers and other apparatus at the extremely rapid rate of millions of volts per second. With this rapid rate of application the voltage may reach a very high value in a microsecond (millionth of a second). Hence, since there is always a delay or lag in the breakdown of insulation, quite peculiar effects result from these voltages. For instance, some remarkable phenomena that take place are: Much higher lightning voltages are usually required to jump a given distance than voltages at normal operating frequency; conductors at normal frequency voltages are often good insulators for lightning voltages; water may be punctured like oil; the wet and dry spark-over voltage of insulators are equal; the lightning discharge has a decidedly explosive effect, etc. In addition to the characteristics just mentioned, a study has also been made of the change in voltage and shape of a lightning wave as it travels over a transmission line at the velocity of light. In order that a laboratory study may be of a practical as well as a theoretical interest, it is necessary to be able to reproduce lightning voltages in the laboratory on a large scale and of known characteristics. This investigation was started some years ago with a 200-kv. generator. The generator has been added to from time to time until now, 2,000,000 volts are available and single lightning strokes can be obtained that increase at the rate of 50 million million volts per second. The power is of the order of millions of kilowatts. It is believed that this generator closely approximates voltage and other conditions that usually occur on transmission lines. The lightning voltages used in this investigation were far in excess of any heretofore produced in a laboratory. This impulse generator discharge must not be confused with that produced by an oscillator. The lightning generator, unlike the ocillator, discharges with a loud sharp report or crack. The photographic study shows the lightning spark-over of insulator strings that are of such a length as rarely to spark-over in practise even in bad lightning country. The photographs of the flashes show all the characteristics of lightning such as a zig-zag path, side flashes, etc. The study of the travel of the lightning wave on transmission lines is of interest. It indicates for instance, a certain protective effect of corona and shows that under certain conditions inductance coils may increase the lightning voltage four fold. Certain phases of the ground wire have also been studied.