Abstract :
The influence of positive and negative d.c. corona products and humidity on the breakdown probability under an impulse voltage of a 45 cm rod-plate gap in air contaminated by the breakdown products was examined using the electrical charge and current oscillograms of the predischarge phase. A positive d.c. corona discharge produced near the gap reduces the breakdown probability. This may be because of the larger corona streamers from the point which, in this case, appears at the very front of the applied surge. However, a negative d.c. corona discharge increases the breakdown probability of the gap by reducing the magnitude of the corona streamer. The same effect was observed for both polarities of the surge voltage applied to the gap. In the absence of d.c. corona discharge, and at voltages giving a low breakdown percentage, the products of the first breakdowns in one series of measurements appear to increase the electric strength of the gap as they increase the magnitude of the initial corona streamer. At higher concentrations of the breakdown products, the electric strength of the gap decreases again, but it always remains greater than the initial. Thus, when the breakdown probability is determined by a small number of trials, the determined quantitative influence of other factors, such as humidity, on the breakdown probability, may be erroneous. With a negative rod, the influence of humidity seems to be almost negligible. However, with positive rod and at high breakdown product concentrations, the breakdown probability appears to be considerably influenced by the humidity. Measurements performed in comparatively small containers to determine the quantitative influence of humidity on the breakdown probability of the air are influenced by the breakdown products and increased by the humidity to greater than the true values.
