Abstract:
At present, the return stroke speed is generally set to some fixed value between 1×10
8 m/s and 2×10
8 m/s when the nationwide lightning location systems are used to invert Cloud to Ground lightning return stroke currents. In this paper, the impact of adopting this method on the accuracy of the return stroke current inversion is discussed. Firstly, the method of using the lightning location system to invert the return stroke current is introduced, and the important influence of the speed of the return stroke on the peak current is analyzed. Secondly, based on the relationship between the negative return stroke current and the return stroke speed proposed by Lundholm et al., the analysis on the negative first return stroke reveals that the return stroke current below 25 kA may be underestimated, and the return stroke current above 160 kA may be overestimated. For negative subsequent return strokes, the return stroke current below 5 kA may be underestimated, and the return stroke current above 32 kA may be overestimated. Then, the return stroke speed observation results of negative first return strokes, negative subsequent return strokes, positive first return strokes and positive subsequent return strokes are summarized. It is believed that it is necessary to treat various types of return strokes in the lightning location system differently. Finally, for individual cases, the inversed return stroke currents and those measured from artificially triggered lightning and transmission towers are compared. Based on large samples obtained in Wuhan from 2012 to 2019, the cumulative probability distribution curve of the system inversed return stroke currents and the curve recommended by IEEE are compared. It is considered that the lightning location system of China's power grid can retrieve reasonable return stroke currents, though the peak current of negative subsequent return strokes may be overestimated to a certain extent, which may be related to the fact that the return stroke speed set by the system is smaller than the actual value.