抄録
Lightning occurs in the Joetsu area on between 30 and 35 days per year. The Japan Meteorological Agency's lightning warnings are valid for as many as 2, 170 hours per year. Moreover, electric power consumers have observed that approximately 65 incidences of grounding (zero-phase voltage of 4.5 kV or more), which are assumed to have been caused by lightning, occurred in the 66-kV systems belonging to utility power suppliers in the Joetsu area. Furthermore, there have been a number of cases of instantaneous power failure (two seconds duration or less) each year that are also assumed to be attributable to lightning. For large electric power consumers, such problems particularly instantaneous power failures which can cause serious damage to production facilities, cannot be overlooked. If predicting such types of lightning could be made relatively easily, electric power consumers would be able to establish production systems that enabled them to anticipate and deal with instantaneous power failure. Lightning prediction systems are therefore particularly important to users of utility power systems.
With the increasingly widespread use of the Internet, it has become possible to readily acquire and utilize cloud-to-ground lightning information that has been made publicly available by Tohoku Electric Power Co., Inc., and Hokuriku Electric Power Co., Inc. Therefore, the authors propose applying a prediction algorithm formulated by them to the assessment index through utilizing such Internet based lightning warnings from the Meteorological Agency and cloud-to-ground lightning information from Tohoku Electric Power Co., Inc., and Hokuriku Electric Power Co., Inc. The number of measured grounding incidences was used to determine the assessment index. During a two-year period from 1999 to 2000, lightning was predicted at the users' end of the Joetsu area's 66-kV utility power supply systems. Measurements indicated a 94.6% success rate in predicting lightning and a 100% success rate for predicting instantaneous and normal power failure, thereby proving the effectiveness of the proposed approach.
