Adapting the statistics of soil properties into existing and future lightning protection standards and guides

Abstract

Risk estimates for lightning faults make use of log-normal distributions for lightning characteristics such as peak current. Simplifications for highly correlated parameters, notably peak current and rate of current rise, justify use of equivalent front time of 2 μs in backflashover calculations for first negative return strokes. The transmission line backflashover rate is also affected by uncorrelated and broad statistical variations in soil resistivity, tower footing impedance Zf and resistance Rf. Statistical properties of Rf from transmission lines in Tennessee USA and Portugal are compared. Modeling of Rf variation using a ten-step distribution from the IEEE Standard 1243 FLASH program is compared with estimates using finer probability step intervals, log-normal and log-logistic models.