A Predictive Indicator of Icing Damage Risk

The ability to predict severe ice storm events is an essential step to prevent or mitigate power grid outages caused by extreme wind and ice loads. We have shown previously that an Ice Accretion Forecasting System (IAFS) can predict severe ice storms several days in advance. However, identifying the storms’ maximum intensities and localizing their epicentres remain challenges to be addressed. In this paper, we describe an icing risk indicator based on the observation that predicted freezing rain fields differ, when compared across time and space. This leads to a concept similar to pseudo-ensembles used for quantitative precipitation forecasts. In addition to precipitation, this new indicator must consider temperature, wind, and fraction of frozen precipitation. The final icing risk indicator is developed using a combination of icing forecasts shifted in time and space, weighted according to the relative distance from the area and time of interest. When combined with power line design parameters, it can provide a probabilistic forecast of areas likely to be affected by ice storms. Performance of the new indicator is illustrated using a case study of an icing event that caused widespread damage to the power distribution infrastructure.