Predicting Risk to Electric Power Distribution Lines

Electrical utilities spend a large fraction of their budget maintaining electrical distribution lines- the low-voltage lines that supply neighborhoods with power. These lines often run on poles among trees, which present a risk of damage when strong winds cause tree limbs to fall on power lines. Good maintenance practices help to mitigate this risk, but given the amenities provided by trees in residential neighborhoods and the expense of burying distribution lines, the risk cannot be brought to zero. However, expense may be saved, and harms to power users reduced by accurate forecasting of locations of highest risk, so that equipment and labor for repair can be brought to bear more quickly. Risk to power lines is a function of wind speed and tree canopy type, but can also be substantially increased by ice burden on trees, itself a difficult forecast problem. In addition very high soil moisture may increase the risk of toppling trees. In this study, we consider the limits to forecast skill of power disruptions due to the limited accuracy of geographic information (power line locations, canopy), wind speed forecasts, and icing forecasts, and identify data and model improvements that would contribute strongly to increased forecast skill for power line risk.