Optimal energy dispatch of the power distribution network during the course of a progressing wildfire

Wildfires are natural phenomena that play a crucial role in many forest and grassland ecosystems. Under favorable conditions, that is, extreme heat, availability of fuel, and high winds, they may spread beyond control and approach city limits. Power lines are vulnerable to wildfires in their vicinity, mainly because of increased conductor temperatures as a result of ambient temperature rising. This may lead to conductor annealing and/or increased sag. A violation of safety clearances of the lines could in turn lead to possible flashovers and onset of cascading failures. To address the situation, the operator may dynamically alter the ratings of the lines in the affected areas in order to reduce the flow of current through them, thereby assisting with conductor cooling. However, with reduced line capacities, the utility may not be able to supply all the loads from the main substation anymore, which may increase the risk of unserved load and load shedding. A solution is put forth in this paper that enables the system operator to dispatch distributed generators, demand responsive loads, and microgrids in order to supply loads under such emergency conditions. Because of the uncertain spread/severity of wildfires, the problem is stochastic in nature and is therefore modeled as two-stage stochastic optimization. First-stage variables determine the generation capacity to be purchased before the onset of the event, while second stage variables represent generation dispatch commands. The efficiency of the proposed solution is verified on a modified version of the Institute of Electrical and Electronics Engineers (IEEE) 123-bus test system. Copyright (C) 2014 John Wiley & Sons, Ltd.