Risk-based generation dispatch in the power grid for resilience against extreme weather events

Abstract

Natural disasters have been considered as one of the main causes of the largest blackouts in North America. When it comes to power grid resiliency against natural hazards, different solutions exist that are mainly categorized based on the time-frame of analysis. At the design stage, robustness and resiliency may be improved through redundant designs and inclusion of advanced measurement, monitoring, control and protection systems. However, since massive destructive energy may be released during the course of a natural disaster (such as a hurricane) causing large-scale and widespread disturbances, design-stage remedies may not be sufficient for ensuring power grid robustness. As a result, to limit the consequent impacts on the operation of the power grid, the system operator may be forced to take immediate remedial actions in real-time. To effectively manage the disturbances caused by severe weather events, weather forecast information should be incorporated into the operational model of the power grid in order to predict imminent contingencies. In this work, a weather-driven generation dispatch model is developed based on stochastic programming to provide a proactive solution for power grid resiliency against imminent large-scale disturbances. Hurricanes and ice storms are studied as example disaster events to provide numerical results. In this approach, the statistics of the natural disaster event are taken into account along with the expected impact on various power grid components in order to determine the availability of the grid. Then, a generation dispatch strategy is devised that helps operate the grid subject to weather-driven operational constraints.