Optimal Resilient Distribution Grid Design Extended Abstract

Natural disasters such as earthquakes, hurricanes, and other extreme weather pose serious risks to modern critical infrastructure including electrical distribution grids. Recent U.S. government sources [1, 2] suggest that new methodologies for improving system resilience to these events is necessary. Here, we focus on developing methods for designing and upgrading these grids to better withstand and recover from these threats. These tools select from a set of potential distribution grid upgrades proposed by planners, e.g. adding lines, transformers, breakers, and microgrids (modeled as distributed generation), so that the post-event damaged grid meets a minimum standard of service (98% of the critical demand is met and 50% of all other load is served) with minimum upgrade budget. The approach currently addresses events that create relatively homogeneous damage, e.g. ice storms. When the damage is known apriori, a mixed integer programming formulation is tractable, and we present some initial results.