Fair and Reliable Reconnections for Temporary Disruptions in Electric Distribution Networks using Submodularity

Abstract

We analyze a distributed approach for automatically reconfiguring distribution systems into an operational radial network after a fault occurs by creating an ordering in which switches automatically close upon detection of a downstream fault. The switches' reconnection ordering significantly impacts the expected time to reconnect under normal disruptions and thus affects reliability metrics such as SAIDI and CAIDI, which are the basis for regulator-imposed financial incentives for performance.We model the problem of finding a switch reconnection ordering that minimizes SAIDI and the expected reconnection time as Minimum Reconnection Time (MRT), which we show is a special case of the well-known minimum linear ordering problem [29] from the submodular optimization literature, and in particular the Min Sum Set Cover problem (MSSC) [19]. We prove that MRT is also NP-hard.We generalize the kernel-based rounding approaches of Bansal et al. [5] for Min Sum Vertex Cover to give tight approximation guarantees for MSSC on c-uniform hypergraphs for all c. For all instances of MSSC, our methods have a strictly better approximation ratio guarantee than the best possible methods for general MSSC.Finally, we consider optimizing multiple metrics simultaneously using local search methods that also reconfigure the system's base tree to ensure fairness in service disruptions and reconnection times and reduce energy loss. We computationally validate our approach on the NREL SMART-DS Greensboro synthetic urban-suburban network. We evaluate the performance of our reconfiguration methods and show significant reductions compared to single-metric-based optimizations.

Sharing the Load: Considering Fairness in De-energization Scheduling to Mitigate Wildfire Ignition Risk using Rolling Optimization

“…Optimized PSPS formulations provide the groundwork for explicitly considering fairness when making line deenergization decisions. Fairness is an increasingly studied topic in many fields, such as artificial intelligence [12] and wireless spectrum allocation [13], as well as in other power systems applications, e.g., energy storage [14], distribution network reconfiguration [15], transmission loss allocation [16], and tariff design [17]. However, incorporating notions of fairness into PSPS decision-making introduces modeling, algorithmic, and computational challenges.…”