Resilience Enhancement of Distribution Grids Against Extreme Weather Events

• Published in: IEEE transactions on power systems Vol. 33; no. 5; pp. 4842 - 4853
• Main Authors: Ma, Shanshan, Su, Liu, Wang, Zhaoyu, Qiu, Feng, Guo, Ge
• Format: Journal Article
• Language: English
• Published: New York IEEE 01-09-2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; decision-dependent uncertainty; Decisions; Distributed generation; Distribution systems; failure probability; Feeders; Hardening; Lower bounds; Maintenance engineering; MATHEMATICS AND COMPUTING; Meteorology; Mixed integer; POWER TRANSMISSION AND DISTRIBUTION; progressive hedging; Reactive power; Resilience; resilience-oriented design; scenario bundling; Stochastic processes; stochastic programming; Switches; Uncertainty; Weather
• ISSN: 0885-8950; 1558-0679
• DOI: 10.1109/TPWRS.2018.2822295

This paper proposes a resilience-oriented design (ROD) technique to protect distribution grids against high-impact but low-probability extreme weather events. The problem is formulated as a two-stage stochastic mixed integer problem. The first stage is to make ROD decisions, i.e., hardening existing distribution lines and deploying ROD resources such as back-up distributed generators and automatic switches. The second stage evaluates the system operation cost during a realized extreme weather event and repair cost after the event. A novel modeling strategy is proposed to deal with the decision-dependent uncertainty of distribution line damage status, which is affected by the first-stage hardening decisions. As both stages have binary variables, a modified and computationally efficient progressive hedging algorithm with scenario bundling is introduced. The algorithm performance is evaluated by calculating lower bounds of solutions. The proposed model and algorithms are demonstrated on 34-bus and 123-bus test feeders.