Multi-period Power System Risk Minimization under Wildfire Disruptions

arXiv2305.02933MaRDI QIDQ6435360FDOQ6435360

Authors: Hanbin Yang, Noah Rhodes, Haoxiang Yang, Line A. Roald, Lewis Ntaimo 

Publication date: 4 May 2023

Abstract: As climate change evolves, wildfire risk is increasing globally, posing a growing threat to power systems, with grid failures fueling the most destructive wildfires. In day-to-day operations, preemptive de-energization of equipment is an effective tool to mitigate the risk and damage of wildfires. However, such power outages have significant impacts on customers. This paper proposes a novel framework for planning preemptive de-energization of power systems to mitigate wildfire risk and damage. We model wildfire disruptions as stochastic disruptions with random magnitude and timing and formulate a two-stage stochastic program that maximizes the electricity delivered while proactively reducing the risk of wildfires by selectively de-energizing components over multiple time periods. We use cellular automata to sample grid failure and wildfire scenarios based on grid-related risks and environmental factors. We develop a decomposition algorithm that can generate adaptive shutoff plans before a disruption occurs. We test our method on an augmented version of the RTS-GLMC test case in Southern California and compare it with two benchmark cases. We find that our method reduces both wildfire damage costs and load-shedding losses over multiple time periods, and our nominal plan is robust against the uncertainty model perturbation.

This page was built for publication: Multi-period Power System Risk Minimization under Wildfire Disruptions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6435360)