Abstract:
The interaction between power systems and wildfires can be dangerous and costly. Distribution grids can be liable for the outbreak of wildfires during extreme weather. In wildfire-prone areas, investment planning should consider the impact of operational actions on wildfire-related uncertainties affecting line failure likelihood. To address this planning exercise, we propose two decision-dependent uncertainty (DDU) aware frameworks to optimize the combination of upgrades to improve the system’s ability to deal with wildfires. The first one is two-stage distributionally robust planning optimization problem with DDU, where the first stage determines optimal switching actions and line investments, and the second stage evaluates the worst-case expected operational cost under a DDU framework designed to account for the endogenous impact of power-flow levels and hardening investment decisions in the line failure probabilities. The second one is a multi-horizon planning framework that integrates these timescales within a single distributionally robust stochastic optimization model to capture both long-term and short-term uncertainties. We develop an exact decomposition algorithm that produces coordinated investment and operational policies across both horizons.
Biography:
Alexandre Moreira is a research scientist in the Energy Technologies Area at Lawrence Berkeley National Laboratory. His research applies mathematical programming techniques to critical questions in power systems planning, operations, and economics, with a particular focus on novel methodologies for transmission expansion planning that account for reliability and resilience. At Berkeley Lab, he leads and collaborates on projects that develop foundational modeling frameworks and tools to (i) formulate risk-based optimization and decision-making models to propose transparent, well-informed “cost vs. risk” tradeoffs for infrastructure planning, and (ii) support the valuation of emerging technologies. He holds BSc and MSc degrees from PUC-Rio, Brazil, and a PhD from Imperial College London.