Sustainable Systems Seminar Lunch Series - Hierarchical Solving of High-Resolution Energy System Models

 

This week's speaker is:

Andreas Mühlbauer, Leavell Fellow and PhD student, Stanford University
"Hierarchical Solving of High-Resolution Energy System Models"

Abstract:

Energy system models that plan highly renewable systems need high resolution in space and time. US grid operators schedule power in 15-minute intervals and dispatch real-time markets every 5 minutes. At such time steps, a full year of a large network quickly becomes too large to solve, so models are coarsened once and their error is rarely known. This talk presents a hierarchical approach similar to the nested grids of weather and climate models: a coarse model of the whole year and system is solved first, and its results serve as boundary conditions for finer levels that are solved in parallel time windows. We test the approach in PyPSA on standard IEEE networks with 14 to 118 buses, which are small enough to solve in full, so every gap to the optimum is exact. At 15-minute resolution, the hierarchy is 2.8–4.0 times faster on the same cores and stays within 0.45–0.9% of the optimum. I close with an outlook on bridging continental and local scales, using the example of zooming into US metropolitan areas.

Bio:

Andreas Mühlbauer is a Leavell Fellow and PhD student in Civil and Environmental Engineering at Stanford University, where he develops optimization frameworks for multi-sector energy systems. His research covers a range of topics in energy system transitions, including renewable integration, infrastructure planning, wind repowering, and carbon management. His current work focuses on renewable microgrids and on hierarchical methods for solving high-resolution energy system models.
Before Stanford, Andreas gained research experience at the Lappeenranta-Lahti University of Technology (Finland), the Potsdam Institute for Climate Impact Research (Germany), and the non-profit Energy Watch Group (Germany), working at the intersection of renewable energy modeling, carbon management technologies, and climate policy.

The topics of this seminar are broad but typically fall under technologies’ scaling potential and impact on (the system of) people, the environment and the economy. A particular focus is placed on the interaction potential of technologies with the energy, water, and material systems. Our goal is to create an intimate, collaborative space for students, postdocs, scientists, and PIs within Stanford across micro-level (material and technology) to macro-level (system) interests. These seminars will provide an opportunity to disseminate insights from your studies, connect with fellow researchers, and strengthen bonds across the community.