Colloquium: Hannes Hobbie "A Massively Parallel Interior-Point Solver for LPs with Generalized Arrowhead Structure, and Applications to Energy System Models"

Abstract:
Traditional congestion management involves the activation of large-scale conventional power plants to relieve grid bottlenecks in electricity transportation networks. The gradual phase-out of these technologies and the development towards a renewable-based generation structure requires novel solutions to ensure system operators have sufficient flexibility means in future electricity systems. These also include new forms of coordination between grid operators and flexibility suppliers, enabling end-user participation, typically owning flexible demand-side applications. However, the increased granularity of future energy systems representation imposes new challenges for the power market modelling community.

In my talk, I will present my latest findings from three selected research papers: First, I will show how distributed computing of the congestion management problem with a newly developed solver can significantly improve solving performance of electricity network models. Secondly, I will discuss how industry decarbonisation can contribute to solving grid congestion by applying fundamental power market modelling. Finally, I will investigate efficiency losses in market-based forms of decentralised flexibility provision considering strategically behaving aggregators using bilevel programming.