January 27, 2022
Many consumers, many producers: The power system of the future will become more decentralized and thus more complex. How can security of supply be ensured despite multiple loads on the grids and how can the energy system be made resilient? These questions are the focus of a joint research project of Jacobs University Bremen and OFFIS – Institute of Computer Science in Oldenburg. It models future loads on the system using game theory and machine learning.
More than two million photovoltaic systems are already generating electricity nationwide, and 15 million e-cars are expected to be on the road by 2030. In addition, there are wind turbines, heat pumps and battery storage systems. Taken individually, the small consumers and generators of renewable energy are not critical for system stability; taken as a whole, they are, due to simultaneity effects. When the sun shines, photovoltaic systems kick in, and at the same time, e-cars charge their batteries as the price of electricity drops. In the future, small private plants will also be called upon in the event of bottlenecks in the power grid, in order to resolve or at least mitigate them. Until now, conventional coal and gas-fired power plants have primarily fulfilled this function, the so called redispatch.
"The grids can only withstand these jumps to a limited extent," said Marius Buchmann, project manager of the Bremen Energy Research working group at Jacobs University. "The possibility of short-term interactions in the power grid, synchronized by market incentives, poses new challenges for grid operators and a need for systematic, simulation-based analysis."
The researchers are looking at these requirements in a digitized and interconnected power system from the perspective of grid operators. In the future, they will rely on system services from distributed plants via various market mechanisms. For example, photovoltaic systems or even heat pumps will contribute to stabilizing the grid voltage at 50 Hertz in the future.
"We are modeling various scenarios for the market rules of these system service markets and are looking at how the behavior of the individual players will affect the stability of the grids under these changing conditions," said Buchmann. Will critical situations arise? And how can these be avoided? To find out, the researchers at Jacobs University, led by Gert Brunekreeft, Professor of Energy Economics, are resorting to game theory. These model results are then merged with OFFIS modeling.
"We combine artificial intelligence techniques, especially from the field of self-learning systems, with proven simulation techniques to simulatively map the interactions between prosumers operating in the market and the electrical infrastructure. In the process, the market participants are to learn bidding strategies independently in the simulation in order to be able to investigate an adaptive and thus realistic reaction of prosumers to market incentives," said Astrid Nieße, Professor of Digitalized Energy Systems at the University of Oldenburg and scientific director at OFFIS. Prosumers are both, producers and consumers of electricity.
The goal of the research project, titled "Resilience in the digitalized power system: market rules for dealing with simultaneity effects in system service markets," is to develop a toolbox. The tools contained in this toolbox are intended to enable key stakeholders to identify and assess the impact of new market rules on stakeholder behavior and thus on the resilience of the power system. The use and further development of open source software and the comprehensive use and provision of open data is a key aspiration of the project. The research project is funded by the German Federal Ministry for Economic Affairs and Climate Protection for a period of three years with a total of around one million euros.
Questions can be addressed to:
Project Manager Working Group Bremen Energy Research
Phone: +49 421 200-4868
Email: m.buchmann [at] jacobs-university.de