Flexibility options in electricity markets with high shares of renewable energies: An agent-based analysis of economic viability and system impacts

Abstract

As part of the energy transition, the electricity sector and its power generation infrastructure must be extensively transformed to reduce greenhouse gas emissions. As renewable energy sources rapidly replace fossil fuel-based electricity generation, intermittent wind and solar power sources require increasing supply and demand balancing. Flexibility options are critical technological solutions for this challenge, however, significant knowledge gaps remain regarding their profitability and deployment in future electricity systems.

This thesis contributes to closing these gaps by conducting agent-based electricity market simulations that target several key research objectives. First, it investigates operational strategies for flexibility options and their economic performance in electricity markets. The analysis reveals that strategies must consider the price impacts of flexibility options on both an individual and collective level. Second, the research examines profitable technical parameters and demonstrates that medium-term storage configurations likely offer the greatest revenue potential in day-ahead markets. This is particularly relevant in scenarios with increased competition between flexibility options. Third, the analysis identifies significant cannibalisation effects when substantial flexibility option capacity is integrated into the system. Due to their impact on electricity price dynamics, the overall revenue decreases, thus affecting profitability.

These findings lead to the following recommendations for future research and policy. Endogenous modelling of large-scale flexibility options is necessary to understand future energy systems, as isolated analysis significantly underestimates market interactions. Additionally, the profitability of flexibility options is strongly linked to cost assumptions, particularly storage costs, which require consideration in investment decisions. Furthermore, planned investments in flexibility options should be accurately monitored, as cannibalisation effects and changing market dynamics will substantially impact profitability.

The endogenous modelling of these complex phenomena represents a key methodological advancement of this thesis. This is achieved through extensions and application of a state-of-the-art electricity market model coupled with machine learning-based electricity price forecasting. This thesis contributes multiple software packages and data sets that adhere to FAIR principles, thereby enhancing reproducibility and facilitating future research. The presented work advances the understanding of the role of flexibility options in energy transition scenarios, while also revealing important avenues for future investigation. Key limitations to be addressed in further research include the focus on day-ahead market analysis rather than multi-market simulations, and simplified consideration of sector coupling dynamics.