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

Felix Nitsch1,* PhD Defense Bochum, 22nd September 2025 Flexibility Options in Electricity Markets with High Shares of Renewable Energies An Agent-based Analysis of Economic Viability and System Impacts 1 German Aerospace Center | Institute of Networked Energy Systems | Energy Systems Analysis * [email protected] Acknowledgements: Kristina Nienhaus on behalf of the Energy Economics Group 2Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Image: ourworldindata.org/electricity-mix Electricity Generation by Source Renewables on the Rise: Germany 3Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Image © by Fraunhofer-Gesellschaft: www.energy-charts.info Renewable Energy Transformation Electricity Generation in Germany in 2015 4Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Image © by Fraunhofer-Gesellschaft: www.energy-charts.info Renewable Energy Transformation Electricity Generation in Germany in 2025 Efficiency Power input/output in kW Flexibility Options (FO) Mitigating Intermittent Electricity Generation 5Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Stored Energy in kWh Charging/ discharging time in h 𝐸=𝑃∙𝑡∙ η Technological Solutions to Provide Flexibility 6Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Image based on Schmidt et al. (2019). 10.1016/j.joule.2018.12.008 State of the Literature Flexibility Options on Future Electricity Markets Background ▪Renewable energy (RE) power plants essential Twidell (2021) ▪Increasing electricity demand by electrification Staffell and Pfenninger (2018) ▪Flexibility becoming more important Michaelides et al. (2020) ▪Several technological solutions considered as Flexibility options (FO) Zöphel et al. (2018) ▪Thorough ex ante analysis necessary for FO investment Keles (2013), Ölmez, Ari, and Tuzkaya (2024) Existing studies ▪Focus on individual FO devices without considering market implications Lund et al. (2015), Elalfy et al. (2024) ▪Assume “central planner” and perfect coordination Mancò et al. (2024) Identified gaps ▪Lack of endogenous modelling of FO impacts on market dynamics Siala et al. (2022) ▪Need for analysis of high RE scenarios while incorporating operational uncertainty Bessa et al. (2019) ▪Limited open science to promote transparency, reproducibility, and wider application Chang et al. (2021) 7Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Research Targets Advancing the Understanding of Flexibility Options in High RES Scenarios Energy Economics T1.1 Identify operational strategies for FOs performing reliably in high RES scenarios T1.2 Evaluate how technical parameters (capacity & power) influence refinancing potential T1.3 Quantify effect of increasing competition in spot electricity markets T2.1 Expand ABM to capture individual and collective impact on system dynamics 8Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 T2.2 Develop modular open-source software to enhance reproducibility and further analyses Energy Informatics ( ) Synthesis Paper Contributions 9Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 T1.2 technical parameters T1.1 operational strategies T1.3 increasing competition T2.1 expand ABM T2.2 modular open software Paper I Paper II Paper III Paper IV ( ) Synthesis Paper Contributions 16 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 T1.2 technical parameters T1.1 operational strategies T1.3 increasing competition T2.1 expand ABM T2.2 modular open software Paper I Paper II Paper III Paper IV Endogenous Modelling of Storage Competition Paper III 17 Nitsch, F., Schimeczek, C., & Bertsch, V. (2024). Applying machine learning to electricity price forecasting in simulated energy market scenarios. Energy Reports, 12, 5268-5279. 10.1016/j.egyr.2024.11.013 Storage Activity Basic Operational Considerations and Implications 18 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Low Prices High Prices C h a r g i n g D i s c h a r g i n g Enhance Electricity Price Forecasts Account for collective impact 19 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Applying electricity market model AMIRIS Evaluation of forecasting performance Generation of scenarios for training and test data Training ML Model Need for robust price forecasts in simulation models All emojis designed by OpenMoji –the open-source emoji and icon project. License: CC BY-SA 4.0 Machine Learning (ML), Temporal Fusion Transformer (TFT) Nitsch et al. (2024). 10.1016/j.egyr.2024.11.013 Electricity Price Forecasting Performative Prediction Forecasting performance in scenarios of varying renewable energy expansion Train Test I Test II Note: Scenarios are considered as parameter variations and shall not be interpreted as definitive and complete future electricity systems Temporal Fusion Transformer (TFT) Nitsch et al. (2024). 10.1016/j.egyr.2024.11.013 20 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 T1.3 increasing competition AMIRIS Open Agent-based Electricity Market Model Schimeczek et al. (2023a). 10.21105/joss.05041 Schimeczek et al. (2023b). 10.21105/joss.05087 Nitsch et al. (2023a). 10.21105/joss.04958 21 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Nitsch et al. (2023a). 10.21105/joss.04958 Nitsch et al. (2023b). 10.5281/zenodo.8382789 Nitsch et al. (2025a). 10.5281/zenodo.14907870 ) Schimeczek et al. (2023a). 10.21105/joss.05041 Schimeczek et al. (2023b). 10.21105/joss.05087 Schimeczek et al. (2023c). 10.5281/zenodo.14273248 Model Development and Simulation Setup Enabling Machine-Learning Based Electricity Price Forecasts 22 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 T2.1 expand ABM T2.2 modular open software Model Coupling AMIRIS PriceForecast Nitsch et al. (2025). 10.2139/ssrn.5320926 23 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 ( ) Synthesis Paper Contributions 24 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 T1.2 technical parameters T1.1 operational strategies T1.3 increasing competition T2.1 expand ABM T2.2 modular open software Paper I Paper II Paper III Paper IV Assessing Future Battery Storage Potential Paper IV 25 Nitsch, F., Schimeczek, C., & Bertsch, V. Profitability of Competing Flexibility Options in Renewable-Dominated Energy Markets: Combining Agent-Based and Machine Learning Approaches. Available at SSRN 5320926. 10.2139/ssrn.5320926 2020 I I 2023 2022 I 2021 I 2024 I 2025 I in review Supplementary Work Main Paper Software & Data Legend Icon by OpenMoji, CC BY-SA 4.0 32 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Contributions of my PhD Discussion Limitations 33 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 Scenario/Systems Uncertainty ▪Pace of RE deployment exceeded projections Creutzig et al. (2017) ▪Dynamics of net-zero system may differ Azevedo et al. (2021) ▪Contribution of sector coupling not sufficiently considered Gaafar et al. (2024) ▪No fundamental market design changes considered Ölmez et al. (2024) ▪Shocks, such as the 2022 energy crisis, not considered Ruhnau, Stiewe, et al. (2023) Profitability Uncertainty ▪Missing important additional revenue Agrela et al. (2022) ▪No complex bidding Signer et al. (2025) ▪But, competition also likely on intraday markets and for system services Deman et al. (2025) Discussion Contextualization 34 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 ▪FO refinancing on spot markets only is challenging Hu et al. (2022) ▪Multi-market bidding increases profitability Miskiw et al. (2025), Agrela et al. (2022), Sorourifar et al. (2020) ▪Especially intraday markets relevant for battery storage Löhndorf, Wozabal (2023) ▪Ideal energy-to-power ratio tends to be lower in literature Schmidt, Staffell (2023) ▪Careful application of ML-based forecasts necessary Keles (2016), Fraunholz et al. (2021), Amor et al. (2024) ▪Accurate modeling of (future) markets essential Ward et al. (2019) ▪Increasing RE shares may amplify price volatility Liebensteiner et al. (2025) ▪Interpreting simulation results as trends rather than predictions Pfenninger et al. (2014) Learnings and Recommendations ▪Use strategies considering technical parameters ▪Define technical parameters with market dynamics in mind ▪Reduce storage cost ▪Explicitly account for competition effects ▪Anticipate novel market dynamics and avalanche effects 35 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22 T1.2 technical parameters T1.1 operational strategies T1.3 increasing competition T2.1 expand ABM T2.2 modular open software ▪Build upon existing models ▪Utilize community benefits ▪Go the “extra mile” in software development ▪Follow FAIR4RS principles Contact: Felix Nitsch, [email protected], German Aerospace Center, Institute of Networked Energy Systems, Energy Systems Analysis Conclusions ▪High need and interest in electricity storage, yet market dynamics unclear ▪Investigation with modular agent-based electricity market simulations ▪Various storage analyses reveal challenging profitability potential ▪Storage strategies essential, especially when self-cannibalization occurs ▪Niches to be filled by different storage systems ▪However, cost reductions or additional revenue streams necessary ▪Enhance storage strategies submitted paper Schimeczek et al. (2025) ▪Detailed assessment of sector coupling impact & application for other FO Acknowledgements: Kristina Nienhaus on behalf of the Energy Economics Group Outlook 0000-0002-9824-3371 0000-0002-9824-3371 36 Imprint Topic Flexibility options in electricity markets with high shares of renewable energies: An agent-based analysis of economic viability and system impacts Date 2025-09-22 Author Felix Nitsch Institute Institute of Networked Energy Systems Credits Content DLR and Felix Nitsch (CC-BY-4.0) DLR Logo, REMix Logo, AMIRIS Logo, and title background image © DLR RUB Logo © RUB Acknowledgements: Kristina Nienhaus on behalf of the Energy Economics Group 37 Felix Nitsch, Institute of Networked Energy Systems, 2025-09-22