Context.

Distribution System Operators (DSOs) in Germany face a surge of investor-led requests to connect Battery Energy Storage Systems (BESSs), with E.ON’s distribution arm alone receiving over 2 000 applications for nearly 100 GW of capacity [1], yet connection studies still rely on worst-case power-flow rules that assume simultaneous feeder peaks, so DSOs either refuse projects or set low export limits on the battery’s incjection power. Investors can reduce their network charges by granting the DSO a right to curtail the battery when congestion threatens, yet that option is still assessed with the same conservative worst-case snapshot. The outcome is an overly restrictive picture of hosting capacity and no clear estimate of how often constraint-driven curtailment would trim the battery’s trading returns. The study aims to deliver a fast, joint decision-support tool that lets DSOs and BESS investors quantify, for any feeder bus of interest, the economically optimal power-energy sizing and the revenue impact of grid-driven curtailment, aligning market profitability with network security.

Methodology

Hourly spot-price, demand, PV and wind time-series are synthesized for an entire year from historical weather and load records, and mapped onto the 20 kV CIGRE benchmark grid. A state-of-the-art intra/inter-period time series clustering algorithm [2] then compresses the 8760-hour data set into 30 chronologically coherent “typical” days. We run the one-year arbitrage model for a series of candidate power ratings, starting from a set minimum size up to a set maximum size, so as to maximize trading profit by charging when spot prices are low and discharging when they are high, while enforcing state-of-charge limits, battery efficiency, voltage and line-loading constraints, connection fees, size-dependent capital and land costs, and cycle-related degradation costs. The solution of this optimization problem gives the optimal energy capacity E* together with the rated power P, the hourly dispatch and the resulting Net Present Value (NPV). Two cases are solved: market-only and grid-aware, the latter enforcing grid constraints and logging curtailed energy times the current spot price as lost revenue.

Results & Conclusions

Scanning the full range of power ratings yields three companion curves E*(P), NPV(P) and curtailed-cost(P), giving DSOs and investors an immediate view of economic and technical hosting limits, map the profit–risk frontier, and pinpoint the power levels where network constraints begin to erode returns. The entire analysis completes in minutes using the Gurobi linear-optimization solver.

References

[1] Enkhardt, Sandra. “German transmission companies had connection requests for 226 GW of big batteries at turn of year,” PV Magazine Global (via ESS News), 14 January 2025.

[2] Kotzur, L., Markewitz, P., Robinius, M., & Stolten, D. (2018). Time series aggregation for energy system design: Modeling seasonal storage. Applied Energy, 213, 123–135. https://doi.org/10.1016/j.apenergy.2018.01.023