Optimizing Wind Power Integration with A-CAES

(A Techno-Economic Case Study of Gotland, Sweden)

Ahmad Romadun, Ardian C. Pratama, Johanna Kusuma, Tanno Kambier

Gotland Wind Power Plant

Gotland, Sweden’s largest island located in the Baltic Sea, has been designated by the Swedish Energy Agency as a pilot region for renewable energy technology deployments and aims to achieve a climate-neutral energy system by 2025. By 2017, local wind power already supplied half of the island’s energy demand, with the remainder imported via submarine cables from the mainland (Nilsson, Gotland). With one of the highest wind potentials in Sweden, Gotland had around 180 MW of installed wind capacity producing 522 GWh annually as of 2020 (Ossa, 2021), and several upcoming onshore and offshore wind projects further strengthen its position as a prime location for an Adiabatic Compressed Air Energy Storage (A-CAES) plant. The island’s subsurface geology, dominated by salt and limestone formations, provides structurally stable caverns and underground reservoirs with low leakage potential, making them suitable for compressed air storage (Sopher, 2019), and ongoing research initiatives and pilot projects continue to assess the technical and economic feasibility of CAES development on Gotland.

The primary objective of this study is to conduct a techno-economic assessment of wind farms integrated with A-CAES systems, focusing on both technical feasibility and the economic viability of the storage system. Specifically, the analysis aims to evaluate net present value (NPV) and revenue streams as well as determine the optimal system sizing. To achieve these objectives, the study is guided by two research questions: (1) What is the maximum revenue and NPV that can be achieved if A-CAES is integrated with the existing wind power plants in Gotland? and (2) What is the optimal size combination if new A-CAES and wind power plants are developed on the island?

System Overview

In the wind farm integrated with A-CAES, electricity from the wind turbines can either be exported directly to the grid or stored in the A-CAES system. The system layout is illustrated in Figure, with a dashed line showing the system boundary and different lines showing different energy flows in the system. In the case that the electricity will be stored, the motor-compressor will convert electricity into compressed air, which will be stored in the underground salt caverns. The compression process produces waste heat, which will be stored in the TES. During periods of high electricity demand or high electricity prices, the stored compressed air is used to generate electricity through an air turbine coupled with a generator. Before the expansion process, the air will be reheated using the heat stored in the TES to prevent ice buildup in the air turbine \cite{roos_analytical_2022}. The electricity will be exported to the local grid connected to the A-CAES plant. The compressor and the expansion turbine will operate based on the operating modes and decision factors. In this study, the electricity grid is outside the system boundary. It is assumed that the electricity grid can consume 100% of the electricity from the Wind-integrated A-CAES system. The wind farms integrated with the A-CAES system operated under five proposed operating modes (OM1–OM5), each defined by specific rules aimed at maximizing both economic performance and wind energy utilization.Three main decision factors determined these operating modes:

Wind power availability – Assesses whether the wind speed is sufficient to generate power.
Electricity price threshold – Evaluates if the current electricity market price exceeds a predefined threshold.
CAES charge status – Checks whether there is available capacity to store or discharge compressed air.

The interaction of these factors resulted in different operational behaviors for the system under each mode (OM1–OM5). The specific equipment operations were defined accordingly.

Result

The analysis indicates that the optimum configuration for the wind-A-CAES system is achieved with the smallest cavern size (S16) and the largest wind turbine capacity (106.47 MW). Integrating A-CAES increases overall revenue compared to a standalone wind farm, but the higher capital cost significantly lowers the Net Present Value (NPV). The configuration with the best NPV corresponds to a smaller storage size, emphasizing direct grid dispatch (Mode 2), while the highest-revenue setup uses a larger A-CAES capacity that enables greater operation in storage and discharge modes (Modes 1 and 3), enhancing flexibility and monetization when electricity prices are high. Financially, although A-CAES raises revenue by about 2%, its CAPEX is 57% higher, resulting in a much lower NPV (€27 million vs €115 million without A-CAES). Hence, the A-CAES-integrated system is not yet economically attractive under current market price conditions.

The images below illustrate the overall analysis workflow and key simulation outcomes. The first set of figures presents the share of operating modes for both the best Net Present Value (NPV) and best revenue configurations, highlighting how different operational strategies influence system performance. The Pareto front analysis was used to identify the optimal balance between economic performance and technical feasibility, guiding the selection of the best design parameters such as cavern size, turbine capacity, and threshold price. The simulation result shown above depicts the system’s dynamic behavior—specifically, wind electricity generation, A-CAES and TES storage activities, and electricity price variations relative to the threshold—demonstrating how the model responds to market signals and storage availability across the analyzed time horizon.

Conclusion

In conclusion, the integration of an A-CAES system enhances the overall revenue of the wind farm across all evaluated scenarios. The configuration achieving the highest NPV scenario using storage increases annual revenue by approximately €245,000 (from €19.95 million to €20.19 million), while the highest possible annual revenue reaches €21.56 million.

The model results indicate that the largest infrastructure setup—combining the maximum wind turbine capacity, compressor/expander size, and the largest (S12) cavern—produces the highest total revenue. Conversely, the configuration yielding the best NPV is achieved with the same wind capacity but with the smallest cavern (S16), minimizing capital costs. This setup prioritizes direct electricity dispatch to the grid (Mode 2), which occurs about 72% of the time, rather than relying heavily on storage operations.

When comparing systems, the NPV for the configuration with A-CAES drops significantly to €27.3 million, compared to €115.1 million for the system without A-CAES, mainly due to the 57% higher CAPEX (€240 million vs €152 million). This implies that the additional revenue generated does not offset the increased investment cost under current market conditions.

Therefore, for private investors, installing A-CAES in a wind farm would only be economically viable under certain circumstances—such as the presence of governmental subsidies, grid stability requirements, or dynamic pricing schemes that allow flexible operation strategies. While A-CAES offers social and technical benefits, such as reducing curtailment, providing peak shaving, and enhancing energy reliability, its economic justification remains limited at the current cost and market price levels.

Future Work

This study examined the potential of A-CAES integration to enhance the economic performance of wind farms in Gotland, focusing on revenue and NPV improvement. Since no utility-scale A-CAES plants currently exist, the model is based on assumptions drawn from existing research, and several future directions are recommended:

System Design Refinement – Future studies could investigate alternative configurations, such as multiple compression and expansion stages, integration of multiple TES units with varying temperature levels, or improved heat exchanger effectiveness. These variations could influence round-trip efficiency and economic outcomes.
Dynamic Dispatch Strategy – The current model applies a fixed threshold selling price. Future work could implement a dynamic pricing approach or integrate electricity price forecasting to optimize discharge timing and increase profitability.
Grid Interaction Studies – The system boundary here excludes the electricity grid. Subsequent research should analyze the impact of A-CAES on grid frequency regulation, ancillary services, and stability.
Sustainability Assessment – Beyond technical and economic aspects, environmental evaluations such as life-cycle assessment, decommissioning, and emission analysis should be conducted to assess long-term sustainability.
Hybrid Operation Strategies – The A-CAES could be charged during low-price periods or by integrating other renewable sources (e.g., solar or biomass), enabling hybrid charging. Such strategies could enhance renewable energy utilization and further reduce curtailment.

Page updated

Google Sites

Report abuse