The North German Basin (NGB) is known for its capacity for geological storage applications. However, due to limited data availability, the data first needs to be collected from different sources and further compilation and processing is needed. This study's primary step is to compile and refine public data, fostering transparency for further scientific exploration. The comprehensive database encompasses five key aspects: geological features (e.g., reservoir thickness), litho-mineralogical characteristics, petrophysical properties (e.g., porosity), surface and risk factors (e.g., fault zones), and DAC-related considerations (e.g., geothermal energy availability).

The secondary step involves adapting and validating a holistic ranking methodology for evaluating sites suitable for direct air capture (DAC) technology in the NGB region. This method incorporates ten distinct criteria for initial site screening and selection. The criteria are divided into five classes and assigned scores based on quantitative or qualitative assessments. Then weighting factors were applied to quantify the relative importance of each criterion. They are determined by analytical hierarchy analysis method, which leveraging pair-wise comparisons to minimize inconsistencies and mitigate bias in expert opinions from a broad research background.

The method is tested on the 91 identified potential sites for underground CO₂ storage, with 4 offshore North Sea traps identified for their favorable geological characteristics and 2 other onshore sites, a gas field in the west of Bremen and a saline aquifer in the north-east of Berlin. Future site-specific studies should be carried out on these four traps.

Our study investigates the subsurface storage of CO₂ in saline aquifers, specifically in the Volpriehausen Sandstone beneath the German North Sea. We conduct numerical simulations using the TOUGH3 simulator with the ECO2N module. The reliable estimation of dynamic storage capacity for CO₂ storage is a challenge due to the lack of measurements for process parameters in the model area.

Therefore, we estimate some parameter ranges from literature and OpenData for Volpriehausen Sandstone from Denmark and the Netherlands. Based on these parameter ranges, sensitivity analyses are conducted to identify important rock parameters. A comprehensive dataset of parameters and corresponding simulation results is generated using Latin Hypercube Sampling.

This sample is used to perform sensitivity analyses and to train surrogate models using machine learning approaches. This allows us to identify relevant process parameters. At a constant injection rate, the injection pressure is affected by the permeability and pore compressibility of the reservoir rock, as well as the reservoir and injection temperature. The storage efficiency is affected by the relative permeability of the reservoir rock. The predictions of the surrogate models are illustrated with 3D simulations.

This research has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement “LEILAC2 - Low Emission Intensity Lime and Cement” GA 884170.