Modeling and measurement of CO2 solubility in salty aqueous solutions and application in the Erdos Basin

• A simple model has been proposed to estimate CO2 solubility in formation brine for temperatures from 313 K to 378 K, for pressures from 50 to 220 bar, and for mixed salt concentrations from 0 to 4 mol/kg.
• Degrees of impact of different cations (K+, Na+, Ca2+, Mg2+) on CO2 solubility have been quantified.
• Brines in five formations in the Erdos Basin have been sampled and analyzed.
• Experiments to measure CO2 solubility in synthetic brine have been conducted in T–P range included in CO2 geological sequestration.

Geological carbon storage (GCS) is recognized as an effective method for mitigating the greenhouse effect. Deep saline aquifers hold the highest potential capacity for CO2 storage. CO2 solubility in salty aqueous solutions under geological sequestration conditions plays a key role in GCS. However, most CO2 solubility studies focus mainly on single-salt solutions (NaCl and CaCl2), and extrapolation of these studies to aqueous solutions with mixed ions is unavailable. To fill the research gap, based on the collection of CO2 solubility data, a semi-empirical thermodynamic model is proposed in this paper to calculate CO2 solubility in aqueous solutions containing K+, Na+, Ca2+ and Mg2+ in the temperature and pressure ranges from 313 K to 378 K and from 50 bar to 220 bar. To describe the CO2–liquid phase equilibrium, the Peng–Robinson equation of state (PR EoS) and the Setschenov equation are applied. The former has been modified to improve its performance in the studied T–P range, and the latter shows excellent accuracy with only three optimized parameters. Before modeling was done, experimental studies were conducted. Brine sampling from five reservoirs potentially chosen for CO2 sequestration in the Erdos Basin was carried out using a monitoring well in the support of the Shenhua Group carbon capture and storage (CCS) site project. The chemical composition of the samples was determined, and experiments measuring CO2 solubility were carried out in synthetic brine with 64 valid data points reported. An analytical method with a simplified sampling technique was chosen. In the range studied, the average absolute deviation of CO2 solubility between the model and experimental results was 2.01%, and the maximum absolute deviation in this study was less than 4.79%. The proposed model and experimental data therefore possess broad adaptability to GCS with satisfactory accuracy.