Diffusivity of Carbon Dioxide in Aqueous Solutions under Geologic Carbon Sequestration Conditions

Abstract

Accurate assessment of the long-term security of geologic carbon sequestration requires knowledge of the mobility of carbon dioxide in brines under pressure and temperature conditions that prevail in subsurface aquifers. Here, we report Raman spectroscopic measurements of the rate of CO 2 diffusion in water and brines as a function of pressure, salinity, and concentration of CO 2 . In pure water at 50 ± 2°C and 90 ± 2 bar, we find the diffusion coefficient, D, to be (3.08 ± 0.03) × 10 −9 m 2 /s, a value that is consistent with a recent microfluidic study but lower than earlier PVT measurements. Under reservoir conditions, salinity affects the mobility of CO 2 significantly and D decreased by 45% for a 4 M solution of NaCl. We find significant differences of diffusivity of CO 2 in brines (0−4 M NaCl), in both the absolute values and the trend compared to the Stokes−Einstein prediction under our experimental conditions. We observe that D decreases significantly at the high CO 2 concentrations expected in subsurface aquifers (∼15% reduction at 0.55 mol/kg of CO 2 ) and provides an empirical correction to the commonly reported D values that assume a tracer concentration dependence on diffusivity.

Diffusion coefficients at infinite dilution of carbon dioxide and methane in water, ethanol, cyclohexane, toluene, methanol, and acetone: A PFG-NMR and MD simulation study

“…Excellent agreement between the results from the present work and literature data (Fig. 3) is found for the systems (CO 2 + cyclohexane), for which only one data set is available in the literature; (CO 2 + toluene), for which the present data basically confirm the recent data of Cadogan et al [41]; and (CH 4 + of CO 2 in water with literature data [7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34], grouped by the used experimental method. Circles: Diaphragm cell, squares: Wetted surface absorber, triangle: In a gross picture, when comparing the results from the PFG-NMR to those from other methods, the former tend to lie rather on the upper end of the band of results than at the lower.…”

“…This is in particular the case for the two systems with the solvent water. For (CO 2 + water), abundant data for the diffusion coefficient at infinite dilution are available [7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34]. Fewer data are available for CO 2 in ethanol [35,10,11,36,37,29,38], cyclohexane [39], toluene [10,40,41], methanol [35,10,11,19,31,42], and acetone [10].…”

Diffusion of CO2 in Magnesite under High Pressure and High Temperature from Molecular Dynamics Simulations

“…The averages of D values increased by 1.5%-4.8% as the CO 2 pressure increased from 10 MPa to 100 MPa under the different pressures. As detected by Raman spectroscopy, the diffusion coefficient of CO 2 in aqueous solutions decreases significantly at high CO 2 concentrations [49]. Our results here show different diffusion behavior from the diffusivity of CO 2 in aqueous solutions under geologic carbon sequestration conditions.…”

Lab on a chip for a low-carbon future

“…39 Microfluidics have provided new insights into CO 2 behavior for subsurface storage, including invasion patterns of CO 2 displacing the fluids in model porous media, [40][41][42][43] characterizing the physical properties of CO 2 in the supercritical state, [44][45][46] and screening CO 2 solubility in various solvents and brines for delivery. 47,48 Although CO 2 mineralization occurs on geological timescales, microfluidic devices have provided new insights into how CaCO 3 precipitates along the transverse mixing zone, thus substantially reducing porosity and permeability, which can limit the formation of more stable carbonate phase, such as calcite. [49][50][51][52] Reproducing the chemical properties of subsurface media, in addition to their structure, can be necessary to assess and predict system reactivity during carbon injection and storage.…”