Impact of Reservoir Permeability, Permeability Anisotropy and Designed Injection Rate on CO₂ Gas Behavior in the Shallow Saline Aquifer at the CaMI Field Research Station, Brooks, Alberta

Abstract

Carbon capture and storage is part of Canada’s climate change action plan to reduce greenhouse gas emissions. The Containment and Monitoring Institute Field Research Station contributes to scientific and technological progress that ensures the secure underground storage of CO₂. In this study, the process of shallow CO₂ gas injection (300 m) and subsequent plume development at the FRS was investigated using numerical simulation. Due to reservoir uncertainties, various sensitivity analyses were performed to illustrate their effects on CO₂ saturation, plume distribution and CO₂ dissolution in a saline reservoir in response to variations in horizontal permeability (k_h), k_v/k_h ratio and CO₂ injection rate. The distance of horizontal migration of the plume post-injection was predicted analytically, and the result was validated against the numerical simulation prediction. Results show that increases in k_v/k_h ratio resulted in increases in both vertical and lateral plume migration and in decreases in dissolution rate and CO₂ solubility. It was also indicated that the subsequent post-injection CO₂ migration rate was independent of both k_v/k_h and previous injection rate. Dissolution varied significantly with changes in horizontal permeability. The model shows that increased horizontal permeability facilitated plume migration vertically and horizontally. Modeled permeability variations in horizontal permeability (k_h) and k_v/k_h ratio had a progressively decreasing effect on plume vertical migration with time, while lateral migration effects increased with time.

Appendix: Analytical Solution of CO2 Plume

The CO₂ plume analytical solution was proposed by Nordbotten et al. (2005):

$$- \frac{\lambda - 1}{{r^{\prime}\left( {\left( {\lambda - 1} \right)b^{\prime} + 1} \right)^{2} }} + 2\varGamma r^{\prime}b^{\prime} + 2\varLambda r^{\prime} = 0,$$

(1)

$$\varLambda \left( {\lambda - 1} \right)^{2} - \varGamma \lambda \ln \left( {\frac{\varGamma + \varLambda }{\varLambda \lambda }} \right) = \frac{{2\lambda \left[ {\varLambda \left( {\lambda - 1} \right) - \varGamma } \right]^{2} }}{\lambda - 1}.$$

(2)

where $\lambda$ is the averaged phase mobility of water and CO₂, defined as $\lambda = \frac{b}{B}\lambda_{c} + \frac{B - b}{B}\lambda_{w}$, where b denotes thickness of CO₂ layer, and B represents total reservoir thickness. $\lambda_{\alpha } = \frac{{k_{r\alpha } }}{{\mu_{\alpha } }}$ is the ratio of relative permeability to fluid viscosity, where $\alpha$ represents each phase, with c for CO₂ and w for water. $\varLambda$ denotes the Lagrangian multiplier.

In addition, $r^{\prime}$, $b^{\prime}$ and $\varGamma$ are dimensionless variables, where $r^{\prime} = r\sqrt {\frac{\pi B\varphi }{{Q_{\text{well}} t}}}$, $b^{\prime} = \frac{b}{B}$, and $\varGamma = \frac{{2\pi \Delta \rho g\lambda_{\text{w}} kB^{2} }}{{Q_{\text{well}} }}$. In these equations, r denotes migration distance, $\Delta \rho$ is the density differential between brine and CO₂, g is the gravitational constant, k is the average permeability of the reservoir, $\varphi$ is the average porosity, t is the injection period, and $Q_{well}$ is the CO₂ injection rate. For simplicity, water and CO₂ viscosities and densities are assumed constant. Fluid properties of CO₂ and brine used in these calculations are shown in Table 1.