(244e) The Investigation of Impact of Mineralogical Heterogeneity of Clay–Calcite Based Fracture–Matrix System for CO2 Storage By Using a Hybrid–Scale Model

CO2 storage and sequestration are considered an effective approach to mitigate greenhouse gas emissions. Injected CO2 in the subsurface formation can easily move upward by buoyancy or through leakage out of the target reservoir through the existing fractures and negatively influences the CO2 sealing capacity and security. While injecting an enormous amount of CO2 into carbonate–rich aquifers, CO2 is effortlessly dissolved in the formation brine under the high pressure, and the subsequently formed CO2–enriched brine can react with the calcite. Reaction–induced changes in pore structure and fracture geometry alter the porosity and permeability, giving rise to the concerns of CO2 storage capacity and security. Previous experimental investigation of a limestone covered with clay coating showed that the clay coating was progressively developed as calcite was dissolved. The dissolution rate of calcite was significantly decreased. The clay swelling decreased the permeability and enhanced the CO2 storage safety capacity. This study aims to analyze the acid–rock interaction, subsequent flow path evolution and ions distribution in the calcite–clay matrix systems during CO2 injection by implementing a multi–scale reactive transport modeling. The dynamic change of calcium ion was tracked because the produced calcium ion concentration decreased the local dissolution rate of calcite dissolution. The injected water mixed with calcium ions was unstable and resulted in the precipitation of CaSO4, causing scale formation. During the injection of CO2–enriched brine, the clay was regarded as nonreactive and considered its swelling effects. Other minerals such as quartz were regarded as nonreactive, while the calcite was reacted with carbonic acid. We elucidated the heterogeneity of mineralogical distribution on the transport pattern and mineral dissolution–induced porosity–permeability relationship. We considered the clay phase of mineralogy in CO2 transport and storage, which gives rise to a more accurate description of the underground geology process and prediction of long–term CO2 exhibition. We implemented immersed boundaries condition (IBC) to describe fluid zones, chemically reacted–phase zone, nonreacted–phase zone, and the plastic–solid swelling phase zone with irregular configurations by the structured grids. The IBC method took advantage of the uniform, structured grids while making it possible to solve the partial differential equations robustly and efficiently. Our simulation results can help establish and calibrate the Darcy–scale or field–scale model to track the CO2 trajectory, design, and optimum CO2 storage project.

To capture this complex sub–surface process, we coupled Darcy–Brinkmann–Stokes (DBS) method with geomechanics of the microporous plastic solid. The calcite phase was regarded as pore–scale in our model, where the clay phase was regarded as Darcy–scale. The DBS momentum equation was used to calculate the fluid flow in the porous media, and a momentum conservation equation for a plastic solid was used to describe clay swelling and deformation. Reactive transport was described by the multi–component advection–diffusion equation. The mass balance equations were used to update the volume fraction of the calcite phase and clay phase. The swelling pressure was obtained by the semiempirical formulation, which was determined by clay volume fraction and sodium concentration in the system. The solid effective viscosity was calculated by Herschel–Bulkley non–Newtonian plastic viscosity models. A series of partial differential equations were discretized by applying the Finite Volume Method (FVM). The multi–scale multi–species reactive transport model was developed based on the pisoFoam framework in OpenFOAM, an incompressible transient flow solver. The discretized algebraic equations were solved subsequently. To validate the numerical model in simulating the multi–species reactive transport, we compared the numerical results with the analytical solution of the Plane Poiseuille flow and Kinetic Decay–Chain, respectively. The numerical results were matched well with the analytical results, showing the confidence of our simulation model.

We analyzed the heterogeneity of mineralogical distribution in the ideal fracture–matrix models. The mineralogical matrixes were symmetrically distributed at the upper and bottom sides of the fracture. Several distribution patterns with the same porosity were considered in the investigation, such as clay coating, striped distribution, staggered pattern, random distribution, and clustered distribution. From the simulation, we have the following interesting findings. First, under different clay–calcite distributions, the local dissolution rate of calcite and calcium ion accumulation were various. For clay coating pattern, only slight dissolution was observed at inner calcite, because the clay distributed around the calcite swelled and subsequently reduced the permeability. The contacted surface area between CO2–enriched brine and calcite decreased. Due to the fluid flow limitation in the fracture, CO2–enriched brine cannot saturate well the fracture. When the calcite was mainly distributed at the downstream area, there was no apparent dissolution of calcite. However, when the calcite was mainly distributed around the inlet area, the calcite showed significant dissolution. Second, we also found that the clay swelling size had no significant relations with their distribution in the system. The clay distributed in the different areas has the same swelling size. However, under the action of fluid flow, the clay showed the deformation in the system. As clay clustered together, only outside layers of clay have obvious deformation. For the random distribution pattern, calcite was fully mixed with clay. Most clay in the system has deformation. Third, we observed meaningful phenomena regarding calcium ion concentration. Calcium ion concentration can reflect the underground reaction rate of dissolution. Total calcium ion concentration increased initially in the system. As calcite dissolved, calcium concentration decreased, and then kept a constant value. Nevertheless, the mineralogical heterogeneity has significant impacts on the total calcium. For example, the clay–coating pattern accumulated more amount in the system. The inner calcite was dissolved and produced calcium, while as the outer clay swelled, the permeability decreased, and significant amount of calcium was accumulated. When calcite was distributed all around the fracture, the calcium ion was not likely to be accumulated in the system. If it was the clay distributed all around the fracture, the generated calcium tended to accumulate around the upper and bottom wall due to the viscosity limitation of the boundaries. The porosity–permeability relationships of different patterns were obtained, and we presented a universal formula to describe the relationship by introducing the geometrical factors of the distribution pattern.

From the research, we found the several mineralogical patterns with better storage capacity, including clay distributions around the calcite, fracture, and upstream side. Clay swelling can significantly decrease the permeability and limited the transport of CO2–enriched brine. However, clay decrease the contacting area between the calcite and acid, preventing the dissolution of carbonate. Understanding the flow pattern and dissolution characteristics can ensure that CO2 has been stored in the target zone and whether CO2–enriched brine leaked through the channels. The pore–scale parameters from this investigation can help establish a Darcy–scale model to track the trajectory of CO2 in the subsurface, which can help enhance the CO2 sealing capacity and guarantee environmental security.