Research Plan/Project Summary
Scienteer Project ID:
Shaunak Pandey
Project Title:
Mathematical Modeling of CO2 Migration and Trapping in Geologic Formations for Risk Assessment of Carbon Sequestration impacting Carbon Footprint
School Name:
School City:
Austin, Texas, US

Over the last 150 years, the average global temperature on Earth has rapidly increased. This detrimental impact has created–among many others–widespread famine, the melting of ice caps, larger, more frequent storms, drought, and poverty. The cause of such disaster lies behind the hands of humans. Due to the heavy reliance on fossil fuel mining and usage, greenhouse gasses are emitted into the atmosphere, including methane, carbon dioxide, nitrous oxide, etc. Geologic carbon sequestration (GCS) has emerged as a promising technology to mitigate the environmental impacts from increased atmospheric CO2. GCS involves the process of artificially injecting CO2 into underground geologic formations such as saline aquifers and depleted oil and gas reservoirs. One of the main advantages of GCS is that it allows us to continue using fossil fuels as a source of energy while mitigating their environmental impacts. Fossil fuels are currently the primary source of energy for most countries around the world, and it is unlikely that we will be able to transition to renewable energy sources overnight.

To guarantee safe and long-term GCS, it is essential to find and determine optimal porous storage formations capped by low-permeability layers called caprocks as geologically sequestered CO2 could potentially encounter unidentified (or unwanted) pathways (e.g., boreholes, microfractures within caprocks, and fault zones), which could serve as CO2 leakage pathways directly connecting deep storage formations to shallow potable aquifers or to the atmosphere. Occurrence of such CO2 leakage from the storage formation could result in the failure of GCS and also has a detrimental effect on public and environmental health.

The flow of CO2 through microfractures is critical to ensure a secure CO2 storage and quantifying potential leakage rates. Rock fractures are naturally rough-walled, which highly affects the CO2 flow behavior and causes local trapping of CO2 on the fracture surface. The physics governing the dynamics of multiphase flow of CO2 and water and capillary trapping of CO2 in rough-walled fractures have remained largely unexplored. In this study, computational fluid dynamics and high-performance computing systems will be leveraged to study the pore-scale controls such as the fracture aperture size, surface roughness, and wettability on transport and trapping of CO2 in rough-walled fractures. 


This experiment will leverage computational fluid dynamics and high-performance computing to explore the dynamics of CO2 transport and trapping in rough-walled fractures focusing on the effect of fracture geometric factors such as aperture, roughness, and wettability. 


Pore-scale two-phase flow simulation of CO2 and water in 3D fracture geometries will be performed using a computational fluid dynamics (CFD) toolbox called OpenFOAM. An OpenFOAM solver called interFoam will be used for two-phase immiscible fluid flow simulations. Prior to running the simulations, a hexahedral mesh (computational grid cells) will be created in the fracture geometry. An OpenFOAM solver called SnappyHexMesh will be used for that purpose. The pre-processing of the simulations including the generation of fracture geometry will be done with Matlab. To improve the computational efficiency, the simulations will be carried out on different computational nodes of the high-performance computing (HPC) system at the Texas Advanced Computing Center (TACC) at UT-Austin. 
The post-processing of the simulation outputs will be performed with a visualization application called ParaView.  


1. Generate fractal geometry via Matlab: 
Use a mathematical model called fractional Brownian motion (FBM) to generate roughness on the fracture surfaces. There is a random parameter called Hurst exponent in the FBM method that controls the degree of roughness (fractality) in the geometry, with a higher value leading to a smoother surface. Using this parameter, different fractures will be generated with various surface roughness. The aperture size of the fracture is another controlling factor for generating different geometries. 

2. Construct computational mesh (grid cells) in the fracture geometry to be used in fluid-flow simulations
Will use a solver called SnappyHexMesh to generate the grid cells in the simulation domain. Coarse and fine meshing was used in order to ensure accuracy while maintaining high computational accuracy. 

3. Pore-scale two-phase flow simulations via OpenFOAM toolbox and high-performance computing 
Will perform pore-scale two-phase flow simulation of CO2 and water in 3D fracture geometries using a computational fluid dynamics (CFD) toolbox called OpenFOAM. An OpenFOAM solver called interFoam will be used for two-phase immiscible fluid flow simulations. To improve the computational efficiency, the simulations will be carried out on different computational nodes of the high-performance computing (HPC) system at the Texas Advanced Computing Center (TACC) at UT-Austin. 

Data Analysis

The visualization of the fluid flow simulation outputs will be done with an application called ParaView. Snapshots of CO2 plume distribution can be created in the fracture at different time-steps of the simulation. An animation video will be created to show the dynamics of CO2 fluid flow in the fracture during the simulation time. 
Using the ParaView software, the saturation profile of CO2 will be exported in all grid blocks of the simulation domain as a .CSV file and plot the time-evolution of CO2 saturation and CO2 trapping in the fractures with different geometries and under different wettability conditions.  

Project Summary
Changes from Original Research Plan


Project Conclusions

Pore-scale two-phase flow simulation of CO2 and water in 3D rough-walled fractures were performed using a computational fluid dynamics (CFD) toolbox called OpenFOAM, for CO2 leakage risk assessment. Fracture geometry, including surface roughness and fracture aperture size, wettability, the ability of a liquid to spread over a surface, measured by the contact angle between the liquid and the surface, and injection rate were manipulated and their effect on the dynamics of CO2 flow was studied. The key findings from the simulation results are concluded as follows.

It was found that the roughness of the fracture surface and the wetting condition form capillary force fields which control the displacement of CO2 and water in the fracture. Aperture size was another factor which controlled the migration of CO2. When the fracture aperture was selected to be tight, CO2 bypassed the water in the high capillary pressure zone. While for a wide water-wet fracture, CO2 tends to migrate in the middle section of the fracture opening and thin film of residually trapped water occurs on the surface of the fracture. When the wetting condition shifted towards less water-wet, a larger degree of CO2 capillary trapping on the surface of the fracture was observed. An unstable and fingering type of CO2 displacement by increasing the injection rate in a rough fracture meaning that the fracture roughness is a key factor in controlling the CO2 displacement regime in fractured reservoirs. The high-fidelity pore-scale simulations using high-performance computing presented in this study provide insights towards dynamics of CO2 migration in fractured reservoirs and elucidate the fact that microscale properties control the macroscale phenomena. Such simulations can be used as a robust tool to explore a broad space of numerous parameters and scenarios to investigate the behavior of stored CO2 under various conditions.

In summary, roughness impacts phase distribution, flow patterns and relative permeability curves in a single rough-walled fracture. Rough surface increases two-phase interference and flow resistance, and an unsteady flow state appears for the non-wetting phase (NWP) when the wetting phase (WP) saturation is small. In addition, the NWP flow can either be lubricated or be resisted by the WP depending on the flow patterns and the WP saturation.

Fracture wettability was assumed to be hydrophilicity in this study, and thus the results may not apply to hydrophobic or mix-wetting fractures. For true naturally fractured reservoirs, the spatial distribution and connectivity of the fracture network may be more complex. Therefore, simulations on gas and water flow through true fracture networks are anticipated to reveal the two-phase interaction characteristics under the influence of various factors such as fracture network patterns and flow rates. The pore-scale network models provide valuable insights into how the fluid properties, aperture characteristics, dynamic flow conditions, and the initial state of the rock-fluid system affect the behavior of fluid flow and capillary trapping in natural rock fractures. These insights are of fundamental and practical importance to better understand multi-phase flow mechanisms and engineer subsurface flow applications, such as oil recovery, carbon sequestration, and contaminant remediation in fractured formations.

Reference Source

Yi, J., Xing, H., Wang, J., Xia, Z., & Jing, Y. (2019). Pore-scale study of the effects of surface roughness on relative permeability of rock fractures using lattice Boltzmann method. Chemical Engineering Science, 209, 115178.

Han, G., Han, W. S., Kim, K.-Y., Park, J. G., Piao, J., & Yun, T. K. (2019). Roles of fault structures and regional formations on CO2 migration and distribution in shallow saline aquifer in Green River, Utah. Journal of Hydrology, 570, 786–801.

Zhu, Q., Yang, Y., Zhang, X., Wang, S., Yang, J., & Zhang, J. (2022). Pore-Scale Simulation of Gas and Water Two-Phase Flow in Rough-Walled Fractures Using the Volume of Fluid Method. Energies, 15(24), 9382.

Transport in Porous Media. (2021). Springer; Springer Netherlands.

Gong, Y., Sedghi, M., & Piri, M. (2021). Dynamic Pore-Scale Modeling of Residual Trapping Following Imbibition in a Rough-walled Fracture. Transport in Porous Media, 140(1), 143–179.