Multiscale clay in sustainable geosystems

Bentonite clay is considered as an engineered barrier in the disposal of high-level radioactive waste because of its unique properties. The thermal gradient generated by radioactive decay is expected to lead to coupled thermal-hydrologic-mechanical-chemical (THMC) processes that may impact barrier performance. We developed methodologies based on high-performance molecular dynamics (MD) simulations to predict the microscale material properties of compacted montmorillonite clay (Zheng et al., ACS, 2023). Simulation predictions were compared with experimental results on bentonite material properties measured on scales of centimeters and days. This research provides new insight into the coupled THMC properties of clay barrier systems and helps the evaluation of clay barriers performance over a long timescale (Zheng and Bourg, ACS Nano, 2023).

The assembly of smectite clay platelets, tactoids, and aggregates leads to the formation of diverse hierarchical structures at nanometer to micrometer scales. Understanding of clay tactoid formation and clay microstructure remains incomplete, however, as evidenced by the existence of significant discrepancies between experimental datasets. In this work, we utilize a new coarse-grained molecular dynamics (CGMD) simulation model to generate equilibrated configurations of 2,000 montmorillonite clay platelets to investigate suspension, aggregation, and disassembly of clay tactoids on a scale of 0.1 micrometers in length. Simulation results are analyzed to predict rheological properties as well as structural properties of smectite (such as the number of platelets per tactoid) as a function of the distribution of exchangeable cations. This research advances efforts to understand the relationship between the microstructure and rheology of smectite clay particles in dilute suspension (Zheng and Bourg, in preparation).

Understanding the coupling between multiphase fluid flow in pores with distinct sizes and solid deformation induced by flow or external stresses is crucial for the development of many important geotechnics. The recent so-called Darcy-Brinkman-Biot (DBB) framework can capture capillary, viscous, inertial, interfacial, and gravitational forces at both the pore and Darcy scales. In this work, we build upon previous studies to extend the DBB framework to compressible non-isothermal fluid flow. The model’s numerical implementation, hybridBiotThermalInterFoam, is achieved in the Computational Fluid Dynamics (CFD) software OpenFOAM. This model is then rigorously validated via an example application in the engineered barrier system (EBS) in a nuclear waste repository. Results show that the new solver is capable of predicting fracture propagation and healing in bentonite buffers exposed to strong thermal fluxes and complex aqueous chemistry conditions. The development in this work creates the first model representing multiphase compressible non-isothermal fluid flow in multiscale deformable porous media (Zheng and Bourg, in preparation).

Faults are key geologic components defining fluid migration pathways in sedimentary basins. I ran experiments to quantify the effects of grain size, porosity, and clay content on the transport properties of smectite-rich fault gouge in faults, including CO2 breakthrough pressure and post-breakthrough CO2 permeability. The results of this work helped quantitatively evaluate fault sealing capability and migration of buoyant fluids through faults in sand-shale sequences (Zheng & Espinoza, JMPG, 2021). In addition, the injection of fluids into a compartmentalized formation induces pore pressure buildup and may result in fault reactivation. I measured the new uniaxial strain unloading compressibility of unconsolidated Frio sand to predict the pressure increase during fluid injection. My results generate useful guidelines for subsurface fluid injections to prevent excessive pressure buildup in storage formations (Zheng et al., ARMA, 2019; Zheng & Espinoza, RMRE, 2021).

Pressure monitoring above the injection zone is a potential method to detect potential CO2 leaks into overlying formations. I applied a compositional simulator coupled with geomechanics to predict pressure changes above the caprock due to both fast hydraulic communication and partially undrained loading. My results reveal that pressure monitoring above the caprock is a feasible technology to track the CO2 plume and interpretation of pressure signals in the field must account for partially undrained poroelastic loading (Zheng et al., ARMA, 2021; Zheng et al., IJGGC, 2022). Moreover, the presence of fault heterogeneity can lead to different trapping capacities for the same structure. I developed two stochastic models including a continuous shale gouge model and a discrete smear model to statistically determine the possible range of CO2 column height in sand-shale sequences. The results enable effective prediction of CO2 column height in the presence of heterogeneous clay smears in faults (Zheng & Espinoza, JMPG, 2022).