The use of geological structures as repositories for the byproducts of fuel consumption such as CO₂ and nuclear waste will perturb the subsurface systems significantly, driving the hydrogeologic and the biochemical cycles far from equilibrium.  Moreover, earth materials are complex and extremely heterogeneous at length scales from 10^-8m to 10⁶m, and they exhibit spatial correlations from the pore scale upwards.  The transport of multiple fluid phases and species within those phases is strongly coupled to flow, reaction, and deformation.  The coupling lead to rich behaviors even in trivial domains.  

Subsurface modeling and simulation have typically relied on macroscopic continuum dynamics for prediction and assessment. Continuum scale models often poorly represent the heterogeneity and complexity existing at the pore scale. Moreover, the averaging inherent in continuum models can eliminate incipient structures from which large-scale patterns emerge, such as preferential flow paths. This lack of a basic understanding inhibits our ability to predict quantitatively the in situ response of rock masses during short-term human activities such as injecting CO₂ into a deep, saline aquifer (tens of years), and during longer-term processes such as CO₂ plume immobilization (hundreds to thousands of years).

To what extent do depositional heterogeneity, diagenetic alteration, and engineering- induced disequilibria each influence flow patterns of reactive perturbations? 
 For some coupled phenomena, a far-from-equilibrium perturbation leads to self-reinforcing behavior that evolves at characteristic length scales.

In bicontinuous material, how does structural deformation of the solid affect the flow behavior through the void?
Pore-scale reconstructions, when combined with modeling and experimental methods, can reveal how microscale flow and deformation processes in complex geomaterials affect the macroscale behavior of rock masses. 

Can pore-scale and continuum-scale models be integrated seamlessly?
Pore-scale models rarely consider their interaction with the continuum scale, and these boundary conditions may control the pore-scale behavior.

How does pore-level heterogeneity of grain surface chemistry lead to large-scale patterns in multiphase fluid flow? 
A complete understanding of the physics and chemistry that enables a non-wetting phase to flow or to be trapped in geologic porous media is essential to any determination that geologic storage is feasible.

Can the influence of interfacial energy on multiphase flow properties be described as a nonequilibrium process?
 Suitably treated nanoparticles exhibit an affinity for the interface between an aqueous and nonaqueous phase and can be used to test long-held assumptions of multiphase flow in porous media.

Can perturbation of the biogeochemical cycle and of the  cause qualitatively different patterns of large-scale transport?
The anticipation of rare events with disproportionately large impacts is crucial to risk assessments and the evaluation of storage system performance.  

Our approach to answering the research questions consists of an integrated experimental, theoretical, and numerical modeling effort. Experiments are being used to determine the patterns of displacement in naturally driven systems, and how these patterns change as a function of rock parameters, flow rates, and chemical enhancements such as nanoparticles.   

Theoretical arguments are exploring different potential analytic upscaling from the micro-scale to the centimeter and meter scale, in attempt to model the correct large scale behavior from the small scale physics.  

 Numerical modeling is proceeding through resolving capillary driven displacements at the pore-scale (with different surface chemistry and geometric pore structures), and integrating these pore-scale models to continuum aquifer displacement models.   

This gives rise to better understanding of coupled physical processes that operate across a vast range of scales, from the atom to large geological formations, from nanoseconds to millennia.

Application of finite element mortars enables inexpensive upscaling of large heterogeneous regions from network models that resolve pore scale features within the regions. We are modeling transport near wells entirely at the pore scale to demonstrate that direct upscaling could give faulty results (Figure 1). This will provide insight to our long-term goals of developing more accurate a-priori upscaling techniques.   

 

 

Figure 1 (a) A. Multiscale simulator that utilizes pore-scale models near a well and continuum grid cells in the remainder of the domain.  (b). Networks coupled using mortars to ensure continuity of interface pressures and fluxes.  

Preliminary core-scale experiments indicated increased resistance when a nonwetting phase (analog to CO₂) invaded rock saturated with nanoparticle-laden brine; this supports the novel concept of treating formations to be self-sealing against leaking CO₂. We will test this concept in fractures and within the CT in the next year. 

Figure 2. Three successive tomograms of gas (cooler colors) invading an initially brine (hotter colors) in a 4” diameter core.    

Application of finite element mortars enables inexpensive upscaling of large heterogeneous regions from network models that resolve pore scale features within the regions. We are modeling transport near wells entirely at the pore scale to demonstrate that direct upscaling could give faulty results. This will provide insight to our long-term goals of developing more accurate a-priori upscaling techniques.          

Developed experimental technique to measure adsorption of surface-treated nanoparticles on materials with well defined pores and surface area; this helps predict how far such particles can propagate into subsurface formations for applications such as self-sealing. We will apply the technique to a wide range of nanoparticles in the coming year. 

Successfully distinguished aqueous dispersions of nanoparticles (1 wt % to 23 wt %) from brine in a 1 ft limestone core using X-ray tomography; this establishes a powerful capability to attribute variations in flow resistance to presence of nanoparticles. We will conduct and image multiphase displacement experiments in the coming year. 

Procured and built new HPHT visualization cell with servo-controlled load frame in which coupled mechanics, multiphase flow, and reactive transport involving super critical CO₂ can be monitored by laser scanning confocal microscopy. We will conduct experiments on crack propagation and multiphase flow in the coming year.

Investigation of chemical and mineralogical interactions between CO₂-laden brine and sandstone at natural CO₂ seeps shows that porosity occlusion by cement is offset by simultaneous opening of fractures; this proxy for long-term fluid-rock interaction in reservoirs of anthropogenic CO₂ suggests a new form of self-regulating emergent behavior may emerge from the mechanical/chemical coupling. We are working on detailed petrography, CT pore-scale imaging, and porosity/permeability analyses of naturally CO₂-altered sandstone.

Emergent Behavior during Radionuclide Transport:  

Anomalous reactive transport in the framework of the theory of chromatography
We analyzed the rapid solute migration of radionuclides in the framework of the theory of chromatography and we defined the conditions under which the phenomenon occurs (Prigiobbe et al. (2012) Transp. Porous Med., accepted).  

Figure 3 Emergence of the anomalous wave from a dispersed interface at Pe = 160 at different PV (pore volume injected). Upper figures show the evolution of the strontium (C_Sr) and the effective anion (C_a) concentration profiles versus the self-similar coordinates (x=x/t), where C_a is the difference between the chlorine and sodium concentrations. The lower figures show the corresponding pH profile and the retardation of strontium.  

Experimental evidence of the anomalous reactive transport of strontium in a porous medium
We performed column-flood experiments using a Sr²⁺-Na⁺-Cl^- aqueous system and a column packed with silica beads coated with hydrous ferric oxide (HFO). Under the conditions defined in Prigiobbe et al. (2012) for the anomalous rapid transport, we observed a retarded front as predicted by the theory and a pulse of Sr²⁺ travelling at the average fluid velocity (anomalous wave) which is not predicted by the thery. Our experiments demonstrate that radionuclides can travel at the average fluid velocity even if the injected high-pH solution is strongly adsorbed by the porous medium.  

Figure 4 Measured concentration histories as a function of PV (pore volume injected). a Sr²⁺ concentration profile exhibits a pulse traveling at the average fluid velocity and a retarded front as predicted by the theory of chromatography; b three concentration profiles of Sr²⁺ and Na⁺, where the latter behaves as conservative tracer, measured during replicate experiments.