(760a) Response of Low-Permeability Carbonate Samples to CO2-Brine Exposure: Experiments & Modeling

Several large-scale subsurface carbon utilization and sequestration projects have been implemented in response to rising levels of atmospheric carbon dioxide (CO₂). Good candidate scenarios include former CO₂-enhanced-oil-recovery fields, which can be transitioned into long-term CO₂ storage sites given the presence of a sufficiently impermeable “seal” formation. Because of the high reactivity of carbonate minerals such as calcite and dolomite under CO₂-acidified conditions, carbonate reservoir formations can be expected to respond to CO₂ injection near the injection zone with generation of new pore space (and potential permeability enhancement), as CO₂-charged fluids partially dissolve carbonate host minerals until chemical equilibrium is reached. However, this same type of reaction can also lead to the formation of less desirable fast fluid pathways if the seal formation also includes carbonate minerals, and these preferential pathways could allow leakage of CO₂ into overlying geologic units.  A full understanding of the impact of chemical and mechanical stresses (e.g., fluid-mineral disequilibria, fluctuating reservoir conditions, etc.) resulting from CO₂ injection, and description and/or prediction of how these stresses will translate into meaningful changes in permeability throughout the storage reservoir, is an ongoing active area of research.

To address these questions, we present results from several core-flood experiments, conducted at in situ storage temperature and pressure, on well-characterized low-permeability carbonate core samples from two different carbon storage sites (Weyburn, Saskatchewan, Canada; Wellington, Kansas, USA). The purpose of these experiments was to explore the effects of physical (pore versus mineral distribution, connectivity of pore space) and chemical (variable CO_2(aq) concentrations, carbonate versus non-carbonate mineral distribution) heterogeneity on the evolution of permeability within these core samples. To quantify these attributes, each sample was imaged via X-ray microtomography both prior to and after each brine-CO₂ core-flooding experiment. Geochemical fluid samples were also collected throughout each experiment and interpreted to define each sample’s approach to specific fluid-mineral saturation states. The range of initial bulk permeability of these carbonate samples spans at least five orders of magnitude; however, it is evident that it is not necessarily the initial permeability value but rather the contrast in permeability between neighboring regions that controls the ultimate evolution of fluid flow within these samples. In addition to the characterization and experimental data, we also present reactive transport model simulations, which were conducted to calibrate the relationships between porosity, permeability, and carbonate dissolution rate laws. Our results, to date, indicate that a continuum-scale model can be used to describe the evolution of permeability in a reactive carbonate-brine-CO₂ system, provided that effective permeability can be reproduced well at the model element scale. Correlations between changing porosity and permeability values can also be constrained using the available experimental data. However, we find that even at the high resolution of the tomography data, difficulties arise in adequately describing the effective reactive specific surface area of minerals within these samples. Thus, additional adjustment of kinetic rate constant values is necessary in order to account for order-of-magnitude uncertainties in surface area estimations.