REACTIVE TRANSPORT MODEL PARAMETERIZATION VIA A CO2 SEQUESTRATION ANALOGUE

This study constrains a reactive transport model using data from Mississippian Madison Group carbonates on the Moxa Arch, which hosts natural supercritical CO₂, CH₄, H₂S and He. Barren parts of the Arch are candidates for future anthropogenic gas sequestration. Extensive petrographic, elemental, isotopic, porosity, permeability and water chemistry data show significant compositional and petrophysical heterogeneity in the Madison on multiple scales. Stratigraphically, the reservoir compartmentalizes into numerous high-porosity (10-30%), 1-100 ft dolomite flow units, which are more abundant in lower Madison 3^rd order sequences (I to III). The units are often separated by lower porosity (0-5%), 1-5 ft limestone barriers with highly variable, fracture-based permeability. Porosity is a function of sedimentary protolith composition, dolomitization extent, and Laramide hydrothermal activity. Storage zone porosity is primary moldic or intercrystalline, while primary interparticle and fracture porosities in other lithologies have flow implications. Primary rock types are variably dolomitized mudstones to grainstones. Clay and silica fractions are only significant near upper and lower Madison contact. Anhydrite, however, is common, especially in lower 3^rd order sequences, and is frequently affiliated with hydrothermal calcite, dolomite and fluorite. Carbonates show evidence of early bacterial (in reduced carbonaceous mudstones) and later thermogenic sulfate reduction (in reservoir dolomites), Fe-rich protoliths yielding pyrite and Fe-poor protoliths yielding H₂S or native sulfur. Given 1) relatively fast dissolution and precipitation kinetics of sulfur phases, 2) their influence on reservoir pH and redox, 3) the role of pCO₂ in sulfate reduction processes, and 4) the occurrence of SOx gases in many carbon waste streams, including SO₄ in the modeling matrix is critical. Reservoir water chemistries reflect basin brines (TDS~ 2e⁴-6e⁴ mg/kg) immersed in a dolomite host, but require correction for CO₂ degassing, and possibly scaling, during transport. The above considerations warrant a minimum 8-component model matrix consisting of HCO₃-H-Na-Cl-SO₄-Ca-Mg-Fe. Flow models should also address to what extent limestone baffles disperse or inhibit CO₂ migration before the upper seal is reached.