Geochemical and mineralogical characterization of the Arbuckle aquifer: studying mineral reactions and its implications for CO[subscript]2 sequestration

Abstract

In response to increasing concerns over release of anthropogenic greenhouse gases the Arbuckle saline aquifer in south-central Kansas has been proposed as a potential site for geologic storage for CO2. Two wells (KGS 1-32 and 1-28) have been drilled to provide data for site specific determination of the storage potential of the Arbuckle. Cores from specific depths within Arbuckle (4164-5130) were utilized for study and flow-through experiments. Examination of formation rocks by thin section studies, SEM, XRD and CT scans was carried out to characterize the mineralogy of the core. Dominant mineralogy throughout the formation is dolomite with large chert nodules and occasional zones with pyrite and argillaceous minerals. Carbonate-silica contacts contain extensive heterogeneity with sulfide minerals and argillaceous material in between. Extensive vugs and microfractures are common. This study focuses on three zones of interest: the Mississippian pay zone (3670-3700), a potential baffle in Arbuckle (4400-4550) and the proposed CO2 injection zone (4900-5050). Drill stem tests and swabbed brine samples collected from 13 depths throughout the aquifer reveal a saline brine (~50,000-190,000 TDS) dominated by Na+, Ca2+ and Cl-. Elemental ratios of major cations with Cl- demonstrate a typical saline aquifer system. Cl/Br ratios reveal mixing between primary and secondary brines within the aquifer. Ca/Cl and Mg/Cl ratios suggest effect of dolomitization within the brines. δ18O and δ2H isotopes and Li/Cl ratios in the brine suggest the separation of upper and lower Arbuckle by a baffle zone. Swabbed waters provide Fe speciation data and reveal the importance of it in the system. Laboratory experiments carried out at 40°C and 2100 psi using formation core plug and collected brine identify reaction pathways to be anticipated when supercritical CO2 is injected. Results showed fluctuating chemistries of elements with Ca2+, Mg2+, Na+ and Cl- increasing during the first 15 hours, while Fe, S, and SO42- decrease. For the next 15 hours a reverse trend of the same elements were observed. Alkalinity and pH show inverse relationship throughout the experiment. We conclude that dominant reactions will occur between brine, CO2 and dolomite, calcite, chert, pyrite and argillaceous minerals. There is no perceived threat to freshwater resources in Kansas due to CO2 injection.