Effect of dissolved H2SO4 on the interaction between CO2-rich brine solutions and limestone, sandstone and marl

Abstract

The effect of pure and impure carbon dioxide (CO2) on geological storage is uncertain. Oxidation of impurities such as sulfur dioxide (SO2) and hydrogen sulfide (H2S), which may be catalyzed by co-injected oxygen or nitrogen oxides, leads to the formation of sulfuric acid (H2SO4) and a decrease in pH of the formation water. We investigated the effect of 0.005 mol L− 1 H2SO4 (corresponding to a worst case of 0.4% SO2 in the flue gas, anticipating total conversion of SO2 into H2SO4) on the reactivity of the reservoir (limestone and sandstone) and cap (marl) rocks of Hontomin (Spain) at P = pCO2 = 10 bar and 60 °C during 24 days using flow-through column experiments. Aqueous element concentrations were measured from fluid extracts obtained periodically throughout the experiments to infer fluid-rock reactions over time. Results were modeled with the CrunchFlow reactive transport code. The added H2SO4 lowered the pH of the injected brine by ~ 1.5 pH units with respect to the pH of ~ 3.6 of the H2SO4-free brine. In both H2SO4-free and H2SO4-rich brine, calcite dissolution fostered gypsum precipitation. A comparison between the reactivity of the rocks reacted in H2SO4-free and H2SO4-rich brine showed that calcite dissolution and gypsum precipitation rates were increased by 27–48% and 25–75%, respectively, in H2SO4-rich brine. Overall rock porosity increments in H2SO4-rich brine were 2.9–3.6%, 3.7–4.8%, and 2.1–2.7% for sandstone, limestone and marl, respectively. Porosities in H2SO4-rich brine were 6%, 23% and 250%, higher, respectively, than under pure CO2. Modeled porosity increments in the acid inlet zone in H2SO4-rich brine for sandstone, limestone and marl were 28–29%, 44–45% and 24–28%, respectively, corresponding to an increase of 1%, 250% and 25%, respectively, relative to H2SO4-free brine. Gypsum precipitation was consistently higher in marl than in limestone and sandstone, indicating kinetically favorable conditions for gypsum precipitation within the cap rock. Our results provide relevant data for long-term storage simulations of impure CO2 injection.