Résumé
In this study, we compare the hydrogeochemical response of two fractured reservoir rocks (limestone composed of 100 wt.% calcite and clast-poor sandstone composed of 66 wt.% calcite, 28 wt.% quartz and 6 wt.% microcline) in contact with CO (sub 2) -rich sulfate solutions. Flow-through percolation experiments were performed using artificially fractured limestone and sandstone cores and injecting a CO (sub 2) -rich sulfate solution under a constant volumetric flow rate (from 0.2 to 60 mL/h) at P = 150 bar and T = 60 degrees C. Measurements of the pressure difference between the inlet and the outlet of the samples and of the aqueous chemistry enabled the determination of fracture permeability changes and net reaction rates. Additionally, X-ray computed microtomography (XCMT) was used to characterize and localize changes in the fracture volume induced by dissolution and precipitation reactions. In all reacted cores, an increase in fracture permeability and in fracture volume was always produced even when gypsum precipitation happened. The presence of inert silicate grains in sandstone samples favored the occurrence of largely distributed dissolution structures in contrast to localized dissolution in limestone samples. This phenomenon promoted greater dissolution and smaller precipitation in sandstone than in limestone experiments. As a result, in clast-poor sandstone reservoirs, the larger increase in fracture volume, as well as the more extended distribution of the created volume, would favor the CO (sub 2) storage capacity. The different distribution of created volume between the limestone and sandstone experiments led to a different variation in fracture permeability. The progressive stepped permeability increase for sandstone would be preferred to the sharp permeability increase for limestone to minimize risks related to CO (sub 2) injection, favor capillary trapping and reduce energetic storage costs. 2D reactive transport simulations that reproduce the variation in aqueous chemistry and the fracture geometry (dissolution pattern) were performed using CrunchFlow. Given that calcite dissolution reaction at pH < 5 is transport controlled, the calcite reactive surface area used in the model had to be diminished with respect to the geometric surface area initially considered in order to fit the model to the experimental data. The fitted reactive surface area was higher under faster flow conditions, reflecting a decrease in transport control and a more distributed reaction in sandstone compared to limestone.