Geochemical effects of an oxycombustion stream containing SO2 and O-2 on carbonate rocks in the context of CO2 storage
Résumé
This paper describes the effects of the injection of a CO2-dominated gas mixture into a geologic reservoir rock through experimental work in the context of limiting greenhouse gas emissions into the atmosphere. The injected gas mixture consists of the exhaust fumes from an oxyboiler without desulphurisation with the following mole fraction composition: CO2 = 0.82, SO2 = 0.04, O-2 = 0.04, N-2 = 0.04 and Ar = 0.06. Corresponding experiments using pure CO2 and N-2 were performed as a benchmark. The rock sample was obtained by drilling to a depth of 4600 m into a low-porosity dolostone reservoir containing micrometric to centimetric fractures in the south-west of France (northern Pyrenees). The fracture network represents the primary volume available for CO2 storage and is partly filled with dolomite separated from the rock matrix by a thin layer of calcite covering the wall rocks. Experimental reactivity of the rock was tested in 2-cm(3) batch reactors in the presence of saline water (25 g/l NaCl) and a gas phase (pure N-2, pure CO2 and gas mixture). Chemical analyses of the reacting solutions indicated that the mineralogic assemblage during exposure to pure CO2 was in equilibrium with the aqueous solution. Raman analyses of the gas phase revealed only the presence of CO2. Optical and electronic microscopy of the resultant solid phases indicated partial dissolution of carbonates and oxidation of the pyrite surfaces. In the presence of the gas mixture, important mineralogic alteration occurred together with the consumption of half of the O-2 and total consumption of the SO2. This high reactivity with the gas mixture leads to the complete dissolution of calcite and partial dissolution of dolomite and the precipitation of anhydrite and barite, particularly in the zones where the calcite was initially present. Similarly, pyrite was completely oxidised to hematite. Analyses of the rock samples indicated partial alteration of the clay minerals in the matrix to potassic beidellite in the experiment involving pure CO2 and solely to vermiculites in the gas mixture experiment. In conclusion, the presence of SO2 in the injection stream associated with the presence of O-2 results in an early strong acidification of the water, which was buffered by the significant reactivity of the carbonates (dissolution of all of the calcite and 6% of the dolomite) and partial alteration of the clay minerals (87% of illite and 100% of smectite) to vermiculites. Pyrite and aqueous Fe from clays were completely oxidised by O-2, resulting in hematite and Fe3+. The mineralogic alteration and consequent volume changes under experimental conditions led to a slight increase in the porosity of the dolomite matrix and an average pore volume loss of 11% in the fractures caused by the replacement of calcite with anhydrite. Due to its high reactivity with carbonate, SO2 can react early during the injection phase. The spatial distribution of calcite in the fractures of the reservoir has to be considered as one of the primary parameters controlling the evolution of the reservoir in terms of injectivity and petrophysical properties, particularly in the zone near the injection wellbores. The integrity of the calcite-rich caprock is, however, ensured due to its thickness (more than 1000 m) and its calcite content, which leads to pH buffering and anhydrite precipitation, which, in turn, induces a porosity reduction and a possible coating of the rock formation.