Résumé

Geologic CO2 sequestration is considered to be the most promising technique to reduce the concentration of greenhouse gases in the atmosphere. Among all the storage options, deep saline aquifers have the greatest potential and due to their worldwide occurrence can play a major role in reduction of carbon dioxide emissions. In sedimentary basins at depths below 800 meters, CO2 usually exists in supercritical condition (scCO2), which means that its temperature and pressure are above 31.1o C and 7.4 MPa, respectively, and it has a liquidlike density (500-800 kg/m3). Carbon dioxide then can be trapped in pore space of the storage formation and by reacting with minerals that form it, as well as dissolve in the in-situ fluids. Choice of a host rock is crucial for proper retention of scCO2, and sandstone reservoirs, which mostly are single-porosity systems, are usually considered. However, in some countries (e.g. Switzerland, Italy and Canada) limestone aquifers are widespread and have to be examined for the possibility of storage. The injection of significant amounts of CO2 into a limestone reservoir has to be carefully treated because it is usually a multipleporosity system with wide permeability variations. Also, because of the dissolution reactions, large pores can be created in calcite-rich limestones, which increases their permeability, but significantly decreases capillary trapping of CO2. In order to study the possibility of CO2 injection and storage in limestones, Calcarenite (Apulian limestone) was tested for full characterization of its poroelastic response under three limiting conditions: drained, undrained, and unjacketed. The rock with interconnected porosity of 33%, permeability equal to 3-5 mD, and less than 2% level of anisotropy was first tested in uniaxial compression. Dry elastic parameters were accurately measured using both strain gages and digital image correlation (DIC) data. Two types of measurements provided the same rock parameters: Young’s modulus E = 7.3 GPa and Poisson’s ratio  = 0.25. Hydrostatic compression experiments were performed on jacketed and unjacketed specimens: dry bulk modulus K was measured to be 4.8 GPa at 6 - 8 MPa hydrostatic pressure and unjacketed bulk modulus Ks˝ was found to be 42 GPa and constant throughout the loading. Biot coefficient was then calculated to be  = 0.89. Furthermore, plane strain compression experiments were performed under undrained conditions. Calcarenite’s Skempton’s B coefficient appeared to be increasing with saturation and the maximum achieved value at 6 MPa effective mean stress was B = 0.70. Undrained bulk modulus was measured to be smaller than the dry bulk modulus of the rock and the results of experiments performed under the same conditions were not reproducible. It means that the limestone behavior was not poroelastic even at low effective mean stress (up to 6 – 7 MPa) and can be related to the partial dissolution in water of the minerals forming the rock. X-ray CT scanning showed the significant increase in the rock porosity for the saturated specimens. Also, the decrease in P-wave velocity was observed with injection of water in the limestone. The experiments with water as the saturation fluid will be compared to the tests where scCO2 is injected; even stronger reduction of poroelastic parameters is expected to be observed.

Détails

Actions