A Hypothetical Scenario for Full-Scale Deployment of Geological Carbon Sequestration: Investigating the Interaction Between Multiple CO2 Storage Sites in a Sedimentary Basin

Most ongoing projects of geological carbon sequestration (GCS) are relatively small in size, with annual injection rates from a few thousand to less than a million tonnes. These projects help build the GCS technology with respect to modeling, monitoring, risk assessment, and mitigation, and have been successful so far in terms of CO2 containment and caprock geomechanical integrity. In the future, GCS will be implemented at full-scale, multiple industrial-size CO2 storage sites in large sedimentary basins to make full use of the potential storage capacity. Simultaneous injection into multiple not-too-distant storage sites will lead to interference between the individual regions of pressure build-up and possible interference between the individual CO2 plumes. The Illinois Basin is used to model the future impact of multiple injection sites in the thick, extensive Mount Simon Formation. The basin-scale model domain of 241,000 km2 covers a core injection area of 24,000 km2, a larger near-field area where significant pressure buildup is expected, and an even larger far-field area for investigating environmental impacts on groundwater resources. The model assumes that there are twenty sequestration sites (spaced 30 km apart) within the core injection area. Three injection scenarios are considered, featuring annual injection rates of 5, 10, and 15 million tonnes of CO2 at each site, respectively. These scenarios correspond to 33%, 67% and 100% of the current single-point large CO2 sources in the relevant states (Illinois, Indiana and Kentucky). The model adequately captures the characteristics of the Mount Simon Formation in the core injection area, which include (1) an overall thickness of 300 to 680 m, (2) an upper unit of sandstone and shale tidally influenced and deposited, (3) a thick middle unit of clean sandstone of relatively high permeability, and (4) a lower arkosic unit of higher permeability (one Darcy) with an average thickness of 90 m. At each site, CO2 is injected into the lower arkosic unit to ensure sufficient permeability to accommodate high injection rates. A three-dimensional unstructured mesh is used for the model, with progressive refinement horizontally from the far-field area to the core injection area and radial refinement toward each injection center within the injection area, as well as progressive vertical refinement from four model layers in the far-field area to 50 model layers for each CO2 plume. The overlying Eau Claire seal and the underlying Pre-Cambrian granite unit are also included in the model. Both the two-phase CO2-brine flow within the twenty CO2 plumes and the single-phase brine flow away from the plumes are simulated using the parallel TOUGH2/ECO2N simulator. Preliminary simulation results are discussed with respect to (1) the dynamic evolution and migration of individual CO2 plumes, (2) the possible interference between different plumes, (3) the interference between individual regions of pressure buildup, (4) the possible impacts of GCS on the groundwater resources at the basin's boundary, and (5) the possibility of caprock damage.