Effect of Heterogeneity on CO2 Sequestration

Summary:
We present a pore-scale tomographic study of CO2 trapping mechanism in heterogeneous sandstone rocks with two different types of lamination. The in-situ experiments consist of scCO2 injection (drainage) and brine injection (imbibition) using a high temperature/pressure triaxial flow cell. X-ray micro Computed Tomography (XCT) is utilised to resolve grains, brine and CO2 in the system.

The purpose of this experimental research is to understand the role that heterogeneity (as a significant outstanding uncertainty) plays in plume migration and scCO2 storage safety in water-wet sandstone rocks. The experimental setup is composed of (1) high pressure/temperature triaxial flow cell (HPTC) and related pumps and hydraulic system (2) x-ray micro computed tomography (XCT) to map fluid distribution in sandstone rocks in 3D. Two cylindrical sandstone cores 36 mm lengths and 12 mm diameters were installed in Australian National University
(ANU) HPTC. The experiments were carried out at flow rates between 0.1 cc/min to 1 cc/min providing a constant pressure differential of 200 PSI across the sub-cores. The scCO2/brine fluid system and injection processes are monitored and recorded using an in-situ setup combined with XCT. Scans of the system are obtained at various stages of the experiment to monitor and record changes in the fluid saturations at elevated temperatures and pressures, representative of sub-surface storage conditions. The working fluids in this experiment were scCO2 and KI brine. The core was initially fully saturated with brine followed by scCO2- saturated 0.5 M KI brine injection as primary drainage, representing the initial migration of CO2 through the storage reservoir. This was followed by a water-flooding step, where brine saturated with scCO2 was injected then chase brine flooding commenced. scCO2 and brine flooding was carried out by injecting at a constant upstream pressure of 1350 PSI, with the downstream back pressure regulator set to 1550 PSI. When brine is artificially injected: some scCO2 is displaced, leaving behind an immobile residual fraction. scCO2 flow rates under these conditions exhibited a period of low flow until approximately 17 pore volumes of scCO2 had been injected; at this point a consistent scCO2 pathways was established through the layers in the sub-core. Injection of scCO2 flow was terminated when the effluent fluid was pure CO2. We performed two different set of initial brine saturation i.e. drainage and imbibition experiments on two different sandstone rocks with clear layering and scattered bioturbated layers of fine elements. XCT scans were acquired at the endpoint of each experimental step via region-of-interest acquisition under helical space-filling trajectory. Each high-resolution scan (voxel size of 4.83 microns) required approximately 12 hours to complete. The tomographic series were reconstructed using in-house algorithms and the 3D datasets were analysed for mapping spatial distribution of CO2 and brine.