Flow Behaviour of Supercritical CO2 and Brine in Berea Sandstone During Drainage and Imbibition Revealed by Medical X-ray CT Images

Yi Zhang,Osamu Nishizawa,Tamotsu Kiyama,Shun Chiyonobu,Ziqiu Xue

DOI: https://doi.org/10.1093/gji/ggu089

IF: 3.352

2014-01-01

Geophysical Journal International

Abstract:We injected Berea sandstone with supercritical CO2 and imaged the results with a medical X-ray computed tomography (CT) scanner. The images were acquired by injecting CO2 into a core of brine-saturated sandstone (drainage), and additional images were acquired during reinjection of brine (imbibition) after drainage. We then analysed the temporal variations of CO2 saturation maps obtained from the CT images. The experiments were performed under a confining pressure of 12 MPa, a pore pressure of 10 MPa and a temperature of 40 A degrees C. Porosity and CO2 saturation were calculated for each image voxel of the rock on the basis of the Hounsfield unit values (CT numbers) measured at three states of saturation: dry, full brine saturation and full CO2 saturation. The saturation maps indicated that the distributions of CO2 and brine were controlled by the sub-core-scale heterogeneities which consisted of a laminated structure (bedding) with high- and low-porosity layers. During drainage, CO2 preferentially flowed through the high-porosity layers where most of the CO2 was entrapped during low flow-rate imbibition. The entrapped CO2 was flushed out when high flow-rate imbibition commenced. Plots of the voxel's CT number against porosity revealed the relationship between fluid replacement and porosity. By reference to the CT numbers at the full brine-saturated stage, differential CT numbers were classified into three bins corresponding to voxel porosity: high, medium and low porosity. Distributions of the differential CT number for the three porosity bins were bimodal and in order with respect to the porosity bins during both drainage and imbibitions; however, the order differed between the two stages. This difference suggested that different replacement mechanisms operated for the two processes. Spatial autocorrelation of CO2 saturation maps on sections perpendicular to the flow direction revealed remarkable changes during passage of the replacement fronts during both drainage and imbibition, changes reflecting the interfingering pattern across the replacement fronts. Although the permeability differences between high- and low-porosity layers were not sufficiently large to disturb the uniform flow of brine, the CO2 concentration in the high-porosity layers may have been caused by the differences of capillary pressure between wide and narrow pore throats, perhaps enhanced by an invasion percolation mechanism in flow-path networks.