Different flow behavior between 1-to-1 displacement and co-injection of CO2 and brine in Berea sandstone: Insights from laboratory experiments with X-ray CT imaging

In this study, we compare the changes in CO2 saturation and pressure drop for two modes of injections: (i) a displacement of brine by supercritical CO2 injection (1-to-1 displacement), and (ii) a forced co-current injection of supercritical CO2 and brine at the same flow ratio (co-injection), respectively, in a laboratory core-flooding experiment using a Berea sandstone sample. The Berea sandstone sample showed a weak bedding structure that consists of high- and low- porosity layers. The main flow direction was set perpendicular to the bedding layers. We utilized the X-ray CT technique to image the rock interior and obtain the information of local porosity and saturation of each voxels in a three dimensional volume. The results show that the co-injection needs a much higher pressure drop to maintain the constant flow rate than the 1-to-1 displacement at similar saturation degrees. Moreover, the co-injection shows significant fluctuations in saturation and pressure drop, whereas the 1-to-1 displacement shows more gradual and monotonous changes. The spontaneous fluctuations in saturation and pressure drop during co-injection are basically coincident in temporal pace, and can be explainable by the intermittent flow of brine and CO2 as shown in differential saturation images. Furthermore, the X-ray images show that the CO2 mainly flows through built flow pathway during the 1-to-1 displacement; whereas the CO2 flows near uniformly and does not strictly rely on pore size and capillarity during the co-injection. CO2 saturation distributes more uniform among image pixels during the co-injection. These dissimilarities between the co-injection and 1-to-1 displacement suggest differences in fluid flow mechanism between them. In the 1-to-1 displacement, the pathway of CO2 flow was created by the forward motion of CO2 over the capillary pressure force. CO2 preferred to first percolating through large-size pores and gradually expanded the percolation region while the CO2 saturation grew. In this process, the displaced brine mainly flowed in its remained phase-pathway-less phase interference occurred. The connection of flow pathway for CO2 naturally satisfied the maintaining of the flow rate. In contrast, during the co-injection, both phases flowed in the pore space. The connection of the phase-pathway was affected by the phase snap-off effect. The transport efficiency of such a partially disconnected flow-pathway was significantly lower than the 1-to-1 displacement case, leading to that much higher drive force of pressure drop was necessary to let the fluids flow at setting rates. Nevertheless, the reachability of CO2 to low-porosity sites during the co-injection was higher than during the 1-to-1 displacement. Our findings are important for understanding of co-injections in applications of relative permeability measurement, and enhanced efficiency in capillary trapping, usage of pore space in CO2 geological storage and in oil recovery.