The impact of small scale rock heterogeneity on CO2 plume migration and residual trapping

The lateral migration of the CO2 plume is often not well predicted by reservoir simulators where small-scale rock heterogeneity is not taken into account. Moreover, small-scale rock heterogeneity has a significant impact on the residual trapping potential of the reservoir rock. In this research, we have measured and characterized the impact of small-scale rock heterogeneity on multiphase flow parameters and residual trapping of CO2. Furthermore, we have developed methods to find upscaled effective parameters that can be used to incorporate the effects of small-scale heterogeneities into larger scale models. This could significantly improve the prediction of plume migration which will increase our confidence in assessing storage capacity and leakage risks. For this purpose, we performed experimental and numerical CO2/water core-flood tests at reservoir conditions for a range of sandstone rock cores (including the Fontainebleau, Berea, Bentheimer, Massillon and several cores of the Paraatte formation) containing different degrees and types of heterogeneity. The experimentally obtained CO2 saturation distributions, visualized with the use of a medical X-ray CT-scanner, were used to construct a model of the sub-core scale permeability field for each of the sandstone rock cores studied. The interplay of gravitational, viscous and capillary forces determines the CO2 saturation distribution and the amount of residual trapping that takes place. The relative permeablity function is flow rate dependent as a result of this. We used the models of the rock cores to simulate the core-flood tests for a range of flow rates and quantified the relative importance of each of the forces at the scale of the heterogeneity. The results show that the direction, magnitude and spatial extent of the heterogeneity impact the local capillary forces and, therefore, the capillary pressure and saturation distribution. From the numerical simulations upscaled flow-rate dependent effective relative permeabilties were obtained which can be used to incorporate flow rate dependency into larger scale models.