Core scale investigation of fluid flow in the heterogeneous porous media based on X-ray computed tomography images: Upscaling and history matching approaches

In this paper, experiments and simulations were performed on outcrop samples from Lagoa Salgada in Rio de Janeiro, Brazil, as a possible analog to one of the most typical Brazilian Pre-salt carbonate reservoirs rocks. The rocks were microbial carbonates where plugs comprising two main facies were sampled, simplified as fine-grained and vugular facies. The plugs were utilized to study the impact of pore geometry with both experi-mental and simulation approaches on recovery factor, saturation profile, and relative permeability estimations. To provide direct visualization of the geometry, description of pore structure, and calculation of concentration profiles, computed tomography (CT) imaging was integrated with experimental measurements of petrography and core flooding. The injection of two pore volumes of formation water resulted in a recovery factor between 28 and 34 percent for the plug samples. Furthermore, based on porosity generated by dry and wet CT, as well as saturation profiles resulting from CT data collected during drainage and imbibition processes along the length of the plugs, it is revealed that the distribution of these properties was diverse and heterogeneous. An algorithm was used to process the 2D tomography images of the samples to remove the region related to the exterior parts. The images were then stacked to create a 3D fine-scale grid to simulate the porous media and the fluid flow by applying rules for the segmentation of rock types, porosity, and permeability estimations of each grid block. Course-scale grids were created by applying upscaling techniques to reduce computation time. Simulated pro-duced fluid cuts for different upscaled models were compared with the experimental results from core flooding. A history-matching technique was then applied to match experimental and simulation results, calculating the relative permeability of two main defined facies and creating an updated model capable of assessing past and present performance and future forecasting. Since relative permeability is essential for accurate simulation, estimating these curves in the heterogeneous pre-salt reservoir considering different facies, greatly influences reasonable future prediction performance and the ability to make informed operational decisions.