A Robust Model To Simulate Dolomite-Matrix Acidizing

Abstract

Summary The two-scale model for simulating carbonate acidizing has gained substantial attention recently. Five studies dealt with matching experimental data studying regular acid. Four studies considered limestone samples, while the fifth examined one dolomite core with face dissolution. The previous work only considered the pore volume (PV) to breakthrough (PVBT) to match experimental results. Researchers assumed linear kinetics for hydrochloric acid (HCl) carbonate reaction and relied on changing Carman-Kozeny exponents to match experimental data. Unlike previous studies, experiments were performed on 6-in.-long and 1.5-in.-diameter vuggy-dolomite cores at two sets of temperatures (150 and 200°F) and acid concentrations (15 and 20 wt% HCl). Computed tomography (CT) was used to scan the cores when dry, wet, and after acidizing. Porosity distribution calculated from the dry and wet scans was used to build a rectangular model with the cylindrical core inscribed inside. Nonlinear reaction kinetics were applied. The acid-reaction rate and diffusion coefficient were modified on the basis of X-ray-fluorescence (XRF) results and effluent chemical analysis. Wormhole 3D shape and experimental PVBT were used to assess the quality of model results. The tuned model was used to simulate a hypothetical 18-in. core as well as large-scale radial experiments to assess its prediction capabilities, and finally the model was used to predict the dolomite-acidizing performance under field conditions. The simulation runs emphasize that the exclusion of the wormhole shape and branching from the matching process results in an unrealistic match. It is important to simulate the cylindrical shape of the core using the actual porosity distribution to capture the wormhole growth, which is increasingly important when the wormhole propagates near the core perimeter. The present study highlights that matching parameters using experimental data yields a trustworthy model that matches both PVBT and wormhole spatial propagation. Accordingly, there is no need for excessively changing the Carman-Kozeny correlation exponents to match the dolomite-acidizing experiments. The current model accurately matches the wormhole propagation inside the core along with the PVBT. This model can be tuned using a few acidizing experiments and then can be used to generate an acid-efficiency curve with a high degree of confidence, thus avoiding the extra experimental cost. The current model was able to match two sets of experiments and follow the experimental trend of longer cores and large-scale radial experiments. It was used to predict acid performance under field conditions. The results show that the optimal PVBT under field conditions is always lower than the one predicted under laboratory conditions; the acid depth of penetration has a significant effect on the acid-efficiency curves; and the vertical flow of acid should be considered in acid-job design.