Enhanced oil recovery in fractured vuggy carbonates

Naturally fractured carbonates contribute substantially to global oil reserves. Waterflood and gas-oil gravity drainage (GOGD) recover oil from the fractured oil-wet carbonates, with limited success due to poor sweep and very low recovery factors. Surfactant flooding has shown a great potential to enhance oil recovery in the oil-wet carbonates by reducing interfacial tension and/or altering wettability. Carbonates are characterized by the wide pore-size distributions. Surfactant EOR cannot be successfully implemented in a fractured, oil-wet, carbonate reservoir unless the reservoir is fully characterized and all of the mechanisms involved in oil recovery are fully understood. NMR T₂ measurement, mercury injection capillary pressure test (MICP), thin-section imaging, and computerized tomography (CT) scanning were conducted in the characterization of vuggy dolomite cores from the field. Both thin section and CT images reveal that the touching vugs and separate vugs co-exist in the core samples. Although the vuggy porosity is estimated to be 85%, the matrix controls the permeability of the core because of poor vug connectivity. MICP and NMR T₂ measurements show multimodal pore-throat and pore-body size distributions. Reconstructed 3D CT porosity maps indicate that the vugs in the field dolomite are large and randomly distributed, while the vugs in the Silurian dolomite are small and densely populated. A single-phase tracer test performed under CT scanner reveals a large porosity variation and the preferential flow paths within the field dolomite core. The mercury withdrawal test and NMR T₂ measurement have indicated that snap-off retains oil in the vugs due to the large aspect ratio pores and the large length-scale of the oil blobs. The imbibition oil recovery from the initially oil-wet field dolomite core is 20% lower (in OOIP) than that from the Silurian dolomite core, mainly because of an unfavorable pore structure in the field dolomite core. A few surfactants were selected as promising candidates for wettability alteration because they possess aqueous stability in hard brine at elevated temperatures and reduce contact angles. The divalent cations in the hard brine significantly suppress the anionic surfactant-mediated wettability alteration. The removal of Ca²⁺, and then Mg²⁺ from the hard brine progressively promotes anionic surfactant-assisted wettability alteration, evidenced by decreasing contact angles. The addition of sufficient amount of divalent ion scavengers, including chelating agents (e.g. EDTA.4Na) and scale inhibitors (e.g. Sodium Polyacrylate) in the hard brine, rescues the anionic surfactant-mediated wettability alteration. We propose that the scavenger reduces the concentration of free divalent cations, and promotes the release of the surfactant monomers, which favors wettability alteration through the surfactant adsorption mechanism. The scavenger- triggered mineral dissolution only weakly contributes to the imbibition oil recovery. Experiments and simulation studies consistently showed the synergy between wettability alteration and IFT reduction in a surfactant-assisted gravity-driven process. The residual oil saturation after gravity drainage is approximately 10~20% higher than that by gravity-driven imbibition if the two processes have the same trapping number N[subscript T], which implies that wettability alteration contributes to oil recovery from the oil-wet carbonates. A critical capillary number was found in the capillary desaturation curve plotted for the spontaneous imbibition tests, not for gravity drainage tests. In a UTCHEM model, wettability alteration is represented by the changes in P[subscript c], k[subscript r] and CDC. The simulation successfully history-matched and also predicted the incremental oil recovery by the surfactant formulations. The sensitivity study carried out in UTCHEM simulation shows the strong effects of fluid density, capillary pressure and vuggy pore structures on oil recovery. Three current available oil recovery prediction models (Hagoort, 1980; Aronofsky, 1958; Gupta and Civan, 1994) were tested against imbibition experiments. Two new analytical models were developed in this work, which significantly improved the quality of matching with experimental oil recovery. The matrix-fracture transfer functions, derived from the analytical oil recovery models, can be implemented in a dual-porosity simulator, providing more accurate numerical simulations of oil production in the fractured reservoirs. Lastly, we investigated the feasibility of using single well tracer test (SWTT) in the fractured reservoirs to determine the ROS or connate water saturation. The fractures studied are mainly small-scale fractures. The effects of fracture and its orientation on SWTTs were studied in four Berea cores with a single fracture in each core, orientated as 90°, 60°, 30°, and 0° against dominant flow direction. A simple Cartesian grid without dual porosity in UTCHEM simulator is adequate to interpret the experimental data. A synthetic field-scale SWTT is not sensitive to the presence of moderate degrees of small-scale fractures. The sensitivity study of fluid drift, representing flow irreversibility in a fractured reservoir, reveals the existence of a critical drift velocity, below which the tracer breakthrough curves (BTCs) are interpretable.