This paper presents results of a systematic laboratory study to investigate how three-phase immiscible flow can affect gravity-drainage gas injection in naturally fractured reservoirs. In the experiments, quasi two-dimensional glass-bead packs represented matrix and grooved glass strips the fractures. A three-phase hydrocarbon-water-alcohol system mimicked immiscible fluids. Conductive and isolated fractures were examined for two different flow rates. Continuous gas injection, water-alternate-gas and simultaneous water-alternate-gas injection schemes were run. A scaling analysis was conducted to quantify effects of viscous, capillary and gravity forces on the displacement performance for each experiment.
Experimental results show that displacements were controlled mainly by capillary forces. The comparison between injection schemes shows that the continuous gas injection scheme yields highest ultimate recovery. An injection efficiency factor is introduced to assess the feasibility of injection process, which demonstrates that the optimum injection occurs through the water-alternate-gas injection. This is because of the minimum gas consumption, the gas phase which improves the microscopic displacement efficiency in gravity-stabilised floods and the water phase that supports the macroscopic sweep efficiency by preventing the gas phase from fingering. It is found that the oil recovery and injection efficiency in the experiment with the isolated fracture are higher than in the experiment with the conductive fracture model.