Gas injection schemes (including WAG and SWAG injections) are among the most widely used Enhanced Oil Recovery (EOR) processes for carbonate reservoirs due to their usually low rock permeability. Because of the complex nature of carbonate reservoirs (structural heterogeneities and mixed- to oil-wet conditions), most of the existing reservoir simulators are not able to adequately model EOR processes in these reservoirs. Performing reliable laboratory experiments in reservoir cores under reservoir conditions is necessary in order to evaluate the performance of these oil recovery techniques and obtain the required data to assess the prediction of the currently available models in commercial simulators.
In this paper, we present the results of a series of coreflood experiments performed in carbonate rocks. Both reservoir core plugs as well as a dolomite outcrop core were used in the experiments. The cores were aged in reservoir crude oil to restore wettability. Gas and water injection experiments were performed under reservoir conditions and with different injection strategies. Separate multi cycle WAG injection experiments were performed starting with gas (WAG-DI) or water injection (WAG-ID) in order to compare the impact of the order of fluid injection. SWAG injection experiment was also performed with an injection gas to water ratio of unity. The injectivity and production data of different injection strategies were compared. Generally speaking, dolomite cores showed less oil-wet characteristics compared to reservoir carbonate cores which showed strongly oil-wet behaviours. WAG injection starting with waterflood and SWAG injection performed better than waterflood or gas injection alone. Ultimate oil recovery by gas injection was considerably less than that obtained by waterflood. As a result of poor efficiency of the gas injection, the WAG injection starting with gas injection (WAG-DI) showed lower recovery performance compared to the waterflood up until the end of 3rd WAG injection cycle. Nevertheless, WAG-DI ultimately outperformed waterflood. Although SWAG injection outperformed the WAG-ID for the first two cycles of injection but the ultimate residual oil saturation was lower for the WAG-ID injection. In addition to lower ultimate oil recovery, SWAG injection test showed more injectivity issues compared to the WAG-ID injection sequence. The observed trends of residual gas and oil saturations in carbonate reservoir cores were different compared with those of water-wet and mixed-wet sandstone rocks reported in our previous studies in which the residual oil to water was reducing linearly with an increase in trapped gas. The results provide useful insights into mechanisms of oil recovery by gas and water injection in carbonate reservoirs and highlight some important differences with sandstone reservoirs.