As heavy oil resources are making an ever-increasing contribution to the energy supply of the world, maximizing recovery from these reservoirs has attracted significant attention in recent years. Leveraging low salinity waterflooding and polymer flooding to improve heavy oil recovery will be the focus of this paper. While low salinity waterflooding performance in light (conventional) oil reservoirs has been investigated extensively, it has not been applied in heavy oil systems.Published literature recorded higher oil recovery by injection of low salinity water. One major component among several mechanisms of this increase in oil recovery is due to the lower residual oil saturation (S orw ) by wettability alteration. Therefore, published data was used to produce a correlation between S orw (high salinity water) and S orw (low salinity water). A sector model with 36,000 grid cells was used to evaluate the performance of several injection scenarios involving low salinity water and polymer under the conditions of a heavy oil reservoir at the depth of 3,000 ft with oil viscosity of 80 centipoise (20° API). The thickness of the reservoir is less than 100 ft.The simulated results showed that low salinity waterflooding could increase the estimated ultimate recovery by approximately 5% of original oil in place (OOIP) while the polymer flood provided a similar increase with better efficiency. Subsequent simulations that combine both low salinity waterflooding and polymer resulted in an additional oil recovery of 7.5% to 10% of OOIP.The results of this study show significant potential for enhancing heavy oil recovery by low salinity waterflooding augmented by polymer injection and multilateral horizontal wells. The commercial viability of such methods to improve heavy oil recovery would make a significant step towards increasing the recovery factor from heavy oil and viscous oil reservoirs. Subsequently, the results would need to be validated by laboratory experiments to establish clear correlation between relative permeability end points with both viscosity and water salinity on heavy oil systems.

Water-Alternating-Gas injection is one of the most applicable EOR techniques in oil reservoirs worldwide. To optimize the WAG injection projects, numerical simulation is the best available tool to aid with such objective. In this paper, we are looking into the simulation process (from core-scale to reservoir-scale) in more details to understand the challenges and try to suggest best practices and guidelines. The first part of this paper will discuss the role of hysteresis in WAG injection based on the results of several three-phase Water-Alternating-Gas injection experiments. Then the significance of three-phase hysteresis in simulating WAG injection will be highlighted. After establishing a good understanding of the role of three-phase hysteresis in WAG injection simulation, the second part will present sensitivity study at reservoirs-scale to identify the main differences in the simulation results. Finally, towards the end of the paper, WAG injection's best practices guideline will be suggested. Currently, the common practice is to estimate three-phase hysteresis parameters (Land trapping parameter [C], secondary drainage gas relative permeability reduction exponent [alpha], residual oil modification factor [a], and three-phase water relative permeability [krw3ph]) from coreflood experiments. Based on several three-phase WAG injection experiments conducted at Heriot-Watt University, we came up with several conclusions. These conclusions are: If the pressure drop (DP) during the three-phase water injection experiment is significantly higher than the DP during two-phase water injection, then a lower three-phase water relative permeability is essential to match the three-phase WAG injection experimental data.If the pressure drop (DP) at the end of two-phase gas injection is lower than the DP at the end of three-phase gas injection, then alpha [α] is critical to match the three-phase WAG injection experimental data.If the experimental results showed significant gas trapping [Sgt] during water injection after gas injection, then Land's trapping parameter is essential to model such trapping behaviour at any scale.These parameters should be obtained from experiments starting with gas injection (GAW) to fit with the definitions accepted into the numerical simulations. Many studies in the literature either ignores the effect of hysteresis in WAG injection or use invalid WAG hysteresis parameters. In this paper, we highlighted the role of hysteresis in WAG injection and clarified the confusion on the procedures to obtain the correct hysteresis parameters. Also, we provided an easy guideline to help the industry in making accurate predictions for WAG injection projects.

To ensure the physics of multiphase flow in porous media is rightly modelled, the balance between capillary, viscous, and gravity forces need to be understood. Capillary pressure (Pc) and relative permeability (kr) are critical parameters representing capillary and viscous forces respectively. Both are typically determined by special core analysis in the laboratory. The importance of application of proper capillary pressure curves for different processes and the consistency between the kr and Pc input are investigated. In this study, we used the idea of Numerical Coreflood Experiment (NCFE) where detailed geology and known oil-water relative permeability and capillary pressure curves are used. Various NCFEs' production and pressure data can be generated for different conditions and used to estimate the kr and Pc. Here, we used a commercial application to back calculate a set of relative permeability and capillary pressure that fits the given production and pressure data for the examined cases. Afterward, we compared the resulting kr and Pc curves with those that were used to generate the NCFEs data to begin with. In this study, we focused on the two-phase oil-water system to assess the process of obtaining a history- matched relative permeability by fitting the core-flood experiment data while setting capillary pressure to zero (ignoring Pc) in the simulation model. We compared different cases that we examined to identify the role and importance of Pc measurements in core-scale and reservoir-scale numerical simulations. The conclusion is that using appropriate input capillary pressure measurement is an essential step to ensure proper representation of the multi-phase flow physics. Ignoring Pc or using inaccurate Pc measurements could lead to inaccurate relative permeability curves and as a results unrealistic production and pressure output. The results of this study can be used by researchers and practicing reservoir engineers in oil and gas industry. NCFE is an inexpensive and easy-to-use technique to evaluate the current experimental procedures and suggest improvements. NCFE can be extended to cover a wider-range of evaluations including the effect of gravity, injection rate and heterogeneities.

The oil recovery factor from oil reservoirs must be increased significantly to meet the ever-increasing demand for energy. Majority of oil reserves worldwide (estimated as 60%) are held in carbonate reservoirs. It is believed that most carbonate reservoirs are mixed-wet to oil wet with high variation in wettability within the reservoir rock. Water-Alternating-Gas (WAG) injection is one of the most applicable EOR techniques known to increase oil recovery factor by 8-20% on top of waterflood alone. To maximize recovery from carbonate reservoirs, numerical simulation is the best available tool to aid with such objective. In this paper, we are presenting a new approach to simulate near-miscible water- alternating-gas injection for mixed-wet reservoirs. The coreflood experiment used in this study was performed on a 65mD mixed-wet sandstone core sample at 38°C (100°F) and 12.69 MPa (1840 psia) where calculated Interfacial tension (IFT) between gas and oil at these conditions is 0.04 mN/m. As recorded in the literature, the current simulation capabilities to model WAG injection behaviour is questionable. Some authors suggested that Land’s gas trapping parameter [C] should be variable during the WAG injection process. In this study, we investigated the best way to match the results of WAG injection experiments performed on a mixed-wet core. Matching the full experimental results were not possible without varying gas trapping parameter [C]. However, by updating the gas trapping parameter, based on our suggested procedure, the match between the simulation results and experimental data improved significantly. This paper highlights the shortcomings of the commercial simulators in modelling WAG injection process. The current WAG hysteresis model can be improved based on the information published in this paper for better WAG injection simulation.

No abstract