Fatehgarh reservoirs in Aishwariya field, located in Barmer Basin of Rajasthan India, have very high CO2 content; ~80-90 mole% CO2 in the gas phase and ~10-40 mole% CO2 in the oil phase. Formation water sample has shown CO2 concentration of ~7500 ppm with a pH of ~5.6. The oil is moderately viscous and aqueous based chemical EOR techniques like polymer/ASP flooding have been identified as the most suitable for Aishwariya's field conditions. The presence of CO2 decreases the pH of a system. The viscosity of polyacrylamide polymers is known to decrease with reduction in pH; however, to the best of author's knowledge, commercial simulators do not have any method to directly model this viscosity dependence on pH. Therefore, a procedure was developed to capture the impact of CO2 on polymer in-situ viscosity in a reservoir simulation model.
An EOS based PVT model was first developed using PVT data from the field. This EOS was then used in a commercial compositional simulator to calculate the phase distribution of CO2 during an injection process. The liquid-liquid partitioning coefficient for CO2 was calculated as the ratio of CO2 mole fraction in oil phase to that in water phase. A similar model was then created in an "advanced process" reservoir simulator which modelled component distributions based on partitioning coefficients. For simplicity, the solution gas was assumed to be entirely composed of CO2. The CO2 component was defined as part of the aqueous phase instead of the default oleic phase. Data which related polymer viscosity to pH were used to create a "Viscosity-pH" table; other data permitted the construction of a "pH-CO2 concentration" relationship. Linking these two tables together allowed the development of a "Viscosity-CO2" function. Defining the CO2as an "aqueous phase salt component" enabled us to model the polymer viscosity variation with CO2 concentration. Simulation runs were then carried out both with and without the effect of CO2 on polymer viscosity and the results were compared.
The developed approach captured the impact of CO2 on the simulated polymer flood performance. It permitted us to quantify the amount of additional polymer that would be required to compensate for polymer viscosity loss in the presence of CO2. It helped to improve our confidence in polymer flood performance predictions for this field despite the presence of a significant amount of CO2.
The paper presents a step by step approach of utilizing commercially available simulation softwares to model the impact of in-situ CO2 on polymer flood performance. The proposed method can be applied in fields with high concentrations of CO2. The method can be extended for fields with lesser concentrations of CO2.