Dynamic Validation of Phase Behaviour Models for Reservoir Studies of Gas Injection Schemes

Abstract

Compositional models based on cubic equations of state (EOS) are generally employed for simulation of the reservoir performance in gas injection processes. Tuning of the EOS to experimental equilibrium data generated at static conditions, PVT data, prior to its use for compositional studies is the current practice in the oil industry. The displacement of oil in a slim tube by the injection gas simulates the phase behaviour resulting from the continuous contact between the two phases in the reservoir more realistically than any other laboratory tests. It has therefore been suggested that the comparison of the displacement data with simulated results of a compositional model using the tuned EOS can be employed to evaluate and, if necessary, further tune the phase behaviour model. As the flow parameters and numerical methods incorporated in the simulation model can strongly affect the prediction results, the value of such an exercise has been seriously questioned. A systematic study of the applicability of the gas-oil displacement data for tuning of the phase behaviour model to simulate gas injection processes was conducted in this study. The flow characteristics of a number of slim tubes packed with glass beads were evaluated by displacing binary oil mixtures with equilibrated gases to measure the relative permeabilities of the phases at no mass transfer conditions. A large amount of data at different interfacial tension values were used to develop relative permeability-saturation correlations as a function of the interfacial tension. Liquid-Liquid displacement tests were conducted to measure the physical dispersion in the slim tube. The information was used to select the grid block/time step size of the simulator to match the numerical dispersion, due to the truncation error, to that of the physical dispersion. The reliability of the developed parameters was confirmed by comparing predicted and measured displacement data using multicomponent and real reservoir fluids at no mass transfer conditions. Having identified all the parameters of the slim tube and simulation model, a series of gas-oil displacements were conducted using multicomponent and real reservoir fluids at miscible and immiscible conditions. The fluids were initially tested at static conditions and the generated experimental data were used to tune the phase behaviour model prior to simulation of the dynamic gas injection tests. The simulated displacement results, for miscible and immiscible conditions, matched the experimental data favourably where the tuned model was capable of adequately predicting the static data covering the whole range of compositional variations. The study highlighted the value of displacement data for evaluating phase behaviour models particularly at conditions where significant mass exchange occurs between the phases. Introduction An essential tool for predicting the performance of gas injection processes is an equation of state (EOS) compositional simulator. The volumetric behaviour calculation algorithm is the principle part of such a simulator in which the phase behaviour changes are modelled by an EOS. Tuning of EOS is necessary for characterising the reservoir fluids and evaluating their volumetric performance at various pressure levels. Such a tuning exercise is performed against static measured data by adjusting the EOS parameters systematically. Typical laboratory data used in the tuning process include conventional PVT data such as flash and differential vaporisation, single contact, and multiple contact vapour- liquid phase equilibria measurements. P. 201^

Effect of Capillarity and Surface Tension on the performance of VAPEX Process

“…Khazam et al (4) conducted a set of slim tube experiments to investigate the effect of IFT on relative permeability curves. They showed that as IFT approaches zero, that is approaching miscibility, the residual oil saturation, the immobile oil left behind, decreases toward zero and the relative permeabilities increase.…”