Development of Gas/Oil Miscibility in Water and Gas Injection

Abstract

Abstract In this paper we use analytical solutions to study the development of multicontact miscibility in simultaneous water and gas (SWAG) injection into a reservoir containing a mixture of hydrocarbons. Application to CO2 storage and enhanced oil recovery is considered. We consider fully compositional one-dimensional (1D), four-component, three-phase flow through a porous medium. Solutions are constructed for a vaporizing gas drive at a pressure below and above the gas/oil minimum miscibility pressure (MMP). We demonstrate that miscibility does not develop when the fraction of water in the injection mixture is sufficiently high and define the minimum gas fraction (MGF) necessary to achieve miscibility. SWAG injection when a mobile water phase is present in the reservoir initially is also studied. We show that the MGF necessary to achieve miscibility depends on the fraction of water present in the reservoir at the start of SWAG injection. In all cases the relative permeability of the supercritical CO2 and water phases determine the MGF. These results indicate that, unlike in miscible gas injection, in SWAG injection the development of miscibility is dependent on the multiphase flow parameters, not just the phase behavior, and highlights the importance of improved relative permeability models. Introduction Oil reservoirs are excellent candidates for geological CO2 storage. They provide a large volume of potential storage space, have excellent geological seals to provide storage security and may have in place some infrastructure that can be used for CO2 injection. High oil prices in recent years make miscible gas injection a potentially cost-effective enhanced oil recovery (EOR) strategy, even without CO2 storage as a project goal. As more power plants are fitted with CO2 capture technologies, streams of compressed, reasonably pure CO2 may be more widely available at costs low enough to allow for use as an injection fluid in EOR. It is also possible that carbon credits may be earned for CO2 that stays in the reservoir. CO2 is often multicontact miscible with light crude oils at reservoir temperature and pressure. Understanding the development of miscibility in the reservoir is critical both to achieve improved oil recovery ensure that injected CO2 remains trapped in the reservoir. Analytical solutions have been the key to understanding the development of miscibility in complex gas/oil displacements (Johns, et al, 1993; Orr, 2007.) They are used in conjunction with slim-tube experiments to tune equations of state so that simulation studies of the field have the correct thermodynmics. Analytical solutions have also been used to predict a priori if a two-phase displacement is going to be sensitive to numerical dispersive effects (Jessen et al, 2004). The drawback in injecting CO2 alone is that the sweep efficiency can be low. Supercritical CO2 has much lower viscosity than oil. As a result, injected CO2 will preferentially invade high permeability flow paths, and additional effects of viscous instability may also reduce sweep efficiency. Furthermore, supercritical CO2 is also less dense than oil, which may lead to gravity override and poor sweep of the bottom of the reservoir. Water is denser and more viscous than the oil in reservoirs targeted for CO2 EOR. By injecting water and gas in alternating slugs or simultaneously, it is possible to use water to limit to some extent the mobility of the injected fluid and improve overall sweep efficiency. LaForce and Orr (2008) have shown that in condensing gas drives injection of water can cause a loss of miscibility when too much water is injected and the CO2 is not able to contact hydrocarbons effectively. It is important that we understand the trade-off between mobility control and the development of miscibility in the 1D flow setting to provide insight and general injection strategies that can be studied in reservoir simulations. In this paper we extend previous work in LaForce and Orr (2008) on miscible condensing gas drives in simultaneous water and gas (SWAG) injection to study the vaporizing gas drive and the effect of mobile water in the reservoir at the start of SWAG injection. First the mathematical model is described. Then solutions are constructed for a vaporizing gas drive at a low pressure and then at the MMP. Finally solutions are constructed for cases when mobile water is present in the reservoir, and the implications, limitations and future directions of this work are discussed.