Application and Comparison of Two Models of Foamy Oil Behavior of Long Core Depletion Experiments

Abstract

Abstract The behavior of gas in foamy oil is one of the key factors for the heavy-oilsolution-gas-drive process. Due to the high viscosity of the oil, and theinteraction between capillary pressure and viscous forces, the depressurizationof live heavy oil below the bubble point causes bubbles of gas to nucleate, grow, cluster, and eventually connect into a free phase. In this paper, two models aimed at simulating foamy oil behavior aredescribed and compared. One model partitions the gas into three groups(solution gas, dispersed gas, and free gas), while the other distinguishes fourgroups (including two types of dispersed gas). In the first model, two kineticequations for mass transfer are defined to describe how one group of gas istransformed into another, while in the second model six kinetic equations formass transfer are proposed. Each model defines in an original way the mobilityof the gas groups, which leads to different relative permeability models. Both models were applied to two sets of long core depletion experiments withsignificantly different depletion rates. A comparison of the matched resultsfrom the two models is presented in this paper. Introduction Over the past several years, a number of efforts have been made tounderstand and to model the solution gas drive mechanism in primary heavy oilrecovery. Depressurized live heavy oil is commonly called "foamy"oil1. As the pressure declines below the bubble point, small bubblesof gas appear. In fairly high viscosity heavy oils, the gas bubbles arelong-lived and remain distributed throughout the oil phase. The resultingdispersion evokes the name "foamy" oil. Foamy oil is not at equilibrium.Eventually, the gas bubbles will coalesce to form a continuous gas phase atequilibrium with the oil phase. Many factors have an influence on the evolutionof the bubbles of gas during the stages of nucleation, growth, and coalescence.They include oil viscosity, containment environment (e.g. type of porousmedium), concentration of dissolved gas in the oil, interfacial properties, andshear rate2–4. These factors also affect the mobility of the oil andgas phases. In the first part of this paper a summary is given of the mechanismsinvolved in foamy oil behaviour. Then, there is a brief description of the longcore pressure depletion experiments that provide a test bed for the twodifferent foamy oil models presented here. Finally, the two numerical models, both using CMG STARS® as a platform, are presented and their matches to theexperiments are compared. Foamy Oil Behaviour There are two key elements to foamy oil behaviour: the availability of thegas phase; and, the mobility of the gas phase. Several schools of thoughtregarding the mechanistic understanding and modelling of foamy oil behaviour ata macroscopic scale have emerged in the literature. One school of thoughtrelies on a pseudo single phase description of foamy oil, sometimes includingnon-equilibrium (supersaturation) effects. A second school of thought rejectssupersaturation as a laboratory phenomenon, irrelevant under field conditions.It argues that a conventional two-phase model with parameters appropriate forheavy oils, such as a higher critical gas saturation and a depressed gasmobility, is more suitable for representing foamy oilbehaviour5.