A Parametric Simulation Study for Solvent-Coinjection Process in Bitumen Deposits

Abstract

Summary In-situ extraction of ultraviscous deposits from the vast bitumen resources in western Alberta, Canada, requires significant water and energy usage, which consequently leads to greenhouse-gas emissions. Currently proven steam-based recovery schemes include cyclic-steam-stimulation (CSS), steamflooding, and steam-assisted gravity-drainage (SAGD) processes, which are accompanied by many economic and environmental challenges. Coinjection of solvent with steam is a technology that has the potential to improve the efficiency of steam processes as well as reduce energy usage and carbon dioxide emissions. In recent years, researchers and industry professionals have attempted to develop the process further by conducting fundamental research as well as field pilot trials, with varying degrees of success. However, the current level of understanding of the process and the knowledge surrounding the fundamental physics and mechanisms involved are not entirely satisfactory. In this paper, a parametric simulation study was performed to address the key aspects of the solvent-coinjection (SCI) process that contribute to further understanding and development of the process. Simulation observations were verified with experimental evidence where available to support the results and conclusions. Effects of several operational and geological parameters were evaluated on the performance of the SCI process, and the relative performance benefits were assessed over normal SAGD operations. These parameters included solvent type, solvent concentration, initial-solution gas/oil ratio (GOR), relative permeability curves, and pay thickness. The results revealed that the optimal solvent should not be chosen only on the basis of mobility-improvement capability, but also under consideration of other operational, phase- and flow-behavioral and/or geological conditions that are set or present. Higher concentrations of solvents showed more energy-saving upsides than rate-acceleration benefits. It was also observed that the reservoir steam-intake rate is still likely to be the prime performance indicator of the SCI process. In addition, SCI showed that the potential exists for accessing more resources, particularly below the producer level. Furthermore, steam trap control on the producer seems to be problematic when used for SCI simulation. With the current well-control capacity of simulators, a higher degree of subcool is likely to be needed to avoid live vapor-phase production from the producer.