Mass Diffusion into Bitumen: A Sub-pore Scale Modeling Approach
Taheri, S., Kantzas, A. and Abedi, J.
CSUG/SPE 138129 presented at the Canadian Unconventional Resources & International Petroleum Conference held in Calgary, Alberta, Canada, October 19-21, 2010.
The increased interest in the production of heavy oil and bitumen has amplified attention paid to solvent-based methods for heavy oil recovery, such as vapor extraction (VAPEX) or miscible flooding.
The diffusion of solvent into oil plays a major role for all the solvent-based recovery methods. Since the diffusion process is governed by the diffusion coefficient, the accurate prediction of mass transfer of the solvent in heavy oil and bitumen is extremely important. The concentration dependency of the diffusion coefficient differs from sample to sample and is determined experimentally in the laboratory. What is measured in the lab as the diffusion coefficient is influenced by different porous medium properties during the solvent injection. Therefore, an effective diffusion coefficient is defined for porous media, which has a dependency on the medium properties.
The main goal of this paper is the prediction of an accurate value for the effective diffusion coefficient from experimentally measured values of diffusion considering the properties of the porous medium.
The medium in this paper can be a micro model pattern, thin section, tomographic image or microscopic picture. The picture is analyzed by an image processing program to distinguish the pore and grain sections. After gridding the pore regions of the picture, virtual porous medium properties are extracted by applying the Navier-Stoke and continuity equations as the governing equations. The diffusion equation is applied to the medium to find the concentration profile of the solvent in the porous medium, and the effective diffusion coefficient of the system is computed from the concentration profile. An extensive investigation of the effects of medium properties on the diffusion coefficient will lead to the capability of predicting the effective diffusion coefficient for other media with different patterns and properties.