Abstract

Molecular diffusion is the process describing the natural mixture of miscible fluids, whose main modeling parameter is the molecular diffusion coefficient. This work aims to evaluate the molecular diffusion coefficient for CO2-light oil systems under different experimental conditions of pressure. Experimental measurements are based on both pressure decay and computed tomography (CT) scan methods. The oil studied is light oil from Brazilian subsalt oil reservoirs. Tests were carried on a specially constructed vertical high pressure cell, from 2.76 MPa up to 28.96 MPa and at 293.15 K. The swelling effect was also evaluated during the diffusion process considering the height variation of the oil column inside the cell.

Molecular diffusion is particularly important for miscible gas flooding processes, as diffusion is a key mechanism controlling the miscibility between oil and gas. The diffusion coefficient determines the rate of mass transfer during the diffusive process that will result in a miscible system. The diffusivity of solvents into light oil in porous media has become of great significance in petroleum engineering, since CO2 injection has been proposed more and more as the enhanced oil recovery method to be applied in the reserves of conventional oils.

Currently, the topic of CO2 diffusion in light oils is scarcely described in the public literature, while diffusion in CO2-heavy oils systems has attracted some attention. Although both diffusion process and swelling effect have a common ground, many differences must be taken into consideration to truly model the mass transfer phenomenon.

Diffusion coefficients were obtained using the CT method and the pressure decay technique throughout Etminan et al. (2013) interface resistance model. Both methods have significative discrepancies in coefficients values. The diffusivities obtained from pressure decay were more consistent with published data. Therefore pressure drop technique seems more robust even for CO2-light oil mixture, while CT technique needs further improvements. Furthermore, as Etminan et al. (2013) model is applicable only for VLE conditions, an improved model will be required for additional LLE situations.

Introduction

Molecular diffusion is the process by which matter is transported from one place in a system to another as a result of random molecular motions (Crank, 1958). It is well established that the diffusion coefficient measured in a two-component system depends on the presence or absence of a chemical concentration gradient. In addition, it also has been explained that there is an asymmetry of molecule movement frequency when a gradient is detected. This asymmetry is due to particle thermodynamics, thus mass flow occurs by diffusion when there is a pressure gradient at that level (Prager, 1953).

In petroleum engineering, molecular diffusion is a key mechanism in solvent-solute process mainly applied for the improvement of petroleum recovery. In the case of gas-based heavy oil recovery, it has been recognized that the molecular diffusion and convective-dispersion are the two important mass transfer processes. The former has a significant role during the soaking time, while the latter is predominant during the time of solvent injection (Tharanivasan, 2006).

The efficiency of the recovery process is highly affected by the rate of dissolution of the injected gas into the target oil, which is controlled by molecular diffusion. So are the resulting changes in the oil properties. Molecular diffusion, whose main characteristic parameter is the molecular diffusion coefficient, is the process describing the natural mixture of miscible fluids. In the recovery processes the diffusion promotes mixing between the injected gas and the oil, prevents viscous fingering, retards gas breakthrough and therefore improves the sweep efficiency for enhancing oil recovery (Ping Guo, 2009).

This content is only available via PDF.
You can access this article if you purchase or spend a download.