Molecular Diffusion in Naturally Fractured Reservoirs: A Decisive Recovery Mechanism

Abstract

Abstract The purpose of this paper is to demonstrate the importance of the Fick's molecular diffusion in fractured reservoirs and to present a numerical solution to predict the related mass transfer. As opposed to porous media where molecular diffusion is generally small, in naturally fractured reservoirs, molecular diffusion may be very important, as the dispersive flux through fractures rapidly increases the contact area for diffusion. Fick's molecular diffusion potential may even override viscous forces when hydrocarbon or inert gases are injected and the fracture spacing is small. The importance of Fick's diffusion is largely illustrated in the present paper showing results from analytical models applied to field cases. A method to predict molecular diffusion is presented and has been implemented in a very recent presented and has been implemented in a very recent dual porosity, fully implicit, fully compositional, EOS based reservoir simulator. One advantage of the proposed algorithm, which is an extension of the proposed algorithm, which is an extension of the Sigmund correlation, when compared to others, is that it requires only the component critical properties and other parameters used in any equation of state. Results from the above referred reservoir simulator are also presented to illustrate the effect of molecular diffusion while injecting separator gas or inert gas in a North Sea volatile oil reservoir. Solutions for molecular diffusion of gas in fractured undersaturated oil reservoirs, using single porosity simulators, are also described in the paper. porosity simulators, are also described in the paper Introduction In fractured reservoirs, the dispersive and segregated flux through fractures tends to accentuate compositional differences between matrix and fractures hydrocarbons, generating molecular diffusion potential. When depleting high relief fractured reservoirs, as for instance in the Middle East, important changes on fluid saturation pressures may occur due to the segregation of liberated gas in the fractures and dissolution of free gas in contact with undersaturated matrix oil. Convectional upward flew and diffusion of enriched oil to the secondary gas cap take place in parallel with downward flow of the resultant heavier oil, after release of excess dissolved gas, with diffusion of solution gas into undersaturated oils brought in contact. This type of diffusion-convection with mass transfer between matrix and fractures hydrocarbons was included some 20 years ago in the numerical simulation of large Iranian Asmari Fields, as described by Klaus Schwabe in reference 1. When dry gas is injected in highly fractured/ highly undersaturated oil reservoirs, gas will be dissolved in the matrix oil, resulting in increasing saturation pressure and oil swelling, together with decreasing viscosity and interfacial tensions. In those conditions, molecular diffusion of gas dispersed through the fractures is the main oil recovery mechanism of matrix oil. In both above referred situations, it is a question of molecular diffusion of gas into oil, or transfer of dissolved gas between enriched and heavier oils through a molecular diffusion mechanism, due to differences on compositional gradients between vapour and liquid phases or between two liquid phases. In reality, apart from a thin fringe of phases. In reality, apart from a thin fringe of matrix oil, which is contacted directly by gas existing in the fractures, molecular diffusion within the matrix blocks is processed through successive liquid fringes, which composition tends to be equalized. Molecular diffusion will stop when final chemical equilibrium is attained throughout the matrix liquid. Gas may eventually be injected above miscibility conditions in fractured reservoirs and vapour-liquid miscibility displacement may take place, also by molecular diffusion throughout the matrix blocks. A good example of molecular diffusion-miscibility displacement is the injection of carbon dioxide in P. 429