[1] Mobile-immobile mass transfer is widely used to model non-Fickian dispersion in porous media. Nevertheless, the memory function, implemented in the sink/source term of the transport equation to characterize diffusion in the matrix (i.e., the immobile domain), is rarely measured directly. Therefore, the question can be posed as to whether the memory function is just a practical way of increasing the degrees of freedom for fitting tracer test breakthrough curves or whether it actually models the physics of tracer transport. In this paper we first present a technique to measure the memory function of aquifer samples and then compare the results with the memory function fitted from a set of field-scale tracer tests performed in the same aquifer. The memory function is computed by solving the matrix diffusion equation using a random walk approach. The properties that control diffusion (i.e., mobile-immobile interface and immobile domain cluster shapes, porosity, and tortuosity) are investigated by X-ray microtomography. Once the geometry of the matrix clusters is measured, the shape of the memory function is controlled by the value of the porosity at the percolation threshold and of the tortuosity of the diffusion path. These parameters can be evaluated from microtomographic images. The computed memory function compares well with the memory function deduced from the field-scale tracer tests. We conclude that for the reservoir rock studied here, the atypical non-Fickian dispersion measured from the tracer test is well explained by microscale diffusion processes in the immobile domain. A diffusion-controlled mobileimmobile mass transfer model therefore appears to be valid for this specific case.

No abstract

No abstract

No abstract

No abstract

No abstract

Summary This paper evaluates the behavior of the dimensionless numbers that gauge the gas/oil interfacial tension (IFT) with respect to the other forces (viscous and buoyancy forces) involved in two-phase flow through porous media. These numbers, referred to respectively as the capillary and Bond numbers, diverge on approach to gas/oil complete miscibility, meaning that viscous and buoyancy forces become dominant over capillary forces. The divergence behaves as a power of the distance to complete miscibility as quantified, for instance, by the difference in densities between the oil and gas phases. The exponents of these power laws are "universal," whereas the prefactors vary between gas/oil systems. This allows the classification of the most common injection gases with respect to their efficiency in reducing the gas/oil IFT. This efficiency increases with the miscibility of the injected gas in the oil: supercritical CO2 is more effective in reducing IFT than off-critical CO2 or CH4, which themselves are more effective than N2. A simple procedure is then introduced to determine the wetting (or spreading) behavior of oil on a porous substrate covered with water, as often occurs in practice. The only inputs required are the composition and densities of the three coexisting phases: water, oil, and gas. When they are not measured, these quantities can be calculated by means of an appropriate equation of state. CO2 turns out to be the most effective for promoting the spreading of the oil on water, followed by CH4 and then by N2.

Gas condensate production and gas injection processes are strongly influenced by the gas/oil (or condensate) interfacial tension and by the wetting behavior of oil on the porous substrate. Oil (condensate) recovery is favored by low gas/oil (condensate) interfacial tensions and by complete wetting of oil (condensate)on the water phase that often covers the porous rock. The relevant parameters are dimensionless numbers that measure the importance of the oil/gas interfacial tension relative to gravity (Bond number) and to viscous (capillary number)forces. Another important parameter is the oil/gas contact angle on the water that often covers the porous substrate: a zero contact angle corresponds to complete wetting of oil, while a finite angle corresponds to partial wetting. These parameters are increasingly used in modern reservoir simulators for estimating oil and gas residual saturations and relative permeabilities. We analyze how these parameters vary with the operating conditions (e.g., pressure) and the nature of the injection gas: CH4, CO2 and N2. On approach to gas/oil complete miscibility, both the capillary and Bond numbers diverge, i.e., viscous and buoyancy forces dominate over capillary forces. For near-miscible conditions, these numbers obey simple scaling laws as a function of the distance to complete miscibility. The prefactors of these scaling laws depend on the particular oil+gas system, which allows to classify the efficiency of the different injection gases. For reducing capillary forces with respect to the viscous or buoyancy forces, supercritical CO2is more advantageous than off-critical CO2 or CH4, which themselves are more advantageous than N2. We also present a simple calculation scheme for predicting the conditions under which the transition to complete wetting of oil on water occurs, i.e., the oil/gas contact angle vanishes. The only ingredients needed are the composition and densities of the three coexisting phases (water, oil and gas), calculated by means of an appropriate equation of state. The essential result is that under typical operating conditions, oils or condensates spread on water, the transition to partial wetting being remote from complete miscibility conditions. CO2 is the most effective in promoting the wetting(or spreading) of the oil phase on water, followed by CH4 and then by N2. Introduction Interfacial forces play an important role in various oil recovery schemes, starting with primary oil recovery. The gas/oil interfacial tension and the wetting behavior of oil in the presence of gas control the distribution of the oil and gas phases within the pore space; therefore these quantities affect the two phase flow parameters: capillary pressure, phase permeabilities and the quantity of oil remaining after drainage with gas. These interfacial properties are strongly dependent on thermodynamic(e.g., pressure or composition) conditions. In the production of near-critical gas condensates or volatile oils and in near-miscible gas injection processes, variations of the gas/oil (or condensate) interfacial tension (IFT) by several orders of magnitude are not uncommon. Upon such variations, the flow regime changes from an emulsion-like flow at very low IFT to a capillary-dominated flow at high IFT. These changes are reflected in the multiphase flow parameters. Phase permeabilities are lower and residual saturations are higher for capillary-dominated flows (high IFT); for very low IFT residual phase saturations tend to zero and relative permeabilities tend to the corresponding phase saturations.1,2 The wetting behavior of oil changes with thermodynamic conditions as well, a transition from partial wetting (i.e., the oil forms lenses on the porous substrate) to complete wetting (i.e., a thick oil layer covers the susbstrate) being expected when the conditions for complete oil and gas miscibility are approached.3 This transition is referred to as the wetting transition. Important changes in the flow processes are expected at the wetting transition.4,5