Effect of capillary pressure on the approach to residual saturation

The approach to residual oil saturation during the immiscible displacement of oil as predicted by the multiphase Darcy equations is studied. It is well known that when the capillary pressure term is neglected, one arrives at the Buckley-Leverett formulation according to which the inlet face attains residual oil saturation instantaneously. This result may, however, be strongly influenced by the inclusion of the capillary pressure term. In this paper it is shown that when the relative permeability and capillary pressure functions have power law dependencies on the saturation deviation from residual oil condition, the long time solution exhibits a power law decay toward residual saturation. Moreover, the power law decay solution is found to be unique and independent of the initial condition. The relationship of this solution to the classical Buckley-Leverett result is shown. Finally, generalization to the time varying flow rate case is addressed. As a verification of the theoretical conjectures, the power law solution is compared with direct numerical simulation of the two phase flow equations.     

Enhancing Immiscible Fluid Displacement in Porous Media by Capillary Pressure Discontinuities

Fluid displacement in porous media plays an important role in many industrial applications, including biological filtration, carbon capture and storage, enhanced oil recovery, and fluid transport in fuel cells. The displacement front is unstable, which evolves from smooth into ramified patterns, when the mobility (ratio of permeability to viscosity) of the displacing fluid is larger than that of the displaced one; this phenomenon is called viscous fingering. Viscous fingering increases the residual saturation of the displaced fluid, considerably impairing the efficacy of fluid displacement. It is of practical importance to develop suitable methods to improve fluid displacement. This paper presents an experimental study on applying the discontinuity of capillary pressure to improve immiscible fluid displacement in drainage for which the displacing fluid (air) wets the porous media less preferentially than does the displaced fluid (silicone oil). The concept involves using a heterogeneous packing system, where the upstream region features large pores and small capillary pressure, and the downstream region features small pores and large capillary pressure. The increase in capillary pressure prevents fingering from directly crossing the media interface, thus enhancing the displacement. The experimental apparatus was a linear cell comprising porous media between two parallel plates, and glass beads of 0.6 and 0.125 mm diameter were packed to compose the heterogeneous porous media. The time history of the finger flow was recorded using a video camera. Pressure drops over the model from the inlet to the outlet were measured to compare viscous pressure drops with capillary pressures. The results show that the fluid displacement was increased by the capillary discontinuities. The optimal displacement was determined through linear regression by adjusting the relative length of the large- and small-pore region. The results may assist in the understanding of fingering flow across the boundaries of different grain-sized bands for the gas and oil reservoir management, such as setting the relative location of the injection and production wells. The findings may also serve as a reference for industrial applications such as placing the grain bands in an adequate series to improve the displacement efficacy in biological filtration.     

Relative permeabilities from two- and three-dimensional pore-scale network modelling

We present a computer study of two-phase flow in a porous medium. The porous medium is represented by an isotropic network of up to 80 000 randomly placed nodes connected by thin tubes. We then simulate two-fluid displacements in this network and are able to demonstrate the effects of viscous and capillary forces. We use the local average flow rates and pressures to calculate effective saturation dependent relative pemeabilities, fractional flows and capillary pressures. Using a radial Buckley-Leverett theory, the mean saturation profile can be inferred from the solution of the fractional flow equation, which is consistent with the computed saturation. We show that the relative permeability may be a function of both viscosity ratio and capillary number.     

Steady-state solutions to Bentsen's equation

Based on experimentally observed phenomena and the physical requirement of a unique value of saturation at any location within a porous medium, a restrictive condition for a valid solution to Bentsen's equation is derived: ?² f/?S ²≤0. The steady-state solution to Bentsen's equation is shown to be identical to the Buckley-Leverett solution to the displacement equation, and the steady-state solution for the fractional flow is shown to be independent of the capillary number. It is proved that under steady-state conditions, the capillary term of the fractional flow equation in the frontal region does not depend on the capillary number. Therefore, the unrealistic triple-valued saturation profile of the original Buckley-Leverett solution resulted because the capillary term was in-appropriately neglected. The break-through recovery efficiency,Τ _bt , is shown to be a function of the capillary number. As the capillary number decreases, the break-through recovery efficiency increases and the maximum value ofΤ _bt can be obtained asN _c → 0. The Buckley-Leverett solution is the limiting solution asN _c → 0.     

Efficiency of Capillary Imbibition Dominated Displacement of Nonwetting Phase by Wetting Phase in Fractured Porous Media

The critical and optimum injection rates as well as the critical fracture capillary number for an efficient displacement process are determined based on the experimental and numerical modeling of the displacement of nonwetting phase (oil) by wetting phase (water) in fractured porous media. The efficiency of the process is defined in terms of the nonwetting phase displaced from the system per amount of wetting phase injected and per time. Also, the effects of injection rate on capillary imbibition transfer dominated two-phase flow in fractured porous media are clarified by visualizing the experiments. The results reveal that as the injection rate is increased, fracture pattern begins to become an effective parameter on the matrix saturation distribution. As the rate is lowered, however, the system begins to behave like a homogeneous system showing a frontal displacement regardless the fracture configuration.     

Multiphase flow with a simplified model for oil entrapment

A computationally simple procedure is described to model effects of oil entrapment on three-phase permeability-saturation-capillary pressure relations. The model requires knowledge of airwater saturation-capillary pressure relations, which are assumed to be nonhysteretic and are characterized by Van Genuchten's parametric model; scaling factors equal to the ratio of water surface tension to oil surface tension and to oil-water interfacial tension; and the maximum oil (also referred to as nonwetting liquid in a three-phase medium) saturation which would occur following water flooding of oil saturated soil. Trapped nonwetting liquid saturation is predicted as a function of present oil-water and air-oil capillary pressures and minimum historical water saturation since the occurrence of oil at a given location using an empirically-based algorithm. Oil relative permeability is predicted as a simple function of apparent water saturation (sum of actual water saturation and trapped oil saturation) and free oil saturation (difference between total oil and trapped oil saturation), and water relative permeability is treated as a unique function of actual water saturation. The proposed method was implemented in a two-dimensional finite-element simulator for three-phase flow and component transport, MOFAT. The fluid entrapment model requires minimal additional computational effort and computer storage and is numerically robust. The applicability of the model is illustrated by a number of hypothetical one- and two-dimensional simulations involving infiltration and redistribution with changes in water-table elevations. Results of the simulations indicate that the fraction of a hydrocarbon spill that becomes trapped under given boundary conditions increases as a nonlinear function of the maximum trapped nonwetting liquid saturation. Dense organic liquid plumes may exhibit more pronounced effects of entrapment due to the more dynamic nature of flow, even under static water table conditions. Disregarding nonwetting fluid entrapment may lead to significant errors in predictions of immiscible plume migration.     

Water removal from porous media by gas injection: experiments and simulation

The flow of a saturated gas through a porous medium, partially occupied by a liquid phase, causes evaporation due to gas expansion. This process, referred to as flow-through drying, is important in a wide variety of natural and industrial applications, such as natural gas production, convective drying of paper, catalysts, fuel cells and membranes. X-ray imaging experiments were performed to study the flow-through drying of water-saturated porous media during gas injection. The results show that the liquid saturation profile and the rate of drying are dependent on the viscous pressure drop, the state of saturation of the gas and the capillary characteristics of the porous medium. During the injection of a completely saturated gas, drying occurs only due to gas expansion. Capillary-driven flow from regions of high saturation to regions of low saturation lead to more uniform saturation profiles. During the injection of a dry gas, a drying front develops at the inlet and propagates through the porous medium. The experimental results are compared with numerical results from a continuum model. A good agreement is found for the case of sandstone. The comparison is less satisfactory for the experiments with limestone.     

Viscous coupling in two-phase flow in porous media and its effect on relative permeabilities

An idealized model of a porous rock consisting of a bundle of capillary tubes whose cross-sections are regular polygons is used to assess the importance of viscous coupling or lubrication during simultaneous oil-water flow. Fluids are nonuniformly distributed over tubes of different characteristic dimension because of the requirements of capillary equilibrium and the effect of interfacial viscosity at oil-water interfaces is considered. With these assumptions, we find that the importance of viscous coupling depends on the rheology of the oil-water interface. Where the interfacial shear viscosity is zero, viscous coupling leading to a dependence of oil relative permeability on oil-water viscosity ratio for viscosity ratios greater than one is important for a range of pore cross-section shapes and pore size distributions. For nonzero interfacial shear viscosity, viscous coupling is reduced. Using values reported in the literature for crude oil-brine systems, we find no viscous coupling.     

Filtration of a three-phase liquid (Buckley-Leverett model)

Buckley and Leverett [1] formulated the problem of the displacement of immiscible liquids in a porous medium and obtained a very simple one-dimensional solution for a two-phase flow. Different generalizations of it are known [2]. In [3, 4], a method of characteristics is proposed for numerical solution of the problem of three-phase flow. Articles [5, 6] consider the problem of the injection (at a given pressure) of two incompressible liquids into a porous stratum previously saturated with a third, elastic liquid. The authors started from the assumption of the existence, for this problem, of zones of three-, two-, and single-phase flow, separated by unknown mobility gradients. The present work investigates the solution for a three-phase flow, analogous to the Buckley-Leverett solution for two phases. It is shown that the character of the degrees of saturation depends essentially on the initial saturation of the porous stratum and on the phase composition of the mixture being injected.Moscow. Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 1, pp. 39–44, January–February, 1972.     

Simulation of capillary pressure curves using bond correlated site percolation on a simple cubic network

A stochastic approach to network modelling has been used to simulate quasi-static immiscible displacement in porous media. Both number-based and volume-based network saturation results were obtained. Number-based results include: number-based saturation curves for primary drainage, secondary imbibition and secondary drainage, fluid distribution data, and cluster trapping history. Using pore structure data of porous media, it is possible to convert the number-based curves to capillary pressure — saturation relationships. Pore size distribution functions and pore shapes which are thought to closely represent Berea sandstone samples were used to predict the capillary curves. The physical basis of these calculations is a one-to-one correspondence between the cumulative node and bond index fractions in the network analysis, and the cumulative number-based distributions of pore body and pore throat diameters, respectively. The oil-water capillary pressure curve simulated for primary drainage closely resembles those measured experimentally. The agreement between the simulated and the measured secondary imbition and secondary drainage curves is less satisfactory.     

In-Situ Capillary Trapping of CO<Subscript>2</Subscript> by Co-Injection

Co-injection of water with CO₂ is an effective scheme to control initial gas saturation in porous media. A fractional flow rate of water of approximately 5–10% is sufficient to reduce initial gas saturations. After water injection following the co-injection, most of the gas injected in the porous media is trapped by capillarity with a low fractional volume of migrating gas. In this study, we first derive an analytical model to predict the gas saturation levels for co-injection with water. The initial gas saturation is controlled by the fractional flow ratio in the co-injection process. Next, we experimentally investigate the effect of initial gas saturation on residual gas saturation at capillary trapping by co-injecting gas and water followed by pure water injection, using a water and nitrogen system at room temperature. Depending on relative permeability, initial gas saturation is reduced by co-injection of water. If the initial saturation in the Berea sandstone core is controlled at 20–40%, most of the gas is trapped by capillarity, and less than 20% of the gas with respect to the injected gas volume is migrated by water injection. In the packed bed of Toyoura standard sand, the initial gas saturation is approximately 20% for a wide range of gas with a fractional flow rate from 0.50 to 0.95. The residual gas saturation for these conditions is approximately 15%. Less than approximately 25% of the gas migrates by water injection. The amount of water required for co-injection systems is estimated on the basis of the analytical model and experimental results.     

A Parametric Model for Predicting Relative Permeability-Saturation-Capillary Pressure Relationships of Oil–Water Systems in Porous Media with Mixed Wettability

A parametric two-phase, oil–water relative permeability/capillary pressure model for petroleum engineering and environmental applications is developed for porous media in which the smaller pores are strongly water-wet and the larger pores tend to be intermediate- or oil-wet. A saturation index, which can vary from 0 to 1, is used to distinguish those pores that are strongly water-wet from those that have intermediate- or oil-wet characteristics. The capillary pressure submodel is capable of describing main-drainage and hysteretic saturation-path saturations for positive and negative oil–water capillary pressures. At high oil–water capillary pressures, an asymptote is approached as the water saturation approaches the residual water saturation. At low oil–water capillary pressures (i.e. negative), another asymptote is approached as the oil saturation approaches the residual oil saturation. Hysteresis in capillary pressure relations, including water entrapment, is modeled. Relative permeabilities are predicted using parameters that describe main-drainage capillary pressure relations and accounting for how water and oil are distributed throughout the pore spaces of a porous medium with mixed wettability. The capillary pressure submodel is tested against published experimental data, and an example of how to use the relative permeability/capillary pressure model for a hypothetical saturation-path scenario involving several imbibition and drainage paths is given. Features of the model are also explained. Results suggest that the proposed model is capable of predicting relative permeability/capillary pressure characteristics of porous media mixed wettability.     

Effect of Initial Wettability on Rock Mechanics and Oil Recovery: Comparative Study on Outcrop Chalks

Surfactant-alternating-gas (SAG) is a favored method of foam injection, in part because of excellent gas injectivity. However, liquid injectivity is usually very poor in SAG. We report a core-flood study of liquid injectivity under conditions like those near an injection well in SAG application in the field, i.e., after a prolonged period of gas injection following foam. We inject foam [gas (nitrogen) and surfactant solution] into a 17-cm-long Berea core at temperature of 90 °C with 40 bar back pressure. Pressure differences are measured and supplemented with CT scans to relate water saturation to mobilities. Liquid injectivity directly following foam is very poor. During prolonged gas injection following foam, a collapsed-foam region forms near the inlet and slowly propagates downstream, in which water saturation is reduced. This decline in liquid saturation reflects in part liquid evaporation, also pressure-driven flow and capillary effects on the core scale. In the collapsed-foam region, liquid mobility during subsequent liquid injection is much greater than downstream, and liquid sweeps the entire core cross section rather than a single finger. Mobility in the region of liquid fingering is insensitive to the quality of foam injected before gas and the duration of the period of gas injection. This implies that at the start of liquid injection in a SAG process in the field, there is a small region very near the well, crucial to injectivity, substantially different from that further out, and not described by current foam models. The results can guide the development of a model for liquid injectivity based on radial propagation of the various banks seen in the experiments.

     

Permanent Fronts in Two-Phase Flows in a Porous Medium

A family of exact solutions for a model of a one-dimensional horizontal flow of two immiscible, incompressible fluids in a porous medium, including the effects of capillary pressure, is obtained analytically by solving the governing singular parabolic nonlinear diffusion equation. Each solution has the form of a permanent front propagating with a constant velocity. It is shown that, for every propagation velocity, there exists a set of permanent fronts all of which are moving with this velocity in an inflowing wetting–outflowing non-wetting flow configuration. Global bifurcations of this set, with the front velocity as a bifurcation parameter, are investigated analytically and numerically in detail in the case when the permeabilities and the capillary pressure are linear functions of the wetting phase saturation. Main results for the nonlinear Brooks–Corey model are also presented. In both models three global bifurcations occur. By using a geometric dynamical system approach, the nonlinear stability of the permanent fronts is established analytically. Based on the permanent front solutions, an interpretation of the dynamics of an arbitrary front of finite extent in the model is given as follows. The instantaneous upstream (downstream) velocity of an arbitrary non-quasistationary front is equal to the velocity of a permanent front whose shape coincides up to two leading orders with the instantaneous shape of the non-quasistationary front at the upstream (respectively, downstream) location. The upstream and downstream locations of the front undergo instantaneous translations governed by modified nonsingular hyperbolic equations. The portion of the front in between these locations undergoes a diffusive redistribution governed by a nonsingular nonlinear parabolic diffusion equation. We have proposed a numerical approach based on a parabolic–hyperbolic domain decomposition for computing non-quasistationary fronts.     

Microtomography of Imbibition Phenomena and Trapping Mechanism

Water imbibition during the waterflooding process of oil production only sweeps part of the oil present. After water disrupts the oil continuity, most oil blobs are trapped in porous rock by capillary forces. Developing an efficient waterflooding scheme is a difficult task; therefore, an understanding of the oil trapping mechanism in porous rock is necessary from a microscopic viewpoint. The development of microfocused X-ray CT scanner technology enables the three-dimensional visualization of multiphase phenomena in a pore-scale. We scanned packed glass beads filled with a nonwetting phase (NWP) and injected wetting phase (WP) in upward and downward injections to determine the microscopic mechanism of immiscible displacement in porous media and the effects of buoyancy forces. We observed the imbibition phenomena for small capillary numbers to understand the spontaneous imbibition mechanism in oil recovery. This study is one of the first attempts to use a microfocused X-ray CT scanner for observing the imbibition and trapping mechanisms. The trapping mechanism in spontaneous imbibition is determined by the pore configuration causing imbibition speed differences in each channel; these differences can disrupt the oil continuity. Gravity plays an important role in spontaneous imbibition. In upward injection, the WP flows evenly and oil is trapped in single or small clusters of pores. In downward injection, the fingering phenomena determine the amount of trapped oil, which is usually in a network scale. Water breakthrough causes dramatic decrease in the oil extraction rate, resulting in lower oil production efficiency.     

Displacement of a Newtonian fluid by a non-Newtonian fluid in a porous medium

This paper presents an analytical Buckley-Leverett-type solution for one-dimensibnal immiscible displacement of a Newtonian fluid by a non-Newtonian fluid in porous media. The non-Newtonian fluid viscosity is assumed to be a function of the flow potential gradient and the non-Newtonian phase saturation. To apply this method to field problems a practical procedure has been developed which is based on the analytical solution and is similar to the graphic technique of Welge. Our solution can be regarded as an extension of the Buckley-Leverett method to Non-Newtonian fluids. The analytical result reveals how the saturation profile and the displacement efficiency are controlled not only by the relative permeabilities, as in the Buckley-Leverett solution, but also by the inherent complexities of the non-Newtonian fluid. Two examples of the application of the solution are given. One application is the verification of a numerical model, which has been developed for simulation of flow of immiscible non-Newtonian and Newtonian fluids in porous media. Excellent agreement between the numerical and analytical results has been obtained using a power-law non-Newtonian fluid. Another application is to examine the effects of non-Newtonian behavior on immiscible displacement of a Newtonian fluid by a power-law non-Newtonian fluid.     

Diffusive Effects on Recovery of Light Oil by Medium Temperature Oxidation

Volatile oil recovery by means of air injection is studied as a method to improve recovery from low permeable reservoirs. We consider the case in which the oil is directly combusted into small products, for which we use the term medium temperature oil combustion. The two-phase model considers evaporation, condensation and reaction with oxygen. In the absence of thermal, molecular and capillary diffusion, the relevant transport equations can be solved analytically. The solution consists of three waves, i.e., a thermal wave, a medium temperature oxidation (MTO) wave and a saturation wave separated by constant state regions. A striking feature is that evaporation occurs upstream of the combustion reaction in the MTO wave. The purpose of this paper is to show the effect of diffusion mechanisms on the MTO process. We used a finite element package (COMSOL) to obtain a numerical solution; the package uses fifth-order Lagrangian base functions, combined with a central difference scheme. This makes it possible to model situations at realistic diffusion coefficients. The qualitative behavior of the numerical solution is similar to the analytical solution. Molecular diffusion lowers the temperature of the MTO wave, but creates a small peak near the vaporization region. The effect of thermal diffusion smoothes the thermal wave and widens the MTO region. Capillary diffusion increases the temperature in the upstream part of the MTO region and decreases the efficiency of oil recovery. At increasing capillary diffusion the recovery by gas displacement gradually becomes higher, leaving less oil to be recovered by combustion. Consequently, the analytical solution with no diffusion and numerical solutions at a high capillary diffusion coefficient become different. Therefore high numerical diffusion, significant in numerical simulations especially in coarse gridded simulations, may conceal the importance of combustion in recovering oil.     

An investigation of hydraulic fracturing. Part II: two-phase flow model

In the previous work presented in Part I (Theoret. Appl. Fracture Mech. 18, 89–102 (1993)), hydraulic fracture in an infinitely large saturated porous medium is analyzed under an assumption of one-phase flow in the medium. The investigation is extended in this paper to the case of a two phase saturated immiscible flow of oil and water in the porous medium. The medium is initially saturated with oil. Flow in the medium is induced by diffusion of water injected into the fracture. The quasi-static growth of the fracture for a prescribed injection rate is analyzed based on the assumptions that the pressure in the fracture is uniform and that the permeating flow in the medium is unidirectional. The constant fracture toughness criterion for plane strain deformation is employed and the effect of capillary pressure is neglected. Empirical formulas are used for the permeabilities of the oil and water phases. It is seen that the distributions of water saturation and pore pressure in the medium are governed by two nonlinear partial differential equations. Numerical solutions are obtained by a finite difference scheme with iterations. It is found that the injected water is restricted within a layer near the surface of the fracture whose thickness is small compared with the length of the fracture. Thus the flow in the medium is governed essentially by the oil phase. To compare our problem with the corresponding problem of one-phase flow, we find that the difference in crack growth in these two problems is small for the ration of kinematic viscosities of the oil and water phases within the practical range. Hence our study confirms the validity of the one phase flow assumption used in the previous work for prediction of hydraulic fracture growth.     

The Role of Capillarity in Two-Phase Flow through Porous Media

When determining experimentally relative permeability and capillary pressure as a function of saturation, a self-consistent system of macroscopic equations, that includes Leverett's equation for capillary pressure, is required. In this technical note, such a system of equations, together with the conditions under which the equations apply, is formulated. With the aid of this system of equations, it is shown that, at the inlet boundary of a vertically oriented porous medium, static conditions pertain, and that potentials, because of the definition of potential, are equal in magnitude to pressures. Consequently, Leverett's equation is valid at the inlet boundary of the porous medium, provided cocurrent flow, or gravity-driven, countercurrent flow is taking place, and provided the porous medium is homogeneous. Moreover, it is demonstrated that Leverett's equation is valid for flow along the length of a vertically oriented porous medium, provided cocurrent flow, or gravity-driven, countercurrent flow is taking place, and provided the porous medium is homogeneous and there are no hydrodynamic effects. However, Leverett's equation is invalid for horizontal, steady-state, forced, countercurrent flow. When such flow is taking place, it is the sum of the pressures, and not the difference in pressures, which is related to capillary pressure.     

Leak-off During Placement of Cr(III)-Partially Hydrolyzed Polyacrylamide Gelling Solution in Fractured Porous Media

During waterflooding of a fractured formation, water may channel through the fracture or interconnected network of fractures, leaving a large portion of oil bearing rock unswept. One remedial practice is injection of a gelling solution into the fracture. Such placement of a gelling mixture (referred as gelant) is associated with leak-off from the face of the fracture into the adjoining matrix. As the gelant gets more crosslinked, the gelant encounters more resistance in flowing into the porous matrix. This article addresses the build-up of flow resistance as the Cr(III)-partially hydrolyzed polyacrylamide gelant, at various stages of crosslinking flows into the matrix. Flow experiments were conducted at constant injection pressure in unfractured Berea rocks that represent a matrix adjoining a fracture. Before entering the core, gelants underwent post-mixing delays, shorter than their gel time. On continued displacement, flow resistance developed that reduced the flow rate further. More delay, after mixing of gelant hastened, the build-up of resistance to flow and the resistance was contained nearer to the inlet face. Effect of flow over fracture face on the build-up of flow resistance in the matrix was also evaluated by conducting displacement of gelant in two fractured slabs. In one case, a part of the injected fluid came out of the fracture outlet with the rest leaking off into matrix. In the other case, all the fluid that entered into the fracture leaked off into the matrix. Build-up of flow resistances in the matrix for the two cases was compared. A simple conceptual model is presented that could explain the flow of gelant and build-up of resistance in porous rock at constant injection pressure.