Elsevier

Chemical Engineering Science

Volume 196, 16 March 2019, Pages 225-246
Chemical Engineering Science

Microscopic theory of capillary pressure hysteresis based on pore-space accessivity and radius-resolved saturation

https://doi.org/10.1016/j.ces.2018.10.054Get rights and content

Highlights

  • Two concepts naturally capture capillary pressure hysteresis with arbitrary cycling.

  • Accessivity describes connectivity between different sized pores.

  • Radius-resolved saturation, unlike saturation, gives pore-scale fluid distribution.

  • Conceptual framework has broader utility in continuum modeling of porous media.

Abstract

Continuum models of porous media use macroscopic parameters and state variables to capture essential features of pore-scale physics. We propose a macroscopic property “accessivity” (α) to characterize the network connectivity of different sized pores in a porous medium, and macroscopic state descriptors “radius-resolved saturations” (ψw(F),ψn(F)) to characterize the distribution of fluid phases within. Small accessivity (α0) implies serial connections between different sized pores, while large accessivity (α1) corresponds to more parallel arrangements, as the classical capillary bundle model implicitly assumes. Based on these concepts, we develop a statistical theory for quasistatic immiscible drainage-imbibition in arbitrary cycles, and arrive at simple algebraic formulae for updating ψnF that naturally capture capillary pressure hysteresis, with α controlling the amount of hysteresis. These concepts may be used to interpret hysteretic data, upscale pore-scale observations, and formulate new constitutive laws by providing a simple conceptual framework for quantifying connectivity effects, and may have broader utility in continuum modeling of transport, reactions, and phase transformations in porous media.

Introduction

From rocks and wood to concrete and catalysts, porous materials vary widely in origin, properties, and applications. Despite their macroscopic appearance as solid objects, porous media are distinguished by their ability to contain fluids internally, owing to their heterogeneous microstructure – the solid matrix occupies only a portion of the macroscopic domain, while the complementary pore space is able to accommodate one or more fluid phases (Bear, 1972).

In the typical case of well-connected pores, diverse physical phenomena in porous media, such as fluid flow, heat and mass transfer, gas adsorption, and phase transformations, may be amenable to homogenized macroscopic descriptions (Torquato, 2013), although the exact connection with microscopic details of the porous medium is not always clear. Remarkably, any continuum model implicitly assumes that the overall effects of the often nontrivial pore-space morphologies (Sahimi, 2011) can be encapsulated in a small number of parameters, e.g., porosity, ϕ, tortuosity, τ, intrinsic permeability, ks, effective thermal conductivity, ke, etc., and the state of any fluid phase can be described by a small number of distributed state variables, e.g., pressure, p, saturation, s, etc. For certain simple physical processes, especially those involving a single fluid phase, simple continuum formulations generally work well, and a rigorous connection between the pore-scale and continuum-scale governing equations can be sought – examples include single-phase flow (Allaire, 1989, Hornung, 2012) and heat transfer (Loeb, 1954, Carson et al., 2005) – although estimating the transport coefficients from microscopic features of the porous medium is still an open research problem (Ranut, 2016, Pabst and Gregorova, 2017, Neithalath et al., 2010, Chareyre et al., 2012, Mostaghimi et al., 2013).

In comparison, it is considerably more challenging to develop continuum models for immiscible multiphase flow in porous media (including unsaturated flow and condensate transport) (Bear, 1972, Scheidegger, 1974, Richards, 1931, Buckley and Leverett, 1942, Childs and Collis-George, 1950, Klute, 1952, Tamon et al., 1981, Lee and Hwang, 1986, Jaguste and Bhatia, 1995, Do and Do, 2001) based on microscopic physics. While it is possible to upscale pore-scale equations by careful averaging (Durlofsky, 1998, Arbogast, 2000, Cushman et al., 2002, Wood, 2009, Li et al., 2006), the resulting model varies depending on the macroscopic state variables selected and the scaling laws assumed for the application considered, not to mention that the mechanisms for pore-scale fluid motions are highly complex and are still actively researched (Lenormand et al., 1983, Pak et al., 2015, Holtzman and Segre, 2015, Zhao et al., 2016). Thus, in formulating continuum models of multiphase flow in porous media, there exists a trade-off between mathematical simplicity and consideration of pore-scale physics. In conventional models that are widely accepted in practice, saturation is the primary state variable; it is used to compute capillary pressure (Brooks and Corey, 1964, van Genuchten, 1980) and relative permeabilities (Corey, 1954, Irmay, 1954, Brooks and Corey, 1964) via empirical constitutive relationships, which are able to fit typical experimental measurements by virtue of having several adjustable parameters (Pinder and Gray, 2008). By recognizing that pore-scale phenomena like viscous flow and capillary equilibrium must first and foremost depend on the microscopic dimensions of pores, some of the most popular continuum constitutive relationships in the literature have also incorporated the concept of a pore-size distribution (Thomson, 1872, Washburn, 1921, Burdine, 1953, Mualem, 1976), conceptualizing the pore space as interconnected pores of various sizes, either implicitly or explicitly (van Brakel, 1975, Quiblier, 1984), such as in the well-known “capillary bundle” model.

However, these conventional models suffer from hysteresis, meaning that the relationships between state variables are non-unique and history-dependent. This suggests that saturation alone cannot fully describe the microscopic state of fluid phases in real porous media, and additional macroscopic state variables are required to capture hysteresis. Many authors have considered extending the conventional constitutive relationship between capillary pressure and saturation, pc(sw), by assuming that pc may also depends on the “rate of saturation”, sw/t (or at least its sign) (Barenblatt, 1971, Luckner et al., 1989, Hassanizadeh and Gray, 1990, Barenblatt et al., 2003, Juanes, 2008), thereby attributing hysteresis to nonequilibrium effects. Some authors have identified “specific interfacial area”, awn, or the interfacial area between fluid phases w and n per unit volume of the porous medium, as a physically relevant state variable on thermodynamic grounds, and have advocated for its inclusion in continuum models to reduce capillary pressure hysteresis (Kalaydjian, 1987, Hassanizadeh and Gray, 1990, Hassanizadeh and Gray, 1993, Beliaev and Hassanizadeh, 2001, Hassanizadeh et al., 2002). This hypothesis seems to hold in many but not all cases, as revealed by micromodel experiments (Cheng et al., 2004, Chen et al., 2007, Pyrak-Nolte et al., 2008, Karadimitriou et al., 2014), lattice-Boltzmann simulations (Porter et al., 2009), pore-network simulations (Reeves and Celia, 1996, Held and Celia, 2001, Joekar-Niasar et al., 2008, Joekar-Niasar and Hassanizadeh, 2012), and analysis (Helland et al., 2007), while the exact form of any new constitutive relationships required may not be completely clear. More recently, Hilfer put forth a new class of continuum models for two-phase flow in porous media (Hilfer, 1998, Hilfer, 2006, Doster et al., 2010) involving four fluid saturation variables, {s1,s2,s3,s4}, as opposed the conventional two, {sw,sn}, for the two fluid phases (in either case, all saturation variables must sum up to unity); we have sw=s1+s2 and sn=s3+s4, where s1 and s3 correspond to “percolating regions” of the respective fluid phases, and s2 and s4 correspond to “non-percolating regions”. By differentiating between the contributions of percolating and non-percolating fluid “subphases”, Hilfer’s model naturally predicts hysteresis as a result of the dynamics of the newly introduced state variables. Because the model is not derived from the principles of microscopic physics, phenomenological assumptions are still required to, say, model the “mass transfer rates” between s1 and s2, or s3 and s4 (namely, percolating fluid regions becoming non-percolating and vice versa), which result in model parameters that may be difficult to physically interpret, though potentially determinable from experiments.

It appears that the intrinsic complexity of multiphase flow in porous media would ensure that any continuum model to come in the foreseeable future will not match the performance of pore-scale methods (pore-network, Lattice-Boltzmann, phase-field, etc.; see this review (Meakin et al., 2009) for instance) in terms of either predictability or connection to first principles. On the one hand, conventional models are preferred for macroscopic simulations and used routinely in practical applications, where hysteresis is either described empirically or neglected altogether, although it is considered crucial for making certain types of predictions (e.g., Essaid et al., 1993). On the other hand, new continuum models, despite having rightly introduced new and physically meaningful state variables (awn, Hilfer’s s1,,s4) so as to naturally predict hysteresis, deviate significantly from conventional models, and involve somewhat unintuitive constitutive laws with phenomenological constants that lack a clear connection to the pore-scale descriptions, possibly due to the emphasis placed on reproducing certain macroscopic observations.

In this work, we take the view that there is great value in identifying new physically meaningful concepts that are relevant to continuum modeling of multiphase flow in porous media. These concepts should be connected to essential aspects of pore-scale physics, yet intuitive enough for a wide range of continuum-scale applications – the pore-size distribution would fall within this category. In the short term, these concepts may be incorporated into conventional continuum models for incremental improvements, while in the long term, they may be subject to pore-scale investigations, and ultimately play a role in future continuum models of multiphase flow.

The first concept we shall propose is the “pore-space accessivity”, denoted by α. Accessivity is a continuum-scale property of porous media that, in the simplest possible fashion, contrasts serial and parallel arrangements of different sized pores. This extends the capillary bundle model, which is known to be insufficient, but still routinely invoked because of its simplicity (Diamond, 2000, Hunt et al., 2013). The capillary bundle picture coincides with the α1 limit in our framework.

The second concept we shall propose is the “radius-resolved saturation”, ψ(F), where 0F(r)1 is the cumulative distribution function (CDF) of the pore-size distribution, where r denotes the pore radius. Radius-resolved saturation would replace saturation as a better continuum-scale descriptor for the distribution of fluid phases in the pore space, where “saturation” ψ is now defined for pores of each particular size given by F, hence “radius-resolved”.

The paper is organized as follows. In Sections 2 Characterization of pore-space morphology, 3 Characterization of fluid distribution, we introduce the concepts of accessivity and radius-resolved saturation, and relate them to existing ideas in the literature. In Section 4, we present a simple statistical theory based on pore branching that leads to simple governing equations involving the proposed concepts. In Section 5, we present simple illustrative examples to highlight the usefulness and limitations of our theory. Finally, we discuss the broader utility of accessivity and radius-resolved saturation and identify outstanding questions and future research directions in Section 6, before concluding in Section 7.

Section snippets

Characterization of pore-space morphology

In this section, we consider macroscopic descriptors for the pore-space morphology of a porous medium. For that purpose, the pore space need not be filled with any particular fluid and can be left “empty” (or, alternatively and conceptually equivalently, filled uniformly with an inert fluid). We will briefly review some existing descriptors and introduce pore-space accessivity at the end of the section.

Characterization of fluid distribution

In this section, we will consider a porous medium whose pore space now hosts multiple fluids. The fluid phases may distribute in any arbitrary fashion that is consistent with pore-scale physics. Our goal is to characterize the distribution of fluid phases at the macroscopic scale. The concepts we propose here are general and can be readily extended to systems involving any number of immiscible fluids, though for simplicity, we shall confine our discussion to the case of two fluid phases, which

Statistical theory

So far, we have proposed accessivity, α, as a continuum-scale descriptor for the connectivity of different sized pores, and radius-resolved saturation, ψw(F), for the distribution of immiscible fluid phases in the pore space. We have qualitatively demonstrated that, in a porous medium with α<1, serial connections between different sized pores may contribute to the accumulation of “metastable” fluids (i.e., wetting phase in “larger” pores, or nonwetting phase in “smaller” pores, relative to a

Illustrative examples

In this section, we present several exploratory case studies to illustrate the implications of our theory. We will examine the behaviors of the key formulae derived, including: rules for updating radius-resolved saturations during arbitrary drainage-imbibition cycles, Eqs. (54), (55); their simplified forms for primary drainage and imbibition, Eqs. (50), (51); and the resulting formulae for conventional saturations, (43) and (45), respectively. Of particular interest is the role of accessivity,

Further discussions

In this section, we will further expound the conceptual usefulness as well as limitations of our theory. We will discuss the broader utility of accessivity and radius-resolved saturation, including how they may be generalized.

Conclusions

We have proposed the pore-space accessivity, α0,1, as a continuum-scale parameter for describing the arrangement of different sized pores in porous media. Defined quantitatively based on the amount of hysteresis observed in macroscopic invasion percolation processes, α can also be interpreted geometrically as the ratio of the length (or volume) scale for pore radius variation to the typical size of a pore instance (the pore space explored by an average meniscus in a control volume) observed at

Acknowledgement

The authors wish to acknowledge funding from Saudi Aramco, a Founding Member of the MIT Energy Initiative. We are also grateful to Samuel J. Cooper for useful discussions.

References (103)

  • R.J. Held et al.

    Modeling support of functional relationships between capillary pressure, saturation, interfacial area and common lines

    Adv. Water Resour.

    (2001)
  • D.N. Jaguste et al.

    Combined surface and viscous flow of condensable vapor in porous media

    Chem. Eng. Sci.

    (1995)
  • R. Juanes

    Nonequilibrium effects in models of three-phase flow in porous media

    Adv. Water Resour.

    (2008)
  • R.G. Larson et al.

    Effects of sample size on capillary pressures in porous media

    Powder Technol.

    (1981)
  • K.-H. Lee et al.

    The transport of condensible vapors through a microporous Vycor glass membrane

    J. Colloid Interface Sci.

    (1986)
  • L. Li et al.

    Upscaling geochemical reaction rates using pore-scale network modeling

    Adv. Water Resour.

    (2006)
  • G. Mason

    The effect of pore space connectivity on the hysteresis of capillary condensation in adsorption-desorption isotherms

    J. Colloid Interface Sci.

    (1982)
  • G. Mason

    The effect of pore lattice structure on the pore size distributions calculated from sorption isotherms using percolation theory

    J. Colloid Interface Sci.

    (1983)
  • N. Neithalath et al.

    Characterizing pore volume, sizes, and connectivity in pervious concretes for permeability prediction

    Mater. Character.

    (2010)
  • M. Parlar et al.

    Percolation theory of vapor adsorption-desorption processes in porous materials

    J. Colloid Interface Sci.

    (1988)
  • M.L. Porter et al.

    Lattice-Boltzmann simulations of the capillary pressure-saturation-interfacial area relationship for porous media

    Adv. Water Resour.

    (2009)
  • J.A. Quiblier

    A new three-dimensional modeling technique for studying porous media

    J. Colloid Interface Sci.

    (1984)
  • P. Ranut

    On the effective thermal conductivity of aluminum metal foams: review and improvement of the available empirical and analytical models

    Appl. Therm. Eng.

    (2016)
  • C. Salmas et al.

    Mercury porosimetry: contact angle hysteresis of materials with controlled pore structure

    J. Colloid Interface Sci.

    (2001)
  • N.A. Seaton

    Determination of the connectivity of porous solids from nitrogen sorption measurements

    Chem. Eng. Sci.

    (1991)
  • J. van Brakel

    Pore space models for transport phenomena in porous media review and evaluation with special emphasis on capillary liquid transport

    Powder Technol.

    (1975)
  • G. Allaire

    Homogenization of the Stokes flow in a connected porous medium

    Asymptot. Anal.

    (1989)
  • T. Arbogast

    Numerical subgrid upscaling of two-phase flow in porous media

  • S. Bakke et al.

    3-D pore-scale modelling of sandstones and flow simulations in the pore networks

    SPE J.

    (1997)
  • G.I. Barenblatt

    Filtration of two nonmixing fluids in a homogeneous porous medium

    Fluid Dyn.

    (1971)
  • G.I. Barenblatt et al.

    The mathematical model of nonequilibrium effects in water-oil displacement

    SPE J.

    (2003)
  • J. Bear

    Dynamics of Fluids in Porous Media

    (1972)
  • A.Y. Beliaev et al.

    A theoretical model of hysteresis and dynamic effects in the capillary relation for two-phase flow in porous media

    Transp. Porous Media

    (2001)
  • R.H. Brooks et al.

    Hydraulic properties of porous media and their relation to drainage design

    Trans. ASAE

    (1964)
  • S.E. Buckley et al.

    Mechanism of fluid displacement in sands

    Trans. AIME

    (1942)
  • N.T. Burdine

    Relative permeability calculations from pore size distribution data

    J. Petrol. Technol.

    (1953)
  • P.C. Carman

    Fluid flow through granular beds

    Trans. Inst. Chem. Eng.

    (1937)
  • M.A. Celia et al.

    Recent advances in pore scale models for multiphase flow in porous media

    Rev. Geophys.

    (1995)
  • B. Chareyre et al.

    Pore-scale modeling of viscous flow and induced forces in dense sphere packings

    Transp. Porous Media

    (2012)
  • D. Chen et al.

    Measurement of interfacial area per volume for drainage and imbibition

    Water Resour. Res.

    (2007)
  • J.-T. Cheng et al.

    Linking pressure and saturation through interfacial areas in porous media

    Geophys. Res. Lett.

    (2004)
  • E.C. Childs et al.

    The permeability of porous materials

  • A.T. Corey

    The interrelation between gas and oil relative permeabilities

    Produc. Mon.

    (1954)
  • W.M. Deen

    Analysis of Transport Phenomena

    (2011)
  • F. Doster et al.

    Numerical solutions of a generalized theory for macroscopic capillarity

    Phys. Rev. E

    (2010)
  • F.A.L. Dullien

    Porous Media: Fluid Transport and Pore Structure

    (1991)
  • L.J. Durlofsky

    Coarse scale models of two phase flow in heterogeneous reservoirs: volume averaged equations and their relationship to existing upscaling techniques

    Comput. Geosci.

    (1998)
  • H.I. Essaid et al.

    Simulation of fluid distributions observed at a crude oil spill site incorporating hysteresis, oil entrapment, and spatial variability of hydraulic properties

    Water Resour. Res.

    (1993)
  • I. Fatt

    Network model of porous media

    J. Petrol. Technol.

    (1956)
  • A. Forner-Cuenca et al.

    Advanced water management in PEFCs: diffusion layers with patterned wettability II. Measurement of capillary pressure characteristic with neutron and synchrotron imaging

    J. Electrochem. Soc.

    (2016)
  • Cited by (9)

    • Effect of particle size and adsorption equilibrium time on pore structure characterization in low pressure N<inf>2</inf> adsorption of coal: An experimental study

      2020, Advanced Powder Technology
      Citation Excerpt :

      These time-related changes may be related to the gas diffusion rates in the pores. According to the experimental results, pores are continuously distributed ranging from several nanometers to several microns, the pore network in the coal matrix is simplified into a system of a series of many cylindrical pores with different pore diameters where the pore size decreases from the surface of the matrix to the internal areas [38,39]. Methane desorption experiments have shown that the diffusion coefficient decreases with the time, indicating that the path of methane diffusion from the matrix to the fractures is varied during the methane desorption [40,41].

    View all citing articles on Scopus
    View full text