Drainage front width in a three-dimensional random porous medium under gravitational and capillary effects
Paula Reis 1, Knut Jørgen Måløy 1,2 · 2025
Read Drainage front width in a three-dimensional random porous medium under gravitational and capillary effects by Paula Reis 1, Knut Jørgen Måløy 1,2 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-13465.txt.
Introduction
Paula Reis and Knut Jørgen Måløy predict how wide the invasion front is when one fluid drains another from a porous solid. Think of water pushed out of rock by air or oil. The front is not a clean sheet but a ragged band whose width sets how much mixing occurs.
Gravity can steady the front while capillary forces roughen it. Their formula builds on percolation theory, the math of connected clusters in random media. Inputs include pore size, the spread of pore throat thresholds, and the stabilizing pressure gradient. Computer runs of an invasion model matched the formula across many flow settings. Reservoir and groundwater work both need front widths right.
Catalog record
| CREATOR | Paula Reis 1, Knut Jørgen Måløy 1,2 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-13465
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Drainage front width in a three-dimensional random porous medium under gravitational and capillary effects Paula Reis 1, Knut Jørgen Måløy 1,2 Abstract A theoretical approach to estimating stable drainage front widths in three-dimensional random porous media under gravitational and capillary effects is presented. Based on the frontier of the infinite cluster in gradient percolation, we propose an expression for the 3D front width dependent on the pore-network topology, the distribution of capillary pressure thresholds for the pore throats, the stabilizing capillary pressure gradient, the average pore size, and the correlation length critical exponent from percolation in three dimensions. Theoretical predictions are successfully compared to numerical results obtained with a bond invasion-percolation model for a wide range of drainage flow parameters. Preprint: arXiv I Introduction The study of drainage in porous media is relevant to several subsurface transport processes and industrial applications [ 1 , 2 , 3 , 4 , 5 , 6 , 7 ] . It describes the displacement of a wetting fluid from the pore space by an immiscible non-wetting fluid, as their interface moves along a path of least resistance. At the pore scale, this movement is subject to the local capillary pressure value P (defined as P=P_{nw}-P_{w} , where P_{nw} is the pressure in the non-wetting phase, and P_{w} is the pressure in the wetting phase) overcoming the capillary pressure threshold P_{t} of a pore body or pore throat. Depending on the spatial distribution of P_{t} values across the porous medium and variations of P along the invasion front, diverse drainage patterns can arise, ranging from compact to ramified and fractal [ 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 ] . Variations in P along the interface are commonly related to different pressure gradients developed within each phase during drainage, often caused by gravitational and viscous forces. Values of P_{t} are inversely proportional to the sizes of pore bodies and throats, and their spatial distribution in the medium can be homogeneous or heterogeneous [ 16 , 17 , 15 ] . Consider a porous medium where P_{t} values for the pore throats are randomly drawn from an arbitrary distribution N(P_{t}) . In a drainage process with constant capillary pressure along the front, the non-wetting fluid invasion of such porous medium is characterized by capillary fingers [ 11 ] , described well by the Invasion Percolation (IP) method [ 18 ] . Similarities between IP and ordinary percolation allow the theoretical estimation of relevant quantities related to drainage processes, such as fluid saturations and trapped cluster size distributions [ 18 , 9 , 11 ] . By associating each pore throat in a porous medium with a bond in a lattice, the minimum capillary pressure for the percolation of a porous medium, P_{crit} , can also be estimated [ 13 , 19 , 20 ] . For this, each pore throat capillary pressure threshold P_{t} is associated with a bond occupation probability p , and P_{crit} is the capillary pressure value that satisfies: p_{c}=\int_{-\infty}^{P_{crit}}N(P_{t})\,dP_{t} (1) where p_{c} is the bond percolation threshold for the lattice representing the porous medium. In the same isotropic random porous medium, gradients in the capillary pressure along the interface may stabilize or destabilize the invasion front, significantly affecting the efficiency of the drainage process [ 8 , 13 , 21 , 22 , 23 , 24 , 25 , 19 , 26 ] . Positive capillary pressure gradients in the direction of the flow lead to the unbounded growth of thin drainage fingers, which bypass most of the defending phase [ 23 ] . Negative gradients limit the invasion front width to finite values, resulting in compact drainage patterns and a more efficient displacement of the defending phase. In this work, we focus our analysis on the drainage of a random 3D porous medium under gravitational and capillary effects, where a less dense non-wetting fluid displaces a denser wetting fluid from the top. Under these conditions, the gradient of capillary pressures at the interface is G=-\Delta\rho g , where \Delta\rho is the fluid density difference and g is the gravitational acceleration. In particular, we investigate the dependence of the stable drainage front width, \eta_{3D} , on the 3D porous medium structure and G . A common approach to investigate gradient stabilized drainage in porous media is to establish a parallel between the fluids’ interface and the frontier of the infinite cluster in the gradient-percolation problem [ 27 , 28 , 29 ] . For 2D systems, Sapoval et al. [27] demonstrated that this frontier takes place in a critical region, exhibiting a width \eta/l of the order of the correlation length \xi , where l is the length between two lattice sites. As further explained in Section II , the analogy between \eta/l and \xi leads to a drainage interface scaling of the type: \eta/a\propto F^{-\nu/(1+\nu)} (2) where a is the average pore length, \nu=4/3 is the percolation critical exponent for the correlation length in 2D, and F is the dimensionless fluctuation number, F=-N(P_{crit})Ga , defined in [ 19 ] . The theoretical prediction in Eq. 2 has been verified by numerous studies of drainage in 2D porous media, both numerically and experimentally [ 13 , 22 , 19 , 30 , 31 , 26 , 32 ] . In 3D porous media, however, the same interface scaling is not expected to be valid. By extending the gradient-percolation problem to 3D lattices, Gouyet et al. [29] demonstrated that the infinite cluster frontier is no longer confined to a critical region but rather contains a critical region, termed the front tail. Only within the length of the front tail, \eta_{t} , does the front present a fractal structure and obey scaling laws obtained from percolation theory. In the context of drainage, Chaouche et al. [33] performed experiments with 3D porous media consisting of packed glass beads and verified that \eta_{t}\propto Bo^{-\nu/(1+\nu)} , where Bo is the Bond number. Similar scaling laws have been proposed by Wilkinson [34] for the maximum length of trapped wetting-fluid clusters, L_{max} , and for the residual wetting-fluid saturation, S_{res} . While both represent relevant drainage parameters, they are insufficient to disclose the full extent of the region spanned by the front. In the next section, we build on the theoretical approach presented by Gouyet et al. [29] to propose an equation for \eta_{3D} in gradient stabilized drainage in three-dimensional random porous media. The prediction is later verified with a modified IP model incorporating capillary and gravitational effects in Sec. III . II Theory In two dimensions, gradient percolation describes the behavior of N\times N lattices where the site or bond occupation probability p varies monotonically along a direction z . Let us assume a percolation problem in a square lattice, where p varies from p(z=0)=1 to p(z=N)=0 . In this lattice, an infinite cluster exists, containing the occupied sites or bonds connected to the boundary where p=1 . The infinite cluster presents a fractal frontier of width \eta , centered at z_{c} , where p(z=z_{c})=p_{c} . Considering that clusters formed in the frontier region should be of the order of the correlation length \xi\propto|p-p_{c}|^{-\nu} (where \nu is the percolation