Hostname: page-component-7b7567c8c-7pwfn Total loading time: 0 Render date: 2026-11-16T06:29:50.917Z Has data issue: false hasContentIssue false

Evolution of solute blobs in heterogeneous porous media

Published online by Cambridge University Press:  29 August 2018

M. Dentz*
Affiliation:
Spanish National Research Council, IDAEA-CSIC, c/Jordi Girona 18, 08034 Barcelona, Spain
F. P. J. de Barros
Affiliation:
Sonny Astani Department of Civil and Environmental Engineering, University of Southern California, Los Angeles, CA 90089, USA
T. Le Borgne
Affiliation:
Geosciences Rennes, UMR 6118, Université de Rennes 1, CNRS, 35042 Rennes, France
D. R. Lester
Affiliation:
School of Civil, Environmental and Chemical Engineering, RMIT University, Melbourne, Victoria 3001, Australia
*
†Email address for correspondence: marco.dentz@csic.es

Abstract

We study the mixing dynamics of solute blobs in the flow through saturated heterogeneous porous media. As the solute plume is advected through a heterogeneous porous medium it suffers a series of deformations that determine its mixing with the ambient fluid through diffusion. Key questions are the relation between the spatial disorder and the mixing dynamics and the effect of the initial solute distribution. To address these questions, we formulate the advection–diffusion problem in a coordinate system that moves and rotates along streamlines of the steady flow field. The impact of the medium heterogeneity is quantified systematically within a stochastic modelling approach. For a simple shear flow, the maximum concentration of a blob decays asymptotically as $t^{-2}$. For heterogeneous porous media, the mixing of the solute blob is determined by the random sampling of flow and deformation heterogeneity along trajectories, a mechanism different from persistent shear. We derive explicit perturbation theory expressions for stretching-enhanced solute mixing that relate the medium structure and mixing behaviour. The solution is valid for moderate heterogeneity. The random sampling of shear along trajectories leads to a $t^{-3/2}$ decay of the maximum concentration as opposed to an equivalent homogeneous medium, for which it decays as $t^{-1}$.

Information

Type
JFM Papers
Copyright
© 2018 Cambridge University Press 
Figure 0

Figure 1.(a) Multi-log-normal hydraulic conductivity field for $\unicode[STIX]{x1D70E}_{f}^{2}=1/4$. The colour scale is logarithmic from red (highest) to blue (lowest value). (b) Spatial distribution of the absolute value of the corresponding Darcy velocity. The colour scale is linear from red (highest) to blue (lowest value). (c) Spatial distribution of the Okubo–Weiss parameter $\unicode[STIX]{x1D6E9}(\boldsymbol{x})$. The colour scale is linear from blue (lowest negative) to red (highest positive value). (d) Snapshots of the solute concentration $c(\boldsymbol{x},t)$ evolving from a disk-shaped initial distribution of size $\unicode[STIX]{x1D70E}\approx 0.05\ell _{c}$ with $\ell _{c}$ the correlation length of the logarithm of hydraulic conductivity, at different times $t>\unicode[STIX]{x1D70F}_{u}$, where $\unicode[STIX]{x1D70F}_{u}$ is the characteristic advection time scale. The Péclet number is $Pe=5.65\times 10^{2}$. The solid black lines superposed in (b–d) denote streamlines of the flow field.

Figure 1

Figure 2.(a) Evolution of the (solid line) maximum concentration equation (4.6) for $\unicode[STIX]{x1D70E}_{11}^{2}=10^{-3}$, $Pe_{\unicode[STIX]{x1D6FE}}=10^{2}$. The dashed lines show the behaviours in the first intermediate (${\sim}t^{-1/2}$), second intermediate (${\sim}t^{-3/2}$) and late time (${\sim}t^{-2}$) regimes. The dash-dotted line denotes the solution for $\unicode[STIX]{x1D6FE}=0$, this means for uniform flow. (b,c) Concentration $c(\boldsymbol{x},t)$ at times $t=10^{-1}$ and $t=10^{2}$. The solid red lines denote the evolution of the initial contour due to shear only.

Figure 2

Figure 3.(a) Evolution of the (solid line) maximum concentration equation (4.6) for $\unicode[STIX]{x1D70E}_{11}^{2}=1$, $Pe_{\unicode[STIX]{x1D6FE}}=10^{4}$. The dashed lines show the behaviours in the intermediate (${\sim}t^{3/2}$) and late time (${\sim}t^{-2}$) regimes. The dash-dotted line denotes the solution for $\unicode[STIX]{x1D6FE}=0$, this means for uniform flow. (b,c) Concentration $c(\boldsymbol{x},t)$ at times $t=10^{-1}$ and $t=10^{2}$. The solid red lines denote the evolution of the initial contour due to shear only.

Figure 3

Figure 4.(a) Evolution of the (solid line) maximum concentration equation (4.6) for $\unicode[STIX]{x1D70E}_{11}^{2}=1$, $Pe_{\unicode[STIX]{x1D6FE}}=10^{-1}$. The dashed lines show the behaviours in the intermediate (${\sim}t^{-1}$) and late time (${\sim}t^{-2}$) regimes. The dash-dotted line denotes the solution for $\unicode[STIX]{x1D6FE}=0$, this means for uniform flow. (b,c) Concentration $c(\boldsymbol{x},t)$ at times $t=10$ and $t=10^{2}$. The solid red lines denote the evolution of the initial contour due to shear only.

Figure 4

Figure 5.Evolution of the dilution index for $\unicode[STIX]{x1D70E}_{f}^{2}=1/4$ and $\unicode[STIX]{x1D70E}\approx 0.05\ell _{c}$ for (from top to bottom) $Pe=5.65\times 10^{2}$ and $5.65\times 10^{3}$ with (blue circles) numerical data, (red solid line) the analytical expression (4.17). The red dashed lines denote the solution for a corresponding homogeneous medium. The black dash-dotted lines indicate the $t^{3/2}$ scaling.