Anomalous dispersion in correlated porous media: A coupled continuous time random walk approach

TL;DR

This paper introduces a coupled continuous time random walk model to explain anomalous dispersion in heterogeneous porous media, linking flow heterogeneity to particle transport behavior and dispersion patterns.

Contribution

It develops a systematic CTRW framework that captures the effects of heterogeneity correlation and distribution on anomalous dispersion in porous media.

Findings

Heterogeneity length scales influence flow and transport velocities.

Broad distributions of heterogeneity points lead to similar dispersion behaviors.

The model predicts particle position and arrival time distributions.

Abstract

We study the causes of anomalous dispersion in Darcy-scale porous media characterized by spatially heterogeneous hydraulic properties. Spatial variability in hydraulic conductivity leads to spatial variability in the flow properties through Darcy's law and thus impacts on solute and particle transport. We consider purely advective transport in heterogeneity scenarios characterized by broad distributions of heterogeneity length scales and point values. Particle transport is characterized in terms of the stochastic properties of equidistantly sampled Lagrangian velocities, which are determined by the flow and conductivity statistics. The persistence length scales of flow and transport velocities are imprinted in the spatial disorder and reflect the distribution of heterogeneity length scales. Particle transitions over the velocity length scales are kinematically coupled with the transition time through velocity. We show that the average particle motion follows a coupled continuous time random walk (CTRW), which is fully parameterized by the distribution of flow velocities and the medium geometry in terms of the heterogeneity length scales. The coupled CTRW provides a systematic framework for the investigation of the origins of anomalous dispersion in terms of heterogeneity correlation and the distribution of heterogeneity point values. Broad distributions of heterogeneity point values and lengths scales may lead to very similar dispersion behaviors in terms of the spatial variance. Their mechanisms, however are very different, which manifests in the distributions of particle positions and arrival times, which plays a central role for the prediction of the fate of dissolved substances in heterogeneous natural and engineered porous materials.