On the statistical properties of fluid flows with transitional power-law rheology in heterogeneous porous media

TL;DR

This study investigates how heterogeneous porous media influence the flow of non-Newtonian fluids with shear-dependent rheology, revealing distinct flow regimes, heterogeneity effects, and fractal statistical properties of the velocity field.

Contribution

It provides new insights into the interaction between permeability heterogeneity and nonlinear rheology, identifying three flow regimes and analyzing the statistical properties of the velocity field.

Findings

Flow rate varies linearly or as a power law with pressure gradient.

Velocity field heterogeneity is higher for shear-thinning fluids.

Regions above rheology threshold exhibit fractal statistical properties.

Abstract

In this work, we study non-Newtonian fluid flow in heterogeneous porous media. We are interested in fluids presenting a specific change in rheology: Newtonian below a certain shear rate and power law above. Since porous media generally exhibit strong spatial heterogeneity at large geological scales, we study the interaction between such inhomogeneity and the nonlinear rheology of the fluid. The coupling between permeability heterogeneity and nonlinear rheology significantly affects the flow. We are particularly in the statistical properties of the velocity field (mean, variance, correlation, etc). Depending on the imposed mean pressure gradient, three macroscopic flow regimes are identified. For a low or high average pressure gradient, the average flow rate increases linearly or according to a power law, respectively. In the latter regime, we observe that the velocity field is more heterogeneous for shear-thinning fluids than for shear-thickening fluids. This is corresponding to a channeling effect of shear-thinning fluids. The intermediate regime corresponds to a progressive and inhomogeneous change of the local rheology. This transient regime is then characterized in terms of pressure gradient range. The flow field is also analyzed statistically. The spatial distribution of the regions above the rheology threshold shows interesting statistical properties. For instance, they exhibit multiscale characteristics (fractal), similar to other critical systems (percolation, avalanches, etc.). More surprisingly, even though some statistical properties are independent of the parameters, an interesting abrupt rotation of the correlations is found for a particular set of parameters. This is explained by using some symmetries of the problem.