Demonstration of Reversible Dispersion in a Darcy-Scale Push-Pull Laboratory Experiment

Abstract

A solute transport experiment was conducted in a 50 cm × 50 cm × 4.4 cm apparatus filled with monodisperse glass beads and glycerin, using Rhodamine 6G dye as the solute. The dye movement was tracked through two cycles of a push-pull flow pattern, alternating between approximately radial outward flow and approximately radial inward flow. Depthaveraged concentration was measured during the experiment and integrated over the angular direction during post-processing to obtain a radial distribution of concentration. The experimental results were replicated using particle tracking, simulating advection and dispersion in the radial direction, including reversible dispersion during extraction. Results show that spreading of the measured concentrations follow a Fickian dispersion model during steps with radial outward flow. During steps with radial inward flow, spreading is reversed and can be modeled as reversible dispersion. The observed reversal in spreading is an indication of incomplete mixing within the pore space.

Experiments and Simulations on Plume Spreading by Engineered Injection and Extraction in Refractive Index Matched Porous Media

“…Thus, this is a limitation of the two‐solute approach to mimic reactive transport. The reversal of reaction is not observed in the numerical simulations because spreading cannot be reversed in a traditional random walk method (Neupauer et al., 2021); thus the numerical simulations show more reaction than the experiments. These differences in experimentally observed and numerically simulated total mass do not, however, diminish the two‐dimensional correlation coefficients between experimentally observed and numerically simulated spatial patterns (Table 4, Table 5).…”

“…The greater difference at step 5 comes from an aspect of physics of the experimental system that cannot be simulated numerically. Specifically, because molecular diffusion of Rhodamine dye in glycerin is negligible within the time scale of the experiment (Roth et al., 2021), spreading of dye can be reversed during flow reversal (Neupauer et al., 2021). Since the experimental results infer the concentration of the reaction product from the overlapping of the plumes of non‐reactive solutes A + P and B + P , the experimentally observed reversed spreading of these non‐reactive solutes leads to a spurious numerically modeled reversal of reaction as well, which is not realistic.…”

“…In the special case of quasi‐two dimensional radial flow from a well, it is possible to derive semi‐analytical expressions for the relative contributions of pore‐scale mixing and mechanical dispersion precisely because the plume interface is always perpendicular to the radial flow direction (Neupauer et al., 2020). Indeed, in the case of reversible radial flow, there is evidence that dispersion itself is reversible, reflecting incomplete mixing within pores (Neupauer et al., 2021).…”