Abstract
The mixing and reactive transport of chemical elements and nutrients is a primary controlling process for biogeochemical cycles and contaminant transport in subsurface environments. Reaction kinetics measured in well-mixed laboratory reactors are much larger than those encountered on the field, and Darcy-scale models based on Fickian transport/mixing often cannot explain field measurements. This is attributed to the heterogeneity of the pore scale velocity field in porous media, and to the interplay, at pore scale, between this heterogeneous advection, molecular diffusion, and chemical reactions. We use a quasi 2D millifluidic setup to investigate that interplay. The setup provides pore scale measurement of the flow velocity field by particle tracking, and of the concentration field over 3 orders of magnitude using a fluorescent tracer. Firstly, I’ll show how Lagrangian velocities can be used to upscale the dispersion and mixing of a passive solute, in good agreement with corresponding tracer tests performed in the same geometry. Secondly, I’ll present results from a reactive transport experiment where the (miscible) defending and invading fluids react, leading to the formation of a third solute at their interface. Two successive kinetic regimes, both featuring a non-Fickian scaling of the product mass with time, are observed. A theoretical model explains how the geometry of the mixing interface limits reactant mixing, thus controlling the reaction kinetics. Finally, I shall address mixing under conditions encountered in the vadose zone: the fingers of solute that develop along the average flow direction are prevented from coalescing laterally (as they would under saturated conditions) by the presence of air bubbles. Consequently, an anomalous temporal scaling of the mean scalar dissipation rate is measured: instead of decreasing as 1/t, it only decreases very slowly for the whole duration of the experiment. This finding suggests that the vadose zone potentially represents a hotspot of chemical reactivity. In the geophysical method electrical resistivity tomography (ERT), a map of the effective conductivity of the subsurface is obtained in order to remotely monitor the migration of electrically-conductive tracers and contaminant plumes. The local effective conductivity, measured at the resolution scale of ERT, is controlled by the spatial distribution of the charged tracer. Obtention of the tracer’s concentration field is done from recorded effective conductivity map based on the assumption of a solute concentration that is homogeneous below the resolution scale. The method is known to usually result in a large discrepancy between the measured and injected masses of solute (of up to 80%). Using a millifluidic setup in which we can measure the concentration field in a two-dimensional (2D) porous medium while measuring the effective conductivity of the medium, we show that the solute mass underestimation encountered in field ERT results systematically from the heterogeneity of the solute at scales smaller than the resolution scale of ERT.
ID riunione Microsoft Teams: 337 340 638 861
Passcode: M8u2z2