Hydrodynamics of solute transport in karst conduits

: Assessment by multi-point dye tracing and numerical modelling

Résumé

The work presented in this thesis manuscript aims at providing a quantitative assessment of the relationship between the hydrodynamics of a karst river and the spatial heterogeneities of breakthrough curves that can be observed when performing a dye tracing. The complex geometries of karst conduits can generate hydrodynamical phenomena that are usually much less present in surface stream, such as sumps, small- or large-scale circular motions, or dead zones. The quantitative assessment of the heterogeneities is done using multi-point tracer tests with a transversal configuration. This consists of placing multiple fluorometers across a river section and quantifying the differences of results. This methodology is applied in various karstic rivers in Wallonia, Belgium, and is combined with the use of both 1-D solute transport model and 3-D Computational Fluid Dynamics (CFD) simulations. The results from the various sites show significant variabilities of tracer tests results across the river section, such as a variable time of first arrival, modal time, modal concentration, and recovery rate. The relationship between these observed variabilities and the hydrodynamics of the river is assessed using the simulated velocity field of the CFD simulations. The results suggest that hydrodynamical phenomena such as a small- or large-scale circular motion, or a dead zone, can have a significant impact on the tracer tests results, even when the tracer cloud is homogeneously distributed among the river section upstream the phenomena. Additionally, the presence of a sump can cause significant spatial variability of tracer tests results. These conclusions are especially significant in an underground lake where the circular motions and dead zones can have large dimensions and generate strong contrast with the main advective flow path. Additionally, the presence of a confluence upstream the section can induce heterogeneities across the river section as the lateral mixing of the tracer clouds coming from each stream can be incomplete. Finally, this study attempts to provide some recommendations of good practice for the placement of field fluorometers when performing a dye tracing in karst rivers. These recommendations are: (i) to ensure that the lateral mixing of the tracer cloud is complete, which is the case when the distance of injection is sufficient, but which can be a challenge when the presence of a confluence cannot be highlighted by speleologists; and (ii) to place the fluorometer in or as close as possible to the main advective flow path. This could be guaranteed by the measurement of the velocity profile. Finally, this study shows that the existence of a diffluence-confluence system can be highlighted by a double peak or a point of inflexion if: (i) the tracer injection is done not too far upstream the diffluence; and (ii) the field fluorometer is placed not too far downstream the confluence. Finally, transversal multi-point tracer tests allow to assess the spatial distribution of the tracer flux by combining results with the velocity profile. This is crucial when dealing with pollutant transport.

la date de réponse	2 juil. 2024
langue originale	Anglais
L'institution diplômante	Universite de Namur
Superviseur	Vincent Hallet (Promoteur), Sabine Henry (Président), Serge Brouyère (Jury), Hervé Jourde (Jury), Jean-Christophe Maréchal (Jury), Amael Poulain (Jury) & Sandra Soares-Frazao (Jury)