Multiscale characterization of transport in heterogeneous porous and fractured aquifer media using innovative heat and solute tracer tests
Nous avons le plaisir de vous convier à la défense de thèse de Monsieur Richard HOFFMANN. Celle-ci se déroulera le jeudi 21 janvier à 14h, via LifeSize (https://call.lifesizecloud.com/3896304).
Cette thèse, réalisée sous la direction du Prof. Alain Dassargues, s'intitule "Multiscale characterization of transport in heterogeneous porous and fractured aquifer media using innovative heat and solute tracer tests".
Résumé de la recherche doctorale
The realistic description of subsurface mass and energy transport in heterogeneous porous and fractured aquifer media requires the accurate assessment of preferential pathways and dynamic matrix processes at multiple scales. Solute and highly diffusive tracer information were combined with the goal of improving the understanding of actual transport processes (e.g., channeling and matrix diffusion) and the predictive capability of groundwater flow and transport models.
The evolution of concentration and temperature as a function of time (breakthrough curves) was investigated after injections of: (1) heat and a solute in alluvial sediments (previously performed); (2) dissolved gases (helium, argon, and xenon), heat, and uranine in a porous/fractured chalk system; (3) salt and water at higher and lower temperatures than the natural groundwater background temperature in a weathered and fractured granite system in India. Observations were interpreted in terms of peak time, peak value, and slope of the tailing (i.e., quantifying the slow decline of the induced anomaly), considering the variation in the diffusion coefficient value of the different tracers used.
Moreover, analytical solutions and numerical models including stochastic simulations were developed to simulate the field observations as close as possible to reality. Spatial heterogeneity could be better constrained when stochastic solutions were used and those solutions allowed to test a Euclidean distance-based sensitivity analyses. The latter relates model input and output uncertainty to reveal key information about the parameters most affecting the model used. Compared to classical sensitivity analyses, e.g., methods in which one input parameter is ultimately changed while the others are held constant, multiple interactions among model input parameters can be considered, making its use very promising for future hydrogeological purposes. The PhD project clearly highlights the importance and usefulness of the new tracer developments as an innovative method (process-based imaging) to realistically evaluate the inherent heterogeneity of the different groundwater systems studied.
