Abstract
Dye tracing experiments are generally set up to identify karst springs catchments or transport properties of heterogeneous or homogeneous aquifers. They are also relevant methods for characterizing flow and transport processes in fractured and karst aquifers, ang get insights about their complex structure and heterogeneity. A precise analysis of breakthrough curves (BTCs) obtained from dye tracing experiments in synthetic pipe networks and fracture networks of know geometry and topology should thus allow a better assessment of the relationship between drainage network properties and transport processes.In this work, solute transport experiments consisting of step tracing are thus conducted, with a focus on dual-conduit structure. Five groups of experiments have been carried out to investigate how the following factors related to the dual-conduit structure may influence the transport processes: length ratio, total length (fixing the length ratio), connection angle, aperture contrast, and flow rate. Numerical models are then applied to fit the experimental BTCs and quantitatively evaluate the solute transport processes. The single-peaked BTCs are simulated considering i) the Advection Dispersion Equation (ADE) model, ii) the Dual Region Mobile Immobile Model (DRMIM), and iii) the Transfer Function Approach (TFA). The dual-peaked BTCs are simulated considering i) the Dual Region Advection Dispersion (DRAD) model, ii) the Weighted Sum Advection–Dispersion Equation (WSADE) model and iii) the DRMIM. Based on these results, a method has been proposed for estimating underground karstic conduit lengths from experimental dual peaked BTCs. For some experiments, it is shown that the TFA allows a better fitting than the ADE model and DRMIM. In order to assess what physical information contain the TFA parameters, we analyse the equivalence between TFA and Advection Dispersion Equation (ADE). We then obtain two equations that, under certain conditions, make TFA and ADE produce identical BTCs and allow expressing the TFA parameters as a function of the variable and parameters considered in the ADE.Finally, a methodology for characterizing a karst conduit network at the scale of an experimental field site where both pumping test and dye tracing experiment have been performed is proposed. This characterization is based on a hydraulic tomography method which considers a discrete conduit network model to explicitly represent the underground conduit network, integrating hydraulic data and dye tracing data to constrain the diameter distribution of the karst conduit network.The results of these lab scale and field scale experiments, together with the proposed numerical tools, should help hydrogeologists to better interpret the results of the tracing tests realized in real karst aquifers and improve their understanding of transport processes. These results may also guide hydrogeologists to select the most suitable model for interpreting their dye tracing experiment.