Experimental and theoretical analysis of closed-flow column experiments : an alternative approach for the investigation of solute transport on the continuum scale

Research of solute transport fundamentally contributes to our understanding of soil functions as most processes in soils are dynamically driven and related to the transient conditions produced by the transport of solutes. For this reason, especially the transport of contaminants is frequently studied with laboratory scale column outflow experiments in an open-flow mode. This thesis presents a complementary approach of conducting saturated column experiments that is characterized by the recirculation of the effluent into the inflow via a mixing vessel and is therefore referred to as closed-flow mode column experiment. Depending on the ratio of the volume of the mixing vessel and the water-filled pore space, a damped oscillating concentration emerges in the effluent and in the mixing vessel. Oscillation frequency, damping and amplitude are thereby governed by the properties of the porous medium and the target substance. It was shown by column experiments with quartz sand that the appearance of oscillations can be controlled by using different mixing vessel solute volumes. The breakthrough data obtained within these experiments was then used to validate a numerical model that was derived by coupling the numerical solution of a transport equation with the model describing the mixing vessel in a loop. This model was used for a comprehensive sensitivity analysis to illustrate the response of the breakthrough curve to changes in the dispersion and the parameters describing adsorption with respect to strength, rate and nonlinearity. Each process thereby produced unique responses that confirm the high information content of closed-flow breakthrough data, which has shown to intrinsically contain information on the water content and the pumping rate as well. Finally, a numerical analysis of inverse parameter determination revealed a massive decrease in parameter uncertainty under optimal conditions, which renders the approach also highly relevant for practical applications.