| This dissertation consists of three self-contained, yet closely related, papers summarizing the results of a systematic study on solute transport in heterogeneous aquifer systems containing well-connected networks of small-scale preferential flow paths (PFPs). The first paper demonstrates that the classical advection-dispersion (AD) model is of limited utility when applied to such aquifer systems. This is so even when contrasts in hydraulic conductivity (K) are small and variance in lnK is less than 0.10. The paper evaluates how well the AD model could be used to represent solute plumes transported through mildly heterogeneous 3-D systems characterized by a well-connected dendritic network of 10 cm-wide high-K channels. The results indicate that as the conductivity contrast between the channels and matrix increased, the simulated plumes in the channel-network system became more and more asymmetric, with little solute dispersed upstream of the plume center, and extensive downstream spreading of low solute concentrations. The second paper evaluates the applicability of the dual-domain mass transfer (DDMT) model, in a fundamental sense, in representing transport processes in the presence of decimeter-scale connected PFPs. As an alternative to the classical AD model, the DDMT model is demonstrated to be more appropriate for characterizing PFP-controlled transport behaviors under certain circumstances. Within each PFP network, the most critical factors that dictate the applicability of the DDMT model are PFP/matrix conductivity contrast, initial source configuration, and molecular diffusion rate. The DDMT model is more effective when the conductivity contrast or the diffusion rate coefficient is larger, or when there is more solute mass initially distributed in the high-conductivity PFPs. The third paper presents a direct association between average properties of dendritic PFP network systems and mass-transfer rate coefficients for the DDMT model. The paper shows that the value of the mass-transfer rate coefficient can be estimated directly using a physically based relation that requires no specific knowledge of the network geometry and no model calibration. The findings from this study significantly improve our current understanding of solute transport in highly heterogeneous media and will contribute to more accurate site characterization and groundwater remediation assessment. |
