An experimental and computational investigation of solute transport and nonaqueous-phase-liquid dissolution in discrete rough-walled fractures

Dissolution of residual nonaqueous-phase-liquids (NAPLs) entrapped in fractured subsurface media can create long term sources of groundwater contamination. Thus, understanding what parameters influence dissolution rates is critical to evaluating environmental risk and developing clean-up strategies. NAPL dissolution is influenced by the coupled processes of fluid flow, dissolved-NAPL transport, interphase mass transfer, and the corresponding NAPL-water interface movement. Predicting how these processes interact in complex fracture networks requires first understanding the dissolution process at the scale of a single fracture. This dissertation investigates solute transport and NAPL dissolution in discrete rough-walled fractures through a combination of physical experiments and computational simulations.; Physical experiments in transparent analog fractures allowed nondestructive measurements of fracture aperture, solute concentration, and NAPL distribution over the entire flow field (15 x 30 cm) during experiments. To allow unambiguous comparisons between experiments and computational simulations, the light transmission measurement system was thoroughly evaluated and a refined measurement protocols were developed. Using these protocols yielded measurements of unprecedented accuracy and spatial resolution (0.015 x 0.015 cm).; Developing fracture-scale descriptions of solute transport and NAPL dissolution requires experiments and/or computational simulations at scales that are considerably larger than the features that define the flow field. In rough-walled fractures, these features are aperture variability and the geometry of the entrapped NAPL. Numerical solution of the three-dimensional equations that govern dissolution at an adequately large scale is currently computationally intractable. Thus, a primary focus of this dissertation has been to develop and evaluate the validity of depth-averaged models of solute transport and NAPL dissolution through direct comparison to physical experiments.; Careful evaluation of the inherent limitations of the depth-averaged modeling approach with respect to flow and transport suggested that we can define conditions under which the depth-averaged model of solute transport is reasonable. Subsequent comparisons of depth-averaged computational simulations of NAPL dissolution to a physical experiment demonstrated that the depth-averaged approach yields excellent predictions of both the evolution of the NAPL distribution within the fracture and the rate of dissolution. These results suggest that the depth-averaged modeling approach can be used to investigate the influence of different parameters on dissolution in discrete fractures. Results of such studies may provide fracture-scale constitutive relationships that can be used in larger scale network models of solute transport and dissolution.