Analytical solutions for contaminant fate and transport in parallel plate fracture-rock matrix systems with poiseuille flow
Junqi Huang,John Christ,Mark N Goltz
DOI: https://doi.org/10.1016/j.jhydrol.2021.126097
2021-05-01
Abstract:Modeling contaminant transport in fractured-rock matrix systems often approximates the effect of the parabolic flow field in the fractures (i.e., Poiseuille flow) on transport by adding a dispersion term to the uniform flow field. In this study, an analytical solution is derived to model contaminant transport in a parallel-plate fractured-rock matrix that explicitly simulates Poiseuille flow in the fractures, eliminating the need for the dispersion approximation. In addition to simulating Poiseuille flow in the fracture, the contaminant transport model developed here includes: (1) two-dimensional contaminant diffusion in the fractures and matrix, (2) first-order decay in the aqueous phase, and (3) rate-limited sorption onto matrix solids. It should be noted, however, that this model, much like the commonly employed Taylor dispersion approximation, neglects macro dispersion, thereby limiting the model's applicability to systems having wide fracture apertures with extremely high flow velocities (P e > 104). Model equations are analytically solved in the Laplace domain and numerically inverted. In addition, analytical expressions for the zeroth, first, and second spatial moments of the concentration profiles along the fractures are derived for both the new Poiseuille flow model as well as a model that approximates the effect of Poiseuille flow on transport by using a dispersion term. The first and second moment expressions are used to quantify how well the dispersion term approximates the effect of Poiseuille flow. Simulations confirm that the dispersion approximation will be adequate for natural fractures at long times. However, if a modeler is concerned with short-time transport behavior or transport behavior in systems with relatively wide-aperture fractures and high groundwater velocities where macro dispersion can be ignored, such as may be found at engineered geothermal systems and carbon capture and storage sites, there may be significant differences between model simulations that explicitly incorporate Poiseuille flow and those that approximate Poiseuille flow with a dispersion term. The model presented here allows the modeler to analytically quantify these differences, which, depending on the modeling objective, may cause the dispersion approximation to be inadequate. Simulations were also run to examine the effect of adsorption rate on remediation of fractured-rock matrix systems. It was shown that moderate adsorption rate constants could lead to very long remediation times, if remediation success is quantified by achieving low concentrations within the fracture.