Transport of polydisperse colloids in a saturated fracture with spatially variable aperture

Abstract

A particle tracking model is developed to simulate the transport of variably sized colloids in a fracture with a spatially variable aperture. The aperture of the fracture is treated as a lognormally distributed random variable. The spatial fluctuations of the aperture are described by an exponential autocovariance function. It is assumed that colloids can sorb onto the fracture walls but may not penetrate the rock matrix. Particle advection is governed by the local fracture velocity and diffusion by the Stokes-Einstein equation. Model simulations for various realizations of aperture fluctuations indicate that lognormal colloid size distributions exhibit greater spreading than monodisperse suspensions. Both sorption and spreading of the polydisperse colloids increase with increasing variance in the particle diameter. It is shown that the largest particles are preferentially transported through the fracture leading to early breakthrough while the smallest particles are preferentially sorbed. Increasing the variance of the aperture fluctuations leads to increased tailing for both monodisperse and variably sized colloid suspensions, while increasing the correlation length of the aperture fluctuations leads to increased spreading.

Contaminant transport in a fracture with spatially variable aperture in the presence of monodisperse and polydisperse colloids

“…11 is y=0.162 m. Finally, within an entrance element, the z-location of a colloid perpendicular to the fracture wall must be specified. It is assumed that the probability of a colloid entering a fracture element at a given z-location is proportional to the local flow rate in the corresponding parabolic velocity profile (Reimus 1995;James and Chrysikopoulos 2000). Furthermore, its centroid must be located a distance of at least d p /2 m from the fracture wall to ensure that it is wholly contained within the fracture.…”

“…A fully implicit finite difference technique is used to calculate the pressure within each fracture element by solving the Reynolds lubrication equation (Abdel-Salam and Chrysikopoulos 1995b;James and Chrysikopoulos 2000;Chrysikopoulos and James 2003). Average velocity components in the x-and y-directions are then calculated for every element using steady-state volumetric fluxes.…”

“…1 and 2, and diffuse isotropically in all directions. The general advective/diffusive particle tracking equations used for colloids are (Tompson and Gelhar 1990;James and Chrysikopoulos 2000):…”

Quantifying fracture geometry with X-ray tomography: Technique of Iterative Local Thresholding (TILT) for 3D image segmentation

“…The 2D steady-state local cubic law model [21,76] was selected for this purpose, and the same numerical schemes as in Deng et al [55] were used. In the application, there was no smoothing; i.e., the resolution of the flow simulation model was the same as the resolution of the image.…”

Colloid-Facilitated Transport of Radionuclides through the Vadose Zone

“…In the past several decades, considerable advances have been made on understanding the processes and mechanisms of colloid deposition and transport through laboratory and field studies, as well as numerical modeling Saiers and Hornberger, 1994;Johnson et al, 1996;McGechan et al, 2002]. It has been recognized that colloid mobilization is a function of many factors, which include properties of colloid and medium phases Zhuang et al, 2004], solution chemistry , and flow conditions [McDowell-Boyer, 1992;James and Chrysikopoulos, 2000;. However, 2 complex natural environmental conditions greatly preclude a complete characterization of colloid transport in vadose zones.…”