Modeling colloid transport in fractures with spatially variable aperture and surface attachment

• Published in: Journal of hydrology (Amsterdam) Vol. 566; pp. 735 - 742
• Main Authors: James, Scott C., Wang, Lichun, Chrysikopoulos, Constantinos V.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.11.2018
• Subjects: algorithms; Colloid transport; colloids; DLVO attachment; hydrology; isotropy; Particle tracking; preferential flow; simulation models; tuff; variance; Wall effects on diffusion; Wall effects on diffusion; DLVO attachment; Particle tracking; Colloid transport
• ISSN: 0022-1694; 1879-2707
• DOI: 10.1016/j.jhydrol.2018.09.047

•Colloid-transport simulations through rough fractures subject to DLVO forces.•Transport depended on how surface potentials correlated with the aperture field.•As surface potentials correlated to apertures, first and second moments were altered. A particle-tracking algorithm was developed to simulate colloid transport subject to wall effects on diffusion as well as colloid surface attachment as described by DLVO kinetics. The effects of spatially variable fracture surface potential, which contributed to spatially variable attachment strength affecting colloid transport, were investigated. The fracture surface potential was assumed to be either positively, neutrally (zero), or negatively correlated with the lognormally distributed local fracture aperture, described with a mean, variance, and isotropic correlation length. The results from several model simulations indicated that wall effects were negligible for the synthetic fractures studied here. When fracture surface potential was negatively correlated with local aperture, colloids were preferentially transported through the fracture, because they tended to enter high-flow, large-aperture regions where they underwent less attachment (have the largest first moment measured upon exit of the first colloid from the fracture). The variance (second moment) increased for flowing colloids when comparing negatively to zero and then positively correlated surface potentials to fracture apertures, because spreading notably increased when suspended colloids were temporarily attached onto fracture surfaces. For colloids attached onto fracture surfaces, both first and second moments decreased from negatively, to neutrally (zero), to positively correlated surface potentials to apertures. This is an intuitive result, consistent with fewer colloids attaching along the larger aperture preferential flow paths.