@MISC{Szymczak_microscopicsimulations, author = {Piotr Szymczak and Tony Ladd}, title = {MICROSCOPIC SIMULATIONS OF THE DISSOLUTION OF ROCK FRACTURES}, year = {} }

Share

OpenURL

Abstract

The results of numerical simulations of dissolution in fractured rocks are reported. The model is microscopic, with a detailed representation of the topography of the fracture. The velocity field in the fracture is assumed to be Stokes flow and is efficiently calculated with an implicit lattice-Boltzmann technique. The transport of dissolved species in the pore spaces is modelled by an innovative random walk algorithm that incorporates the chemical kinetics at the solid surfaces. The simulated morphological changes in a complex fracture are compared with laboratory experiments. The processes leading to dissolution of a fractured rock by a reactive fluid depend on a subtle interplay between chemical reactions at mineral surfaces and fluid motion in the pores. The complex geometry of a typical fracture makes both numerical and theoretical calculations very demanding. Existing models of fracture dissolution are rarely constructed on a microscopic (pore-scale) level. Instead, various approximations are usually resorted to in order to make the analysis more tractable. For example, instead of computing a three-dimensional velocity field by solving the Navier-Stokes equations, the Reynolds (or lubrication) approximation is used [1, 2, 3], so that the volumetric flow rate is proportional to the cube of the local aperture. It has been shown [4, 5, 6] that the Reynolds equation may significantly overestimate the flow, especially for fractures of high roughness and small apertures. The complicated topography of the fracture is also the reason why the transport of the dissolved material from the walls into the bulk of the fluid is usually accounted for in a simplified way, with the effects of convection assumed to be adequately expressed by a Sherwood number for transport in