A second order in space accurate implicit scheme for time-dependent advection-dispersion equations and a discrete fracture propagation model are employed to model solute transport in porous media. We study the impact of the fractures on mass transport and dispersion. To model flow and transport, pressure and transport equations are integrated using a finite-element, node-centered finite-volume approach. Fracture geometries are incrementally developed from a random distributions of material flaws using an adoptive geomechanical finite-element model that also produces fracture aperture distributions. This quasistatic propagation assumes a linear elastic rock matrix, and crack propagation is governed by a subcritical crack growth failure criterion. Fracture propagation, intersection, and closure are handled geometrically. The flow and transport simulations are separately conducted for a range of fracture densities that are generated by the geomechanical finite-element model. These computations show that the most influential parameters for solute transport in fractured porous media are as follows: fracture density and fracture-matrix flux ratio that is influenced by matrix permeability. Using an equivalent fracture aperture size, computed on the basis of equivalent permeability of the system, we also obtain an acceptable prediction of the macrodispersion of poorly interconnected fracture networks. The results hold for fractures at relatively low density. © 2011 American Physical Society.
Bibliographical noteKAUST Repository Item: Exported on 2020-10-01
Acknowledged KAUST grant number(s): KUK-C1-017-12
Acknowledgements: We thank the sponsors of the ITF project on Improved Simulation of Flow in Fractured and Faulted Reservoirs, and the Technology Strategy Board (TSB), for supporting this research. This work was partially supported by the Center-in-Development Award to Utrecht University (No KUK-C1-017-12) by King Abdullah University of Science and Technology. Dr. Matthew Piggott receives our thanks for scientific feedback and comments. We thank the Imperial College High Performance Computing Service for the use of their cluster.
This publication acknowledges KAUST support, but has no KAUST affiliated authors.