Physical and numerical fracture network models have been used to analyze the transport of conservative solutes in systems of parallel‐sided fractures. The processes controlling dispersion in fracture systems that are explicitly simulated by the numerical model are (1) development of a velocity profile within individual fractures, (2) transverse molecular diffusion between streamlines, both within fractures and at fracture junctions, and (3) advection with the bulk fluid through a system of fractures with a range of hydraulic gradients and apertures. The first two processes, referred to as microdispersion, are often assumed to be secondary to the third, referred to as macrodispersion. The validity of this assumption is, however, highly dependent on the hydrodynamics of the system under consideration. Data collected from a physical model of a fracture network are used to validate a numerical model that explicitly simulates all three transport processes. The numerical model is then used to evaluate the relevance of microdispersion processes in a system where macrodispersion is significant.
Read full abstract