Use of Discrete Fracture Network Statistics for Construction of Two-Dimensional Continuous Time Random Walk Model

Abstract:

The Continuous Time Random Walk (CTRW) method provides a framework for modeling non-Fickian transport behavior through heterogeneous media by employing probability distributions to generate particle jump lengths and residence times spanning multiple orders of magnitude. In this work, we seek to formulate and parameterize a 2D CTRW directly from attributes of fracture networks with complex geometry. A Discrete Fracture Network (DFN) model is used to produce data on plume evolution over multiple spatial scales by synthetically generating a fracture network based on known fracture characteristics and conducting flow and particle tracking simulations under steady-state boundary conditions and flux-weighted particle migration. DFN Fracture segments, defined as the linear distance between two fracture intersections, are analyzed to define the distribution of jump lengths. The time for particles to migrate along these segments is recorded by the DFN model and is used to define a distribution of waiting times. These distributions provide a basis on which to formulate a CTRW to predict the migration of inert particles on a continuum of scales. The performance of the CTRW in simulating transport at multiple spatial scales is obtained by comparing spatial moments of the DFN plumes with CTRW solutions.