Power-law velocity distributions in fracture networks: Numerical evidence and implications for tracer transport

Scott Painter, Vladimir Cvetkovic, Jan Olof Selroos

Research output: Contribution to journalArticlepeer-review

58 Scopus citations

Abstract

Velocity distributions in two- and three-dimensional networks of discrete fractures are studied through numerical simulations. The distribution of 1/v, where v is the velocity along particle trajectories, is closely approximated by a power law (Pareto) distribution over a wide range of velocities. For the conditions studied, the power law exponents are in the range 1.1-1.8, and generally increase with increasing fracture density. The same is true for the quantity 1/bv, which is related to retention properties of the rock; b is the fracture half-aperture. Using a stochastic Lagrangian methodology and statistical limit theorems applicable to power-law variables, it is shown that the distributions of residence times for conservative and reacting tracers are related to one-sided stable distributions. These results are incompatible with the classical advection dispersion equation and underscore the need for alternative modeling approaches.

Original languageEnglish
Article number1676
JournalGeophysical Research Letters
Volume29
Issue number14
DOIs
StatePublished - Jul 15 2002
Externally publishedYes

Fingerprint

Dive into the research topics of 'Power-law velocity distributions in fracture networks: Numerical evidence and implications for tracer transport'. Together they form a unique fingerprint.

Cite this