Skip to Main content Skip to Navigation

Caractérisation et modélisation des écoulements dans les milieux fracturés

Abstract : Managing an aquifer relies on the understanding of its physical properties. We propose in this study a computational approach in order to characterize the physical structure of aquifers and the corresponding hydraulic properties. We focus on heterogeneous fractured media where flow is highly channeled. We first investigate the fracturing properties two Swedish sites by defining the corresponding DFN model. The DFN model framework is a power-law length distribution with a density term that depends on orientation, which parameters are calibrated with observations of fractures (lengths, orientations) by correcting the sampling bias. Hydraulic properties of heterogeneous porous and fractured media are then described by two channeling indicators based on statistical properties of flow and characterizing the distance between channels and their actual effective length. These indicators are used to relate the hydraulic properties to the physical structure of the medium. Finally, we develop a strategy to identify the position and the transmissivity of main flow channels from head data by solving the corresponding inverse problem. We show that this identification is possible provided that 1- data are remotely sensitive to the main structures underlying flow channels, 2- the parameterization focus on these structures and 3- the strategy is carefully set up. All these approaches can be combined to define a relevant model of the hydraulic properties of complex media.
Complete list of metadatas
Contributor : Isabelle Dubigeon <>
Submitted on : Friday, March 26, 2010 - 2:29:54 PM
Last modification on : Friday, April 5, 2019 - 8:03:16 PM
Long-term archiving on: : Friday, October 19, 2012 - 10:47:42 AM


  • HAL Id : tel-00467417, version 1


Romain Le Goc. Caractérisation et modélisation des écoulements dans les milieux fracturés. Sciences de la Terre. Université Rennes 1, 2009. Français. ⟨tel-00467417⟩



Record views


Files downloads