3D fracture network models from DOM : extraction , modelling , scales and uncertainties

Sophie Viseur and Juliette Lamarche and Jules Fleury and S{\'{e}}bastien Chatel{\'{e}}e and Nadjet Bachtarzi. ( 2017 )
in: 2017 Ring Meeting, pages 150569, ASGA

Abstract

Numerical outcrops are key data for modelling the 3D spatial organization of fracture networks, especially in the case of fracture corridors. Indeed, few studies have been focused on characterizing and modelling the spatial organization of fracture corridors. Fracture cor- ridors are particular geological structures within which the fracture density is particularly high. Recently, it was highlighted that the internal architecture of fracture corridors is more complex than expected: 1) the fracture density may vary within the fracture corridors; 2) other geological structures such as breccia are also encountered. Fractures have major impacts on fluid flow in carbonates. Geostatistical approaches were proposed to generate 3D fracture networks for reservoir characterization. They require in- formation about fracture size, orientation and density. As only 1D data are available to describe fractures in subsurface, 3D fracture models obtained from outcrop studies are then of paramount importance for reservoir analog characterization. In this paper, an approach is proposed to automatically extract fracture planes from DOM and to build the 3D models. Fracture patches are first extracted. Second, these patches are correlated depending on their relative position and orientation. The correlation can be per- formed deterministically or stochastically, in order to propose different possibilities in case of uncertainties on the fracture continuities. Third, the 3D model is built using Gocad. An interesting work is to compare the fracture density obtained from fractures depicted thanks to the outcrop geometry and the ones depicted along scan-lines. In this study, 3D Digital Outcrop Models (DOMs) were acquired from outcrops in the South- East of France. They were used as a support for interpreting and modelling fracture networks to study the fracture patterns in corridors. The proposed approach was applied onto these DOMs to constrain 3D modelsof fractures.

Download / Links

BibTeX Reference

@INPROCEEDINGS{Viseur2017,
    author = { Viseur, Sophie and Lamarche, Juliette and Fleury, Jules and Chatel{\'{e}}e, S{\'{e}}bastien and Bachtarzi, Nadjet },
     title = { 3D fracture network models from DOM : extraction , modelling , scales and uncertainties },
 booktitle = { 2017 Ring Meeting },
      year = { 2017 },
     pages = { 150569 },
 publisher = { ASGA },
  abstract = { Numerical outcrops are key data for modelling the 3D spatial organization of fracture networks, especially in the case of fracture corridors. Indeed, few studies have been focused on characterizing and modelling the spatial organization of fracture corridors. Fracture cor- ridors are particular geological structures within which the fracture density is particularly high. Recently, it was highlighted that the internal architecture of fracture corridors is more complex than expected: 1) the fracture density may vary within the fracture corridors; 2) other geological structures such as breccia are also encountered. Fractures have major impacts on fluid flow in carbonates. Geostatistical approaches were proposed to generate 3D fracture networks for reservoir characterization. They require in- formation about fracture size, orientation and density. As only 1D data are available to describe fractures in subsurface, 3D fracture models obtained from outcrop studies are then of paramount importance for reservoir analog characterization. In this paper, an approach is proposed to automatically extract fracture planes from DOM and to build the 3D models. Fracture patches are first extracted. Second, these patches are correlated depending on their relative position and orientation. The correlation can be per- formed deterministically or stochastically, in order to propose different possibilities in case of uncertainties on the fracture continuities. Third, the 3D model is built using Gocad. An interesting work is to compare the fracture density obtained from fractures depicted thanks to the outcrop geometry and the ones depicted along scan-lines. In this study, 3D Digital Outcrop Models (DOMs) were acquired from outcrops in the South- East of France. They were used as a support for interpreting and modelling fracture networks to study the fracture patterns in corridors. The proposed approach was applied onto these DOMs to constrain 3D modelsof fractures. }
}