Genetic-like Modeling of Discrete Fracture Networks.

in: Proc. 30th Gocad Meeting, Nancy

Abstract

Fracture networks exert a strong influence on subsurface flow and therefore call for appropriate technology in a wide variety of geoscience fields. This paper reports on our investigations to generate stochastic fracture models that are consistent with patterns observed on outcrops and fracture growth principles. The main idea was to develop a technique which simulates evolving fracture networks. It results in a genetic-based propagation method which grows iteratively 3D fractures. The algorithm relies on heuristic rules imitating the mechanics of fracture initiation, propagation, interactions and termination. As compared to mechanical models, our approach can generate a large number of fractures. Also, the geometry and flow properties of our discrete fracture networks (DFN) are expected to be closer to those observed on natural analogs than with classical DFNs made of planar fracture which tend to underestimate connectivity.

Download / Links

BibTeX Reference

@INPROCEEDINGS{HenrionGM2010,
    author = { Henrion, Vincent and Caumon, Guillaume },
     title = { Genetic-like Modeling of Discrete Fracture Networks. },
 booktitle = { Proc. 30th Gocad Meeting, Nancy },
      year = { 2010 },
  abstract = { Fracture networks exert a strong influence on subsurface flow and therefore call for appropriate technology in a wide variety of geoscience fields. This paper reports on our investigations to generate stochastic fracture models that are consistent with patterns observed on outcrops and fracture growth principles.
The main idea was to develop a technique which simulates evolving fracture networks. It results in a genetic-based propagation method which grows iteratively 3D fractures. The algorithm relies on heuristic rules imitating the mechanics of fracture initiation, propagation, interactions and termination. As compared to mechanical models, our approach can generate a large number of fractures. Also, the geometry and flow properties of our discrete fracture networks (DFN) are expected to be closer to those observed on natural analogs than with classical DFNs made of planar fracture which tend to underestimate connectivity. }
}