Genetic-like modeling of fracture network constrained by static and connectivity data: Methodology and Case Study.

Francois Bonneau and Guillaume Caumon and Judith Sausse and Philippe Renard and Vincent Henrion. ( 2011 )
in: Proc. 31st Gocad Meeting, Nancy

Abstract

This paper reports on our investigations to integrate reservoir connectivity data in a geneticlike modeling of discrete fracture networks (DFN). DFNs are built using a three-step-workflow. Initialization: an object-based simulation initializes fracture seeds according to a fracture density map. Propagation: the growing direction is defined by neighboring fractures. Termination: the propagation stops when the fracture gets its final dimension or when it is connected to another fracture. Location and density of fractures are then iteratively updated using a fast-marching algorithm in order to match DFN connectivity and the observed one. This methodology is applied to the fractured granitic geothermal heat exchanger at Soultz-sous-Forˆts (France). Initial fracture e density map is deduced from microseismicity data, and then DFNs are modeled honoring static data and connectivity information.

Download / Links

BibTeX Reference

@INPROCEEDINGS{BonneauGM2011,
    author = { Bonneau, Francois and Caumon, Guillaume and Sausse, Judith and Renard, Philippe and Henrion, Vincent },
     title = { Genetic-like modeling of fracture network constrained by static and connectivity data: Methodology and Case Study. },
 booktitle = { Proc. 31st Gocad Meeting, Nancy },
      year = { 2011 },
  abstract = { This paper reports on our investigations to integrate reservoir connectivity data in a geneticlike modeling of discrete fracture networks (DFN). DFNs are built using a three-step-workflow.
Initialization: an object-based simulation initializes fracture seeds according to a fracture density map.
Propagation: the growing direction is defined by neighboring fractures.
Termination: the propagation stops when the fracture gets its final dimension or when it is connected to another fracture.
Location and density of fractures are then iteratively updated using a fast-marching algorithm in order to match DFN connectivity and the observed one.
This methodology is applied to the fractured granitic geothermal heat exchanger at Soultz-sous-Forˆts (France). Initial fracture e density map is deduced from microseismicity data, and then DFNs are modeled honoring static data and connectivity information. }
}