Use of different mesh types to run flow simulations - reservoir application

Margaux Raguenel and Pierre Samier and Gilles Darche. ( 2021 )
in: 2021 RING Meeting, ASGA

Abstract

The understanding of the subsurface, to predict rocks and fluids behaviour and extract natural resources, calls for its numerical representation. 3D models are designed to describe multi-scale information, going from the geometry of the subsurface with all its geological features to operational constraints such as well trajectories. Once discretized into a 3D mesh, this is also the support to run physical simulations, which are a mathematical representation of the multi-phase flow dynamics and geomechanical behaviors occurring underground. Therefore, each model is the result of several choices and compromises done to obtain the best possible representation depending on the problem to solve. There is no general answer to the following issues: Which mesh type is more adapted to which geological context, to which physical phenomena, or to which operational constraints? In this perspective, the choice of the type of mesh used appears as a key parameter to represent the subsurface, and each one must be tested to determine its potential benefits. Each of them also triggers several adaptations to the standard industrial workflow to run physical simulations. \\ This work presents the global workflow as well as the needed adaptations regarding each type of mesh. Results of a two-phase fluid flow simulation run on a realistic study on a full reservoir discretized with three mesh types is presented.

Download / Links

BibTeX Reference

@INPROCEEDINGS{RAGUENEL_RM2021a,
    author = { Raguenel, Margaux and Samier, Pierre and Darche, Gilles },
     title = { Use of different mesh types to run flow simulations - reservoir application },
 booktitle = { 2021 RING Meeting },
      year = { 2021 },
 publisher = { ASGA },
  abstract = { The understanding of the subsurface, to predict rocks and fluids behaviour and extract natural resources, calls for its numerical representation. 3D models are designed to describe multi-scale information, going from the geometry of the subsurface with all its geological features to operational constraints such as well trajectories. Once discretized into a 3D mesh, this is also the support to run physical simulations, which are a mathematical representation of the multi-phase flow dynamics and geomechanical behaviors occurring underground. Therefore, each model is the result of several choices and compromises done to obtain the best possible representation depending on the problem to solve. There is no general answer to the following issues: Which mesh type is more adapted to which geological context, to which physical phenomena, or to which operational constraints? In this perspective, the choice of the type of mesh used appears as a key parameter to represent the subsurface, and each one must be tested to determine its potential benefits. Each of them also triggers several adaptations to the standard industrial workflow to run physical simulations. \\ This work presents the global workflow as well as the needed adaptations regarding each type of mesh. Results of a two-phase fluid flow simulation run on a realistic study on a full reservoir discretized with three mesh types is presented. }
}