Stochastic Simulation of Cave systems in Reservoir Modeling.

in: Proc. 27th Gocad Meeting, Nancy

Abstract

Open cave systems in aquifer or collapsed paleocave systems in reservoir have a significant impact on hydraulic behavior because heterogeneities such as fractures, conduits and caves, constitute high permeability regions controlling fluid circulation. Exact location and geometry of these features are tediously assessed due to lack of data and complexity of the underlying physical processes. This paper proposes a new stochastic method to simulate possible location and geometry of cave system. Because fractures form preferential flow path, and therefor, initiate the development of conduits by dissolution processes, discrete fracture networks are first generated. A screening test is performed to remove fractures wich are not likely to yield large karsts. The distance function to the resulting fracture network is then computed. Finally, multiple realizations of distance cutoff values are generated by sequential gaussian simulations, and used to discriminate between cave occurences and host-rock.

Download / Links

    BibTeX Reference

    @INPROCEEDINGS{P309_Henrion,
        author = { Henrion, Vincent and Caumon, Guillaume and Vitel, Sarah },
         title = { Stochastic Simulation of Cave systems in Reservoir Modeling. },
     booktitle = { Proc. 27th Gocad Meeting, Nancy },
          year = { 2007 },
      abstract = { Open cave systems in aquifer or collapsed paleocave systems in reservoir have a significant impact
    on hydraulic behavior because heterogeneities such as fractures, conduits and caves, constitute high
    permeability regions controlling fluid circulation. Exact location and geometry of these features are
    tediously assessed due to lack of data and complexity of the underlying physical processes. This
    paper proposes a new stochastic method to simulate possible location and geometry of cave system.
    Because fractures form preferential flow path, and therefor, initiate the development of conduits by
    dissolution processes, discrete fracture networks are first generated. A screening test is performed
    to remove fractures wich are not likely to yield large karsts. The distance function to the resulting
    fracture network is then computed. Finally, multiple realizations of distance cutoff values are generated
    by sequential gaussian simulations, and used to discriminate between cave occurences and
    host-rock. }
    }