Modélisation des incertitudes concernant la géométrie des failles. Impact sur les études volumétriques des réservoirs pétroliers

Magali Lecour. ( 2000 )
INPL, Nancy, France

Abstract

Uncertainty studies are currently a key point for petroleum exploration. lndeed, available data for deep reservoirs modelling are often uncertain, scattered and inaccurate. 50, resuIting uncertainties concern many parameters: the reservoir geometry, the facies distribution and rock properties. In order to minimize risk, the exploration phase must give, not only one, but severaI models scanning all the possible range for the zone of interest. However, data concerning fauIt geometries are less precise than horizon data, fauIt modelling must respect more mechanical rules and structural model updating is more difficuIt in faulted domain, because more contacts must be preserved. Consequently, only few studies on fauIt geometry uncertainties have been performed at the time. The work presented in this thesis, concerns one part of fauIt geometry uncertainties: the modification of their geometry (location, dip, and shape in map view) around a reference model, at constant fauIt network scheme. Most of this work consisted in creating new data structures: • a fauIt object that stores uncertainty estimations and allows for a fast computation of new locations for each fauIt point. The simulation method is based on the Pfield technique and try to preserve the initial geometry at best • a contact manager storing all the contacts between the structural model interfaces (fauIts, horizons and boundary of the model). This structure allows also the model updating after each simulation loop • a fauIt network simulation manager, that not only simulate new geometries for the fauIts, but also update the structural model afterwards. The main goal of this work is to propose a global modular algorithm. This algorithm can give either the possibility to follow the evolution of reservoir volume estimations during the simulation pro cess, or the possibility to select several realizations of the structural model geometry. The la st option enables resevoir engineers to study other uncertainties like rock properties (porosity, permeability), or oil-water contacts.

Download / Links

BibTeX Reference

@PHDTHESIS{Lecour00MI,
    author = { Lecour, Magali },
     title = { Modélisation des incertitudes concernant la géométrie des failles. Impact sur les études volumétriques des réservoirs pétroliers },
   chapter = { 0 },
      year = { 2000 },
    school = { INPL, Nancy, France },
  abstract = { Uncertainty studies are currently a key point for petroleum exploration. lndeed,
available data for deep reservoirs modelling are often uncertain, scattered and inaccurate.
50, resuIting uncertainties concern many parameters: the reservoir geometry,
the facies distribution and rock properties.
In order to minimize risk, the exploration phase must give, not only one, but severaI
models scanning all the possible range for the zone of interest. However, data concerning
fauIt geometries are less precise than horizon data, fauIt modelling must
respect more mechanical rules and structural model updating is more difficuIt in faulted
domain, because more contacts must be preserved. Consequently, only few studies
on fauIt geometry uncertainties have been performed at the time.
The work presented in this thesis, concerns one part of fauIt geometry uncertainties:
the modification of their geometry (location, dip, and shape in map view) around
a reference model, at constant fauIt network scheme.
Most of this work consisted in creating new data structures:
• a fauIt object that stores uncertainty estimations and allows for a fast computation
of new locations for each fauIt point. The simulation method is based on the
Pfield technique and try to preserve the initial geometry at best
• a contact manager storing all the contacts between the structural model interfaces
(fauIts, horizons and boundary of the model). This structure allows also the
model updating after each simulation loop
• a fauIt network simulation manager, that not only simulate new geometries for
the fauIts, but also update the structural model afterwards.
The main goal of this work is to propose a global modular algorithm. This algorithm
can give either the possibility to follow the evolution of reservoir volume estimations
during the simulation pro cess, or the possibility to select
several realizations of the structural model geometry. The la st option enables resevoir
engineers to study other uncertainties like rock properties (porosity, permeability), or
oil-water contacts. }
}