ABSTRACT: From 3D Geological Models to Multi-phase Fluid Flows in Reservoirs: Some Recent Improvements

Guérillot, Dominique , IFP, Rueil Malmaison, France

Last 15 years, new concepts such as high resolution sequential stratigraphy and new software tools using geostatistics allow today to deliver to the reservoir engineer 3D models of its reservoir for a large variety of geological environments.

This presentation will illustrate some key issues in the methodology developed recently by our teams to make a better use of these models in fluid flow simulations:

Meshing (or Gridding): because of the complexity of geological objects, reservoir simulators needs more flexibility to properly follow these objects when their sizes is such that their impact on the fluid flow behaviour cannot be 'summarised' in the averaged petrophysical properties associated to the cells of this meshing;

Averaging petrophysical properties is another important pathway to handle carefully for simulating properly the fluid flow in highly heterogeneous reservoirs. This averaging (or up-scaling) problem, although is not always solvable using classical Darcy's equations coupled with the concept of capillarity, is particularly difficult to handle for fractured reservoirs when the size of these fractures are of the same order of magnitude or greater than the cell size;

Integrating dynamical data: inversion method, classically called history matching has made recent advances in the context of geostatistical methods. Bayesian inversion is one possible issue but an original method called scenario matching brings an interesting alternative.

This methodology will be illustrated through several case studies.

AAPG Search and Discovery Article #90913©2000 AAPG International Conference and Exhibition, Bali, Indonesia