Skip to Main Content
Book Chapter

5: Knowledge Representation in a Workstation for Reservoir Analysis

By
G. M. E. Lafue
G. M. E. Lafue
Andersen Consulting, Plazza Pablo Puiz Picasso, Torre Picasso, 28020 Madrid, Spain. Formerly Schlumberger Well Services.
Search for other works by this author on:
D. J. Pajot
D. J. Pajot
Schlumberger Well Services, 8311 North FM 620 Rd., Austin, TX, 78726.
Search for other works by this author on:
A. J. Winchester
A. J. Winchester
Services Techniques Schlumberger, 50, Avenue Jean Jaures, B.P. 362, 92541 Montrouge Cedex, France. Formerly Schlumberger Well Services.
Search for other works by this author on:
Published:
January 01, 1991

Abstract

When analyzing complex reservoirs, interpretation professionals have long felt the need for an integration of all forms of data. The interpretation system described in this paper integrates applications from geophysics, petrophysics, geology, and reservoir engineering around a common database that contains both the data and the resulting interpretation model.

The database includes both large numeric files (e.g., to store measurements for seismic cubes or sections, well logs or maps) and declarative, symbolic descriptions of these files (e.g., the field or wells to which they belong or how they were derived from one another). The symbolic descriptions, called knowledge bases, are implemented in an object-oriented Knowledge Representation System that made possible the integrating of notions that come from several disciplines and that are in constant evolution. Object-oriented modeling was also used to develop and execute a fast graphics code and to support tight interactions between graphics and the knowledge bases.

A system was built to analyze surface seismic and log data in an interactive environment. The system uses Knowledge Representation techniques to represent data and model in knowledge bases. The results of single well log interpretation are first used to calibrate surface seismic data, and to define at the well locations the reservoir components to be reconstructed. Seismic horizons are identified, interpreted, and mapped from seismic data, and are then used to model the geometry of the reservoir units. Log data provide measurements of the reservoir properties to be mapped within the components. Finally, the components are stacked together and their properties are interpolated on a simulation grid.

You do not currently have access to this article.

Figures & Tables

Contents

Geophysical Developments Series

Expert Systems in Exploration

Fred Aminzadeh
Fred Aminzadeh
Search for other works by this author on:
Marwan Simaan
Marwan Simaan
Search for other works by this author on:
Society of Exploration Geophysicists
Volume
3
ISBN electronic:
9781560802532
Publication date:
January 01, 1991

GeoRef

References

Related

Citing Books via

Close Modal
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close Modal
Close Modal