Datapages, Inc.Print this page

ABSTRACT: Reservoir Characterization Utilizing Well Log and Seismic Data

REBEC, TONY, BRUCE CORNISH, and MIKE CURTIS, Halliburton Geophysical Services, Inc., Sugarland, TX

Accurately estimating the volume of in-place hydrocarbons and optimizing their recovery requires a detailed reservoir characterization. This objective demands integrating numerous data types, including geologic, seismic, and borehole information, into high-resolution spatially distributed estimates of such important reservoir parameters as porosity, lithology, fluid content, etc. A constrained inverse method by which the diverse data types can be integrated into a consistent model is presented.

The high resolution integrated inversion technique delineates reservoir stratigraphy from the combined data sets by wide band parameter estimation while allowing the analyst full interactive control over the process. Non-uniqueness is reduced by "a prior" knowledge contained in an initial petrophysical model, together with associated constraints and estimates of data uncertainty. This initial model, consistent with the analyst's geologic knowledge, represents the reservoir in terms of its fluid and lithologic parameters. It incorporates the high vertical resolution of the available wells together with the gross areal configuration of the reservoir's structure as determined from the seismic data. The model is interactively updated, within its constraints, using Bayesian inversion until its estimated responses approximate the various observed data types.

In order to retain the required high vertical resolution, the procedure does not model the earth as a series of equal time or depth thickness layers but rather represents each interface as a distinct event using a delayed pulse formulation. This not only reflects geologic insight but also allows the quantitative estimation of the reliability of the resulting model on a unit-by-unit basis. Interpreters may thus review the quality of the results at a particular zone without guesswork.

The technique is demonstrated with a case history of sand reservoirs in a clastic basin. The results are consistent with well, geologic constraints, and seismic data, and illustrate the great improvement in resolution and continuity obtained from this integrated technique over that available from conventional inversion methods.


AAPG Search and Discovery Article #91016©1992 AAPG-SEPM-SEG-EMD Pacific Section Meeting, Sacramento, California, April 27-May 1, 1992 (2009)