--> Abstract: Facies Based Modeling of a Highly Heterogeneous System - Kokongo Field, Cabinda, Angola, by C. J. Harrison, D. M. McKay, T. E. Sneed, M. Skalinski, J. P. Popek, and B. R. Power; #90942 (1997).

Datapages, Inc.Print this page

Abstract: Facies Based Modeling of a Highly Heterogeneous System - Kokongo Field, Cabinda, Angola


HARRISON, C.J., D.M. MCKAY, T.E. SNEED, M. SKALINSKI, J.P. POPEK, and B.R. POWER

The Kokongo Field located offshore Cabinda, Angola, began production in November 1994. The Field has an estimated 1.4 billion barrels of oil in place in the thick (over 2000ft), highly heterogeneous mixed siliciclastic/carbonate Pinda Formation. This reservoir is characterized by thin flow units and widely variable lithofacies. Modeling the reservoir for volumetric estimates and flow simulation requires innovative approaches to lithofacies determination, petrophysical analysis, and simulation model construction if the essential characteristics are to be preserved.

The model-building process was driven primarily by an extensive core-based data set, which included detailed lithofacies descriptions, depositional environment interpretations, and petrophysical measurements. These data were extended into non-cored sections via mineral-based formation evaluation methods and multivariate statistical analysis. The core and wireline data were then consolidated into stratigraphically consistent groups which provided the basis for subsequent facies-based geostatistical modeling of the reservoir. Fine scale (2000 layer) geostatistical models which honored the well data and the depositional environments interpreted from core were constructed using GoCad/G2 software. Complex porosity-permeability and saturation relationships were incorporated in the model by including facies group as a controlling factor.

The fine-scale geostatistical models of the Field contain approximately 5 million cells. This size of model is presently impractical for `real world' reservoir simulation. Selective coarsening of the models was accomplished using dynamic scale-up methods, giving final models with less than 100000 active cells which still accurately represent the flow properties of the reservoir.

AAPG Search and Discovery Article #90942©1997 AAPG International Conference and Exhibition, Vienna, Austria