AAPG Bulletin, V. 83 (1999),
No. 1 (January 1999),
P. 19-42.
Modeling a Vuggy Carbonate
Reservoir, McElroy Field, West Texas1
Kaveh Dehghani,2
Paul
M. (Mitch) Harris,2 Kelly A. Edwards,3 and William
T. Dees3
©Copyright 1999. The American Association of
Petroleum Geologists. All Rights Reserved
1Manuscript received July 15, 1997;
revised manuscript received April 27, 1998; final acceptance May 27, 1998.
2Chevron Petroleum Technology Company,
1300 Beach Boulevard, La Habra, California 90631-6374; e-mail: kdeh@chevron.com;
hapm@chevron.com
3Chevron North America Production
Company, 15 Smith Road, Claydesta Plaza, Midland, Texas 79705; e-mail:
kaed@chevron.com; wtde@chevron.com
Many of our co-workers at Chevron provided
input at various stages in our work. We are especially grateful to Minhtrang
Doan for conducting some of the simulation runs; Ed Donovan for SEM photos
and thin section preparation; Leon Roe, Robert Meyer, Robert Ehrlich, and
Alan Bernath for their help and constructive comments. Chevron Petroleum
Technology Company and Chevron North America Exploration and Production
Company gave permission to publish.
ABSTRACT
The McElroy field produces approximately
17,000 BOPD (barrels of oil per day) under a mature waterflood from the
Permian Grayburg Formation. The main pay zone in the reservoir is primarily
peloidal dolograinstones/packstones with interparticle/intercrystalline
porosities. The central portion of the field is more heterogeneous because
of thin high-porosity and high-permeability vuggy zones. The occurrence
of these zones is confirmed by core description and measurements, porosity
logs, tracer studies, and injectivity measurements. These thin high-porosity
and high-permeability vuggy zones diminish waterflood effectiveness and
leave millions of barrels of bypassed oil in the lower permeability matrix.
A method was developed to identify the vuggy zones
on logs, create geostatistical models of porosity and permeability incorporating
the vuggy zones, and characterize them in simulation models. The methodology
involved the following: (1) developing a log trace to identify zones of
high secondary porosity, mainly vuggy porosity, in the area of the field
that was modeled, (2) creating a detailed geostatistical model (1 million
cells) of total porosity using well-log data, (3) creating a geostatistical
permeability model based on total porosity, (4) creating a separate detailed
geostatistical model of secondary porosity, and (5) superimposing exceptionally
high permeability in areas of the permeability model defined by high secondary
porosities.
The detailed permeability models were scaled-up
to 12,000-cell models for simulation studies. The models incorporating
vuggy permeability distributions showed a far superior history match of
primary and waterflood processes than did models that did not incorporate
vuggy permeability; these models also showed good-quality history matches
for individual wells. Successful history matching of the simulation models
validates our method and indicates that core data underestimate the permeability
of vuggy zones due to sampling and measurement issues.