Interdisciplinary Analysis of Fluvial Dominated Deltaic Reservoirs: Glenn Pool Field, Creek County, Oklahoma

Gokay Bozkurt, Liangmiao Ye, Asnul Bahar

Glenn Pool Field, located on the Northeastern Oklahoma Platform, has been under production since 1905. However, it still contains 80% of the original oil in place, due to the high reservoir complexity. The field study is focused on a 160-acre block (Self Unit) which was under primary production, then was gas pressurized and has been under water flood ever since. At present, a 99% water cut makes it necessary to develop a better management plan. A detailed multi-disciplinary (Geology, Geophysics, Petroleum Engineering) reservoir study with state-of-the-art technology was undertaken in this mature, marginal oil field.

Self Unit was studied with conventional methods prior to drilling a cooperative project well (Uplands Self #82). Stratigraphic framework of the Glenn Sand reservoir has been established through a series of northeast to southwest stratigraphic cross sections. Six discrete genetic intervals (DGI) for the 120-140 ft thick reservoir were delineated from well log correlations. Channel-fill, splay, channel-mouth bar, levee and interdistributary mudstone facies were recognized from well log profiles and core analysis. Attempts were made in simulating geology using sequential indicator simulation methods; results were not entirely satisfactory.

Self #82 was drilled in late December, 1993. The project objectives for drilling the well were: 1) evaluate reservoir predictions; 2) collect data using conventional and advanced technologies. Pre-drilling facies architectural characterization was reasonably successful. In addition to the conventional well log suites and core, advanced technologies including microresistivity imaging log and crosswell tomography were applied. Simulation of the DGI distribution was undertaken using truncated Gaussian simulation method. Simulation results strongly agree when comparing probability input distributions with the output distributions. Porosity distribution was simulated using the simulated annealing method. For a selected well location (Self #82) the comparison between simulated and core poro ity is very good although core data were not used as conditioning data. The permeability distribution was transformed from porosity using a conditional distribution approach and was validated using well test data. Crosswell transmission and migrated reflection tomography images between Self #82 and three offset wells constrained lateral reservoir continuity.

A reservoir management plan, which is to be implemented in the last quarter of 1994, was developed from reservoir performance simulation, well test data, facies architecture and crosswell tomography.

AAPG Search and Discovery Article #91020©1995 AAPG Annual Convention, Houston, Texas, May 5-8, 1995