Integrated geostatistical three-dimensional modeling and simulation of Mississippian St. Louis carbonate reservoir systems in Kansas

In the Hugoton Embayment of southwestern Kansas, St. Louis Limestone reservoirs have produced over 300 million barrels of oil. Reservoir units consist of relatively thin (<4m) oolite skeletal grainstone shoals that have a scattered distribution, resulting in exploration and production challenges. The geometry and distribution of facies control the heterogeneity of the reservoirs, which results in relatively low recovery efficiencies. Quantifying the spatial distribution of oolitic deposits can contribute to understand sedimentologic processes and efficient management of oolitic reservoirs.; Rock facies classification, external facies geometry, and internal rock property distribution are fundamental to reservoir simulation and future hydrocarbon prediction of oolitic reservoirs. Lack of high resolution seismic data provides a significant challenge. An integrated geostatistical approach is presented in this study using available well data to improve stratigraphic modeling of oolitic systems and corresponding streamline simulation of selected oolitic reservoirs in the St. Louis Limestone of southwestern Kansas. This approach investigates neural network and stochastic simulations to integrate different types of data (core data, log data, top data and production data), at different scales (vertical, horizontal, fine scale core data, coarse well-log data) and degrees of quantification (facies, log, well data) to improve models of St. Louis (Mississippian) carbonate reservoir systems in selected fields (including Archer, Big Bow and Sand Arroyo Creek fields) of southwest Kansas.; The application of three-dimensional models provides insights into facies distribution patterns, external and internal geometry, controls on the distribution of oolitic deposits, and the sedimentologic processes that formed reservoirs in the St. Louis Limestone. This study provides: (1) 3D stochastic simulations of facies distribution of St. Louis oolitic reservoirs; (2) improved reservoir framework models for carbonate shoal reservoirs; (3) improved understanding of spatial distribution and variability of rock properties; (4) 3D visualization of the St. Louis carbonate reservoir systems; (5) streamline simulations of the static geostatistical models to rank and determine the efficacy of the modeling procedure; (6) better understanding of factors controlling the facies distribution and the production of hydrocarbons within carbonate shoal reservoir systems. Geostatistical 3D modeling methods are applicable to other complex carbonate or siliciclastic reservoirs in shallow marine settings.