A better understanding of a Uinta Basin channelized analog reservoir through geostatistics and reservoir simulation

Abstract

The Green River Formation is located in the Uinta basin of northeastern Utah. It contains several reservoirs that can be classified as lacustrine such as the Altamont-Bluebell and Red Wash. Lacustrine reservoirs are abundant in other provinces in the world such as China, Southeast Asia, Brazil, West Africa, and the Caspian Sea. Even though they can contain important accumulations of hydrocarbons, our understanding of the primary controls on fluid flow within these systems is still not clear. This ambiguity leads in some cases to inefficient recovery of hydrocarbons in such reservoirs. This study is aimed at clarifying the effects of heterogeneities in channelized reservoirs on fluid flow. It uses a multidisciplinary approach combining geologic knowledge with reservoir engineering. It involves the geologic modeling and fluid flow simulation of a channelized outcrop of the Green River formation. The study of this outcrop provides insights for modeling, understanding, and possibly predicting the behavior of channelized oil and gas reservoirs. Results show that the number of channels in the model can have a significant effect on performance. The rock properties in these channels and the channel paths are also important factors that determine the recovery efficiency. Other findings include the effect on performance of vertical anisotropy in a channelized reservoir. We discovered that an isotropic reservoir performs better than an anisotropic one and that the well perforation interval is extremely important when comparing the performance of several anisotropic cases. Finally, we investigated the effects of the recovery strategy on performance in a channelized setting. We found that waterflooding yields better results than any of the other recovery techniques analyzed. Sensitivity runs with different waterflood patterns indicated that a staggered line drive results in the best performance in the analog channelized reservoir we modeled, as it allows for the best recovery factor in the least amount of time. The results of this work can be used qualitatively to predict performance in a channelized setting but their use is limited quantitatively because of the issue of scale, i.e. the outcrop width is much less than typical interwell scale.