Computer Modelling Group Ltd.

Low primary recovery percentages from unconventional reservoirs have long motivated interest in Enhanced Oil Recovery (EOR) for these reservoirs, resulting in numerous simulation studies and injection pilots. However, performance from injections pilots has typically been disappointing compared to the simulations, suggesting that reservoir permeability and heterogeneity are not adequately described in the reservoir simulation models.

In this study, a simulation and history-matching approach was used to quantify the permeability matrix over a six-section, nine-well area. Twelve years of production data were history-matched, using a combination of pressure-dependent permeability and enhanced permeability to represent natural fractures or other high-permeability features. Also, the performance of a failed injection pilot was history-matched to determine the level of reservoir heterogeneity needed to explain the pilot failure.

Based on this study, a reservoir description capable of matching twelve years of production and injection history has been developed. Formation properties in the high-permeability streaks capable of causing the disappointing injection pilot performance have been quantified. Recovery has been forecast to depletion, and EOR under hydrocarbon gas injection has been forecast for a variety of scenarios. Optimal operating strategies and recommendations for technology development to mitigate early breakthrough are made. Realistic cost estimates were made for each scenario, and economics were run for each recovery method. These results give insight into the economic potential of enhanced oil recovery in the Elm Coulee Bakken formation. Recommendations for favorable tax treatment and scheduling of expenses/investments are made.

Developing the permeability matrix using the history matching approach is a novel and versatile way of quantifying unconventional reservoir properties. However, it is important to match both injection and production data, since the permeability vector appears to have pressure-dependent effects. The effect of controlling injection thief zones by controlling local wellbore outflow is quantified, and a need for in situ permeability modification of fracture thief zones has been determined.