The Burgan Sand reservoir is the largest and most prolific reservoir in the Greater Burgan field in Kuwait. As development of this reservoir proceeds, greater attention is being paid to the more heterogeneous upper part of the reservoir, the Upper Burgan Sands, which contain significant, multi-billion barrel, resources. A conceptual Field Development Plan was developed to enable an assessment of performance under waterflooding.
Due to the complexity of the Upper Burgan Sands and the very large areal extent of the Greater Burgan field, it was considered impractical to model the reservoir in appropriate detail in the full field simulation model. An alternative modeling approach was adopted. This involved developing a number of part field geological models which were subsequently used to develop high resolution dynamic simulation models.
This paper describes the challenges of choosing an appropriate range of part field models that capture variation in geological characteristics and of hydrocarbon properties, conditioning these models to available dynamic data and accounting for the historical production.
Particular attention is paid to three problems. First, to rank the models in terms of key features that would be expected to govern reservoir performance. These included internal heterogeneity, average reservoir quality, connectivity to the underlying Middle Burgan Sands and variation in fluid properties. Second, to assign the selected models to represent different parts of the reservoir. Third, to ensure that historical performance, including water influx inferred from surveillance data, is accounted for in predicting future performance.
The estimates were determined by a hybrid formulation, combining an analytic and simulation approach, suited for rapid computations and multi-scenario generation. Simulation results of water-cut versus recovery have been integrated with standard analytic expressions for fully developed pattern waterfloods. Scenarios investigated include the determination of ultimate recovery, a phased out recovery according to production and sweep of lower intervals, sensitivities to drilling rate, and a waterflooding scheme prioritized on area production potential.
Based on the results from the part field models, performance estimates of the Upper Burgan Sands reservoir have been made. Volumes forecasts and associated well numbers have been predicted.
A new reservoir simulation and analytical formulation has been developed to enable rapid predictions of waterflood scenarios. The hybrid formulation has proved to be significantly faster, resulting in a much quicker turnaround time compared to a traditional simulation study.