Umiat field, located in Alaska North Slope (ANS) poses unique development challenges because of its remote location and permafrost within the reservoir. This hinders the field development, and further leads to a potential low expected oil recovery. The objective of this work is to assess various possible well patterns of the Umiat field development, and perform a detailed parametric study to maximize oil recovery and minimize well costs using statistical methods.
Design of Experiments (DoE) is implemented to design simulation runs for characterizing system behavior using the effect of certain critical parameters, such as well type, horizontal well length, well pattern geometry, and injection/production constraints on oil recovery. After carrying out simulation runs using a commercially available simulation software, well cost is estimated for each simulation case. Response Surface Methodology (RSM) is used for optimization of well pattern parameters. The parameters, their interactions and response are modeled into a mathematical equation to maximize oil recovery and minimize well cost.
Economics plays a key role in deciding the best well pattern for any field during the field development phase. Hence, while solving the optimization problem, well costs have been incorporated in the analysis. Thus, based on the results of the study performed on selected parameters, using interdependence of the above mentioned methodologies, optimum combinations of variables for maximizing oil recovery and minimizing well cost will be obtained. Additionally, reservoir level optimization assists in providing a much needed platform for solving the integrated production optimization problem involving parameters relevant at different levels, such as reservoir, wells and field. As a result, this optimum well pattern methodology will help ensure optimum oil recovery in the otherwise economically unattractive field, and can provide significant insights into developing the field more efficiently.
Computational algorithms are gaining popularity for solving optimization problems, as opposed to manual simulations. DoE is effective, simple to use and saves computational time, when compared to algorithms. Although DoE has been used widely in the oil industry, its application in domains like well pattern optimization is novel. This paper presents a case study for the application of DoE and RSM to well optimization in a real existing field, considering all possible scenarios and variables.