Mixed integer optimization in well scheduling

Dimitrios I Gerogiorgis, Vassileios D Kosmidis, Efstratios N Pistikopoulos

Research output: Chapter in Book/Report/Conference proceedingChapter (peer-reviewed)

Abstract / Description of output

This Chapter presents a novel, mixed integer nonlinear programming (MINLP) model for the well scheduling problem, where the nonlinear behavior of the reservoir, wells, pipelines and surface facilities has been incorporated into the mathematical formulation. The well scheduling problem is formulated as a snapshot optimization problem with an objective function that expresses the maximization of an economic index. Discrete decisions here include the operational status of the wells, the allocation of wells to manifold or separators and the allocation of surface flowlines to separators. Continuous decisions include the well oil flowrates, and the allocation of gas-to-gas lift wells.
A three-step solution strategy is proposed for the solution of this problem, where logic based relations and piecewise linear approximations of oil field wells are integrated in the MINLP formulation. The model is solved following an Outer Approximation (OA) class algorithm. A number of examples are presented to illustrate the performance and business value of the proposed strategy; a remarkable increase in oil production of up to 10% is demonstrated, compared to results obtained via widespread heuristic methods. A further increase of 2.9% can be achieved by dynamic optimization based on explicit consideration of the multiphase flow within the reservoirs of a particular oil field.
Original languageEnglish
Title of host publicationEncyclopedia of Optimization
PublisherSpringer US
Number of pages24
ISBN (Print)978-0-387-74758-3
Publication statusPublished - 2009


Dive into the research topics of 'Mixed integer optimization in well scheduling'. Together they form a unique fingerprint.

Cite this