Abstract
Offshore field development activities are commonly constrained by the capacity of production facilities available at the platform. It is a challenge for practitioners to find the best development and operational strategies to maximize field production in the presence of such constraints. Additional difficulties are raised when the often large geological uncertainties inherent to field development activities need to be accounted for throughout the search for the best strategy. In this work we present how computer-assisted optimization can help practitioners tackle these challenges.
In this work, we focus on the particular problem of selecting the optimal subset(s) of wells to be subject to certain operational actions taking place throughout the field production life-cycle. We leverage robust and efficient field development optimization framework (EVEReST) based on stochastic gradients and its flexibility for customization to new types of decisions. We extend the mathematical parametrization based on a single set of well priorities, so far used to optimize time-static drilling order and well selection optimization. We apply it to multiple sets of well priorities which facilitate the formulation of the large combinatorial time-dynamic well subset selection optimization problem in terms of continuous control variables more suitable for gradient-based optimization techniques.
The extended method is applied to two different real-life field case studies based on an offshore heavy oil field. In the first case, we use the well priorities to determine how to optimally select and swap the 14 producers (out of a total of 18) to be put on stream to satisfy constraints on the available variable speed drivers (VSDs) to provide power to the ESP pumps required to operate the producers. In the second case study, we tackle the problem of converting producers into polymer injectors for EOR purposes. Firstly, simulation scenarios indicate that such conversion has a positive impact on the overall project business case. However, deciding which 2 wells (out of a total of 5) to convert into polymer injectors, combined with when and how much polymer to inject while respecting the limits of available capacity of polymer injection facilities, leads to a complex optimization problem.
In both case studies, optimization found improved and non-trivial strategies, resulting in significant increases in the economic objective function (+107 million USD and +318 million USD in terms of NPV) and providing new insights to asset engineers into the operations of the field. This showcases the strengths of the EVEReST framework to tackle real-life optimization problems accounting for complex constraints and uncertainties which are critical to deliver solutions of practical value to the asset teams.
