Evolutionary Algorithm to Support Field Architecture Scenario Screening Automation and Optimization Using Decentralized Subsea Processing Modules

Abstract

Manual generation of test cases and scenario screening processes, during field architecture concept development, may produce a limited number of solutions that do not necessarily lead to an optimal concept selection. For more complex subsea field architectures, which might include processing modules for enhancing pressure and thermal management for the production network, the number of configuration cases and scenarios to evaluate can be extremely large and time and resource-consuming to handle through conventional manual design processes. This paper explores the use of evolutionary algorithms (EA) to automate case generation, scenario screening, and optimization of decentralized subsea processing modules during field development. An evaluation of various genetic operators and evolution strategies was performed to compare their performance and suitability to the application. Based on the evaluation results, an EA using structural uniform crossover and a gradient plus boundary mutation as the main variation operators was developed. The methodology combines EA and an integrated modeling approach to automate and optimize the concept selection and field architecture design when considering decentralized subsea processing modules.