Session: 11-13-01 Digitalization of Subsurface, Well Systems, Subsea Systems and Operations

Paper Number: 78647

78647 - Designing Subsea Processing Systems Using a Hybrid Genetic Algorithm 

Designing a subsea production system is a task that takes several months to complete by an experienced and multidisciplinary team of engineers. Many factors play a key role in the decisions, such as the production rates, oil economics, current technology available, reservoir characteristics, and more. Optimization techniques can be used to tackle this problem, to thoroughly explore the search space and determine designs that provide the largest profit, provide insights and ensure the efficient use of resources. In this work, we propose using a genetic algorithm combined with a gradient search method to find optimal designs and support subsea production system design.

The genetic algorithm is formulated to optimize the structure of the production system, while the gradient method solves the continuous non-linear variables related to flow rates, reservoir deliverability, equipment capacities, and others. The study case is based on Goliat field. This hybrid approach is compared with solutions obtained with exact methods.

The method successfully finds subsea configurations that maximize the net present value (NPV) in short running times.  The difference in NPV between the subsea system designs found is modest.  The method proposed finds a similar optimum solution to the one obtained with the exact method, although with differences in the running time. Many of the observations can be generalized to other cases beyond the specific study case. The results bring interesting insights for understanding subsea production system design. The methodology presented provides an advancement toward modelling and automated decision-making in subsea production system design.

Presenting Author: Leonardo Sales NTNU

Authors:

Leonardo Sales NTNU
Johannes Jäschke NTNU
Milan Stanko NTNU