Hybrid of PSO and CMA-ES algorithms for joint optimization of well placement and control

Summary Oilfield development related decisions such as well placement and production control settings are crucial for commercial success of any project. Planning is done to maximize return on investment or fluid recovery, and involves reservoir simulation studies. Large scale field planning involves many variables including well location and design, as well as bottom hole controls. Reservoir simulation workflows employ optimization algorithms to search for well settings which maximizes our objective. This work presents efficacy of a hybrid evolutionary optimization algorithm for solving high dimensional well placement and control optimization problem, and demonstrates its application using Olympus benchmark. Particle swarm optimization (PSO) is first used in standalone mode to solve joint optimization problem. Two different objective functions, weighted sum of cumulative fluid (WCF) and net present value (NPV) of discounted cash-flow, are used for rigorous analysis and comparison. Next, PSO algorithm is used with another popular optimization algorithm, covariance matrix adaptation – evolution strategy (CMA-ES), in hybrid mode. Hybrid optimization run is made by transferring the best result from PSO algorithm to CMA-ES as its starting point for further improvement. Hybrid PSO-CMA-ES algorithm is more effective in handling high-dimensional and multi-modal optimization problem.

[1]  Lei Liu,et al.  Particle swarm optimization algorithm: an overview , 2017, Soft Computing.

[2]  Antonio Bolufé Röhler,et al.  Measuring the curse of dimensionality and its effects on particle swarm optimization and differential evolution , 2014, Applied Intelligence.

[3]  David H. Wolpert,et al.  No free lunch theorems for optimization , 1997, IEEE Trans. Evol. Comput..

[4]  Yudong Zhang,et al.  A Comprehensive Survey on Particle Swarm Optimization Algorithm and Its Applications , 2015 .

[5]  E. Della Rossa,et al.  Introduction to the special issue: Overview of OLYMPUS Optimization Benchmark Challenge , 2020, Computational Geosciences.

[6]  Xian-Huan Wen,et al.  Large Scale Field Development Optimization Using High Performance Parallel Simulation and Cloud Computing Technology , 2018, Day 3 Wed, September 26, 2018.

[7]  Nikolaus Hansen,et al.  The CMA Evolution Strategy: A Tutorial , 2016, ArXiv.

[8]  John D. Hedengren,et al.  Review of Field Development Optimization of Waterflooding, EOR, and Well Placement Focusing on History Matching and Optimization Algorithms , 2017 .

[9]  Mathias C. Bellout,et al.  FieldOpt: A powerful and effective programming framework tailored for field development optimization , 2020, Comput. Geosci..

[10]  Christian L. Müller,et al.  Particle Swarm CMA Evolution Strategy for the optimization of multi-funnel landscapes , 2009, 2009 IEEE Congress on Evolutionary Computation.

[11]  Xin Lin,et al.  Hybrid of PSO and CMA-ES for Global Optimization , 2019, 2019 IEEE Congress on Evolutionary Computation (CEC).

[12]  Maurice Clerc,et al.  The particle swarm - explosion, stability, and convergence in a multidimensional complex space , 2002, IEEE Trans. Evol. Comput..