A Survey on the Application of Machine Learning and Metaheuristic Algorithms for Intelligent Proxy Modeling in Reservoir Simulation

Ng, Cuthbert Shang Wui; Amar, Menad Nait; Ghahfarokhi, Ashkan Jahanbani; Imsland, Lars

doi:10.1016/j.compchemeng.2022.108107

Cited by 21 publications

(18 citation statements)

References 105 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These models can predict time-dependent variables at 100 to 1000 times faster speeds than traditional simulators. This acceleration in computation velocity via ML methods maintains an equivalent level of functionality (Ng et al, 2023).…”

Section: Benefits and Limitations Of MLmentioning

confidence: 98%

“…provides some examples of different ML algorithms. Among these various algorithms, supervised learning is most applied in the oil and gas industry (Ng et al, 2023). Furthermore, the enhancement of the ML process involves the optimization techniques to determine optimal values for control parameters, including the spreading coefficient, number of neurons, biases, and weights.…”

Section: Summary Of Machine Learning Approachesmentioning

confidence: 99%

“…Many literature reviews have been conducted in the past to summarize the application of ML in the oil and gas industry (Ng et al, 2023). However, no study on global research trends analyzed the dominant input parameters and evaluated the research work on CO2-EOR projects.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Comprehensive Summary of the Application of Machine Learning Techniques for CO<sub>2</sub>-EOR Projects

Du,

Salasakar,

Thakur

2024

Preprint

View full text Add to dashboard Cite

In recent years, machine learning (ML) techniques have emerged as an efficient and effective technology within the petroleum industry. This paper focuses on the current application of ML in enhanced oil recovery (EOR) through CO2 injection, which exhibits promising economic and environmental benefits of climate-change mitigation strategy. Our comprehensive review explores the diverse use cases of ML techniques in CO2-EOR, including aspects such as minimum miscible pressure (MMP) prediction, well location optimization, oil production and recovery factor prediction, multi-objective optimization, Pressure-Volume-Temperature (PVT) properties estimation, Water Alternating Gas (WAG) analysis, and CO2-foam EOR, from 101 reviewed papers. In this comprehensive review, we catalog relative information, including the input parameters, objectives, data sources, train/test/validate information, results, evaluation, and rating score for each area based on criteria such as data quality, ML building process, and analysis of results. We also briefly summarized the benefits and limitations of ML methods in petroleum industry applications. Our detailed and extensive study could serve as an invaluable reference for employing ML techniques in the petroleum industry. Based on the review, we found that ML techniques offer great potential in solving problems in the majority of CO2-EOR areas involving prediction and regression. With the generation of massive amounts of data in the everyday oil and gas industry, machine learning techniques can provide efficient and reliable preliminary results for the industry.

show abstract

Section: Benefits and Limitations Of MLmentioning

confidence: 98%

Section: Summary Of Machine Learning Approachesmentioning

confidence: 99%

See 1 more Smart Citation

A Comprehensive Summary of the Application of Machine Learning Techniques for CO<sub>2</sub>-EOR Projects

Du,

Salasakar,

Thakur

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Optimization algorithms are available to help investors find the Worst asset allocation in their portfolios to minimize risk and achieve expected returns. Optimization algorithms have been further applied with the development of deep learning (DL) and machine learning (ML) [17][18][19]. In the ML and DL, parameter optimization which improves the performance and accuracy of the model in machine learning models can be achieved with optimization algorithms.…”

Section: Introductionmentioning

confidence: 99%

A novel meta-heuristic algorithm for high-dimensional problems: Rhinopithecus Swarm Optimization

Zhou,

Wang,

Zhou

et al. 2024

Preprint

View full text Add to dashboard Cite

This paper introduces a novel meta-heuristic algorithm know as Rhinopithe-cus Swarm Optimization (RSO) to deal with the high-dimensional problems. The RSO is inspired by the different social behaviors of different subgroups in the rhinopithecus swarm. This algorithm categorizes the population into king, matures, adolescences and infancy based on fitness values before each iteration. At iterations they are given different search methods including vertical migration , concerted search and mimicry due to their social division of labor. The experiment was independently conducted 36 times on the highest dimension recommended by CEC2017. The results show that comprehensive performance of RSO is higher than 8 well-known meta-heuristic algorithms including DBO, BWO, SSA, AVOA, WOA, ARBBPSO, GTO, and HHO. RSO ranked first in the CEC2017 with a score of 1.655 on 29 benchmark functions using Ferideman test, which is 42.8% better than the second-ranked algorithm SSA. Overall, it 1 is concluded that RSO outperforms well-known algorithms and can better solve high-dimensional optimization problems

show abstract

“…The nature-inspired algorithm is the epitome of gradient-free algorithms. Its successful integration with the ML-based proxy models has been displayed in several pieces of literature in reservoir and production engineering [13][14][15]. In this study, two optimization algorithms are used: the Genetic Algorithm (GA) and the Particle Swarm Optimization (PSO).…”

Section: Introductionmentioning

confidence: 99%

Fast Well Control Optimization with Two-Stage Proxy Modeling

Ghahfarokhi

Wiranda

2023

Energies

Self Cite

View full text Add to dashboard Cite

Waterflooding is one of the methods used for increased hydrocarbon production. Waterflooding optimization can be computationally prohibitive if the reservoir model or the optimization problem is complex. Hence, proxy modeling can yield a faster solution than numerical reservoir simulation. This fast solution provides insights to better formulate field development plans. Due to technological advancements, machine learning increasingly contributes to the designing and building of proxy models. Thus, in this work, we have proposed the application of the two-stage proxy modeling, namely global and local components, to generate useful insights. We have established global proxy models and coupled them with optimization algorithms to produce a new database. In this paper, the machine learning technique used is a multilayer perceptron. The optimization algorithms comprise the Genetic Algorithm and the Particle Swarm Optimization. We then implemented the newly generated database to build local proxy models to yield solutions that are close to the “ground truth”. The results obtained demonstrate that conducting global and local proxy modeling can produce results with acceptable accuracy. For the optimized rate profiles, the R2 metric overall exceeds 0.96. The range of Absolute Percentage Error of the local proxy models generally reduces to 0–3% as compared to the global proxy models which has a 0–5% error range. We achieved a reduction in computational time by six times as compared with optimization by only using a numerical reservoir simulator.

show abstract

A Survey on the Application of Machine Learning and Metaheuristic Algorithms for Intelligent Proxy Modeling in Reservoir Simulation

Cited by 21 publications

References 105 publications

A Comprehensive Summary of the Application of Machine Learning Techniques for CO<sub>2</sub>-EOR Projects

A Comprehensive Summary of the Application of Machine Learning Techniques for CO<sub>2</sub>-EOR Projects

A novel meta-heuristic algorithm for high-dimensional problems: Rhinopithecus Swarm Optimization

Fast Well Control Optimization with Two-Stage Proxy Modeling

Contact Info

Product

Resources

About