Machine Learning-Driven Prediction of CO<sub>2</sub> Solubility in Brine: A Hybrid Grey Wolf Optimizer (GWO)-Assisted Gaussian Process Regression (GPR) Approach

Machine Learning-Driven Prediction of CO<sub>2</sub> Solubility in Brine: A Hybrid Grey Wolf Optimizer (GWO)-Assisted Gaussian Process Regression (GPR) Approach

The solubility of CO<sub>2</sub> in brine systems is critical for both carbon storage and enhanced oil recovery (EOR) applications. In this study, Gaussian Process Regression (GPR) with eight different kernels was optimized using the Grey Wolf Optimizer (GWO) algorithm to model this impo...

Full description

Saved in:
Bibliographic Details
Main Authors: Seyed Hossein Hashemi, Farshid Torabi, Paitoon Tontiwachwuthikul
Format: Article
Language:English
Published: MDPI AG 2025-08-01
Series:Energies
Subjects:
CO<sub>2</sub> solubility
Gaussian process regression
Grey Wolf Optimizer
ARD kernels
carbon storage
enhanced oil recovery
Online Access:https://www.mdpi.com/1996-1073/18/15/4205
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The solubility of CO<sub>2</sub> in brine systems is critical for both carbon storage and enhanced oil recovery (EOR) applications. In this study, Gaussian Process Regression (GPR) with eight different kernels was optimized using the Grey Wolf Optimizer (GWO) algorithm to model this important phase behavior. Among the tested kernels, the ARD Matern 3/2 and ARD Matern 5/2 kernels achieved the highest predictive accuracies, with R<sup>2</sup> values of 0.9961 and 0.9960, respectively, on the test data. This demonstrates superior performance in capturing CO<sub>2</sub> solubility trends. The GWO algorithm effectively tuned the hyperparameters for all kernel configurations, while the ARD capability successfully quantified the influence of key physicochemical parameters on CO<sub>2</sub> solubility. The outstanding performance of the ARD Matern 3/2 and ARD Matern 5/2 kernels suggests their particular suitability for modeling complex thermodynamic behaviors in brine systems. Furthermore, this study integrates fundamental thermodynamic principles into the modeling framework, ensuring all predictions adhere to physical laws while maintaining excellent accuracy (test R<sup>2</sup> > 0.98). These results highlight how machine learning can improve CO<sub>2</sub> injection processes, both for underground carbon storage and enhanced oil production.
ISSN:1996-1073

Similar Items