What is it about?
This study demonstrates that machine learning especially boosting-based ensemble models can predict underground rock permeability much more accurately than traditional methods, improving reservoir characterization and supporting safer, more efficient geological CO₂ storage.
Featured Image
Photo by Anton Lukin on Unsplash
Why is it important?
This study shows that machine learning can provide faster and more accurate estimates of underground rock permeability than traditional methods. Better permeability predictions help engineers select safer CO₂ storage sites, estimate storage capacity more reliably, and design more effective carbon capture and storage projects to reduce greenhouse gas emissions.
Perspectives
This study demonstrates that machine learning, combined with geological knowledge and explainable artificial intelligence, can substantially improve permeability prediction for geological CO₂ storage. By integrating petrophysical measurements, feature engineering, data augmentation, and ensemble learning, the proposed workflow offers a robust framework for reservoir characterization, injectivity assessment, and storage capacity evaluation in heterogeneous sedimentary formations.
Manawar Pervaiz
University of Science and Technology of China
Read the Original
This page is a summary of: Machine Learning-Based Permeability Prediction for CO₂ Storage: A Case Study from the Petrel Sub-basin, Australia, Geomechanics for Energy and the Environment, July 2026, Elsevier,
DOI: 10.1016/j.gete.2026.100867.
You can read the full text:
Contributors
The following have contributed to this page
