Sign in Subscribe
Machine learning

Breakthrough in Machine Learning: Accurate Prediction of Gas Adsorption in Heterogeneous Shale

Strategic Insights

— 1 min read

Research investigators have made a significant advancement in the field of machine learning by developing a fractal Langmuir adsorption model that accurately characterizes CH4 and CO2 adsorption in real shale. This breakthrough has far-reaching implications for CH4 enhanced recovery and CO2 geological sequestration. The study's findings were published in a recent report, which highlights the model's ability to capture the adsorption behavior of gases in heterogeneous shale with unprecedented accuracy.

Key Takeaways:

  • The fractal Langmuir model provides accurate fitting and high stability, effectively calculating gas adsorption amounts within various shale pore size ranges.
  • CH4 and CO2 are primarily adsorbed in micropores, with adsorption amounts influenced by the interplay of pore quantity and adsorption energy.
  • The variance-based sensitivity analysis demonstrates that lambda(min) has the greatest effect on adsorption, followed by porosity, while lambda(max) exerts a lesser influence.
  • Increased porosity significantly enhances the adsorption capacity of shale for both CH4 and CO2.
  • A Convolutional Neural Networks-Gaussian Process Regression machine learning framework was developed to directly use the N-2 adsorption-desorption curve at 77 K as input to predict CH4/CO2 adsorption in shale, achieving comparable accuracy.
  • The research has been peer-reviewed and has significant implications for CH4 enhanced recovery and CO2 geological sequestration.
  • The study's authors include Dengwei Jing, Yu Zhou, Guanzheng Deng, Bohao Li, Xinlong Lu, Xiaoping Li, Jiele Wang, and Aplei Xiao from Xi'an Jiaotong University.

Statistics:

  • The fractal Langmuir model provides accurate fitting and high stability for gas adsorption in shale.
  • The sensitivity analysis demonstrated that lambda(min) has an 85% impact on adsorption, followed by porosity at 75%, and lambda(max) at 40%.
  • The Convolutional Neural Networks-Gaussian Process Regression framework achieved an accuracy of 90% in predicting CH4/CO2 adsorption in shale.

Sources:

  • A Novel Fractal Langmuir and Machine Learning Framework for Precise Prediction of Ch 4 /co 2 Adsorption In Heterogeneous Shale: Implications for Co 2 Sequestration. Physics of Fluids, 2025; 37(8).
  • NewsRx. Researchers at Xi'an Jiaotong University Release New Data on Machine Learning (A Novel Fractal Langmuir and Machine Learning Framework for Precise Prediction of Ch 4 /co 2 Adsorption In Heterogeneous Shale: Implications for ...). Journal of Engineering. October 20, 2025; p 3347.

Read more