Data driven based deep learning for optimizing carbon storage and methane adsorption in unconventional shale gas reservoirs; Journal of Environmental Chemical Engineering; Vol. 13, iss. 3

Data driven based deep learning for optimizing carbon storage and methane adsorption in unconventional shale gas reservoirs; Journal of Environmental Chemical Engineering; Vol. 13, iss. 3

Bibliographic Details
Parent link:Journal of Environmental Chemical Engineering.— .— Amsterdam: Elsevier Science Publishing Company Inc.
Vol. 13, iss. 3.— 2025.— Article number 116901, 15 p.
Other Authors: Wang Yongjun, Thang Hung Vo, Al-Mudhafar W. J. Watheq, Dai Zhenxue, Hemeng Zhang, Davoodi Sh. Shadfar, Zhang Tao
Summary:Shale gas is a critical component of the global energy supply, necessitating accurate gas-in-place evaluations for optimized extraction. In unconventional shale reservoirs, gas adsorption is the primary storage mechanism. This study introduces a novel one-dimensional Convolutional Neural Network (1D-CNN) to predict gas adsorption. The 1D-CNN model was compared with traditional machine learning models, including Random Forest (RF), Extra Trees (ET), and Gradient Boosting (GB). Using a dataset of 352 samples with key features—pressure, temperature, TOC, and moisture content—the 1D-CNN achieved superior predictive accuracy, with a mean absolute error (MAE) of 0.015 and R² of 0.9949, outperforming other models. Sensitivity analysis identified moisture content as the most influential factor affecting adsorption. Validation using unseen methane adsorption data demonstrated that the PIDL model generalizes well and produces physically consistent predictions. Overall, the PIDL-enhanced 1D-CNN model advances methane adsorption prediction by integrating domain-specific physical laws into a deep learning framework, facilitating more efficient shale gas extraction and improved carbon sequestration practices. This hybrid approach not only reduces experimental costs but also provides consistent and reliable results, offering a robust solution for real-world applications
Текстовый файл
Language:English
Published: 2025
Subjects:
Gas adsorption
Shale gas reservoir
CO2 sequestration
Deep learning
Convolutional Neural Network (CNN)
CCUS
Carbon neutrality
Next generation reservoir simulator for CCS
труды учёных ТПУ
электронный ресурс
Online Access:https://doi.org/10.1016/j.jece.2025.116901
Format: Electronic Book Chapter
KOHA link:https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=680359

Similar Items

Similar Items