View in EDS

Research on Prediction Methods of Deep Coalbed Methane Content Based on Geophysical Logging.

Saved in:
Bibliographic Details
Title: Research on Prediction Methods of Deep Coalbed Methane Content Based on Geophysical Logging.
Authors: Feng, Peng1,2,3 (AUTHOR), Li, Song2,3 (AUTHOR) lisong@cugb.edu.cn, Tang, Shuling2,3 (AUTHOR) tangshuling@cugb.edu.cn, Tang, Dazhen2,3 (AUTHOR), Zhong, Guanghao2,3 (AUTHOR), Yang, Qiang4 (AUTHOR), Zhou, Guoxiao5,6 (AUTHOR)
Source: Arabian Journal for Science & Engineering (Springer Science & Business Media B.V. ). Jun2026, Vol. 51 Issue 11, p14187-14202. 16p.
Subjects: Coalbed methane, Geophysical well logging, Prediction models, Back propagation, Support vector machines, Multiple regression analysis, Random forest algorithms
Geographic Terms: China
Abstract: Accurate assessment of deep coalbed methane (CBM) content is crucial for CBM extraction. This study conducted an analysis of the correlation between logging parameters and deep CBM content within the Daning area of the Ordos Basin, China. Four deep CBM content predictive models—Multiple Linear Regression (MLR), Support Vector Regression (SVR), Random Forests (RF), and BP Neural Network (BPNN)—were developed and compared. Additionally, the distribution of CBM content in the study area was further analyzed based on the SVR model. The results indicate that well caliper logging (CAL), compensated neutron logging (CNL), sonic-interval transit logging (DT), and comprised density logging (DEN) exhibit stronger correlations with deep CBM content compared to other logging parameters. The predictive accuracy of the MLR, BPNN, SVR, and RF models, constructed using the four logging parameters, surpasses that of single logging parameter relationships. The average relative errors for the training and testing data of SVR model are 10.23 and 9.12%, respectively, while those for the other models exceed 11%. Comparatively, the SVR model exhibited the highest predictive accuracy and the most robust generalization capability among the four models. The CBM content in the Daning area shows an overall increasing trend from the northeast to the southwest within the study area, aligning with the increasing depth of the coal seam. This is attributed to a significant increase in free gas within the deep coal, compensating for the decrease in adsorbed gas caused by temperature effects. [ABSTRACT FROM AUTHOR]
Copyright of Arabian Journal for Science & Engineering (Springer Science & Business Media B.V. ) is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:Accurate assessment of deep coalbed methane (CBM) content is crucial for CBM extraction. This study conducted an analysis of the correlation between logging parameters and deep CBM content within the Daning area of the Ordos Basin, China. Four deep CBM content predictive models—Multiple Linear Regression (MLR), Support Vector Regression (SVR), Random Forests (RF), and BP Neural Network (BPNN)—were developed and compared. Additionally, the distribution of CBM content in the study area was further analyzed based on the SVR model. The results indicate that well caliper logging (CAL), compensated neutron logging (CNL), sonic-interval transit logging (DT), and comprised density logging (DEN) exhibit stronger correlations with deep CBM content compared to other logging parameters. The predictive accuracy of the MLR, BPNN, SVR, and RF models, constructed using the four logging parameters, surpasses that of single logging parameter relationships. The average relative errors for the training and testing data of SVR model are 10.23 and 9.12%, respectively, while those for the other models exceed 11%. Comparatively, the SVR model exhibited the highest predictive accuracy and the most robust generalization capability among the four models. The CBM content in the Daning area shows an overall increasing trend from the northeast to the southwest within the study area, aligning with the increasing depth of the coal seam. This is attributed to a significant increase in free gas within the deep coal, compensating for the decrease in adsorbed gas caused by temperature effects. [ABSTRACT FROM AUTHOR]
ISSN:2193567X
DOI:10.1007/s13369-025-10428-y