CNN-based assessment of the degrees of fracture development in bedrocks and classification of bedrock weathering zones

Authors

• LEI Jian, School of Earth and Environment, Anhui University of Science and Technology, Huainan 232001, ChinaFollow
• LU Haifeng, School of Earth and Environment, Anhui University of Science and Technology, Huainan 232001, China
• FAN Yufei, School of Earth and Environment, Anhui University of Science and Technology, Huainan 232001, China

Abstract

Background Bedrock weathering zones exhibit distinct vertical gradational characteristics. Accurately identifying the degrees of fracture development in the zones is critical to the safe production of mines in shallowly buried coal seams covered by thick water-bearing unconsolidated layers. Methods Focusing on the Qinan coal mine in Anhui Province, this study proposed a hybrid convolutional neural network (CNN) model, which enabled feature extraction using a synergistic architecture consisting of conventional convolutional layers and inception modules. Specifically, based on conventional log data and calculated shale content and fracture development index, combined with core observations, this study constructed a dataset containing labels that indicated four degrees of fracture development: fractured, high, medium, and low degrees. The synergistic architecture was then designed, in which conventional convolutional layers and inception modules were employed to extract local subtle features and conduct multi-scale feature fusion, respectively. Subsequently, hyperparameters, including the sequence length and learning rate, were optimized through grid search. Finally, the CNN model was trained to determine the degrees of fracture development intelligently. Results and Conclusions Compared to the random forest (RF), XGBoost, and support vector machine (SVM) models, as well as other CNN models, the hybrid CNN model demonstrated superior performance, with an accuracy reaching 92.71% and a weighted average F1 score of up to 93% on the test set. Furthermore, the proposed model can accurately capture the log response characteristics of fracture development in rocks, significantly enhancing the efficiency and continuity of log-based fracture assessment. The hybrid CNN model was applied to 25 boreholes in a certain mining area of the study area. Accordingly, a 3D geological model of weathered zones in the mining area was created. With an average root mean square error (RMSE) of 8.9 m in the prediction of the top boundary depths of various weathering zones, the geological model clearly exhibits the spatial distribution patterns of completely, strongly, and moderately weathered zones, together with slightly weathered to unweathered zones, thereby providing a quantitative geological basis for grouting engineering, along with the prevention and control of roof water hazards, in the coal mine.

Keywords

bedrock weathering zone, convolutional neural network (CNN), 3D geological modeling, log interpretation, fracture development index, intelligent classification

DOI

10.12363/issn.1001-1986.25.12.0908

Recommended Citation

LEI Jian, LU Haifeng, FAN Yufei, et al. (2026) "CNN-based assessment of the degrees of fracture development in bedrocks and classification of bedrock weathering zones," Coal Geology & Exploration: Vol. 54: Iss. 7, Article 16.
DOI: 10.12363/issn.1001-1986.25.12.0908
Available at: https://cge.researchcommons.org/journal/vol54/iss7/16

Reference

[1] STACEY T R,WESSELOO J. Design and prediction in rock engineering:The importance of mechanisms of failure,with focus on high stress,brittle rock conditions[J]. Rock Mechanics and Rock Engineering,2022,55(3):1517−1535.

[2] 许延春,马子民,李小二,等. 基岩风化带性质对顶板突水溃砂的影响研究:以赵固一矿为例[J]. 煤炭科学技术,2023,51(7):64−71. XU Yanchun,MA Zimin,LI Xiaoer,et al. Research on the influence of the nature of the weathered bedrock zone on the roof water bursting and sand bursting:Taking Zhaogu No. 1 Mine as an example[J]. Coal Science and Technology,2023,51(7):64−71.

[3] LEBEDEVA M I,BRANTLEY S L. Weathering and erosion of fractured bedrock systems[J]. Earth Surface Processes and Landforms,2017,42(13):2090−2108.

[4] WANG Qi,GAO Hongke,YU Hengchang,et al. Method for measuring rock mass characteristics and evaluating the grouting-reinforced effect based on digital drilling[J]. Rock Mechanics and Rock Engineering,2019,52(3):841−851.

[5] ZHANG Wenquan,LEI Yu,SHAO Jianli,et al. Simulation of the activation of mining faults and grouting reinforcement under thick loose layer and thin bedrock[J]. Scientific Reports,2022,12:17049.

[6] 邢茂林,郑士田,石志远,等. 注浆改造厚含水砂层提高开采上限技术及应用[J]. 煤田地质与勘探,2023,51(5):113−122. XING Maolin,ZHENG Shitian,SHI Zhiyuan,et al. Technology of raising upper limit of mining by grouting reconstruction in thick water-bearing sand layer and its application[J]. Coal Geology & Exploration,2023,51(5):113−122.

[7] 隋旺华. 近松散层采掘抗渗透破坏评价方法Ⅰ:临界水力坡度[J]. 煤田地质与勘探,2023,51(2):175−186. SUI Wanghua. Evaluation method of resistance to seepage failure due to mining near unconsolidated aquifers Ⅰ:Critical hydraulic gradient[J]. Coal Geology & Exploration,2023,51(2):175−186.

[8] 朱茂林,刘震,刘惠民,等. 东营凹陷北带基岩风化壳储层发育特征及控制因素[J]. 地学前缘,2024,31(3):324−336. ZHU Maolin,LIU Zhen,LIU Huimin,et al. Development characteristics and controlling factors of bedrock weathering crust reservoirs in the northern belt of the Dongying sag[J]. Earth Science Frontiers,2024,31(3):324−336.

[9] 陈心路,王粤川,彭靖淞,等. 基于测井–录井资料评价变质岩风化壳结构:以渤海南部渤中X构造为例[J]. 新疆石油地质,2019,40(2):181−187. CHEN Xinlu,WANG Yuechuan,PENG Jingsong,et al. Comprehensive evaluation of weathering crust structure of metamorphic rocks based on log data:A case from Bozhong X structure in the southern Bohai Sea[J]. Xinjiang Petroleum Geology,2019,40(2):181−187.

[10] 廖海博,刘红岐,崔云江,等. 基于常规测井的变质岩裂缝识别方法研究[J]. 海洋地质与第四纪地质,2023,43(4):189−198. LIAO Haibo,LIU Hongqi,CUI Yunjiang,et al. Research on fractures identification method of metamorphic rock based on conventional logging[J]. Marine Geology & Quaternary Geology,2023,43(4):189−198.

[11] 申威,唐军,吴进波,等. 潜山型储层纵向分带测井表征方法及应用[J]. 中国石油勘探,2023,28(2):144−153. SHEN Wei,TANG Jun,WU Jinbo,et al. Logging characterization method and application of vertical zonation of buried hill reservoir[J]. China Petroleum Exploration,2023,28(2):144−153.

[12] DONG Shaoqun,WANG Leting,ZENG Lianbo,et al. Fracture identification in reservoirs using well log data by window sliding recurrent neural network[J]. Geoenergy Science and Engineering,2023,230:212165.

[13] SINGHA D K,SARKAR A. A comprehensive study on fracture identification and fractures properties estimation from well log data in the hydrocarbon bearing formation using semi-automated machine learning and dual porosity model[J]. Pure and Applied Geophysics,2025,182(9):3613−3637.

[14] 郭建宏,杜婷,张占松,等. 基于支持向量机与地球物理测井资料的煤体结构识别方法[J]. 物探与化探,2021,45(3):768−777. GUO Jianhong,DU Ting,ZHANG Zhansong,et al. The coal structure identification method based on support vector machine and geophysical logging data[J]. Geophysical and Geochemical Exploration,2021,45(3):768−777.

[15] 肖航,张占松,郭建宏,等. 基于随机森林结合地球物理测井资料的煤体结构识别方法及应用[J]. 科学技术与工程,2021,21(24):10174−10180. XIAO Hang,ZHANG Zhansong,GUO Jianhong,et al. Coal structure identification method based on random forest combined with geophysical logging data and its application[J]. Science Technology and Engineering,2021,21(24):10174−10180.

[16] NADEGE M N,SHU Biao,KOUASSI A K F,et al. Fracture toughness prediction using well logs and extreme gradient boosting based on particle swarm optimization in shale gas reservoir[J]. Engineering Fracture Mechanics,2025,315:110759.

[17] RAJABI M,BEHESHTIAN S,DAVOODI S,et al. Novel hybrid machine learning optimizer algorithms to prediction of fracture density by petrophysical data[J]. Journal of Petroleum Exploration and Production Technology,2021,11(12):4375−4397.

[18] 丁燕,杜启振,YASIN Q,等. 基于深度学习的裂缝预测在S区潜山碳酸盐岩储层中的应用[J]. 石油物探,2020,59(2):267−275. DING Yan,DU Qizhen,YASIN Q,et al. Fracture prediction based on deep learning:Application to a buried hill carbonate reservoir in the S area[J]. Geophysical Prospecting for Petroleum,2020,59(2):267−275.

[19] 邵蓉波,史燕青,周军,等. 地球物理测井反问题机器学习数据集的构建方法研究[J]. 地球物理学报,2023,66(7):3086−3101. SHAO Rongbo,SHI Yanqing,ZHOU Jun,et al. Construction of machine learning data set for geophysical logging inversion[J]. Chinese Journal of Geophysics,2023,66(7):3086−3101.

[20] LU Guoqing,ZENG Lianbo,LIU Guoping,et al. Subsurface natural fracture identification using an integrated ensemble learning method[J]. IEEE Transactions on Geoscience and Remote Sensing,2025,63:4700716.

[21] TABASI S,TEHRANI P S,RAJABI M,et al. Optimized machine learning models for natural fractures prediction using conventional well logs[J]. Fuel,2022,326:124952.

[22] WANG Zihao,CAI Yidong,LIU Dameng,et al. Characterization of natural fracture development in coal reservoirs using logging machine learning inversion,well test data and simulated geostress analyses[J]. Engineering Geology,2024,341:107696.

[23] 董少群,曾联波,冀春秋,等. 超深层致密砂岩裂缝测井识别深度核方法[J]. 地学前缘,2024,31(5):166−176. DONG Shaoqun,ZENG Lianbo,JI Chunqiu,et al. A deep kernel method for fracture identification in ultra-deep tight sandstones using well logs[J]. Earth Science Frontiers,2024,31(5):166−176.

[24] DONG Shaoqun,ZENG Lianbo,WANG Leting,et al. Fracture identification in shale reservoir using a deep learning method:Chang 7 reservoirs,Triassic Yanchang Formation[J]. Geoenergy Science and Engineering,2024,238:212853.

[25] LI Chao,ZOU Changchun,PENG C,et al. Intelligent identification and segmentation of fractures in images of ultrasonic image logging based on transfer learning[J]. Fuel,2024,369:131694.

[26] 边会媛,姬嘉骏,段朝伟,等. 基于贝叶斯优化的CNN–BiLSTM–Attention的煤体结构识别方法[J]. 煤田地质与勘探,2025,53(11):34−43. BIAN Huiyuan,JI Jiajun,DUAN Chaowei,et al. Coal body structure identification method based on Bayesian-optimized CNN-BiLSTM-Attention[J]. Coal Geology & Exploration,2025,53(11):34−43.

[27] TIAN Miao,LI Bingtao,XU Huaimin,et al. Deep learning assisted well log inversion for fracture identification[J]. Geophysical Prospecting,2021,69(2):419−433.

[28] DONG Shaoqun,HAO Jingru,ZENG Lianbo,et al. A deep learning object detection method for fracture identification using conventional well logs[J]. IEEE Transactions on Geoscience and Remote Sensing,2024,62:5920716.

[29] 徐碧兰,何建华,吴炎峰,等. 多源数据与地质逻辑融合的深层页岩储层构造裂缝测井智能识别[J]. 成都理工大学学报(自然科学版),2025,52(5):931−950. XU Bilan,HE Jianhua,WU Yanfeng,et al. Intelligent logging identification of structural fractures in deep shale reservoirs based on the fusion of multisource data and geological logic[J]. Journal of Chengdu University of Technology (Science & Technology Edition),2025,52(5):931−950.

[30] 强志成,叶传红,刘祥强. 祁南煤矿某采区新生界含水层与隔水层分布特征及其水力联系分析[J]. 科技创新与应用,2025,15(19):97−100.

[31] 尹哲,郭晋. 祁南煤矿基岩及其风氧化带工程地质特征[J]. 中外企业家,2012(8):86−87.

[32] 曹伟林. 祁南煤矿86采区四含砂砾岩分布区群孔抽水试验与疏降可行性研究[D]. 徐州:中国矿业大学,2022. CAO Weilin. Study on group drilling pumping test and feasibility of depressurizing in sandy conglomerate distribution area of the fourth aquifer in 86 mining area of Qinan coal mine[D]. Xuzhou:China University of Mining and Technology,2022.

[33] ZHAO Fengda,ZHAO Zhuoyi,LYU Hongjin,et al. Lithology identification of imbalanced well log data based on diffusion model and multiscale CNN[J]. Mathematical Geosciences,2025,57(7):1283−1304.

[34] HA T N,LUBO–ROBLES D,MARFURT K J,et al. An in-depth analysis of logarithmic data transformation and per-class normalization in machine learning;application to unsupervised classification of a turbidite system in the Canterbury Basin,New Zealand,and supervised classification of salt in the Eugene Island minibasin,Gulf of Mexico[J]. Interpretation,2021,9(3):T685−T710.

[35] KIM M,KIM H S,CHUNG C K. A three-dimensional geotechnical spatial modeling method for borehole dataset using optimization of geostatistical approaches[J]. KSCE Journal of Civil Engineering,2020,24(3):778−793.