基于机器学习的页岩微观参数识别及其断裂离散元模拟

Machine learning based identification of shale micro-parameters and discrete element simulation of its fracture behavior

摘要

在油气开采过程中对页岩参数进行准确识别是提升水力压裂与水平钻井定量分析精度的关键难题。针对页岩成分复杂、非均质性强的特点,基于CT扫描构建数字岩心,通过灰度处理提取结构体积占比并建立几何模型,引入基体与层理表征其非均质特征;结合深度神经网络(DNN)与遗传算法(GA)建立非均质页岩微观与宏观参数之间的跨尺度双向关联模型,基于离散元开展页岩单轴压缩模拟,并定量分析其微观参数对宏观力学行为的影响。结果表明,所建立的跨尺度双向关联模型能够有效预测页岩的宏观力学参数并反演识别其微观参数,基于预测所得微观参数的页岩单轴压缩模拟可实现对页岩断裂行为的预测。

Abstract

During the process of oil and gas production, the accurate identification of shale parameters has become a key challenge in improving the quantitative analysis accuracy of hydraulic fracturing and horizontal drilling. In response to the characteristics of complex composition and strong heterogeneity in shales, a digital core based on CT scanning was constructed in this study. Through grayscale processing, the volumetric proportions of structural components were extracted to establish a geometric model, and the matrix and bedding planes were introduced to characterize its heterogeneous features. By integrating deep neural networks (DNN) and genetic algorithms (GA), a cross-scale bidirectional correlation model between microscopic and macroscopic parameters of heterogeneous shale was established. Furthermore, uniaxial compression simulations of shale were conducted using the discrete element method (DEM), and the influences of micro-parameters on macroscopic mechanical behavior were quantitatively analyzed. The results demonstrate that this model enables the prediction of macroscopic mechanical parameters and the identification of microscopic parameters in shales. The uniaxial compression simulations can successfully predict the shale fracture behavior.

关键词

非均质页岩 / 深度神经网络 / 遗传算法 / 离散元 / 裂缝预测

Key words

heterogeneous shale / deep neural network / genetic algorithm / discrete element method / fracture prediction

参考文献

[1] 刘豪, 刘怀亮, 刘宇, 等 . 页岩气多级压裂断层动态滑移规律研究[J]. 石油机械, 2024, 52(2): 65-74.

[2] LIU Hao, LIU Huailiang, LIU Yu, et al. Research on dynamic slip of fault resulted from multistage fracturing of shale gas reservoir[J]. China Petroleum Machinery, 2024, 52(2): 65-74.

[3] CHO J W, KIM H, JEON S, et al. Deformation and strength anisotropy of Asan gneiss, Boryeong shale, and Yeoncheon schist[J]. International Journal of Rock Mechanics and Mining Sciences, 2012, 50: 158-169.

[4] KIM H, CHO J W, SONG I, et al. Anisotropy of elastic moduli, P-wave velocities, and thermal conductivities of Asan Gneiss, Boryeong Shale, and Yeoncheon Schist in Korea[J]. Engineering Geology, 2012, 147/148: 68-77.

[5] 郭旭升, 魏志红, 魏祥峰, 等 . 四川盆地侏罗系陆相页岩油气富集条件及勘探方向[J]. 石油学报, 2023, 44(1): 14-27.

[6] GUO Xusheng, WEI Zhihong, WEI Xiangfeng, et al. Enrichment conditions and exploration direction of Jurassic continental shale oil and gas in Sichuan Basin[J]. Acta Petrolei Sinica, 2023, 44(1): 14-27.

[7] 姚军, 王萌, 樊冬艳, 等 . 考虑层理缝岩性差异的页岩油藏压裂水平井动态分析方法[J]. 中国石油大学学报(自然科学版), 2024, 48(5): 91-102.

[8] YAO Jun, WANG Meng, FAN Dongyan, et al. Dynamic analysis of fractured horizontal wells in shale reservoirs considering lithological differences of bedding fractures[J]. Journal of China University of Petroleum (Edition of Natural Science), 2024, 48(5): 91-102.

[9] CUNDALL P A . A computer model for simulating progressive, large-scale movement in blocky rock system[C]. Nancy: Proceedings of the International Symposium on Rock Mechanics. 1971: 129-136.

[10] CUNDALL P A, STRACK O D L . A discrete numerical model for granular assemblies[J]. Géotechnique, 1979, 29(1): 47-65.

[11] ZHANG X, SI G, BAI Q, et al. Numerical simulation of hydraulic fracturing and associated seismicity in lab-scale coal samples: a new insight into the stress and aperture evolution[J]. Computers and Geotechnics, 2023, 160: 105507.

[12] BOCK S, PRUSEK S . Numerical study of pressure on dams in a backfilled mining shaft based on PFC3D code[J]. Computers and Geotechnics, 2015, 66: 230-244.

[13] OETOMO J J, VINCENS E, DEDECKER F, et al. Modeling the 2D behavior of dry-stone retaining walls by a fully discrete element method[J]. International Journal for Numerical and Analytical Methods in Geomechanics, 2016, 40(7): 1099-1120.

[14] WANG Z J, JACOBS F L, ZIEGLER M . Experimental and DEM investigation of geogrid-soil interaction under pullout loads[J]. Geotextiles and Geomembranes, 2016, 44(3): 230-246.

[15] 申海萌, 李琦, 李霞颖, 等 . 川南龙马溪组页岩不同应力条件下脆性破坏特征室内实验与数值模拟研究[J]. 岩土力学, 2018, 39(增2): 254-262.

[16] SHEN Haimeng, LI Qi, LI Xiaying, et al. Study on the brittle failure characteristics of the Longmaxi formation shale in Southern Sichuan under different stress conditions: laboratory experiments and numerical simulations[J]. Rock and Soil Mechanics, 2018, 39(sup2): 254-262.

[17] TANG W, ZHAI C, YU X, et al. Dynamic Brazilian splitting experiment of bedding shale based on continuum-discrete coupled method[J]. International Journal of Impact Engineering, 2022, 168: 104289.

[18] LIU M D, INDRARATNA B N . General strength criterion for geomaterials including anisotropic effect[J]. International Journal of Geomechanics, 2011, 11(3): 251-262.

[19] GU Z W, KONG X Q, LIU J L, et al. A coupled deep learning and DEM modelling approach for transverse crack prediction in UD-GFRP composites[J]. Composites Science and Technology, 2025, 269: 111234.

[20] TIE J, MENG W, WEI M, et al. Loading rate and mineralogical controls on tensile strength of rocks: a machine learning view[J]. Rock Mechanics and Rock Engineering, 2023, 56(8): 6119-6125.

[21] DENG S, PAN H Y, WANG H G, et al. A hybrid machine learning optimization algorithm for multivariable pore pressure prediction[J]. Petroleum Science, 2024, 21(1): 535-550.

[22] GU Z, DING X, HOU X, et al. A genetic evolved machine learning approach for 3D DEM modelling of anisotropic materials with full consideration of particulate interactions[J]. Composites Part B: Engineering, 2023, 250: 110432.

[23] 毕开原. 我国页岩气钻完井关键技术现状及发展趋势[J]. 石油机械, 2024, 52(11): 12-20, 28.

[24] BI Kaiyuan. Current status and trend of key technologies for shale gas well drilling and completion in China[J]. China Petroleum Machinery, 2024, 52(11): 12-20, 28.

[25] HE H Y, ZHENG L L, ZUO Y J, et al. Study on the relationship between macroscopic failure and micro-crack evolution in interbedded sandstone[J]. Geomechanics and Geophysics for Geo-Energy and Geo-Resources, 2022, 8(2): 79.

[26] POTYONDY D O, CUNDALL P A . A bonded-particle model for rock[J]. International Journal of Rock Mechanics and Mining Sciences, 2004, 41(8): 1329-1364.

[27] Numerical Modeling in Micromechanics via Particle Methods-2004[C]// Proceedings of the 2nd International PFC Symposium. Kyoto: CRC Press, 2004.

[28] SHI Q W, MISHRA B . Discrete element modeling of delamination in laboratory scale laminated rock[J]. Mining, Metallurgy & Exploration, 2021, 38(1): 433-446.

[29] XU Z H, WANG Z Y, WANG W, et al. An integrated parameter calibration method and sensitivity analysis of microparameters on mechanical behavior of transversely isotropic rocks[J]. Computers and Geotechnics, 2022, 142: 104573.

[30] SUN B W, YANG S Q, XU J, et al. Discrete element simulation on failure mechanical behavior of transversely isotropic shale under two kinds of unloading paths[J]. Theoretical and Applied Fracture Mechanics, 2022, 121: 103466.

[31] NAWI N M, ATOMI W H, REHMAN M Z . The effect of data pre-processing on optimized training of artificial neural networks[J]. Procedia Technology, 2013, 11: 32-39.

[32] YIN P F, YANG S Q, GAO F, et al. Experiment and DEM simulation study on mechanical behaviors of shale under triaxial cyclic loading and unloading conditions[J]. Geomechanics and Geophysics for Geo-Energy and Geo-Resources, 2023, 9(1): 10.

[33] ALI FAIZI S, KWOK C Y, DUAN K . The effects of intermediate principle stress on the mechanical behavior of transversely isotropic rocks: insights from DEM simulations[J]. International Journal for Numerical and Analytical Methods in Geomechanics, 2020, 44(9): 1262-1280.

基金资助

国家自然科学基金项目(12572119)

山东省自然科学基金面上项目(ZR2024MA086)

山东省优秀青年科学家基金项目(2024HWYQ-049)