Application and challenges of deep learning technology in seismic data-based reservoir prediction
LUO Di1,2, WANG Hongbin1,2, CAI Feng1,2, WU Zhiqiang1,2, SUN Yunbao1,2, LI Qing1,2
1. Key Laboratory of Gas Hydrate, Ministry of Natural Resources, Qingdao Institute of Marine Geology, China Geological Survey, Qingdao, Shandong 266237, China; 2. Laboratory for Marine Mineral Resources, Laoshan Laboratory, Qingdao, Shandong 266237, China
Abstract:Traditional seismic data-based reservoir prediction technology fails to meet the demands of refined reservoir evaluation. Deep learning has strong feature extraction and high-dimensional data processing capabilities and has been extensively applied in seismic data-based reservoir prediction with promising results in recent years. This paper delved into the application and progress of deep learning technology in seismic data-based reservoir prediction, analyzed the challenges encountered during practical implementation, and proposed future research directions. The conclusions are as follows: ①In terms of qualitative hydrocarbon detection, deep learning technology facilitates the comprehensive utilization of multi-attribute seismic data to improve the efficiency and accuracy of prediction results. In terms of quantitative prediction, it enables a more precise approximation of the intricate nonlinear relationship between seismic data and targets, thereby achieving a refined quantitative evaluation of reservoirs. ②The application of deep learning technology faces several challenges. The issues such as insufficient label data and unbalanced samples lead to overfitting and poor generalization ability of the model; the complex model results in high computational costs; the “black box” feature of the model makes the prediction results lack physical interpretability; there is no evaluation criteria for qualitative prediction model and high-precision quantization algorithm for uncertainty. ③Future research should prioritize addressing challenges related to insufficient data availability and limitations of deep learning, such as constructing geophysical knowledge maps, effectively integrating and sharing multi-source data and knowledge, and combining deep learning with other machine learning algorithms such as feedback reinforcement learning, so as to provide more reliable technical support for hydrocarbon exploration and development.
骆迪, 王宏斌, 蔡峰, 吴志强, 孙运宝, 李清. 深度学习技术在地震储层预测中的应用及挑战[J]. 石油地球物理勘探, 2024, 59(3): 640-651.
LUO Di, WANG Hongbin, CAI Feng, WU Zhiqiang, SUN Yunbao, LI Qing. Application and challenges of deep learning technology in seismic data-based reservoir prediction. Oil Geophysical Prospecting, 2024, 59(3): 640-651.
SANG K H, YIN X Y, ZHANG F C. Machine learning seismic reservoir prediction method based on virtual sample generation[J]. Petroleum Science, 2021, 18(6):1662-1674.
[2]
LIU X, SHAO G, YUAN C, et al. Mixture of relevance vector regression experts for reservoir properties prediction[J]. Journal of Petroleum Science and Engineering, 2022, 214:110498.
[3]
SONG L, YIN X, ZONG Z, et al. Semi-supervised learning seismic inversion based on Spatio-temporal sequence residual modeling neural network[J]. Journal of Petroleum Science and Engineering, 2022, 208, Part D:109549.
[4]
YU S, MA J. Deep learning for geophysics:current and future trends[J]. Reviews of Geophysics, 2021, 59(3):e2021RG000742.
[5]
LIU Lihui, LU Rong, YANG Wenkui. Seismic lithofacies inversion based on deep learning[J]. Geophysical Prospecting for Petroleum, 2019, 58(1):123-129.刘力辉, 陆蓉, 杨文魁. 基于深度学习的地震岩相反演方法[J].石油物探, 2019, 58(1):123-129.
[6]
SONG L, YIN X, ZONG Z, et al. Two-stage semi-supervised learning inversion for reservoir physical parameters[J]. Journal of Petroleum Science and Engineering, 2022, 216:110794.
[7]
DAS V, POLLACK A, WOLLNER U, et al. Convolutional neural network for seismic impedance inversion[J]. Geophysics, 2019, 84(6):R869-R880.
[8]
YANG F, MA J. Deep-learning inversion:a next-generation seismic velocity model building method[J]. Geophysics, 2019, 84(4):R583-R599.
[9]
XUE Y J, WANG X J, CAO J X, et al. Hydrocarbon detections using multi-attributes based quantum neural networks in a tight sandstone gas reservoir in the Sichuan Basin, China[J]. Artificial Intelligence in Geosciences, 2021, 2:107-114.
[10]
ZHANG K, LIN N T, YANG J Q, et al. Predicting gas-bearing distribution using DNN based on multi-component seismic data:quality evaluation using structural and fracture factors[J]. Petroleum Science, 2022, 19(4):1566-1581.
[11]
ZHANG K, LIN N, TIAN G, et al. Unsupervised-learning based self-organizing neural network using multi-component seismic data:Application to Xujiahe tight-sand gas reservoir in China[J]. Journal of Petroleum Science and Engineering, 2022, 209:109964.
[12]
LIN Niantian, ZHANG Dong, ZHANG Kai, et al. Predicting distribution of hydrocarbon reservoirs with seismic data based on learning of the small-sample convolution neural network[J]. Chinese Journal of Geophysics, 2018, 61(10):4110-4125.林年添, 张栋, 张凯, 等. 地震油气储层的小样本卷积神经网络学习与预测[J]. 地球物理学报, 2018, 61(10):4110-4125.
[13]
WANG Di, ZHANG Yiming, ZHANG Fanchang, et al. Quantitative prediction of tight sandstone sweet spots based on deep learning method with prior information constraints[J]. Oil Geophysical Prospecting, 2023, 58(1):65-74.王迪, 张益明, 张繁昌, 等. 利用先验信息约束的深度学习方法定量预测致密砂岩"甜点"[J]. 石油地球物理勘探, 2023, 58(1):65-74.
[14]
GAO J, SONG Z, GUI J, et al. Gas-bearing prediction using transfer learning and CNNs:an application to a deep tight dolomite reservoir[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19:1-5.
[15]
JÚNIOR D A D, BATISTA DA CRUZ L, OTÁVIO BANDEIRA DINIZ J, et al. Detection of potential gas accumulations in 2D seismic images using spatio-temporal, PSO, and convolutional LSTM approaches[J]. Expert Systems With Applications, 2023, 215:119337.
[16]
LIM B, YU H, YOON D, et al. Machine learning derived AVO analysis on marine 3D seismic data over gas reservoirs near South Korea[J]. Journal of Petroleum Science and Engineering, 2021, 197:108105.
[17]
ZHANG Yan, ZHENG Xiaodong, LI Jinsong, et al. Unsupervised seismic facies analysis technology based on SOM and PSO[J]. Chinese Journal of Geophysics, 2015, 58(9):3412-3423.张䶮, 郑晓东, 李劲松, 等. 基于SOM和PSO的非监督地震相分析技术[J]. 地球物理学报, 2015, 58(9):3412-3423.
[18]
WANG Y J, WANG L J, LI K H, et al. Unsupervised seismic facies analysis using sparse representation spectral clustering[J]. Applied Geophysics, 2020, 17(4):533-543.
[19]
LIU S, NI W, FANG W, et al. Absolute acoustic impedance inversion using convolutional neural networks with transfer learning[J]. Geophysics, 2023, 88(2):R163-R174.
[20]
WANG Zefeng, LI Yonggen, XU Huiqun, et al. Comparative analysis of three seismic impedance inversion methods based on deep learning[J]. Oil Geophysical Prospecting, 2022, 57(6):1296-1303.王泽峰, 李勇根, 许辉群, 等. 基于深度学习的三种地震波阻抗反演方法比较[J]. 石油地球物理勘探, 2022, 57(6):1296-1303.
[21]
ALEARDI M, SALUSTI A. Elastic pre-stack inversion through discrete cosine transform reparamete-rization and convolutional neural networks[J]. Geophysics, 2020, 86(1):R129-R146.
[22]
WEI C, GUO X B, TIAN F, et al. Seismic velocity inversion based on CNN-LSTM fusion deep neural network[J].Applied Geophysics,2021,18(4):499-514.
[23]
ZHANG H, ZHANG G, GAO J, et al. Seismic impedance inversion based on geophysical-guided cycle-consistent generative adversarial networks[J]. Journal of Petroleum Science and Engineering, 2022, 218:111003.
[24]
HONGFA K, YONGGUANG C, YI L. Data processing of small samples based on grey distance information approach[J]. Journal of Systems Engineering and Electronics, 2007, 18(2):281-289.
[25]
ZHANG X, YU L, YIN H, et al. Integrating data augmentation and hybrid feature selection for small sample credit risk assessment with high dimensionality[J]. Computers & Operations Research, 2022, 146:105937.
[26]
LIU G, ZHANG L, WANG Q, et al. Data-driven seismic prestack velocity inversion via combining residual network with convolutional autoencoder[J]. Journal of Applied Geophysics, 2022, 207:104846.
[27]
ARAYA-POLO M, FARRIS S, FLOREZ M. Deep learning-driven velocity model building workflow[J]. The Leading Edge, 2019, 38(11):872a1-872a9.
[28]
WANG Y Q, WANG Q, LU W K, et al. Seismic impedance inversion based on cycle-consistent generative adversarial network[J]. Petroleum Science, 2022, 19(1):147-161.
[29]
XIE P, HOU J, YIN Y, et al. Seismic inverse modeling method based on generative adversarial networks[J]. Journal of Petroleum Science and Engineering, 2022, 215, Part B:110652.
[30]
WANG Z, WANG S, ZHOU C, et al. Dual wasserstein generative adversarial network condition:a generative adversarial network-based acoustic impedance inversion method[J]. Geophysics, 2022, 87(6):R401-R411.
[31]
WANG P, SI F, FAN W, et al. Data enhancement for data-driven modeling in power plants based on a conditional variational-adversarial generative network[J]. Industrial & Engineering Chemistry Research, 2021, 60(24):8829-8843.
[32]
FENG Yongji, CHEN Xuehua. Seismic data reconstruction based on partial convolution and attentional mechanism adversarial network model[J]. Oil Geophysical Prospecting, 2023, 58(1):21-30.冯永基, 陈学华. 融合部分卷积和注意力机制对抗网络模型的地震数据重建[J]. 石油地球物理勘探, 2023, 58(1):21-30.
[33]
BISWAS R, SEN M K, DAS V, et al. Prestack and poststack inversion using a physics-guided convolutional neural network[J]. Interpretation, 2019, 7(3):SE161-SE174.
[34]
RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics-informed neural networks:A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics, 2019, 378:686-707.
[35]
VASHISTH D, MUKERJI T. Direct estimation of porosity from seismic data using rock-and wave-phy-sics-informed neural networks[J].The Leading Edge,2022, 41(12):840-846.
[36]
DUAN Youxiang, CUI Lele, SUN Qifeng, et al. Deep learning seismic waveform inversion based on the forward modeling guidance of wave equation[J]. Oil Geophysical Prospecting, 2023, 58(3):485-494.段友祥, 崔乐乐, 孙歧峰, 等. 波动方程正演引导的深度学习地震波形反演[J]. 石油地球物理勘探, 2023, 58(3):485-494.
[37]
DAS V, MUKERJI T. Petrophysical properties prediction from prestack seismic data using convolutional neural networks[J].Geophysics,2020,85(5):N41-N55.
[38]
FENG R. Estimation of reservoir porosity based on seismic inversion results using deep learning methods[J]. Journal of Natural Gas Science and Engineering, 2020, 77:103270.
[39]
ZHONG Z, SUN A Y, WU X. Inversion of time-lapse seismic reservoir monitoring data using cyclegan:a deep learning-based approach for estimating dynamic reservoir property changes[J]. Journal of Geophysical Research:Solid Earth,2020,125(3):e2019JB018408.
[40]
LI Mingxuan, HAN Hongwei, LIU Haojie, et al. Permeability prediction and uncertainty quantification base on Bayesian neural network and data distribution domain transformation[J]. Chinese Journal of Geophysics, 2023, 66(4):1664-1680.李明轩, 韩宏伟, 刘浩杰, 等. 基于数据分布域变换与贝叶斯神经网络的渗透率预测及不确定性估计[J]. 地球物理学报, 2023, 66(4):1664-1680.
[41]
HAMPSON D P, SCHUELKE J S, QUIREIN J A. Use of multiattribute transforms to predict log properties from seismic data[J]. Geophysics, 2001, 66(1):220-236.
[42]
SUN Yongzhuang, HUANG Yun. Application of multi-task deep learning in reservoir shear wave prediction[J]. Progress in Geophysics, 2021, 36(2):799-809.孙永壮, 黄鋆. 多任务深度学习技术在储层横波速度预测中的应用[J]. 地球物理学进展, 2021, 36(2):799-809.
[43]
LI Duo, YANG Senlin, REN Yuxiao, et al.Multi parameter inversion of seismic elastic wave data based on multi task learning[C]//Proceedings of the 2020 China Earth Science Joint Academic Annual Conference, Chongqing, 2020, 80-83.李铎, 杨森林, 任玉晓, 等. 基于多任务学习的地震弹性波数据多参数反演[C]//2020年中国地球科学联合学术年会论文集, 重庆, 2020, 80-83.
[44]
LI D, PENG S, GUO Y, et al. Progressive multitask learning for high-resolution prediction of reservoir elastic parameters[J].Geophysics,2023,88(2):M71-M86.
[45]
FENG R, GRANA D, BALLING N. Uncertainty quantification in fault detection using convolutional neural networks[J]. Geophysics, 2021, 86(3):M41-M48.
[46]
GAL Y, GHAHRAMANI Z. Dropout as a Bayesian approximation:representing model uncertainty in deep learning[C]//Proceedings of the 33rd International Conference on Machine Learning, New York, NY, USA, 2016, 1050-1059.
[47]
FENG R. Physics-informed deep learning for rock physical inversion and its uncertainty analysis[J]. Geoenergy Science and Engineering, 2023, 230:212229.
[48]
ZOU C, ZHAO L, XU M, et al. Porosity prediction with uncertainty quantification from multiple seismic attributes using Random Forest[J]. Journal of Geophysical Research:Solid Earth, 2021, 126(7):e2021JB021826.
[49]
SONG C, LU M, WANG Y, et al. New networks findings from Tsinghua University outlined(a dynamic time warping loss-based closed-loop CNN for seismic impedance inversion)[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60:5925313.