Application of Machine Learning in Reservoir Characterization of the Tight Sandstone

作     者：Shiba K C 

作者单位：中国石油大学(北京) 

学位级别：硕士

导师姓名：Zhao Jianguo

授予年度：2022年

学科分类：0820[工学-石油与天然气工程] 12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081803[工学-地质工程] 081104[工学-模式识别与智能系统] 08[工学] 0818[工学-地质资源与地质工程] 0835[工学-软件工程] 082002[工学-油气田开发工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

主      题：Machine Learning Reservoir Characterization Neural Networks Support Vector Machines Random Forest Simultaneous Seismic Inversion 

摘      要：Reservoir characterization is the process of accessing the viability of the reservoir with the determination of the physical reservoir parameters ***,permeability,lithology,formation thickness,saturation,*** data types *** data,well-log data,and petrophysical core data are manipulated individually or integrated together to gain a better insight into these reservoir ***,approaches like empirical,and model-driven deterministic methods including soft computing methods like machine learning（ML）have been established to determine these *** learning is a widely accepted and appreciated approach in reservoir *** machine learning algorithms are successfully employed and proven to be better performing than the conventional methods of reservoir parameter *** subpar performance of conventional methods and their time and capital inefficienc y aspires to machine learning usage in reservoir *** this study,reservoir parameters:S-wave velocity,porosity,permeability,and shale percentage are predicted using machine learning regression algorithms:Support vector regression（SVR）,Random Forest（RF）,and Deep neural network（DNN）.S-wave velocity is predicted using well logs and porosity,permeability,and shale are predicted using both well logs and seismic *** study makes use of the data from a tight sandstone reservoir in the Hangjinqi area of the Ordos basin in *** establishes the applicability of ML algorithms for a heterogeneous reservoir with a comparative ML regression model analysis.S-wave velocity is predicted from well log attributes:acoustic（AC）,density（DEN）,and gamma-ray log（GR）with a good performance by all three ***,SVR outperforms the remaining with an accuracy of 0.972 *** model performance is validated with a blind well test as ***,Permeability,and Shale percentage were predicted with inputs P-impedance,S-impedance,V/V ratio,and density.A