Review of Artificial Intelligence for Oil and Gas Exploration: Convolutional Neural Network Approaches and the U-Net 3D Model

Deep learning, especially through convolutional neural networks (CNN) such as the U-Net 3D model, has revolutionized fault identification from seismic data, representing a significant leap over traditional methods. Our review traces the evolution of CNN, emphasizing the adaptation and capabilities of the U-Net 3D model in automating seismic fault delineation with unprecedented accuracy. We find: 1) The transition from basic neural networks to sophisticated CNN has enabled remarkable advancements in image recognition, which are directly applicable to analyzing seismic data. The U-Net 3D model, with its innovative architecture, exemplifies this progress by providing a method for detailed and accurate fault detection with reduced manual interpretation bias. 2) The U-Net 3D model has demonstrated its superiority over traditional fault identification methods in several key areas: it has enhanced interpretation accuracy, increased operational efficiency, and reduced the subjectivity of manual methods. 3) Despite these achievements, challenges such as the need for effective data preprocessing, acquisition of high-quality annotated datasets, and achieving model generalization across different geological conditions remain. Future research should therefore focus on developing more complex network architectures and innovative training strategies to refine fault identification performance further. Our findings confirm the transformative potential of deep learning, particularly CNN like the U-Net 3D model, in geosciences, advocating for its broader integration to revolutionize geological exploration and seismic analysis.

A 3D attention U-Net network and its application in geological model parameterization

To solve the problems of convolutional neural network–principal component analysis(CNN-PCA)in fine description and generalization of complex reservoir geological features,a 3D attention U-Net network was proposed not using a trained C3D video motion analysis model to extract the style of a 3D model,and applied to complement the details of geologic model lost in the dimension reduction of PCA method in this study.The 3D attention U-Net network was applied to a complex river channel sandstone reservoir to test its effects.The results show that compared with CNN-PCA method,the 3D attention U-Net network could better complement the details of geological model lost in the PCA dimension reduction,better reflect the fluid flow features in the original geologic model,and improve history matching results.

The Method and Practice of Constructing 3D Geological Model from Coalfield Exploration 2D Maps

3D geological modeling is an inevitable choice for coal exploration to adapt to the transformation of coal mining for green, fine, transparent and Intelligent mining. In the traditional Coalfield exploration geological reports, the spatial expression form for the coal seams and their surrounding rocks are 2D maps. These 2D maps are excellent data sources for constructing 3D geological models of coal field exploration areas. How to construct 3D models from these 2D maps has been studying in coal exploration industry for a long time, and still no breakthrough has been achieved so far. This paper discusses the principle, method and software design idea of constructing 3D geological model of an exploration area with 2D maps made by AutoCAD/MapGIS. At first, the paper analyzes 3D geological surface expression mode in 3D geological modeling software. It is pointed out that although contour method has unique advantages in coal field exploration, TIN (Triangular Irregular Network) is still the standard configuration of 3D modeling software for coal field. Then, the paper discusses the method of 2D line features obtaining elevation and upgrading 2D curve to 3D curve. Next, the method of semi-automatic partition is introduced to build the boundary ring of the surface patch, that is, the user clicks and selects the line feature to build the outer boundary ring of the surface patch. Then, Auto-process method for fault line inside of the outer boundary ring is discussed, it including construction of fault ring, determining fault ring being normal fault ring or reverse fault ring and an algorithm of dealing with normal fault ring. An algorithm of dealing with reverse fault ring is discussed detailly, the method of expanding reverse fault ring and dividing the duplicate area in reverse fault into two portions is introduced. The paper also discusses the method of extraction ridge line/valley line, the construction of fault plane, the construction of stratum and coal body. The above ideas and methods have been initially implemented in the “3D modeling platform for coal field exploration” software, and applied to the 3D modeling practice of data from several coal field exploration areas in Ningxia, Shanxi, Qinghai, etc.

3D Geological Modeling with Multi-source Data Integration in Polymetallic Region:A Case Study of Luanchuan,Henan Province,China

The development of 3D geological models involves the integration of large amounts of geological data,as well as additional accessible proprietary lithological, structural,geochemical,geophysical,and borehole data.Luanchuan,the case study area,southwestern Henan Province,is an important molybdenum-tungsten -lead-zinc polymetallic belt in China.

Framework system and research flow of uncertainty in 3D geological structure models

Uncertainty in 3D geological structure models has become a bottleneck that restricts the development and application of 3D geological modeling.In order to solve this problem during periods of accuracy assessment,error detection and dynamic correction in 3D geological structure models,we have reviewed the current situation and development trends in 3D geological modeling.The main context of uncertainty in 3D geological structure models is discussed.Major research issues and a general framework system of uncertainty in 3D geological structure models are proposed.We have described in detail the integration of development practices of 3D geological modeling systems,as well as the implementation process for uncertainty evaluation in 3D geological structure models.This study has laid the basis to build theoretical and methodological systems for accuracy assessment and error correction in 3D geological models and can assist in improving 3D modeling techniques under complex geological conditions.

Alternative 3D Modeling Approaches Based on Complex Multi-Source Geological Data Interpretation

Due to the complex nature of multi-source geological data, it is difficult to rebuild every geological structure through a single 3D modeling method. The multi-source data interpretation method put forward in this analysis is based on a database-driven pattern and focuses on the discrete and irregular features of geological data. The geological data from a variety of sources covering a range of accuracy, resolution, quantity and quality are classified and integrated according to their reliability and consistency for 3D modeling. The new interpolation-approximation fitting construction algorithm of geological surfaces with the non-uniform rational B-spline(NURBS) technique is then presented. The NURBS technique can retain the balance among the requirements for accuracy, surface continuity and data storage of geological structures. Finally, four alternative 3D modeling approaches are demonstrated with reference to some examples, which are selected according to the data quantity and accuracy specification. The proposed approaches offer flexible modeling patterns for different practical engineering demands.

An Approach to Computer Modeling of Geological Faults in 3D and an Application

3D geological modeling, one of the most important applications in geosciences of 3D GIS, forms the basis and is a prerequisite for visualized representation and analysis of 3D geological data. Computer modeling of geological faults in 3D is currently a topical research area. Structural modeling techniques of complex geological entities contain- ing reverse faults are discussed and a series of approaches are proposed. The geological concepts involved in computer modeling and visualization of geological fault in 3D are explained, the type of data of geological faults based on geo- logical exploration is analyzed, and a normative database format for geological faults is designed. Two kinds of model- ing approaches for faults are compared: a modeling technique of faults based on stratum recovery and a modeling tech- nique of faults based on interpolation in subareas. A novel approach, called the Unified Modeling Technique for stratum and fault, is presented to solve the puzzling problems of reverse faults, syn-sedimentary faults and faults terminated within geological models. A case study of a fault model of bed rock in the Beijing Olympic Green District is presented in order to show the practical result of this method. The principle and the process of computer modeling of geological faults in 3D are discussed and a series of applied technical proposals established. It strengthens our profound compre- hension of geological phenomena and the modeling approach, and establishes the basic techniques of 3D geological modeling for practical applications in the field of geosciences.

A 3D Geological Model Constrained by Gravity and Magnetic Inversion and its Exploration Implications for the World-class Zhuxi Tungsten Deposit, South China

The Zhuxi tungsten deposit in Jiangxi Province,South China,contains a total W reserve of about 2.86 Mt at an average grade of 0.54 wt%WO3,representing the largest W deposit in the world.Numerous studies on the metallogeny of the deposit have included its timing,the ore-controlling structures and sedimentary host rocks and their implications for mineral exploration.However,the deep nappe structural style of Taqian-Fuchun metallogenic belt that hosts the W deposit,and the spatial shape and scale of deeply concealed intrusions and their sedimentary host rocks are still poorly defined,which seriously restricts the discovery of new deposits at depth and in surrounding areas of the W deposit.Modern 3 D geological modeling is an important tool for the exploration of concealed orebodies,especially in brownfield environments.There are obvious density contrast and weak magnetic contrast in the ore-controlling strata and granite at the periphery of the deposit,which lays a physical foundation for solving the 3 D spatial problems of the ore-controlling geological body in the deep part of the study area through gravity and magnetic modeling.Gravity data(1:50000)and aeromagnetic data(1:50000)from the latest geophysical surveys of 2016-2018 have been used,firstly,to carry out a potential field separation to obtain residual anomalies for gravity and magnetic interactive inversion.Then,on the basis of the analysis of the relationship between physical properties and lithology,under the constraints of surface geology and borehole data,human-computer interactive gravity and magnetic inversion for 18 cross-sections were completed.Finally,the 3 D geological model of the Zhuxi tungsten deposit and its periphery have been established through these 18 sections,and the spatial shape of the intrusions and strata with a depth of 5 km underground were obtained,initially realizing―transparency‖for ore-controlling bodies.According the analysis of the geophysical,geochemical,and geological characteristics of the Zhuxi tungsten deposit,we discern three principles for prospecting and prediction in the research area,and propose five new exploration targets in its periphery.

Coping with uncertainties through an automated workflow for 3D reservoir modelling of carbonate reservoirs

Reliable 3D modelling of underground hydrocarbon reservoirs is a challenging task due to the complexity of the underground geological formations and to the availability of different types of data that are typically affected by uncertainties. In the case of geologically complex depositional environments, such as fractured hydrocarbon reservoirs, the uncertainties involved in the modelling process demand accurate analysis and quantification in order to provide a reliable confidence range of volumetric estimations. In the present work, we used a 3D model of a fractured carbonate reservoir and populated it with different lithological and petrophysical properties. The available dataset also included a discrete fracture network(DFN) property that was used to model the fracture distribution. Uncertainties affecting lithological facies, their geometry and absolute positions(related to the fault system), fracture distribution and petrophysical properties were accounted for. We included all different types of uncertainties in an automated approach using tools available in today’s modelling software packages and combining all the uncertain input parameters in a series of statistically representative geological realizations. In particular, we defined a specific workflow for the definition of the absolute permeability according to an equivalent, single porosity approach, taking into account the contribution of both the matrix and the fracture system. The results of the analyses were transferred into a 3D numerical fluid-dynamic simulator to evaluate the propagation of the uncertainties associated to the input data down to the final results, and to assess the dynamic response of the reservoir following a selected development plan. The "integrated approach" presented in this paper can be useful for all technicians involved in the construction and validation of 3D numerical models of hydrocarbon-bearing reservoirs and can potentially become part of the educational training for young geoscientists and engineers, since an integrated and well-constructed workflow is the backbone of any reservoir study.

3D modeling of geological anomalies based on segmentation of multiattribute fusion

3D modeling of geological bodies based on 3D seismic data is used to define the shape and volume of the bodies, which then can be directly applied to reservoir prediction, reserve estimation, and exploration. However, multiattributes are not effectively used in 3D modeling. To solve this problem, we propose a novel method for building of 3D model of geological anomalies based on the segmentation of multiattribute fusion. First, we divide the seismic attributes into edge- and region-based seismic attributes. Then, the segmentation model incorporating the edge- and region-based models is constructed within the levelset- based framework. Finally, the marching cubes algorithm is adopted to extract the zero level set based on the segmentation results and build the 3D model of the geological anomaly. Combining the edge-and region-based attributes to build the segmentation model, we satisfy the independence requirement and avoid the problem of insufficient data of single seismic attribute in capturing the boundaries of geological anomalies. We apply the proposed method to seismic data from the Sichuan Basin in southwestern China and obtain 3D models of caves and channels. Compared with 3D models obtained based on single seismic attributes, the results are better agreement with reality.

Automatic Generation Method of Geological Cross-Sections in Dredging Engineering Based on 3D Solid NURBS Models

An automatic generation method of geological cross-sections in dredging engineering based on 3D geological solid models is presented.The 3D geological models are built applying the non-uniform rational B-splines(NURBS) technique,and a 2D profile can be calculated and generated automatically through Boolean operation to meet the demands of dredging projects.Moreover,an automatic marking method for geological attributes is put forward based on database technology,and the geological attributes include the profile name,scale,horizontal and vertical relative coordinates,geological lithology,and 2D standard lithology legend.At the same time,the automatic marking method can also provide an interactive mode for geological engineers to edit and modify the profile in the modeling system.Practical engineering applications show that the automatic generation method is a simple,flexible,fast and precise visual graphics rendering process that can create 2D standard profiles automatically and efficiently.This method also provides a convenient support tool for geological engineering digital analysis.

Integrating artificial neural networks and geostatistics for optimum 3D geological block modeling in mineral reserve estimation:A case study

In this research, a method called ANNMG is presented to integrate Artificial Neural Networks and Geostatistics for optimum mineral reserve evaluation. The word ANNMG simply means Artificial Neural Network Model integrated with Geostatiscs, In this procedure, the Artificial Neural Network was trained, tested and validated using assay values obtained from exploratory drillholes. Next, the validated model was used to generalize mineral grades at known and unknown sampled locations inside the drilling region respectively. Finally, the reproduced and generalized assay values were combined and fed to geostatistics in order to develop a geological 3D block model. The regression analysis revealed that the predicted sample grades were in close proximity to the actual sample grades, The generalized grades from the ANNMG show that this process could be used to complement exploration activities thereby reducing drilling requirement. It could also be an effective mineral reserve evaluation method that could oroduce optimum block model for mine design.

3D Integrated Geological Modeling in Tongshan Copper Deposit,Heilongjiang Province,China

In this paper,3D-GIS reconstruction and interpolation approach,additional virtual borehole technology and BP network technology are used to explore the concealed ore body.The virtual borehole has same function as reality borehole due to the multi-information check and validation in

基于3Dmine的矿山防治水分析应用模型建立与研究

矿山三维建模是建设数字矿山的重要手段,文章通过3Dmine矿业工程软件创建了莱新铁矿防治水分析应用模型。在收集矿山地质勘察、巷道、钻孔、涌水量等资料的基础上,进行了数字化工作并建立相关数据库,从而构建了矿山三维地质模型、矿体模型、巷道模型及钻孔涌水模型4个部分,并利用涌水来衡量模型的合理性。模型实现了真实直观的井下近矿体帷幕工程三维可视效果,将有利于矿床水文地质条件分析、防治水工程的设计与研究以及工程稳定性评价等工作。

3D stratum modeling in urban geological information visualization

A conceptual model called virtual borehole is presented in practice engineering.Based on borehole data and the geologic section and by using GMS 6.0,MAPGIS software,the 3D stratum model of urban geologic information is established.The visual expression and analysis of 3D geologic data can be better realized in the open model established.

Application of 3D Reservoir Geological Model on Es1 Formation, Block Nv32, Shenvsi Oilfield, China

Three-dimensional geological modeling of reservoirs is an essential tool to predict reservoir performance and improve the understanding of reservoir uniqueness in Es1 formation. The paper focuses on the use of petrel software to build three-dimensional reservoir geological model which characterizes and assesses block Nv32 that located in the west of the Shenvsi oilfield in the south of Cangzhou city, Hebei province of China, and has an oil-bearing area of 1.4 km2. This study is depending on integration data from well logs of 22 wells which provided from geology, geophysics, and petrophysics to identify and provide precise depict of the subsurface internal structure and the reservoir heterogeneity. Input data was used to build the structural model, sedimentary facies model, petrophysical properties (porosity, permeability, saturation, and N/G model, and finally to determine the reservoir volume. The lithological facies were simulated using the assigned value method. Moreover, Petrophysical properties (Porosity, permeability, oil saturation and net to gross) were constructed for each zone using the Sequential Gaussian Simulation method to guide the distribution of petrophysical properties of Es1 formation, block Nv32. Statistical analysis of the porosity, permeability, oil saturation and N/G model present that the porosity occurrence distribution is mainly concern between 0.2% - 36.39% of block Nv32 with an average porosity value of 17.5%, permeability between 0.017 mD to 974.8 mD, having an average permeability of 59.44 mD, oil saturation between 0.00 to 0.95 having an average value of 0.22, and N/G is mainly concentrated between 0.01 to 1.00 within an average value of 0.61. This research has indicated the reliability of the three-dimensional model technique as a suitable tool to provide a sufficient understanding of petrophysical distribution. The south-western and north-western indicate that oilfield is very promising an exploratory well should be drilled to find out the thickness and size of the reservoir.

基于SURPAC模型的采空区与露采边坡相互影响的FLAC^(3D)分析

采用SURPAC建立了瓮福磷矿穿岩洞矿段采场的三维地质模型,利用接口程序将模型导入FLAC3D对采空区与露采边坡之间的相互影响进行了计算,并在边坡坡顶、坡脚与采空区四周布置了监测点以监测各点应力与位移的变化。计算结果表明:采空区的存在极大地改变了采空区周边与边坡的应力分布;使得采空区周边水平位移出现明显增加,并且改变了采空区位置周边的竖向位移场,对各个采空区之间的区域的影响最为显著,使得回弹量减小;边坡与采空区监测点的应力与位移随着边坡露采的进行出现规律性的变化;采空区的存在极大地改变了岩体塑性区的分布,对于有采空区边坡,塑性区分布范围随着开挖的进行有减小的趋势,但坡面上出现明显的拉伸破坏区,不利于边坡稳定;采空区使得露采边坡的安全系数降低。

Study upon Modeling and Visualization of 3D Geologic Body Based on Generalized Tri-Prism

According to the characteristics of bore data,a model of 3D geologic body with generalized tri-prism as the primitive modeling element is constructed while the modeling process and key algorithms of modeling are presented here in detail.Using this method,the original bore data go through Delaunay triangulation to generate irregular triangular network on the surface,and then links stratum segments on the adjoining bores in session to form tri-prisms which would be pinched out.Finally stratified 3D geologic body model is built by an iterated search which searches for consecutive layer of the same property.The result shows that this method can effectively simulate stratified stratum modeling.

Deposit 3D modeling and application

By the aid of the international mining software SURPAC, a geologic database for a multi-metal mine was established, 3D models of the surface, geologic fault, ore body, cavity and the underground openings were built, and the volume of the cavity of the mine based on the cavity 3D model was calculated. In order to compute the reserves, a grade block model was built and each metal element grade was estimated using Ordinary Kriging. Then, the reserve of each metal element and every sublevel of the mine was worked out. Finally, the calculated result of each metal reserve to its actual prospecting reserve was compared, and the results show that they are all almost equal to each other. The absolute errors of Sn, Pb, and Zn reserves are only 1.45%, 1.59% and 1.62%, respectively. Obviously, the built models are reliable and the calculated results of reserves are correct. They can be used to assist the geologic and mining engineers of the mine to do research work of reserves estimation, mining design, plan making and so on.

Regional 3D geological modeling along metro lines based on stacking ensemble model

This paper presents a regional 3D geological modeling method based on the stacking ensemble technique to overcome the challenges of sparse borehole data in large-scale linear underground projects.The proposed method transforms the 3D geological modeling problem into a stratigraphic property classification problem within a subsurface space grid cell framework.Borehole data is pre-processed and trained using stacking method with five different machine learning algorithms.The resulting modelled regional cells are then classified,forming a regional 3D grid geological model.A case study for an area of 324 km2 along Xuzhou metro lines is presented to demonstrate the effectiveness of the proposed model.The study shows an overall prediction accuracy of 85.4%.However,the accuracy for key stratigraphy layers influencing the construction risk,such as karst carve strata,is only 4.3%due to the limited borehole data.To address this issue,an oversampling technique based on the synthetic minority oversampling technique(SMOTE)algorithm is proposed.This technique effectively increases the number of sparse stratigraphic samples and significantly improves the prediction accuracy for karst caves to 65.4%.Additionally,this study analyzes the impact of sampling distance on model accuracy.It is found that a lower sampling interval results in higher prediction accuracy,but also increases computational resources and time costs.Therefore,in this study,an optimal sampling distance of 1 m is chosen to balance prediction accuracy and computation cost.Furthermore,the number of geological strata is found to have a negative effect on prediction accuracy.To mitigate this,it is recommended to merge less significant stratigraphy layers,reducing computation time.For key strata layers,such as karst caves,which have a significant impact on construction risk,further onsite sampling or oversampling using the SMOTE technique is recommended.