Natural fractures play a crucial role in the storage and seepage of oil shale. However, identifying fractures using conventional logging techniques presents challenges due to complex response characteristics and severe data imbalance. Here, we propose a highly accurate integrated ensemble learning method, called BSI-XGBoost, for identifying the natural fracture development, which combines several steps including Isolation Forests (iForest), Synthetic Minority Oversampling Techniques (SMOTE), and Extreme Gradient Boosting (XGBoost), and incorporates rock brittleness as a controlling factor in the model construction process. The proposed model effectively addresses several challenges encountered in fracture identification, including complex logging response characteristics, low precision and recall of fractured labels, and excessive sensitivity of ensemble learning to noise. To do so, the relationship between fracture density and brittle mineral content is analyzed through core analysis and XRD. Then, conventional logging and rock brittleness are used as features for training the model. Herein, by screening the outliers of iForest, SMOTE oversampling, and feature selection, optimal hyperparameters of the model are obtained through the grid search method. The results demonstrated that using BSI-XGBoost, the testing set achieved an accuracy of 92.45%. Comparatively, this accuracy is 4.86% higher than the original XGBoost model and 3.73% higher than the B-XGBoost model, which incorporated brittleness curves but did not include oversampling and outlier removal. Collectively, this workflow provided an effective method for intelligent identification of fractures in oil shale with high accuracy based on easily accessible conventional logging curves.

Identification of natural fractures in continental shale is a problematic task by conventional logging. To address this issue, a method known as Frac-gPCC that combines graph pooling, graph construction, and node classification to train the identification model is proposed in this study. This method integrates existing geological knowledge into the graph structure and model network construction, captures the topology information of a single fracture and fractured zones through the calculation of graphs. The model, effectively addresses the issue of incorrectly identifying samples as non-fracture when the number of fracture samples is significantly less than non-fractures. Concurrently, this method utilizes a large number of unlabeled samples effectively to incorporate them in model training, avoiding the problem of limited labeled samples. The identification process is divided into three steps: first, reconstruct the logging curve and use the graph pooling section to filter and fuse the node information within a certain depth interval to enhance the characteristics of the tool response to fractures. Second, it integrates the relationship between fractures and lithologies, as well as the spatial distribution characteristics of the strata, into the structure of the global graph. Finally, the nodes of the constructed global graph are classified through node classification section, which naturally supports combination generalization. This method is applied in the Fengcheng Formation of the Mahu Sag, western China. The results showed that the identification accuracy for testing data reaches 91.96%. Collectively, this reflects the superiority of Frac-gPCC in fracture identification, providing a successful workflow for continental shale characterization.

Fracture data acquired from drill core and borehole image logs require corrections for the bias due to fracture orientation, that are usually achieved by the Terzaghi correction technique. Previous studies often approximate the wellbore as a one-dimensional scanline, assuming that the length of the core axis within the sampling range is equal to the scanline length. This study refers to the commonly used workflow as the original Terzaghi correction method which is known to perform poorly when the angle (θ) between the core axis and the fracture is small. To address this issue, we propose an extension of the Terzaghi correction method that also considers the core diameter and is a function of both the fractures and the host layer dipping angle. The new method resolves the fracture density problem by selecting a new direction for the scanline perpendicular to the fracture and calculating the projection length of the sampling space in this direction. The fracture spatial arrangements and observed number of sampled fractures mainly affect the performance of this new approach. However, possible errors in the new correction method pertaining to the equidistant fracture density should decrease with increasing number of sampled fractures. It is found that an acceptable correction range exists for equidistant fracture density, though, when the corrected density is not within the acceptable correction range, the observed sampled fractures will not be equal to the true observed value. This means the corrected fracture density would be in an unacceptable range of errors. Moreover, the new method can provide results within the acceptable range when the angle between the fracture and the core axis is less than 20 °. At the same time, this new method is free from other disadvantages in the original Terzaghi correction method, particularly, when the ratio of layers’ thickness to the core diameter is low. Therefore, the improved approach presented here is especially applicable to thin layers (no more than two times the core diameter) and conditions where the angle between the fracture and the core axis is less than 20 °, which can contribute to fracture density characterization in the subsurface.

The natural fractures, e.g., the bedding-parallel fractures (BPF), have a significant impact on the storage space and horizontal permeability of continental shale oil reservoirs. However, the conventional logging response of BPF is complex. To solve the problem of BPF density prediction, we propose the Edge Generation Weighted Graph Convolutional Network (EG-WGCN) method, which is an improved Graph Convolutional Network (GCN). The new method integrates the relationship information between the fractures and lithology into the edge generation method, and adds the generated edge weight information into the message passing process, thereby effectively improving the accuracy of fracture density prediction. The prediction process is divided into three steps: first, using conventional logging sampling points as vertices, establish vertex sets for each lithology and single lithologic layer. Second, based on the depth sequence of a single lithologic layer and the Euclidean distance of the same lithologic vertex sets, the connection between vertices is established and overlapping edges generated by the above methods are given higher weights. Third, the vertices in the constructed graph are classified through graph convolutional layers, which are integrated into the edge weight information. This method is applied to the fracture density prediction of continental shale oil reservoirs in the Mahu Sag of the Junggar Basin, western China. The results show that the accuracy of our proposed EG-WGCN method in testing data is 95.13%. Compared to the GCN (boundless) without using the edge generation method and EG-GCN with the edge generation method but without the weighted aggregation mechanism, model accuracy is improved by 14.48% and 1.22%, respectively. In summary, this method proves that if a proper ML model is used, conventional logging data can become valuable for predicting BPF density with high accuracy even when cores are missing.

Lacustrine shale oil reservoirs of the Fengcheng Formation in the Mahu Sag, Junggar Bain, NW China contain abundant oil resources, with the highest daily oil production of a single well reaching over 700 barrels. Based on core and thin section observation along with borehole image interpretation, this study integrated paleotectonic stress regime, fluid inclusion analysis, and calcite U-Pb dating to clarify fracturing periods and discuss the influence of fractures on shale oil accumulation. Results show that natural opening-mode fractures are wide-spread in the Fengcheng Formation, which play an important role in oil accumulation. Fractures are developed in several stages and filled by three minerals: reedmergnerite, quartz and calcite. Reedmergnerite and quartz fillings show similar fluid inclusions and petrographic features, with numerous hydrocarbon inclusions compared to aqueous ones, yet, such inclusions were less observed in calcite fillings. Fractures in the study area were formed in two separate stages. In the first period, due to the NW-SE compression of the Hercynian movement, the WNW-ESE and NNW-SSE striking fractures were formed in the Late Permian, with Th values of 95-115 degrees C. The first stage of fractures coincided with hydrocarbon generation, providing migration pathways for the oil. In the second period, due to the nearly S-N compression of the Indosinian movement, the NNW-SSE and NNE-SSW striking fractures were formed in the late Triassic, with Th values of 120-140 degrees C. The second stage of frac-tures was formed subsequent to major hydrocarbon generation and expulsion period, hence their contribution to hydrocarbon migration and accumulation was found relatively minor. Moreover, fractures with the nearly E-W, ENE-WSW and WNW-ESE strikes intersect with the present-day maximum horizontal principal stress at a small angle, with large apertures and good connectivity, which should have a positive impact on shale oil accumulation.

The continental shale oil reservoir of Fengcheng Formation in the northern slope area of Mahu Sag, Junggar Basin, Western China is very heterogeneous in lithology. Thus, the complex response characteristics of conventional logging and limited core availability in the study area has led to major challenges in lithology identification. Therefore, to resolve lithology identification by well logs in continental shale oil reservoirs, a graph neural network (GNN) method named GraphSAGE is used to train the lithology identification model based on a constructed graph, which connects the samples with adjacent depth and similar log response features on operator intention. The identification process is divided into two parts: first, based on the formation depth sequence and affinity propagation clustering method, the vertical distribution of the stratum and nodes logging curve similarity information are integrated into the graph structure, which structurally represents the conventional logging curves as graph instead of well logs as input data; Second, the nodes of the constructed graph are classified by GraphSAGE, which naturally supports combination generalization and improves sample complexity accompanied by strong relational inductive bias. To examine the effectiveness of GraphSAGE for lithology identification, a conventional log dataset labelled by direct core observations from two separate wells in Muhu Sag are used. The identification results showed that the accuracy of GraphSAGE for the lithologies exceeds 90% of the testing data, especially for transitional lithology such as dolomitic mudstone, silty mudstone and tuffaceous fine sandstone. Compared with the commonly used machine learning methods such as SVM, RF and XGBoost, GraphSAGE was more accurate in lithology identification, matching core observations. Collectively, this reflects the superiority of graph neural network in conventional logging lithology identification and effective means provided for lithology identification of continental shale oil reservoir in the Mahu Sag.

The deep continental shale of the Permian Lucaogou Formation in the eastern Junggar Basin is abundant in oil resources, characterized by extensive development and various types of natural fractures. Based on core observations, image log interpretation, and thin section analyses, this study identified fracture types and analyzed the development characteristics of different types in deep continental shales. Moreover, the main controlling factors affecting fracture development and distribution were discussed by integrating statistical and experimental analysis. Results show that natural fractures in deep continental shale oil reservoirs can be grouped as non-bed-parallel and bed-parallel fractures based on their relationship with bedding. Non-bed-parallel fractures are subdivided into subvertical fractures (≥80°) (translayer and intralayer fractures) and moderately dipping fractures (<80°), and bed-parallel fractures are further divided into bedding fractures and stylolites. Since silty and dolomitic rocks appear widely as interlayers and increase the rock brittleness, the deep continental shales are more prone to developing non-bed-parallel fractures. Moreover, intralayer fractures with high dip angles are more developed within these thinner mechanical units and commonly terminate at the mechanical interface, while the translayer fractures are usually less developed. Deep continental shales with higher TOC and argillaceous content are more conducive to forming bed-parallel fractures because of their stronger plasticity, more densely developed laminar structure, and easier formation of abnormally high pressure during the hydrocarbon generation and evolution. In addition, due to the frequent interaction of different types of laminae and the stronger compaction and pressure solution produced by the larger burial depth, the bed-parallel fractures, especially the stylolites, are more developed as a prominent feature in deep continental shales.

The Lower Permian Fengcheng Formation in the Mahu Sag develops a set of organic-rich alkaline lacustrine shale strata, which is a key area for shale oil exploration and development. As an important storage space and seepage channel for shale reservoirs, natural fractures have an impact on shale oil enrichment, production and development effect. In this study, the types and characteristics of natural fractures were first analyzed using core, thin section and imaging logging data. On this basis, combined with the distribution of fractures in single wells, the vertical distribution law of fractures is discussed. Finally, the planar distribution of fractures is evaluated using different seismic attributes such as coherence, curvature, likelihood, and AVAz. The results showed that three types of fractures are existed, including transformational shear fractures, intraformational open fractures and bed-parallel shear fractures, with intraformational open fractures being the most developed. The development degree of fractures in different layers has obvious differences, mainly controlled by lithology and brittle mineral content. The basalt and tuff are developed in the Feng 1 Member, with low carbonate mineral content, resulting in a relatively low degree of fracture development. The dolomite and argillaceous dolomite are developed in the Feng 2 Member and the Feng 3 Member, with high carbonate mineral content and brittleness, resulting in a high degree of fracture development. Additionally, the closer to the fault, the higher the degree of fracture development. On the plane, the fracture zone develops near the main and secondary faults, with the trend mainly oriented in the E-W direction and approximately parallel to the direction of the faults. The width of the fracture zone is largest in the central and southern part of the study area. These fractures are fault-related and are caused by regional stress fields resulting from the activity of the main-secondary faults.

Maximum burial depth of the Permian Fengcheng Formation, which contains significant oil resources in the Mahu Sag, exceeds 4500 m constituting a deep reservoir. Fractures are widespread in the deep lacustrine shale, providing important storage space and fluid conduits. Based on cores, borehole image logs, and petrographic analysis, we identified and analyzed the types and characteristics of fractures and conducted X-ray diffraction (XRD) and Cathodoluminescence (CL) tests to investigate the factors influencing fracture development. We found faults and three types of opening-mode fractures: bed-bounded fractures, unbounded fractures and bed-parallel fractures. Mechanical stratigraphy (depositional rock types modified by diagenesis) and proximity to faults, are main controls of opening-mode fracture occurrence and attributes. Specifically, Dolomitic mudstone with a high carbonate content shows the highest fracture density of 5.24 m(-1). Bed-bounded fractures are mainly developed within brittle layers, with fracture density lower in thicker beds. On the other hand, unbounded (tall) fractures primarily occur near reverse faults (0-1500 m), with orientations parallel to extension directions implied by the kinematics of nearby faults. Additionally, the effectiveness of early-formed fractures as fluid conduits is likely modified (reduced) by cementation and enhanced by dissolution. Fractures have a positive influence on oil production, corresponding to high fracture density and aperture in the vertical direction.

Natural fractures in continental shale oil reservoirs of the Fengcheng Formation in the Mahu Sag show multi-scale characteristics, which leads to complex seismic responses and difficult identification. In order to establish fracture prediction models with good performance in these reservoirs, this study uses seismic attributes such as post-stack coherence, curvature, likelihood, and pre-stack AVAz to predict the multi-scale fractures, including main-secondary faults, large-scale fractures, and medium-small scale fractures in continental shale oil reservoirs. The final prediction results are superimposed on the plane to clarify the multi-scale fracture distribution law of the Fengcheng Formation in the Mahu Sag. Seismic prediction results show that natural fractures in the upper sweet spot of the Fengcheng Formation are more developed, especially in the northern and central platform areas, and they are mainly near E-W strikes. With the increase of the primary-secondary fault distance, the fracture density gradually decreases. Natural fractures obtained by seismic prediction are consistent with the fractures interpreted by image logs, which can be used to effectively predict fractures for continental shale oil reservoirs in the Mahu Sag of the Junggar Basin and other areas with a similar geological background.

马海东地区下干柴沟组下段砂岩储层具有良好的油气勘探前景.基于铸体薄片、扫描电镜和核磁共振T2谱曲线分析,研究了马海东地区下干柴沟组下段砂岩储层孔隙结构特征及主控因素.研究表明:下干柴沟组下段4砂组储集岩发育残余粒间孔、粒内溶孔、成岩缝和孔隙缩小型喉道,孔隙结构以Ⅰ类大孔中喉型和Ⅱ类中孔细喉型为主,中—大孔比例较高,孔隙连通性较好;2砂组储集岩发育残余粒间孔、粒内溶孔和弯片状、管束状喉道,孔隙结构以Ⅲ类细孔微喉型为主,小型孔隙占据主体,孔隙连通性较差.指出沉积环境是储集岩孔隙结构形成的前提条件,机械压实和胶结作用封堵、缩小孔喉,构造作用和溶蚀作用连通、扩大孔喉,导致4砂组与2砂组储层孔隙结构存在明显差异.研究成果为马海东地区下干柴沟组下段砂岩储层评价与预测提供了依据.