Figure - available from: Earth Science Informatics
This content is subject to copyright. Terms and conditions apply.
Source publication
Image well logging is an intuitive approach to identify fractures of reservoir for oil and gas exploration. However, these logging images are rare and nonannotated. A method of unsupervised segmentation network based on convolutional neural network is adopted to automatically extract pixels pertaining to fracture information in this paper. We propo...
Similar publications
As the northwestern area of the Junggar Basin is a key area for oil and gas exploration, the sedimentary facies of the Jurassic formations in the Wuerhe area has long been a focus of research. The target strata are Jurassic strata, including five formations: the Lower Jurassic Badaowan and Sangonghe, the Middle Jurassic Xishanyao and Toutunhe and t...
Accurately measuring wettability is of the utmost importance because it influences several reservoir parameters while also impacting reservoir potential, recovery, development, and management plan. As such, this study proposes a new formulated mathematical model based on the correlation between the Amott-USBM wettability measurement and field NMR T...
The oil and gas industry is a vital component of the global energy landscape. However, it is also a sector characterized by complex engineering challenges and risks, with maintaining well integrity being paramount among them. The ability of a well to avoid the uncontrolled escape of formation fluids over its lifespan is known as well integrity, and...
![Fig. 1. An illustrative example of a raster well-log. Image taken from [2].](profile/M-Quamer-Nasim/publication/377649813/figure/fig1/AS:11431281227984998@1709817965299/An-illustrative-example-of-a-raster-well-log-Image-taken-from-2_Q320.jpg)
Raster well-log images are digital representations of paper copies that retain the original analog data gathered during subsurface drilling. Geologists heavily rely on these images to interpret well-log curves and gain insights into the geological formations beneath the surface. However, manually extracting and analyzing data from these images is t...
A technique is proposed that calculates derivative and volatility attributes from just a few well log curves to assist in brittleness index predictions from sparse well-log datasets with machine learning methods. Six well-log attributes are calculated for selected recorded well logs: the first derivative, the moving average of the first derivative,...
Citations
... In this paper, Liu et al.'s (2021) model is used to calculate the permeability shown by reservoir logs. A model for calculating the logging permeability is established by comparing well core analysis data in the study area. ...
Natural fractures are effective storage spaces and the main seepage channels of tight reservoirs, and the spacing and role of structural fractures of different scales in tight reservoirs are also different. However, the opening or height of underground natural fractures is often difficult to determine directly using seismic data or imaging logging, which restricts the characterization, modeling, and analysis of the formation mechanisms of reservoir fractures. This paper uses the morphology of water absorption profiles and conventional logging data to calculate the ratio of the isotope intensity of a single well to natural gamma values from logging and characterize the types of water absorption profiles. The size (opening and height) of the fractures is closely related to the height of the water absorption profile and the fluctuation in the water absorption intensity. The calculated rock mechanics parameters along a single well, combined with the internal friction angle of the rock and the fracture strike, are used to determine the magnitude and direction of the paleostress of the Y290-Y414 Block in the southeastern Ordos Basin during the Himalayan period. By establishing a single-well geomechanical model to simulate the three-dimensional distribution of paleostress and paleostrain during the formation of fractures, a database model for constructing the scale of fractures is established. Through correlation analysis, a model for predicting the strength and thickness of the water absorption profile is established. Finally, the three-dimensional distribution pattern of the scale of structural fractures is predicted, and the reliability of the prediction results is verified using dynamic development data. The research results have certain reference significance for explaining the mechanism of fracture formation, characterizing the scale of underground fractures, and modeling reservoir fractures.
... Also edge detection algorithms Fernandez, Rodriguez-Lozano et al. 2017, Dorafshan, Thomas et al. 2018) and 66 wavelet transform (Subirats, Dumoulin et al. 2006, Ouma andHahn 2016) (Ai, Jiang et al. 2023). (Zhang, Wu et al. 2021 (Dorafshan, Thomas et al. 2018). Edges or cracks are defined as sharp intensities (1) ...
Modern pavement management systems depend mainly on pavement condition assessment to plan rehabilitation strategies. To assess pavement damages conventionally, manual inspection are performed by trained inspectors. This can be time-consuming and a source of risk for inspectors. Moreover, manual inspection can highly affected by the state of mind of inspectors. To overcome such problems, image based inspection using smartphone camera combined with image processing methods can be used. This combination is relatively cheaper and easier to use. This research proposes an automatic crack detection and mapping program for rigid pavement which can automate visual inspection process. The program consists of various image processing techniques that are used to identify and detect cracks from images. Detected cracks are defined in a pixel coordinate system. Cracks coordinates are converted from pixel coordinates to global coordinates in order to compute their lengths using Global Navigation Satellite Systems (GNSSs) data. The performance of the program was assessed with field study. Cracks quantification process is performed to determine crack lengths and areas. The results show that the precision, recall and accuracy values for the program’s image processing algorithm are 57.00%, 98.81% and 65.22% respectively.
... has unique advantages including intelligence, systemization, and strong learning ability. Related machine learning algorithms have wide application in the field of earth science and engineering practice and have demonstrated excellent results (Aïfa, 2014;Anifowose et al., 2015;Imamverdiyev and Sukhostat, 2019;Liang et al., 2022;Sabah et al., 2019;W. Zhang et al., 2021). In recent years, machine learning using logging data modeling for lithology identification has become popular research area . proposed an improved method based on Hidden Markov Model and Random Forest to obtain a new hidden feature from the elastic parameters. The method is superior in handling different classes of crossover data and p ...
Lithology identification using geophysical log information is vital for log interpretation and reservoir evaluation. As a result of the highly similar features for log curves that characterize complex lithology, there is significant information redundancy regarding the process of lithology identification. In addition, as a result of the highly nonlinear characteristics of log curves, the mapping relationship with lithology has certain ambiguities and uncertainties, which affect the lithology prediction results. Combining principal component analysis (PCA) and the fuzzy decision tree (FDT) model, we propose a new intelligent lithology identification method that is capable of effectively solving these problems well. However, because of the inaccuracy for empirically set parameters, an adaptive fuzzy decision tree algorithm based on particle swarm optimization (PSO-FDT) was proposed after analyzing the main features of the fuzzy decision tree and using an improved particle swarm optimization (PSO) algorithm to determine the relevant parameters. Compared with the FDT algorithm which determines parameter values empirically, the performance of the PSO-FDT has been significantly improved. Finally, the proposed PSO-FDT model was verified using test data. Experiments confirm that the proposed model is more effective than other lithology identification models. The identification accuracy for all lithologies was equal to or greater than that of the other methods. In addition, the overall accuracy was improved by at least 9.71%.
... As an alternative approach, various supervised (Karimpouli et al. 2020;Lu et al. 2020;Lee et al. 2021) and unsupervised (Taibi et al. 2019;Zhang et al. 2021) machine learning techniques were implemented for fracture identification and segmenting porous samples (Chauhan et al. 2016). Among the unsupervised approaches, encoder-decoder networks in the form of CNNs received much attention during fracture identification in the DRP workflow (Varfolomeev et al. 2019;Hong & Liu 2020;Karimpouli et al. 2020;Kim et al. 2020;Lu et al. 2020;Lee et al. 2021). ...
Image segmentation remains the most critical step in Digital Rock Physics (DRP) workflows, affecting the analysis of physical rock properties. Conventional segmentation techniques struggle with numerous image artifacts and user bias, which lead to considerable uncertainty. This study evaluates the advantages of using the random forest (RF) algorithm for the segmentation of fractured rocks. The segmentation quality is discussed and compared with two conventional image processing methods (thresholding-based and watershed algorithm) and an encoder–decoder network in the form of convolutional neural networks (CNNs). The segmented images of the RF method were used as the ground truth for CNN training. The images of two fractured rock samples are acquired by X-ray computed tomography scanning (XCT). The skeletonized 3D images are calculated, providing information about the mean mechanical aperture and roughness. The porosity, permeability, flow fields, and preferred flow paths of segmented images are analyzed by the DRP approach. Moreover, the breakthrough curves obtained from tracer injection experiments are used as ground truth to evaluate the segmentation quality of each method. The results show that the conventional methods overestimate the fracture aperture. Both machine learning approaches show promising segmentation results and handle all artifacts and complexities without any prior CT-image filtering. However, the RF implementation has superior inherent advantages over CNN. This method is resource-saving (e.g., quickly trained), does not need an extensive training dataset, and can provide the segmentation uncertainty as a measure for evaluating the segmentation quality. The considerable variation in computed rock properties highlights the importance of choosing an appropriate segmentation method.
... The automatic detection and characterization of planar geologic features in core and borehole images are bases of automatic core orientation by comparing planar geologic features. Previous studies suggested that simple threshold segmentation methods are effective only in images with obvious differences between planar geologic features and background, such as the fracture segmentation in ultrasonic and borehole optical images (fractures tend to appear as dark sinusoidal curves) [22]- [24]. To accurately segment fractures from ultrasonic images with much background noise, previous researchers introduced methods such as mathematical morphology [25] and ant colony algorithm [26], [27], which greatly improved the effects of fracture segmentation. ...
Drilling cores, the most direct and accurate first-hand data reflecting the subsurface geological information, are of unclear orientation for most boreholes. The image feature comparison method is one of the most widely used methods for core orientation. However, this method is mainly conducted by manual comparison, which is labor-intensive, time-consuming and inaccurate. In this paper, we propose an automatic core orientation method by comparing the planar geologic features (one of the most common features) in drill-core scans and microresistivity images. This method first automatically detects planar geologic features by combing the vertical gradient-based Otsu threshold segmentation, Hough transform and optimal sinusoid extraction and then determines correction difference of the core orientation by comparing the dip azimuths of the optimal sinusoids from the two types of images. This method is successfully applied to actual data. The automatic core orientation results are compared with the manual core orientation results. The results show that this method can eliminate human errors and has higher accuracy, especially for core scan images with non-distinct features. The proposed method has wide application potential in the fields of hydrocarbon exploration and development, scientific drilling research, and basic geological surveys.
... As CNNs have begun to be used in rock engineering field, several studies have been made using this technique for the study of fracture recognition, including fracture segmentation in coal image (Lu et al. 2020;Karimpouli et al. 2020), and ultrasonic well logging image (Zhang et al. 2021), and fracture detection in borehole image (Dias et al. 2020). However, few studies have been applied CNNs to rock fracture segmentation in outcrop survey, where there are various interfering objects and features. ...
Accurate recognition of rock fractures is an important problem in rock engineering because fractures greatly influence the mechanical and hydraulic properties of rock structures. However, existing image segmentation methods for identifying rock fractures tend to be limited to handling only very simple fracture images, despite many real cases containing interfering objects or features such as dark surfaces, stripes (e.g., from foliation), infilling materials, scratches, shadows, and vegetation. Here, we propose a novel deep convolutional neural network to construct the first model that is applicable in the field. After selecting U-Net, a simple and powerful network for semantic segmentation, as a baseline network, we tested network architectures by applying atrous convolutions and extra skip connections to develop an optimal network specialized for rock fracture segmentation. The rate of erroneously detecting non-fracture objects or features was reduced by using the atrous convolution module, and more skip connections were appropriately added to increase the detection rate of the actual fractures. The model's performance gradually improved as these new techniques were added to the original model. Contrast-limited adaptive histogram equalization and a fully connected conditional random field were applied before and after the network, respectively, to enhance the model’s performance. Evaluation of the proposed model using raw images of diverse site conditions shows that it can effectively distinguish rock fractures from various interfering objects and features. The source code and pre-trained model can be freely download from GitHub repository (https://github.com/Montherapy/Rock-fracture-segmentation-with-Tensorflow).
... In petroleum geology, researchers have used DNN models like DeepLabv3+ and SegNet to automatically identify fracture areas in electric imaging images and calculate parameters such as fracture dip for reservoir evaluation. This approach has effectively improved efficiency and reduced the influence of human subjective factors [22,23]. However, research on applying deep learning methods to gravel recognition remains relatively rare. ...
Gravels are widely distributed in the Baikouquan formation in the Mabei area of the Junggar Basin. However, conventional logging methods cannot quantitatively characterize gravel development, which limits the identification of lithology, structure, and sedimentary facies in this region. This study proposes a new method for automatically identifying gravels from electric imaging images and calculating gravel parameters utilizing the salient object detection (SOD) network. Firstly, a SOD network model (U2-Net) was constructed and trained using electric imaging data from the Baikouquan formation at the Mahu Sag. The blank strips in the images were filled using the U-Net convolutional neural network model. Sample sets were then prepared, and the gravel areas were labeled in the electric imaging images with the Labelme software in combination with image segmentation and human–machine interaction. These sample sets were used to train the network model, enabling the automatic recognition of gravel areas and the segmentation of adhesive gravel regions in the electric imaging images. Based on the segmented gravel results, quantitative evaluation parameters such as particle size and gravel quantity were accurately calculated. The method’s validity was confirmed through validation sets and actual data. This approach enhances adhesive area segmentation’s accuracy and processing speed while effectively reducing human error. The trained network model demonstrated an average absolute error of 0.0048 on test sets with a recognition accuracy of 83.7%. This method provides algorithmic support for the refined evaluation of glutenite reservoir logging.
Modern pavement management systems depend mainly on pavement condition assessment to plan rehabilitation strategies. Manual inspection is performed by trained inspectors to assess pavement damages conventionally. This can be cost-intensive, time-consuming, and a source of risk for inspectors. An image-based inspection using a smartphone is adopted to overcome such problems. This paper proposes an automatic crack detection and mapping program for rigid pavement, which can automate the visual inspection process. The program uses Global Navigation Satellite System (GNSS) data recorded by smartphones and various image processing techniques to detect crack lengths and areas in images. The performance of the program was evaluated by a field study. A crack quantification process was performed to compare the manually measured values and crack lengths obtained from the program. The results show that the program can detect other types of distress, such as pop-outs and punch-outs. This method can achieve satisfactory performance compared to the effort and costs spent.
Real-time monitoring of wellbore status information can effectively ensure the structural safety of the wellbore and improve the drilling efficiency. It is especially important to recognize the wellbore fractures and identify their parameters, which motivates us to propose a wellbore fracture recognition and parameter identification method using piezoelectric ultrasonic and machine learning. To realize a Self-model Emission Detection (SED), we innovatively utilize a single transducer to act as both an actuator and a sensor, allowing for the efficient acquisition of ultrasonic echo signals of the wellbore. For fracture recognition, we use the wavelet packet transform to extract features from the ultrasonic echo signal, while constructing a Convolutional Neural Network (CNN) model for fracture recognition. Then, we establish the relationships between the fracture width-depth parameter and the echo signal, including the peak value as well as the arrival time difference. The experimental results show that the proposed method effectively recognizes the fractures from the ultrasonic echo signal of the wellbore. At the same time, the established function truly reflects the relationship between the fracture parameters and the echo signal. Therefore, the proposed method can provide an identification function for quantitative monitoring of wellbore fracture parameters. Moreover, the functions can be used as a reference for other structural health monitoring, which has good application prospects.
