Integrated Porosity Classification and Quantification Scheme for Enhanced Carbonate Reservoir Quality: Implications from the Miocene Malaysian Carbonates

The pore system in carbonates is complicated because of the associated biological and chemical activity. Secondary porosity, on the other hand, is the result of chemical reactions that occur during diagenetic processes. A thorough understanding of the carbonate pore system is essential to hydrocarbon prospecting. Porosity classification schemes are currently limited to accurately forecast the petrophysical parameters of different reservoirs with various origins and depositional environments. Although rock classification offers a way to describe lithofacies, it has no impact on the application of the poro-perm correlation. An outstanding example of pore complexity (both in terms of type and origin) may be found in the Central Luconia carbonate system (Malaysia), which has been altered by diagenetic processes. Using transmitted light microscopy, 32 high-resolution pictures were collected of each thin segment for quantitative examination. An FESEM picture and a petrographic study of thin sections were used to quantify the grains, matrix, cement, and macroporosity (pore types). Microporosity was determined by subtracting macroporosity from total porosity using a point-counting technique. Moldic porosity (macroporosity) was shown to be the predominant type of porosity in thin sections, whereas microporosity seems to account for 40 to 50% of the overall porosity. Carbonates from the Miocene have been shown to possess a substantial quantity of microporosity, making hydrocarbon estimate and production much more difficult. It might lead to a higher level of uncertainty in the estimation of hydrocarbon reserves if ignored. Existing porosity classifications cannot be used to better understand the poro-perm correlation because of the wide range of geological characteristics. However, by considering pore types and pore structures, which may be separated into macro- and microporosity, the classification can be enhanced. Microporosity identification and classification investigations have become a key problem in limestone reservoirs across the globe.

Contribution of the ALBION Dynamic Analogue in Understanding the Diversity of Fluid Flows in Fractured Carbonate Reservoirs. The Example of the LSBB Instrumented Site

Carbonate reservoirs exhibit an extreme geological heterogeneity inducing a great diversity of fluids flows. Grasping the plurality of flows and the corresponding geological features require data scarcely available from subsurface hydrocarbons fields and even rarely acquired together on outcrop analogues. Among the different sites of the ALBION R&D project, the LSBB underground research laboratory provides outstanding access to both fractured limestone and groundwater dynamics through several experimental areas, including a 3.8 km long tunnel, which penetrates the Barremian-Aptian Urgonian formation to a maximum depth of 519 m. This paper gives an overview of the data acquired and the different works carried out on the LSBB site. From this synthesis, it draws lessons on the characterization of outcrop analogues and some insights for the modeling of fractured carbonate reservoirs. The quantity and diversity of the data acquired on the LSBB site allow: (i) the construction of nested multi-scale geological models, (ii) the comparison of measurements of different physical properties to better characterize the reservoir properties of the fractured rock, (iii) a multi-scale and multi-support approach to heterogeneity. Defining a common geological framework (facies model, rock type classification, inventory of structural objects, etc.) appears to be an essential step, possibly iterative, for the coupled interpretation of the various acquisitions and the extrapolation of results. Building a common geological model as a framework for interpretation help cross-fertilisation between geoscience domains. However, despite the huge amount of data, performing relevant and parsimonious rock typing remains a delicate exercise. This reminds us of the great uncertainties that can exist in establishing rules and concepts from limited data sets, such as those classically available for operational studies. Beyond the characterization of the depositional environment, the observations emphasize the importance of understanding the structural and diagenetic history, which leads to different rock types and current reservoir properties, to successfully define such a rock classification. Furthermore, the organization of flow paths within the fractured medium and its evolution over geologic time condition the processes of diagenesis and karstification. Hydrological processes and history must therefore be taken into account in this genetic reconstruction.

Integration of 3D Volumetric Computed Tomography Scan Image Data with Conventional Well Logs for Detection of Petrophysical Rock Classes

Summary Core measurements are used for rock classification and improved formation evaluation in both cored and noncored wells. However, the acquisition of such measurements is time-consuming, delaying rock classification efforts for weeks or months after core retrieval. On the other hand, well-log-based rock classification fails to account for rapid spatial variation of rock fabric encountered in heterogeneous and anisotropic formations due to the vertical resolution of conventional well logs. Interpretation of computed tomography (CT) scan data has been identified as an attractive and high-resolution alternative for enhancing rock texture detection, classification, and formation evaluation. Acquisition of CT scan data is accomplished shortly after core retrieval, providing high-resolution data for use in petrophysical workflows in relatively short periods of time. Typically, CT scan data are used as two-dimensional (2D) cross-sectional images, which is not suitable for quantification of three-dimensional (3D) rock fabric variation, which can increase the uncertainty in rock classification using image-based rock-fabric-related features. The methods documented in this paper aim to quantify rock-fabric-related features from whole-core 3D CT scan image stacks and slabbed whole-core photos using image analysis techniques. These quantitative features are integrated with conventional well logs and routine core analysis (RCA) data for fast and accurate detection of petrophysical rock classes. The detected rock classes are then used for improved formation evaluation. To achieve the objectives, we conducted a conventional formation evaluation. Then, we developed a workflow for preprocessing of whole-core 3D CT-scan image stacks and slabbed whole-core photos. Subsequently, we used image analysis techniques and tailor-made algorithms for the extraction of image-based rock-fabric-related features. Then, we used the image-based rock-fabric-related features for image-based rock classification. We used the detected rock classes for the development of class-based rock physics models to improve permeability estimates. Finally, we compared the detected image-based rock classes against other rock classification techniques and against image-based rock classes derived using 2D CT scan images. We applied the proposed workflow to a data set from a siliciclastic sequence with rapid spatial variations in rock fabric and pore structure. We compared the results against expert-derived lithofacies, conventional rock classification techniques, and rock classes derived using 2D CT scan images. The use of whole-core 3D CT scan image-stacks-based rock-fabric-related features accurately captured changes in the rock properties within the evaluated depth interval. Image-based rock classes derived by integration of whole-core 3D CT scan image-stacks-based and slabbed whole-core photos-based rock-fabric-related features agreed with expert-derived lithofacies. Furthermore, the use of the image-based rock classes in the formation evaluation of the evaluated depth intervals improved estimates of petrophysical properties such as permeability compared to conventional formation-based permeability estimates. A unique contribution of the proposed workflow compared to the previously documented rock classification methods is the derivation of quantitative features from whole-core 3D CT scan image stacks, which are conventionally used qualitatively. Furthermore, image-based rock-fabric-related features extracted from whole-core 3D CT scan image stacks can be used as a tool for quick assessment of recovered whole core for tasks such as locating best zones for extraction of core plugs for core analysis and flagging depth intervals showing abnormal well-log responses.

Mineral Rock Classification Using Convolutional Neural Network

The main aim of the research is to build a model that can effectively predict the type of mineral rocks. Rocks can be predicted by observing it is colour, shape and chemical composition. On-site technicians need to apply different techniques on rock sample in order to predict rock type. Technicians need to apply different techniques on rock samples, so it is a time-consuming process, and sometimes the predictions may be accurate, and sometimes predictions may be false. When predictions are false, it might show a negative impact in several ways for workers and organization as well. We considered an image dataset of rock types, namely Biotite, Bornite, Chrysocolla, Malachite, Muscovite, Pyrite, and Quartz. We applied CNN (Convolutional Neural Network) Algorithm to get a better prediction of different mineral rocks. Nowadays, CNN is mainly used for image classification and image recognition tasks.

Intelligent rating method of tunnel surrounding rock based on one-dimensional convolutional neural network

Accurate prediction of surrounding rock grades holds great significance to tunnel construction. This paper proposed an intelligent classification method for surrounding rocks based on one-dimensional convolutional neural networks (1D CNNs). Six indicators collected in some tunnel construction sites are considered, and the degree of linear correlation between these indicators has been analyzed. The improved one-hot encoding method is put forward for transforming these non-image indicators into one-dimensional structural data and avoiding the sampling error in the indicators of surrounding rock collected in the field. We found that the 1D CNNs model based on the improved one-hot encoding method can best extract the features of surrounding rock classification indicators (in terms of both accuracy and efficiency). We applied the well-trained classification model of tunnel surrounding rock to a series of expressway tunnels in China, and the results show that our model could accurately predict the surrounding rock grade and has great application value in the construction of tunnel engineering. It provides a new research idea for the prediction of surrounding rock grades in tunnel engineering.

Automated Classification of High-resolution Rock Image Based on Residual Neural Network

The identification and classification of high-resolution rock images are significant for oil and gas exploration. In recent years, deep learning has been applied in various fields and achieved satisfactory results. This paper presents a rock classification method based on deep learning. Firstly, the high-resolution rock images are randomly divided into several small images as a training set. According to the characteristics of the datasets, the ResNet (Residual Neural Network) is optimized and trained. The local images obtained by random segmentation are predicted by using the model obtained by training. Finally, all probability values corresponding to each category of the local image are combined for statistics and voting. The maximum probability value and the corresponding category are taken as the final classification result of the classified image. Experimental results show that the classification accuracy of this method is 99.6%, which proves the algorithm’s effectiveness in high-resolution rock images classification.

Optimization Analysis of Excavation Procedure Design of Underground Powerhouses under High In Situ Stress in China

To recommend the excavation procedures and design parameters for underground powerhouses, excavation procedures of fifty-one underground powerhouses in China were summarized and analyzed based on in situ stress conditions. Firstly, the complex stress environment in China was introduced and fifty-one underground powerhouses with their engineering scale, size, lithology, rock classification and in situ stress level were listed in detail. Subsequently, to evaluate the influence of in situ stress levels on excavation procedure design, the correlation between excavation procedures and in situ stress level in three main excavation zones were analyzed accordingly. Moreover, to provide the excavation design recommendations, the strength–stress ratio (SSR) was promoted to analyze and recommend the design parameters, and the blasting excavation design based on the stress transient unloading control was also supplemented. The results show that excavation procedures have different priorities under different in situ stress levels, and the design parameters show an obvious relationship with in situ stress levels. Moreover, the excavation procedure parameters are suggested to adjust accordingly under different SSR. The discussion of influencing factors and specification ensures its rationality and accuracy. It is believed that the summary and recommendations can provide a good reference for excavation procedure optimization of underground powerhouse under high in situ stress.

Research on Classification of Fine-Grained Rock Images Based on Deep Learning

Rock classification is a significant branch of geology which can help understand the formation and evolution of the planet, search for mineral resources, and so on. In traditional methods, rock classification is usually done based on the experience of a professional. However, this method has problems such as low efficiency and susceptibility to subjective factors. Therefore, it is of great significance to establish a simple, fast, and accurate rock classification model. This paper proposes a fine-grained image classification network combining image cutting method and SBV algorithm to improve the classification performance of a small number of fine-grained rock samples. The method uses image cutting to achieve data augmentation without adding additional datasets and uses image block voting scoring to obtain richer complementary information, thereby improving the accuracy of image classification. The classification accuracy of 32 images is 75%, 68.75%, and 75%. The results show that the method proposed in this paper has a significant improvement in the accuracy of image classification, which is 34.375%, 18.75%, and 43.75% higher than that of the original algorithm. It verifies the effectiveness of the algorithm in this paper and at the same time proves that deep learning has great application value in the field of geology.