Natural Resources Research最新文献

High-mature organic-rich shale (HMORS) has substantial resource potential, and its reservoir heterogeneity is essential for shale gas resource evaluation and exploration. In this research, to characterize quantitatively the complex pore structure of HMORS in detail, we conducted monofractal and multifractal analyses using N₂ adsorption–desorption data from the Lower Permian (LP) HMORS in the Lower Yangtze South Yellow Sea, which is a prospective target for shale gas exploration. We also aimed to discuss the correlation, controlling factors, and application effects, to provide a new scientific analytical tool for characterizing the pore structure heterogeneity (PSH) of HMORS. The upper, middle, and lower sublayers of the LP are dominated by siliceous shale, clay shale (ClS), and clay shale and clay-mixed shale (ClS–ClMS), respectively. The monofractal dimensions D1 and D2 calculated by the Frenkel–Halsey–Hill model were not notably correlated, indicating that they are independent. The D1 of H3-type HMORS was significantly higher than its D2, while D1 and D2 of the H2 type were similar, indicating that slit-shaped pores have higher surface roughness than the internal structural complexity, whereas ink-bottle pores do not differ substantially. The monofractal study revealed that the overall PSH of HMORS is controlled primarily by calcareous minerals, and that of the ClS is also influenced by total organic carbon. The multifractal analysis revealed that the low-probability measure areas controlled the full-size pore size distribution heterogeneity of HMORS. The monofractal model can characterize ClS–ClMS with ink-bottle pores, and the multifractal model can characterize ClS with slit-shaped pores. In addition, D1 and the multifractal parameters were not significantly correlated [a_-10- a₁₀, Hurst index (H), a₀- a₁₀ and a_-10- a₀], whereas D2 correlated negatively with a₀-a₁₀, which had opposite a_-10-a₀ and H, indicating that the pore connectivity of the internal PSH of HMORS can be improved. Compared to monofractal analysis, the multifractal model has enhanced applicability in characterizing the PSH of HMORS quantitatively, which is of great significance for the study of widely developed HMORS with huge shale gas exploration potential in South China.

As the intensity and depth of coal mining grow year by year, coal seam gas pressure increases and stope structures become more complex, which can easily cause coal and gas outburst. During the process of coal and gas outburst, a large amount of coal is broken and ejected, seriously threatening the safety of workers and coal mine production. Therefore, a multifunctional coal and gas outburst physical simulation test system was used to carry out three outburst tests under different gas pressures to study the particle size distributions and fragmentation characteristics of the ejected coal. The results showed that the relative intensity of outburst increased with gas pressure, but the increase rate decreased. Gas pressure also played a role in promoting the coal crushing. For the crushing product, the R–R (Rosin–Rammler) distribution model with high COD (coefficient of determination) was used to calculate the comminution energy at 0.35 MPa, while the fractal distribution model with high COD was used at 0.85 MPa and 2.00 MPa. When gas pressure increased, the basic shape of the R–R model curve remained unchanged, the probability density curve of fractal model changed from concave to nearly straight and then to convex and the basic shape of the cumulative distribution curve of fractal model remained constant. The values of α (uniformity coefficient) and x_e (characteristic particle size) impacted on the R–R model and the values of D_f (fractal dimension) and x_max (maximum particle size) impacted on the fractal model. Within a certain error range, the comminution energy could be approximated. The comminution energy increased with gas pressure, and the potential energy of crushing product decreased with the value of the n related to the crushing mechanism. There was a strong linear relationship between relative intensity of outburst and comminution coefficient. The combination of experiments and machine learning provided a new direction for outburst prediction and prevention at coal mine sites.

Ensemble learning (EL) is a machine learning paradigm where multiple learning algorithms (base learners) are trained to solve the same problem. This study provides a comprehensive evaluation of widely used EL algorithms, including bagging, boosting, and stacking, highlighting their significant advantages in terms of accuracy and generalization of mineral prospectivity mapping (MPM). This study tested mapping of prospectivity for gold deposits in the Qingchengzi Pb–Zn–Ag–Au polymetallic district using single machine learning algorithms and EL algorithms. According to the critical and favorable geological factors for magmatic-related medium-temperature hydrothermal lode system for gold deposits, five targeting criteria were extracted from multi-source geoscience datasets (i.e., geological map, gravity and magnetic datasets, stream sediment geochemical datasets) for mineral prospectivity mapping. The receiver operating characteristic curve, the area under the curve, and learning curves were used to evaluate the performance of the tested single and ensemble machine learning algorithms. The results demonstrate that the stacking model, which combines multiple base models for hierarchical feature extraction, achieves the best predictive performance. The concentration–area fractal model was used to outline the prospective areas predicted by the EL algorithms, clarifying areas with very high prospectivity for gold mineralization in the study area.

To investigate fully the poroelastic effect on apparent permeability in coal micro/nanopores, a multi-mechanism apparent permeability model coupling the gas slippage effect and the poroelastic effect is hereby constructed on the strength of the lattice Boltzmann method. The contributions of the permeability of gas slippage, surface diffusion, and viscous flow were investigated. The results showed that the gas transport was controlled by surface diffusion in micro/nanopores with initial sizes of less than 10 nm. Under a low pore pressure, the contribution share of gas slippage permeability to the apparent gas permeability decreased exponentially as the pressure rose. When the pore pressure ascended, the dynamic apparent permeability ratio (i.e., the ratio of the apparent permeability affected by the poroelastic effect to the initial apparent permeability) was subjected to the slippage effect initially and dominated by the poroelastic effect later. Additionally, the slippage effect’s contribution to the apparent permeability ratio plunged under a lower pore pressure, but such decrease slackened as the pore pressure grew to a higher value. During coalbed methane (CBM) recovery in low-permeability coal seams, the slippage effect’s contribution to the CBM recovery production surges first, then falls slowly, and finally restores to a slow increase, and its contribution is enhanced in micro/nanopores with smaller average pore sizes.

With the continuous development of unconventional natural gas resources, the formation mechanisms of different types of gas reservoirs have become a hot topic of current research. The migration mechanisms of gas in various types of conductive media play a crucial role in studying the formation and distribution of different types of gas reservoirs. In studying natural gas migration, the pressure difference between the source and reservoir and buoyant force are generally considered the main driving forces for gas migration, while the resistance mainly comes from the capillary pressure of the reservoir. In studying capillary pressure, a circular shape is typically used as the basic model for pores or throats. The magnitude of the capillary pressure is inversely proportional to the radius of the pore or throat. However, this study conducted experiments on gas migration in circular pore models, fracture models, sandstone rock models, and pore-fracture dual models. The experimental results showed that the aspect ratio of the migration medium has an important impact on gas migration. In spaces with high aspect ratio, the gas can undergo deformation during migration, significantly reducing the capillary resistance it encounters, and under certain conditions, capillary pressure can also become a driving force for gas migration. In circular spaces, the buoyant rise of gas must satisfy the condition that connected free water can form above and below the gas column, and water can freely flow downward during the gas column's ascent. Otherwise, even if the buoyant force experienced by a continuous gas column of a certain height exceeds the capillary force of the pores, it is difficult for gas to migrate. In pores of reservoir rocks, gas often migrates in the form of bubbles, making it difficult to form a continuous gas phase, and so gas migration under buoyant force is relatively difficult. However, gas migration is easier in fractures and faults with high aspect ratio. Faults are important pathways for gas migration from deep to shallow layers, and they are also crucial for studying the correlation between shallow gas reservoirs and deep enriched gas reservoirs. This paper proposes that the aspect ratio of the migration space positively affects gas migration from the perspective of capillary pressure, improving the existing models of natural gas migration and accumulation. This is significant for understanding the formation mechanisms of different types of gas reservoirs. However, this study primarily focused on quantitative research. Further research is needed to explore the numerical relationship between the aspect ratio of pore spaces and capillary pressure, as well as the specific impacts of factors such as the density and viscosity of two-phase fluids on the experimental results and the evaluation methods of the aspect ratio of reservoir pores.

In mineral exploration, geochemical anomaly detection aims at identifying areas where geochemical properties differ from the surrounding areas, indicating possible mineralization. Robust outlier detection can help better identify potential anomalies. However, standard outlier detection techniques tend to work only in low-dimensional and Gaussian space, hence the need of a more robust outlier detection technique that can be used in the space of geochemical elements, which has high complexity and dimensionality. In this paper, a novel machine learning-based outlier detection technique is proposed. The masked autoregressive flow (MAF) was used to model the density of the high-dimensional geochemical space. Once successfully trained, the MAF provides a Gaussian space on which standard outlier detection techniques (here robust Mahalanobis distance) can be applied more successfully. The proposed method was applied to a high-quality lake sediment geochemical data acquired in Quebec, Canada, in an area with known Li–Cs–Ta (LCT) pegmatites. Results are very encouraging, with the detection of many of the known occurrences of LCT pegmatites and the discovery of potential new targets for further exploration. Hence, the method described here can be used to explore for LCT pegmatites.

In the realm of oil and gas exploration, accurate identification of lithology is imperative for the assessment of resources and the refinement of extraction strategies. While artificial intelligence techniques have garnered considerable success in lithology identification, existing methodologies encounter difficulties when addressing highly heterogeneous and geologically intricate unconventional oil and gas reservoirs. Specifically, they struggle to account for the dynamic variations in sample characteristics across spatial dimensions and temporal sequences. This separate treatment of spatial and temporal dynamics not only confines the precision of fluid prediction but also significantly attenuates the robustness of the models. To address this challenge, we propose the spatiotemporal network (STNet), a dual-branch deep learning framework that integrates seamlessly spatial feature graph methods with time-sequential prediction methods. By employing a graph structure that accounts for spatial characteristics to capture the complex spatial relationships within logging data, and by utilizing a temporal model to discern the dynamic properties of time series data, this dual-mechanism framework enables a more comprehensive understanding of the multidimensional attributes of subsurface fluids, thereby enhancing the accuracy of lithology identification. Experimental results from multiple wells in different regions of the Tarim and Daqing oilfields demonstrate that STNet not only achieves detection accuracy exceeding 95% but also exhibits strong generalizability. The results indicate a significant improvement in the accuracy of lithology identification compared to seven other advanced models. Integrating both temporal and spatial elements of logging data provides a new perspective for enhancing fluid prediction capabilities.

This paper proposes an extension of the traditional multigaussian model, where a regionalized variable measured on a continuous quantitative scale is represented as a transform of a stationary Gaussian random field. Such a model is popular in the earth and environmental sciences to address both spatial prediction and uncertainty assessment problems. The novelty of our proposal is that the transformation between the original variable and the associated Gaussian random field is not assumed to be monotonic, which offers greater versatility to the model. A step-by-step procedure is presented to infer the model parameters, based on the fitting of the marginal distribution and the indicator direct and cross-covariances of the original variable. The applicability of this procedure is illustrated with a case study related to grade control in a porphyry copper-gold deposit, where the fit of the gold grade distribution is shown to outperform the one obtained with the traditional multigaussian model based on a monotonic transformation. This translates into a better assessment of the uncertainty at unobserved locations, as proved by a split-sample validation.

The problem addressed in this work is the conditional simulation of a regionalized variable that is modeled as a realization of a non-monotonic transform of a Gaussian random field. As an alternative to Markov Chain Monte Carlo methods that often suffer from a slow convergence to the target distribution, we propose the use of sequential Monte Carlo approaches, with different variants of particle filtering. These variants are tested on synthetic and real datasets, to showcase their applicability and effectiveness under a proper setup of the importance sampling strategy, visiting sequence, number of particles, block size and kriging neighborhood used. The real case study, which deals with the simulation of gold grades in a porphyry copper-gold deposit, shows that the multi-Gaussian model based on a non-monotonic anamorphosis better assesses uncertainty than the traditional model based on a strictly monotonic anamorphosis, and that a moving neighborhood implementation of sequential Monte Carlo approaches can be successful, opening the door to applications to large-size problems in spatial uncertainty modeling.

Rockburst prediction significantly affects the development and utilization of underground resources. Currently, an increasing number of artificial intelligence algorithms are being applied for rockburst prediction. However, owing to the scarcity of data for certain rockburst grades, machine learning models have struggled to accurately train and learn their characteristics, resulting in bias or overfitting. In this study, 321 worldwide cases of rockbursts were collected. Seven indices considering both rock mechanics and stress conditions were selected as input parameters for the model. To address the issue of limited data for certain rockburst grades, the Synthetic Minority Over-sampling TEchnique (SMOTE) algorithm was used for comprehensive oversampling and synthesis of the rockburst data. The theoretical rationality of this method was corroborated by the Spearman’s correlation coefficient. Additionally, the model hyperparameters were optimized using the Bayesian optimization method, and an improved eXtreme gradient boosting (XGBoost) rockburst prediction model (SM–BO–XGBoost) was established. The constructed SM–BO–XGBoost model was compared with decision tree, random forest, support vector machine, and k-nearest neighbor classification machine learning models. The results showed a significant improvement in the prediction accuracy for the None and Strong rockburst categories, which had limited data in the original rockburst dataset. To address the poor interpretability of the XGBoost model, the SHapley Additive exPlanations (SHAP) method was introduced to explain the constructed model, and to analyze the marginal contributions of different features to the model output across various rockburst grades. The SM-BO-XGBoost model was validated using field rockburst records from the Xincheng and Sanshandao gold mines. As indicated by the results, the model demonstrated favorable performance and applicability, with wide potential for predicting engineering rockbursts.