Geological factors are key elements to control shale gas enrichment and influence the accurate estimation of shale gas reserve. However, the impact of the main geological factors, such as porosity, mineralogy, and organic matter, on marine shale gas enrichment and reserve calculation has not yet been fully clarified. Herein, we measured gas adsorption, porosity, mineralogical composition, and total organic carbon content of the marine shale samples from the Jiaoshiba area of Fuling gas field in Sichuan Basin, South China, and investigate the relationships between the geological factors and the adsorbed gas content. The results show that adsorbed gas content is positively correlated with total organic carbon and porosity; the larger specific surface area of samples with more clay minerals essentially contributes to shale gas enrichment. Additionally, the sealing of faults imposes a significant impact on shale gas accumulation. The probability volume method was applied to calculate the shale gas reserve. The reserves of P₉₀ (the most pessimistic reserve), P₅₀ (the most likely reserve), and P₁₀ (the most optimistic reserve) were calculated, respectively, which provides useful information to reduce the risk in shale gas development.

Multiple-point geostatistics (MPS) has more advantages than two-point geostatistics in reproducing the continuity of geobodies in subsurface reservoir modeling. For fluvial reservoir modeling, the more continuous a channel, the more consistent it is with geological knowledge in general, and fluvial continuity is also of paramount importance when simulating fluid flow. Based on the pixel-based MPS algorithm Snesim, this study proposes a method that utilizes multiple search trees (MSTs) to enhance simulation continuity in 2D fluvial reservoir modeling. The objective of the MST method is to capture complete data events from a training image (TI), which aims to achieve enhanced continuity in fluvial reservoir sublayer modeling. By resorting to search neighborhoods based on their proximity to the central node of the data template, multiple data templates that correspond to the MSTs will be generated. Here, four data templates were generated by arranging the relative search neighborhood coordinates in ascending and descending order with respect to the central node. Parallel computing was tried for the construction of the search trees. This work calculated the conditional probability distribution function (CPDF) of the simulating nodes by averaging the CPDFs derived from the MSTs, and double retrieval was employed to filter out the search trees that possessed an inaccurate local CPDF for the simulating nodes. In addition, the connected component labeling (CCL) method was introduced to evaluate the simulation continuity in MPS. The results indicated that the MST method can enhance the simulation continuity of the Snesim algorithm by reproducing the fine connectivity of channel facies in 2D fluvial reservoir modeling.

The variogram function is a crucial tool for quantifying spatial correlation and a key component of Kriging interpolation, directly influencing the accuracy of interpolation results. However, traditional variogram fitting methods, such as least squares and weighted polynomial regression, often face limitations when handling complex or non-stationary spatial structures, which can lead to unstable fitting outcomes. To address these challenges, this paper proposes an automatic fitting method for experimental variogram functions based on deep learning. The variogram fitting process is inherently a nonlinear optimization problem, where the goal is to optimize the alignment between experimental and theoretical variogram functions. In this study, we construct a deep learning model designed to capture the nonlinear characteristics of experimental variograms with the objective of minimizing fitting errors and improving both the accuracy and stability of the fitting process. Extensive experimental results demonstrate that the proposed neural network model can effectively learn to fit a variety of experimental variograms and accurately represent different theoretical variogram functions. Compared to traditional methods, this approach consistently achieves superior fitting performance across multiple datasets, validating its effectiveness and robustness under diverse conditions. The stronger generalization capability and improved accuracy of the proposed method significantly enhance the automatic fitting of experimental variogram functions.

Training image (TI) is important for multipoint statistics simulation method (MPS), since it captures the spatial geological pattern of target reservoir to be modeled. Generally, one optimal TI is selected before applying MPS by evaluating the similarities between many TIs and the well interpretations of target reservoir. In this paper, we propose a new training image optimization approach based on the convolutional neural network (CNN). First, candidate TIs were randomly sampled several times to obtain the sample dataset. Then, the CNN was used to conduct transfer learning for all samples, and finally, the optimal TI of the conditioning well data is selected through the trained CNN model. By taking advantage of the strong learning ability of CNN in image feature recognition, the proposed method can automatically identify differences in spatial features between the conditioning well data and the samples of the training image. Hence, it effectively resolves the difficulty of spatial matching between discrete datapoints and grid structures. We demonstrated the applicability of our model via 2D and 3D training image selection examples. The proposed methods effectively selected the appropriate TI, and then the pretreatment techniques for improving the accuracy of continuous TI selection were achieved. Moreover, the proposed method was successfully applied to training image selection of a discrete fracture network model. Finally, sensitivity analysis was carried out to show that sufficient conditioning data volume can reduce the uncertainty of the optimization results. By comparing with the improved MDevD method, the advantages of the new method are verified in terms of efficiency and reliability.

Strong heterogeneity, complicated lithology, and chaotic seismic reflection characteristics are all common features of glutenite reservoirs. It is challenging to pinpoint the interior lithology and quantitatively describe the heterogeneity of the single-stage glutenite. To explore the distribution and superimposition features of subaqueous fans, the upper Es4 in the Y229 region of Dongying sag was used as an example. We have developed a multitrend fusion constraint modeling method based on deposition patterns. First, the truncated Gaussian simulation method is used to establish the sedimentary subfacies model of the subaqueous fan. Second, one lithologic probability volume is generated according to the proportion of various lithology of different sedimentary subfacies. Then, the lobes model is constructed using the object-based simulation method and the quantitative parameters from outcrop and flume sedimentation simulations. In addition, a different lithology probability volume is determined based on how far a particular lobe is from the centerline. The two probability volumes are combined to create the integrated lithology probability volume, which reflects the planar trend and the internal differences of different lobes. To constrain lithologic modeling, the integrated lithology probability volume is used. By comparing the model results with a single-trend constraint, the findings indicate that the multitrend integration constraint modeling method may more accurately depict the internal variability of the glutenite reservoir. In addition, the lithologic model built on this foundation is consistent with the depositional model.