Fractal dimension is a critical parameter to evaluate the heterogeneity of complex pore structure in tight sandstone gas and other low permeability reservoirs. To quantify the fractal dimension of tight sandstone at various pore size classes and evaluate their implications on mineral composition and nano pore structure parameters, we conducted an integrated approach of N₂ adsorption/desorption experiment (N₂-GA), X-ray diffraction (X-RD), and field emission scanning electron microscopy (FE-SEM) on Sulige tight sandstone reservoirs. By comparing the nine types of fractal dimensions calculated from N₂ adsorption data, we put forward the concept of “concentrated” fractal dimensions and “scattered” fractal dimensions (D_N2, D_N3, D_N5, D_N7 and D_N8) for the first time according to its concentration extent of distribute in different samples. Result shows that mineral composition has a significant influence of a different level on specific surface area (SSA), pore volume (PV), and fractal dimensions (D_N), respectively, where the “scattered” fractal dimension is more sensitive to certain specific property of the reservoir, including mineral content and the specific surface area contribution rate (S_r) of type II mesopores (Mesopore-II: 10~50nm). In addition, three type of hysteresis loops were distinguished corresponding to different pore shape combination of N₂-GA isotherm curve, which reveals that pore structure heterogeneity is mainly controlled by inkbottle-shaped pores and the volume contribution rate (V_r) of mesopores in this study area. These findings could contribute to a better understanding of the controlling effect of pore heterogeneity on natural gas storage and adsorption.

As increasingly large‐scale and higher‐resolution terrain data have become available, for example air‐form and space‐borne sensors, the volume of these datasets reveals scalability problems with existing GIS algorithms. To address this problem, a kind of serial algorithm was developed to generate viewshed on large grid‐based digital elevation models (DEMs). We first divided the whole DEM into rectangular blocks in row and column directions (called block partitioning), then processed these blocks with four axes followed by four sectors sequentially. When processing the particular block, we adopted the ‘reference plane’ algorithm to calculate the visibility of the target point on the block, and adjusted the calculation sequence according to the different spatial relationships between the block and the viewpoint since the viewpoint is not always inside the DEM. By adopting the ‘Reference Plane’ algorithm and using a block partitioning method to segment and load the DEM dynamically, it is possible to generate viewshed efficiently in PC‐based environments. Experiments showed that the divided block should be dynamically loaded whole into computer main memory when partitioning, and the suggested approach retains the accuracy of the reference plane algorithm and has near linear compute complexity.

As the reservoir and its attribute distribution are obviously controlled by sedimentary facies, the facies modeling is one of the important bases for delineating the area of high-quality reservoir and characterizing the attribute parameter distribution. There are a large number of continental sedimentary reservoirs with strong heterogeneity in China, the geometry and distribution of various sedimentary microfacies are relatively complex. The traditional geostatistics methods which have shortage in characterization of the complex and non-stationary geological patterns, have limitation in facies modeling of continental sedimentary reservoirs. The generative adversarial network (GANs) is a recent state-of-the-art deep learning method, which has capabilities of pattern learning and generation, and is widely used in the domain of image generation. Because of the similarity in content and structure between facies models and specific images (such as fluvial facies and the images of modern rivers), and the various images generated by GANs are often more complex than reservoir facies models, GANs has potential to be used in reservoir facies modeling. Therefore, this paper proposes a reservoir facies modeling method based on GANs: (1) for unconditional modeling, select training images (TIs) based on priori geological knowledge, and use GANs to learn priori geological patterns in TIs, then generate the reservoir facies model by GANs; (2) for conditional modeling, a training method of “unconditional-conditional simulation cooperation” (UCSC) is used to realize the constraint of hard data while learning the priori geological patterns. Testing the method using both synthetic data and actual data from oil field, the results meet perfectly the priori geological patterns and honor the well point hard data, and show that this method can overcome the limitation that traditional geostatistics are difficult to deal with the complex non-stationary patterns and improve the conditional constraint effect of GANs based methods. Given its good performance in facies modeling, the method has a good prospect in practical application.

Three-dimensional (3D) geological property modeling is used to quantitatively characterize various geological attributes in 3D space based on geostatistics with the help of computer visualization technology, and the results are often stored in grid data. The 3D geological property modeling includes two main components, grid model generation and property interpolation. In this review article, the existing grid generation methods are systematically investigated, and both traditional and multiple-point geostatistical algorithms involved in interpolation methods are comprehensively analyzed. It is shown that considering the numerical simulation of oil reservoirs, the orthogonal hexahedral grid remains the most suitable grid model for simulations in petroleum exploration and development. For the interpolation methods aspect, most geological phenomena are nonstationary, to simulate various types of reservoirs; the main development trends are increasing geological constraints and reducing the limitation of stationarity. Both methods have certain constraints, and the multiscale problem of multiple-point geostatistics poses a main challenge to the field. In addition, the deep-learning based method is a new trend in geological property modeling.