Ball-and-stick model is widely employed in simulating pore network within reservoirs. However, given a broad range of pore scales and diversity of structural types in reservoirs, whether this model is universally applicable remains inadequately validated. Constant-rate mercury intrusion (CRMI) is one of the key methods for studying pore-throat structures. By employing the configuration theory and analytic hierarchy process (AHP), the configurations of CRMI curves were classified, the corresponding hierarchical architectures of reservoir pore systems were interpreted, and the applicability of ball-and-stick model in reservoir pore network simulation was examined. The results indicate that CRMI curves can be divided into Configuration A and B regions, corresponding to micron-scale pores and nano-scale pores, respectively. In Configuration A region, as the mercury injection pressure increases, the mercury injection saturation in pore network, pores, and throats rises monotonically, indicating a binary pore-throat structure in the micron-scale pores, with a pore/throat ratio higher than 1. Here, the ball-and-stick model is applicable. In Configuration B region, as the mercury injection pressure increases, the mercury injection saturation in pore network and throats increases monotonically, while the mercury injection saturation in pores remains constant. This suggests that nano-scale pores have no binary pore-throat structure, and are dominated by throats, with a pore/throat ratio of 1. In this region, the ball-and-stick model is inapplicable, while a capillary tube model is more suitable. The ball-and-stick model and capillary tube model can be combined to fully simulate reservoir pore networks. The poorer the reservoir physical properties, the more applicable the capillary tube model for pore network simulation.
