Permeability evolution in hydrate-bearing porous media is a key factor controlling gas production efficiency during natural gas hydrate exploitation. In this study, laboratory experiments were conducted using sand-packed tubes filled with quartz sand and glass beads to systematically investigate the variation of liquid-phase permeability with hydrate saturation. The effects of pore structure, particle size, and initial gas injection pressure on hydrate formation and permeability reduction were analyzed. Furthermore, experimental results were compared with four commonly used permeability models, including the Kozeny model, the Dai model, the Masuda model, and the parallel capillary model. The results show that permeability decreases continuously with increasing hydrate saturation in both porous media, and the most rapid decline occurs at low saturation levels between 0 and 9%. Under the same conditions of 20–40 mesh and an initial pressure of 6.0 MPa, the pressure drop rate in the quartz-sand-packed tube reaches 1.062 kPa per minute, which is about 2.35 times higher than the 0.451 kPa per minute observed in the glass-bead-packed tube, indicating a faster hydrate formation rate and stronger permeability reduction in quartz sand. In addition, both increasing particle mesh size and raising the initial gas injection pressure significantly promote methane consumption and hydrate formation. Model comparison results demonstrate that permeability reduction is strongly dependent on pore structure. The Kozeny pore-filling model, the Dai model (M = 3), and the Masuda model (N = 8) show good agreement with the glass-bead data, whereas the Dai model (M = 8), the Masuda model (N = 15), and the pore-center form of the parallel capillary model better describe the quartz-sand system. In contrast, models based on particle-surface coating show poor agreement in both media. These findings indicate that permeability reduction is primarily controlled by pore-space occupation and flow-path restriction rather than uniform surface coverage. The results suggest that hydrate growth is more likely to occur in pore centers and critical pore-throat regions, although this conclusion is based on macroscopic model comparison and requires further validation by pore-scale observations. This study provides a quantitative basis for model selection and improves the understanding of permeability evolution in hydrate-bearing porous media.
Yang et al. (Fri,) studied this question.