Permeability is a key factor influencing the recovery of gas resources from gas hydrate reservoirs. During the production process, reservoir permeability increases as the hydrate saturation decreases, which causes difficulties in permeability prediction. However, predictions using classical permeability models do not provide a good match to most of the experimental data. This study proposes a new Cubic model to calculate the permeability increase induced by hydrate dissociation. A series of permeability experiments with variable hydrate saturations were conducted to validate the provide models. Results show that most experimental permeability data points fall between two fitted line segments of the provided models: i.e. the Cubic pore wall coating and Cubic pore center occupying models. These results also indicate that the behaviors of pore wall coating and pore center occupying hydrates coexist in the process of hydrate synthesis and thus a hybrid growth pattern provides better explain for the experimental and simulated results in this study. Additionally, different hydrate growth patterns in the experimental samples reveal that in the reservoirs with low porosity and low permeability, the coating hydrates appear to be more common than center occupying hydrates. The proposed model can be applied to predicting permeability change during gas production, and it also helps understand microscopic hydrate growth mechanisms. • Proposed a Cubic model to estimate permeability change during hydrate dissociation/synthesis. • The proposed Cubic model was validated by experimental data from two artificial sandstones. • Hydrates hybrid grow pattern including pore center occupying and pore wall coating was found in sandstones. • In low permeability, the coating hydrates appear to be more common than center occupying hydrates. [ABSTRACT FROM AUTHOR]