Accurate estimation of permeability for reservoir simulation and production is challenging in carbonate rocks due to the diversified pore structures resulting from deposition and diagenetic modification. A significant amount of residue gas is expected in the Puguang field, China. We use a shear-frame flexibility factor from a rock-physics model as an index to quantify its spatial variation of pore structure and constrain the estimation of permeability in this field. The pore-structure index is established and used to classify various permeability-porosity trends at well locations where core and log data are available. It is found that when , the pore type is dominated by intercrystalline pores with large pore-throat sizes and high connectivity, the permeability-porosity relation is ; when , the pore type consists of isolated moldic pores, the permeability-porosity relation is , where , , , , , and are constants, which are 80, 18, 3, 0.004, 5.5, and 1.6, respectively, for the studied gas reservoir. Rock-physics-based seismic inversion is then applied to quantify the spatial variation of pore type and permeability. The inversion results indicate that regional stratigraphy has a paramount control on the distribution of pore type and permeability. Moldic pores () are widely distributed near the unconformities, whereas the large intercrystalline porosity and microporosity are distributed below and above the unconformities due to the sea-level regression and transgression, respectively. It is concluded that production problems may occur in porous and permeable intercrystalline-porosity zones and the exploration of residual gas remains in the high porosity yet less permeable moldic-porosity zones.