“…As an alternative to experiments, some researchers rely on fitting empirical relative permeability models (Joseph et al, ) to pore network modeling results (Kang et al, ; Katagiri et al, ; Mahabadi & Jang, ; Mahabadi, Dai, et al, , Mahabadi, Zheng et al, ; Wang et al, ), which, although useful, is limited by the lack of mechanistic models to describe petrophysical properties in addition to the computational expense required in developing a new pore network model for each scenario. The empirical relative permeability models used in pore network modeling, particle‐based 3‐D packs (Katagiri et al, , ), or otherwise (Yoneda et al, ) essentially involve fitting relative permeability curve separately for each hydrate saturation, which gives different model parameters for each hydrate saturation. The major drawback of these empirical models, besides several other limitations as discussed in Singh et al (), is that they rely on fitted parameters for relative permeability and capillary pressure functions at studied gas hydrate saturation to predict relative permeability and capillary pressure at other gas hydrate saturations.…”