The deep saline aquifer and depleted gas reservoirs are suitable candidates for CO2 storage. The estimation of residual gas and corresponding flowing behavior are major parameters for successful CO2 trapping in geological formations. In this article the residual gas trapping phenomenon in pore scale has been studied using Micro-CT images. In macro scale, the results of imbibition experiments on core plug samples have been considered to determine residual gas saturation (RGS) and the corresponding water relative permeability. In the field scale, pressure transient data in water leg has been used to determine the water relative permeability at RGS in a tight carbonate reservoir in Middle East. In the proposed workflow, an artificial intelligence approach is applied to determine absolute permeability profile. Then, a numerical dynamic single well model is built to calculate the water relative permeability. The comparison of core and pressure transient data shows that the absolute permeability and water relative permeability from core imbibition test are significantly overestimated while the results of gas-water simulation in pore scale confirmed the real field transient data. The proposed method can be used by industrial reservoir simulator to define rock/fluid interaction parameters in simulation of CO2 storage in depleted underground gas reservoirs.