In recent years, the greenhouse gas control has been a hot issue. A solution of injecting supercritical CO 2 (scCO 2 ) in Longmaxi shale gas formation (in Southwest China) is proposed in this study. To evaluate the feasibility of this method that storing CO 2 and enhanced gas recovery (EGR) simultaneously, we first carried out isothermal adsorption experiments on samples from Longmaxi formation to describe the adsorption behavior of CH 4 and CO 2 , and a generalized Ono-Kondo Lattice (OK) model was applied to predict adsorption amount of pure CH 4 /CO 2 and their binary mixtures under supercritical, high-pressure conditions. In addition, discrete fracture network (DFN) model was adopted to characterize the complex hydraulic fracture networks constructed from micro-seismic monitoring (MSM) data and engineering analysis. The porous flow in shale matrix was modeled with the multiple interacting continua (MINC) and fractal theories. A multiscale compositional numerical model based on unstructured tri-prism grids was finally developed and solved by control volume finite element (CVFE) method. The simulation results of base and CO 2 injection cases presented that scCO 2 huff ‘n’ puff in Longmaxi formation might be an effective method for CO 2 storage and EGR, and the orthogonal experiments were designed to optimize scenarios in the field application and obtain a maximum balanced result of EUR and CO 2 storage capacity (CSC) for multistage fractured horizontal well (MFHW) with complex fracture networks in shale. • The OK model was applied to predict pure CH 4 /CO 2 and their mixtures adsorption amount under supercritical conditions. • The hybrid DFN and MINC model was established to describe the flow in fractal porous media and complex fracture system. • The CVFE simulator was successfully conducted for comprehensive analysis of mechanisms of scCO 2 storage and EGR in shale.