Abstract
Macroscopic convection is severely limited in low‐permeability reservoirs, and aqueous‐phase chemical reactions are frequently neglected in existing CO
2
miscible flooding models. To address these gaps, this study develops a diffusion–reaction coupled model that simultaneously tracks five distinguishable components—CO
2
, oil, miscible mixture, carbonic acid, and water—and couples the reversible CO
2
–water–carbonic acid reaction with multiphase mass transfer. A water‐saturation‐dependent dynamic diffusion coefficient is introduced to characterize spatial variations in effective diffusion capacity. Numerical results show that all concentration fronts advance at a scaling, confirming diffusion‐dominated transport consistent with analytical solutions and pore‐scale simulations. The dynamic diffusion coefficient improves the continuity and spatial stability of the miscible front and broadens the effective oil mobilization range. The reversible CO
2
–water–carbonic acid reaction buffers local concentration gradients, producing a cooperative “diffusion–reaction buffering–rediffusion” migration pattern; it acts as an internal regulator rather than replacing diffusion as the dominant mechanism. Validation against analytical solutions, literature trends, and mass conservation confirms the model’s physical consistency. This framework offers a theoretical basis for mechanistic studies and parameter optimization of CO
2
miscible flooding in low‐permeability and tight reservoirs.