The subsurface storage of CO₂ in damaged reservoirs presents unique technical challenges, demanding the development of new mathematical models. A key issue is the velocity-dependent relative permeability observed during CO₂-water displacement, a result of the shifting balance between viscous, capillary, and gravitational forces. Since fluid velocity changes substantially from the near-wellbore to the far-field, these dynamic permeability functions are critical for accurately predicting reservoir and well behavior, particularly where formation damage is a controlling factor in flow. Analytical models offer a streamlined means of forecasting reservoir behaviour in CO₂ geosequestration, though they typically rely on simplifying assumptions. In contrast, numerical modelling captures detailed spatial and temporal variations in fluid properties but demands extensive computational resources and large datasets. This study adopts a hybrid methodology: numerical simulations in Petrel are employed to characterize the velocity-dependent behaviour of relative permeabilities, while a newly developed analytical model estimates sweep efficiency and injectivity at the reservoir scale. The analytical framework extends the classical Buckley-Leverett CO₂–water displacement theory by incorporating spatially variable fractional flow. This study examines damaged heterogeneous reservoirs initially saturated with brine. The proposed hybrid model introduces two key innovations: (i) velocity-dependent relative permeability functions that capture flow regimes ranging from viscous to gravity-dominated, and (ii) an upscaling solution for CO₂–water displacement processes. The sensitivity analysis demonstrates how heterogeneity, formation damage, and the density of the injected fluid influence the vertical distribution of CO2 within the reservoir. These factors play a crucial role in determining the solubility and residual trapping of carbon dioxide during geosequestration.