A bubble-population-balance foam-modeling technique is developed to investigate how carbon dioxide (CO2) foam behaves rheologically and propagates in a field-scale radial system. The modeling technique is based on pore-scale events and honors three different foam states (weak, strong, and intermediate) and two steady-state strong-foam-flow regimes (high- and low-quality) measured in corescale experiments. The model parameters are first obtained from a fit to laboratory-coreflood experimental data, and then the mechanistic model is applied to different types of CO2 foams, ranging from gaseous to supercritical-CO2 foams, represented by various mobilization pressure gradients. The results from the fit to existing coreflood data show that a reasonable match can be made satisfying multiple constraints, such as hysteresis exerted by three foam states, non-Newtonian flow behavior caused by gas trapping and shear-thinning rheology, and bubble stability in different capillary pressure environments. When applied to field-scale scenarios, supercritical-CO2 foams requiring low mobilization pressure gradients propagate much farther than gaseous-CO2 foams, far enough to make use of promising supercritical-CO2 foams in the field. This study, for the first time, theoretically demonstrates why supercritical-CO2 foams should be preferred in the field compared with gaseous N-2 or CO2 foams. The companion paper to extend this study to full-field-scale foam propagation in conjunction with gravity segregation is Izadi and Kam (2018).