Pool scrubbing is an important mitigation measure for reducing radioactive aerosol release under nuclear reactor accident conditions. Its efficiency of aerosol removal is governed by the coupled dynamics of rising bubbles and aerosol transport within the bubble. Although computational fluid dynamics (CFD) offers a powerful framework for studying these interacting processes, detailed CFD investigations of aerosol removal in pool scrubbing are still limited. In this work, aerosol scrubbing in a rectangular bubble column is numerically investigated using an Euler–Euler multi-fluid approach, with particular emphasis on interphase momentum transfer and particle removal modelling. Drag is the dominant interfacial force and strongly influences bubble rise velocity, gas residence time, and aerosol removal efficiency. The results show that the effective drag coefficient is significantly reduced under pool scrubbing conditions. Classical drag models, such as the Ishii–Zuber and Tomiyama models, which perform well for pipe flows and bubble columns, underpredict bubble rise velocities in pool scrubbing columns. This behaviour is attributed to their inability to capture the strong turbulence generated by high gas injection velocities and large bubbles’ wake. Based on recent findings by Salibindla et al. [1], improved predictions of bubble rise velocity and void fraction distributions are achieved. Aerosol particles in the gas and liquid phases are treated as separate species, with particle removal modelled as mass transfer from gas to liquid. Four literature models (Fuchs [2], Kunsek [3], Powers [4], and Fujiwara [5]) are evaluated for different gas injection rates and particle sizes. The results reveal a strong particle-size dependence. To address existing limitations, a particle-size-dependent correction is introduced into the centrifugal impaction term of the Powers model, resulting in significantly improved removal predictions across all cases.