The hydrodynamic modelling of granular anaerobic digestion reactors, such as the Upflow Anaerobic Sludge Blanket (UASB) reactor, is strongly influenced by the complex interactions between gas and liquid phases. These reactors are often considered self-mixed because of the impact of gas production on the reactor mixing. However, accurately capturing this gas-liquid flow behaviour in computational fluid dynamics (CFD) models remains challenging [1]. This study aims to evaluate the sensitivity of gas–liquid Eulerian–Eulerian CFD simulations to key modelling choices and numerical parameters, with a particular focus on interfacial force formulations. An air–water system is employed as a hydrodynamic surrogate to isolate flow phenomena relevant to UASB reactor design. Simulation results are compared against experimental measurements of water and air velocities obtained under identical operating conditions using Particle Image Velocimetry (PIV) and shadowgraphy techniques. The flow in the reactor was characterized as a homogeneous dispersed-bubbly regime, with spherical, non-breaking bubbles. Ansys Fluent R25.2 is used for the simulations. A mesh sensitivity study is first performed using a baseline model. The influence of two different drag force correlations (Grace et al. and Schiller–Naumann) is studied [2]. In addition to drag forces, lift and wall lubrication effects are also investigated to account for bubble distribution and near-wall hydrodynamic interactions. The wall lubrication effects are evaluated through a comparative assessment of different formulations, using the Antal et al. model as a reference and comparing it against the Lubchenko model with Shaver–Podowski correction [3]. Initial results indicate that the baseline model reproduces key experimental trends. The final analysis aims to identify the most suitable modelling strategies for gas–liquid interactions in UASB reactors operating in the homogeneous bubbly flow regime. Furthermore, the mass-source modelling approach will also be verified, with the goal of being implemented in the next stage of the simulations.