The formation of free surface vortices at intakes results in air entrainment into the headrace system of hydropower plants. This reduces efficiency, may damage turbines and pumps, and increases the likelihood of transient phenomena in pressurized waterways. To prevent these losses and damages, operators of hydropower plants maintain a minimum submergence level above the intake. As a result, a significant part of storage volume is not usable leading to lower energy storage capacity and income. In this study, a two-phase 3D CFD model was implemented in OpenFOAM to examine the influence of five different intake geometries on air entrainment and critical submergence. For this purpose, eleven different model configurations were used. Among different RANS turbulence models, the k-Ȧ SST model was chosen. For validation of the CFD model, the data of a physical model were compared with the simulation results. It can be shown that, in the case of low circulation above the intake, unsteady RANS models cannot properly capture the dynamics of free surface vortices. The RANS approach tends to overestimate the turbulent kinetic energy and the turbulent viscosity of the weak circulation region above the intake induced by the geometry where the kinetic energy is indeed low. This results in excessive kinetic energy losses and the already weak circulation in the model is not sufficient to develop either surface dimple formation or full air core vortices. For two model configurations, the simulations show reasonable results for the estimation of the air entrainment. Furthermore, the results confirm that the distance of the intake opening from the front wall and the inclination of the front wall have a significant influence on air entrainment and critical submergence.