Introduction Inland and seagoing vessels are equipped with (bow) thrusters. The use of these thrusters can cause scour of the bed alongside a quay wall. In order to assess the consequences of local scour due to bow thrusters on the design of the structure it is desirable to know the dimensions of the scour holes as a function of in time. Analytical relations for the calculation of scour, only determine the scour depth, but do not involve remaining dimensions of the scour hole and the development of scour in time. Model description A three dimensional quasi-steady-state numerical model is developed, which describes the scour development alongside a vertical quay wall induced by thrusters of a vessel. The flow simulations, using the Realizable k-? Model in the Computational Fluid Dynamics package OpenFOAM, provide the flow properties near the bed in order to calculate the bed shear stress. In the flow simulation the influence of the rotation of the propeller is neglected. A boundary adjustment technique is applied in order to move the mesh near the bed. In case the critical shear stress is reached, the morphology plays a role and the bed changes. This critical shear stress and the bed change are calculated, by applying an empirical relation for the erosion rate of sand, which is valid for both the high and low velocity regime of the flow. Every time step the bed level is updated and the hydrodynamics are calculated for the updated bed level. Results & Validation The hydrodynamics are validated separately from the model with erosion by applying a fixed (i.e. non-erodible) bed and comparing the results regarding flow velocity in the numerical model with physical experiments and analytical expressions. Several cases are elaborated with different distances between quay and ship, and keel clearances. It appears that for relative large distances between ship and quay the numerical model overestimates the near bed velocities. The erosion itself is validated with full scale tests regarding thruster induced scour, performed in the Port of Rotterdam. A maximum scour depth of 2.2 meter after 6 departures of the vessel is calculated with the numerical model, which is rather conservative compared to the measurements where a maximum depth of 1.75-1.85 meter is measured. The difference in results is probably due to presence of clay and silt in the Port of Rotterdam, however the numerical model is developed using an empirical relation for sand. Sensitivity From a sensitivity analysis it appears that the maximum scour depth is sensitive for the amount of fine sediment which is present in the soil. Coherent with this the porosity of the soil and the dilatation of the soil during erosion play an important role in the magnitude of the erosion rate. Recommendations It is recommended to improve the modelling of the propeller induced flow and to integrate this flow model with the current erosion model. Beside that an improved mesh near the bed is recommended. In this case flow properties, regarding flow velocity and turbulence, can be averaged over the water column in order to involve the higher order moments for the calculation of the extreme bed shear stresses. The length of this water column depends on the mixing length of the turbulence. In addition, it is striking that the width of the scour hole in both calculations and measurements, are smaller than the width of common applied bed protections. From an economical point of view it might be interesting to assess whether the current bed protection design is not too conservative.