Sediment deposition near to a river confluence is a crucial challenge for navigation since it can strongly affect the bed morphology and thus the barges traffic. The prediction of particle deposition is thus important in order to optimize the dredging operations which are necessary to maintain a sufficient draft. In the present study, we numerically examine the transport and deposition process of sediments at Scarpe and Escaut confluence (north of France) using an Euler-Lagrange approach. This method is known to be able to accurately predict the behaviour and spatial distribution of inertial particles in turbulent fluid flow. Nonetheless, contrary to the standard method (based on convection-diffusion equation for sediment concentration) usually used to predict the behaviour of sediments in such a complex configuration, it has an higher computational cost. In our study, the mean fluid flow is predicted by solving the 3D Reynolds-Averaged Navier-Stokes (RANS) equations. The trajectory of each sediment is then computed from the particle equation of motion with a stochastic dispersion model which predicts the sediment/turbulence interaction.