Disruptions reduce the lifetime of future fusion machines. In view of the construction of a DEMOnstration fusion power plant (DEMO), a solid physical basis for plasma disruptions must be developed. In particular, unmitigated (hot) Vertical Displacement Events (VDEs) must be avoided, as they apply large electromagnetic forces and heat loads to the plasma wall. A promising mitigation technique is based on the injection of impurities into the plasma. This causes the plasma to radiate much of its thermal energy on a millisecond time scale (thermal quench, TQ) and the vertical force on the wall can be greatly reduced [1]. However, the increase in the electrical resistance of the plasma at the end of the TQ leads to the establishment of large toroidal electrical fields. These accelerate the electrons to relativistic velocities, producing so-called runaway electrons (REs). By carrying a large fraction of the pre-disruption plasma current, the REs can locally damage the first wall, including deep melting, by applying large thermal loads to it. In the present work, numerical simulations are performed to predict the power deposition of a formed RE beam in a plausible scenario for a “European DEMOnstration fusion power plant” (EU-DEMO). We study the formation and termination of a RE beam using a magnetohydrodynamic model coupled to a RE fluid model. In postprocessing, we evolve the RE markers along the magnetic field lines calculated in the previous fluid simulations to track the RE markers lost to the wall and estimate the energy deposited. Our aim is to contribute to the construction of an EU-DEMO by testing the goodness of the design of the upper limiter, as originally conceived by the “DEMO team”, in protecting the wall from the RE. We also show studies where different upper limiter geometries are considered with respect to the original one. This allows us to address the question of how the upper limiter can be better modelled to better protect the wall [3]. [1] N. Schwarz et al, Nuclear Fusion 63, 126016 (2023) [2] M. Hoelzl et al, Nuclear Fusion 61, 065001 (2021) [3] F. Vannini et al, in preparation.