The future tokamak ITER is expected to run routinely in type-I ELMy H-mode confinement regime. In fact, the ELMs are necessary in order to rid the plasma of its impurities. However, it is expected that ELMs in ITER could release up to 20MJ, which is not bearable for the presently designed tungsten divertor. Hence, some techniques have been developed to control ELMs (RMPs, pellets, kicks), but there is at present poor understanding of how these tools provoke or mitigate the ELMs. In fact, the ELMs themselves are not fully understood. The linear stability of ELMs has been well established in the last two decades, but the understanding of the nonlinear properties of ELMs still requires some effort. The nonlinear simulation of ELMs is already in reasonable qualitative agreement with experimental observation. In particular, simulations show a filamentation of the plasma edge into the Scrape-Off Layer, and large amounts of energy arriving on the divertor in the form of small heat-flux structures near the strike point. In sight of using the simulations to obtain some predictions of ELMs in ITER, the MHD code JOREK should first undergo a quantitative validation against experimental data. This paper aims at a first step towards this validation, by presenting simulations of ELMs for a given plasma pulse from JET. The long term goal being to extend the simulations to multiple shots analysis, both for JET and for other tokamaks.