This note provides an evaluation of the ASTEC V0.2 code capability in simulating typical severe accident scenarios. One base case (the H2 sequence) has been chosen. Afterwards, different accident evolutions are studied with the aim of testing the code in extreme conditions: e.g.assuming, at the vessel failure, either a high or a low vessel pressure,various degrees of hydrogen combustion, etc.. The analysis is accomplished using ASTEC V0.2 code in a multi-compartment(four compartments) mode. Main conclusions are: an acceptable computing time (it increases with the number of compartments and with the activation of modules as the combustion one, as well as with the size of compartment junctions) and, after some model corrections, an improvement of code robustness. The target is to provide a data base in order to judge the ASTEC v0.2 code physical model behaviour by comparing current ASTEC predictions to similar ones obtained by ESCADRE mod1.2 (mono-compartment mode).