An analysis of the PKL midloop tests E3.1 and F2.2 run 2 was performed with the TRACE (TRAC/RELAP Advanced Computational Engine) and RELAP5/MOD3.3 codes. Both tests allow study of the phenomenology and different accident management actions after a loss of the residual heat removal system at midloop conditions with the primary side closed. A comparison of the results obtained with both codes and the experimental data shows that in general, the main phenomena are well reproduced. The good results obtained allow one to confirm that the modelization methodology is adequate for this kind of transient. However, there are still a few phenomena that are not well predicted, like pressurizer water level behavior.