Steam condensation has turned into a key safety mechanism in the removal of heat from the containment atmosphere under hypothetical accident conditions of the next generation of nuclear reactors. The new configurations and conditions foreseen have required further studies focusing on various condensing scenarios. At the University of Wisconsin an extensive experimental plan to investigate condensation related to AP600 containment has been undertaken. In addition a model capable of simulating the tests and, eventually, predicting the heat transfer under the anticipated conditions in the case of a postulated accident is being developed. In this paper two different approaches based on the heat and mass transfer analogy are evaluated: the so called ''standard'' approach and the Peterson's type model. Assuming free convection conditions, the model predictions are compared to two sets of experimental data corresponding to different compositions of the noncondensable gas. It has been found that Peterson's approach is limited to small temperature jumps across the boundary layer. However once this drawback is removed from the formulation consistency between both approaches has been observed, and a close agreement with the data when suction is accounted for, is noted. Further work is required, however, to test the model at overpressure conditions and to make it capable of dealing with wall orientations other than vertical.