Hello Raul,
If the boundary conditions and the model settings are identical, then the corresponding solutions should become identical eventually. There are a couple of points to consider:
1) Convergence assessments. Please ensure that the solutions are deeply converged not only in terms of residuals' values, but also in some monitors for variables and parameters that are critical to fuel-cell operation: fuel and oxygen utilization, minimum and maximum values for molar fractions in the reaction zones, minimum and maximum temperatures. Please ensure that the double-precision solver is used and that the convergence criteria is lifted from 1e-3 to "none" for all solved equations
2) Please check the presence of inert non-reacting species in the reaction zones, such as nitrogen and water vapour on the "wrong" side. These species can get initialized/patched within the porous layers and it might be difficult for them to get out of there. They can easily reduce the usage of the cell active area and change the outcome of the simulation.
Hope this helps,
Aleksey.