Unfortunately, Carlos is correct. Version 6, and earlier versions of FEFLOW, allow a "parameter expression" on elemental properties only, two of which are actually boundary conditions (in/outflow on top/bottom and source/sink). Recharge could be used to simulate ET if your model can be set up to simulate ET (negative recharge) only in layer one. Otherwise, you would have to use the source/sink "material" but this is further complicated by its odd units: your parameter expression would have to account for the control volume associated with this "material" (which could be specified and accessed via an elemental user data distribution).
I once tried doing something similar using a parameter expression on source/sink and it did not work so well, which leads me to think that simulating ET this way might not work so well except in simple cases.
There are no "parameter expressions" available for "Boundary Conditions (BC)"
Another solution, which would probably work well, would be to program an IFM module to handle ET using flux or well nodes (well nodes would be easiest). Let me know if you want help with such an approach.
I think the ET video would still be applicable, but the URL is broken and has not been fixed.
Pete