During some studies with a OGS-RICHARDS_FLOW model, I was wondering if it is possible to use GRADIENT boundary conditions (bc) in combinations with #Curves (.rfd). Well, what should I say: it somehow worked!
However, there is something you should know and take care if you do so:
If you use e.g an hydro-static z-gradient along a surface with Curve 1:
GRADIENT 1 0 9810
and you combine it with curve definition in the RFD file like
; time value
then at first the appropriate gradient value is assigned to the mesh node (or RHS vector) and multiplied with the time-related value from the RFD file, which is 2 in the above example.
Looking in the code tells me that this happened just by chance and parameterization of the curves values will be therefore not straightforward. However, if your original curve values can be transferred to multiplication values then it should be fine.