Farfield Results on a structure which requires Open boundari

Hello mates,

I have a very tricky question for you today:

I need to simulate an antenna (loop) inside a pipe full of water. Obviously the antenna has been protected by a proper radome.
I need to know the farfield of the overall structure, but a problem occured.
Suppose that the plane of the loop is the yx and the pipe's axis is along z. The boundary conditions I have set are the following:

ymax, ymin, xmax, xmin --> Open add space
zmax,zmin --> Open with PML gamma=1e-6

The boundaries in zmax and zmin have been chosen so that the pipe looks has it were infinite. I cannot add space in such directions because otherwise the material's changing (water->free space) would cause a very high reflection and the pipe behaves as a resonator (this is NOT what I want).

When I look the farfield results with the above boundaries, the CST gives me a warning: Directivity<1 farfield innacurate. I have tried also to delete the imaginary part of the loop with a lumped element at the frequency of the farfield monitor in order to resonate the antenna, but the problem remains.

Can someone help me please?

Regards,

Luca