Radiation efficiency very low on Gap-Coupled Patch (MWS)

Hi there !
I am experiencing some problems with a gap-coupled simulation that I setup...
S11 params are good (resonant freq ~ 865MHz) but radiation efficiency is very low (~1% !)... I'm using a probe to feed the patch (modeled as a discrete port).
I'm attaching the geometry if someone has time to check it out :)
Any help would be much appreciated !
cheers !
mog.

I don't think your patch antenna is a good radiator. You have 10/20 mm width of each patch, do you really expect them to radiate highly? I'm also confused.
Since lenght of your patches are approx. half a wavelength, you have resonance at 865 MHz.
Please let me know how much eff. do you expect.

Thanks for your comments :)
That's what I also thought myself at first... Difficult to achieve high efficiency rates but 1-3% sounds extremely low to be true...
On the other hand, directivity of the patch seems normal. Doesn't that mean that it is working well as a radiator (at least the radiation is shaped normally).. Am I totally wrong on that assumption ?
Expected efficiency isn't known.. The design was used after the attached paper (exact representation)...
Thanks again !
mog.

ya, I don't guess either about the expected efficiency.
Directivity plot seems ok however it does not mean that your antenna is a good radiator. Because directivity is defined by "radiation at a certain direction averaged over all directions" -- means it considers only the power radiated but does not take into account input power to the antenna.
regards,
Shameem