CST accuracy control beyond the mesh density

Hello,
In CST, is there a relation between the calculated Radiation Efficiency and the accuracy of the results?
How one would control that?
Thanks

Your question is not clear.
If you have >100% radiation efficiency, something is obviously wrong.
If you have a lossless antenna model, radiaton efficiency must be 100%.
For a lossy antenna, we don't know what efficiency to expect.

My radiation efficiency is -2dB. I use materials with no losses. Return loss about -10dB
Not sure why the radiation efficiency is so low. I seem to remember that if the meshing is too sparse that could be a problem. I am using 25 cells/wlgs. The convergence is about -30dB

That's obviously wrong. For calculating radiation efficiency, the total radiated power is calculated by integrating over all the boundaries. I'm not a CST user, but in some EM tools there are settings for field sampling angle range (theta and phi) and resolution. Maybe your angle sweep range for efficiency calculation misses some direction.
I would not expect that the mesh has any effect on accuracy of calculated efficiency.

Thanks, that confirms my suspicions