IESVE 2019 gives me totally wrong solar gains from windows.

[blumentopferde]

Dear community!

I have very persisting and annoying problems with IESVE 2019 and I suspect that these are related to SunCast:
I get obviously wrong simulation results for solar gains. This problem occurs in thermal zones that are attached to other zones or geometry (but not in all of them), I never had this issue in free standing zones.

The problem is as follows: Solar gains are too low (less than 10% of the incident solar radiation at a g-value of 0.5) and seemingly independent of the orientation: I get similar solar gains from north-facing windows as from south facing windows, as if only diffuse gains were calculated.

I have already tried different suggestions from the forum, all without success:
- I saved all files on a local folder instead of a network folder
- I have reduced the size of the model and centred it over the origin
- I have deleted existing shading files from the SunCast Folder
- I have created a model report, but it didn’t find any geometry problems.
- I have tried some key-in commands to repair flawed geometry, such as: clipopenings, build, model, rebuild, rmh, tidy, tidysurfs and tidyunbound, but without success.
- I exported the geometry to a GEM-file and imported it into a new file.

None of this works - The solar gains stay weirdly low. I only get plausible results for free-standing zones without any adjacent geometry.

What can I do to get accurate results?
Can anyone help me with this?