Hi,
One quick comment on the passive scalars: they are usually used to evaluate contaminant dispersion or mean age of air. I don’t think your specific application is a good candidate for passive scalars.
In any case, if these openings would be “open” in real life, I would not expect any walls blocking these paths:
For example, this ventilation project that I posted on forums some time ago.
So those “natural convection inlet/outlet” BCs that you defined to those openings should not exist. If you want to model no wind around the hut, I’d recommend setting natural convection inlet/outlet to the top and sides of the enclosure instead:
