large vortex on outlet when gravity accounted for
 

I have not been able to figure this problem out. I am trying to model a simple rectangular room with a window (inlet) on one end, and an open outlet (discharge to ambient) on the other end. I also have a hot plate somewhere on the floor, near the window. Density is calculated according to Boussinesq approximation.

In the output, the outlet will have a large vortex (with axis perpendicular to gravity, and on the plane of the outlet), where the velocities are about two or four times that of the entering jet (the vortex is there even if I remove the jet and the hot plate). I have large incoming velocities near the ceiling, and large outgoing velocities near the floor.

In any simulation I have run, where I have a vertical outlet, I have the same vortex.

I think that the answer could come from having a hydrostatic pressure inside the room and a constant pressure in the outlet. If this were true though, that would mean that I cannot use "outlet" as a boundary condition. What to use then?

What am I doing wrong?

Any feedback will be greatly appreciated!