I think generally a vapor barrier is always required on the "warm" side of an insulated wall or ceiling to avoid condensation whenever the dew point temperature falls somewhere in the wall cavity. In some climates the warm side can be either the interior or exterior, also depending on season
15 or so years ago Here in massachusetts it was popular amongs dry-wallers to omit VBs in the ceiling, supposedly to aid in avoiding over-humidifing the living space in winter.
I put 6 MIL poly in our ceiling below the insulation anyhow. With dogs, cats and kids the doors are open so often that we get plenty/too many air changes
The problem we face is that on the 3rd and 4th floors of a 4 story building (the bottom two floors are open parking deck), condensation is forming on air grills located in the drop ceiling. This only occurs on days that are over 100 degree F. The design temp is 96 Degrees F. We are trying to determine what is causing excess moisture and/or heat to find its way into the space above the drop ceiling. We are using a systematic approach and making a list of possible items which may cause this to occur. One is the absence of a roof vapor barrier. If we calculate that one is needed, then the absence of one could be a contributing factor...if the roof does not need it per the calculation, then we ignore that item. Project is in South Carolina.
There is a good practical article on roof design in the latest ASHRAE journal.
However, I doubt that the roof would be causing a uniform problem on the 3rd and 4th floor though. Basic physics check - Is the supply air temperature (i.e. grille temp) less than the dewpoint of your space. Is your space temp and humidity within spec? Is your AHU/RTU doing as it should? Has someone recently umm...messed with it?
Is outside air leaking in and condensing on the grilles? If this is happening closer to the perimeter or near doors this would be my first check.
Is there water being carried over from your coil - face velocity? If the coil is not adequately dehumidifying the air, you have another contribution to the problem. Is the moisture limited to the face of the grilles, or all through the ductwork?
