I posted a topic recently about a cathedralized attic, but I need to post another one because half of my roof is a nearly-flat addition.  The roofing for both sections of my roof (the 12/12 in the front, 0.5/12 in the back) has already been done, so installing rigid insulation on the exterior is not an option  Unfortunately I know more now than I did at the start of my reno a year ago.

So, I'm currently working on a viable solution for the roof that involves venting.  There is a 12" wide soffit surrounding the flat roof addition, and the framing has been done with cross purlins: on top of every 2x12 joist is a sloping 2x6 that was cut in a wedge shape from 0" to full height 5 1/8".  That creates a 'peak' in the centre of the roof which slopes down to each side.  The 'cross-purlin' effect was created by placing 2x2's perpendicular to the 2x6's (on top of them) to allow air to travel perpendicular to the joists underneath the roof deck.  If you look up in the centre of the room, you would see: 2x12 joist; 2x6 purlin; 2x2 cross purlin; plywood deck.  The architect intended that air would be able to travel in both directions. 

From reading about flat roof ventilation I've learned to be cautious about its likelihood for success, since the deck heats up a lot in the summer and that hot air won't have a natural 'chimney' to escape from.  Although spray foaming and closing off the vents is certainly an option, I'd like to go with a lower cost alternative if possible.  I'm doing all the work on the house myself so I can afford to pay attention to the details. 

***An important consideration here as well is that the 12/12 roof section is in the front of the house and the 0.5/12 ('flat' roof) is in the back.  That flat roof section "butts into" the peak of the sloped roof but it is slightly lower, so that the vented section of the flat roof 'drains' into the cavity provided by the area above the collar ties in my cathedralized attic ceiling.  I'm hoping this will be advantageous to the venting capabilities of the flat roof, since the 'chimney effect' ventilation in the front sloped roof will hopefully pull air from the flat roof as well.  I realize this may be hard to visualize, so I included a photo from the front of the house - you can see one half of the addition poking out behind the chimney to get an idea.  If necessary, I'll go over there to snap a few more pictures and add them to this post.

Anyway, I'm posting this because one contractor I consulted with suggested that the hot air under the roof deck will "spread out like water" as it expands, and since there are soffits on all four sides of the flat roof addition and plenty of space from the top of the 2x12 joists (which will be filled with Roxul) to the underside of the deck for air to move in I won't have a problem.  One key question I have is whether or not to put an air barrier above the Roxul, using strips of Typar taped between the joists, or cut pieces of plywood, or rigid foam.  My temptation would be the Typar, since it's easiest to work with and also the most vapour permeable (to avoid a double vapor barrier).  That way (in my thinking), the soffit air can travel through the purlins and cross purlins without risk of being pulled into the insulation by pressure differentials (I'm thinking about stack effect in the summer drawing air downwards, then being stopped by the 6 MIL V.B. and forming condensation).

Experiences with venting low-slope roofs???
Thanks :)

I can't see the photo.

If there is at least 1.5" of gap above the insulation it'll have a pretty-good chance of working. It sounds like he lower soffit of the sloped-roof is well below the lowest soffit of the flat roof, so there WILL be stack-effect convection through the venting of the flat roof as long as you don't install any type of ridge-venting at the top of the 12:12 pitch. But it should only have open soffits on the side opposite that which intersects the sloped roof ventilation- if you open up the other sided there won't be guaranteed flow at the far back corners and low-end of the flat roof, since the bulk of the convection would likely escape via the sides.

Rock wool is fairly dense and air retardent compared to fiberglass, but putting a Typar air barrier above it would still add some performance at the temperature extremes. Unfaced XPS or EPS foam even at 1-2" thickness would still be vapor permeable enough to not create a moisture-trap with the (presumably 6-mil poly) vapor barrier on the interior side. But since it's thermally bridged by the rafters it doesn't add a whole lot to the whole-assembly-R. If you go that route, even 3" of EPS (~R12+ @ 0C ) would be sufficiently permeable, but not 3" of XPS. For an application like that using reclaimed roofing EPS from commercial demolition & re-roofing makes it affordable- about 1/4-1/3 the cost of virgin stock.

The "spread out like water as it expands" theory has no science behind it. First, it takes quite a dramatic increase in temp to get even a 2x multiplier to the volume of air, and so what if it SPREADS out? It has to MOVE out, and be displaced by OUTDOOR AIR for there to be any moisture-purging of the roof deck (which is the primary function of a ventilated roof.)