Ventilation channels below roof deck vs above
One accepted assembly for the insulated roof when using fiberous insulation is to build ventilation channels, eve to peak, under the roof deck and between the rafters. It’s considered best practice to use caulk to airseal the ventilation channels from the interior side. (See 2017 FHB house baffle article, link at end of post). The insulation batts or loose fill go underneath the baffles and then drywall attached the rafters.
Now, consider a structure built with the roof deck sealed to wall sheathing, and all seams are taped, etc to make the sheathing and roof deck one continuous airsealed layer. Now 2 crossing layers of furring are added to the roof deck to create eave to peak ventilation and modest overhangs, and a metal roof is fastened to that. Fibrous insulation is installed between the rafters and drywall on the rafters.
Are these two assemblies any functionally different for moisture management?
They don’t seem any different to me, but the first assembly is normal while I’ve never seen the second suggested. It seems like the second method would offer several advantages; it’s quicker, less sheathing used, the structure is dried in quicker (if using ZIP or similar) , air sealing is easier, and you can eliminate the soffit vents and use Coravent or similar between the furring strips.
For reference, 2017 FHB house site built channels with loose fill cellulose:https://www.finehomebuilding.com/2017/08/02/ventilation-for-the-prohome-roof
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Unless you put your sheating on the inside or you are using some permeable sheating (fiberboard for instance) this is a dangerous assembly in terms of moisture. You will get condensation under your sheating unless your air sealing is perfect because it is on the cold side of the roof.
The problem is the permeability of the sheating. If it doesn't let enough vapor pass, there is no way it can reach the other side of the panel and be vented out of the attic.
Just skip the sheating and install some permeable membrane instead like the Proclima Solitex Mento or Siga Majvest. Rigid moisture proof fiberboard panels would be the best but you have to import them from Europe.
Thanks for the reply, Alex. Is the issue really the permeability? From what I've read, the ventilation channels can be built from OSB or rigid foam, which are really not more permeable than ZIP, or sheathing with a WRB.
Yes permeability is the issue and you should use vapor permeable baffles. The reason vapor impermeable baffles don't cause problems most of the time is because they are not air sealed and they usually don't cover the entire roof deck so moisture is still able to reach the vent channel anyway. But that's not a good design and it really depends on how air tight your ceiling is. If tons of moisture is able to reach your impermeable baffles, you will have problems.
Also keep in mind that the thickness of the material is proportional to its vapor permeability. 1/4 plywood is much more vapor open than 3/4 plywood and doesn't classify as a vapor barrier.
Would you consider the 2017 FHB house assembly, I linked to in my post, not safe? They use 7/16 OSB and throughly airseal the channels.
You should check this article : https://www.greenbuildingadvisor.com/article/site-built-ventilation-baffles-for-roofs
It seems vapor impermeable baffles are not a big concern but I think foam would be a safer choice than OSB.
Martin's article "Site Built" says, yes, use vapor permeable baffles if airsealing them.
It's hard for me to find a solid answer, but some searching is indicating bare plywood or OSB sheathing has a perm rating under 10.
Under 10 perm rating is "semi permeable" not "permeable".
So....it seems the FHB article is not presenting a safe assembly with 7/16 OSB baffles, airsealed, and loose fill celluose.
>"They don’t seem any different to me, but the first assembly is normal while I’ve never seen the second suggested."
They're pretty different. AlexPoi has that part right. The vapor permeance of the roof deck is less than 1 perm (it would qualify as a "vapour barrier" under Canadian code definitions), and it will accumulate moisture from interior moisture drives in winter when cold.
It would be difficult to meet structural codes without structural roof sheathing though (that part would need to be engineered.)
Dana, but what is the permeability of the ventilation channels, built from plywood, OSB, or foam, and airsealed?
Regarding structural roof sheathing, my house, as built in 1978 in Vermont without codes, has no roof sheathing but only purlins with metal roofing installed. My questions arise from contemplating what I will do when I replace the roof. I am leaning towards adding sheathing, but also contemplating leaving the attic vented and unconditioned or making it unvented and conditioned.
Tim,
The primary aim of a ventilated roof is to protect the sheathing from moisture build-up. A roof with the vent channel below the sheathing is designed to both dissipate the moisture before it hits the sheathing, and to dry any that does accumulate in it. When the vent channel is above the sheathing, moisture has to move through the sheathing before being dissipated.
I'm not convinced impermeable baffles are a good idea, or that air-sealing them makes any sense. I wish someone would re-visit that advice, especially with reference to how important wind-washing is or isn't.
Dana,
What would be a good way to make a cathedral ceiling work if you don't want to put the ventilation under the deck? I'm trying to plan a roof right now and the type of construction Tim describes is appealing to me because it seems simpler. Would his plan work if you used all vapor permeable materials? I'm thinking of doing interior air-barrier like intello plus -> mineral wool batts -> sheathing -> solute mento plus -> ~2inches comfort board -> furring for roof -> metal roof
I'm in San Jose California so we don't have a very humid environment, or a cold winter, so not sure if I'm going in the right direction trying to plan for I guess a vapor open building?
I would put the sheating on the inside, use it as an air barrier and skip the exterior insulation unless your attic is not truss framed. In that case, you'll need the insulation as a thermal break. And since you are in a seismic zone, you'll need to check with an engineer.
Tim asks:
>"Dana, but what is the permeability of the ventilation channels, built from plywood, OSB, or foam, and airsealed?:"
With half-inch thick or thinner baffle materials the permeance of the baffle hardly matters, since the 1/2" of rafter is sufficiently high permeance to provide a drying path for the cavity fill. Half inch OSB is under 1 perm when dry, about 5 perms when the moisture content is high enough to support mold, but the half inch long path through a milled lumber rafter at the edge of the baffle is a few perms even when dry.
Alex asks:
>"What would be a good way to make a cathedral ceiling work if you don't want to put the ventilation under the deck? I'm trying to plan a roof right now and the type of construction Tim describes is appealing to me because it seems simpler. Would his plan work if you used all vapor permeable materials??
Hi-perm materials create a wintertime moisture accumulation problem in the roof deck in cold climates. You can get away with it in US climate zone 2 though. For zone 3 a smart vapor retarder on the interior can work in some cases, but not necessarily if using "cool roof" high SRI roofing.
My thinking about vent channels has changed over the years. Initially, I was thinking of them as channels where air flow from wind/stack effect would blow through and dry the underside of the roof.
This does happen to some extent, but the more important function is to reduce the dew point under the roof deck to near ambient air. If the dew point consistently stays bellow the roof deck's temperature, you won't get any condensation, thus nothing to dry.
I think it is this above mechanism that is important in colder climates, this is why roof with sub optimal venting (things like side vented flat roofs) will still work even though there is no real airflow within the channel.
So if you look at the roof from this perspective, there is a big difference between top and bottom vented roofs. This is also why fully air sealed vent channels are bad idea unless made from very permeable materials (ie fiberboard or good house wrap).
You can top-vent a fiber filled roof, but it has to have a good warm side air and vapor barrier. Even then, you would want something more permeable such plywood above.
In a cold climate, best to have the exterior side be 5x more vapor permeable than the interior side. Sheathing+underlayment or a vent baffles+joists assembly may meet this (or not).
I wouldn't trust air leaks around the edges of vents baffles to provide adequate vapor movement. But maybe someone has hard data showing this is reliable across the entire vent area, even with cellulose (which tends to plug air leaks).
Air sealing is critical and two air barriers are better than one.
Does anyone have a link to an existing assembly using dense packed cellulose in a cathedral ceiling? I was wondering the same thing, and seem to read a lot of conflicting information. I've read the importance of a really tight 6 sided air barrier for the roof on GBA, but unsealed baffles would seem to be the opposite of this. If plywood baffles are fine because they are not sealed, are we then to assume that the above deck ventilation channel would be fine if the bottom layer of sheathing was not sealed and leaky? Wouldn't this drive interior air into the assembly from the stack effect?