I’ve been reading up on thermal bridging and high R-value roofs, and mulling over a theoretical assembly in my head, but since my understanding of all this is pretty surface level I was wondering if someone knowledgeable would be interested in weighing in.
Given that the ideal location for the vapor barrier on a wall in a cold climate (zone 5) is somewhere within the assembly in order to protect it, but not too close to the cold side of the wall, it seems to follow that this would apply for the roof as well – there are certainly many instances of leaky vapor barriers full of can lights and vents so moving it away from the ceiling seems like a good idea.
I’ve been trying to work out an assembly with minimal thermal bridging, a continuous transition of a protected vapor barrier from wall to roof, and minimal or no rigid insulation. It seems to me a “Mooney wall” type roof could be constructed if the rafters (no overhangs) were filled with R=30 BIBS and sheathed with plywood, vapor barrier applied, then 2×8 purlins installed perpendicular (except to form the eave overhangs – maybe staggered here) and also filled with R=30 BIBS. You could sheath that with something that is vapor permeable, add an AWRB, a drainage mat or other air gap, and install metal roofing panels or something over the top.
My thinking is that separating the top and bottom of the roof structure would allow the vapor barrier to be continuous without worrying about the pesky overhang, and the protected location within the assembly would allow it to dry to both sides. This is probably a terrible idea, since I haven’t seen it done or proposed, but I’d like to learn WHY. Thanks in advance!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
The vapour-barrier doesn't need to be entirely continuous or sealed, so it doesn't need protecting, or moving further into the assembly. Locating it at the interior surface is fine. If it is moved into the insulation layer, where it is still safe from condensation depends on climate, but the rough rule of thumb is no more that 1/3 of the distance from the interior - which would pose a problem for the assembly you are suggesting.
The air-barrier needs to be continuous, sealed and protected. It could be moved to where you suggest, but if it is the lower layer of sheathing , it should be kept permeable, and not treated as a vapour-barrier.
Over-framing a second layer above the lower sheathing is certainly possible. It does become a lot harder structurally when the members run horizontally, and are a deep as 2"x8"s, especially on steep pitches. If the intent is to form the overhangs and reduce thermal bridging with this layer, running something much shallower (2"x4"s), and running them offset but still in the same direction as the rafters below would make more sense. Using batts on this outer layer makes the assembly weather dependent in a way using them on the interior of the roof underlay doesn't. Here in the PNW it might mean being unable to build the roof for several months.
The other problem with the assembly is it doesn't have the required ventilated air gap. It needs to be something more than a drainage matt (most codes want at least 1" clearance), so yet another layer - or you could use the upper framing for this, but you would lose the benefit of a thermal break.
Stepping back a bit: Unfortunately I don't think that it yields much that more commonly used, simpler and less expensive roof assemblies don't.
Thanks, that gives me food for thought. I clearly need to read up more on air barriers vs vapor barriers, it’s possible I’m confusing them. We do 2x8 purlins over porch roof rafters all the time here so I’m not worried about the structural considerations, but I definitely hadn’t seen specifics on an air gap requirement (although I’ve seen it shown in some assemblies) so I will look into that more as well. Thanks!
I have a follow up question: how much of a gap is required behind siding or under roofing to allow an assembly to dry to the outside? I assume that’s what you meant by 1 inch air gap, but a lot of rain screen assemblies use 3/8” furring strips so I wasn’t sure - maybe it varies?
Bigger gap required under roof as air doesn't move as efficiently as behind siding on a vertical wall.
True, but most rain-screen walls aren't open at the top, and don't try and move air - yet are still effective. That wouldn't be true of similarly built roofs.
It's worth remembering that roofs are not simply sloped walls, their requirements are quite different.
Because the greatest moisture threat to walls is from the outside, the rain-screen gap behind siding is primarily about providing a capillary break. Venting is a secondary benefit, like re-distributing moisture, and providing a drainage plane. That's why the gap can be as little as 1/4" - although practically it's hard to build one that narrow and still keep the cavity clear.
Because roofs a) have an impermeable outer surface, and b) are subject to the stack effect bringing more moist interior air into them, they need a deeper gap capable of providing an effective ventilation channel from eaves to peak. Most codes set the minimum depth of that at 1". Effective vent channels on low sloped roof need to be deeper.