Vapor-Open Membrane for Unvented Attic
Unvented attics that are not insulated with ccSPF have been prescribed a vapor diffusion port to take care of the interior moisture migrating upwards to the sheathing. However, ive seen this used primarily in asphalt roofing (shingles and felt) where that combination provides essentially a vapor barrier above the deck, trapping moisture in the sheathing.
So my thought is that if I have a vapor open membrane (solitex mento, 38 perms) above a plywood roof deck, and have a rain screen with a ridge vent, that would essentially convert my entire roof into vapor diffuser as vapor has nowhere to become trapped. Does this make sense in a hot, humid climate?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Alberto,
The reason vapor diffusion ports are used for asphalt shingles, is *because* they're impervious. The VDP is there to provide a permeable membrane so that vapor pressure can reach equilibrium. Just to make sure it's clear, the VDP is only a membrane spanning empty space, it's not some membrane sheet covering the top section of plywood.
On the other question - if you have a vapor open membrane (38 perm) over plywood (1.X perm) with a ridge vent (10^100 perm) you will have significant drying capability, should the deck ever get wet. Are you talking about an over-roofed, unvented assembly here, where the venting is between the channels of the rain screen, or a vented attic with solitex as the underlayment for shingles?
Kyle,
Sorry for not clarifying that, it is an unvented attic with a rain screened metal roof, the "ridge vent" i am referring to is the metal ridge cap. So yes the only venting is occurring at the rain screen.
No worries, I was pretty sure that was that case. A few more questions,
1. What's the approximate location? You don't have to be specific, but something like "southern Texas" would probably help narrow in for some advice that's climate specific.
2. Is there an attic space, or finished living area under the roof? That may have an effect on a few things.
3. If you're not insulating with ccSPF, what did you have planned? Cellulose, fiberglass, mineral wool, or some blend?
Just remember that water, like anything else, flows from high to low concentrations. If the conditions favor moving from the attic space to the ourside, it'll do that. However, the opposite could also happen, and drive vapor diffusion in the opposite direction. BSC (~2003 ish) has a few articles on vapor drive through sheathing, thought it's claimed to be due to the water absorbed in the space between shingle layers. I'm not sure this would be the case with a metal roof, and the capillary break with a large rain screen.
Kyle,
1) Im in northern Honduras which resembles CZ1
2) Small attic space, more of a service cavity, not a lot of volume
3) Fiberglass only, ccSPF is out of the question as there are no suppliers down here and cost would be prohibitive anyway. Polyiso (for above deck insulation) is available only by importing which is doable but adds significant cost to the assembly. This is the reason im trying to figure out if a fiberglass only assembly could work with the correct drying potential of the assembly. BTW we have also spec'ed air exchange in the attic space via an ERV.
Sheathing can reliably stay warmer than 45 F year round to avoid condensation. Not sure if interior humidity can stay reliably within 50% on those cold days though, as our mini-splits would cycle a lot.
Thanks for the info.
Given the temperatures and rainfall totals I was able to find on the internet, I don't see a reason an assembly like this couldn't work. It has bidirectional drying capability, is vented, and wouldn't appear to trap moisture. There's practically no chance of condensation forming even in your coldest month.
Data I've seen suggests that asphalt shingle assemblies (and metal roofs) are about .6 perms. What releases more moisture - a small 20 perm diffusion port or an entire roof covered with vapor permeable underlayment and .6 perm of asphalt shingles?
Good question, Kyle. By rainscreen I'm thinking we are talking either metal roof on wooden battens (criss-crossed or diagonal, or we're talking overroof (membrane, battens, sheathing, asphalt).
One consideration (perhaps?) is that diffusion can be both ways ... but if the port is only at the ridge then the warm moist air will migrate there and thus maintain an inside --> outside gradient even when it is humid outside.
But that will probably happen even if the whole roof is vapor permeable.
OTOH, does anyone think one might get inward diffusion lower on the roof (outside-->inside) as the moisture migrates to the ridge, leaving the lower part of the air in the attic relatively dry?
James,
Yes, excuse me for not clarifying that. It is a metal roof on battens, providing venting between the panels and the solitex membrane. I was also wondering whether diffusion can happen both ways. I guess its a question of whether the vapor pressure differences between interior and exterior are most significant than the humid air buoyancy. This is climate zone 1 BTW.
Alberto, yes, that's a good option, often called an over-vented roof. You still want to limit how much moisture can get into the assembly and with metal roofing you need to know that night sky radiation will create condensation dripping from the bottom of the roofing panels during certain conditions, so make sure the Mento is installed in a watertight manner.
A similar approach is the eliminate the sheathing entirely and just use Mento directly over the rafters, called a Sarking Membrane system. It's not as hard to do as some who haven't tried it think it is, but it is different than typical approaches and the roof framing needs to be braced a different way.
Micheal,
The Sarking method seams very interesting, however, I suspect that air sealing the mento overlaps and the roof-wall connection would be very challenging. Will look more into it, thanks!