Can I apply open cell foam under a vented roof deck?
I am planning to build a number of homes with some affordable housing groups in my area and have run into a significant expense with the application of closed cell foam to the underside of the roof decks to create conditioned attic space for mechanical systems. Constraints and additional details:
– Climate Zone 4, on the edge of Zone 5
– the homes are generally going for Energy Star levels of performance
– the attics must remain conditioned
– the homes will max out in size at about 1500 sq. ft.
– to preserve the warranty on our shingles, we plan to install the primary roof deck over trusses, lay flat 2×4’s on top of the primary roof deck, and then install a secondary roof deck on the 2×4’s, creating a 1 1/2″ vent channel for soffit and ridge vents
– in Virginia we can install ceiling insulation at R-49 if the insulation extends over the top plate and modeling (BeOpt) indicates negligible improvements if we install R-60. (about .5 %)
Here’s the question(s):
Given that we are building a vented roof assembly regardless of insulation strategy, can we use open cell foam with an intumescent coating in lieu of closed cell?
What are the considerations for and risks of using open cell?
I have very little experience with spray applied foams – what are my biggest blind spots to watch out for?
For context, if we can make the switch to open-cell, the savings would be nearly $9000 per home – at 56 homes, this type of savings could mean housing 2 more households, possibly 4 if we can leverage donations and local volunteers effectively.
Thanks in advance!
Nat
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Nat,
Yes as long as you condition the attic. The venting above the sheathing won't save the roof if there is excessive humidity in the attic. By conditioning I mean not just include it in the thermal boundary, but actively condition it reduce humidity.
https://www.greenbuildingadvisor.com/article/high-humidity-in-unvented-conditioned-attics
To save a bit more, I'd suggest substituting a membrane for the lower sheathing. That way you have a vent channel separated from the foam by a high-perm, material and avoid two layers of sheathing.
https://www.greenbuildingadvisor.com/article/building-a-vaulted-high-performance-and-foam-free-roof-assembly
Malcolm,
First, thanks for the quick reply!
Second, the attics will absolutely be conditioned! I appreciate the emphasis on the fundamentals. I will also get some zonal readings when I run my blower door to ensure I don't have a "confused space".
Third, the vent channel assembly is probably at the limits of our labor force here. A number of these homes (if not the majority) will be modular and the roofs will be pre-sheathed in the factory and "unfolded" after the boxes are placed, so the extra expense in materials should be made up for in labor savings. I would love to try a membrane based assembly since I saw it in the PGH book, and if I ever personally stick-frame another small house with a cathedral ceiling, I will probably give it a shot. Do you have a preferred membrane you would recommend?
Thanks again!
Nat,
With a well conditioned attic and also providing a vent space above I think that roof should be problem free.
I agree. The membrane covered trusses may be a bridge too far for most framing crews. If I did go that route on a project, given that most membranes and house-wraps are much more vapour-open than OSB or plywood, I'd probably opt for something like Tyvek Commercial simply for it's strength.
Noted and appreciated! Best!
Important bit for this assembly to work well is the main roof deck (the one with spray foam on it) needs to be permeable. Be careful that if any underlayment is installed on it, it needs to be permeable. Regular felt works great and there are a number of permeable synthetic options out there as well.
The permeability of the 2nd roof deck under the shingles doesn't matter. There you can use any underlayment your roofers prefer.
PS. If you ever want to do a sarking membrane type roof, I would design for purlins instead of rafters. This will let you at walk on the roof while placing the membrane.
If the roof is vented you don't have to use foam at all. Fiberglass, cellulose, wood fiber and mineral wool would all work. Open-cell is less problem-prone than closed cell but search for "spray foam failures" here or elsewhere online and you'll find plenty of stories.
The best way to build over-vented roofs like the one you propose is to eliminate the sheathing directly on the trusses and install a high-quality WRB (aka housewrap) instead. You can search for "sarking membrane roof." It's unconventional and may require additional truss bracing, but it's not prohibited by typical building codes. (Walls are a different story.)
Thanks Michael.
I would love to get away from foam altogether if I can; I just have to make the case for it. (PGH is open next to me - thank you for that contribution!)
Given that this assembly is cathedral ceiling adjacent, am I correct in thinking that we need to install a vapor control layer such as Intello on the underside of the insulation? This would allow for blown-in insulation, but may require some accommodation in our roof framing/truss design.
Michael,
I'm a bit confused as to how these attics using trusses, with permeable insulation following the roofline - whether vented below, above, or using a vapour diffusion port - should be built.
It seems like all would be vulnerable, or at the least inefficient without a warm-side air-barrier, but I don't see how practically you would provide that - and I don't see it mentioned in Joe L's article on these assemblies.
https://buildingscience.com/documents/guides-and-manuals/gm-2101-guide-building-conditioned-unvented-attics-and-unconditioned
Malcolm,
That is kind of the crux of my most recent question - I couldn't easily wrap my head around how to get away from foam without framing it like a cathedral ceiling and providing an interior vapor control layer directly below the insulation. I've done cathedrals this way on a couple occasions with ridge beams, but I am not sure there would be any advantage or cost savings in this scenario given that the attic will be mechanical space and we will have a flat ceiling below.
I think I may be able to discuss some options with my truss designer to provide a tray ceiling type truss that transitions into a scissor/sloped truss - think room-in truss with really short knee walls, extra depth along below the top chord to accommodate R-49 fluffy stuff, and no central bottom chord. If I can drape an air-barrier over my top plate, I could then connect it to a warm side air barrier on the bottom side of the truss. On the interior of the air barrier, we could ledger perpendicular to the truss and install ceiling joist in place of a typical bottom chord. Extra framing, but nothing our crews haven't done before.
Or, we could just use open cell foam and an intumescent coating, which may be more practical and cost effective.
Like I said, I would love to stay away from foam, but I also want to build these houses simply and have some constraints I need to work within and be realistic about.
Thanks for the conversation.
I think you are on the right track. I would modify it slightly by starting with essentially a raised heel scissor truss shape and add a collar tie to it. Essentially building a room truss with low ceiling and sloped walls so it can all be insulted with loose fill.
You can sheath the inside of this space including the floor with zip with taped seams which becomes the air barrier.
If you select the inner slope of the truss to be 4:12 or less you can now loose fill the attic space for low cost insulation.
No spray foam and no extra roof deck. The inner ZIP does add a bit of labor but if you can limit the size of the space (ie add a tie up high to the truss to reduce room height or change the width of the inner room), it won't be all that much.
You will have to add blocking at the bottom between the truss members for air barrier continuity down to the ceiling bellow.
Akos,
I like that thought, but may use a membrane for the air barrier instead of zip, and just construct a catwalk for servicing mechanicals instead of flooring the entire attic floor. I will keep iterating and discuss options with my truss designer. Thanks for the input,
Nat
Nat,
The reason I suggested ZIP is that it can be done by the framers. This means after rough framing you have the attic space ready for the HVAC install. Might be a bit more cost up front but saves on staging as a single trade can handle the prep.
The floor doesn't need full coverage nor does it need to be taped. The blocking at the bottom does need to be well sealed though so that is where I would focus your effort.
Nat and Malcolm, yes, sorry, I should have mentioned that--by code any air-permeable insulation installed in a moderate or cold climate needs an air barrier on the interior side.
If your roof trusses will have a lot of webs in the way, getting the interior air control layer actually airtight can be pretty challenging. It's possible, but it takes effort. Open cell foam expands enough to fill in around the webs, though the installer does need to pay attention to avoid air bypasses.
Vapor control is separate from air control. I've heard Dr. Joe say more than once that although building codes allow open-cell foam in roofs because it's not air-permeable, the fact that it's vapor-permeable means that in cold climates moisture will migrate through. I believe upcoming codes will be updated with that information.
Malcolm, I see now that you had mentioned the sarking membrane system in his first comment. Josh's article is great.
I think these are my conclusions for moving forward:
1. Closed cell is not necessary if there is a vent space between the two roof decks (or roof deck and sarking)
2. Detail well for air control to mitigate moisture movement due to infiltration/exfiltration
3. Provide an appropriate vapor control layer to mitigate moisture movement due to diffusion
4. Provide adequate mechanical control in the conditioned attic to maintain appropriate/ideal temperature and humidity levels
5. Don't forget the fundamentals of structure, fire control, water management, and constructability
Unless you all think I am missing something significant, I think I've got it from here.
Once again, thanks for the conversation. I often feel like I don't have anyone to question, challenge, or affirm my thoughts and designs here in Southwest Virginia. I really appreciate the GBA community and the individuals in it. All the best,
Nat