Questioning this Exterior Rigid Foam on Roof Assembly
I have finally found a roofer after trying to get one all summer but he is questioning our original plan and is concerned it isn’t safe.
I have an old cabin that currently has a 2×6 t&g ceiling that is also the roof deck. It has very low standing space on the second floor interior and I do not have any room to add insulation on the interior side. The cabin is fairly small, 750 square footprint, with another 400 on top and a cathederal ceiling/loft situation. Zone 6b with a very long winter in the mountains at 6200′. From the interior out is:
– 2×6 t&g deck
– tar paper (very old and crumbling)
– 3 layers of rigid insulation
– metal roofing
The roof is a gambrel/barn style roof, with about 4′ on top at a 20 degree pitch, and then 10′ at a 55 degree pitch. The way the structure is built up from the roof deck is funky (to me anyways, not a roofer but used to be a carpenter). It is:
– 2×2 running horizontally across the roof deck
– 2×4 (except ripped down so actually 1.5″ x 3″) running vertically, attached to the 2×2.
There are three layers of 1.5″ rigid foam cut and cobbled between this lattice, and held in with 1×4 strapping running horizontally with the roofing metal screwed to that.
The metal is in bad shape, so I figured as long as it was coming off I should add more insulation to the top. I bought a bunch of reclaimed foam, 2 pallets of 4.5″ and 2 pallets of 1.5″, enough to do the entire roof with 6 extra inches (I was originally only planning 4.5 but picked up the 1.5″ for next to nothing on craigslist). The plan the roofer and I were originally discussing (that he’s saying won’t work as the date for work nears) was going to be:
– Remove the metal and 1×4 horizontal strapping.
– Build up 6″ of foam.
– Add plywood deck with ice and water shield
– have 2×4 purlins running vertically (for an air gap and to screw metal to) and screw through them, all the way down through the deck and into the beams. From the top of purlin, to achieve 2″ of embedment into the beam would be a 16″ screw.
He is concerned the screws won’t have enough holding power on the steeper pitch since they will be going through 10.5″ of foam before they hit wood and is also questioning the added complexity and cost. Another option I was considering is to hit the 2″x3″ vertical member which would lessen it to the screw going through 6″ of foam before hitting wood, and also could reduce to a 12.5″ screw if we embed all the way into that 2×4 but then we’d really have to hit the mark. If I had to ditch 1.5″ of foam it wouldn’t make me too sad as I didnt pay much for it, so that could possibly reduce the complexity here?
Additionally, I have read a few comments on the long comment thread on the “how to install rigid foam on top of roof sheathing” article, and I am wondering if the fact that I do not have an air barrier between my roof deck and the foam is a problem? Does this roof need to be taken down to the 2×6 deck and re-started from there with a peel and stick membrane? Any help is really appreciated. Thanks.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Hi Jacob, I'm giving your question a bump in hopes a reader with first-hand experience will weigh in. It just so happens I was talking with Jordan Goldman from Zero Energy Design yesterday about detailing exterior rigid foam--it is part of their go-to wall and roof assemblies. I will also be talking with the builder, Mark Doughty, who manages their projects; I will see if he has any thoughts on your situation. On another note, good for you for scoring recycled foam from Craigslist. People regularly recommend going that route but I don't often hear of people having done it.
Thank you Kiley, really appreciate it. One of the longest comment chains i've seen here, spanning from early 2015 to this month is on the topic of exterior rigid foam on rooves. I have read all the comments and there are lots of useful ones there, but exterior foam with no interior insulation is rarer, and over 6" seems to be very rare. I posted on that thread back in June but didnt catch anyones attention.
I will ping Michael Maines for you--I bet he would have something useful to say.
Jacob, I had a similar experience with a roofer a couple of years ago on a renovation I was building. Months earlier they had said they could do my assembly, but when they got to the jobsite and saw the printed details they decided it was too complicated. After my crew and I built it, I understood why—it’s more like fine carpentry than what roofers usually deal with.
On another project that I designed and someone else built, we had a very similar situation to yours, with an exposed timber frame inside a gambrel roof and the same build-up that you have, except it had full sheathing and asphalt shingles instead of metal roofing. Our approach, to save time, money and other resources, was to strip the roofing, cover the sheathing with a Siga membrane connected to the other new air control layers we added at the walls, and run new 2x furring horizontally and 2x6 “rafters” vertically 24” o.c.. We placed 6” mineral wool batts between the new “rafters”, leaving a vent space, then sheathed and installed new roofing. I’ll add a couple of photos and a drawing. The drawing is not my best work, it's more of a sketch, but it might (or might not) help explain what we did.
If you want to stick with your general approach, I think it could work but it’s very hard to hit a small target through that much foam. Some screw manufacturers, such as Heco Topix, offer engineering assistance to design a screw pattern that can handle vertical loads. When the load exceeds the bending strength of the screw, they specific some installed on an angle to create a truss effect. In your case that could be some very long screws, and even harder to hit your target, but from an engineering standpoint it's possible to do what you planned.
Thank you for the reply Michael. To clarify - you stripped the roofing down to the very bottom deck, i.e. the one sitting on the exposed beams, or just stripped the shingles, put on a membrane, and then built that roof structure over top?
Jacob, yes, the contractor kept the existing foam and sheathing in place.
> the screws won’t have enough holding power
SIPs don't use screws all the way through, so perhaps an adhesive would be useful?
> plywood deck with ice and water shield
Consider vapor permeable underlayment over most of the roof to allow some upwards drying into your air gap - avoid a moisture-trap design. Also make sure that interior air can't leak upwards far enough into the assembly to condense.
How often the cabin is heated also has a big effect.
My concern is the "poor condition" tar paper is apparently the vapor barrier from inside to the cobbled foam assembly...
What you are proposing is a complicated build that would be outside the skills of most roofers. About the only way to pull this off is to split it up into pieces that a roof and a framer can handle and have each only do the part they are good at.
The most important part for your type of roof is getting a good warm side air barrier.
This means the ice and water layer should go directly over the T&G to seal it up.
You also need to deal with the gaps between the T&G boards where they pass over exterior walls. The simplest way I've found to seal these up is to drill a 3/8" hole from the outside over the wall top plate and inject spray foam (see attached photo). This step should not be skipped as all those T&G grooves add up to a pretty big hole and a very large air leak.
Once the roofer covers the T&G with ice and water, your house is under roof again and can take a bit of time to install the rest of the layers.
The simplest is to nail some 2x3 or 2x4 on flat over your rafters (careful to hit the rafters so the nails are not visible from bellow) and fill the space between the 2x with cut and cobble strips of your 1.5" foam.
Over this install the 4.5" foam. The new deck can now be screwed down with 6" screws into the 2x on flat bellow. The 2x on flat are a much wider target so it is easy to hit it even with the long screws.
Cover with a synthetic underlayment and install your new roof over it.
Make sure to picture frame around the outside of the roof with 2x ripped to the thickness of your foam and roof angle. You don't want any spots where the foam is exposed and critters can make their way in.
Your roofer can do the peel and stick plus the final underlayment and roofing. You want a decent carpenter to do the picture framing, the 2x strapping and the new roof deck install.
This is a great post.
Is there a reason to do the picture framing rather than just letting the fascia cover the foam? Or are you essentially considering the picture frame the sub-facia?
The picture frame does become the sub facia. You also always need something something solid to nail things into.
Thick foam roofs are ideal spot for critter nests. Anything you can do to protect the foam is worth the effort.
Thanks for the reply - I recognize it's not what they usually do, it's just driving me crazy that he made it seem like something he could do and now he's saying he won't warranty any work he does in this fashion and doesn't think it'll hold. Back when I worked as a carpenter, I was on a job where we used very long screws once for a rather heavy corten steel facade and I acknowledge hitting studs over a distance can be tricky.
Can you expand on the importance of the warm side air barrier please, this is a point I am a bit confused about. After reading every article and comment thread I could find on this topic I was originally thinking that this roof would come down to the deck and I would do this, but the roofer in our first several discussions convinced me this wasn't necessary and to just ice and shield on the deck above the purlins (for water proofing, I know this isn't an air barrier). I understand that from a air sealing perspective, onto the t&g is the best, but how important is this from a "safe" (i.e. won't mold/condensate) perspective? Warm air will be traveling out the t&g, but there will then be many layers of seams staggered rigid foam, and then there will be a gap between the top of rigid and bottom of new roof deck due to the purlins. The idea of the ventilation gap is to wash away the warm air and greatly reduce/eliminate ice dams right? So bottom t&g roof deck will be kept plenty warm and will receive no condensation there, and the air shouldn't be able to travel all the way up to the top deck in a manner that would cause problems. What am I missing?
If I take the roof down to the deck, I would guess that the rigid foam that is in there currently will probably not be worth re-using/ we won't be able to. It is probably around 25 years old and brittle. If I go with the route of building up on top of the existing structure I will still have the benefit of the 4.5", de-rated as it may be. Will the air sealing benefits outweigh the extra 4.5"? Or would it simplify this whole thing if I just tear the inside structure out, air seal, build up 6" of reclaimed foam, thus reducing the complexity of the project.
As for the wall connection - there are a few different things going on. There is an area where the t&g hits a 2x4 sitting on the top log. I have furred out 2x4 walls on the interior of the log walls and I plan on spray foaming that connection. There is another area where there is a loft where there is a floor that is t&g as well that extends past the log ends and the roof t&g intersects with that. You can see light (I seriously can't imagine how cold this house was prior to this reno) so what I did was frame pony walls on the interior (as it's dead space really anyways since the roof intersects with it at a 55 degree angle) and plan on spray foaming that as well.
In heating dominated climates, interior air leaks are your biggest source of moisture in the roof.
For this type of structure, interior air leaks are also your bigger source of heat loss. Your existing ~R20 roof assembly is actually not terrible, going up to an R50/R60 roof will save you less energy than controlling the air leaks.
The problem with T&G is you get 3D air flow paths. Air can leak in through the grooves in the boards and flow along the grooves outside. Spraying the bottom by the walls will help but won't stop all the airflow. It might not seem like a lot but 3/16" gap x 30 boards x 1.5" thick board is big hole.
Since these air leaks are straight into the cold soffit area, you can get condensation and drips on the bottom of your ceiling near the walls.
So before adding any extra rigid insulation above, you need to figure out how to air seal the ceiling. You can seal above the 3" EPS but you also have to run this air barrier down to the spray foamed mini walls you are building.
If you tape all the seam of the top layer EPS (or peel and stick over the whole thing) and if you can cut through the many layers of foam along the perimeter and also spray foam from the top all the way down to your mini walls you should be able to get a reasonable air seal. You would still have to do drilling+SPF to seal all the grooves between the boards as well.
The 4.5" iso could now go over the existing insulation and top with new metal roof. There is not much point for the extra 1.5" though.
I would still be temped to take everything down, do a very thorough job air sealing and install the 1.5+4.5 of iso.
Again, thank you very much for your help. I am leaning more towards taking the roof down to the deck because the roofer has now informed me that he won't even have a carpenter available for the job. This job is scheduled for a month from now, so what I am considering now is seeing if they can come do the tear down and put down ice and shield. I could then take on the carpentry myself and go with the cut and cobble of the 1.5" foam between the 2x4 on flat and 4.5" over that as you described.
I am struggling a bit to wrap my head around what you're describing for air sealing the t&g over the walls, however. Would you drill a hole at each t&g joint above the wall from on top of the roof once everything is stripped to the deck and hit it with spray foam?
The proper way to air seal a new build T&G ceiling is to stop the T&G half way across the wall top plate, carry the peel and stick above the roof down to the wall plate and out to the sheathing for air barrier continuity. The T&G for the soffit then continues from the half way point at the wall plate out.
In older construction the T&G is usually all that is holding up the soffit, so this most likely won't be an option and way too much work anyways, thus my suggestion of holes+foam.
Drilling a hole at each T&G joint plus foam sounds tedious but it is actually pretty quick. If you don't have it already, get a gun for the canned foam, makes this type of job much easier.
Most reclaimed roofing polyiso is not waterproof. Take care to protect the iso when it is up on the roof in case it rains. The water won't make it into the house through the ice and water but the iso can get saturated and it takes forever for it to dry out.