Cellulose on Top of Roof Deck
My question is would I have any problems using cellulose insulation on top of my first layer of roof decking?
I am designing a passive house in climate zone 4A following the guidelines of Joe Lstiburek’s perfect wall concept. For the walls I am going to go with a variation of the double stud wall that can be found at https://buildingscience.com/documents/enclosures-that-work/high-r-value-wall-assemblies/high-r-value-double-stud-wall-construction
My plan is to build the interior wall first and set truss rafters with no overhangs on top of the wall then run ZIP sheathing up the wall and onto the roof with no breaks. I’ve seen tons of unvented roof designs that use either rockwool comfortboard or rigid foam board. These are expensive materials. What I would like to do is put 2×6 rafters 3″ off of the face of the ZIP roof sheathing. I would put a 3″ shim under the 2×6 rafter’s every 4 feet so the truss rafters support the roof and the 2×6 rafter simply allows me to fasten a second layer of sheathing over the 8.5″ of cellulose as well as build my facia and soffit. I will have a vapor diffusion port at the ridge as well. Would this design cause and moisture problems? Am I correct in assuming that if building science recommends using cellulose outside of the ZIP sheathing on the wall than I can use it on the roof? Thanks!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Mick,
You can use permeable insulation in un-vented roofs with a diffusion port in climate zones 1, 2 and 3. Unfortunately, outside of those they don't work or meet code.
This article explains both how roofs differ from walls, and why the assembly doesn't work in colder climates: https://www.greenbuildingadvisor.com/article/a-researcher-looks-at-insulated-roof-assemblies
That aside (suppose you were in a warmer climate zone), I'm struggling to see what having the insulation above the trusses buys you? How is it any different than just insulating the trusses and avoiding the additional complications of an over-roof and two layers of sheathing?
My limited at best understanding is that applies with no insulation on top of the sheathing. What I am describing is installing the cellulose on top of the ZIP sheathing to raise the ZIP temperature and eliminate condensation because the hot humid air from in the house is stopped at the ZIP sheathing but does not condensate because it is kept warm with the insulation on top. But venting the above deck insulation as to not trap moisture. Same thing that Joe does in this roof but I am proposing to use cellulose instead of mineral wool. In this article when discussing the vapor diffusion port Joe says "We don’t “trap” any moisture with this assembly. Neat eh? It gets better. This is not limited to the south. This works everywhere – with the caveat that you need to add enough insulation top-side of the structural deck to control the temperature of that condensing surface" https://buildingscience.com/documents/building-science-insights-newsletters/bsi-096-hot-and-wet-dry
Mick,
"What I am describing is installing the cellulose on top of the ZIP sheathing to raise the ZIP temperature and eliminate condensation but venting the above deck insulation as to not trap moisture."
But you aren't venting. Venting implies removing moisture through air movement. You are proposing using a vapor diffusion port. And unfortunately they don't work, and don't meet code in colder climates.
Again leaving that aside. How would your roof practically differ from one where you sheathed the underside of the trusses or rafters, insulated the cavity and then sheathed the top as a substrate for the roofing? Same two layers of sheathing with the bottom one warm enough to prevent condensation. Same un-vented cavity filled with permeable insulation, with a roof deck above.
There would be two differences that I see. The first being that by sheathing and air sealing the underside of the truss you would still have major thermal bridging at each truss meaning that condensation would still be able to occur along the length of the truss.
The second issue that I see is that there is no convenient way of connecting that sheathing to the wall sheathing to make the air barrier continuous.
But forget the vapor diffusion port for now. Would there be any issues using the cellulose on top of the ZIP sheathing to control the condensation risk on the underside?
Mick,
What I was getting at comparing the two assemblies is that the sheathing we worry about is the one above permeable insulation. That's the one which can accumulate moisture and become a problem. What you are calling the roof deck and protecting from getting too cold isn't the one to worry about. It can't accumulate moisture in any conditions. The problem is the upper one. It can be kept safe by having thick enough impermeable insulation (foam) under or over it, but if you want the cavity to be filled with cellulose or batts, that cavity needs venting. Practically, and by code.
I was planning on having more cellulose under the ZIP sheathing as well. My understanding is that with no insulation on top of the ZIP there would be condensation. But by putting more cellulose on top of the ZIP it keeps it warm and will not condensate. The second layer of plywood that the shingles nail to will not condensate because there will be no hot humid air coming from the house because the ZIP stops it.
While doing more research I found a 90min video here: https://www.greenbuildingadvisor.com/article/the-bs-beer-show-unvented-roofs-and-fluffy-insulation
That describes almost exactly what I was describing except with only one layer of sheathing and they uses I Joists. I attached a picture. In your Expert opinion do you think I would have a problem doing this but putting another layer of sheathing on the Joists? At the ridge leaving the top 6" or so of plywood off and putting a vent there? Just like Joe did in his cold climate in the second picture that I attached.
Mick,
Every roof has a warm side air-barrier (like your lower ZIP) . They aren't enough to keep them safe.
Watch the link you posted from 26 minutes on - and especially Kohta's summing up at 36:21. He says what you are proposing is risky, he can not recommend it, and it doesn't meet code.
The first picture is of a vented assembly with a sarking membrane above the insulation used as baffles. That's completely different from what you are suggesting. If you replace the permeable sarking membrane and venting with a second layer of sheathing, then yes - that sheathing will be at risk of moisture accumulation.
I can only reiterate yet again. What you are suggesting is risky and doesn't meet code. The explanations I have offered, and links I have posted for some reason don't seem to be able to convince you of that. I don't know what more I can say.
We have a good idea what roofs work (and are code compliant).
https://www.greenbuildingadvisor.com/article/five-cathedral-ceilings-that-work
I'm baffled as to why people insist on inventing their own alternatives and building risk into their projects.
I am sorry I am very confused. At 37:17 Kohta started discussing what you called the vented assembly, and says the results were "boring and beautiful" and can say nothing but praise for this assembly he does not say they are risky or does not meet code. All I want to know is that if I put a layer of plywood (around 10 perms) between the I joists and the sarking membrane in that assembly will that retard the vapor enough to not allow it to dry and cause moisture problems?
I am sorry if I keep asking the same thing but I am just trying to understand. Thanks
Mick,
Kohta is describing the roof you posted in your first photo in post #6. That is not what you are proposing. It is a vented roof with a sarking membrane. You are not proposing a vented roof or a sarking membrane. You want to use a second layer of sheathing under your roofing instead.
The moisture that may accumulate and move through your permeable insulation (Kohta talks about where that moisture comes from in the video you posted) will find its way to the upper sheathing, and will have no way of drying without a vent channel. In warm climates a vapour diffusion port is enough to deal with the moisture, but not in where you live. That's why the code precludes having permeable insulation in contact with the roof sheathing unless you have sufficient impermeable insulation above. Even if you find my explanations unconvincing, surely that alone should be enough for you to pause and change course?
Mick, I helped develop the Ecocor assembly you're showing and I co-host the BS+Beer Show. Malcolm is correct; that assembly has a single layer of sheathing; everything above that is vapor-open and vents to the outdoors. The cross-strapping over the Mento membrane is critically important, to allow drying. The gap created by the cross-strapping is vented conventionally at the soffit and ridge.
While Bill Hulstrunk has had luck with dense-packing unvented roofs, and I have learned a lot from him, I don't know of any other building science experts who think it's a good idea and in that show I even got Bill to admit that it's riskier than using a vented assembly.
Malcolm is correct, the roof you propose (unvented fluffy over roof) won't work.
Took me a while to wrap my head around it so I can see where you are coming from.
For a robust unvented assembly what you want is the condensing plane, that is where the insulation temperature is around ambiant air dewpoint, to be inside impermeable insulation. You need impermeable insulation (ie rigid or spray foam) to prevent air getting to there and condensing. If the air can't get to this plane you won't get condensation and you have a very happy unvented assembly. I've built a number of unvented roofs with exterior polyiso and so far they have worked great.
The problem with fluffy insulation is it allows air movement, so in case of your cellulose over-roof, the condensing plane will be somewhere inside this cellulose. That means that any moisture that makes it into this space can condense and eventually saturate the cellulose. Without any drying path, this will cause your roof to eventually fail. The diffusion vent you propose will help but not guaranteed to keep you out of danger.
The solution is simple, vent your over roof. Instead of rafters, build it as purlins. Fill it with fluffy, run a sarking membrane over it (this could be regular house wrap, permeable roofing underlayment or fiberboard sheathing) then install 2x sleepers running from soffit to ridge over the purlins. Install your roof deck over this. The 2x will form the vent channel for the roof. If you use high density batts instead of loose fill, you can skip any membrane over the purlins.
You now have a happy vented roof that is guaranteed to work and is code compliant.
Akos,
Thank you. I appear to have reached the end of any useful response in this discussion.
Deleted
Akos,
Thank you for your solution. I was planning on using the rafter tails to support my soffit, facia, and the top of my second exterior wall. Is there a way to do this while still running them as rafters?
Is this not accurate?
Mick,
What should be the permeability of the middle roof deck? How well does that need to sealed? How big should the diffusion vent be?
I can guess but can't answer those. Neither will you find it in any code book. I guess you can ask Joe L if you can afford the engineering fee.
As for your earlier question. You can frame the purlins the same you would do for rafters. Depending on snow load, you stop 1 or 2 purlins from the edge and ladder frame the overhangs. The 2x on flat vent strapping can continue over this framing. If this is too thick, you can trim the bottom of the ladder framing in the overhang area as you don't need any insulation there. You would run your wall cladding up the extra couple inches to cover this step.
Akos,
Great solution thank you very much for your time!
Under the 2021 code a roof in Zone 4 needs to be R-60. Unless I'm missing something, Mick is proposing 8.5" of insulation which is going to be R-34.
Even without the code requirement, if you're going to do double-stud walls it seems like a skimpy roof relatively.
I am only talking about above the roof deck. I will still be putting more insulation in the truss bays.