vented mineral wool roof, is air barrier desirable over mineral wool?
I’m about to undertake a top-down reinsulation of a roof area of my house which has seriously mouse-compromised fiberglass. rafters will be reinsulated with mineral wool and topped with an uninterrupted layer 1.5″ Rockwool Comfortboard to increase R value and block through-framing conduction. then 2x vent channel nailers running up over rafters, plywood, some sort of membrane, and finally reinstall metal roofing. most of roof is cathedral, and there will be an area at peak where thicker Rockwool batts will follow the collar ties, with no Comfortboard in that area (but 2nd layer of batts laid over collar ties, so again, no through-framing).
my question is whether it is advisable to install some sort of vapor-permeable air barrier (Tyvek, etc) on top of the Rockwool to reduce air-washing of top surface of insulation. most of the top layer area will be the denser Comfortboard, but in collar tie area it will be batts. my contractor’s initial thought was that this would be an unnecessary complication and I get that. still, wondering what others think about this. if the air washing of batts, in particular, is seen as a concern, I could deepen the batts in the collar tie area as an alternative to Tyvek.
also, the ceiling should be fairly airtight – there is a well detailed 6 mil poly vapor barrier with no penetrations other than a block chimney, and the poly
will be inspected for damage and repaired as needed as part of the reinsulation.
thanks!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Roy_Settgas,
From the data I have seen by RDH, I don't think either mineral wool batts or board experience enough loss of R-value through wind-washing to warrant a separate secondary air-barrier.
In your stack-up there is a problematic layer. Comfortboard 80 compresses too much to be laid directly on the rafters without some continuous support. Unfortunately, roofs with exterior insulation end up with two layers of sheathing, or a lower one and strapping above.
thanks for that heads-up, Malcolm. do you have a sense of what sort of issue the compression of the Comfortboard creates? a plane of structural weakness? would the screws that fasten the strapping/nailers through the Comfortboard end up unsupported in a way that they would be prone to shear? seems to me that some compression of the Comfortboard where it bears on rafters would be tolerable there in terms of insulating capacity, but if there are unacceptable structural consequences, or if the compression is severe, that's a whole other issue. I'm constrained in terms of thickness I can add to this roof area and another layer of sheathing isn't real easy to fit in ...
the other approach I can think of that would avoid both compression and thickness concerns would be to horizontally cross-strap with 2x4s between the sheets of Comfortboard. then the ventilating strapping/nailers would bear on the cross-strapping rather than on the Comfortboard. I'd end up with a small amount of through-framing, but this would probably be preferable to another layer of sheathing.
Fasteners have very much reduced shear strength when passing through anything squishy. The reason is that the fasteners will tend to bend severely as they migrate through the squishy material under load. The best case scenario here is that the fasteners all bend and your surface shifts noticeably, worst case can be fastener failure. You do NOT want to put bending moments on the fasteners. One way to avoid this is to put some fastners in at an angle so that the "sliding force" will tend to pull the fastener in the straight out direction instead of sideways in shear. In this way, those angled fasteners help handle the "sliding force" so that the other fasteners are only holding the pieces together in one plane. Hopefully that makes sense.
It's better to use some hard points (solid blocks) periodically for support. That squishyness of the Comfortboard can also cause a ripply effect with anything fastened over it, since the fasteners tend to pull things in in the immediate area of the fastener, but the Comfortboard will press out on that fastened thing between the fasteners. The result is wavy-looking furring strips, for example.
Since you have a vented assembly, have you considered polyiso under the rafters instead of Comfortboard (hopefully I understand your proposed assembly correctly)? This way the foil facing on the polyiso acts as a vapor barrier to limit the amount of moisture that can get up into the roof assembly, but the vent channels can still carry away any moisture that does manage to sneak up there. This can make for a pretty robust assembly, and it will cost less than Comfortboard.
Bill
Roy_Settgas,
The mineral wool faces the same problem as foam on rafters, which is why Zip-R has not been approved as roof sheathing. There are the structural difficulties that Bill spoke about, but also the possibility of the fastener heads may end up protruding from the sheathing as the boards underneath compress under load. The live loads may be uniform like snow, or the point loads of workers.
That whole layer above the rafters needs to remain in the same relationship with the structure below over time. Movement can compromise the integrity of the metal roof above at flashing, trim, and penetrations.
Bill & Malcolm - thanks for your understandings! that all makes sense to me.
Bill - with regard to your suggestion of polyiso instead of Comfortboard, this is work all being done from top without disturbing existing ceilings. Comfortboard is top layer, on top of rafters, and my understanding is that I could get myself in trouble with condensation on bottom of any sort of foam layer on top of the insulation profile, unless it's thick enough to bring the lower surface above potential dewpoint. I'm in a cold climate and constraints on how much I can build up the roof left me concluding I wouldn't be able to make that foam thick enough to avoid bottom side condensation potential. so the breathable Comfortboard on top was my solution.
so, my updated plan is that the existing rafter structure will be infilled with Rockwool batts. a layer of 1.5" Comfortboard will go over that and, in light of your input, 2x4 purlins run between the 2' wide rows of Comfortboard. then 2x4 nailers up-down the roof to create venting channels. finally, plywood and metal roofing. this will take a bit more 2x4 lumber than my original thought of a continuous Comfortboard layer, and create a minor amount of through-framing heat loss, but in light of my depth constraints and the structural issues you've pointed out, seems like my best design choice. and the purlins will also give me nice support for my fly rafters, an issue I hadn't resolved with the original plan.
thanks again - Roy
You'll have only a very minimal amount of thermal bridging doing what you describe, since the only thermal bridges will the "+" shape areas where the horizontal purlins contact both the top of the rafters below and the bottom of the 2x4 nailers above. I wouldn't worry too much about those small "squares" of thermal bridging.
BTW, you're correct about polyiso if used above the insulation in a too-thin layer. I was thinking you wanted to go UNDER the rafters, where it wouldn't be a problem. In your assembly, insulating only from above, something very vapor open like Comfortboard is your only real safe option.
Bill
Roy,
That sounds like a good plan to me. See what the prices are for 2"x3"s. You maybe able to save a bit.