I recently purchased a 1940 cape cod-style house originally built as a two bed one bath in the Kansas City area. In the late 80s a builder renovated the attic into a master bedroom with office and bath. The renovation included a small shed dormer, but most of the newly created living space was created by finishing the roof rafters as vaulted ceilings with knee walls. The renovation spans the entire upstairs (~40 ft.) The rafters are 2x4 timbers and are insulated with craft-backed fiber glass insulation. There is no ventilation for the underside of the decking and the roof was replaced less than five years ago after hail damage (previous owner.) I purchased the house in winter and I didn't notice the intrusion of stale air into the living spaces until the roof heated in the spring and summer sun. A family member has asthma and I’d like to eliminate all traces of the unpleasant odor and hopefully restore neutral air quality. I have had several contractors and a building engineer look at the space and generally the solutions seem to be first to demolish all of the dry wall on the vaulted areas and then either: Spray foam the entire underside of the roof and allow the knee wall spaces to become ‘conditioned’ areas hopefully eliminating the higher temperatures behind the knee walls and thus remove the odor. Alternatively, use Iso board insulation to create a vaulted ceiling that’s a sandwich of drywall -> vapor barrier->iso board -> empty channel for air -> decking -> roofing material. This would also require me to add ridge and soffit venting for the entire house. Additionally, there’s no overhanging soffit on this house so the roof needs to be cut back to allow the installation of a vented drip edge. Another solution is to tear off the roof and add a ventilation solution from the outside. This turns out to be quite a bit more expensive than the other options and the roof is fairly new. I think that I also need to consider that there are 4 48”x48” roof windows on the East side of the house and they present a barrier to the venting solution for those 8 rafter channels. The cons I see to the spray foam are expense and potential sensitivity to the off gassing (since we have an asthmatic living in the home.) Also, I’m skeptical about whether this will actually work for conditioning the knee wall voids. As for the venting, 2” Iso has an R-10 rating which is probably less than the fiber glass that’s currently there. I also haven’t seen Iso being used for this type of application when I look for examples. Either solution is fairly resource intensive so I can’t afford to implement a solution that doesn’t fix the issue. Any input or guidance is welcome. Thanks.

IRC 2012 code min for roofs in US climate zone 4 (K.C.'s climate zone) is now R49. You can't get anywhere near there with 2x4 cavity fill.

The "whole assembly-R" of 2x4 framing with R13 cavity fill is about R10, due to the thermal bridging of the framing, and that's including the R-value of the interior gypsum and the roof deck, and presumes air-tight construction. Sandwiching continuous R10 polyiso between sheets of gypsum would be a performance upgrade, since that adds up to about R11-R12 when you count the dual-layer of gypsum + roof deck. But it's still way below code. If it's 2x4 framing and you have an air space between the fiberglass & roof deck you are probably looking at R7 econobatts..

Closed cell foam as cavity fill between framing is a waste of good foam, since no matter what the R-value of the foam is, the thermal bridging of the rafters is unimpeded. The high R/inch foam budget is best applied to the exterior of the framing, where it reaps the full R value.

In climate zone 4 as long as at least 30% of the total roof-R is on the exterior of the assembly, the roof deck is moisture safe without being vented from below, and it needs nothing more vapor retardent than standard latex paint on the interior side. Both open cell foam and closed cell foam have outgassing potential that could prove to be a step in the wrong direction here. It's probably best to insulate between the rafters with R15 rock wool or R15 fiberglass snugged up against the roof deck, with only latex-painted gypsum board on the interior, and installing at least R7.5 as rigid foam over the exterior. A couple inches of EPS would deliver R8+, and since the rafters would then be thermally broken by R8 continuous foam, you'd be at least half-way to code-min performance.

If you have the budget, going with 4" of exterior polyiso would get you up to R38-ish, but a thermally broken R38, which would perform pretty close to a code-min R49 between joists or rafters. If you go that route, install a sacrificial nailer-deck of half-inch OSB through-screwed to the rafters 24" o.c. (screws penetrating the rafters by 1.5") , staggering the seams of the foam with those of the OSB, and put standard roofing felt & shingle layup. If the roof ever leaks you'll be swapping out a few sheets of OSB when you re-roof, but that's pretty cheap in the grand scheme of a 20-25 year shingle lifecycle.

Before insulating the interior side, use a blower door oar large window fan pressurizing/depressurizing the house to find and fix all the big air leaks.

If you're not going to insulate above the roof deck, blow cellulose in mesh between the rafters, then install a "smart" vapor retarder membrane such as Intello Plus or Certainteed MemBrain, then install the interior side gypsum board. (You'll have to roll the pillowed out mesh flat first.) The vapor retarder allows ample drying rates for the roof deck as it releases it's wintertime moisture accumulation, but it also slows the wintertime uptake. This gives it quite a bit if moisture resilience.

In an unvented roof assembly you can't use true vapor barriers like poly sheeting or foil facers on the interior, since that would otherwise create a moisture trap at the roof deck, putting it at long term risk of rot. It's worth chasing down a source for MemBrain (usually cheaper than Intello) if you don't intend to insulate above the roof deck.

If you want to put a layer of foam on the interior side to boost performance, use UNFACED EPS (not XPS, not polyiso, no foil or plastic facers) and no more than 3", which would still have sufficient drying capacity to also fill the cavity. With R15 rock wool and 3" of EPS you'd be at about R28 center-cavity, but it would outperform R30 between rafters due to the R12-R13 thermal break of the foam under the rafters.

Thanks for all of the info - I'm working through your response and reading up on the suggestions you've made.

Currently, I don't think I can spare the expense of replacing the exterior roofing to add insulation.

Is it possible to install continuous 1 or 2" 4x8 iso board insulation over the top of re-insulated rafter spaces? Then apply the wall board on top of that? Then the thermal bridging would be reduced to the sheet rock screws. So then the assembly would be wall board->2" iso board->insulated rafters->air channel->deck->shingles?

Currently, I don't think I can spare the expense of replacing the exterior roofing to add insulation.

Is it possible to install continuous 1 or 2" 4x8 iso board insulation over the top of re-insulated rafter spaces? Then apply the wall board on top of that? Then the thermal bridging would be reduced to the sheet rock screws. So then the assembly would be wall board->2" iso board->insulated rafters->air channel->deck->shingles?

Yes, it is commonly followed with "furring" (1x2s or 1x4s) screwed through the EPS into the rafter, with the drywall screwed to that. The screws holding the furring should be stronger structural screws. You'll be better served with permeable EPS foam rather than XPS or polyiso, which don't breathe. Unless you are using the board as a vapor barrier iself. The insulation between rafters should not have a vapor barrier (paper-backed batt, etc). If it does, this should probably be removed so there is no double-VB where moisture can be trapped. If you don't want to remove the paper, I think you can use EPS over the top, and put your furring/drywall over that. I refer to Dana for the final word. I suppose you can screw through the EPS to mount the drywall directly but it seems it would be harder to do so holding a large sheet of drywall while trying to hit a rafter through 2in of foam.

I started to close the loop on this topic by following up on the SIP installation on the exterior just to see what it entails and the costs. I just talked to the two biggest roofers in the KC area and neither work with SIP. I'm surprised. (we have hail damage often in the spring here and it would be a good opportunity to consider replacing the deck with maybe SIP if a plane of the roof needs to be repaired.)

If you go with rigid foam on the interior side it's better to use unfaced EPS than foil-faced polyiso. With foil or plastic facers on the interior side you then be lock yourself out of the possibility of filling the vent space with blown insulation and adding exterior insulation above the roof deck at some later date.  Foil facers are true vapor barriers, and roofing layups are too- the ventilation channel would be needed forever if you go low-permeance on the interior.  EPS is still somewhat vapor permeable, and at 2-3" 1.5lb density "Type-II" EPS would still have plenty of seasonal drying capacity toward the interior. At the same time if air-tight that much EPS would also be sufficiently vapor-tight to limit wintertime moisture accumulation even if you filled up the rafter bays with fiber insulation and blocked the ventilation.

Beyond SIPs, there are also nailbase panels with OSB only on one side of either polyiso or EPS, from any number of manufacturers. With simple roof designs this can be a quick & easy way to go, and unlike a true SIP, repairing a section at some later date would not require replacing an entire roof pitch. eg:

http://www.hpanels.com/

http://www.atlasroofing.com/nailable-polyiso-roof-insulation

http://www.achfoam.com/NailbaseRoofInsulation.aspx

Ok, after more research I'm becoming convinced. If I add a SIP to the exterior of the roof, do I have to tear off the current shingles, or can I add the insulation right on top of the current assembly?

the key to adding foam is that it has to be tight to the roof so cold air cannot get between the existing roof and the additional insulation. Leaving the existing shingles will add hundreds of feet of air pockets where cold air can and will enter that space & make the additional insulation meaningless. So yes, you have to rip off the shingles.

Makes sense - if the decking is in fine shape does it also need to be removed as well or can we screw the new sip directly to it (I suppose this would double the osb layer.)

Yes you can screw the sips THROUGH the decking into the rafters. the screws should penetrate 1-1/2-2" into the rafters.

I'm writing up some plans for how to divide on conquer this project in parts that I can budget for. Reading more about spray foam, I think with some due diligence we can find a product that won't cause any issues for the chemical sensitive person here. My main issue to solve by going with foam first is to eliminate the air movement and smell of the hot deck that's coming inside. Then begin adding exterior sip in sections on the roof (waiting for the hail season to pass first.)


Can you guys help me further. 1) If I go with a closed cell spray foam on the interior side - do I need to fill the 2x4 cavities (several passes and more expensive) with a greater depth of foam or can I just leave a gap between the new drywall and sprayed foam. 2) does the spray foam permanently glue the current osb deck down so tight that it can't be removed ever without breaking the foam? 3) John's Manville's core bond 3 looks pretty compelling but it's expensive. Currently I've received quotes to install standard closed cell (Icynene) for about $2.60-3.00 per sq ft (my understanding is that the application is 2" depth roughly R-14 for the foam alone) does that seem reasonable for price? Any comments appreciated.

1. Yes, but...
2. yes; you might be able to pull it off, but it may be in pieces
3. Icynene was open cell - have they added a closed cell product?

Let us know when you find a spray foam that does not off-gas; did not know such a product existed. Otherwise you may well regret using spray foam.

The installer(s) didn't mention which Icynene product - I just Googled their site to find them:


1) not sure what you mean

3) http://www.icynene.com/en-us/architects/products/product-portfolio-USA

I'm not as concerned about the installation gasses - I'm more concerned about the long-term air quality

With ANY 2-part expanding foam if the chemical mix proportions or tempertature is off during installation there can be long term outgassing relevant to the chemically sensitive (or even the rest of us.) This is not a rare phenomenon, even though most installations are fine.

Icynene sells both open & closed cell products. The closed cell goods are blown with HFC245fa, which is a powerful greenhouse gas (about 1000x CO2) that takes decades to dissipate. Their open cell goods (as with most open cell foams) is blown with water, which has low toxicity and dissipates quickly. Icycene also has a couple of higher-density water-blown foams (MD-R-200 and MD-R-210) which are basically the same chemistry as their open cell goods, but run about R5/inch instead of R3.7-ish/inch (or R6+/inch for the HFC blown closed density stuff.)

Installing a high R/inch foam between R1-R1.2/inch framing is a waste of good foam, since the thermal performance is undercut by the severe thermal bridging of the framing. If you install 2" of R6.5/inch foam in a cavity the center-cavity R might be R13, but the framing fraction (about 15% of the total area in a roof assembly) is only R2-R2.4. That means the framing is conducting about as much heat as all of the foam area. If instead you installed R13 of open cell foam (about 3.5") the framing fraction's R value rises to R3.5-R4.2, and the framing is conducting only about half as much as the foam does. Same center-cavity R-value, but a significant difference in overall performance.

Thanks Dana that's a great consideration.

Is this still the case if the remainder of the cavity is filled with glass or cellulose?

My construction is 2x4 so the entire cavity is filled in this case.