I'm trying to decide how to insulate our cathedral ceilings, cantilevers and rim joists for our new home build. I'm using ICF's on the foundation addition, SIP walls, radiant in-floor heat, as well as high efficiency windows and anything else I can do to be enviromentally friendly.

 I was looking into Waltite and stumbled across Envirofoam. They use Polarfoam 7300-0 SOY spray foam and it was around $4/sq foot vs $6.75 for the Waltite.

 I called and asked some questions and I just wanted more input in the differences. First he said it's about an R7/inch and that I would only need 4" giving me an R28. I thought the roof had to be closer to R40 up here in Calgary, Alberta but he said that would be throwing money away. He also told me it is a better vapour barrier than Waltite and it has a lifetime warranty.

 Next call was to another contractor that said he can give me R40 with the 1/2 pound for about the same cost of $4.50/ sq foot and that the 2 lb isn't worth the additional cost even though he does sell it.

 I like the idea of the Polarfoam but is R28 enough in a city that ranges from -30 to +30 degrees C? I want a tight house and I don't mind paying a bit extra if it means the best product.

Thanks for any input,
Sean

Anybody who is telling you that R28 vs. R40 for Calgary is only selling high $/R-value material (2lb foam fits that description.) Half pound foam (Icynene, et al) is a much better value in $/R, making R40 much more cost effective. The guy who sells both half-pound and 2 pound stuff is giving you the honest economics.

Half pound foam (any vendor, soy based or petro based) by itself doesn't provide sufficient vapor-retardency to meet Canadian building codes though. In the US most places allow spraying on vapor-retardent latex directly on half-pound foam, or on the interior finish wall/ceiling to lower the permeability to acceptable levels- not sure if that works for the inspectors in Calgary (it might.) In thicknesses over 2" it's a very good AIR-barrier though- it provides the air-tightess, even if it needs a little bit of help on the water-vapor issues in cold/very-cold climates.

2lb foam adds significant rigidity and toughness to the structure, and has high R-value per inch. But it makes roof leaks difficult to detect & correct, etc. It has it's uses (retro-insulating very thin rafters, basement interiors etc.) but it sounds as if you have sufficient space to fill in R-value with the half-pound goods.

If you want to superinsulate the roof, consider putting 3" of iso board on the exterior of the roof decking (or walls, for that matter). It has the same R-value per inch as 2lb foam but it's much lighter (doesn't add much to the structural load requirements of the rafters) and a fraction of the cost of spray foam. It has very little in the way of structural strength, and you'd need to long-screw/bolt a nailer layer of OSB/plywood/etc. over it to install the final roofing, but it's fairly cost-effective boostl. (It's becoming a fairly common retrofit method in the US- and for new construction too.)

Adding a true R20-R22 (3" thick iso) to an ~R40 half-pound foam insulated decking from above nearly doubles the clear-surface R-value, since it covers over the thermal-bridging of the rafters with a high R-value & uniform thermal break. (It would MORE than double the performance of a roof with R28 of 2lb foam between the rafters. ) Use fiber-faced iso, not the more commonly seen foil-faced stuff, since the foil is a powerful vapor barrier, and if you have a code-requred interior vapor barrier it would create a moisture trap. (Fiber-faced polyisocyanurate is very commonly used as roof insulation in flat or low-slope/high-span roofs due to it's light weight & high R-value, but also for it's very high operating tempeture (an issue with hot-mopped & membrane roofs in hot sunny climates). It's highly permeable- simliar to housewrap material even in 3" thicknesses.)

There are many vendors of the stuff, but here's one example of what I'm talking about:

http://www.glasscellisofab.com/sheets/isocyanurate/data_sheets/NRG.pdf

Alternatively, there are EPS products available in various thicknesses that already come with an OSB nailer-deck bonded to one side that might be used. It's only about R4/inch though- a 4" thick version would be only (only? :-) ) R16 or so, suitable for applying atop another roofing deck. (I'd use iso myself, but thought I'd mention that it's not the only stuff suitable for the application.)

sgo70,
I had the same q's as yourself for I have the same scenario as you do for my new build.
Here's what I found out.
First off insulate r-40 fiberglass, then vapor 6mil. or better, 2'' eps w/tapped seems, straping to hold it up and serve as backing for drywall.
Roof will be 6/12 and cathedral 4/12, therefor in my case I will have air circulation and room to do so.
Ottawa,On.
Cheers!

Wow, great information guys. I'll have to read into this a little more when I get home from work. The Thermapan rep is going to give me some info about doing a SIP roof, I'll keep updating what I find out.

 

Thanks again,

Sean

SIP roofs have a spotty track record- it's important to get the ridge-seams right, and there are plenty of mechanical issues to getting/keeping it right. Most SIP homes in my region use rafters or trusses, and get sealed with spray foam even if/when other insulation materials are used. There are some high-R value sips around, but price/performance wise I'm not sure they're a better value than going with a more conventional structural roof insulated from below, with layers of iso above to bring it up to whatever R-value you need.

R40 SIPs are available from several manufacturers, and it'll outperform a raftered roof insulated to R40 by quite a bit due to the lack of thermal bridging. I'm a bit leery of them at higher thicknesses due to a history of "issues", but I may be trading on old information. (A guy I work with lives in a somewhat-problematic R25 SIP house constructed in 1989- my ancient 1923 stick-built house thermally outperforms his, and it shouldn't. He also chases ant-problems annualy, a problem now allegedly solved by the inclusion of borates in the foam & skins.)

You might search the SIP forum on this site for issues/solutions surrounding SIP roofs.

Without a doubt the closed cell spray foam from polar foam is amongst the best that you can install. I would strongly suggest against using 1/2 pound foam in a cathedral ceiling, especially Icynene. If you do use half pound foam you should have a 6 mil vapour barrier that is well sealed and taped , and a 2 1/2 inch ventilation air space. We install spray foam, we were a dealer for Icynene and found it to be a very over rated product that is completely ineffective. We had a lot of issues with the company and consequently refuse to install their product. I am not sure what the code requirements are for Calgary in a cathedral ceiling, in Ontario the code requires R28 for a cathedral ceiling

The best installation is with 4.5 inches of closed cell spray foam which gives you the R28, vapour barrier is not required. The international building code allows direct application of closed cell spray foam to the underside of a roof deck, however a 2/12 inch air space is not a bad idea. If ever the roof leaks, the air path will allow the roof sheathing to dry out, also if the roof ever requires replacement, the sheathing can be removed without damaging the foam.

The thing with closed cell spray foam is that once you reach about three inches of foam you really don't gain any more efficiency by adding additonal depth or R value. R values are a good comparison for fiber based materials, but they don't take into account the air and vapour barrier abilities with closed cell spray foam. The R value is really only one number, it indicates the resistance to heat transfer, it doesn't tell the whole story. It is sort of like buying a piece of property, the vendor tells you it is a 200, well does that mean 200 sq/ feet/ 200 linear feet or even 200 acres, it does not tell you the whole story. Closed cell spray foam is the same when it comes to R value, there is so much more than just R value.

I suggest you spend the money on closed cell spray foam, the house will reward you for the life of the building. The extra cost to increase the mortgage to cover the cost of the foam will easily be covered by the utility savings. There is absolutely no comparison between closed cell spray foam and fiberglass batts. Fiberglass is an old technology that is difficult to install properly, is prone to air infiltration, air leakage and grossly over rated R values.

The best way to understand how effective closed cell spray foam is, is to think about a styrofoam coffee cup, you can fill that cup with boiling hot water, too hot to drink, but you can comfortably hold the cup with out burning your hand.. How thick is that styrofoam, maybe 1/8 inch thick. Styrofoam has an R value of about R5 per inch, now think about four inches with an R value of about R6- R7 per inch. There will be no heat transfer through that much foam.

Finally why go to the expense of building with ICF and then put an inferior insulation in the ceiling it doesn't make sense. That is like wearing a snowpants in -30 weather and a t shirt with no jacket.

i used 5 inches of closed cell foam for my cathedral ceiling ( 29 ft ) in ontario, all spray foam contractors i spoke with recommended closed cell over 1/2 lb foam very happy with my choice

Self-serving distortions from a high-density foam vendor, sez me.

First, 4.5" isn't an arbitrary number- it's related to the legal depth that can be applied in a single pass spray due  to heat buildup & fire hazard in the installation process.

Second, it doesn't create an R28 clear-surface roof structure- not even close.  With 1.5" wide breaks of R4-R5 at every rafter, the clear-surface R value is closer to R20-R22 (actual R value depends on rafter spacing.)  You can get that with 3" of iso on top of the roof deck with NOTHING on the underside of the roof deck. 

Third, once you have even 1" of 2lb. foam you have the perfect air barrier, after which it's ALL about R value, not infiltration.  (Is R6 "efficient enough? Don't think so...)  There's an argument that it's not COST EFFECTIVE WITH 2LB FOAM to go any deeper than 3", but that's due to it's significantly higher cost, and decidedly NOT because "...once you reach about three inches of foam you really don't gain any more efficiency by adding additonal depth or R value."  Once its air-sealed doubling the depth, cuts the heat loss from that surface in half, which is VERY CLEARLY more efficient.  (Of course, if you don't air-seal the house to a reasonable level and it's a real leaker, a well-insulated wind-tunnel is hard to differentiate from an UNinsulated wind tunnel. :-) )

You bet there's SO much more to it than just R-value- R28 worth of half pound foam will thermally outperform R28 of 2lb foam in a rafter installation 'cuz the depth of the thermal-bridging at the rafters ends up at a higher value.  (The bridges are now ~R8 instead of ~R5.)  Sure, for more money & material you can apply a skim coat of 2lb foam to the sides of the rafters as well to reduce the bridging, but...  it's more money & material, eh?  Yes, half-pound foam will need a vapor retarder on the interior (which could be an inch of 2lb foam over the half-pound foam, if that's in the budget), but for the raw cost of installing the sheets of poly (or sprayed on vapor retardent paints, where allowed) you BET you can end up with more efficiency per dollar than doing it all in 2lb foam.  Per unit volume, 2lb foam may be ~1.7 times the R value of the half pound stuff, but it's ~ 2.5x the cost, which makes it roughly 50% more per equivalent R-value depth, but when clear-surface bridging is included (which is the true measure of performance, not the center-cavity value) it's closer to 75% more expensive.  What does a vapor barrier cost to install?

Don't get me wrong, 2lb foam is great stuff (I use it), but don't get sucked in by self-serving claims & distortions. You'll hear similar claims from half-pound foam vendors too, about how "...you don't really need more than 6 inches..." etc., but if they're only comparing themselves to low-density fiberglass it's pretty much a straw-man argument.  Save the BS, show me the numbers.

Any time you hear "...more R value than that won't increase efficiency..." it's time to first check your wallet, then grab the manure fork...  The fine print that might support such a ridiculous statement is up to your eyeballs, and generally stinks.

Attics over flat ceilings are often the easiest part of the exterior building envelope to insulate. They are usually
accessible and have ample room for insulation. However, many homes use cathedral ceilings or have attic knee walls
that present unique insulation requirements. It is important to insulate all types of ceilings properly.

Crane Trucks

Dana1 Some rather outrageous claims from someone who doesn't know all the facts.

In the style of your response.

First, how can I be self serving, when I have absolutely nothing to gain from simply offering my opinion to a well informed consumer.

Second, 4.5 inches is not an arbitrary amount, it is also not the amount that can be installed in one pass. In fact most manufacturers recommend lifts of no more than two inches in one pass.

Third, How can you possibly claim R28 of half pound foam will thermally outperform R28 of closed cell foam, that is simply ridiculous. The thermal bridging through rafters is the same regardless of the depth of insulation in between them. Also wood has an R value of R1 per inch. Cathedral ceilings in Canada are typically at least 2x10 or 2x12 depending on the span.

Fourth one inch of closed cell spray faom is not a vapour barrier, you actually need 1 5/8 to meet the perm rating.

If I suggest to a customer that they don't need more than four inches, that is simply because I am trying to save the customer money. I am not going to rip them off by saying they need 5, 6, or 7 inches of foam, why install more foam than necessary when there is really no difference in performance beyond 3 - 4 inches. Maybe take a look at infrared thermal images across properly installed closed cell spray foam and half pound foam. Again there is no comparison between the two. The two pound closed cell spray foam simply stops the transfer of heat at three inches. The half pound foam does not perform any better than fiberglass, and comes nowhere close to closed cell foam. Yes half pound foam does require a vapour barrier. One that is completely sealed. Any small penetrations may allow moisture into the foam where it accumulates, it may eventually become saturated. Believe me I have removed enough of this stuff to know how it reacts with moisture.

Insulateright, your credibility will suffer if you don't back off on making absolute statements such as:

"The two pound closed cell spray foam simply stops the transfer of heat at three inches."

and:

"The half pound foam does not perform any better than fiberglass,..."

Both of these statements are incorrect. In the first case, heat transfer by conduction is inversely proportional to R, which is 1/U, U being the heat transfer coefficient. If you double the foam thickness, you cut the heat transfer in half for any given delta T across it. That is simple physics. Depending on the type of closed cell foam, three inches of it plus the R of the sheathing, either comes close to or slightly exceeds the R20 requirement for a wall typically required for a northern climate. In the view of many, R20 is yesterday's "adequate." For a hot roof in a northern climate, R20 doesn't even meet code.

That other statement, about open cell foam not "performing" better than fiberglass, has to be qualified. In terms of just R per inch, OC foam clearly is better than low-density FG batt, although not dramatically so, and arguably the same for a high-density batt. "Performance" for sure must cover "keeps the heat in, cold air infiltration out." That is where any foam is so much better than FG, since a few inches of even OC foam is a very effective air barrier, whereas FG is no air barrier at all.

As to depth of insulation between attic rafters, it really does matter. A more conductive material such as wood, at R1 to 1.2 for typical framing lumber, is like a conductive fin sticking into the heated space. The effective R of the wood member is not increased by adding to the uninsulated depth into the heated space. When the "whole area" R of the roof assembly is calculated, the U*A of the framing is based on the insulated depth of the rafter, not the whole depth. Actually, extra uninsulated wood attached to the insulated part makes heat loss worse, due to that "fin effect." Imagine filling in the space between the aluminum fins covering the pipe in baseboard heater sections with some insulation. You reduce the heat transfer as more of the fin surface is covered.

Sure, I'll confess to not knowing the particulars about the exact maximum depths per layer for different manufacturers, or exactly which depth crosses the Class-I vapor barrier. 

But your very statements about the thermal bridging is self-contradictory:

The portion of the stud/rafter either inside or outside of the insulation boundary offers effectively ZERO insulation (at most another R0.5- I'll trust Oak Ridge data on this one. ;-) )  The thermal bridging between the two insulation types is NOT the same- it's dependent on the depth of the wood in contact with the insulation, the full depth of the rafter or stud is irrelevant.  (A 2x100 stud will not make a thermal bridge of R100, and thus be more insulative than your 4.5" of foam!)

A wood stud/rafter  passing  through 4.5" of 2lb foam at ~R1/inch makes a thermal bridge of ~R4.5-R5

A stud/rafter passing through 8" of half-pound foam, at ~R1/inch makes a thermal bridge of ~R8-R8.5, for higher clear-surface R-value than with 4.5" 2lb foam with it's R5 bridges.

4.5"of isoboard (or equivalent-R EPS) on top of the roof deck has effectively no thermal bridges (call it R28- same as center-cavity), with a higher clear-surface R value than either under-deck spray application.

The clear-surface with the highest-R thermal bridges has lower heat loss.

There is a HUGE amount of data demonstrating that half-pound foam at "equivalent" R values exceeds the performance of low density fiberglass in ceiling/rafter applications, particularly at high delta-Ts, with the cold side up, as in a ceiling in a heating climate.  (I'll trust Oak Ridge on this one too.)

Saturation of the insulation due to vapor trapping is indeed possible with half-pound foam, no doubt, but that 's a construction layup issue.  Saturation of the OTHER materials is also possible, especially with 2lb foam, due to it's high vapor retardency, care has to be taken that it be applied to the "correct" side of the structure, and that sufficient drying capacity is designed/built into the other surface.  (Tough to accomplish in roof structures in snowy climes, to be sure.)

Thermal imaging is not the best way to measure the performance differences between a well-insulated structure and a better-insulated structure. It's sensitivity is sufficient to find the insulation gaps or large temperature differences at thermal bridges, but not enough for the very small differences between say, R20 & R40 unless there's a HUGE delta-T between in the indoor & outdoor temps.  (I s'pose you could "see" it in the images if you brought the interior temps up to 250F or something.)  But you can measure it easily in the heating/cooling bills. 

FWIW: Infra-red imaging systems are great for finding fires, but they're not high-precision thermocoulples, and only measure the radiated heat flux. Siding & roofing materials are convection-cooled- it takes a significant difference in heat flux behind them to show up in a IR image, since it requires significant differences in surface temp.  The surface temperatures of the surface materials is measured in 100ths of a degree even at the thermal bridging. The differences in surface temps you'd see beween clear-Rs of R20-R200 are extremely small- beyond the capabilities of the instrument, and below the "noise" of variations due to ground/sky radiation or variances of emmissivity of the materials (even differences in the emmissivity of paints can sometimes show a huge apparent difference) but that doesn't mean there's not much difference in heat loss between an R20 wall and an R200 wall because the IR image doesn't change much.  With normal interior/exterior delta-Ts, yes, you can see the difference between R3.5-R6 studs and R20 center cavities with IR imaging, but not the difference between R20 & R28. And painting the house with low-emmissivity aluminum paint isn't going to offer a huge insulation gain either, even though the IR image suggests otherwise.

That said, if you can show me well documented IR images of "equivalent" & high R-value (over R20)  half-pound vs. 2lb foam under identical conditions & construction to demonstrate the point, let's see 'em! (Yes, I'll look at 'em!)

r-20 of fiberglass at minus 30c is like an r-6 at the most. and insulation is only as good as its air barrier. 80% of heat loss is from air to air exchange. so why would you use an insulation that used air to insulate its self and alows air to move through it. doesn't mattter how cheap it is it gives heat the easiest way to ascape. now 1/2 pound foam is an air barier it is 3.5 and inch of the so called r- value ( its not a relavant measurement anymore of an insulation's effectiveness) but then you have to go and add a vapour barrier to that still. the slam on half pound is that it is open celled and is called that for a reason, it will breath or allow air to pass through it. now some of you might think well this guy just said its a air barrier, and i know i did but an air barrier only has to surpass a minimum air perm test. now 2 pound foam on the other hand is the caddilac of insualtion. no need for vapour barrior it meets and exceeds it with polar foam on osb with just 1 inch. air cannot pas through it becaused it is CLOSED CELLED unlike the open celled half pound and the air filtering fiberglass insulation. its seemless just like half pound but what good is seemless with a million little holes for air to travel through and out the roof in your case. and just for some topic intrest i will let you in on something that has been done in arizona. industerial freezers there use to use r-80 of fiberglass and run there a/c units at close to peak capacity. now they use 5 inches of spray foam 2 pound not half pound to insulate the same building. thats only r-33! now here's the kicker, the same a/c units that were running almost at full are running close to half capacity with only r-33 . now what i said before abour r-value not being an accurate measurement of an insulations effectiveness might make sense when r-33 of 2lbs spray foam out performs r-80 of fiberglass. doesn't make sense right? but its a reality i have to live with everyday explaining to people until a new rating system comes out that will give a real world rating of insulation.