Im in the process of building a new ICF home and cant decide how I want to insulate the Attic. House will be one story ranch with cathedral ceiling in half the house. No mechanical in the attic. I was quoted on spray foaming the roof deck and sealing off the ventilation. I was also quoted for spray celluose and normal venting. The spray foam is almost 3 times higher. I just cant determine which route would be most beneficial. Any input would be helpful, thanks for your time.

Posted By cpd159 on 22 Aug 2013 08:40 AM
Im in the process of building a new ICF home and cant decide how I want to insulate the Attic. House will be one story ranch with cathedral ceiling in half the house. No mechanical in the attic. I was quoted on spray foaming the roof deck and sealing off the ventilation. I was also quoted for spray celluose and normal venting. The spray foam is almost 3 times higher. I just cant determine which route would be most beneficial. Any input would be helpful, thanks for your time.

Got a Zip code? (Local climate matters- a LOT!)

Posted By cpd159 on 22 Aug 2013 08:40 AM
...I was quoted on spray foaming the roof deck and sealing off the ventilation. I was also quoted for spray celluose and normal venting. The spray foam is almost 3 times higher.

I'm surprised the difference in cost wasn't greater than that.  Were you quoted the same R value for each?  I'm not talking about the funny numbers that many foam installers throw out like 3" of closed cell foam between the rafters (~R18 in the cavity, closer to R15 overall due to thermal short circuits of the wood framing) is equivalent to R38 of any other type of insulation.  How many inches of cellulose?  How many inches of foam?  Closed cell or open cell foam?

Dana, zip code is 47933. Arkie, the quotes were as comparable as cellulose to spray foam can be.....I guess 5.5" spray foam roof deck, r-40 cellulose I just dont really understand sealing everything up, it makes me nervous. Ive read and read and its not clear to me probably because I can be dense sometimes. The hvac was to be figured with a HRV for this type of insulation sealing but I dont want problems for my family down the road. Yet I want the most efficient home I can afford...

Crawfordsville is on the warm edge of zone 5. (My brother is a Wabash alum- I've been to your neighborhood several times.)  To be code-legal in an unvented roof assembly would require R20 of insulation above the roof deck + fiber below the roof deck up to the code min, or R20 of closed cell foam applied to the underside of the roof deck, with the rest of the R in fiber, with NO interior side vapor retarders.

It's possible to cheat the code a bit and get away with it, using a 1" shot of closed cell on the underside of the roof deck, and using 1.8lb density sprayed Spider or ~2.5lb density wet-sprayed cellulose to bring the R-up to code.  The IRC prescription for R20 on the exterior has a lot of real science behind it, but requiring the use of that much on the interior side fails to take into account the vapor retardency of the foam itself, only it's air retardency.  At 1" closed cell polyurethane is on the edge of being a minimal class-II vapor retarder, and it doesn't wick liquid moisture either, so  not much moisture ends up the roof deck over a winter the way it would with only R6 ABOVE the roof deck, with air-and-vapor permeable insulation below it John Straube et al from the Building Science Corp ran a series of WUFI simulations just how protective different stackups are in different climates, summarized here. If you look carefully at Table 3, the 1" ccSPF + Spider worked just fine in Chicago, which means it should work for you too.

When you have one side of the house cathedral ceilng, the other a vented attic insulated at the floor you have an awkward air-barrier to seal & insulate, namely the interior gable-partition between the cathedralized & vented attic. That gable also has to be insulated higher than exterior walls for reasonable cooling season performance, since the vented attic might hit 120F on days when the outdoor temps are in the 90s.   For those reasons it's worth condidering doing the whole thing with 4" of rigid polyiso on the exterior, with 5-6" of fiber (or open cell foam) snugged up against underside of the roof deck for the whole house.  The installed cost of open cell foam is only about half to 2/3 the cost of closed cell at any given R-value, and is sometimes cheaper than cellulose or Spider.  In my neighborhood it works out to about 12-13 cents per R per square foot, installed. Roofing iso runs 8-11 cent/R-foot (R20 costs about $2 a square foot for the material), and there will be some scrap rate- depends on the number of dormers & valleys you have to contend with, and there is a labor cost, so it'll usually come in at ~15 cents/R-foot installed- more if there is a lot of cutting and high scrap rate.  Closed cell foam usually runs 18-20 cents/R-foot, installed.  YMMV.

But it's hard to come close to the 3-5 cent/R-foot economics of deep cellulose on the attic floor, and the other option is to lose the cathedral ceiling concept, go all vented attic, meticulously seal the attic floor and drop R60-75 in there.

Don't worry about sealing up the place too tightly- you need to worry about NOT sealing up the place sufficiently!  Random air leakage is a bigger cause of high indoor mold spore counts & building component rot than any air quality issues you get from super-tight homes.  And don't count on the insulation to be a primary air sealant, since it doesn't address the long skinny leaks between framing top plates, or between the bottom plate & subfloor etc.  ALL houses are better served by active ventilation systems like HRV, even if they're pretty leaky, since the air leakage isn't necessarily where the ventilation is needed, and not always the best quality air, since it may be coming through soil, moldy studwall bays, dirty garages & basements, etc.   Code-min air tightness under IRC 2012 is three air changes per hour at 50 pascals pressure (3ACH/50) a level of air tightness that pretty much guarantees you'll need to have a ventilation system, but that's a GOOD thing.  PassiveHouse max leakage is 0.6ACH/50, for energy use reasons, but it's hard to rationalize the amount of detailing you have to do to get there unless you are needing to meet the performance spec. For an IRC 2012 code-min house 3ACH/50 is plenty tight, and not too hard to build.

Dana, yes Crawfordsville is the home of Wabash College, a real good school. I am past the scope of losing the vaulted ceilings as the truss package is been ordered. I looked again at the quote from the spray foam company. This may mean something to you but I'm really not sure. LD-C-50 at 5.5" nominal to roof deck, 3.5" at gable ends and 3.5" devising wall between house/garage The other quote was R-40 stabilized insulation in the entire attic space. The vaulted ceiling he believed he could get into and spray the stabilized insulation as well. Spray 9,161.35 Stabilized 4,125.00 House is 1920 sq ft ranch with gavel ends ICF construction from the footer to the roof truss. Single story ranch on full basement

Just as I suspected. I googled LD-C-50 and that appears to be Icynene Classic open cell foam - see link below. The stated R value is 3.7 per inch. At 5.5" thick, you are only looking at R20 in the cavities between the truss chords. Probably closer to R18 overall factoring in the much lower R value of the wood truss members. At R40 for the cellulose, you are getting twice as much R value with the cellulose for 1/3 the price.

http://www.icynene.com/architects/products/product-portfolio/classic-ld-c-50

LD-C-50 is a half pound polyurethane. Half-pound foam (or stabilized cellulose) on the underside of the roof deck without adding R20 above the roof deck would be a code violation under the IRC, since half-pound foam (and cellulose) are VERY vapor permeable, and the seasonal moisture cycling at the roof deck would be intense.

If you go that route use a "smart" vapor retarder such as JM MemBrain or Intello Plus on the interior side to limit the rate of moisture uptake over the winter when the average roof temps are cold, but releases that moisture quickly when the roof temps warm up in spring. If that's the plan, plank sheathing is the best bet, OSB is the worst, since moisture adsorption rates and mold growth capacity of solid wood is dramatically lower than OSB. Plywood is somewhere in-between. Whatever you do, DON'T put sheet polyethylene anywhere in the stackup!

http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_8_sec006.htm

If the contractor can "dense pack" the cellulose into the cathedralized ceiling for less money than the Icynene, that would be more protective of the wood, and have better cooling-season performance to boot.

Though it's at the grey area of the code, cellulose or open cell foam with a smart vapor retarder (detailed to be air tight) on the interior is a relatively safe assembly.

But there's more:

Having just R20 between rafters in a cathedralized ceiling would also be a code violation under IRC 2012 (and even IRC 2006) in US climate zone 5. If your current code is based on IRC 2006 or IRC 2009 you need R38 as a minimum:

http://publicecodes.cyberregs.com/icod/irc/2009/icod_irc_2009_11_sec002.htm

http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_11_sec002.htm

If you bumped your rafters in the cathedralized ceiling section to 2x12s and did a full cavity fill of dense-packed cellulose or open cell foam you'd meet the letter of the code on R-values at least. That's 2x the amount of foam though, likely close to 2x the price. It would be cheaper and better to put 1-2" of CLOSED CELL foam (MD-C-200, if Icycnene) under the roof deck or better yet, 3-4" of 2lb semi-open cell (MD-R-200), with damp-sprayed stabilized cellulose for the rest, and skip the interior side MemBrain. And while that too would be technically a code violation, it would be even less likely have issues. At 4" thickness MD-R-200 is about 1 perm, a minimal class-II vapor retarder, but even 3" would be protective enough. MD-C-200 is about 1.2 perms @ 1", but ~0.6 perms @ 2", the low side of having any drying capacity. The closed cell foams are also blown with HFC245fa, which has a very hefty environmental footprint (but worth it for its other properties in this stackup), whereas MD-R-200 is blown with water, just like LD-C-50, and fairly low-impact.

Putting only R13 in the partition-gable is on the skimpy side- it's below code min since from an insulation point of view it's an exterior wall. To meet code min you could add 1" of rigid foil-faced polyiso on the attic side of the assembly, but for cooling season performance 2" is better. (The attic is a lot hotter than the outdoor air.)

With an R20-22 ICF walls in a zone-5 location it makes long term sense to go with ~R50 (IRC code-min) at the cathedralized roof- R60+ in the attic floor cellulose. In zone 5 it also makes sense to put 2" of EPS (R8) under the basement slab (R12-R16 if it's a radiant-floor), butting the foam up against the interior foam of the ICF, floating the slab.

Dana1 , I'm putting 2" under the basement slab and still undecided on heating the floor. I'm going to have bedrooms in the basement and don't want to make it uncomfortable. As of now I have a 5" slab figured with the tubing to go in but not a heat source for it now. The HVAC engineer designed a system for the entire house with the floor not being heated, I would only use it so the floor would not be cold on my kids feer because we are going to stain the concrete floor and may not cover it with flooring. As far as my insulation I really appreciate all the input. It's a shame the insulation contractors don't let me know when they are not not using code requirements. Maybe I'll just go old school and vent the roof and fill the attic up with blown in cellulose. I'm so confused on all this.

I just received a third estimate..... Here's a copy I want to offer you a hybrid quote that I would use in my own home. This uses three inches of closed cell foam for the vaulted ceiling portions of your home and R-49 cellulose in the flat parts of the attic along with 2 inches of foam on the kneewall at the transition from vault to flat attic. This would be installed for a quote of $7,450.00. The alternate is to proceed with foam on the entire roof deck which yields a quote of $10,100.00. Spray-in cellulose is a good choice for the garage walls since it is a good air seal and has a high fire resistance rating. The cellulose would also be used in the attic of the garage to give an R-value of 19. Total Garage quote is $1,250.00. An alternate of fiberglass blankets in the wall portion would lower the quote to $780.00.

cpd,I am putting cathedral ceilings in as well and am wondering what the best thing to do is as well.

"This uses three inches of closed cell foam for the vaulted ceiling portions of your home"

Is this contractor saying to use ONLY 3" closed cell?

"In zone 5 it also makes sense to put 2" of EPS (R8) under the basement slab (R12-R16 if it's a radiant-floor), butting the foam up against the interior foam of the ICF, floating the slab."

I am planning 4" EPS under garage and basement both.

"With an R20-22 ICF walls in a zone-5 location it makes long term sense to go with ~R50 (IRC code-min) at the cathedralized roof- R60+ in the attic floor cellulose."

"It would be cheaper and better to put 1-2" of CLOSED CELL foam (MD-C-200, if Icycnene) under the roof deck or better yet, 3-4" of 2lb semi-open cell (MD-R-200), with damp-sprayed stabilized cellulose for the rest, and skip the interior side MemBrain."

Hmm seems the suggestion is to not meet code. I am curious how we meet code R-38 in Z5 without breaking the bank as well. I was planning a 3:12 light colored metal roof in Chicago (that faces due N) so I am worried about this. However I struggle to see the validity of the mold issue in the article you cited in BuildingScience:

http://www.buildingscience.com/documents/reports/rr-1001-moisture-safe-unvented-wood-roof-systems

It is dry in Chicago in the winter. Very dry. When they state moisture levels are these inside or outside the home? R38 batt is code yet that is also untested. How does R38 batt perform?

I see references to R49 batt as a high density or high performace batt that is not as thick but can't find a product anywhere. I have TJI 14" rafters, so I can go to possibly 13" leaving 1" vent space. I might double up R15 and R30 for 13" of HD batt = R45...

I'm saying if you're going to violate code by putting less than the IRC minimum R between the underside of the roof deck and fiber, do it with 1-2" of closed cell foam or 3" of the 2lb semi-open goods from Icynene.

With 3" of R7/inch closed cell on the underside of the roof deck you have sufficient low-permeance-foam- R to take the total up to R50 on the underside of the roof deck, but it is WOEFULLY shy of a reasonable total R-value if that's all there is, since the thermal bridge at the rafters is now a performance-robbing ~R3.5.

At 4" of the 2lb semi-open Icycnene it would be also fully code-complaint on the IRC prescriptive foam/fiber ratio, and provide better drying rates for the roof deck than the 3" closed cell solution. (It's probably cheaper too, or at least comparable in price.) Even 3" wouldn't present a problem (since it's only 1.3 perms @ 3"), but wouldn't quite meet the letter of the code.

In the RR-1001 document the only relevant air moisture is the interior RH. The outside humidity has no relevance in an unvented assembly. There is no way for roof decks to dry toward the exterior through ~0.1 perm felt & asphalt shingles that are covered with rain, ice or snow half the winter. Even in summer the shingles & felt will retain moisture between them after rains, and you can't dry through a puddle, no matter how thin, let alone a layer of very low permeance material.

With a light colored metal roof you can still do OK if you use only #15 lb felt on the roof deck (not #30), and mount the roofing on 2x purlins so that it's at least cross-vented between the roof deck and metal, giving it substantial capacity to dry toward the exterior. But the comparatively low 3:12 aspect doesn't give it a huge convective drive for purging moisture from the purlin cavities. But mounting it snug to the deck with a slip-surface underlayment won't cut it unless you take the interior side down to a half-perm or less (3" of generic closed cell is about 0.4 perms), which is also a borderline moisture trap since the metal is zero perms, and would normally have ~0.2 perm or less to the underlayment layers.

The primary critical points to consider are the average winter temperature of the roof deck compared to the dew point of the interior air, and the vapor retardency of the intervening layers. Whenever the roof deck is below the dew point of the interior air it will adsorb moisture at a rate determined by the vapor retardency of what's between them. If you put a poly vapor barrier on the interior any moisture that gets in never gets out, but if you leave it at only 3-5 perms (latex paint, with cellulose or open cell foam cavity fill) the rates are high enough to be a problem for the roof deck with 35%RH 70F interior air (dew point ~40F). If you reduce that to about 1-1.5 perms (either with MemBrain or 1" of closed cell foam or 3" of MD-R-200) the roof deck is going to be OK, in that climate.

The other thing to consider the stackup is the average winter temperature at the foam/fiber boundary, since that's the "first condensing surface", and it won't wick or adsorb moisture quickly. If the average temp is too cool that's minor air leaks from the interior side into the cavity are potentially a problem for the fiber. This is where cellulose really rocks, since unlike fiberglass or mineral wool it can adsorb water into it's hollow fiber structures without losing effectiveness or sustaining damage. But at sufficiently high density (say 1.8lbs/cubic foot) the new-school blowing wools like Spider or Optima are air retardent enough that it won't be much of a problem either/

For any batt solution code would require both a minimum 1.5" vented air gap between the fiber and roof deck, AND at least a class-II (<1 perm) vapor retarder on the interior. Low density R38s are junk, highly air-permeable, and suffer convective losses of performance at the wintertime temperature extremes. High density "cathedral ceiling" batts or rock wool batts are fairly air retardent, and actually GAIN performance (as does cellulose) at the temperature extremes. But any batt solution has fit issues that must be addressed and the framing is never perfect either. Even the best batt installations have at least some thermal-bypass gaps and a few compressions, whereas blown/sprayed goods always conform to every anomaly, for near-perfect fit.

Most folks in zone 5 meet R38 code min with unfaced low density batts between 2x12 joists, or lower profile higher density cathedral ceiling batts between 2x12 rafters using soffit-to-ridge venting under the roof deck. Some times it even performs OK, but it's not a high performance solution, and R38 is a pretty tepid lid to put on a mid-to-high performance enclosure, in an area where R49 is likely to become code-min by the time you're ready to sell the place. An R38 batt with a kraft facer on the ceiling side snugged up to the roof deck without the vent gap is a code violation that would have significantly worse moisture handling performance than any of the assemblies simulated in the RR-1001 document, since it's 100x as air permeable as 1.8lb Spider or 2.5lbs cellulose. The kraft facer runs ~0.4 perms, and makes it a near-moisture trap, though when it's really soaked it'll be more vapor permeable. Experience suggests you really DO need the vented gap (or an additional exterior R20) to be safe with R38 batts.

The better performance solution for hitting R50+ with the metal roofing would be to put R20-R24 in paper or fiberglass faced rigid roofing polyiso (3.5-4" thick) above the roof deck and R26-R30 cellulose snugged up under the roof deck (a full cavity fill on 2x8 or 2x10 rafters), no interior vapor retarders, and mount the metal roofing on purlins above the foam, holding the foam in place by through-screwing the purlins to the rafters with pancake head timber screws (eg FastenMaster HeadLok). Meets code, roof deck can dry at a high rate, is protected by the moisture buffering capacity of the cellulose, and every body is happy. R20 polyiso runs about $2 per square foot (10 cents per R-foot) f.o.b. the distributor's lot, and installs pretty easily on simpler roof lines with crews who know how to do it.

Dana1, thanks for the input but to be honest I'm more confused now than I was. Well I mean since I understand a little more about insulation now I'm disappointed my subs have estimated less than minimum code specs. Could you provide me a recommendation of what I should request? I could even email you a set of PDF blueprints of my house.

If you have to put the shingles, on and have not had it done yet, have you looked into a rigid foam on top of the roof deck between the deck and the shingles? Seems much cheaper to go that route and simply use cellulose on the underside of the deck. I will probably have this done when my roof needs replacing.

For any batt solution code would require both a minimum 1.5" vented air gap between the fiber and roof deck, AND at least a class-II (<1 perm) vapor retarder on the interior. Low density R38s are junk, highly air-permeable, and suffer convective losses of performance at the wintertime temperature extremes. High density "cathedral ceiling" batts or rock wool batts are fairly air retardent, and actually GAIN performance (as does cellulose) at the temperature extremes. But any batt solution has fit issues that must be addressed and the framing is never perfect either. Even the best batt installations have at least some thermal-bypass gaps and a few compressions, whereas blown/sprayed goods always conform to every anomaly, for near-perfect fit. Most folks in zone 5 meet R38 code min with unfaced low density batts between 2x12 joists, or lower profile higher density cathedral ceiling batts between 2x12 rafters using soffit-to-ridge venting under the roof deck. Some times it even performs OK, but it's not a high performance solution, and R38 is a pretty tepid lid to put on a mid-to-high performance enclosure, in an area where R49 is likely to become code-min by the time you're ready to sell the place. An R38 batt with a kraft facer on the ceiling side snugged up to the roof deck without the vent gap is a code violation that would have significantly worse moisture handling performance than any of the assemblies simulated in the RR-1001 document, since it's 100x as air permeable as 1.8lb Spider or 2.5lbs cellulose. The kraft facer runs ~0.4 perms, and makes it a near-moisture trap, though when it's really soaked it'll be more vapor permeable. Experience suggests you really DO need the vented gap (or an additional exterior R20) to be safe with R38 batts.

First off, your responses are very appreciated Dana. I am sure I speak for many. Thank you for sharing your knowledge.

The metal roof wil be on 2x purlins, on plywood sheathing (which you don't mention but is required since the polyiso is not sheathing). Then there will be TJI14s as rafters with a vent gap. I am considering stacking R30+R15 HD batt with a polyiso 1" board inside on the ceiling with a vapor barrier. While not a R60 roof, I think this will be a very good roof - much better than current building practices by far.

If I put polyiso on top, I would HAVE to seal it so the thermal break of the polyiso doesn't get wasted with a ventgap. This whole thing will depend on my budget and willingness to do the work. I will likely do the roofing deck myself if i choose the R20 Polyiso route.

I just found these, FS25 batts for exterior walls R-25 come in 5-1/4" for a 2x6 stud and come in 16 and 24 inch widths..so I can use them for the walls as well as the warm side batt for the cathedral two-batt stack...will have to compare their costs...


http://www.thermafiber.com/Portals/0/PDF/FS-15%20and%20FS-25%20Data%20Sheet.pdf

With a light colored metal roof you can still do OK if you use only #15 lb felt on the roof deck (not #30), and mount the roofing on 2x purlins so that it's at least cross-vented between the roof deck and metal, giving it substantial capacity to dry toward the exterior. But the comparatively low 3:12 aspect doesn't give it a huge convective drive for purging moisture from the purlin cavities. But mounting it snug to the deck with a slip-surface underlayment won't cut it unless you take the interior side down to a half-perm or less (3" of generic closed cell is about 0.4 perms), which is also a borderline moisture trap since the metal is zero perms, and would normally have ~0.2 perm or less to the underlayment layers.

If I use a ss metal roof that has built in ridges/folds rather than a flat profile along the bottom will that eliminate the need for the 2X purlins?

I ask b/c with the solar PV and solar thermal panels up there, installation and inspection will have people walking on the roof and with 2x purlins there will not be support under their feet.

With TJIs you have an even bigger fit issues when using batts, since batts are typically designed for the nominal 1.5" thickness timber dimensions, not TJI webbing. Almost every high performance house I've seen with TJI rafters has used blown, not batts, and with good reason.

The reason for the purlins is to give 1.5" of guaranteed vent space, just on the top side of the stackup rather than the underside. Any water that blows through the seams can then dry quickly. Anything less than a 1.5" would be very iffy, particularly on low-angle roofing.

Try to avoid true vapor barriers- they're a double edge sword, since they fully block all drying in one direction, and you have to then design it fairly vapor-open on other side of the stackup, which may or may not have seasonal moisture issues. With the IRC prescribed R20+ on the exterior it's fine it the R20 is vapor-tight, but that means you not only can, but SHOULD use nothing more vapor retardent than latex paint on the interior side.

Fiber faced roofing iso is usually 2lbs density, and can be walked on (carefully.) The foil faced goods are typically 1.5lbs density and not-so-walkable.

The racking systems for the solar should rightly go up as the roofing is installed, or you can add a less-structural deck above the purlins with the proper slip-surface underlayment for the metal roofing, leaving that deck fully vented to support walking on the metal roofing.

"With the IRC prescribed R20+ on the exterior it's fine it the R20 is vapor-tight, but that means you not only can, but SHOULD use nothing more vapor retardent than latex paint on the interior side."

This is the only logical and practical conclusion in most of our work, both new and renovation. Depending on a perfect wall assembly is to depend on the impossible. It is fairly easy to build and verify the quality of an exterior installed Thermax for instance, and once done leaving the interior to the vagaries of normal construction techniques. The same hold true for foam, though with less certain quality control.