Thanks Dana for your detailed reply.

Just for clarification, I'm in FL. Do you still think that's not enough R value for my area.

I'm a little surprise to hear you say use foam on the underside of the roof deck AND cellulose in the ceiling. Am I understanding you correctly because I thought that was a no no? i was planning on foamimg the underside of the roof deck only. But I am concerned that it will not be enough R value. The company claims it's R 4 per inch for the open cell foam.

It's already hot as heck here and I don't want to be under insulated. thanks again

Sorry for the identity-confusion (so many have chimed in I'm losing track of the characters- I need a chart! :-) ).

R4/inch is an overestimate, but isn't too far off. 3.6-3.8/inch is the typical range for half pound foam- I wouldn't count on performance better than R17 at R4.5. (And yes R16-R17 is probably enough in FL, especially if you use light colored roof cladding or exterior-side radiant barrier to further reduce the cooling load.)

In FL the humidity profiles in roof/walls are flipped from what it is in NY/NE, with the humidity drive primarily from the exterior, not interior. IIRC, in New England the ratio of foam/cellulose R-values have to be somewhat balanced if the cellulose is also up between the rafters, but can be much more relaxed if it's open-blow. I'd have to think about it (maybe even do the math) to know if/when cellulose on the attic floor in FL it becomes a problem with a foam-sealed attic. (My first reaction is that it doesn't, but I've learned not to necessarily trust my gut on these issues in hot-humid climate areas, where I haven't lived & worked much.) If the ceiling isn't vapor-sealed and has reasonable inward-drying capacity I'd think that with minimal outside air infiltration into the foam sealed attic the humidity profile through the cellulose on the floor would be just fine- there wouldn't be a condensing region within the cellulose since the cool side is humidity-controlled via the AC and the warm side of the cellulose would have higher capacity. But I may be missing something here... (cood be rong, offen am...)

I was curious about the R4/inch claim on the Demilec since it seems well over typcal half-pound foam performance but it appears it's been independently verified:

http://www.demilecusa.com/Repository/File/agribalance_rvalue_chart_081108.pdf

Looks like your 4.5" will perform around R20-ish- VERY high compared to generic half-pound foam, but note that the test samples were blown at 0.7lbs/ft^3, not 0.5lbs/ft^3 (7/10ths , not half pound) so it SHOULD outperform half pound foam if it's 7/10ths pound foam the cell sizes are smaller and the percentage of open cells lower. If the blowing equipment temps & chemistry is set up for getting that 0.7lb density, your 4.5" will achieve the performance of 5.5" of generic half-pound foam.

Interesting, can the staller change the equipment to alter the foam weight? That's a concern since the R value is dependent on .7 lbs/ft foam. If he changes the machine to "stretch" his material the R value is going to be affected - negatively.

Btw. My friend who recommended this type of insulation had his roof rafters foamed with this product. Since his ceiling was already filled with cellulose he left it in place in conjunction with the roof foam. He is very happy with the results and no mold issues "yet".

 This was kind of like something I was thinking. My roof trusses are designed to be a scissor truss so I was considering putting 2 lb spray foam on the roof and then filling the lower section with cellulose or batt insulation. I think the trusses are about 16" deep, I thought that that would give a double vapour barrier though. I guess I could use 1/2 pound spray and then put the vapour barrier below the batts though.

Would this be acceptable in a fairly dry climate like Calgary, Canada?

Sean

Half-pound foam isn't sufficiently vapor retardent enough to meet Canadian code, even in fairly thick applications.  2lb foam DOES form a vapor barrier at a couple of inches.  If you're going to use 2lb foam you may get the best value out of it on the attic floor to vapor-seal it, and applying cellulose & batting above that.  Adding a even a foot of half-pound foam to the roof deck won't vapor-seal it unless you spray it with vapor-retardent paint, but it will form a very good air-barrier there.  Without vapor retardent paint 5" of half-pound foam has a permeance of about 10 perms, whereas 2" of 2lb foam is ~0.75-1 perms, making it a class-1 vapor retarder, but it takes over 4" to approach the permeance of a poly vapor barrier.  (I'm not sure exactly where Canadian code draws the permeance line for vapor barrier requirements- it might be 1 perm, but may be 0.1 perms.)

If you apply vapor-barrier insulation as the exterior layer of a multi-insulation-type stackup in a heating dominated climate, it needs to be a large fraction of the overall R-value such that the average dew-point of the interior air's relative humidity in the temperature profile through the insulating layer(s) occurs in the foam than the fiber, since it's an inward-drying stackup.  This can be controlled to some extent by keeping the interior humidity very low, which isn't the best for human health & comfort. If the vapor-retardent insulation is an interior layer, the outer material will dry toward the exterior and the dew point within the material will always be the great-outdoors (unless breaches in the vapor/air barrier allows more humid interior air to infiltrate into the fiber insulation than it's drying capacity can handle.)

For example, if the interor air is 20C, with 30% RH, then according to the psychrometric chart (if you were at sea level, which you aren't in Calgary), at the point where the averaged daily temperature in the insulation is ~12C you could have condensation were vapor & air allowed to pass somewhat unimpeded the interior into the insulation.

http://upload.wikimedia.org/wikipedia/en/0/05/PsychrometricChart-SeaLevel-SI.jpg

But if the vapor barrier is on the interior side with some freer exchange of vapor/air from the exterior, the exterior air will never be above 100% RH, and as one follows the profile through the insulation the temperatures are warmer, and relative humidity lower. Even with 10%RH @ 20C (uncomfortable and not conducive to health), the dew point occurs at about the 8C depth- you'd still need to make the 2lb foam at the roof deck pretty thick to be able to use it with a large amount of fiber insulation at the interior.

As long as the relative humidity of fiber-insulation-entrapped air stays below ~60% (averaged for the season/year) the potential for growing mold & rot on adjacent wood is quite low, but at 70% RH and up you can almost count on it.  This is why placing the vapor barrier on the WARM side and allowing drying toward the cool side is important.  Air from the warm side, has higher humidity than can be maintained in the cooler layers of the insulation, (the relative humidity will rise as the air cools) and it will condense there.  Seasonally the warm & cool sides flip, but in all parts of Canada the cool side is the exterior, for the preponderance of the year.  There's more ambiguity in the lower 48 of the US, leading to a lot of confusion in the discussion.

If it averages 20C 30%RH indoors, and the 12 mid-winter weeks average 0C outdoors, the simplest model says the average 12C dew point is 12/20ths= 60% of the R-value in from the exterior.  If your total R value is R50, 60%, (R30 minimum or ~4.5") of it needs to be the 2lb foam, if it's at the roof deck.  But if you lower the interior temp & RH you can get away with less.  But if you put at least 2" of 2lb foam at the interior next to the ceiling you can pretty much pile up the rest with whatever permeable insulation you like, as long as you don't vapor-seal the exterior over it.

What's the average winter temp in Calgary and how much 2lb foam do you really need if you're putting it at the exterior? Beats me- this was just simplified model to use as an example of how to think about it.  It's just easier & more reliable to design for strong drying toward exterior, with the vapor barrier on the warm side if you're tying to save some money by combining insulation types.

You may find this collection of slides useful in following the above blather:

http://www.sprayfoam.org/uploads/pages/4507/Vapor%20Barriers%20&%20Vapor%20Retarders.pdf

I tried to google-up better documents and best-practices rules of thumb, etc, but this one covered most of the territory, if not in great depth.  But the real takeaway message is found on P22 of the document:

Install the building materials such that relative vapor retardance increases toward the side with the higher absolute humidity (usually the warm side)

When this cannot be done, install a vapor retarder such that the vapor retarder is positioned as close to the side with the highest absolute humidity as possible and the vapor retarder has an installed perm rating substantially less than that of the next lowest component.

Putting 2lb foam on the "wrong" side of the multi-insulation stackup can be designed around, but isn't ideal.

Wow Dana, thanks for taking the time to post that up. I gotta admit, I read it about four times and stared at my wall for 20 minutes. I was thinking of doing the 2lb foam to what the quote said was R40, seems a little high to me from what I've read. This would leave me with a large cavity that I thought I could just "fill up" with some other form of insulation. I may not have stated that all our ceilings are cathedral style using parrallel chord trusses (edit-I mistakenly said scissor trusses) therefore the extra space.

 If I understand correctly I want the vapour barrier on the inside like normal insulation. The only way I could figure how to do this would be to drywall my ceilings, spray foam, fill with batts, and then add my exterior sheathing?.?.? I'm sure I'm confused now.

So now I'm thinking I should just spray the truss heels and joist ends, put in soffit and ridge vents and then stuff that whole area with blown cellulose or batts.

 I feel so lost, I'll go look for pictures.

Thanks again,

Sean

PS. sorry I had to edit a few things.

Only advantage to a hot attic that I can think of is killing the termites.

All,

Thank you for all of the information contained in this thread.

I am building a house in the St. Louis, MO Area. It is about 7000 sq ft finished atrium ranch with 4000 sq ft on the main floor and 3000 sq ft on the lower level. I am using two ground source heat pump systems (4 ton and 3 ton). Originally I was planning on having R-49 blown in cellulose in the attic and blown in fiberglass blanket in the walls. After doing quite a bit of reading it seems like going with Icynene in the walls would be better option. I am doing an extensive amount of home automation, and audio / video distribution and electricians did not place all of the low voltage wiring up high enough that it will not be buried by the cellulose, and I expect that I will want to change overtime so given this and the fact that it appears applying the Icynene to the underside of the roof deck and creating a semiconditioned attic has some clear advantages.

I spoke with the insulation contractor and they expressed a couple of concerns:

1) Potential additional load on the HVAC system. From some of the previous posts, I get the impression that this should not be a big concern.
2) Some type of ventillation of the attic

I know if I go this route I will need mechanical ventiilation for the house to insure the proper amount of fresh air is brought into the house, do I need to make sure I now have the attic participate the air return system?

All thoughts are appreciated.

Bryan

A good heat load calculation will give you a good idea if there will be any significant increase on the HVAC load. I think it'll be minimal at most.

Consider putting return grills in the ceilings of all your rooms, except kitchen, bath, & utility. Duct them individually to the ventilator. Draw the remainder of the exhaust air into the ventilator from the attic. That will insure air movement through the attic, keeping it conditioned, without mixing attic air into the living space.

Regarding # 1: Where does the "extra" load for the HVAC come from, in his mind?  The additional extra exterior surface area is minimal, but the air-sealing against infiltration is more assured/easier to achieve, pulling less hot-humid air into the space. In practice sealing & insulating the attic tends to decrease, not increase the AC load.  Ventilation provides only EXTEMELY modest cooling by itself, and tends to draw conditioned-air into (then out of) the attic, increasing the net load (both heating & cooling.)

Regarding #2 :  Ventilating the attic for what reason?  Ventilating the attic has very minimal effect on roofing material temperature & lifespan. Ventilating the attic may purge it of wintertime moisture, but unless the attic floor is perfectly sealed it can draw humid interior air into the attic to condense on the wood- the ventilation is an imperfect "solution-problem", fixing the problem that it causes, but at an energy cost.  By naking the pressure boundary at the roof deck there is then only very modest (in comparison to stack-effect) convection pressures to pull conditioned air into the semi-conditioned attic space.  If high humidity in that space becomes an issue in winter, mechanical dehumidification of the very limited space is cheap & effective.

Read & understand this, print it out for the contractors/inspectors if it becomes an issue for discussion:

http://www.buildingscienceconsulting.com/resources/4-Understanding_Attic_Ventilation.pdf

Separate mechanical ventilation isn't as-important for non-occupied spaces, but keeping the humidity controlled is a good idea.  In semi-conditioned spaces inside the pressure boundary but only partially inside the thermal boundary even small amounts of mechanical ventilation will be counterproductive from an energy-use point of view. Providing ventilation while you're actively re-wiring up there makes sense, but otherwise not.

You may want to monitor the relative humidity in that space for the first year or two, but R49 of blown cellulose makes a significant hygric buffer- it's likely to moderate itself just fine without mechanical dehumidification until/unless there's a roof leak. An RH of 60% or less is fine- no need to go lower.  Over 70%RH the mold potential just explodes.  Keep this in mind for your HVAC systems & controls for your conditioned spaces too.  Going with an ERV instead of an HRV is probably worthwhile in St. Louis, and controlling air conditioning via humidistat might be considered. Oversizing the AC reduces it's effectiveness at dehumidification- if anything UNDERSIZING it 15-20% and controlling it with a humidistat instead of (or in conjunction with a thermostat) may provide better comfort.

I am learning a lot in this thread. How would you compare this tyvek approach http://www.tamlyn.com/index_files/Page2485.htm to spray foam? Its appeal to me is its DIYability with blown cellulose. I was going to blow the insulation over the ducting, which would only be about 4 HRV runs.

Using housewrap to tighten up roof venting requires close attention sealing the edges which may be quite difficult to get perfect at the soffit end. (Watch their installation video- they don't show that in very good close-up detail, but it would of-necessity be labor intensive.) I doubt it would be quite as tight as using foam, but using a vapor permeable air-barrier between the soffit/ridge vents and the open attic space will reduce infiltration losses significantly and DOES allow the attic space to shed humidity via the venting scheme.

The Tyvek Attic Wrap is also a radiant-barrier, which will reduce cooling loads somewhat- not enough to care in an R49 cellulose-insulated attic, but measurably in an R38 or less fiberglass insulated attic. The air-barrier aspect will be a much bigger determinant of it's ultimate performance determinant than it's reflectivity. Getting edge & seam seals perfect is key.

Be sure to mastic-seal all joints & seams over your AC duct runs before you blow the cellulose over it. Fixing duct-leakage issues after the fact would be a real pain.

The good thing about DIY (he says cavalierly) is that "labor intensive" costs the same as the other kind. Thanks.

Dana1,

If I go with the Icynene on the underside of the roof deck, I will not be using the blown in cellulose in the attic.

Thank you very much for your comments. I will review the reference you provided, and may be back with some more questions. :)

Bryan

Damaceld,

The ERV system will be in the lower level and integrated as I understand it at the larger Geothermal unit for the house. It will exchange air through vents on the lower level at the rear of the house. Do I need to have one of the returns open up into the attic?

Bryan

Sorry for the slow response but I was out of town for a couple of days. I thought it would be a cinch to find a reference to show you what should be done, but I come up with nada. I was expecting to find an illustration at BuildingScience.com to show attic air circulating through the heating/cooling system just the same as they show for crawl spaces, but I don't. But, all the references to sealed insulated attics refer to them as conditioned attic. Most of the illustrations show the furnace & air handler, and ducts, in the conditioned attic. Obviously conditioned means that the air in the attic goes through the heating/cooling system, probably by way of duct leakage in most cases.

My garage is ICF construction and is insulated under the roof. Right now I have no vents between the garage space and the garage attic, which, in compliance with codes, is separate from the house attic. I can tell it is hotter and more humid up there than in the garage space. I plan to add vents in the ceiling. Also, any closed volume acts as a moisture pump, sucking air with higher humidity in during the cool of night, and expelling warmer, lower humidity, air out during the heat of the day. It's virtually impossible to make a closed space absolutely air tight, especially in a house. As the air moves in and out more vapor is brought in than is expelled, causing the humidity level to climb.

So, I would say you need to move the attic air through the heating/cooling system, but, of course, at nowhere near the rates as through the living space. A small supply and return, maybe even as small as 4", should be sufficient to keep the air in the attic reasonably dry and fresh. A return duct opening into the attic and one or more ceiling vents into the attic would probably work good.

A conditioned attic will:

Add more space/surface area you are heating/cooling
Reduce shingle life about 10%
Cost more in insulation (all foam)
Not seal any better than 1-2" of spray in foam + cellulose on the attic floor
Cause greater problems with ice dams

But also:

Be better in a hurricane or fire
Solve the issue of ducts in the attic losing/gaining heat/air

Add two more:

Makes the attic space into a great conditioned storage space, if size permits.
Makes it a whole lot easier to make changes to wiring, add multimedia wiring, add ceiling lights, etc.

The amount of "extra" surface area is in most cases quite small (~15-25% for low & moderate pitched roofs) with a correspondingly small increase in heat load (an increase in heat load that can be fully mitigated by slightly more insulation.)

10% would be an upper bound on reduced shingle life, not the "typical" number.  The color of the shingle can make as much or more difference in peak temperatures and shingle life.  Shingles are primarily convection & radiation cooled from the outside.  There's typically ~R1 or greater insulative value from the roofing felt & roof decking already, so you don't get a huge shingle-cooling effect from attic ventilation even in the best of cases.

In my region (MA) foamed roof deck is the cure for, not the cause of ice damming problems.  Ice dams are a heat-loss problem- fixing the heat loss with foam at the roof deck itself is reliable (from either above or below the roof deck.)

But foam IS more expensive, and foam sealing the roof deck with 1-2" with cellulose on the floor can indeed still get you there- with caveats: 

In heating dominated climates if there is even small-moderate amounts convective infiltration beween the now-sealed attic and conditioned space there is a possibility of condensation & frost in the "in-between" space, since the drying-to-exterior capacity is much reduced.  This can be avoided if the foam comprises enough of the total R-value that the average wintertime dew point occurs within the foam layer.  More foam lowers the delta-T that would drive the convection currents as well.  Sealing the attic floor & access doors/hatches perfectly is often difficult, and penetrations created later are common (from new wiring/plumbing/recessed lights, etc.)