What would the \best foam for a flat roof? I have a 400 square foot addition with a flat roof that is built on 9.5" engineered trusts. The roof is 3/4" t&g sheeting with a paint on membrane type coating, light gray. My foamer has speced open cell foam which will give me about an r-38, but I'm wondering if going with a closed cell for a better r value would be worth it. The house is in central MD and the biggest concern is the heat gain in the summer. Which product would product would be best to reduce the amount of solar gain. The walls are 2x6 with open foam and a ZIP R-6 (r-6.6) panel on the outside covered with hardie plank. Marvin Low-E windows. Thanks,

Posted By arcamm on 20 Sep 2014 02:57 PM
What would the \best foam for a flat roof? I have a 400 square foot addition with a flat roof that is built on 9.5" engineered trusts. The roof is 3/4" t&g sheeting with a paint on membrane type coating, light gray. My foamer has speced open cell foam which will give me about an r-38, but I'm wondering if going with a closed cell for a better r value would be worth it. The house is in central MD and the biggest concern is the heat gain in the summer. Which product would product would be best to reduce the amount of solar gain. The walls are 2x6 with open foam and a ZIP R-6 (r-6.6) panel on the outside covered with hardie plank. Marvin Low-E windows. Thanks,

Closed cell foam would be an environmental disaster due to the high global warming potential of the HFC245fa blowing agent (about 1000x CO2).  It has to be installed in 2" lifts and allowed to cool between lifts to avoid shrinkage/adhesion problems or self-igniting as it cures.  Also, at anything over R25-R30 it is also on the extremely vapor-tight side, and you run the risk of trapping moisture in the roof deck.

Open cell foam is blown with water which has a very low global warming potential.  At R38 it is sufficiently vapor tight to be protective of the roof deck in an MD climate, but not so vapor tight as to cause a problem.  Open cell too has to be installed in lifts of typically 5", but instead of 3-4 lifts to hit R38 you're looking at only 2 lifts.

The better way to do it is with rigid foam on the exterior, or split it between exterior & interior. R20 polyisocyanurate only adds 3.5" to the height of the roof deck, and with even R30-ish (8") of open cell foam on the interior you would would be at or above IRC 2012 performance.  It's only a bit over half the cost per R of closed cell foam, and is often cheaper than open cell foam, depending on labor rates how simple the roof design is, etc. For a  single pitch flat roof it's probably going to come in cheaper per R than open cell foam, but could be approaching closed cell foam rates if it's a nightmare of multiple dormers & valleys, with a high scrap rate and high labor content due to all of the odd-angles.

Thanks for the info. Looks like it's open cell. The roof deck is already coated with the roof membrane so adding to the exterior is out.

Would I be right to guess that adding an inch or two of ridge foam to the under side of the deck with the open sprayed on it would be a bad idea?

By the time you cut'n'cobbled the rigid foam in there any cost advantage would have disappeared in the labor cost, and you would have potential issues with long term adehesion. Applied to a clean wood surface open cell foam is like glue, on other materials maybe not.


Typical half-pound foam these days comes in at 15-20 perms @ 2" and tests at ~R3.7/inch. So you'll need 10" to hit R38-ish, which will deliver 3-5 perms vapor permeance. If you go less than 8" it's worth spray-applying "vapor barrier latex" directly onto the foam, whiich will bring the total vapor permeance to 5 perms or less, which is where it needs to be. Higher density open cell foam tends to be lower perm/higher R/inch. This is a typical half-pound foam:

http://www.tailoredfoaminc.com/technical-data/data-files/icynene-ld-r-50/icynene-ld-r-50-spec-sheet.pdf

The same manufacturer makes a water-blown semi-open cell foam that runs R5.2/inch, that would still be sufficiently vapor open at R50 (10") to not be a moisture trap. It's rated 1.3 perms @ 3", which would be about 0.4 perms @ 10", which is still very reasonable drying potential.


http://www.icynene.com/sites/default/files/downloads/ICYNENE-MD-R-200-Specification-Sheets-US_0.pdf

(Don't get it mixed up with their HFC blown R7/inch closed cell stuff MD-C-200, which is very low permeance, and blown with HFC245fa.) It's hard to find installers who handle the MD-R-200, and it would cost as much as closed cell, but it would be a higher-R option that would work without becoming an environmental disaster.

Dana, do you feel that closed cell is simply not a good product, due to the environmental impact? I'm using open cell, but was trying to figure how to easily add to the R value of the roof. Sounds like foam board on the exterior is far better?

cathsand
I replied to your post about using open cell foam. You might want to read that. I was surprised by the problems cited in the report. Seems like there isn't a good fit out there but depending where you live maybe there is.

Please post what you decide is a solution and how it turns out. I am currently planning on just insulation on top of my ceiling unless some foam solution makes sense...

Posted By cathsand on 24 Sep 2014 08:22 PM
Dana, do you feel that closed cell is simply not a good product, due to the environmental impact? I'm using open cell, but was trying to figure how to easily add to the R value of the roof. Sounds like foam board on the exterior is far better?

At high-R closed cell polyurethane blown with HFC245fa does more lifecycle climate damage than any carbon-reduction it might achieve over the next 50-100 years.

At high-R closed cell polyurethane is also close to being a true vapor barrier, which can create moisture trap issues that need to be considered when designing a stackup.  Anything over 2" requires a more detailed analysis of the drying paths.

At lower-R when used for vapor-permeance control, air sealing, or dew point control within an assembly (say, in a studwall flash'n'batt)  it can be very useful stuff.

Closed cell foam on the underside of the roof deck limits the drying rate of the roof deck.  Insulation on the exterior of the roof deck keeps the roof deck warmer (= dryer), and with a sufficiently high exterior R relative to what insulation you install on the interior such that the average temp at the roof deck is above the typical wintertime dew point average of the interior-space air, you can leave the interior side relatively vapor-open. That ability for the roof deck to dry quickly to the interior makes the assembly more resilient.

Not all rigid foam board is as benign as EPS or polyisocyanurate.  All XPS (pink, blue, green board) manufactured in north America is blown with a mixture of agents, the dominant component of which is HFC134a, which has an even higher global warming potential than HFC245fa (about 1400x CO2 compared to about 1000x.  Polyiso & EPS are blown with pentane, at about 7x CO2.)

When using polyiso on the exterior for dew-point control in colder climates (really starting from the cold edge of US climate zone 4 or higher) it's important to de-rate it from it's labeled R-value.  The performance of polyiso peaks when the average temperature thorugh the foam is 50-60F, but when the average temp is below freezing it falls off a performance cliff to about half the labeled-R. In climate-zone 4 and lower simply using R5.5/inch (instead of the labeled R6-R6.5/inch) would be good enough.  When the average  mid-foam temp is going to be lower than 30F, splitting the total foam thickness with a layer of EPS on the exterior would yield higher mid-winter performance, something closer to the labeled R-values, since at foam temps below 30F EPS will have a higher R/inch than polyiso.  EPS performance increases fairly linearly with falling temperature, and doesn't have the wild curve to it's thermal conductivity seen with polyiso.  The brown line in this graph is polyiso, the black line is EPS.



Note that the crossover between polyiso & high density EPS is at a mid-foam temp of about 7C/45F. The crossover point for Type-I EPS would be below 40F.

Since the dew point of mid-winter conditioned space air is typically 35-40F, you want the average temp at the roof deck to be above 40F.  While doing it all in EPS would make the design simpler, the higher performance of the polyiso at even modestly higher temps boosts the seasonal drying rate of the roof deck during the shoulder seasons than it would be with an all EPS solution, or even during warmer mid-winter afternoons, and an all-EPS solution is significantly thicker in zones 5 & higher too.

OK, that's more information than you asked for- I'll try to keep it at the overview-primer level.

See more here.

Rigid rock wool panels can be used on the exterior as well, but are only available in a limited number of thicknesses/R-values, and have a comparatively low compressive strength- it won't be a walkable roof.  For insulating wall sheathing it's pretty good though.

Very helpful, thank you Dana. As it stands, I intend to add EPS to the roof, but likely will just stick with open cell foam for the walls. Living south of Atlanta, I think that will be a big improvement over standard construction, and hopefully will leave room in the budget for a PV array.

PV keeps getting cheaper year-on-year. In my neighborhood grid tied commodity-panel PV is well under $4/watt, and high-efficiency panels barely over that price point.  I expect it to hit $3/watt or less before the 30% Federal Income Tax credit steps back to 10% on 1 January 2017, which would put the actual installed cost in the $2/watt range.  By 2020 it's likely to be under $2/watt range even without incentives.

Georgia is one of a handful of states that does not allow third-party ownership of rooftop PV, but net-metering is currently law there.  The statewide vertically integrated electrical utility monopoly has tried to stick fees on small-time PV owner, but those efforts have pretty much failed so far:

http://www.renewableenergyworld.com/rea/...-rate-case

That doesn't mean they won't try again, and as the installed base of PV grows they may get some sort of rate or fee relief from regulators, at which point the real fun begins.  At a buck or buck-fifty a watt installed price and $200/kwh battery costs it will become cost-effective for folks with the real estate and the finances to un-plug from the grid, which will push the grid maintenance and generator financing costs onto a smaller group of ratepayers, which will in turn make it more attractive for others to un-plug.   Given the very substantial sunk-cost Georgia Power has in new and expensive generating resource (the Vogtle nuclear plant under construction being just part of their portfolio), the utility is at substantial risk of holding stranded assets- more so than many utilities in other locations. 

Utilities such as GP have long been protected with guaranteed profits on their capital investments, which incentivized them to build more bigger better generating capacity, but they are not well equipped for competition, which is what they get with PV.  At $2/watt the unsubsidized lifecycle cost of power from PV is below the residential retail rates in most places, including Georgia. The financial sanity of building Vogtle isn't exactly clear, since well before it's anticipated lifecycle is up, the costs of grid defection will have a lower cost to the ratepayer than any amortized cost of power coming out of the thing, even in the most optimistic of scenarios.  Barclays downgraded the bond ratings of US utilities broadly earlier this year on that type of rationale, and analyses from other financial sector types such as Sanford Bernstein, CitiGroup and Morgan-Stanley all pretty much confirm the market forces that are creating a stranded-assets risk for the utilities.  GP may want to "own it all", and do their own utility-scale PV, but it's unlikely that the regulators would give them the power dissallow private PV from hooking up to the grid.

Bottom line- you might want to put the PV in on day 1, to protect yourself from some of this, especially if it can be rolled into the financing of the rest of the house rather than a separate financing deal (at likely higher rates), and locking into net-metering before it gets re-written with less favorable terms.  GP may struggle as they collapse financially,  but you won't necessarily be completely at their mercy if you have your own generating assets.  I suspect they will be forced to write-down their stranded assets, maybe even stiff the bond-holders, but there's only so much they can stick the ratepayers with the bill before simply unplugging becomes a more attractive option.

Thanks for reaffirming my determination to install PV with the build. It is hard to be sure which way to turn. If you had to choose between PV and mini splits, which investment would you make?