Correct way to use closed-cell insulation on a roof?
I’m rehabbing an old 3-story brick building in Pittsburgh. The roof is a flat roof but pitched from front to back. I got estimates to spray closed-cell on the underside of the roof and essentially encapsulate the structure. There is a ‘space’ between the 3rd-story ceiling and the roof, so if I drywall the 3rd-story ceiling, there would then be an air space above the drywall. I’m also planning on installing a new rubber roof. So I guess I have several questions:
1. Is applying closed-cell to create a barrier at the underside of the roof the correct way to go?
2. If this is correct, would installing 2 inches of ISO on top of the roof sheathing and underneath the rubber roof be necessary/recommended?
3. If the closed cell did not give me the required R-value, could I add fiberglass bats between the ceiling joists or would the added insulation need to go up against the closed-cell.
4. If I closed-cell the underside of the roof and drywalled the 3rd-story ceiling, would that airspace between be okay?
I’ve had conflicting advice. One was to make my thermal barrier at the ceiling of the 3rd floor…ie…put fiberglass bats in the ceiling joists and and drywall the ceiling, and caulk any seams to create an air barrier. If I did it this way would I then have to vent the space above the ceiling? I’m going to have closed cell applied to the rest of the house.
Any information would be greatly appreciated.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
John: keep in mind the ready availability of reclaimed xps foam for use under the rubber roof.
John,
I suggest that you start by reading this article: Insulating Low-Slope Residential Roofs.
If you still have questions after reading that article, feel free to post them here.
Thanks for being so responsive. Unfortunately I have to get to work. That looks like a really pertinent article and I will look more closely at it later. I briefly looked at it and picked out this sentence:
•You can install a more moderate thickness of closed-cell spray polyurethane foam on the underside of the roof sheathing, supplemented by a layer of vapor-permeable insulation below that.
Would that vapor-permeable insulation be right up against the closed cell or would it be in the ceiling joists, thus creating an air space?
And would the ISO on top of the roof sheathing be necessary if the closed cell is used underneath?
Thanks again. You guys are really helpful.
John,
Q. "Would that vapor-permeable insulation be right up against the closed-cell or would it be in the ceiling joists, thus creating an air space?"
A. Right up against the closed-cell spray foam. (That's a code requirement.)
Q. "Would the polyiso on top of the roof sheathing be necessary if the closed cell is used underneath?"
A. No, as long as the closed-cell spray foam meets the minimum R-value requirements for spray foam as shown in Table R806.5 of the 2012 IRC, and as explained in my article, How to Build an Insulated Cathedral Ceiling. That means that your spray foam needs a minimum R-value of R-20 in your climate zone (Zone 5).
Alternatively (and as explained in that article), you can install enough polyiso (i.e., R-20) above the roof sheathing to meet the minimum requirements shown in Table R806.5 of the 2012 IRC, and make up the difference (to meet the minimum R-49 code requirement for roof insulation) with fluffy insulation installed directly against the underside of the roof sheathing.
Great thanks. One more question to finish this up. Assuming I used enough closed-cell below the sheathing to meet code; would the air space between the closed cell and and the 3rd story ceiling drywall create any problems? I plan on using closed cell on the walls as well, so this space would be air sealed off from the outside.
By far the better way to go in a Pittsburgh climate(US zone 5) is to put 4" of foam on top- 2" of polyiso on the roof deck, and 2" of EPS above that , and putting up to R30 as cellulose or rock wool snugged up under the roof deck (high density fiberglass being a third-best option.) That stackup would put you at IRC 2012 code min. If you want to go higher R than that, you must proportionally increase the foam & fiber R values to maintain at least the same foam/fiber ratio. (Just more foam is fine, but more fiber demands more foam.)
The dual foam is needed make the best use of the uprating/derating properties with temperature of each, to achieve fairly flat R20-ish performance to the stackup independent of outdoor temp. If the outer 2" is polyiso that layer would dramatically under perform it's rated R, and would in fact underperform EPS during the mid-winter period when you need it the most. The EPS layer keeps the other 2" of polyiso warm enough to perform near it's rated R. Even though 4" of polyiso would outperform the dual-material stackup during the shoulder seasons, the EPS + polyiso stackup outperforms it from a dew-point control perspective.
Closed cell polyurethane foam (ccSPF) isn't all that green, since it's blown with a powerful greenhouse gas HFC245fa, which is ~1000x CO2. Both polyiso and EPS are blown with pentane at about 7x CO2. XPS is blown with HFC134a, which at ~1400x CO2 is even more damaging than closed cell polyurethane. When it's possible to design it out of the stackup it's better to avoid both ccSPF & XPS, and go with the greener materials.
Open cell foam is blown with water, which is pretty low impact, and at equal R-values it air seals better than closed cell too, but it is an order of magnitude more permeable to water vapor. Painting it with "vapor barrier latex" brings it down some, but only to about 5 perms, not the ~0.5 perms the paint performs at when applied to sheet-rock.
Going with reclaimed foam from commercial demolition & re-roofing is greener than any of it, as Stephen Sheehy suggested. For reclaimed XPS, for design purposes derate it to R4.2/inch , which is it's fully-depleted performance (after the blowing agent has already disappeared, which takes a handful of decades or so.) Fiber-faced roofing polyiso is typically R5.5/inch rather than R6/inch, and use R4/inch for any EPS where you are uncertain of it's density. Both EPS and fiber-faced polyiso have pretty stable R values after the first year, not the long-term decay seen by XPS.
Virgin stock polyiso & EPS run about 10 cents per R per square foot ( 2-3 cents/R-foot for reclaimed goods), compared to 17-18 cents/R-foot for ccSPF, or 12-13 cents for open cell. On a flat roof with low scrap rates rigid board is definitely more cost effective.
John,
Q. "Assuming I used enough closed-cell below the sheathing to meet code; would the air space between the closed cell and and the 3rd story ceiling drywall create any problems?"
A. No.
And by the way: Dana is right that installing rigid foam above the roof sheathing is a better way to go than using spray foam underneath, for two reasons:
1. The rigid foam does a better job of addressing thermal bridging through the rafters, and
2. The rigid foam is more environmentally friendly.
Boy, that seems like alot of stuff on the roof. Is it a no no when using closed cell underneath the roof to put a couple of inches of rigid foam on top the sheathing to help minimize thermal bridging?
If it is a no no, would you recommend anything between the sheathing and the rubber roof? It seems like the wood would want to 'sweat' being against the rubber.
John,
What you call "sweating" is usually called condensation. Condensation occurs when warm humid air contacts a cold surface. If you put an adequate amount of rigid foam insulation above your roof sheathing, your sheathing will stay warm and dry, and won't be subject to condensation or moisture accumulation.
According to most experts, it's not a good idea to sandwich your roof sheathing between two layers of impermeable foam insulation. It's best to choose to put your foam on one side, not both.
For many reasons, it makes more sense to put rigid foam above your roof sheathing than it does to put spray foam under your roof sheathing.
Splitting the difference can work, but it's not as resilient as putting all of the foam-R above the roof deck.
As long as you limited the interior side closed cell foam to 2" there is still a reasonable drying path for the wood, but it's about 0.5 perms instead of 3-5 perms.
Membrane roofs generally require some sort of underlayment to relieve the mechanical stresses as things change dimensions with temperature & moisture content. That's true whether it is applied over a foam layer or directly on the roof deck.
A four inch foam-over is pretty easy on most roofs. Once you're over 6" it can be more complicated.
Wow, this is pretty involved stuff. I haven't made up my mind yet, and despite all your wonderful advice I'm still leaning towards trying to get enough R-value by spraying closed cell under the sheathing. Just to be clear, if I did it this way, then i would not want to put any foam at all on top of the sheathing?
The other option of installing layers of foam, polyiso, and EPS on the sheathing...I was a little confused as to the sequence of the layers starting at the sheathing. Would you please go over this again. Is there any way I could attain the required R-value this way? I was trying to avoid climbing into the rafters to install fiberglass or other insulation. Appreciate your time, and thank you.
No, you would still want to put more foam on the exterior. With a membrane roof there is absolutely ZERO drying toward the exterior. Adding exterior foam does not reduce the drying rate from zero down to less than zero. Zero is as low as it gets, but it's zero with or without exterior foam.
But with exterior insulation the average temperature of the roof deck is higher, which limits its maximum moisture uptake from the interior moisture drives, and starts the drying season earlier. The relatively low vapor permeance of the interior side closed cell foam slows how fast moisture can move into the roof deck, but it also slows how quickly it dries in warmer weather.
If you put enough foam on the exterior the average temp at the roof deck stays that much warmer, and correspondingly the roof deck stays that much drier, which means it's OK to allow the vapor permeance toward the interior to be an order of magnitude higher. By having a fairly vapor open drying path, the assembly is quite a bit more resilent. Even if it somehow gets wet, it dries out as much in ONE WEEK as it would TWO OR THREE MONTHS through 2" of closed cell polyurethane. It's not a total moisture trap- the roof deck can indeed dry, but it's a wicked-slow ordeal through 2" of ccSPF.
That is why 4" of foam on the exterior is a far superior approach than 2" of foam between the membrane & roof deck and 2" of foam under the roof deck.
John, if you do install R20+ rigid insulation above the deck, then you don't need any spray foam, and you can make up the rest of the R value with cavity insulation (ideally blown cellulose or blown fiberglass) as mentioned above. Isn't this more economical than installing spray foam? And you avoid a "foam sandwich" that could result in moisture trapped in the deck, which would eventually turn to compost. Why would you still want spray foam?
I am starting to be swayed. I took a closer look at the underside of the roof deck and it looks like there is room to get in there and install vapor permeable insulation (preferably fiberglass batting I would assume) to supplement rigid above the deck. I need to work out the details of the stack noted above, because I have a feeling the roofer won't know. I believe it went this way: roof sheathing, 2" polyiso, 2" EPS, 4" foam...Is this correct?...A couple simple questions to go with...Can the roof be walked on with this stuff? And I assume I need to ensure vent piping is long enough to reach above these layers?
John,
That's 4 inches of rigid foam total (2 in. polyiso + 2 in. EPS = 4 in. total).
Guess what I am having a hard time understanding is if there is zero drying to the outside with a membrane roof, and one of the insulating recommendations is to install closed cell to the underside to a recommended depth to meet code/R-value, which is probably at 3" or more...then it's probably not drying to the inside either at that thickness. I know nothing about this, so sorry if my question are annoying.
John,
You are correct that there is very little drying to the interior when you install closed-cell spray foam on the underside of your roof sheathing. That's one reason why the spray-foam approach is still somewhat controversial (at least for conservative builders), and one reason why most experts recommend that installing rigid foam on top of the roof sheathing makes more sense.
Needless to say, plenty of people are installing closed-cell spray foam on the underside of roof sheathing. If you are comfortable with that approach, go right ahead and do it.