Q: I am renovating and old home with an uninsulated, non-vented walk up attic built in 1848. The shingle roof is less than 10 years old. The simple gable roof has 2×6 rafters. I’d like to apply 2 in. of closed-cell foam and then install 3 ½ in. of R-15 rockwool and add an Intello “smart” vapor retarder. The finished ceiling will be1/2” nickel gap shiplap.
Will this approach prevent condensation on the underside of the spray foam? I’m also considering using strips of foam board with taped joints and foamed edges – as an alternative to spray foam. I live in Westchester, NY in Climate Zone 5. — Andy
A: Combining spray foam with fiber insulation in an unvented cathedral ceiling (or wall) is often referred to as “flash and batt.” This is a pretty easy way to boost your R-value and get an airtight air barrier without a lot of fussy detailing or extra-deep framing. It provides many of the benefits of closed-cell foam (higher R-value, tight air barrier) without the cost of filling the whole cavity with foam.
The approach also has the downsides of spray foam – high cost and fact that most spray foams use a blowing agent that contributes to global warming. Some of the newer foams use less harmful blowing agents. Open-cell foam is more eco-friendly, but not recommended for this application because of its permeability to moisture.
Dew Points & Condensation
With flash-and-batt insulation, you need to use enough foam to reduce the potential for condensation on the bottom surface of the foam. This is the first cold condensing surface where water vapor in household air may condense into liquid water. The goal is to keep the interior surface of the foam above the dew point of the interior air for most of the winter. So even if indoor air leaks into the roof cavity, no significant condensation will form.
The colder the climate, the more foam you will need. It works like this: if 50% of the total R-value is in the foam and 50% is in the rockwool, then the temperature on the interior face of the foam will be 50% of the temperature difference between inside and outside. So if it’s 0° F outside and 70° F inside, the temperature of the foam surface will be 35 degrees F.
If 40% of the total R-value is foam, then the temperature on the interior surface of the foam will be 28° F (40% X 70).
Code Requirements
The IRC building code requires that the average monthly temperature on the interior face of the foam during the three coldest months be above 45° F (about 7° C) . In Climate Zone 5, that works out to R-20 foam with a total R-value of R-49, based on the 2021 IRC. The key issue is the ratio of foam to fibrous insulation. In this climate zone, the foam should account for at least 41% of the total R-value (see Table 1 below).
In your case, you have a total roof R-value of about R-25 and a foam R-value of R-12 to R-13 (for 2 in. of closed-cell foam). That puts your foam R-value well above the recommended 41% ratio for DOE Climate Zone 5.
This assumes an aged R-Value of R-6 to R- 6.5 per inch for closed-cell foam, a fairly conservative number. The measured R-value of spray foam after 6 months ranges from just under R-6 to just over R-7 depending on a host of variables, including the thickness of the foam (thicker is better), the material sprayed to (metal is better than wood or concrete), the formulation and application of the foam, and the test method. Tests conducted by the US Navy and others on closed-cell spray foam after 5-10 years show R-values ranging from R-5.8 to R-6.2 per inch for 1-3 inches of foam applied to a wood substrate (read more on test results).
Since you have no control over most of these variables, it’s best to assume a long-term R-value of about R-6 per inch.
Air Barriers and Vapor Retarders
There is no real consensus on the need for an air barrier and vapor retarder below the insulation.
Some contractors and energy experts say that you don’t need an interior air barrier or vapor retarder because the foam provides an excellent air barrier and is warm enough that you don’t need to worry about condensation.
With properly applied spray foam, and the right ratio of foam to fluff, this may be true. If you use “cut-and-cobble” foam board, however, it is harder to achieve a perfect air barrier. Also the wood rafters act as thermal bridges, and will tend to be colder than the interior foam surface.
In addition, workmanship on most job sites is less than perfect (and wood and buildings move over time). So I always prefer a belt-and-suspenders approach. For those reasons, I still recommend a tight interior air barrier and a moderately permeable Class III vapor retarder (1-10 perms) with flash-and batt-systems. Ordinary latex or enamel paint on drywall works fine as the vapor retarder.
A smart vapor retarder like Membrain or Intello is also a good choice as it can provide a good air barrier, if installed carefully, in addition to a vapor retarder that becomes highly permeable when wet. This is an ideal product for this approach and would be a good solution with your shiplap ceiling.
It’s important to retain some drying potential to the interior in case the cathedral ceiling cavity gets wet despite your best efforts. The source of the moisture may be a roof leak rather than condensation from indoor air.
So a low-permeable vapor barrier like poly sheeting should never be used below foam insulation, whether it is foam alone or foam plus fiber. Any moisture that gets trapped between two impermeable materials will linger, leading to mold growth, wood decay, and possible insect damage.
Cut & Cobble
You also mentioned that you might cut strips of foam board to fit between the rafters as an alternative to spray foam. This is definitely cheaper if you provide free labor. However, this approach, sometimes called “cut and cobble” is very labor intensive and getting the required air seal takes effort.
You will want to tape all foam-to-foam joints with a high-quality construction tape that is suitable for the type of foam used. Also, you’ll need to foam the outside edges where the foam board meets the wood framing. Standard home-center canned foam is not up the task. A larger professional-style foam applicator is needed for the job. And wear gloves, mask, and throwaway clothing as spray foam is nasty stuff and impossible to remove from clothing.
So, yes, it’s doable, but definitely get a price from a spray foam contractor and you might think twice about doing this yourself. — Steve Bliss, Editor, BuildingAdvisor.com
Read more on:
Foam Insulation Thickness on Walls.
Insulating Cathedral Ceilings with Spray Foam
Insulating with Foam Board
Preventing Condensation in Cathedral Ceilings
Sealing Unvented Cathedral Ceilings
Robert F. says
Can Rigid Foam Between The Rafter Work As Baffles?
I’m in Delaware which is zone 4. I am finishing an attic room with cathedral ceilings. I need to allow air flow from the soffits to the ridge vent. My plan is to use 2 in faced closed cell rigid foam cut to fit in between the rafters, with 2 in spacers to hold the insulation away from the roof. This creates the needed air channel. I’ll seal all joints with spray foam. The rigid foam has facing on both sides, the foil side that will face the roof
In doing so, there is a gap of about 3-4 inches from the foam board to the drywall. I plan to fill this gap with unfaced insulation so as not to create a second moisture barrier. Most likely Rockwool as I hate working with fiberglass.
Does the above approach sound ok?
Bob
buildingadvisor says
If you look at Table 1 in the article above, you’ll see that the foam should account for between 26% and 40% of the total R-value, depending on which part of Zone 4 you are building in. You are well within that guideline with R-12 to 13 for the foam and R-12 to 13 for 4 inches of rockwool. So your foam layer is providing about 50% of the R-value.
Whether your total R-value of 24 to 26 meets your local energy code is a separate question.
The goal is to keep the bottom face of the foam above the dew point of the interior air for all but the very coldest days of the winter. This keeps condensation on the bottom of the foam to a bare minimum, even if your ceiling air barrier is less than perfect – although you should still make it as airtight as you can.
The approach you are taking is often used without a vent space, but you have the best of both worlds — virtually no condensation in the roof plus a vent space below the sheathing to permit drying in case the sheathing or framing ever gets wet from a roof leak.
The upward-facing foil will give you a little boost in summer as well due to the radiant barrier effect, although over time it tends to collect dust and its value is reduced.
Spence says
Will Closed-Cell Foam Stop Cathedral Ceiling Condensation?
Thank you for the information on this. I am having some condensation issues with my vaulted roof. It has Tongue and groove pine with no vapor barrier between the rafters and the insulation in the ceiling is only R 30 faced batts. After very cold days I get condensation due to the water vapor condensing on the roof decking, turning to ice and then melting when it warms up.
My thoughts on how to correct this:
My roof structure does not allow for roof venting due to how it was framed.
8 inches of closed Cell Foam spray in the 12 inch rafter space which is an R-value of 48. this closed cell foam will be sprayed directly on the bottom side of the roof decking creating a “hot roof.” I am in Zone 6 for insulation.
On the inside I will install drywall and paint 2 semi gloss coats of paint to help with vapor on the interior. Then I will put my tongue and groove pine ceiling back up.
Will this solve the issue?
Thanks in advance on the advice.
buildingadvisor says
A hot roof design, like you are describing, should perform fine with 8 in. of closed-cell foam. Just make sure you hire an experienced insulation contractor, as it takes some skill to maintain quality control and to avoid gaps or foam that pulls away from the framing as it dries.
You could fill the rest of the rafter space with unfaced batts for the extra R-value and still be well within the limits shown in the table above. The goal is to have the underside of the foam (the “principle condensing surface”) above the dew point of the indoor air on all but the coldest winter days.
I would install the drywall in an airtight manner as an extra precaution – that is high-quality sealants or gaskets around the perimeter of the ceiling and sealant or foam at any other gaps or openings. While the foam is, in theory, airtight and not in need of a second air barrier, I always assume that construction on site will be less perfect than the drawings.
Since this assembly can only dry to the interior, I like to use a Class III vapor retarder (1-10 perms), so even three coats of standard latex paint is OK. Another, more expensive option, is to use a smart vapor retarder like Membrain or Intello – sometimes required by code.
If the t&g pine is finished, that will also provide an additional bit of a vapor retarder. Even though t&g is very leaky stuff, vapor retarders still work with holes and gaps. A vapor retarder that covers 90% of a surface is 90% effective. An air barrier, however, needs to be close to 100% sealed to do its job.
Stephen says
How About Sandwiching Fiberglass Between Two Layers of Foam?
This is a well written article! Thank you!
Have you ever heard of anyone making a foam-and-batt sandwich? E.g., A hot roof with a flash-foam coating under the sheathing, fiberglass batting fill, then closed-cell foam flashed over the batts (under the drywall) to seal the whole kit and kaboodle?
Thoughts?
buildingadvisor says
No, this is not a detail I’ve seen — nor would I recommend it. Hot roofs are designed to dry to the interior and you are preventing this with the bottom layer of foam.
Potential problems:
1) By adding foam below the fiber insulation, you are cooling the foam under the sheathing. If any indoor moisture reaches the underside of the sheathing it may condense. It will be trapped between two layers of foam.
2) In the event of a roofing or flashing leak that wets the roof cavity, you are trapping water between two layers of foam.
I tend to be conservative with building techniques. I prefer to stick with materials and details that have stood the test of time. I’ve seen too many innovative building systems fail badly over the years. However, if you want to live dangerously, and think you can get a perfect air barrier with your bottom layer of foam – and believe your roof is never going to leak – then you could give this a try.
Stephen says
I think your thoughts on the matter are sound. In order to minimize potential condensation issues I decided to pony up and spray the necessary r-value/% in my hot roof. I have already done a flash coating and just need to finish it off with the proper %. I really appreciated the links to the zone/climate figures for cathedral ceilings. I am in zone 8, so it says I should have a min. 71% of rvalue in foam before I hit my batt insulation. Riddle me this…
I have 2×12 rafters I am filling. If I put in another 5” of foam over my existing flash coat I should have a closed-cell rvalue of around 42 (if we are at r7/inch.) The leftover cavity space should then support an r19 or r21 batt. …however…this will lower the effective ratio of my sprayfoam to batt rvalue. By my calculations this brings my relative foam to batt rvalue %’s down to something like 62% foam and 38% batt.
To keep the 71% foam to 29% batt ratio it sounds like I would have to put a minimum of r51 foam in. …Does this sound right to you? If I don’t hit my %..does this mean I still risk more condensations?
I sure don’t want to spray any more than I have to for cost reasons…but I also don’t want to wind up with condensation issues even after such an expensive process. I have thought, too, about spraying to r39/r42 and then cut and cobble for the remainder. I do have a lot of salvaged foam from some 5 ½” sips panels. …but man…what a pain to cobble.
Thoughts?
buildingadvisor says
Turns out you shouldn’t have slept through algebra class after all?
Looks like you are doing the math properly, but assuming R-7 per inch for spray foam is a bit on the optimistic side. Published numbers range from R-5 to R-7 per inch, but most research points to R-6 as a more realistic estimate.
And, yes, if you skimp on the ratio – with less foam than recommended – you risk having more condensation form on the bottom of the foam. If you read the footnotes in the table above, you’ll see that the ratios are based on the average dewpoint of the interior air during the three coldest months of the year – a pretty conservative standard.
With an airtight air barrier at the ceiling level, a moderate (Class III) vapor retarder (2-3 coats of latex paint on drywall), and reasonable indoor humidity levels, the amount of condensation should be minimal, even if you skimp a little on the foam. But given the realities of construction, it’s best to assume that nothing will be built perfectly. So I’d use R-6 per inch for the spray foam in your calculations.
And, yes, cut-and-cobble is a pain. If you go this route, invest in a professional foam applicator – makes a world of difference.
Stepthn says
If you are willing and it is not too much trouble, would you please provide me with a little further counsel regarding our previous discussion..
After spraying apx. 6 inches to my ceiling cavities I am a bit pessimistic with my final depth of insulation. I’m no professional at spraying foam and was limited to using the 600 bft kits, so my spray patterns and consistency per cavity does vary somewhat. Even though I had experience doing it before, It took me 2 kits in to get into my groove. …and wow…does it make a huge difference when you have the kits at a toasty 80 deg. (was difficult doing this in the winter!)
The end result was that, while most cavities are right at my target depth of 6” or more, there are some variances that dip well be low that. These “valleys” are sporadic but cause me to question the final step.
We had discussed the recommended charts of apx. 70% of r-value (minimum needing to be foam) if the remaining is batted. If I assume a conservative average of r-30 for my foam effort, that puts me at apx. R43 total w r13 being my batt target. …however, I’m not one that likes to hover at minimums out here for good reason, so I decided to calculate w r-30 being 80% of my total r. The result is total of 37.5 w 7.5 being the remaining batt estimate.
Now….i don’t know about you, but I have r38 batts to work w, so peeling off a measly 7.5 and trying to put this thin layer of cotton candy up just seems like such a PAIN…and maybe even unnecessary.
What all that rambling was to get to was this: If I just close up the ceiling with my wall that has an r30-r45 average (depending on where you look 😃), and leave the extra 4-5 inches void (using 11 1/4:” 2×12 joists), do I risk getting condensation on the inside of my ceiling ply? More risky without batting or less? I’m currently using plywood for the close-up w 3 layers of a good semi-gloss latex paint. I’m planning to have all my lighting outside of the insulated space, and get everything as airtight as possible.
Thank you for your time,
Stephen in Alaska
buildingadvisor says
If I follow you correctly, you’re asking whether it’s preferable to have a few inches of fiberglass on the interior side of the foam – or no fiberglass. In terms of condensation, no fiberglass is preferable as it results in a higher temperature at the lower surface of the foam. The goal is for the lower surface of the foam (the first condensing surface) to be above the dew point of the interior air. So the warmer the foam surface, the better.
Think of it this way: If you hang an unfaced fiberglass batt across the inside of a window overnight, you’ll have a lot of condensation on the glass in the morning. The insulation will make the glass very cold, below the dewpoint of the household air.
It sounds like your ceiling is painted plywood with three coats of semi-gloss. I’m not sure if you plan to seal the joints in the plywood, but as long as the foam surface is warm enough, you shouldn’t have to worry about condensation in the ceiling cavities.
FWIW, the main reason that building codes require more insulation in ceilings than walls is that it’s cheaper to pile insulation in an attic than it is in walls. It has very little to do with hot air rising. In hot climates (not yours) extra attic insulation also helps with air conditioning due to high temps in the attic. However, the same logic does not apply to cathedral ceilings, which cost about the same as walls to insulate. Bottom line: with cathedral ceilings, ceiling R-value should be comparable to wall R-value for optimal cost effectiveness.
Vickie says
Can I Put Fiberglass 4 ft. Below Foam?
I have metal roofing on 1/2-inch OSB and about 1/2-Inch closed cell insulation on the underside of the OSB. Is it ok to put fiberglass batting 3 or 4 feet below the foam, on top of the sheetrock ceiling?
buildingadvisor says
With a gap that large between the fiberglass and the foam, you’re building is gaining little or no additional insulation value from the foam. This is due to convection currents and probably air leakage in the roof cavity. If the space between the fiberglass and foam is vented, you will definitely gain no insulation value from the foam.
It’s possible that the foam will be a little warmer than the underside of the metal roofing, reducing the likelihood of condensation on the underside of the metal. However, this is best addressed by building an airtight ceiling plane to keep household moisture from leaking into the roof cavity.
Also, take a look at Table 1 in the article above to find the minimum foam thickness for your climate zone. Then put the fiberglass insulation tight against the foam for the best thermal performance.