GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Icynene & moisture

user-1146678 | Posted in General Questions on

OK some of you may remember my previous question/story about Icynene, we now have more data and more questions;-)

Here’s the beef:
I am removing Icynene OC foam from under the roof because it still stinks but here’s something that’s very odd.
It’s a 1920s house with wood planks all over, a new metal roof installed right over 3 layers of self adhering premium underlayment.
Icynene was sprayed 6 in underneath deck to fill the rafter space about 1.5 years ago.
Now, here’s what I measured.
The Icynene was left open and exposed to the interior for about 9 months now. I removed the spray foam using a mini-stripper (which appeared to be the best way so far..) from the top all the way down.

The Discovery:
The moisture meter shows the wood deck moisture is over 25%!
Now, farther down closer to the floor I have a section where the foam was covered using XPS until just before I took the foam out. The moisture there after exposing the roof deck is approx. 10%.
Rafters and ridge are all “dry”, about 5-7%

The day I did this the weather was warm 60s F and very dry around 20% RH.

*Why did the deck not dry towards the interior ‘as advertised’? I would have expected the covered section to be ‘wetter’…..
*Is 25% moisture in hardwood an issue? mold risk?

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. GBA Editor
    Martin Holladay | | #1

    1. What is your climate?

    2. In a cold climate, the phenomenon you note is well understood. Cold roof sheathing + a type of insulation that is vapor-permeable = wet sheathing.

    3. What's with the three layers of peel & stick! That's crazy.

  2. user-1146678 | | #2

    thank you!
    1. DC area. There are no vents or gaps in roof system
    2. it's sort of a mixed climate
    3. I had to spank the roofers a couple of times...they nailed it wrong and stapled it.....incredible

  3. GBA Editor
    Martin Holladay | | #3

    During a cold winter, moisture will definitely accumulate in the roof sheathing if you install open-cell foam like Icynene without an interior vapor retarder.

    Joe Lstiburek used to recommend spraying the interior side of the cured spray foam with vapor-retarder paint. Then he tested that assembly, and found out that the vapor-retarder paint is worthless as a vapor retarder when it is installed on a rough, bumpy, porous surface like spray foam. Now Joe recommends covering the cured foam with gypsum drywall, and then painting the drywall with vapor retarder paint -- or else using closed-cell spray foam, which avoids the problem of vapor transmission.

  4. user-1146678 | | #4

    That's correct Joe does indeed recommend a vapor retarder. I had written about this issue last year in this forum. I had the entire icynene ceiling tightly sealed with XPS and great stuff....until one day in May the drops started coming down and left everyone perplexed as to what's going on.
    After giving this a lot of thought and reading up on the issue, I suspect that roof may have been leaking. But I guess I need to wait until the first warm and humid days of this year before I know for fact my new setup is acceptable...

    OK, the idea with gypsum board is a little strange though because you would leave a big air gap behind it...lots of opportunity for vapors to find their way through.
    I remember when I demoed the entire area way before the icynene was installed, the mold on the 40y drywall was dense on the paper tape between the sheets and on the entire back side of the gypsum board. It seems to pass moisture really well. I find it hard to believe that paint would withstand for long especially around the seams.
    We used to have fiberglass insulation. After 40y it was black like coal....and the dense packing didn't help much against air infiltration. The mold was good evidence of how the vapor moved and where it came from.

    Anyways but I would have expected the deck to dry quickly since the perm rating of icynene is supposed to be rather high around 10 for 6in. and the foam was exposed to very dry air that day.

    Another theory some one expresses was that the foam itself acts like the wood as a vapor "battery" and amplifies the effect....."hydrophilic"?

  5. Expert Member
    Dana Dorsett | | #5


    "Another theory some one expresses was that the foam itself acts like the wood as a vapor "battery" and amplifies the effect....."hydrophilic"?"

    Like polystyrene foam (XPS or EPS) polyurethane foam (eg Icynene) is quite hydrophobic, unlike wood or polyisocyanurate. The reason the roof deck wood's moisture concentration hits 25% is because it's carrying 100% of the wintertime moisture accumulation from the air leaks and vapor diffusion from the interior side. By putting 3 layers of low permeance underlayment on the exterior of the roof deck and no ventilation under the roof deck it provides zero drying capacity toward the interior. By putting 2" of XPS on the interior (difficult to air seal adequately in a retrofit) it only provides a mere ~0.5perm path to drying toward the interior. When the sun heats the roof deck baking out some of it's the moisture, the Icynene passes the moisture easily toward the interior, but it then condenses into liquid on the far less permeable XPS. The problem isn't the Icynene, it's your overall stackup.

    Whether you can succeed in air sealing gypsum any better than you did the XPS is an open question. Leaging the Icynene in place installing a "smart" film type vapor retarder (eg Certainteed Membrain, et al) rather than vapor retardent paint or XPS as the interior vapor retarder would be a better solution, since it limits the rate of wintertime moisture accumulation, but allows rapid trying when the humidity in the cavity insulation rises. If you want a thermal break on the rafters, use (far more permeable) UNfaced EPS rather than XPS which at 2# thickness is about as permeable than standard latex paint on gypsum- sufficiently permeable to not reduce the drying performance of a smart vapor retarder's drying mode.

  6. itserich | | #6


  7. user-1146678 | | #7

    well let's assume for the sake of argument that OCSP performs as advertised and that it's indeed hydrophobic (this is something I'm not 100% convinced since I did a couple of experiments with Icynene....)
    then you wouldn't be able to put any closed cell under the roof deck in my situation.
    Moisture would find its way in, saturate the wood and have nowhere to escape in the summer months.....?
    So far I have taken down half of the smelly 18month old foam and replaced it with 2" XPS. Naturally I am very curious what will happen in the summer. I installed RH meters behind the boards for that purpose. So far not much over 85% RH at around 95F when the spring sun beats on the roof..... The peak temp last year was 130F under the roof deck but I installed the XPS in fall so I have no data for summer performance yet....

    How much water are we talking about, and what do you think is likely to happen next? Having the XPS installed should shield well against winter moisture going up there in the first place, right?

  8. Expert Member
    Dana Dorsett | | #8

    You can certainly get liquid water into open cell foam if you soak it, but it doesn't wick or retain water the way polyiso (or wood) does. It has no real moisture buffering capacity- even though moisture can condense inside the insulation layer, saturating it, that's temporary, and subject to drying by gravity- it doesn't "hold on" to the water the way wood or cellulose does, and the liquid water in open cell polyurethane can even leave via gravity, not just vapor diffusion.

    If you let the roof deck dry to well under 15% you can then put an inch or so of closed cell polyurethane on it, which becomes about a 1-perm vapor retarder and R6 of insulation, and pretty guaranteed to be air-tight to the wood. The rate of moisture accumulation into the wood is very slow through 1-perm of vapor diffusion alone, so over the course of a mid-Atlantic winter it simply can't hit the 25% level, but any moisture that finds it's way in has plenty of time to dry over the rest of the season- as long as you don't make the rest of the stackup too vapor tight. Cut'n'cobble XPS can be made to work, but it's much harder to make air tight, with higher risk of air-transported moisture making it into the roof deck, and at 2", has only about half the drying rate of an inch of closed cell polyurethane. For the money, an inch of closed cell spray foam against the roof deck with

    Be clear about what RH is- it's moisture levels only relative to the temperature. For data logging purposes it's better to use an absolute moisture measure like wet-bulb temperature or dew-point, to eliminate potential confusion. An 85%RH @ 95F would an extremely TORRID level of moisture, a wet-bulb temp of 91F= dew-point temp of 90F, which means that moisture will condense on (or adsorb into) anything 90F or cooler. That is more humidity than is usually found outdoors anywhere except the worst days in tropical rain forests! If that is truly the moisture levels you're seeing in the attic, you have roof leakage issues, not mere releases of stored accumulation from wintertime air leaks or vapor diffusion.

  9. user-1146678 | | #9

    RH or even absolute moisture depends on the actual volume of air surrounding the measurement.
    If you take a piece of freshly baked bread and seal it in a cubic foot plastic container, vapors will condense immediately inside the box.
    Keeping the bread out in your kitchen won't cause condensation on your windows, even if you house is perfectly sealed. The difference is obviously the amount of air available and its capacity to hold water.

    Hence, the actual air volume is a variable here that shouldn't be neglected. When we place an XPS board against the roof deck, at best we have a 1" air gap.
    Say it's 1" deep, 21" wide and 8 feet long. That would be roughly 2000 cubic in => 1.1 cu. ft. -> 0.033 cu meters.
    At room temperature and 100% RH that would be 20 gram of water per cu. meter * 0.033 cu meters actual volume => 0.66 grams of water -> a drop of water

    That's it.
    Say the roof reaches 130F and the RH hits 99% (still no condensate), that would increase the actual water content to 4.18 grams of water which is a teaspoon
    So I am expecting a worst case of one teaspoon of water to condensate behind each 8ft board.

    am I misunderstanding the physics behind it? whats your opinion?

  10. user-1146678 | | #10

    OK I have to add it also depends on how much water the wood is emitting as it heats up.....then it could be more or less of an increase.
    If the wood is already fully saturated with water at room temp (but it isn't in this case), then as temp goes up it will start emitting vapors. So if I understand correctly, as long as the wood isn't fully saturated at room temp, we shouldn't have an issue at 130F, or will we?
    According to the tables I found online, we can expect a reduction of about 15% of the wood's water holding capacity when going from freezing point to 130F...

    It's a confusing yet very interesting subject I have to admit :)
    Thank you for your input!

  11. Expert Member
    Dana Dorsett | | #11

    I was assuming you were measuring the humidity in the full attic space, not a small cavity between the foam & roof deck. It's almost irrelevant what the RH in that small cavity is, and it may be easier/better to monitor the moisture content of the roof deck itself, since that's what you really care about.

    But to be precise, the absolute % moisture content/dewpoint of the air is not volume dependent, though it will determine the grams/cubic-meter. But it will be the same ratio for a 1cc space or a cubic kilometer space.

    But in that small unvented cavity you will NEVER experience summertime moisture condensing as liquid in the cavity- you should be expecting ZERO ccs/teaspoons/drops of condensate. The temperature of the surface of the foam facing the cavity will always be the same as the roof deck- as the roof deck cools it then re-adsorbs the moisture that it released during the heat of the day. Even though the wood can be holding many times the amount of water contained in the tiny volume of air in the cavity, if you get liquid water in that air space, it can only be from bulk-moisture leakage, not condensation. If the wood wasn't there, yes, you would get condensation cycles whenever the temp dropped below 90F. But in this case the wood is both the source & destination of the moisture in the cavity- it can't release more moisture than it can adsorb, so you get adsorptions cycles with the changing temperatures, not liquid-condensation cycles. During the heat of the day when the dew point of the cavity air exceeds that of the larger attic space, some of that moisture diffuses through into the attic, but it's a fairly slow process through 2" of XPS. At the same R-value you'd be getting on the order of 4-5x the moisture release were it unfaced EPS rather than XPS.

    In winter there is a remote possibility of experiencing condensation in that small cavity if the wood becomes fully saturated and can't adsorb more moisture, but that would only happen if the wood started out the cooler season already at near-saturation (well over mold-inducing levels), or if you had substantial air leakage into the cavity from the interior.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |