GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Musings of an Energy Nerd

Open-Cell Spray Foam and Damp Roof Sheathing

When open-cell spray foam insulation is installed on the underside of OSB roof sheathing, the sheathing sometimes gets damp

The roof sheathing on this 13-year-old Icynene-insulated roof was rotten. The problems with this roof in Des Moines, Iowa, were described by builder Mark Parlee in his 2010 article, “Repairing a Rotting Roof.” The darker (rotting) sheathing is above the home's cathedral ceiling, while the undamaged sheathing in the foreground is above the attached garage.
Image Credit: Mark Parlee

UPDATED on July 8, 2015

Now that insulation contractors have been installing spray foam insulation on the underside of roof sheathing for several years, we’re beginning to accumulate anecdotes and data on successful installations and failed installations. The anecdotes and data are enough to provide a few rules of thumb for designers and builders who want to install spray foam on the underside of roof sheathing.

Increasingly, building scientists are investigating why OSB roof sheathing on many spray-foam-insulated roofs stays damp for months at a time. Most of these damp-sheathing problems involve open-cell foam rather than closed-cell foam.

I’ve been reporting on wet-sheathing problems arising from the use of open-cell spray foam since 2005, when I wrote two articles on the topic for Energy Design Update (“Vapor Retarders and Icynene,” April 2005, and “Every Failure Holds a Lesson,” July 2005). As originally understood, the problem with open-cell foam was that it is vapor-permeable, and therefore allows moisture in the interior air to diffuse through the insulation and reach the cold roof sheathing during the winter.

Five years later, Mark Parlee, an Iowa builder, wrote a seminal article on an Icynene-insulated roof with rotten roof sheathing. His article, “Repairing a Rotting Roof,” was published in the June 2010 issue of the Journal of Light Construction. One of the factors that contributed to the failure described by Parlee was high indoor humidity.

Open-cell foam under roof sheathing can be risky

At a recent building science conference in Florida (Conference on Thermal Performance of the Exterior Envelopes of Whole Buildings XII, December 1-4, 2013), two academic papers were presented that shed light on questions surrounding the moisture content of roof sheathing that has been insulated on the underside with spray polyurethane foam.

One paper discussed a field study that found that even in a relatively warm climate (South…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial

116 Comments

  1. BetterBuildingWorks | | #1

    Could a coat of primer...
    ...on the sheathing prior to the installation of open-cell spray foam of the roof deck reduce the risk of saturating the OSB? We have enclosed several attics over the past two years with open-cell in our mixed climate and have not observed these issues...yet, at least.

    Closed cell sounds like it performs much better in practice but the cost is significantly higher and diminishes ROI but of course we don't want to jeopardize the lifespan of construction materials either. Has anyone else experienced this in a mixed climate or tried another method?

  2. GBA Editor
    Martin Holladay | | #2

    Response to Chase Counts
    Chase,
    Frankly, I'm skeptical of whether a coat of primer would make much difference. The best way to answer the question, of course, would be to experiment with primer in two adjacent test huts.

    Q. "Has anyone else experienced this in a mixed climate?"

    A. We have reliable reports of saturated roof sheathing adjacent to open-cell spray foam from Montana, Iowa, Vermont, and Massachusetts. My article reports on evidence of damp sheathing in a building in South Carolina.

  3. Richard Beyer | | #3

    Comment
    Martin did you or a third party inspector inspect the spray foam to determine if it was installed in accordance with the manufacturer's written standard? I'm betting industry will defend this issue and blame it on the installation of the spray foam insulation or the roofing contractor.

  4. user-963733 | | #4

    All climates?
    This warning against open cell foam in all climates, sure seems to clash with what Joe Lstiburek has promoted for a decade. Do I understand correctly?

    There have been some trials in very dry climates, such as Phoenix AZ. Is it a mystery how this foam design can be risky in such a dry climate? It is to me, and would be really nice if someone could explain in elementary detail how it would work in a hot-dry climate.

    I do know about 15 years ago, there was a series of house experimentally built in the Cinco Ranch neighborhood of Katy TX, just west of Houston. It was part of the Building America program, and Pulte was the builder. Definitely in the hot-humid climate category. Apart from a paper or two, the results have been mysterious and little publicized. Some other builders who might be in a position to know, hint darkly about roof decking rotting, and lawsuits resulting -- I get the impression news is being withheld on this matter. If anyone can shed any light on lessons learned here, please share what you know!

  5. GBA Editor
    Martin Holladay | | #5

    Response to Richard Beyer
    Richard,
    The bulk of this article reports on two scientific papers, each written by a team of three researchers. The papers were presented at a conference I attended in Florida.

    One paper referred to moisture accumulation in the roof sheathing of a test facility in South Carolina; I don't know if the spray foam in that building was checked by "a third party inspector," but it's reasonable to assume that the researchers at the test facility did their best to assure that the spray foam was well installed.

    The other paper reported on a computer modeling study. The computer model assumed, as far as I know, that the foam was installed well.

    The phenomenon of vapor diffusion through open-cell foam (even well-installed open-cell foam) in cold climates has been well documented. Problems generally don't occur unless there is high indoor relative humidity; however, such houses exist.

    The modeling study listed many factors that affect the moisture content of the roof sheathing on this type of roof. As I reported, here are the factors that raise the risk: north roofs are riskier than south roofs; attics with tight ductwork are riskier than attics with leaky ductwork; attics with leaky ceilings are riskier than attics with tight ceilings; and homes with high indoor RH are riskier than homes with low indoor RH.

    In all climate zones, open-cell foam was riskier than closed-cell foam.

  6. GBA Editor
    Martin Holladay | | #6

    Response to Mark Johnson
    Mark,
    Q. "This warning against open-cell foam in all climates, sure seems to clash with what Joe Lstiburek has promoted for a decade. Do I understand correctly?"

    A. Yes.

    Concerning your question about "mysterious and little publicized" failures in Texas, I'm not sure what you are referring to. I know that the Building Science Corporation helped a builder solve high indoor humidity problems in some Building America homes in Texas by installing stand-alone dehumidifiers (in addition to the split system air conditioners).

    I also know that there have been clusters of problems with wet roof sheathing in low-slope roofs in the desert Southwest. The main cause of these failures was night sky radiational cooling that make the roof sheathing colder than the outdoor air temperature. The source of the humidity that rotted those roofs was interior moisture. These roofs were insulated with fiberglass batts -- once again, a "breathable" insulation that contributed to the failure.

  7. Ken Levenson | | #7

    smart vapor retarder can help
    An inboard smart vapor retarder and airbarrier membrane like INTELLO or DB+ can help. As the roofing is vapor impermeable and insulation is vapor permeable (breathable), a control layer inboard with the property of these membranes, providing a vapor retarder in winter to prevent wetting but be vapor open in summer to allow drying inward, while also providing airtightness, can make them robust assemblies. I might suggest Alex Wilson's GBA blog post regarding them: https://www.greenbuildingadvisor.com/blogs/dept/energy-solutions/smart-vapor-retarders
    And full disclosure, my company, 475 High Performance Building Supply, provides these materials to the US market.

  8. Richard Beyer | | #8

    Question for Martin
    Martin, Thank you for your response. I think you just spelled out what the manufacturer's claim of fault will be. "It's the building owner and/or designer's problem", not the product, when thing's go south as this article illustrates. I'm currently working on a multi-million dollar home situated on a major waterway in Connecticut where open cell spray foam is installed in the walls/roof and closed cell was installed in the floor above the crawl space. Note:The floor is above a conditioned crawl space with flood gates. Here's the kicker... the room is @ 12 feet wide X @ 22 feet long with a custom spa / hot tub which holds 1400 gallons of heated water which is Ozone treated. The room does not have windows which open but does have a door to a patio and two interior doorways @ 60 inches wide with double doors. The roof assembly is cedar shake. sheathing, 2" air space, cleats on the sides of the rafters and luon attached to the cleats. Open cell is attached to the luon. What's your opinion? Do you think this home owner may have future trouble?

  9. user-626934 | | #9

    Regarding the South Carolina testing...
    Martin, regarding the South Carolina test facility, you reference William Miller as believing that the moisture accumulation was perhaps due to summertime air leakage depositing moisture inside the attic and then subsequent night-time vapor diffusion from inside to outside. If it was an air-tightness problem, then this would seem to be a problem with the application….not the spray foam itself. Did they not test/inspect the air-tightness of the attic?

    Regarding Miller's other hypothesis of post-rain, solar-driven moisture accumulation in the South Carolina study - IF this were the main driving force, wouldn't closed cell foam be even worse, since the moisture would be driven into the OSB, but stopped (more or less) from drying to the inside?

  10. user-963733 | | #10

    The Cinco Ranch Houston project
    I am attempting to attach a copy of a 2003 BSC paper titled "Residential Dehumidification Systems Research for Hot- Humid Climates". Authors are A.F. Rudd, J.W. Lstiburek, P. Eng, and K. Ueno. This is exactly the project I was wanting to describe. It documents trying to help a builder (Pulte) with dehumidifiers and several other methods. When they refer to the 3 homes with cathedralized attic insulation, those very same researchers designed the original construction, so they were addressing their own problems. To be fair any home in this climate could easily have high indoor humidity, often due to high infiltration and feeble air conditioner setup.

    Of course they do not document their own built homes had moisture problems, I think they may have had a clue at the time of this paper. But I don't know with any certainty, it is sad when hearsay is the best news source one has.

    It seems to me generic good advice is not to be an early adopter, to wait until a new material or technique has proven itself a decade in the field before embracing it. Everybody tells me ducts in conditioned space is the only intelligent way to go, but evident to me is there are multiple ways to do that and only some ways don't turn around and bite you.

    Locally I have been hearing that exposed foam in the attic could be considered a fire hazard unless kept behind another layer of sheet rock. Allegedly this has the status of code violation (not forgetting code enforcement is very lax in TX). If this bears out, the added cost may tip the economic balance away from this concept. In its place I have heard emphasis on duct sealing and ceiling-to-attic air tightness, while tolerating HVAC equipment in an unconditioned attic. Of course ducts and even equipment could be brought inside, at a cost.

    Interesting topic. Hope this helps.

  11. Expert Member
    MALCOLM TAYLOR | | #11

    Response to Mark Johnson
    It appears that being an early adopter of both double wall or foam insulation techniques proffered by many as being solutions to increasing building envelope efficiency would have left you with sheathing moisture problems. Those who stood on the sidelines and watched things shake out may have the last laugh.

  12. wjrobinson | | #12

    OSB is not a quality material. So don't expect it to be!!!
    OSB is low cost low quality material. It rots very easily. It gets too wet very easily. Customers today use a home with out thinking for one second about such things as moisture. I have customers telling me pasta can only be cooked without a lid. I have customers running humidifiers 24/7 in every room. I have customers with fifty plants along a wall of windows in a room just big enough for the plants. I have customers that don't go in their cellar ever that never knew their sump system stopped working a few years back.

    I also have homes that have open cell with not even close to a problem in the sheathing. Just an average home with less than out of control inhabitants, no huge plant set up, no humidifiers gone wild..... well... you get the idea.

    Careless house use and crappy materials.... are the problem.... I will conduct the study for a dollar. Save your money.

    Build a crappy house... live in it out of control... let's call the lawyers and start some million dollar studies. That's the 21st century selfie pic.

    aj

  13. GBA Editor
    Martin Holladay | | #13

    Response to Ken Levenson (Comment #7)
    Ken,
    You are right: an interior vapor retarder is recommended when open-cell spray foam is installed on the underside of roof sheathing in cold climates, and GBA has consistently recommended the use of interior vapor retarders in this type of application for any builder who chooses open-cell spray foam.

    Nevertheless, it's still true that open-cell spray foam is riskier than closed-cell spray foam in all climate zones.

  14. GBA Editor
    Martin Holladay | | #14

    Response to Richard Beyer (Comment #8)
    Richard,
    It certainly sounds as if the builder or designer of the project you describe is taking on a lot of risk.

    It makes no sense to insulate the floor above the crawl space; it would have been better to insulate the crawl space walls (and to seal the crawl space vents).

    Using open-cell spray foam for the roof assembly above a spa / hot tub room is extremely risky. It sounds like the builder has included a ventilation channel above the open-cell spray foam, which helps a little, but depending on the vent channel to keep the roof out of trouble is risky.

    The best way to insulate this type of wet room is to put all of the insulation above the roof sheathing. This is usually done by following the PERSIST method: covering the top of the roof sheathing with peel-and-stick (rubberized asphalt), followed by several layers of rigid foam with staggered seams.

  15. GBA Editor
    Martin Holladay | | #15

    Response to John Semmelhack (Comment #9)
    John,
    Q. "Regarding the South Carolina test facility: ... If it was an airtightness problem, then this would seem to be a problem with the application, not the spray foam itself. Did they not test/inspect the airtightness of the attic?"

    A. Good question; I don't know. However, it's fair to ask another question: if sealing an attic with open-cell spray foam can leave enough air leakage paths to cause this type of problem (and I know that the mechanism is, at this point, hypothetical), and if the situation was unnoticed by researchers until problems developed, then couldn't the same situation happen at hundreds or thousands of other homes?

    Q. "Regarding Miller's other hypothesis of post-rain, solar-driven moisture accumulation in the South Carolina study - IF this were the main driving force, wouldn't closed cell foam be even worse, since the moisture would be driven into the OSB, but stopped (more or less) from drying to the inside?"

    A. Yes. The other researcher who has looked into this phenomenon is Armin Rudd, who sometimes calls this "ping-pong" moisture (because it comes and goes as the sun comes out). Rudd blamed inward solar vapor drive for the phenomenon a decade ago, and then began to have his doubts about what was going on when some data didn't fit the theory. At that point he announced, "It would be great to have a million dollars for a study to figure out what is going on in these roofs," but the money didn't materialize. The last time I spoke to Rudd about the issue, he said that it was still a mystery.

  16. GBA Editor
    Martin Holladay | | #16

    Response to Mark Johnson (Comment #10)
    Mark,
    I am very familiar with the study you refer to -- "Residential Dehumidification Systems Research for Hot- Humid Climates" -- because I wrote an article on the study for the January 2003 issue of Energy Design Update. My article was called, "Houston Study Looks At Dehumidification Options."

    Frankly, I'm not sure what the relevance of that study is to the issue discussed here. The researchers note that it's important to control interior humidity in a tight, well-insulated home with high-performance windows. The researchers were surprised to find high indoor humidity at some Building America homes they were involved with in Houston. They developed several solutions to the problem and recommended implementing the best performing option, which was also the least expensive (installing a stand-alone dehumidifier).

    What's the point? Interior humidity levels can sometimes be surprisingly high. That fact should only make installers of open-cell spray foam more nervous, not less nervous.

    You are certainly right that spray foam needs to be covered with an intumescent coating or, even better, with drywall to reduce the fire hazard. You are also correct that many building inspectors have been looking the other way rather than requiring compliance with these requirements.

  17. GBA Editor
    Martin Holladay | | #17

    Response to Malcolm Taylor (Comment #11)
    Malcolm,
    On the "early adopter" problem: it's certainly true that building scientists learn more every year, and are able to refine their recommendations based on better knowledge. It's also true that prominent building experts have made recommendations that needed to be revised once research refined our understanding.

    But we don't really have the luxury of building poorly insulated, leaky houses as we wait for more research results. Fortunately, most homes with double-stud walls and most homes with spray-foam-insulated roofs are performing quite well. The ones with problems are usually the ones with high indoor humidity -- a factor that builders can't really control. So there is reason for concern, but not panic.

  18. GBA Editor
    Martin Holladay | | #18

    Response to A.J. Builder (Comment #12)
    A.J.,
    You are right that plywood stands up to moisture better than OSB. But if researchers are finding very high moisture content readings in roof sheathing -- and they are, in some cases -- then those high levels will be enough to rot plywood eventually. I don't think that switching from OSB to plywood is the entire answer to this issue.

    You're a builder. You say that some of your clients keep their indoor humidity very low, and some keep their indoor humidity very high. Does that fact worry you? Do you ever have sleepless nights worrying about your clients with humidifiers and 50 house plants?

  19. Expert Member
    MALCOLM TAYLOR | | #19

    Response to Martin,
    The way building scientists or proponents of new technologies report their recommendations seem to share certain characteristics with the way medical discoveries or health advice are made public - that is as facts as opposed to working hypothesis. At any given point they seem sure they know the whole picture with all the attendant consequences, and this almost always turns out to be untrue. I'm just suggesting a bit less enthusiasm and a bit more humility might save us all some future grief.

  20. GBA Editor
    Martin Holladay | | #20

    Response to Malcolm Taylor
    Malcolm,
    As I'm sure you know, one of the strengths of the scientific method is that our body of knowledge is always being refined in light of new data. This Renaissance approach was disturbing to the medieval mindset, which assumed that knowledge was fixed and unchanging.

    Ideally, American students would get a good grounding in the scientific method in the 7th and 8th grades, so that citizens would have a firm basis for understanding reports of scientific findings. Sadly, that grounding is lacking.

    Building science is a relatively new branch of physics, and we learn more every year. As a journalist, I don't think it would be responsible to withhold data or research results just because I know that current recommendations may be overthrown in 10 years. Readers of this website deserve to know the most recent data available. I have to trust our readers to understand that, as in any scientific field, new data may trump current understandings in the future.

  21. jmhoerter | | #21

    Response to Martin Comment #5
    Martin,
    In my opinion, some of the inputs that raise the risk of high moisture content are dependent on one another so they do not hold up independently. Specifically, you say attics with leaky ceilings are riskier than attics with tight ceilings. Building science suggest that when you create an unvented attic, you must remove the vapor retarder and insulation from the attic floor to ensure there is plenty of air and humidity exchange between the living space and the attic. Assuming you keep your home at a reasonable humidity level then the humidity in the attic is controlled. Now, if you have high humidity levels in your home, then the humidity in your attic will also be high. In this case, possibly having a tight ceiling is better, but not when the humidity levels in your home are controlled. I would like to say that this research further emphasizes the importance of removing the insulation and vapor retarder from the floor when creating an unvented attic, a step that is sometimes ignored due to added labor. Your comment saying a leaky ceiling increase the risk suggests keeping the insulation and vapor retarder on the floor is a benefit.

  22. GBA Editor
    Martin Holladay | | #22

    Respose to Jason Hoerter
    Jason,
    You are right that the various factors that affect the moisture content of OSB above a sealed attic are interrelated. The researchers I cite are aware of these interrelations, and WUFI software attempts to take these interrelations into account.

    The major mechanism for moisture transfer from the conditioned space below an attic into the attic above is air leakage. Vapor diffusion plays a relatively small role -- something to keep in mind when considering whether the presence or absence of a vapor retarder on the attic floor matters.

  23. wjrobinson | | #23

    Yes Martin as to risky business thoughts
    Anyone's risks and most risks cross my thoughts daily. That is me. Calculated educated risk taking is something of a passion and part of my work and play. Whether flying a hang glider off a thousand foot cliff edge, surfing with sharks somewhere near or far, or trying to talk a client into adding antifreeze to a hydronic setup so that there is no surprise to find out setting back a tstat too far or running a woodstove so much that a back room freezes... floods...

    Anyway... we all in the business... know water water water... from where to there... because of (you name it).. creates lots of economic activity.

  24. wjrobinson | | #24

    Share what you know
    As to the problems homeowners cause... or I cause... that is a main reason why many of us come to sites like GBA. Very rewarding to actually solve water problems... insulation problems... and or relay some thoughts as what someone might be doing that maybe they should not do... like be the cause of too much water in a structure doing damage. Ultimately how a home is managed and maintained is out of my hands.

    Some people just need to boil a few gallons of water daily. After I share my thoughts they are free to boil gallons or not... reap and they shall sow... someone jump in and explain me..lol

    I sleep like a rock by the way.

  25. dankolbert | | #25

    Last paragraph
    I'm not sure I draw such a broad conclusion from your story, Martin. I've always been nervous about open cell SF for the reasons it seems to be causing trouble - it seems to accept moisture faster than it distributes or dissipates it. I'm not sure it's reasonable to tar cellulose or fiberglass with the same brush.

  26. GBA Editor
    Martin Holladay | | #26

    Response to Dan Kolbert
    Dan,
    There is nothing in my last paragraph to indicate that cellulose and fiberglass behave the same way as open-cell spray foam. Every material has different properties.

    I think it's fair to say that if a builder tried to install fiberglass batts or blown-in fiberglass directly against the underside of roof sheathing -- without any rigid foam on top of the roof sheathing -- then the results would be even worse (more risky) than is the case with open-cell spray foam.

    So, if we want to discuss tarring with the same brush, I'll go right ahead and tar fiberglass with the same brush. Here's my warning: don't install fiberglass insulation directly against the underside of roof sheathing unless you have installed a significant layer of rigid foam above the roof sheathing.

    The situation with cellulose is a little more complicated, and I address the ins-and-outs of the debate in my article, How to Install Cellulose Insulation.

    The main point I was trying to make in my concluding sentences is that "breathability" is not an unalloyed virtue. Whenever you are designing a floor assembly, wall assembly, or roof assembly, you have to use your brain. Advocates of "breathable" assemblies often fail to use their brains. Sometimes it's good to choose materials that are vapor-permeable; other times, you definitely want to choose materials that are vapor-impermable.

    So my message is: when choosing materials, you have to think.

  27. Joe Lstiburek | | #27

    Peer review should have addressed the limitations of the papers
    The studies mentioned from Clearwater are misleading. The Des Moines and cold climate failures are a separate issue.

    Open-cell foam does not work without a vapor retarder in Climates Zones 5 and higher. That is why the code prohibits this. This is because the vapor openness of the open-cell spray foam is overwhelmed by high interior levels of moisture. The fact that many installers and manufacturers promote this and that the code is not enforced is troubling but it is not a building science issue. Des Moines should not have happened because open-cell spray foam should not have been used. People "get away" with open-cell foam in cold climates if the interior moisture levels are low. "Getting away" with things is not a sound business or building science strategy.

    The Clearwater papers are more troubling because peer review should have addressed the limitations of the work.

    Unvented attics like unvented crawlspaces are unfortunate terms but for reasons of history and popular usage we are stuck with them. The correct terms should be conditioned attics and conditioned crawlspaces. The misunderstanding arising from this terminology confusion is painfully obvious in the Clearwater work.

    Both conditioned attics and conditioned crawlspaces must have a mechanism of moisture removal just as conditioned houses must have a mechanism of moisture removal. In houses we do this with either dilution ventilation or mechanical means - either an air conditioner or a dehumidifier depending on load. During the winter dilution ventilation is used. During the summer in humid climates the air conditioner does the job - except in ultra-low-load buildings where supplemental dehumidification is necessary.

    Conditioned attics must be coupled to the conditioned part of the house in humid climates in order to control the moisture levels in the attic. This is particularly necessary with open-cell foam due to its moisture storage characteristics.

    The moisture in the unvented attic with open-cell foam studied by Oak Ridge was interior moisture. It did not come through the shingles and roofing paper and roof sheathing through solar driven moisture. We showed this in our Houston work a decade ago. We told them this. This should have been picked up by peer review.

    There was no mechanism for moisture removal in this attic - a simple approach is the use of a supply air duct coupled with a return path. This was not done. In many cases it has proven to be unnecessary when supply ducts are leaky and ceiling planes are leaky - typical construction - the air change provided between the attic space and the house with leaky ducts and a leaky ceiling serves to remove the moisture in the attic. Where supply ducts are tight and where the ceiling plane is tight this mechanism does not work and an active supply duct and deliberate return path is necessary. This should not be news to experienced researchers. It is pretty well understood by many installers using spray foam. But it is apparently not well understood by everyone and clearly a code change would be helpful. No air change measurements via tracer gas were done to provide information about the lack of communication between the attic and the house. Modeling was used by individuals with little or no experience with this type of construction. The modeling was reviewed by people with even less. The comments on moisture sources and energy impacts was also misleading.

    Unvented attics do not generate moisture (note the change in term). Moisture is not passing through the attic roof deck into the attic. Unvented attics do not add an additional moisture load to the house. There is no difference in the moisture load in a house with a vented attic and an unvented attic. The moisture load in houses with vented attics and unvented attics is clearly the same in houses with equal airtightness and equal foundation construction and equal interior occupancy and therefore interior moisture loads. There is therefore no difference in energy load to control this moisture. So what is going on? Unvented attics provide a complication that can be a modeling and measurement dog's breakfast.

    An unvented attic does not affect the amount of moisture in a house except in an indirect way. Unvented attics often lead to lower air change in building enclosures because they typically result in tighter construction. During the winter that means that an increase in dilution ventilation is required via mechanical means to bring the air change up to the air change equal to that of a more leaky house. During the summer the unvented attic leads to a reduced air change which reduces the quantity of exterior moisture entering the building and therefore it typically reduces the interior moisture load - which is a good thing. But the unvented attic also often reduces sensible heat gain because ducts are within the conditioned space and this can lead to a part load humidity issue from oversized air conditioning systems. In any event in buildings with equal air change an unvented attic does not increase interior moisture loads but can lead to an increase in interior relative humidity if the air conditioning system does not remove this moisture. This is therefore a sizing and equipment issue not a vented unvented attic issue.

    The ping-pong effect has been mentioned and discussed by myself for over a decade. It is of no significance if a moisture removal mechanism is in place. It is not an issue with closed cell foam for obvious reasons - closed cell foam does not store moisture. Moisture laden air is less dense and more buoyant than dry air. Heated air is more buoyant than unheated air. That moisture migrates upwards into attic spaces should not be a mystery. In materials such as open-cell foam (and cellulose) some of this moisture is stored particularly at night and during the winter (so we have both diurnal and seasonal effects). During the day the sun drives this moisture back into the attic. During the summer the sun drives this moisture back into the attic. If this moisture is not removed the quantity increases. Note that this moisture comes from the inside. In sloping roofs the moisture marches up the slope with each "ping" due to buoyancy and tends to accumulate at ridge areas first. We noticed this in cellulose insulated unvented roofs in Houston. The issue in Houston was resolved by removing the insulation at the ridge areas to allow the assembly to dry and to provide a mechanism of moisture removal in the attic spaces - air change or dehumidification. With open-cell foam the quantity of moisture stored is less with each cycle or ping than cellulose but over time if this moisture is not removed it can be quite substantial.

    The second study is a modeling study that does not address the moisture removal mechanism nor buoyancy. It is fundamentally flawed for these reasons.

    To condem open-cell foam based on ignorance, hearsay, poorly designed experiments and modeling approaches that are beyond their range of applicability in the face of several decades of successful field experience is a disservice to the industry. I have no particular love for open-cell foam or for closed cell foam or for cellulose or for fiberglass. I think they all have a place and I recommend the use of all of them. But each has its limitations and when used their limitations must be accounted for.

    Respectfully yours,
    Joseph Lstiburek

  28. GBA Editor
    Martin Holladay | | #28

    Response to Joe Lstiburek
    Joe,
    Thanks very much for your knowledgeable and useful comments.

    You and I are in agreement on most of the points you raise, although my emphasis probably differs from yours. When it comes to disagreements between prominent researchers who challenge whether published ASHRAE papers have been properly peer-reviewed, I'm going to have to step back a bit and wait for the academic controversy to be settled over the next few months. Clearly, your perspective differs from that of the six authors cited. Such disagreements are healthy and usually fruitful.

    In “Roof and Attic Design Guidelines for New and Retrofit Construction of Homes in Hot and Cold Climates,” the three authors (William Miller, Andre Desjarlais, and Marc LaFrance) wrote, "Field measurements imply that some of the moisture from a previous rainstorm migrates to the underside of the shingles and underlayment. Irradiance drives moisture from an earlier rainstorm into and through the OSB deck and open-cell foam." Note the use of the verb "imply." The researchers posited that the data implied that inward solar vapor drive was responsible for the elevated moisture levels. They did not state that the matter was settled.

    After the paper was published, William Miller sent me an e-mail in which he stated, “We still are working to determine the mode by which moisture crosses the envelope... I believe it occurs in two different paths:
    1. The sealed attic is not airtight and outdoor moisture penetrates. There is no ventilation and night-sky radiation causes the trapped humid air in the attic to diffuse into the spray foam.
    2. The roof deck becomes wet during the evenings because of night-sky radiation and subsequently the condensate is forced into the roof deck by irradiance the next day.
    Both modes result in moisture storage in the spray foam, and the stored water begins to damage the wood deck.”

    I believe that I accurately reported that the three researchers had not definitively pinned down the source of the high moisture content in the roof sheathing.

    You wrote, "Open-cell foam does not work without a vapor retarder in Climates Zones 5 and higher." I believe that I have stressed that fact for many years, and repeated that fact in this article. GBA recommendations have reflected that fact for years.

    You wrote, "Conditioned attics must be coupled to the conditioned part of the house in humid climates in order to control the moisture levels in the attic. This is particularly necessary with open-cell foam due to its moisture storage characteristics." It could be argued that your recommendation flows naturally from the warnings voiced in my article. When open-cell spray foam is installed on the underside of roof sheathing, designers and builders need to pay attention to the RH of the air in the attic, or the roof sheathing can get damp. We're on the same page here, Joe. If we don't pay attention to these issues, builders can get into trouble.

    You wrote, "The ping-pong effect has been mentioned and discussed by myself for over a decade. It is of no significance if a moisture removal mechanism is in place." Fair enough. One way to stay out of trouble is to include a moisture-removal system in the attic. Another way to stay out of trouble is to use closed-cell spray foam.

    To lower the RH in attic air, you recommend a "mechanism for moisture removal in this attic - a simple approach is the use of a supply air duct coupled with a return path." In essence, your recommendation mirrors an observation made by researchers Simon Pallin, Manfred Kehrer, and William Miller, who noted that "a high duct leakage has a positive effect on the MC of the OSB, most likely due to the dehumidifying effect of the HVAC cooling coils, which, by a higher rate of air leakage, will have a higher influence on the vapor content of the attic air during the operating cooling mode.” Your recommendation to include a small supply air register in the attic, along with a small return-air grille, has the same drying effect as leaky ductwork.

    You wrote, "To condemn open-cell foam based on ignorance, hearsay, poorly designed experiments and modeling approaches that are beyond their range of applicability in the face of several decades of successful field experience is a disservice to the industry." I certainly agree with your warning about modeling studies that are beyond their range of applicability, which is why I wrote, "When interpreting the results of computer modeling studies, it’s important to remember that modeling results sometimes differ from field study results."

    But I stand by my statement that installing open-cell spray foam on the underside of roof sheathing is riskier than installing closed-cell spray foam. You have explained that the risk can be lowered by conditioning the attic air (using the HVAC system to lower the RH of the attic air). That's certainly a helpful recommendation, and I will certainly add your recommendation to my concluding remarks.

    But it's worth emphasizing that the need to condition the attic air is related to the riskiness of open-cell foam. And it's also worth emphasizing that if three Oak Ridge National Lab researchers can end up with a test facility with damp roof sheathing because they failed to properly condition the attic air, then the same error could easily be made by Bob the builder.

  29. dankolbert | | #29

    Now you tell me
    "So my message is: when choosing materials, you have to think."

  30. GBA Editor
    Martin Holladay | | #30

    Response to Dan Kolbert
    Dan,
    You'd be surprised how often builders and designers forget to follow that simple, useful principle.

  31. dankolbert | | #31

    Sadly
    No, I wouldn't. I do reno, thus playing clean-up.

  32. MParlee | | #32

    smart vapor retarder can help by Ken Levenson
    Ken, The difficulty in installing a smart vapor barrier in an attic would be very great. If it is not installed to perfection then it would not work because the vapors will migrate to the leak points. If there is a spray applied smart barrier then that changes things.
    This discussion is a good one and leads me to believe it would be a good idea for me to get back over and take some moisture readings at various places on this home.

  33. GBA Editor
    Martin Holladay | | #33

    Response to Mark Parlee
    Mark,
    While air barriers have to be nearly perfect to work well -- a small hole can defeat the purpose of an air barrier -- vapor retarders don't have to be perfect. Even when a vapor retarder has holes covering 5% of the material, it is still 95% effective at controlling vapor diffusion.

    I am wary of any spray-applied vapor retarders when these are applied to cured spray foam. Researchers at the Building Science Corporation have tested vapor-retarder paints on cured spray foam, and found that the paints are ineffective at limiting vapor diffusion. For more information on this issue, see Joe Lstiburek Discusses Basement Insulation and Vapor Retarders.

  34. Allan27 | | #34

    Attics
    Here in Houston I've used open cell foam in attics and walls in my new homes for about 5 years, spraying the foam between rafters from the top plates up to the ridge. I've not had any callbacks or complaints regarding moisture. My AC contractor has always insisted in having a supply duct (with return) in the attic "dumping" some conditioned air in the attic. Honestly I always wondered why he did this, I guess in retrospect he is doing the right thing.

  35. user-611979 | | #35

    How does this relate to cathedral ceilings?
    I'm trying to get my head around how this would relate to a cathedral ceiling situation. We are in Atlanta (CZ2), do some second story additions/attic build outs, and use open cell foam for this on a regular basis. It would seem to me that this fully conditioned attic situation should remove the extra moisture from the attic air. Thus, it would seem to leave the main source of any moisture at the roof deck to be coming from being driven through the asphalt shingles, through the underlayment, into the decking. I know this wasn't necessarily the main focus of the studies, but looking at the situation I described would seem to be able to isolate how much moisture is coming through from above.

  36. GBA Editor
    Martin Holladay | | #36

    Response to Eric Kronberg
    Eric,
    The three researchers (William Miller, Andre Desjarlais, and Marc LaFrance) aren't sure where the moisture is coming from; they are hypothesizing.

    In Comment #27, Joe Lstiburek expresses his view that the source of any moisture is the interior air, so that if you are able to control the indoor relative humidity, you shouldn't have any problems (at least in warm climates). That advice is consistent with field experience.

    If you are a builder who is worried that there is no way to control occupant behavior (and after all, the occupants are the ones who are in control of the indoor relative humidity), then you may feel that it's worth switching to closed-cell spray foam.

  37. Expert Member
    ARMANDO COBO | | #37

    Manufacturers, Installers & Building Officials
    I've been teaching and preaching about this practice of 5.5" open-cell foam (R20) sprayed under the roof decking without any rigid foam above the roof decking for years. It's a common practice in TX, NM and the whole SW. I've talked to Manufacturers and Retailers, and their response is that their open-cell foam becomes air-impermeable at 2.5" or 3.5", so 5.5" (R20) is OK to use under the performance code (R806.4.5.1), even though the Builder may not be following the performance code.

    Just last year at Summer Camp, I spoke with two manufacturer's reps about the same issue, and both told me that since their O.C. was air-impermeable at 2.5" or so, and as long as they can "get away" with it, they are in the business to sell foam.

    Building Officials are buying into this theories and allowing Builders to follow those practices. It's all a shame. I do hope more studies and reports bring to light how bad this practice can be!

  38. JoeUSAF | | #38

    My builder is just completing
    My builder is just completing the dry wall installation and taping of my new Energy Star to be qualified home in Andover, CT. The home has engineered 2 X 6 wall stud construction with plywood sheathing on both the unvented cathedral roof and the walls. Fortunately, the cathedral ceiling has 4" of closed cell spray home directly under the roof sheathing followed by open cell spray foam to fill the remainder of the rafter cavities covered by 1 1/12 " of XPS and finally dry wall. I hope this configuration prevents the rotting of the roof sheathing which is covered by #30 felt and lifetime high reflectivity asphalt shingles.

    After reading the discussions about "Open-Cell Spray Foam and Damp Roof Sheathing", I am wondering if the same priniciples of drying apply to my walls. My walls have the following configuration; vinyl siding, 3/4" foil faced polyiso carefully taped at seams, Tyvek, 1/2" plywood sheathing carefully taped at seams. open cell spray foam filling the entire 2 X 6 cavity, followed by drywall. I had intended to coat the interior side of the wall sheahing with 2" of closed cell spray foam followed by open cell spray foam. I did not do this because it was advised in much of the latest home building literature, "not to apply impermeable insulation on both sides of wall sheathing." which would prevent drying in both directions Although we are in the process of finishing the taping of the drywall, which itself introduces high mositure content,, my builder and I have taken relative humditiy readings and found high readings of 55 - 60 % in this unfinished home. This home is heated by two minisplits and will be ventilated by and ERV for which I am trying to also obtain an HRV conversion core.. My windows are fiberglass triple pane with extensive glass patio sliders and french doors onto a deck facing my west lake view.

    My worry is that the relative humidity of the home will stay at too high a level and penetrate the permeable open cell spray foam in the walls causing rotting of the wall sheathing? Although it is too late now, I wonder if I should have used closed cell spray foam on the interior side of the wall sheathing to prevent this rotting? Fortunately, the simple solution is to install a dehumidifier in the cathedral living room. This may be a solution to the potential rotting of the wall sheathing but I am disappointed that I did not get the wall insulation configuration correct from the beginning. I believed as expert recommendations did also, that my existing configuration would allow the wall sheathing to dry to the interior!!! I am interested in any comments. Thank You.

  39. GBA Editor
    Martin Holladay | | #39

    Response to Joseph Poland
    Joseph,
    In your climate zone (Climate Zone 5), the minimum R-value for rigid foam installed on the exterior of a 2x6 wall is R-7.5. (To learn why, read Calculating the Minimum Thickness of Rigid Foam Sheathing.)

    You have chosen to install 3/4-inch-thick polyiso, which has an R-value of R-4.5 or R-3.8 (depending on whether you want to de-rate the polyiso for its reduced performance in cold temperatures). That's not enough to keep your plywood sheathing above the dew point during the winter.

    So you made a design error. You can probably stay out of trouble, though, if you control your indoor relative humidity during the winter. Keep your house dry, and your walls will probably be OK.

    If it's not too late, you might consider doubling the thickness of your exterior rigid foam.

  40. wjrobinson | | #40

    Joseph
    Dehumidify your home starting now. Winter air is very dry, in the past we kept windows cracked during the taping painting phase. Today with very tight construction use a dehumidifier. Also do not heat with moisture creating portable heaters. My choice is to use a standard furnace either a temporary one or the one to be installed in the home. Just don't run air through the ducts when sanding and creating dust.

    Your walls should be fine. I have two homes with open cell only that are fine. They have ventilation via always on with bathroom lights bath fans (Panasonic) and they have air conditioning and they do not have humidifiers or large sources of moisture creating sources like 50 house plants for example.

    All the past info on GBA says that walls have less problems than cathedral ceilings.

    So, Joseph, dehumidify to get your new building moisture out now and maybe for the next year or two keep an eye on it, Run the bath fans 24/7 right now.... get your numbers down toward 30% somewhere far away from 60%.

    New home moisture takes time to remove.... be patient, it will lower.

    Edit; Martin's advice is the prudent advice. But, my experience has been with open cell and no exterior foam and all is perfectly fine. On the other hand I have been involved with a few rigid foam installations that were so rotted that all wood framing had to be replaced where water accumulated in materials adjacent to the rigid foam. Granted the homes had water intrusion but the rigid foam (foil faced) made those area (like closed cell foam) vapor barriers and so once wet they didn't dry and they them rotted to point of becoming potting soil. I should post the pics someday.

    The one thing I know for where I live is a highly insulated cathedral ceiling using rigid foam needs to be vented above the rigid.

  41. JoeUSAF | | #41

    Thank You Martin,EPA Energy
    Thank You Martin,

    EPA Energy Star Version 3 (Rev.06) Thermal Enclosure Rater Checklist states in "item 4.4: Reduced thermal bridging at above-grade walls separating conditioned from unconditioned space, (rim / band joists exmpted) using one of the following options: 4.4.1 Continuous rigid insulation, insulated siding, or combination of the two; >= R-3 in Climate Zones 1 to 4, >= R-5 in Climate Zones 5 to 8." The manufacturer of my foil face polyiso rates it at R-5 at 3/4".

    Although manufacturers claim that open cell spray foam becomes air impermeable at 3-4 inches, is this an industry proven fact? If this is the case, then is it true that mositure will never come in contact with my wall sheathing despite the relative humidity of the interior rooms? Thank You.

  42. wjrobinson | | #42

    One word; Buoyancy
    "Buoyancy"... like in 1967... "plastic"

    Thank you again Joe L. for another insight worthy of unlimited gratitude. I will pass it on. Went to one of your seminars way back when at West Point sponsored by NYS. If any readers get a chance make sure you get to see Joe speak. The only problem I had Joe is you are way too entertaining to remember any of the information instead over the humor!

    Buoyancy

    I love learning about building science.... green is not the reason I still come here... building science is.

    We unfortunately lost one of the best advocates of green at this site. Wish he enjoyed his green world only and not smacking around anyone who did not worship at his feet. I miss his very insightful semi natural green advocating thoughts. Many here do.

    Simon and Martin... Joe ended his post.... respectfully yours... I wish the tone of both of your posts did too especially you Simon. You did not need to exclamation bomb Joe in your post. You and I have to check our egos sometimes or maybe with me all the time. Discuss with Joe, I love it... but be nice... Joe is almost a God... to me anyways... I have to meet you sometime and find out if you too are worthy of God status... not yet Simon. Point me to your seminars... I'll see if I can raise your status for me. Take this well...get a laugh from it... life is best with several chuckles per hour... have a few on me.
    ;)

  43. synergytodd | | #43

    Open Cell is Zone 3 is Imperative
    We are located in Alabama and we spray Icynene open cell as well as closed cell foam. We size and design the HVAC systems where we install the spray foam. I completely disagree with your article. I think there is no question as to open cell foam being the superior product in Zone 3 where we are located. There are multiple reasons that I don't have the time or care to list. However, I would like to mention that when you write misguided and misinformed articles like this you create a disservice to the entire building industry. I will spend hours upon hours explaining to customers that happen to Goggle your article why it is completely wrong.

  44. GBA Editor
    Martin Holladay | | #44

    Response to Joseph Poland
    Joseph,
    The building code (and Energy Star checklists) provide minimum guidelines. They do not state best practices. Notice that the Energy Star checklists says that the rigid foam will be >= the stated values (in other words, greater than or equal to). It's your choice. (The correct answer is "greater than," not "equal to.")

    The polyiso manufacturer labels their product as R-5. That's true during the summer, but not true during the winter. For more information, see In Cold Climates, R-5 Foam Beats R-6.

    Even if the polyiso were a true R-5, it still wouldn't have a high enough R-value to keep your wall sheathing above the dew point during the winter.

    Q. "Although manufacturers claim that open cell spray foam becomes air impermeable at 3-4 inches, is this an industry proven fact?"

    A. Yes, it is a fact.

    Q. "If this is the case, then is it true that moisture will never come in contact with my wall sheathing despite the relative humidity of the interior rooms?"

    A. There are two mechanisms for vapor transport from your interior environment to your wall sheathing: exfiltration (escaping air) and vapor diffusion. Your open-cell foam stops the first mechanism but not the second (because it is vapor permeable).

    The fact that you won't have any air movement is a fact in your favor, which is one reason why I think your wall will probably be OK if you can keep your interior RH at a reasonable level. However, you will still have vapor diffusion through your spray foam, and the moisture will still accumulate in your sheathing. Unfortunately, the 3/4-inch polyiso reduces the ability of the sheathing to dry to the exterior without being thick enough to keep it warm, so the polyiso is working against you.

  45. GBA Editor
    Martin Holladay | | #45

    Response to Todd Witt
    Todd,
    You wrote, "I completely disagree with your article."

    This article is a report on two academic papers that were presented at a conference in Florida. Many GBA readers are interested in the latest research, and we do our best to report it here.

    The conclusions of the two research teams, which I believe was accurately reported here, are open to debate, as are the conclusions of any scientist. Joe Lstiburek has raised a few objections to the researchers' conclusions; and one of the authors of one of the papers (Simon Pallin) has responded to Lstiburek in a rebuttal.

    I'm not sure whether you disagree with the fact that I reported on these academic papers, or whether (like Lstiburek) you want to dispute the conclusions of the researchers. If you think that the researchers have made errors, you need to point them out.

  46. JoeUSAF | | #46

    Martin, " Thank You Again"
    Martin, " Thank You Again" for a very detailed, scientific, and understandable explaination. Unfortunately, part of the reason we did not use thicker foil faced polyiso on the exterior of the sheathing is the large additional labor involved with building out the window fames to accommodate the larger thickness of insulation.

    Would an internal polyethylene vapor barrier installed just before putting up the drywall have solved my potential vapor diffusion problem? Also would using an impermeable paint to cover my drywall solve this same potential problem?

  47. GBA Editor
    Martin Holladay | | #47

    Response to Joseph Poland
    Joseph,
    Either MemBrain (a smart vapor retarder) or vapor-retarder paint would probably be a good idea. If your drywall is already up, it's probably too late for the MemBrain. But you might want to use the vapor-retarder paint as a primer.

  48. wjrobinson | | #48

    Overkill
    Martin, and

    Overkill

    Martin, and Joseph... good is good enough... more than good is not gooder at least with all the open cell walls sprayed in my neck of the woods.

    It's like one bolt in a wing strut doing the job... but two bolts weakening the whole assembly. I actually have dealt with this situation with a customer at one time.

    aj

  49. user-958947 | | #49

    Question to Joe L's post #27---Removal of ridge insulation
    Joe, I was with you until you got to the part about solving the Houston house problem by eliminating attic insulation along the ridges. Are you speaking of "bare" ridges---like looking at the bottom of exposed sheathing? I can see how this would dry due to total exposure, but are there other issues to worry about with a bare deck?
    I'm a big fan, so please take this in the spirit in which it's intended---educate me.
    I have a home in zone 2B with OC foam, conditioned attic. If I take RH readings along the attic high points and the RH is too high, then you're suggesting I need to add a supply and return duct or a dehumidifier. At what point do I start thinking about ripping insulation off the ridges?

    [Editor's note: To read a response to this comment, as well as other comments, click the "2" box below to proceed to page 2.]

  50. GBA Editor
    Martin Holladay | | #50

    Response to John Walls
    John,
    Like you, I look forward to Joe Lstiburek's response. I seem to recall that Joe told me, several years ago, that they were experimenting with a vapor-permeable ridge cap to address the issue of damp roof sheathing in unvented, sealed attics. However, I can't find a reference to that remedy in any published documents.

    One clue to Lstiburek's thinking can be found in a 2005 paper by Armin Rudd (a Building Science Corp. researcher). The cases under discussion were experimental attics insulated along the roof line with fiberglass batts and cellulose, not open-cell spray foam, but the discussion is relevant to Lstiburek's point.

    Rudd wrote, "There appears to be migration of moisture up to the highest points in the attic due to moisture buoyancy and air movement due to thermal buoyancy. Where hip or valley rafters, ridge boards, or roof peaks exist, the moisture tended to concentrate there as evidenced by some observations of elevated wood moisture and rusted fasteners. However, the measured data showed elevated dew points with increasing height only during daytime. This indicates that moisture removal by the space conditioning system was sufficient to equalize humidity conditions to that of the actively conditioned living space each night. While the space conditioning system (cooling plus dehumidification) can remove this moisture, it is prudent to eliminate the moisture load by installing a vapor retarding roof underlayment beneath the composition shingles. Such roof underlayment should be used instead of traditional #15 lb felt roofing paper, which has a water vapor transmission of about 6 imperial perm."

    Rudd's recommendation that builders should specify vapor-impermeable roofing underlayment for this type of attic was based on the (possibly mistaken) assumption that inward solar vapor drive through asphalt shingles was responsible for some of the elevated moisture levels in roof sheathing above these sealed attics. Since Rudd's paper was published, Lstiburek has backed away from his earlier recommendations; his current position is that the source of this moisture is the interior air, not inward solar vapor drive through asphalt shingles.

  51. Allan27 | | #51

    Solar Vapor Drive
    Martin:

    In a hot humid climate, say after a rain with 98% humidity, temperature is high 90's, sun beating down on a comp roof, do you think vapor can be driven through shingles into an attic? Seems like there should be some testing somewhere to determine this.

  52. GBA Editor
    Martin Holladay | | #52

    Response to Allan Edwards
    Allan,
    I guess it's time to dust off an old article I wrote over a decade ago. The article appeared in the January 2003 issue of Energy Design Update. I'll quote from the article:

    "... Armin Rudd and Joe Lstiburek, principals at the Building Science Corporation, first noticed the phenomenon of solar vapor drive through asphalt shingles in Houston, where they were monitoring twenty houses as part of the Building America program. The houses all had unvented cathedralized attics, with insulation following the roof slope.

    "“In May of 2001 I noticed that we had high dew points above the insulation and below the roof sheathing,” said Rudd. “The readings were well above those outside -- as much as 100 degrees dew point. The relative humidity was still low enough not to cause us stress about mold, but the dew points were high during the heat of the day. At first I thought that humid air was migrating up toward the peak of the roof from thermal buoyancy, because there was a gradient in the readings, with higher dew points at the peak and less at the eaves. So we did some more air-sealing work, but we got the same readings after the air sealing. The moisture was in every rafter bay -- it seemed to be wherever you checked it -- so I decided it probably wasn’t air leakage. The other thing that clued me in was that on days of cloudiness or rain, the dew point stayed normal, but it went way up every day there was sun. I began to suspect it might be solar vapor drive through the shingles, but at first we wrote it off, because, as we said, ‘The shingles are asphalt, so they’re impermeable.’”

    "Because of night-sky radiational cooling, roofs are typically wet with dew at dawn, and some of the dew is drawn between the laps of the shingles. When the sun heats the roof -- often to 150°F or higher -- the moisture evaporates and is driven into the asphalt felt and roof sheathing. “The phenomenon of solar vapor drive through asphalt shingles exists,” says Lstiburek. “It’s there. We should have known about it all along, and we would have if any of us had bothered to look at a psychrometric chart -- duh.”

    "Whether or not the moisture driven into a roof assembly by the sun will cause any problems depends upon the roof construction details. According to Rudd, the unvented roofs in Houston were designed to exclude exterior moisture, so the solar-driven water vapor amounted to a “flaw.” Lstiburek does not anticipate any performance problems from solar-driven moisture into cathedralized attics, where the moisture can dissipate into the attic space, but he feels it could be a problem in cathedral ceilings with painted gypsum board as the interior finish. “The phenomenon matters a great deal with unvented cathedral-ceiling roofs,” says Lstiburek. “In a cathedral ceiling, the moisture build-up could be cumulative. And we’re finding that this will happen as far north as Chicago.”

    "To prevent solar vapor drive into roof assemblies, Lstiburek and Rudd advocate installing a vapor barrier between the asphalt shingles and the roof sheathing. Ordinary polyethylene, which is degraded by sunlight and dangerously slippery, should never be installed on a roof. Lstiburek recommends the use of a vapor-barrier roof underlayment called Tri-Flex 30 (manufactured by Flexia Corp., Brantford, Ontario, Canada).

    "Lstiburek says that switching from asphalt felt to Tri-Flex 30 not only solved the Houston inward vapor-drive problem, it also solved the problem of asphalt shingle buckling. Although Danny Parker, a senior researcher at the Florida Solar Energy Center, concluded that the mysterious ridges in asphalt shingles installed over insulated roofs were caused by plywood sheathing exposed to high temperatures (see EDU, May 2001), Lstiburek and Rudd are convinced that solar-driven moisture, not high temperature, causes the ridges. Although the exact mechanism is still uncertain, the ridging appears to be caused by either swollen asphalt felt or the swollen edges of the roof sheathing panels. Whatever it is that pushes up the shingles, the installation of a vapor barrier over the sheathing appears to solve the problem.

    "The phenomenon of solar vapor drive through asphalt shingles was recently confirmed in a laboratory experiment conducted by Bruce Davis, senior building science specialist at Advanced Energy in Raleigh, North Carolina. Davis performed the tests using a 4 foot by 12 foot mock-up of a shed roof. Installed indoors, the 12-in-12 pitch roof was built with 1/2-inch OSB sheathing, #15 asphalt felt, and asphalt shingles. “We placed an unfaced fiberglass batt in one rafter bay, blocked off the bay at the top and bottom with one-inch foam board, and covered the interior side of the cavity with a continuous roll of 24-inch duct mask [polyethylene],” wrote Davis. The air-conditioned lab was kept at typical indoor conditions, and conditions in the rafter bay were monitored with temperature and relative humidity sensors.

    "“The engineers here at Advanced Energy built a 4 foot by 12 foot infrared emitter array and placed it along the roof with around 12 to 16 inches of space between the emitters and the roof,” Davis wrote. “A sensor in the shingles controlled the array such that the temperature of the roof deck was never heated above 180°F. At start-up, the roof cavity measured 70.2°F and 52.8% relative humidity, or approximately 58 grains of moisture.” After spraying the roof with moisture, turning on the “sun,” and spraying on additional moisture, the moisture level in the rafter bay increased to 114 grains, at which point the “sun” was turned off. About half an hour later, after the temperature around the test assembly had cooled to normal room temperature, the researchers noticed condensation on the inside of the polyethylene covering the rafter bay.

    "Most builders will probably be surprised to learn that asphalt shingle roofing behaves more like Tyvek than poly. “Of course, this information does not reach a conclusion, but only opens up the need for more work to determine just what this would mean in real buildings in different climates,” says Davis."

    - - - -

    Ten years later, the clues presented in my EDU article represent, as far as I know, about as much as researchers know about this problem. There are still unanswered questions, and one still encounters researchers willing to either confirm or deny that the phenomenon of inward solar vapor drive through asphalt shingles exists. The matter remains unresolved.

  53. Allan27 | | #53

    Moisture in Attic
    Thanks Martin, I am going to re-read your article. I typically use I&WS on my roofs, maybe that is helping? However, if you are using spray foam at the rafters (I use 2x8 as standard and spray 7-1/2"), sealing the attic, have 15 or so tons of AC running in the living area dehumidifying the air, no air barrier or insulation between the attic and living area, seems moisture in the attic would not be a serious problem. What would be the source of moisture in the attic?

    As I mentioned we do add some conditioned air to the attic, I guess my AC contractor does it as a safety measure, I think I will discuss this with him.

  54. GBA Editor
    Martin Holladay | | #54

    Response to Allan Edwards
    Allan,
    It sounds like you are following Joe Lstiburek's advice. As long as the air in your attic is dry, I don't think that you have any reason to worry.

  55. MrPallin | | #55

    Response to Joe Lstiburek, Comment #27
    ###

    My comments were intended to add value to the discussion and clarify misunderstandings regarding the presentation and paper that was presented in Clearwater, FL. However, it is better that if further questions remain, please feel free to contact me directly.

    ###

  56. wjrobinson | | #56

    Huber Zip R sheathing solution possibiity
    More and more I see the value of Zip R sheathing in regard to building a safer moisture tolerant assembly for walls and ceilings.

    Another Idea... spray closed cell 2" thick at the ridge only to save costs. Closed cell and Zip R do the same thing, they isolate the OSB from interior moisture. The added benefit of Zip R is a layer of continuous insulation.

    Buoyancy

  57. GBA Editor
    Martin Holladay | | #57

    Response to A.J. Builder
    A.J.,
    This is the second time that you have suggested the use of Huber Zip sheathing as a solution to the problem of damp sheathing.

    You have more faith in the product than I do. I don't think we have enough evidence yet to say that Zip sheathing won't rot -- especially if it is installed in a damp location.

  58. Joe Lstiburek | | #58

    More information on inward solar vapor drive through shingles
    This solar driven moisture discussion is missing the issue at hand. Let me fill in some gaps on solar driven moisture so that the focus returns to the main issue. We thought solar driven moisture was a problem in the late 1990's with unvented roof assemblies. We were worried about it. When I wrote the Florida Building Code language for unvented attics I included a requirement for impermeable roofing membranes - just to be safe. The FBC was the first jurisdiction to adopt unvented roofs in 2001. We studied the issue more in the early 2000's. By mid 2006 our Houston work showed that it was not a problem which is why when I wrote the code language for unvented roofs that was adopted in the 2006 International Residential Code there was no requirement for impermeable roofing paper. Unfortunately we did not publish a peer reviewed paper in any journal about this conclusion. I just talked about it in all my presentations and thought with the code revision in place we were done with the issue. Similarly, for the issue of moisture accumulation bouncing up the roof to the ridge - for which I coined the phrase "ping pong" water. I didn't publish that either just talked about it. Didn't think much about it at the time. Talked to you about it if I recall.

    I wrote a paper for the ASHRAE Journal that was published in April, 2006, "Understanding Attic Ventilation" that talked about my concern, which turned out to be unfounded, about solar driven moisture. Wasn't sure about things when I wrote the paper. By mid year it was clear that it was not an issue. But the paper was out. I did not issue a correction. In retrospect I should have. The paper covered many things, the solar driven moisture piece, a couple of paragraphs, seemed incidental to the point of the paper. Besides, with the code change in place, I thought who cares? We posted it on our website as it originally ran never thinking anyone would care about solar driven moisture. A couple of months ago I re-edited it to take out the solar driven moisture piece when I re-read it. I though to myself, I thought I deleted that stuff years ago. Never did. I just thought that with me taking about it not being an problem all these years folks would know.

    The unpublished Houston work lead to some exciting (to me anyway) possibilities of using cellulose for passive solar dehumidification. If this moisture (from the interior) pings and pongs its way up to the ridge maybe we could put a ridge vent up there (vapor open but air closed - a vapor diffusion vent) and keep the house dry using solar radiation and vapor diffusion rather than by air conditioning or by dehumidification. So, because of this "possibility" I didn't write up all the Houston stuff, including our conclusions on solar driven moisture being a non-issue. I wanted to see how far we could go with this other stuff before writing about it and letting the cat out of the bag. Yes, pride can be a problem. We built some test houses and found that it (the "possibility") worked in concept but not in practice - not enough moisture was removed to make any difference to the interior moisture load - but it sure kept the roof sheathing dry. This information allowed us to recommend diffusion vents for SIP roof failures. All this is written up in Building America research reports. We didn't post them because it didn't work from a practical perspective. But we sure learned stuff. The solar driven moisture thing was more or less forgotten.

    But the other stuff wasn't forgotten. We are currently using the approach of diffusion venting and ping pong water in some test buildings in Chicago, Houston and Orlando to see if we can construct unvented roofs with cellulose and fiberglass rather than low density or high density spray foam. This approach also might solve the concerns with dense pack cellulose in retrofits in cold climates. Wouldn't it be neat if you could dense pack and just add a diffusion vent and be done with it? Anyway, we are currently working the problem. We will publish even if it does not work.

    So, back to the issue at hand. Solar driven moisture is not the source of the moisture that is being found in some open cell spray foam roof assemblies. The moisture source is interior moisture. Unvented attics need to be conditioned to remove moisture if hygroscopic insulations are used such as open cell foam, cellulose and to less an extent fiberglass. One of the best means of accomplishing this is via air change to the occupied part of the house. Incidental duct leakage and leaky ceiling construction is not reliable enough - even though this approach has been successful in the vast majority of situations. A supply duct and a return path works well. Is this an energy penalty? The moisture load no. It is interior moisture that needed to be dealt with regardless of the type of roof construction. But unvented roof assemblies that are conditioned have a thermal load that can be substantially greater that that compared to vented roofs and that is an energy penalty. They only make sense if there is no other practical means of achieving airtightness in a roof assembly or if there is no other practical means of dealing with ductwork in attics. Where unvented attics make sense low density spray foam works well when a means of moisture removal is provided - and where limits are placed on vapor diffusion in Climate Zones 5 or higher.

    -- Joseph Lstiburek

  59. Joe Lstiburek | | #59

    Response to Mr. Walls (Comment #50)
    Mr. Walls,
    In Houston we removed about 6 inches of cellulose on both sides of the ridge beam and the upper 1/3 of hip rafters exposing the underside of the sheathing. The moisture that ended up there via pinging and ponging evaporated (yes another pong) and was removed via air change into the occupied part of the house below. Not all was removed at once, with each pong some of the moisture pinged back to the sheathing. Think of it as a magic moisture dance where a chair and a dancing couple is removed or kicked off the island at each cycle. I think I maybe mixed a few metaphor there. We could have also installed a diffusion vent along the ridge to accomplish the same thing. The problem was the hips. Hadn't got that one figured out at the time diffusion vent wise. We do now but I will save the detail for one of my upcoming ASHRAE Journal Columns. Thinking of calling it Mr. Ping Meets Mr. Pong....

    Now, to your house in particular. The ping pong thing does not work real well with OC foam. It is not as vapor open as cellulose and a great deal more airtight. Ripping away foam at the ridge will not help you because the moisture is likely to be diffused over a much larger surface area and not concentrated at the ridge unlike with cellulose and fiberglass. The best bet is to add a supply and return air duct (and an air cycler to make sure you get some air change every couple of hours). You will not need a dehumidifier and you will not need to remove your foam.

    Best regards,
    Joseph Lstiburek

  60. Allan27 | | #60

    Moisture in Attics
    If the issue is moisture in attics, why is the best solution to add supply and return rather than a dehumidifier. It seems you might be adding precious conditioned air during low humidity periods and the conditioned air being added is wasted.

    Allan

  61. GBA Editor
    Martin Holladay | | #61

    Response to Allan Edwards
    Allan,
    You don't want to install a dehumidifier in your attic for two reasons:

    1. Having an extra device (a dehumidifier) in your attic means that the homeowner has to know it's there, check on it now and then to be sure it is working, and fix it when it is broken. Eventually it will break, but in most cases the homeowner will never know.

    2. The energy penalty that accrues from conditioning your attic is real but small. After all, the conditioned air that is released into the attic is within the home's thermal boundary. The air is indoors. It's not like you are blowing conditioned air into a vented attic.

  62. Allan27 | | #62

    Thanks
    Thanks Martin, good points. Also when you include the actual cost of the dehumidifier, added electrical circuit, installation, and cost to operate, I think you are correct. The added supply and return involves no moving parts to worry about.

    I guess my next question is how much added air is needed, what humidity levels are acceptable, perhaps these are calculations I need to do with my HVAC contractor.

    Allan

  63. Allan27 | | #63

    City Inspectors
    By the way, as I mentioned my HVAC sub has been adding conditioned air to our attics. I discussed this with him yesterday and he mentioned he recently had a job where a City of Houston inspector disallowed this because he was concerned about "gases" in the attic being pulled into the system. Just another battle for contractors to deal with.

    Allan

  64. wjrobinson | | #64

    Zip R-sheathing, not Zip sheathing Martin
    "ZIP System R-Sheathing is a five-in-one insulated panel designed to provide bulk water, thermal, air, and moisture resistance while delivering strength and durability. Installed with ZIP System tape to create a seamless airtight seal, ZIP System R-Sheathing's insulated panel provides greater energy efficiency and reduced heating and cooling costs."

    This product encapsulates the OSB sheathing so NO MOISTURE gets to the OSB.

    It HAS to be MUCH better than using standard crappy OSB.

    Martin, you resist this idea like you have purchase put stock betting against Huber which by the way you can't due since they are a privately held company.

    My opinion is that plywood has worked well for me, OSB... super cheap to buy and you get what you pay for... crap that can't handle wetting, and then there is Zip R-sheathing... protected on BOTH sides from moisture. BAM. What more could I ask for.

    Martin, I don't appreciate your doubt dismissive attitude toward ZIP R-sheathing. I will continue to push this idea forward till someone can show me one failure of this product from moisture. One.

    And as to failures, the spray foam company I use near me has sprayed hundreds if not thousands of homes with open cell. Last I asked Doug, my contact, about problems they have had one problem with one closed cell install that was minor.

    This whole thread is interesting but in my area I have not heard of or visited one failure of open cell moisture damage.

    It would be helpful to have a list of failures in my area to personally look into. I do not know of one. I do know of a few rigid foam failures and have worked myself on the repairs.

    Putting rigid foam on one side of OSB is much less safe than a product like Zip R-sheathing that has bonded materials on both sides of the OSB to keep water off the OSB.

    The rigid foam on OSB failures make total sense. The rigid is NOT bonded to the OSB and the other side ot the OSB is not protected from water either. It just about HAS to fail Martin.

    Joe L... I wish you would look into Zip R-sheathing and what I am posting. I am throwing my logic at this. You could throw in your decades of building science logic. Thank you in advance if you do.

    aj

    PS photos,,, Choose cheap get crappy rotten OSB. Choose quality... get fully encapsulated moisture safe Zip R-sheathing. Show me one sheet of Zip R-sheathing that looks like the crappy OSB GBA posted with this blog. One.

  65. wjrobinson | | #65

    Zip R-sheathing
    Zip R-sheathing, how can you not like it.

    Rotted OSB, never again.

  66. kwhe642 | | #66

    Cathedral Ceilings
    Was wondering if someone could expand about this effect on cathedral ceilings, finished with sheet rock and/or T&G, that have open cell applied. Is it the same, similar, or less?

  67. GBA Editor
    Martin Holladay | | #67

    Response to Kyle Wheeler
    Kyle,
    It's hard to generalize, because there are so many variables -- as well as a few unknowns that await further research.

    In general, the usual advice for builders in Climate Zone 5 or anywhere colder is that a cathedral ceiling with open-cell spray foam requires a vapor retarder on the interior side. The most dependable way to provide this is with a layer of gypsum drywall painted with vapor-retarder paint.

    If the ceiling consists of tongue-and-groove boards without an air barrier, then there will be more moisture flow from the interior through the open-cell foam than would be the case with gypsum drywall covered with vapor-retarder paint -- so I would definitely call the tongue-and-groove option riskier.

  68. RZR | | #68

    Just got done reading all the

    Just got done reading all the responses to the noted articles above. All I got to say what massive confusion, some contradictions, some lessons learned. I see one guy on the same page I am with improved material designs such as Advantech that take the complex designs by builders and contractor’s following gurus advice that can change on a whim, down to a simple working level reducing or eliminating the need for write-ups like this or complex remedies to band aide materials that should have never been placed in the environment there are in obviously. Materials like Advantech that consolidate all these products and practices makes perfect sense. Martin’s question concerning quality control and field data of such materials is a valid point. There has to be an industry means per spec/code control to validate quality and consistency in production or manufacturing (eg: OSB, foam, insulation, etc). The OEMs indi test is a start. I have been involved in supplier qual testing, we go out look at and approve the production run based on a number of units, if there is a change in approved process the manufacturer has to obtain approval or make the client, AHJ aware.

    Builders and contractors are the last ones you want to depend on to ‘use their brain’ or make material assembly design choices as Martin suggest. I know of Engineers that would struggle with the write ups. The ones that wrote them admit they don’t have enough data to make industry wide conclusions (eg: solar radiant transfusion, accurate WUFI models, etc). Designs that reduce the need for down stream user knowledge other than follow the OEMs installation procedures vs designing complex assemblies that interact well is risky not the material (eg: OC foam), dummy proof sort of speak. If the builder wants an assembly give them a factory designed, controlled, and qualified sandwich or composite construction design with simple installation procedures

  69. mmclarke | | #69

    Insulation recommendations
    Martin,

    Thank you for a very informative (and timely) article! We are working with an architect on an addition in Iowa (basement and first floor). She originally recommended open-cell foam insulation for entire addition after I told her we wanted closed-cell spray foam. Her firm does not recommend use of closed-cell insulation on the roof because it is too impermeable and if there is a leak, the rafters will absorb the water and rot before the leak is noticed. I forwarded your article and she is now recommending open-cell on walls and asking what we want to use on the roof (given the data you presented). I am inclined to request closed-cell spray foam insulation throughout the addition (including the roof), but haven't found any reference to closed-cell in this discussion. Do you have a preferred insulation product for walls and roof in Iowa? I appreciate hearing your thoughts on this! Thank you.

  70. user-962101 | | #70

    Metal roof instead of shingles
    Not trying restart this discussion, but if I'm planning new construction, 6/12 roof, conditioned attic, what is the best arrangement and do I still have the concern with the moisture (OC , CC foam) accumulation on the foam . I live in South Mississippi, 70 miles from the Gulf Coast and am interested in a conditoned attic and double stud wall dense pack cellulose.

  71. GBA Editor
    Martin Holladay | | #71

    Response to A.K. Harrison
    A.K.,
    Although many builders in your climate zone have successfully used open-cell spray foam on the underside of the roof sheathing, the evidence presented in this article shows that there is reason to believe that closed-cell spray foam is less risky.

    The evidence is not definitive. It's your choice.

  72. rmrswims | | #72

    Icynene on roof deck
    I am in central Indiana and had my roof deck sprayed with Icynene in the fall. While our heating bills actually went down in spite of a cold winter so far, I keep thinking "why heat the attic?" and was thinking of adding rolls of fiberglass insulation. This article and comments make me think I shouldn't because I need to encourage air flow from the house to the attic - am I understanding correctly? I think the Icynene is open cell. How can I keep tabs on whether or not my roof deck is rotting? Do I cut out some Icynene each year to check on the roof deck and re-spray with a can of foam? Thanks for your answers.

  73. GBA Editor
    Martin Holladay | | #73

    Response to Rachel Stutsman
    Rachel,
    If you are in Climate Zone 5 -- most of central and northern Indiana is located in Climate Zone 5 -- then you need to install a vapor retarder on the interior side of your cured spray foam to prevent interior moisture from diffusing through the foam and accumulating on your cold roof sheathing.

    While building scientists used to recommend installing vapor-retarder paint on the interior side of the cured foam, further research has shown that vapor-retarder paint won't work in that application.

    To protect your roof sheathing from moisture, you need to install a layer of gypsum drywall on the interior side of your cured foam, and then you need to install at least one layer of vapor-retarder paint on the gypsum wallboard.

    For more information, see Joe Lstiburek Discusses Basement Insulation and Vapor Retarders.

  74. SkyT | | #74

    Perfect Real World Example
    I have a house located in RI which we built in 2008 using open cell foam throughout, exterior wall, ceiling, etc. We were in the process of finishing a small area of our attic as a playroom for our daughter when we found some voids in the foam on the roof sheathing while cutting it back for sheetrock. After further examination we found that the voids were quite large in areas where we were only seeing a small hole, think of an iceberg, the majority of it is beneath the water. The roof sheathing was completely wet and had mold forming on the wood. We immediately contacted the foam contractor and they began sampling several areas in the open attic. We found in many instances the foam was not adhered to the roof which meant there could be as little as 2” of foam. The problem is you cannot see this until you cut into the foam. We also sampled a gable wall in the attic and found a separation of roughly 2” however there was no condensation or mold.

    The house was supposed to get 5.5” of foam in the walls and 7.5” in the ceiling. The walls are constructed of 2x6 while the roof is all scissor trusses and a few areas of 2x10 joists. Much of the roof is cathedral as well. The roof has asphalt singles over felt underlayment with ice & water along the edges. The sheathing on the roof is Advantech while the walls are regular OSB. The house is wrapped with the typical house wrap. We do not have an HRV installed. The heat is a 3 zone hydro air system with hydronic radiant floors in the bathrooms and living room. One air handler is in the basement while the other two are in the attic space. We also have central air connected to all three air handlers all with their own AC unit outside. Our house is kept at around 45% humidity during the winter.

    We had a blower door test done and they said the house is very tight, which I would expect. They noticed a cold draft coming in through the attic access panel while running the test. We also saw some cold areas near the soffits which brought us outside the house. There were some icicles which formed in certain areas along the soffit. I always assumed it was just the normal snow hitting the gutters and thought nothing of it. After using the IR camera on one area we noticed a bit of temperature difference along the bay where the ductwork for a vent ran through the ceiling in the master bedroom. On the outside lined up perfectly with the same bay were quite a few icicles so I took off the vinyl soffit to find the entire soffit filled with mold.

    My best guess is the separation of the foam from the roof is allowing cold air in through the soffits where it is forming condensation along the roof sheathing which is then running back down the roof sheathing into the soffit. Along with the voids and separation the roof is so poorly insulated from the temperature differences that we have an extreme buildup of condensation on the roof. I really hope someone can shed some light on what is happening because there is a lot of finger pointing going on but no one is taking responsibility for the problem.

  75. GBA Editor
    Martin Holladay | | #75

    Response to Sky T
    Sky,
    If the spray foam didn't adhere to the substrate, and if you have opened up the foam and found voids, and if some of the foam areas are thin, it doesn't sound like there is a mystery here. It sounds like the spray foam installer did a substandard job and should be held responsible.

    The spray foam installer should have installed an adequate amount of foam, without voids, adhered to the substrate. Moreover, to conform to code, the foam should have been installed to an adequate thickness to meet (at least) minimum code requirements for R-value, and, if required by your local inspector, should have been protected by an ignition barrier for fire safety. Moreover, in your climate zone, a vapor retarder should have been installed on the interior side of the cured foam.

    If the foam is thin, there is no mystery as to why your roof sheathing is wet and moldy. Interior moisture is diffusing through the thin foam and is accumulating in the cold roof sheathing.

  76. ablondguy | | #76

    Low Pitched Roof Assembly In Regards To Latest Research
    Martin,
    "Adding more fuel to the fire, with conflicting research and opinions of the research, similar to theories on what causes cancer that have gone back and forth", simply put, it seems that if you are going to have a very tight structure, with very tight duct work, without coupling, conditioning or dehumidifying the air in the conditioned attic, which seems counter productive from an energy stand point, closed cell spray foam would be the recommended, safest choice for the underside of the roof decking. And, if so, I want to absolutely make sure that this would be the case for a low pitched, 1/4" per foot sloped roof, with Advantech Sheathing (while this is an OSB product, we think it would be superior to not only regular OSB but plywood as well for moisture concerns), Prosoco Cat 5 w/R Guard Joint Seam or similar WRB, polysio board (varying thicknesses to creates slopes), Densdeck Duraguard and a PVC membrane roof, in Austin, TX, which almost always, - currently - seems to go with open cell. Also, "to stir the coals more", how would 2" of closed cell followed by however many inches of open cell fair in this discussion as a combo approach? Please advise on all.

  77. GBA Editor
    Martin Holladay | | #77

    Response to Norman Luke
    Norman,
    Prosoco Cat 5, to the best of my knowledge, is a liquid-applied water-resistive barrier (WRB) that is usually specified for walls, not roofs.

    To stop water penetration on a low-slope roof, you should depend on the roofing, not on a WRB. If you want improved airtightness in your roof assembly, you would normally either tape the sheathing seams, or depend on the closed-cell spray foam that you mention in your comment. (Remember, as well, that most types of low-slope roofing, including membranes like EPDM, are already airtight.)

    If you are installing polyisocyanurate of adequate thickness above your roof sheathing, there is no need to install any closed-cell spray foam on the underside of your roof sheathing. Most experts would advise against this type of "foam sandwich" -- one that seals the roof sheathing from both sides.

    If you want to use spray foam on the underside of your roof sheathing, you can skip the polyiso. As noted in this article, closed-cell spray foam is less risky than open-cell spray foam in this location.

    To read about all of the different approaches to insulating a low-slope roof, I suggest that your read Insulating Low-Slope Residential Roofs.

  78. benrush | | #78

    "•Adding ventilation channels above the roof sheathing"?
    Mr. Holladay,
    In your recent response to my question at Mr. Bailes's blog about a spray-foamed attic with 2 furnaces and open gable vents, you referred me to your post about Creating a Conditioned Attic- where I followed another link to this post...
    You wrote "•Adding ventilation channels above the roof sheathing" and "•To further limit your risk, consider installing ventilation channels above your roof sheathing." I don't understand. Do you mean ventilation channels BELOW the sheathing? How would one create a ventilation channel ABOVE the sheathing?
    Also- is there a way for me to get an email notification when you reply? -or do I need to keep checking back- to look for new comments?
    Thank you for the really terrific info.
    Ben

  79. GBA Editor
    Martin Holladay | | #79

    Response to Ben Rush
    Ben,
    Q. "How would one create a ventilation channel ABOVE the sheathing?"

    A. The answer to your question can be found in this article: How to Build an Insulated Cathedral Ceiling. In that article, I wrote:

    "If you prefer, you can locate your ventilation channels on top of the roof sheathing rather than under the roof sheathing. You can create 1 1/2-inch-high ventilation channels above the roof sheathing with 2x4s installed on the flat, with the 2x4s located above the rafters, 16 inches or 24 inches on center. Although this approach is less fussy than installing vent baffles underneath the sheathing, it usually costs more, because most types of roofing require a second layer of plywood or OSB on top of the vent channels."

  80. benrush | | #80

    Got it. Crystal clear.
    Got it. Crystal clear. Thanks!
    Ben

  81. Richard Beyer | | #81

    Dr. Joe Theory
    Ping - Pong! I read Dr. Joe's theory. I have a new one for him to tackle. If his theory holds substance, how is he going to deal with moisture in an attic with this new spray foam marketing?

    https://youtu.be/ByMU0_tpnAc

    Create a "sealed" oxygen starved attic in the event of a fire? Certainly you can not acclimate the attic's rh to the living space without penetrating the ceiling with duct work. I only see one way to pull this off. Someone needs to install a duct door assembly on a sensor which closes the ducts in the event of a fire and dramatic temperature changes. ( Don't forget to share the royalties with me for this idea!)

  82. Tedkidd | | #82

    Stop playing thermostat ping pong?
    So might this be a recommendation for those with open cell:

    "Aggressively downsize equipment, design for continuous operation, and abandon setback so dew point is more continuously managed?"

    Who feels the suggestion to continuously maintain steady latent and sensible temperatures might help people avoid these moisture management tipping point issues? Also solves peak load problems, reduces overall consumption, (http://bit.ly/rickchitwood) and makes IAQ quality a lot easier to manage.

    But then again conventional wisdom claims 500 sf per ton and setback strategy "saves lots"...

  83. Dean Brenneman | | #83

    Florida Question
    I live in Florida where the Florida Building Code requires a self-adhering roof underlayment. As far as I know, these are all vapor impermeable. And I would like to foam-insulate my attic. I am concerned that Closed Cell foam will create a problem with vapor barriers on both sides of the plywood sheathing; even though the plywood sheathing is nominally dry, there must be some residual moisture, right? In this case, would I be better-off using Open Cell foam, thereby allowing the plywood to dry to one side?
    I should note that my AC unit and ducts are in the attic and are very leaky, so that should help. I could also open a supply and return if needed (but how will I know it is needed?).

    My roofer doesn't want me to use foam at all, but retain a vented attic instead. That doesn't make sense to me from a hurricane/wind-driven rain perspective.

    Can anyone advise me on these issues?

  84. GBA Editor
    Martin Holladay | | #84

    Response to Dean Brenneman
    Dean,
    Assuming that you are correct about the requirements of the Florida Building Code, you have several options.

    Open-cell spray foam will certainly work, as long as you follow the advice in my article. As I noted, "If you choose to install open-cell spray foam against the underside of roof sheathing in a humid climate, Joe Lstiburek recommends that the HVAC system be designed to condition the attic air and lower humidity levels in the attic." That means that you should probably include a supply air register and a return air grille in your attic.

    If you prefer to use closed-cell spray foam, nothing is preventing you from including a ventilated air space between the underside of your roof sheathing and the top side of your insulation. This can be accomplished by installing sturdy site-built ventilation baffles in each rafter bay (made, for example, by installing 1"x1" sticks and thin plywood) before the closed-cell spray foam is installed.

    For more information on your options, see How to Build an Insulated Cathedral Ceiling.

  85. Dean Brenneman | | #85

    Florida Question
    Thank you Martin - very helpful answer!

  86. ufgatorpa | | #86

    North Florida - Best Practice Recommendation
    Mr. Holliday and Mr. Lstiburek (and others):
    I have read for the past two months all that I can find from you (both) on building in coastal, hot and humid climates areas (like Jacksonville, FL). We are pre- new construction, and have contemplated open cell foam in the attic and walls of the entire house. I can not find any spray foam installers in my area that recommend any closed cell in my application. Given that there are no hard facts on this issue at this time, where would a lay person like myself turn to ensure that I am going to end up with the current best practices for insulation and roofing underlayment for my area? The majority of new homes here (Northeast FL) are 2x6 frame with stucco, and asphalt shingles. My chosen builder does use Zip systems for the walls and roof deck. I'm happy to pay for a qualified consultant or purchase any current publications you feel best guide someone in my position not wanting to have a big mistake. As you can imagine, this information is very technical for the lay person, and trying to ensure that I get the best I can at this time seems elusive. Thanks in advance for your help!

  87. GBA Editor
    Martin Holladay | | #87

    Response to Rob Brown
    Rob,
    You should choose your insulation contractor the same way you choose any other contractor. Ask for references, and call the references. As for proof of liability insurance. As for evidence that the contractor has been trained and certified to install spray foam insulation. (Such training is usually provided by spray foam manufacturers.)

    You can also ask your G.C. (if you have one) or builders you trust.

    If you decide to create an unvented conditioned attic using open-cell spray foam, you should follow the recommendations in this article. In other words, the attic should be conditioned by including a supply register and a return air grille connected to your forced-air heating and cooling system.

  88. ufgatorpa | | #88

    North Florida - Best Practice Recommendation
    Mr. Holliday: Thank you for your reply. Assuming we do go with the unvented conditioned attic with OCSF and run supply and return, what is the best practice for roofing underlayment material type when using asphalt shingles with Zip System roof decking? I've read some of the threads on this, and wanted to be clear on the proper material to be used, since it sounds like the synthetic peel and stick may worsen the moisture problem if it develops. Would a 30# traditional felt be advisable? Thanks again!

  89. GBA Editor
    Martin Holladay | | #89

    Response to Roy Goodwin
    Rob,
    Most brands of synthetic roofing underlayment are vapor-impermeable. Manufacturers of these products all agree that they cannot be installed over unvented roof assemblies.

    If you plan to construct an unvented roof assembly, use a traditional vapor-permeable roofing underlayment (asphalt felt).

    For more information on this issue, see Synthetic Roofing Underlayments.

    1. mhunt11 | | #108

      What if you have metal panels on a 1:12 slope- an Unvented assembly with closed cell spray foam on interior (sprayed on dry sheathing). Would you need a vapor permeable ice and water shield or does it not matter since the metal is vapor impermeable anyway? Is it fine to use a vapor impermeable underlayment even though the roof would not breath at all?

      1. swooley | | #111

        I have this exact question as well. Did you ever find an answer?

        1. MartinHolladay | | #112

          M. Hunt and Swooley,
          If you choose to install vapor-impermeable roofing underlayment on an unvented roof assembly, you are choosing to install a product contrary to the manufacturer's installation instructions. That's a code violation, since all building codes in the U.S. require materials to be installed in accordance with manufacturers' instructions.

          If your roofing is vapor-impermeable, and you don't include a ventilated air gap between the roofing underlayment and the roofing, there really isn't any logical reason to pay extra for vapor-permeable roofing underlayment. But the code doesn't always follow logic.

          1. Expert Member
            MALCOLM TAYLOR | | #113

            Martin,

            I'm not sure that's true of all impermeable underlayments anymore. Some of the high temperature ones specifically made for metal roofs make no mention of restrictions on un-vented assemblies, and some reference installation directly onto foam board.

          2. MartinHolladay | | #114

            Malcolm,
            Thanks for your comment. You're right that I haven't delved into the issue in a few years. If a manufacturer of vapor-impermeable roofing underlayment allows its product to be installed on an unvented roof assembly, then my comments clearly don't apply.

  90. user-741168 | | #90

    On hygric buoyancy
    In the recent Building Science column on attics, Joseph Lstiburek describes “hygric buoyancy.” He states: “In language that will irritate a physicist, the ‘moisture-laden air floats up to the top of the attic.’”

    Heck, I’ll take that bait. Consider air as a mixture of dry air and water vapor. Using partial pressures, we can do a vertical distribution of both of these gases according to the barometric formula as a function of height and differing molecular weight. When I run the numbers for 20 degC air, 10m of height, and a humidity ratio of exactly 0.01 at the bottom, I get a humidity ratio of 0.010005 at the top. This is consistent with my observations that vapor pressure is very uniform in a closed building, horizontally and vertically.

    If that’s true, then the argument calling for providing supply and return within the attic fades away. Which brings us back to the point where application of insulation directly to the underside of the roof deck is workable, effective and should not require special HVAC accommodation for vapor control. (Readers should be aware that code acceptance of these roof assemblies was made possible by Dr. Lstiburek’s efforts.)

    .

    1. picwa | | #105

      Are you calculating water content rather than measuring it?
      There is theory and then there are actual tests: Leave a two bottles whose insides are wetted, one upside-down in a dish rack, the other right-side up. The one right-side up will dry more quickly, even though it has no chance for further gravity drainage.
      I also note that water vapor, rather than being trapped at ground level, rises up to where it is cool and forms clouds.
      Or we could take a hygrometer and actually measure it, in air that is suitably still.

  91. Chaubenee | | #91

    Moisture issues from wet basements, gas-burning appliances?
    Could many of these issues be the result of: wet basements, no or little air conditioning or dehumidification, the use of unsealed gas burning appliances such as old school gas water heaters and boilers, gas cooktops, etc? I go into a lot of humid houses where in summer months there is little or no air conditioning being used, at least not enough to dry the places out. I think AJ Builder addresses this when he talks about people with 50 house plants and steam showers, etc.

  92. GBA Editor
    Martin Holladay | | #92

    Response to Joe Suhrada
    Joe,
    Yes, of course high indoor humidity can contribute to these problems. I noted that fact in the fourth paragraph of my article.

  93. Kevin S | | #93

    What to do now?
    I am the owner of a house remodeled about 7 years ago in Huntsville, Alabama. At that time, we added a floor in one part of the house and created a cape-code style roof with dormers and cathedral ceilings, and closed in an existing attic on another part of the house. We foamed both attics on the underside of the rafters and the insides of the gables with open cell foam.

    I have had several problems. 1) High humidity levels in the attics. Last summer I bought a couple of GE dehumidifiers from home depot. When I turned them on, one read 85% humidity and the other 60%. 2) a couple of roof leaks that the foam has concealed. When we finally got around to addressing them, the foam was fully saturated with water and some of the OSB sheathing was completely decayed. (This happened in a location separate from the two attics.) Recently found another leak in the old attic that has plywood sheathing. The plywood is delaminating. Looking over the rest of the house, it appears I have a section where the OSB sheathing has expanded and bowed up.

    The house has elevated levels of mold spores of various types.

    We've fixed one of the roof leaks (multiple times) and the resulting structural damage. Also cleaned the house for mold.

    Through this process, the open-cell foam we have removed seems to do a great job of absorbing water and holding it without dripping -- the air-side skin seems to hold the water. Is this how all open-cell foams work?

    I'm quite concerned that my mold levels may be contributed to by other hidden roof leaks. (And as a result, all of my expensive mold remediation will be for naught if we don't remedy the source.)

    So what next? It's virtually impossible to check for leaks from below. Tried a thermal camera with no success. Should I pull off all of the roofing from the top to check the condition of the sheathing? And if I do that, should I build in a ventilation channel before I put it back?

    (I can't seem to find anyone here locally that feels confident coming up with a plan to deal with foamed attics. Perhaps I haven't talked to the right person, but I'm running out of time. Any advice or direction to someone experienced in this area is appreciated.

    Kevin

  94. Andrew_C | | #94

    reply to Kevin
    You can read a bit more about Lstiburek's BSI-088 Venting Vapor over at BuildingScience. It starts w/ "ridge rot" in SIPS, and a vapor diffusion vent as a fix. I think they've used this technique w/ success in climate zones 1-3. Regardless of the hypotheses as to why it works, this approach has seemed to work in practice.
    This may not be applicable to your situation, but it may provide some insight that's useful.

    Good luck.

  95. Jon_R | | #95

    > #90 If that’s true, then
    > #90 If that’s true, then the argument calling for providing supply and return within the attic fades away.

    I disagree. There are various other ways to get high %RH in a sealed attic.

    [Editor's note: Jon R's comment is apparently a response to Bill Rose (Comment #90).]

    1. Jon_R | | #99

      > #90 This is consistent with my observations that vapor pressure is very uniform in a closed building,

      See here for data showing that floor/ridge vapor pressure isn't at all uniform.

  96. GibBarrus | | #96

    Open vs. Closed
    I am building myself a house in the mountains of North Carolina. We have pretty cold winters and also pretty humid summers. The house will have a radiant floor and no central AC, so no duct work. It is a bungalow style with living space in a large, common framed roof system, no real "attic space" to speak of. Roof pitch is 8/12 framed with 2"x12"s and I-joists. I built it without venting assuming I would insulate with spray foam. After talking to a few insulation contractors I was considering open cell foam for the roof. Reading your article here makes me think that paying the extra for closed cell might be worth the difference.

  97. GBA Editor
    Martin Holladay | | #97

    Response to Gib Barrus
    Gib,
    My advice to GBA readers is that it isn't a good idea to install open-cell spray foam on the underside of roof sheathing. For more information on this topic, see "High Humidity in Unvented Conditioned Attics."

    If your house will have sloped insulated roof assemblies -- that is, cathedral ceilings -- and you want to make these roof assemblies unvented, by far the best approach is to install an adequately thick layer of rigid foam on the exterior side of your roof sheathing. For more information on this approach, see "How to Install Rigid Foam On Top of Roof Sheathing."

  98. Expert Member
    Dana Dorsett | | #98

    Is the enviromental cost going to be worth it? @ Gib Barrus
    Closed cell foam has about 2x the polymer per R of open cell foam, and is most often blown with HFC245fa (instead of water, like open cell), an extremely powerful greenhouse gas. There are some closed cell foams blown with much lower impact HFO1234ze (Demilec Heatlok High Lift, Lapolla Foam Lok 2000 4G)), but the industry standard is still HFC245fa.

    The mountains of NC are in both US climate zone 4A and climate zone 5A. Which are you located in?

    In zone 4A if 30% or more of the total R is on the exterior of the roof deck, the vapor permeance of what's under the roof deck barely matters. In zone 5A it takes 40% or more. For an R49 roof the IRC prescribes R15 on the exterior for zone 4, R20 for zone 5 to be able to use class-III vapor retarders on the interior. That would be 3" of roofing polyisocycanurate (or a 3.5" nailbase panel) or 4" (or 4.5" nailbase panel). Securing the foam in place can be done by through screwing a nailer deck over the foam to the structural roof deck.

    Fiber faced roofing polyiso is commonly used in commercial roofing, and reclaimed goods from demolition and re-roofing are widely traded, typically at less than 1/3 the cost of virgin stock goods. Polyisocyanurate is blown with low-impact hydrocarbons, (usually pentane), and roofing polyiso can assumed to be in the R5.6-R5.8 per inch range (though some older CHFC blown goods will perform better than that). Using reclaimed good is the greenest insulation of all, no new polymer is being made, the environmental hits have already been taken, and re-use is just piling on to the benefit side of the cost^benefit balance.

    Searching the local craigslist for rigid + insulation usually turns up a few sources, eg:

    https://asheville.craigslist.org/search/sss?query=rigid+insulation

    https://boone.craigslist.org/search/sss?query=rigid+insulation

    https://greenville.craigslist.org/search/sss?query=rigid+insulation

    Rigid foam above the structural roof deck protects it far better than closed cell under the roof deck, since it stays warmer & drier, and can dry readily to the interior. The nailer deck above needs a standard roofing lay-up, but even in the event of a roof leak it's the nailer, not the structural roof deck that takes the brunt of the damage.

  99. Jon_R | | #100

    The Pallin simulation used open cell spray foam with no vapor retarder. A Class II vapor retarder is required by IRC code in zones 5-8, making the results and conclusions irrelevant for these zones and only conditionally relevant in others. Not a trivial difference - we are talking about an order of magnitude difference in the amount of wetting (~.4 perms vs ~4 perms).

    The Miller paper has similar problems (no retarder, poor attic conditioning), leaving us with what data when it comes to properly built/operated homes with a vapor retarder under open cell foam? Lstiburek who says open cell is fine with proper conditioning and this which also says it is fine?

    “Open-cell foam is risky in all climate zones” is a good example of invalid generalization
    or out of context. At minimum, it needs the important qualifier "without a vapor retarder".

  100. home26 | | #101

    Hi Martin,

    Big fan, I would really appreciate some advice on a project I'm doing in my home.

    I am renovating my house built in 1976 in south Florida climate zone 1 – Hot and humid

    I currently have a vented roof with a floor insulated attic and HVAC ducts in the attic. I have been researching unvented conditioned attics and wanted to find the best solution

    I’ve read the articles on Green Building Advisor “Creating a Conditioned Attic” and “Open-Cell Spray Foam and Damp Roof Sheathing”.

    I would really appreciate some advice to make sure I’m going about this the right way

    My plan is to use open-cell spray foam on roof rafters and condition the attic.

    Will the following steps solve the risk of damp roof sheathing?

    1) My roof was recently redone with asphalt shingles, so the roof is new and without leaks.

    2) Add an HVAC supply to the attic (What R-value open-cell should I install in my climate zone? How many CFM should I supply to the attic space? Do I need to add a return?)

    3) Add a whole-house dehumidifier and a supply to the attic (Do I need to install the air-handler and dehumidifier in the attic? Or can I just run supply and leave the air-handler in the house?) Will an Ultra-aire 120h be enough or should I use the SD12?

    4) Add an ERV (Do I need to add a supply to the attic?) Which unit do you recommend?

    Some more info about the project: My house is about 3500 sq ft and I want to change the HVAC to a VRF system. What is your opinion? I am also replacing all my windows to double-pane and low-e glass to get as tight as possible. Should I use a Mitsubishi VRF? Or another brand like Carrier Infinity?

    Should I just use Mini splits all over the house? If so how do I dehumidify?

    I have spoken with a couple of mechanical engineers and they both advised against everything I am doing. They both say to keep a vented attic and leave everything as-is saying it is dangerous.

    Their claim is humidity issues and fresh air change. When I explained I was adding a dehumidifier and an ERV they said it won’t solve the problems without explanation why.

    Can you recommend a mechanical engineer who agrees and can help design this project properly?

    I appreciate your help

    Thank You

    1. GBA Editor
      Martin Holladay | | #102

      Home 26,
      You have too many questions for me to answer. If you continue reading articles on GBA, you'll find answers to many of your questions. (Here is a link to one article you should read: "High Humidity in Unvented Conditioned Attics.")

      To get an answer to a specific question from the talented GBA community, you'll need to pare your questions down to one or two (instead of 12) and post the question(s) on GBA's question-and-answer page. Here is the link: Q&A page.

      Briefly:
      1. In Florida, running ERV will not lower the indoor humidity (except in very cool, dry weather). For most of the year, especially in summer, running an ERV adds humidity to your indoor air.

      2. If your central air conditioner is properly sized and properly installed, you shouldn't need a whole-house dehumidifier.

      1. home26 | | #103

        Sorry for all those questions I will keep it short.

        Climate Zone 1a 3300 sq ft home

        I am planning on installing open-cell spray foam on the roof rafters and seal the attic. I read the article about damp sheathing and was hoping to avoid it.
        1) Do I need to add HVAC supply to the attic to avoid humidity? If so how many CFM?
        2) By creating a tight envelope will I need an ERV or a whole-home dehumidifier to bring in fresh air and keep moisture low?

  101. GBA Editor
    Martin Holladay | | #104

    Home 26,
    Q. "Do I need to add an HVAC supply to the attic to avoid humidity?"

    A. Yes.

    Q. "If so how many CFM?"

    A. 50 cfm for each 1,000 square foot of ceiling area.

    Q. "By creating a tight envelope will I need an ERV?"

    A. No necessarily. But you need some sort of ventilation system. More information here: "Designing a Good Ventilation System."

    Q. "Do I a whole-home dehumidifier?"

    A. Usually not, unless something is wrong with your cooling system.

  102. DGrothe | | #106

    Question: Could you use a Heat Pump water heater ducted into the attic as apposed to a supply/return from the HVAC to solve the high humidity issues with open cell spray foam?

    1. charlie_sullivan | | #107

      It makes sense that that could be a part of the solution. It would work best if the water heater was on a floor right below the attic, so the ducts would be short.

  103. hondolobo33 | | #109

    I've seen several studies of the "vapor diffusion port" concept used with air permeable insulation to reduce risk, but never combined with open-cell spray foam installations. Would the tendency of moisture to migrate to the ridge stay true with an open-cell spray foam roof installation, and if so, would you think that a vapor diffusion port could reduce risk for ocspf roofs, particularly cathedral ceilings?

    1. GBA Editor
      Martin Holladay | | #110

      Hondo,
      I don't know the answer to your question, and I don't want to guess. That said, it's probable that the spray foam would inhibit drying of damp OSB. Vapor diffusion ports aren't used for spray-foam assemblies.

      In my article, "Vapor Diffusion Ports," I wrote, "According to Lstiburek, this approach works with any air-permeable insulation materials: 'Works with fiberglass batts. Works with netted cellulose. Works with adhesive spray-applied fiberglass. Works with fiberglass blown into a netted semi-permeable sheet.'”

  104. isuclones | | #115

    I've invested a lot of time reading up on unvented/conditioned attics, thinking that I wanted to convert my existing house (Omaha/5A) to unvented as part of an upcoming remodel. Really want to do it, had been struggling between open and closed cell. As part of that I came across this article, and that got me started reading about roofing underlayments. I have a 10 year old DaVinci roof with Titanium UDL 30, impermeable at 0.05 or 0.06 depending on source.

    So ... there's no way I should do spray foam directly to the roof deck, correct? Even open cell would not allow for enough permeability to keep the sheathing dry?

    I'm guessing that this has been done on a lot of retrofits without it ever being thought about, but probably no reason to risk it.

    I can't get all of those hours of research back!

    Thanks

  105. MartinHolladay | | #116

    IsuClones,
    I'm not familiar with DaVinci roofing products, but these roofing products appear to be plastic imitation shingles. I doubt whether these plastic shingles are vapor-permeable.

    In other words, it hardly matters how vapor-permeable your roofing underlayment is. Your sheathing is unlikely to dry to the exterior through plastic shingles, regardless of the roofing underlayment you choose.

    For more information on this issue, see "Sandwiching Roof Sheathing Between Two Impermeable Layers."

Log in or become a member to post a comment.

Related

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |