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

Forget Vapor Diffusion — Stop the Air Leaks!

Building codes are finally beginning to recognize that air barriers matter more than vapor barriers

Slowly moving in the right direction. Recent versions of model building codes — notably the 2009 IRC — have loosened vapor retarder requirements and tightened air-barrier requirements. Older versions of the code, including the 1992 CABO, were unjustifiably obsessed with vapor retarders and silent on air barriers.
Image Credit: Martin Holladay

Last week’s blog answered some common questions about vapor retarders. This elicited a comment from Bill Rose, research director of the Building Research Laboratory at the University of Illinois in Champaign. “We might imagine a future in which the building code sections that address the vapor barrier would all go blank,” Rose wrote. “I bet most readers would be able to design excellent buildings that perform well and are quite durable, without using the words ‘vapor barrier’ at any point in the process.”

Rose is a renowned building scientist and an expert on the history of vapor barrier requirements in building codes. These often counterproductive requirements were introduced with little scientific evidence to support their usefulness. As Rose points out, the abandonment of all code references to vapor retarders would appear to be a low-risk proposal, since it’s very rare for any problems to be caused by the diffusion of water vapor from the interior of a home towards the exterior.

An erroneous obsession

While building codes have historically shown a curious and unjustified obsession with vapor barriers, they have (until recent years) almost entirely ignored the need for air barriers. Unfortunately, when code officials decided to require vapor barriers but ignore air barriers, they got it exactly backwards.

Things are looking up, and codes are belatedly beginning to address the need for residential air barriers. In section 402.4.2, the 2009 International Energy Conservation Code (IECC) will, for the first time, require blower-door testing of some new homes. (Builders who don’t want to arrange for blower-door testing can instead comply with the new air-sealing requirements by an alternate method that requires visual inspection of 17 areas known to contribute to air leakage.)

So we’re now in an interesting period of transition. Pressured by building scientists to eliminate nonsensical vapor retarder…

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


  1. Doug McEvers | | #1

    Making buildings airtight is the key to performance and durability, ACH50 of 1.5 or less is very realistic in new construction. Great things happen from a comfort standpoint when air movement through the building envelope is under control.

  2. Bill Rose | | #2

    vapor barriers and oxycontin
    Hi Martin,

    Just a few comments before I take off for spring break, when I’ll be incommunicado for a week. Thanks for picking up on this issue.

    ASTM will be having a Condensation Conference in October in San Antonio. I’ve been invited to give the keynote, although my keynote material is still under review. My main claim is that materials can be wetted in two ways: 1) by chilling them and 2) by surrounding them with higher humidity. When the vapor barrier was invented, back in the late 30s, there was a conscious effort to obscure and mask the first way, combined with an effort to highlight the second way. (Was this because the insulation industry was running the show? Who knows. Who cares.)

    I am critical of the vapor barrier prescriptive requirement not in order to promote air barriers. I am critical of vapor barrier prescriptive requirements because it masks the very important relationship between temperatures of materials and wetness of materials. I can do a pretty good moisture survey using an infrared camera. Cold stuff gets wet; warm stuff tends to be dry. I tell students that if they seem to forget this relationship to go do their laundry. Cold means wet. Warm means dry.

    I guess I need to make a presentation slide that says in big bold letters: Capillary materials do not exhibit condensation at the dewpoint. Take a can of beer from the fridge on a muggy day. Condensation. Liquid water. Take a block of 2x4 from the same fridge and set it next to the can of beer. Same temperature, same surrounding vapor pressure. No condensation. No liquid water. Something happens to the 2x4 but it isn’t condensation. Sorption happens. Sorption also happens at vapor pressures above and below the dewpoint. Get picky? Below freezing materials are far less sorptive, but they exhibit frosting, not condensation. Fully saturated materials may exhibit condensation (liquid water) but they’re no longer capillary materials. So I’ll repeat: Capillary materials do not exhibit condensation at the dewpoint.

    You often hear the litany of wetting processes: bulk water, convection, diffusion, with each one being 100 times greater than the next. Fine. And when people mention diffusion the thoughts immediately turn to aquariums and high occupancy and museums (and Fick’s Law and vapor barriers and trap trap trap). Here’s a wetting process: take the 2x4 from the fridge and set it on a scale. It gets heavier. The added weight is additional water sorbed from the air. Where should this process fit among the three listed above? Is it unimportant? I don’t think so.

    Comments like these above are unsurprising to people who do building investigations and to people who run good hygrothermal models. But they are surprising to many designers who watched two lines cross on a dewpoint chart, or to consultants who like to wag their fingers at designers.

    I’ll close the same way I closed the previous thread: we’re a bunch of prescription junkies. Are there replacement drugs for those coming down off “one perm vapor barrier on the warm side”? Sure, and they are quite good, and much less harmful. What I’m suggesting is to imagine (in John Lennon’s sense) foregoing prescription entirely, until we know that our prescriptions really do deliver performance. Sorry I’ll be gone for a week. I feel like Eris (goddess of discord) tossing the golden apple.

  3. Jesse Thompson | | #3

    Air leaks & water leaks
    I would also hope that a positive side effect of greater scrutiny towards air leakage is that it will help us out tremendously with our even larger building problem, water leaks.

    We still don't have very good protocols for ensuring that we have discovered potential water leaks in new construction, but spending more time blower door testing our shells will have to help find all the gaps in the flashings that would otherwise have been missed.

    I hope.

  4. GBA Editor
    Martin Holladay | | #4

    Well, maybe
    You're probably right that a blower-door test will reveal some potential water leaks. But it won't reveal them all.

    Some of the best water-management details are old-fashioned lapped details — where the flashing or membrane near the top of the roof or wall laps over the flashing or membrane lower down. These laps work perfectly without being airtight. Not only that -- a blower door test will be unable to distinguish between a proper lap and a reverse lap.

    Sometimes there's no substitute for an intelligent, conscientious builder.

  5. Frank | | #5

    17 areas
    you mentioned, "... alternate method that requires visual inspection of 17 areas known to contribute to air leakage."
    could you pass a link to these 17 areas?

  6. GBA Editor
    Martin Holladay | | #6

    Table 402.4.2 of the IECC
    You're talking about Table 402.4.2 of the IECC.
    You can find it here:

    Look on page 21 of this 23-page document.

  7. peter in | | #7

    so in your opinion, a wall extruded foamboard between an exterior air barrier under hardiboard siding and interior drywall in north ga mountains does not need an interior vapor barrier for the winter. does it also mean that an exterior vapor barrier really doesn't matter in summer because the difference in indoor/outdoor temp. is negligible? please elaborate for me. thanks

  8. GBA Editor
    Martin Holladay | | #8

    Vapor retarders in Georgia
    Whether or not you need a vapor barrier depends in part on your local code, which you should research. According to the 2006 IRC and IECC, interior vapor retarders are not required in climate zones 1 through 4, an area that includes Georgia.

    If you have air conditioning and if you have a reservoir cladding (brick veneer, stucco, or fiber-cement), your walls may experience inward solar vapor drive during the summer. As long as your walls have no interior vapor retarder (and no vinyl wallpaper), the walls will be able to dry to the interior during the summer.

    For new construction in your climate, reservoir claddings (especially brick and stucco) are best installed over rigid foam sheathing, which reduces the chance of inward solar vapor drive.

  9. homedesign | | #9

    Rethinking PERSIST for Hot/Humid or Mixed/Humid
    Taking into consideration the "new emphasis" on Air barriers vs Vapor barriers.......
    Don't you think that a PERSIST wall in Austin Texas would be better off with a structural sheathing air control layer and some "crinkled" housewrap or building paper... and eliminate the "full coverage" Peel & Stick membrane?
    The typical (Ray Moore) Persist Wall already has foil faced polyiso to address Solar Vapor Drive.
    Why not eliminate the peel & stick membrane?

  10. GBA Editor
    Martin Holladay | | #10

    Peel & stick for PERSIST
    In a PERSIST house, the main purpose of the peel & stick membrane is indeed to act as an air barrier. It's certainly possible to create an air barrier with structural sheathing, if that's your preference.

    However, if you use structural sheathing instead of peel & stick for this purpose, you'll lose several desirable attributes of peel & stick, which:
    1. acts as a temporary roofing during construction and after the next hurricane, and
    2. seals around penetrations, including fastener penetrations.

  11. Ed Welch | | #11

    Exterior Foam vs. a vapor permeable wall...
    It is possible to achieve very low air infiltration in pretty standard stick framed construction, if the builder is very conscientious about the air barrier and very detailed in sealing all penetrations. Then you have both... an airtight structure and a vapor permeable wall.

    My point is or my question... what is the best assembly when the ventilation system is not maintained?.....isn't it safer to use vapor permeable materials in a wall assembly, allowing it to dry in both directions? If we start using rigid foam on the exterior of new buildings and retrofits, allowing only drying to the interior, aren't we creating a less forgiving structure when homeowners forget to maintain their ventilation systems.

  12. GBA Editor
    Martin Holladay | | #12

    Interior humidity levels
    I think you're worrying that if the homeowners disable their ventilation system, interior humidity levels will rise, increasing the chance of condensation in the wall.

    I think your worry is misguided. If the designer installed enough exterior foam, then the interior face of the foam sheathing (or the interior face of the OSB or plywood sheathing under the foam, if there is any) will never get cold enough to permit condensation.

    Moreover, many homes never see high indoor humidity levels in winter.

    If you want a bulletproof wall — if you're the kind of person who stays awake at night worrying about condensation in your wall cavities — then you should probably invest in a PERSIST wall.

  13. Ed Welch | | #13

    I guess I do worry...

    Yes, I guess I do worry about high humidity levels in the home when ventilation systems fail to provide the adequate air changes. What is the most forgiving system when people fail to maintain vital components of the 'house as a system' approach? I guess you are saying that in an airtight assembly, whether it is vapor permeable or not, occupants will suffer if ventilation is not maintained. And an airtight, vapor impermeable assembly using something like rigid foam will likely be more forgiving for the wall assembly because it is less likely that condensation will occur on the exterior sheathing. And we get better energy efficiency by eliminating thermal bridging, etc.

    I will look more into the PERSIST wall assembly.

  14. GBA Editor
    Martin Holladay | | #14

    The thicker the exterior foam, the more robust the wall
    If you can put all of your insulation on the exterior side of the home's structure — and if your insulation is all foam — you're good to go.

    ICFs also make a very robust wall. Or you can build a CMU (concrete block) wall, with all of the foam insulation attached to the exterior of the wall.

    You can read more about PERSIST walls here:
    "Getting Insulation Out of Your Walls and Ceilings."

  15. adkjac upstateny | | #15

    Bill Rose sorption and...
    12" densepack cellulose... is that a perfect world of noncondensing sorption?

  16. hillbilly2000 | | #16

    Martin, I enjoy your blog! I can't believe the smarts, dedication, and code understanding you bring to building science. Thanks. I read your blog religiously and I'm stimulated to learn more. PS In your 4th paragraph you say IECC when you mean IEEC. It's written out so we all figured it out. Thanks and see you next month. Best, Alan Ostner, Fayetteville, Arkansas

  17. GBA Editor
    Martin Holladay | | #17

    Thanks for your kind words, and thanks for pointing out the discrepancy in my blog. I meant to write International Energy Conservation (not Efficiency) Code; I've corrected the error.

  18. Anonymous | | #18

    Vapor barrier in the basement
    Our home is over a year old and we have framed in the basement walls, added inculation and a poly vapor barrier. Now that we are starting to finish off the basement and opened the wall for electrical wireing, we found all the walls "soping" wet. There was no accoustical sealent along the bottom of the barrier, along the floors but there were everywhere else - cornres, along the top, and around the windows.
    Question: Should the walls be sealed at the bottom as well as the top ?? I have had answers both way, and would like to hear from you on this.
    We live near Watertown NY, and we experiance cold winters and warm summers.

    Thanks for your input. Peter

  19. GBA Editor
    Martin Holladay | | #19

    No polyethylene, no fiberglass
    When it comes to insulating the interior of basement walls, there are two important rules:
    1. No fiberglass insulation, and
    2. No polyethylene anywhere.

    You don't mention what kind of insulation you used, but if it's fiberglass and if it's sopping wet, it's time to put the insulation and the polyethylene vapor barrier in a dumpster.

    Your basement walls can be insulated with any kind of rigid foam: extruded polystyrene (XPS), expanded polystyrenene (EPS), or polyisocyanurate. If you prefer, you can hire a contractor to spray your walls with closed-cell spray polyurethane foam.

    According to the latest version of the International Residential Code, basement walls in your climate zone need a minimum of R-15 insulation — that means at least 3 inches of XPS foam.

    Learn more here:

  20. Chris | | #20

    PERSIST vs ZIP system
    PERSIST vs ZIP system, are these two things not essentially one in the same? An adhesive membrane covering the roof and walls versus a plastic coated OSB with seams taped. I understand that screws penetrate the OSB and the tape may fail over time etc etc.

    My second question is: Would an EIFS (1" XPS) wall with open-cell spray foam insulated cavities (R-16) (rafters sprayed as well) and an ERV/dehumidifier in Dallas, TX along with a Grace membrane on the walls and roof be a recipe for a humidity disaster??

  21. Chris | | #21

    Or will the dehumidifier do
    Or will the dehumidifier do the job its designed to do and allow me to sleep at night?

  22. GBA Editor
    Martin Holladay | | #22

    PERSIST and Zip
    The Zip system is a proprietary sheathing system from a company named Huber. They sell special treated OSB and a proprietary tape for sealing the OSB seams. The Zip system does not require a builder to use any particular type of insulation.

    PERSIST, on the other hand, is an acronym (Pressure Equalized Rain-Screen Insulated Structure Technique) for a construction method, not a brand name. No company controls the PERSIST method -- it is an open-source concept. PERSIST refers to a type of construction that places all of the insulation on the exterior side of the wall sheathing and roof sheathing. It has nothing to do with the Zip system.

  23. GBA Editor
    Martin Holladay | | #23

    Is this a recipe for disaster?
    Whether or not the system you describe is a recipe for disaster or a 100-year system depends entirely on its details. Without knowing a lot more details, it's hard to evaluate the robustness of the building.

  24. Chris | | #24

    I guess I should have
    I guess I should have clarified my questions a bit more...the PERSIST wall system utilizes a self adhered membrane to cover the roof and walls, for example lets say Grace Ice and Water membrane. Just thinking about the air/water/vapor barrier membrane covering the OSB, how is this principle any different from the Huber ZIP system, theoretically?

    With regard to my second question, I guess this is more about feasibility of membrane use on the exterior than that of a full PERSIST wall. I am in the planning phase and I am designing a tight, energy efficient home. The home will be stick build on slab on grade foundation with all wall and roof cavities filled with open-cell spray foam insulation. Walls and roof concept from exterior to interior are: Stucco, lathe, XPS, Tyvek Drainwrap, ZIP system, studs, insulation; Roof tile, Grace membrane, OSB, rafters, insulation. The home will be heated and cooled with a geothermal heat pump. What other details are important?

    My question is regarding replacing the Tyvek drainwrap and ZIP system with the Grace membrane and regular OSB and thus totally encapsulating the exterior in membrane. With most of the insulation being inside the membrane would this shift the condensation point to the interior side of the OSB or within the spray foam insulation? If so, will the whole house dehumidifier coupled to the HVAC duct be sufficient to alleviate any humidity/condensation issues?

  25. GBA Editor
    Martin Holladay | | #25

    The permeance of Zip system sheathing
    I don't know the permeance of Zip System sheathing off the top of my head; you can probably contact Huber to find it out. It's certainly possible — even probable — that Zip System sheathing is more permeable than Grace Ice & Water Shield.

    Anyone who is contemplating covering a roof or wall with Ice & Water Shield needs to think carefully about the wall and roof assemblies' drying potential. The maximum drying potential exists if you don't put any insulation in the stud bays or between the rafters. That's how PERSIST walls can dry to the interior.

    If you're spraying foam on the inside of the sheathing, you're certainly reducing the ability of the sheathing to dry to the interior. Why not avoid potential problems by substituting asphalt felt for the Ice & Water Shield?

  26. Chris | | #26

    With the amount of

    With the amount of rain here I prefer a water barrier over asphalt felt paper but it doesn't have to be a self-adhered membrane. The main reason this is fascinating is for the air tight sealing ability, which accounts for a large portion of the PERSIST method benefits.

    Thanks for the replies, that is exactly what concerns me. Open-cell spray foam has a high vapor permeance so I was wrestling with if condensation would even be a problem, not to mention the dehumidifier.

  27. GBA Editor
    Martin Holladay | | #27

    Don't mix and match
    If you like the PERSIST system, build a PERSIST house. Put all of the insulation outside of the sheathing.

    That's my advice. However, other builders have done what you have done -- they've put some of the insulation on both sides of the sheathing. It's possible, but:
    1. Do your dew point calculation.
    2. Be sure that the wall assembly has as much drying ability as possible in both directions -- a rainscreen on the exterior, and as little insulation as possible on the interior. Of course, no interior polyethylene.

  28. HDendy | | #28

    #1 Do your dew point calculation
    Great discussion! It would be nice to see a write up on dew point calculations as it applies to this conversation. That seems like THE crucial step to prevent designing problems. I love the ideas of the PERSIST walls, but see "some insulation on both sides of the sheathing" as the most likely scenario we'll see in most places....and embraced by most clients.

  29. wjrobinson | | #29

    Wall rot failures with insulation on both sides of OSB
    One of the worst failures of a wall that I have renovated was a wall that had board foam external to OSB and fiberglass batt internal to OSB. 2x4 wall in Upstate NY. Not sure where water came from but the wall was completely rotted out. almost all the 2x4 framing was mulch hiding behind drywall and trim etc.

    The tighter we build homes the more we all have to get the details correct. PERSIST style walls have very good drying potential built into them.

    Bruce Brownell in Edinburg NY builds a wall that is similar to a PERSIST. Though it is his own design and I don't think ice and water shield is a key component. Hundreds of builds with minor issues at air leak spots.

  30. wjrobinson | | #30

    Bill Rose sorption comments please continue
    Would like to see a dedicate Bill Rose thread on sorption including a list of walls that work and don't work.

    Martin if there is a blog to do with Bill Rose and sorption, please point us. Thanks

  31. GBA Editor
    Martin Holladay | | #31

    Respons to AJ
    I read your post here in Westford, Mass., where I just heard Bill Rose give a presentation on a variety of building science topics. He held up a mirror and asked the audience, "Can you get condensation on a mirror?" and he answered his own question: "Of course you can get condensation on a mirror."

    Then he held up a sponge. "Can you get condensation on a sponge?" And he answered his own question, "Well, yes -- you can get condensation on a sponge -- if you use the sponge to wipe the mirror."

    Then he told a story. "On a hot humid day, I take a beer out of the refrigerator. I put the cold beer on the table. Do I get condensation on the beer can? Of course. Now I go back to the refrigerator and I take a chunk of 2x4 that I keep in the fridge and I take that out and put it on the table. Do I get condensation on the 2x4?"

    Finally, he said, "So I got a call from someone who said, "I've got a problem, Bill. I'm getting condensation in the inside of my wall cavities." Bill answered, "You've got condensation in your walls? Why are you building your wall out of beer cans?"

    He advised the audience that the word "condensation" is often misused to talk about sorption.

  32. bruce | | #32

    Wall rot failures with insulation on both sides of OSB by aj

    What do you think caused the problem aj builder described in the 8-2-10 note? Do think the foam board on the outside of the OSB had anything to do with it

  33. GBA Editor
    Martin Holladay | | #33

    Response to Bruce
    Q. "What do you think caused the problem AJ Builder described in the 8-2-10 note?"

    A. There is no way to tell without a site visit. Perhaps AJ Builder has a theory.

  34. Rick | | #34

    air barrier 80 yr old tile block/brick vennier
    Hello I am helping to repair/remodel a two story tile block with brick vennier building,and full/damp basement.Located in northern ohio.I belive the 1st floor has 1 inch of foam between wood furring strips,This is to remain in place. Also the vynal windows are not sealed around the perimiter.There is a drop ceiling.Forced air furnace/and ac on 2nd floor only. I have placed 2x4 studs about 2 inch away from the plaster parging on the tile block. Can i use 6inch paper backed fiberglass that will touch the old wall,or stay with 31/2 inch. I plan to us a 4mill plastic air/vapor barrier on the walls and ceiling,with sealant and tape, then drywall,The floor is old hardwood,and will draw air from below.I would Enjoy any suggestions you might have as i am a jack of all trades master of none. Thank You Rick

  35. GBA Editor
    Martin Holladay | | #35

    Response to Rick
    1. If you know that the house has air leaks, then your most important task is to perform air sealing work. If you aren't sure where the air leaks are, then hire a home performance contractor with a blower door to show you.

    2. Don't install interior polyethylene in Ohio, It will cause more problems than it solves.

    3. Six-inch fiberglass is better than 3 1/2 inch fiberglass. But almost any other type of insulation -- cellulose, spray foam, or rigid foam -- is better than fiberglass.

  36. Robert Hronek | | #36


    The rotted walls you talked about, I dont think it was so much the insulation on both sides but the attnetion to detail. For starters there was fiberglass in the walls, My guess is it had leaky walls combined with porous fibeglass about for air to carry large amounts of water vapor. Then there was probably a plastic vapor barrier, What details were there outside to make the walls air tight and what was the cladding.

    In reading Joe L's Mind the Gap it seams as if very small gaps can provide drainage and vapor relief. and have a very minimal effect on the R value of rigid insualtion.

    I think there are different approaches whether it is double wall, persist or some other system. It is critical that you are aware of the proper way to use whatever system you have chosen.

    On the walls AJ is talking about I would say that there was a way for water to get in but not out. A small hole in a plastic vapor barrier will in a wall that is not air tight will allow a lot of water vapor in but not out. The problem with plastic is that it is rarely sealed around electrical boxes. We have all felt home much air moves around electrical outlets couple with each stud being drilled and you have the perfect design to fill the wall with vapor and a chance to condense on never get out.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |