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

The 2012 Code Encourages Risky Wall Strategies

Building codes promote “R20+R5” walls in cold climates — a recipe for mold and possible rot

R-5 foam on the exterior side of a 2x6 wall? That approach will work in Virginia, but it won't fly in Vermont.
Image Credit: Owens Corning

Builders who follow the prescriptive requirements of the 2012 International Residential Code (IRC) in Climate Zone 6, 7, or 8 are required to install a minimum of “20+5 or 13+10” wall insulation. What does this mean? According to an explanatory footnote in the code, the “First value is cavity insulation, [and the] second is continuous insulation or insulated siding, so ‘13+5’ means R-13 cavity insulation plus R-5 continuous insulation or insulated siding.”

Here in Vermont (a Climate Zone 6 state), builders have been framing walls with 2x6s for at least 35 or 40 years. Nobody installs R-13 insulation in walls in Vermont, so the most likely way that Vermont builders will comply with this code provision is to install R-20 insulation between the studs and R-5 rigid foam insulation on the exterior side of the wall sheathing.

While this approach meets minimum code requirements, it violates a tenet of good wall design: namely, that any rigid foam installed on the exterior side of wall sheathing needs to be thick enough to keep the sheathing above the dew point during the winter. (For more information on this issue, see Calculating the Minimum Thickness of Rigid Foam Sheathing.)

If you want to eliminate an interior vapor barrier, the code requires thick foam

Cold wall sheathing is more likely to be damp than warm wall sheathing, so the “20+5” requirement is problematic. There’s more, however: the 2012 IRC still maintains antiquated vapor barrier requirements. In section R702.7, the code notes that “Class I or II vapor retarders are required on the interior side of frame walls in Climate Zones 5, 6, 7, 8 and Marine 4.” That requirement has always been unfortunate, but it has proven to be hard to change.

Of course, if you’re building a wall with exterior rigid foam, the wall can no longer dry to the exterior; it needs to be able to dry…

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. lecolevielle2002 | | #1

    cavity insulation
    Mr. Holladay, for the purposes of your discussion, are you assuming that the cavity insulation is fiber glass batting, cellulose, or some other comparable insulation? That is, would you have the same concerns, if the cavity insulation were closed cell foam insulation?

    (I'm thinking in particular of Zip R-Sheathing over closed cell foam cavity insulation).

    Many thanks for your helpful explanations!

  2. GBA Editor
    Martin Holladay | | #2

    Response to Lecole Vielle
    Closed-cell spray foam insulation solves one problem -- vapor diffusion from the interior to the cold sheathing during the winter -- but introduces another: it prevents damp sheathing from drying inward.

    In general, most experts recommend against this kind of "foam sandwich" that prevents the OSB or plywood sheathing from drying in at least one direction.

  3. lecolevielle2002 | | #3

    Prevention of damp sheathing from drying
    If I have understood you correctly, Mr. Holladay, prevention of damp sheathing from drying is a problem anytime one uses closed cell spray foam insulation behind sheathing (in turn behind some kind of vapor barrier). That is, the problem is that there are effectively two vapor barriers: the R-Sheathing's ployiso insulation and the closed cell spray foam.

    At, you said "Zip-R sheathing has a permeance that is below 1 perm; for all intents and purposes, it is a vapor barrier.

    You are correct that the polyiso layer of Zip-R sheathing faces inward. Since any building with Zip-R sheathing can't be expected to dry outward, adding closed-cell or open-cell spray foam on the interior side of the polyiso layer doesn't change anything -- at least not with respect to the drying direction.

    So you can go ahead and use spray foam between the studs if you want to."

    Perhaps, the better alternative is to opt for regular Zip sheathing and closed cell foam, as opposed to Zip R-Sheathing. That might allow any moisture to dry out. The HuberWood website says, "The ZIP System® Sheathing overlay protects against water intrusion while providing an optimal permeance level (12-16 perms) to allow panels to properly dry out."

    Going with the alternative would mean suffering from thermal bridging, of course. Boy, it's never easy.

  4. Gregory La Vardera | | #4

    easy solution & better solution
    The easy solution is for the code to prohibit foam exterior insulation where the R value of the exterior layers is less than 1/3 of the total R value of the exterior + cavity. This would be easy to implement, one line in the code, perhaps a foot note on the two charts shown.

    A better solution would be to require a design certification by a Building Scientist for any wall system using exterior foam insulation layers. Because one thing you've made clear is that its complicated, and that you really do need an expert to design an assembly correctly with exterior foam. And it seems right to have those building science experts take responsibility and liability for those assemblies. Then problems will be greatly reduced and when foam is put outside, it will be done right.

    And builders who don't wish this extra expense can use interior side vapor control as they always have, and use permeable exterior insulation where they need those extra layers. And no special certification would be needed for these well understood systems.

  5. GBA Editor
    Martin Holladay | | #5

    Second response to Lecole Vielle
    You wrote, "Prevention of damp sheathing from drying is a problem anytime one uses closed-cell spray foam insulation behind sheathing."

    That's not a hard-and-fast rule. Many types of wall sheathing and siding allow drying to the exterior, for one thing. For another, even when there is no drying to the exterior -- for example, in the case of a cathedral ceiling with vapor-impermeable roofing above the sheathing -- it sometimes makes sense to install closed-cell spray foam on the interior side of the sheathing, even though there is no way for the sheathing to dry. If the sheathing is very dry on the day that the spray foam is installed, the risks may be worth taking.

    The situation with Huber Zip R-Sheathing, which you bring up, is a special case. As you pointed out, the rigid foam layer of this sandwich product is on the interior side of the Zip R-Sheathing, so installing closed-cell spray foam on the interior side of this product isn't particularly risky. The OSB layer faces the exterior, and can still dry (somewhat) to the exterior.

  6. James Steel | | #6

    Mineral Wool Semi-Rigid exterior insulation
    In Canada, similar requirements are in place in cold climates. In B.C. in particular, the exterior R-5 is most often met with semi-rigid mineral wool board insulation. It's vapor permeable. A vapor-permeable WRB is applied to the sheathing inside of the mineral wool exterior insulation. This has worked well to both limit thermal bridging and allow wood frame structures to dry to the outside, as well as inside. Foam isn't the only way to achieve R-5 C.I. (continuous insulation) on the outside of the building.

  7. GBA Editor
    Martin Holladay | | #7

    Response to James Steel
    My article mentions the mineral wool approach, and GBA has published many other articles on the topic as well:

    Installing Mineral Wool Insulation Over Exterior Wall Sheathing

    Installing Roxul Mineral Wool on Exterior Walls

    Mineral Wool Boardstock Insulation Gains Ground

    Wrapping an Older House with Rock Wool Insulation

  8. Expert Member
    Dana Dorsett | | #8

    it's not that easy (response to Gregory LaVardera)
    A one size fits all solution would be at once too much and not enough.

    A 1/3 of total-R requirement would be more than needed for US climate zones 5 & lower, and would be insufficient protection for zones 6 & up, and would even violate the current IRC prescriptives for zone 6 & 7 per Table R702.7.1.

    Putting R19 or R20 cavity fill in a 2 x 6 wall would imply a requirement of only R10 on the exterior using the 1/3 rule, whereas R702.7.1 prescribes a minimum of R11.25, an R15 for zones 7 & 8. The R13 + 5 works fine for dew point control in US climate zones 4 & 5 (if a bit on the minimalist side for zone 5.)

    Rigid rock wool under rainscreened siding rather than foam is a reasonable solution when cheating Table R702.7.1, and R6 Type II EPS is a better solution than R7.5 foil faced polyiso or R5 XPS for compliance in zone 5, due to the much higher vapor permeance of rock wool & EPS at those thicknesses. At R6/1.5" Type-II EPS is as vapor permeable as many grades of exterior house paint (paint that could be legally & safely applied without a rainscreen over 2x6 R20 in most of zone 5), and nearly as vapor permeable as interior latex paints. The fact that R6 EPS over an R20 cavity fill actually performs north of R7 whenever the outdoor temps are low enough to be concerned about sheathing dropping below the interior dew point temps helps too. R20 + 6 EPS(unfaced) makes it with or without a class-II or class-I in zone 5, despite violating Table R702.7.1 prescriptive, and R20 + R7.5 foil faced polyiso (which meets the prescriptive) is much riskier, due to the far-below-rated-R performance of polyiso at sub-freezing foam temps, and the extremely low vapor permeance of the facers.

    But it's hard to include this much nuance in a simple & clear code prescription.

  9. James Steel | | #9

    Re: Response to James Steel
    Thanks for your response. I didn't miss your mention of mineral wool. And I've read most all the GBA articles on mineral wool. My point is that, if I understand the new code correctly, the requirement is for R-5 continuous insulation. Not R-5 foam. I agree that further clarification on what type of exterior insulation should be used would be great. But this is also self regulating. If you don't want your building to rot, probably shouldn't build a hybrid wall assembly with R-5 exterior foam. If the code were to read "R-5 vapor permeable exterior insulation", the foam industry would have a conniption. After all, if enough exterior foam is on the outside then there isn't a problem... No? So the problem isn't foam?

  10. GBA Editor
    Martin Holladay | | #10

    Second response to James Steel
    Thanks for your comments. You are right, of course.

    In particular, you are correct to point out that ambiguous or deliberately obfuscatory code language is often the result of squawking by manufacturers and other so-called "stakeholders" with economic interests to protect.

  11. James Steel | | #11

    Re: Second response to James Steel
    You read into that exactly where I was going!

  12. Gregory La Vardera | | #12

    Well the discussion is
    Well the discussion is convincing me that the way to go is for exterior foam assemblies be certified by a building scientist on a project by project basis. Too many variables for a simple code prescription.

    The code should only prescribe assemblies with exterior permeable insulation, drying to outside, with interior vapor control, and that vapor control sheet of the variable permeability type if there is air conditioning.

    Makes it more complicated for builders on the one hand, but keeps the code clean and simple.

  13. wjrobinson | | #13

    The code could have complete prescriptive plan details sets
    If the code had a book of allowed plan detail sheets builders, subs, archs homeowners and inspectors all would have life easier and have more time for tubing summer rapids and shooshing the slopes or surfing the waves off Maui.

    How to start this? Make one set of complimentary home plan details for each climate. Add more each year. Any manufacturer could also write up complete home detail sheets, as they would then be liable for there success.

    The problem has always been that homes are not built from one companies detail sheets. Some one of us starts pulling together this detail and that manufacturer and a little GBA and a little Joe L. and a little ajbuilder...

    We're lucky any homes are built properly. Here's an example of whatever... In my area every production home is built with poured concrete foundations, no foam under the slabs or on the walls, what they do is build a stud wall off the concrete a few inches and then they fiberglass insulate this wall, then they put roll metalized paper over this wall two rows horizontally applied, no taped seams, and at the top of the wall the air space is open to the underfloor. The rim joist... 1/4" of closed cell spray foam. The electric panel is mounted to the concrete wall no insulation, wall insulation interrupted air gapped all around. It all looks great to the untrained average homeowner but myself or Joe L or Martin would not think any of it was right at all. AND THIS IS STANDARD HERE AND IS EXACTLY WHAT OUR CODE OFFICERS WANT TO SEE. WE HAVE TO HAVE A VAPOR BARRIER OVER OUR PINK INSULATION IN OUR BASEMENTS. Try to find Martin saying that is the way to do it.

    GBA should publish complete home plan details, a package of details that work together. With a list of to do's and not to do's. Sell the set of details for any of us to design a home with or give to our home designer. Joe L. should do it. But really THE PRESCRIPTIVE ASPECT OF THE CODE SHOULD BE A COMPLETE DETAIL LIST FOR A COMPLETE HOME. We all still can build other than prescriptive via the performance part of the code.

    Prescriptive codes need to be complete sets of unified details.

    All the options we have as builders is where any of us get in trouble. I know many of us love having options and freedom to do our own ideas. I do. That is always possible now and would be with my idea of how to fix the codes.

  14. GBA Editor
    Martin Holladay | | #14

    Response to AJ Builder
    Green Building Advisor is not in the business of developing or selling home plans; plenty of other companies already do that.

    But GBA has a very large library of details, fully accessible for downloading by any GBA Pro member (that is, a subscriber).

    Here is the link to our detail library: GBA Detail Library.

  15. Expert Member
    MALCOLM TAYLOR | | #15

    Response to AJ
    Our code does just that with wall and ceiling assemblies that are required to meet a certain standard for fire and acoustic separations. It provides pages of different assemblies using various materials. I don't see why they couldn't do the same for exterior wall and ceiling assemblies required to meet new energy

  16. wjrobinson | | #16

    Martin disconnected detail sheets add to the problem
    Anyone who publishes home construction details without adding a tool that groups compatible details is making things worse not better.

    Like Malcolm says, the code in NY has us submit a passed Reschek. All that would have to be done would be to add detail sheets to reschek that then it either allowed or disallowed the group of detail sheets selected.

    We not talking Mars mission planning... It could be done the same way building code is developed and published. I could ever do it, PHD not.

    Must go now and help eat a pig and monitor proper completion of all beer on site.

  17. Gregory La Vardera | | #17

    Sounds crazy.
    No, the idea of creating a code book of wall assemblies is crazy. Every state adopts the code at a different rate, each would have to adopt each wall assembly. There are so many products, so many possible assemblies, and so many local climate variations that defy the climate zones. For exterior insulation interior drying walls the climate zones are not granular enough to broadly adopt standard assemblies, as you can for interior side vapor control in cold climates.

    Think about it - what you are describing is like the UL Listing book of fire rated assemblies. Except now they are going to be indexed over 6 climate zones plus marine and arid overlays. It would be a monumental undertaking. Its not rocket science, but it is a whole lot of science, by a whole lot of industry players, who will no doubt dig in heels to oppose such regulation. They really don't care if you get mold or rotted sheathing, only that you buy their product.

    Exterior side insulation/vapor control/interior drying is an unforgiving and non-resilient assembly. Its easy to get it all wrong, and even when its a little wrong it can work very badly. Certification by qualified Building Scientists is the only reliable way to have some assurance that the assembly will be properly designed for the application. Even that measure would be greatly resisted. Face it, we are likely to just have the confusing mess outlined in this article.

  18. wjrobinson | | #18

    Nuts Greg, nuts.
    The detail sheets exist today. Assembling a set for all climates would take Joe L. ten minutes.

    And you still would have the right to build a performance based home that you had a professional put together the plan. which is your building scientist idea.

    I love the idea.

    Joe L has really already done this if you buy his regional books you can see he is just about doing this with those publications which is my building scientist/building code idea.

    I would have no problem with handing the whole inspection system to Joe. Done.

    If you have used Rescheck to certify a home, you would see that it with just a bit of added programming could be used to self collate a group of compatible thermal moisture details.

    I should get this aj-Rescheck-building detail grouping software built and sell it.

  19. jinmtvt | | #19

    mmmm ..
    Unfortunately, enforcing this fanfold exception would be tricky, because it would require building inspectors to know how to research the vapor permeance ratings of building products.

    Is this an intended joke ??

    buildings inspectors = n00bs ??!!

  20. GBA Editor
    Martin Holladay | | #20

    Response to Jin Kazana
    Some fanfold insulations have facings. It can therefore be hard to determine the vapor permeance of a sample of fanfold insulation by looking at it.

    Of course, the federal government could pass a new federal law, the Permeance Rule (along the lines of the R-Value Rule which requires insulation products to be labeled with R-values). If such a law were passed, manufacturers of certain building products, including fanfold foam insulation, could be required to label their products with a perm rating.

    In the absence of such a law, I don't see it happening.

    And I don't think it's a good idea to have a code requirement that is hard to verify when an inspector visits a job site. So this issue is more complicated than it first appears.

  21. Lo3H5B3Ck7 | | #21

    Mineral wool
    This exactly why we recommend using mineral wool in a rain screen application. First the mineral wool is vapor diffuse. A properly applied air and moisture barrier is the first step. Beyond that the mineral will dry to the outside of the wall assembly if it does get wet. Second mineral wool's thermal performance holds its performance over a wide variety of temperature extremes. This helps eliminate a moving dew point target. Lastly, a good 3/4" to 1" air gap created with vertically applied furring strips allows for bulk water run off and air movement behind sheathing, siding, etc. This allows for proper drying and the prevention of heat build up.

  22. user-1081937 | | #22

    Why codify it?
    I think the key phrase in Martin's article for me is that the code is simply adopting building science-less rules of thumb based on light frame construction which is proven to be cheap when your only using 2x4s and 2x6s. Now that we are getting beyond what those cavities can take for r-values the prescriptive approach is not very viable. And code struggles with non prescriptive assemblies by nature. Greg's suggestion that we treat these higher r-value assemblies with a more specialized review and design strategies seems to be the natural evolution. The alternative is many rotting boxes and lawsuits, which is what drives code to a large extent historically.

    Why not treat wall assemblies like we would structural loads and have them properly engineered on each and every job for the location? Sure it drives the cost up but much less so as engineers learn the physics of vapor diffusion and retention. And of course a lot cheaper than failed buildings.

    I also think this should force us to reevaluate the foam wall system to get to where we want to go- robust, durable and low cost assemblies. Double walls, larson trusses, or other built out cavities which can be blown with low cost products like cellulose combined with smart vapor barriers starts to make a lot more sense than mindlessly adding expensive foam to the old fashioned 2x wall.

    Come to think of it this sounds like what they have been doing in Europe for some time... humm.

  23. user-1119494 | | #23

    simple solution?
    Sure, the situation is complex and different assemblies will work well in one situation and fail miserably in another. Why not prescriptively allow work-anywhere solutions like PERSIST, but require others to get analyzed and stamped?

  24. GBA Editor
    Martin Holladay | | #24

    Response to Dustin Harris
    Thanks. Your suggestion has merit. But any code change that appears to favor one wall system as the default solution is likely to face resistance from those who prefer a different wall system -- especially if the new default is one like PERSIST, which doesn't resemble the type of residential walls that are usually built in the U.S.

    It's interesting to contrast your suggestion with Gregory La Vardera's suggestion. The two suggestions are mirror images of each other, just as an ICF wall is a mirror image of a Thermomass wall.

  25. Mica239 | | #25

    How do SIPS (structural insulated panel system) rate in this "risky wall strategies" discussion. I am considering SIPS for a building site (zone 7) that is not vehicle accessible (and therefore, reduce the number of trips taken to bring building materials to the work site).

  26. GBA Editor
    Martin Holladay | | #26

    Response to Joel Cooper
    Some SIP buildings have suffered OSB rot. When this occurs, the affected area is usually the exterior OSB facing, near a SIP seam. The cause of this rot is a poorly sealed seam that allows air leakage. Exfiltrating air carries moisture in it, and the moisture condenses when it reaches a cold surface (for example, asphalt felt or cold OSB) during the winter.

    In general, the risk of this type of rot can be greatly reduced if the SIP installers pay attention to air sealing. In addition to using caulk and canned spray foam as directed by the SIP manufacturer, it's wise to also seal all SIP seams with high quality tape on the interior. For more information on this issue, see How to Protect Structural Insulated Panels from Decay.

    Finally, the airtightness of SIP seams should always be verified as part of the construction process. This verification is done with a blower door and a smoke pencil (or, in come cases, with a blower door and an infrared camera).

  27. rfaesy | | #27

    Why Not Just "R-25"?
    What if the code for walls said "R-25" instead of "R-20+5 or R-13+10"? This would remove the faulty building science (for 20+5), state a total minimum R-value that could be achieved through multiple approaches, and could be the basis for sound building science support through other market approaches. Could this solve the conundrum?

  28. GBA Editor
    Martin Holladay | | #28

    Response to Richard Faesy
    Your suggestion would work, of course, and it is one of the options I mentioned in my article.

    I wrote, "If code writers want to allow R-25 walls, that’s fine. There are ways to build R-25 walls that don’t cause rot. But code writers will have to come up with consistent method for calculating the R-value of walls — one that is easy for code enforcement officers and builders to understand and use. Up until now, no one has wanted to do that. (This task is complicated — but it’s possible.)"

    For example, do enforcement officials have to take thermal bridging through studs into account when calculating R-values? If they ignore thermal bridging, what do they do about buildings with steel studs? Does the code need to include a table listing the R-values per inch of a variety of insulation materials? How do you calculate the R-value (or measure the thickness) of bumpy and uneven applications of spray polyurethane foam? Is the R-value of insulation between studs treated identically to continuous insulation, or must its R-value be downgraded by a framing factor? Do code enforcement officials have to calculate a building's actual framing factor, or will there be a published default value for an average framing factor?

    There is another issue to wrestle with if code writers accept your proposal: will 20+5 still be acceptable, or will the code decide to outlaw 20+5 in Climate Zone 6?

  29. wjrobinson | | #29

    After reading all the goofy ideas my idea of approved assemblies is by far the simplest and simple to implement.

    Simply aj

  30. Dana1 | | #30

    R25 center-cavity is not R25
    The reason R20+ R5 continuous insuiation (not thermally bridged) is due to the fact that R25 center-cavity as a low-densty batt in 2x8 framing is also R25 nominal, but dramatcally underperforms R20 in a 2x6 framed wall with R5 continuous insulation. The "whole wall" R of R20 cavity fill is about R13 after factoring in the thermal bridging of the framing. Adding the sheathing, gypsum & siding adds anotyher R1-R1.5 or so, bringing it up to R14-R15, then adding R5 of continous insulation brings the whole assembly up to about R20 performance. R25 cavity fill in a 2x8 16" o.c. assembly without continuous insulation barely makes R17-ish performance after the thermal bridging. Energy codes are about the total whole-wall U-factor/R-value, and not center cavity.

  31. GBA Editor
    Martin Holladay | | #31

    Response to Dana Dorsett
    You wrote, "Energy codes are about the total whole-wall U-factor/R-value, and not center cavity."

    Really? What makes you think that? I don't know of any evidence to support your contention.

    In regions where the prescriptive code calls for a minimum of R-13 wall insulation, code enforcement officials have always deemed builders to be in compliance with the code if the builders install batts labeled "R-13" between the studs.

  32. Expert Member
    Dana Dorsett | | #32

    The evidence is found in Table R402.1.1, with a bit of analysis
    The U-factors/whole-wall-R of the alternate assemblies are very similar.

    Take the wall prescriptives for zones 3- 5: At typical framing fractions an R20 wall comes in at about R14-R15 after factoring thermal bridging and adding other layers, R13 + 5 comes in at about R14.5-R15- a nearly identical U-factor, and which very similar to the R13/R17c.i. mass wall prescription for zone 5. The prescribed center cavity R is either R20, or (R13 + R5=)R 18. That's a 10% difference in R value measured at the center cavity, but nearly identical whole wall performance.

    For zones 6 & higher they prescribe R20+5, which comes in around R19-R20 whole-wall, after framing + other layers, as well as R13 +10, which also comes in at R19-R20 whole-wall, which is identical to the prescriptive R20 (if more than 50% of the insulation is on the interior) of a mass wall for zone 6, or R19/R21 for zones 7+. R20+5 has a center cavity R of R25, whereas R13 +10 comes in at R23, but the whole-wall performance is within R0.5 of one another.

    In zones 1-2 the prescriptive is R13, which is merely the most commonly available batt insulation that fits in 2x4 cavites. They could have prescribed R11s which are slightly cheaper but I suspect the rationale for the denser R13 batts had to do with infiltration. Wall-R has very little effect on cooling loads, but air infiltration does.

    BTW: The R25 2 x 8 framed wall example may only hit R17 at a 25% framing fraction (16" o.c. spacing), but would come close to R20 at a 15% framing fraction (as can be had using some advanced framing techinques at 24" o.c. stud spacing.).

  33. GBA Editor
    Martin Holladay | | #33

    Response to Dana Dorsett
    I understand your analysis, but I think you are being extremely charitable when you attribute a concern with whole-wall R-values to code writers.

    The most charitable version I can come up is that there is no consistency in how the code approaches wall R-values. For decades, the code has referred to walls with R-13 batts as R-13 walls.

    Now that the code has introduced 20+5 and 13+10, your analysis makes sense. But to this day, the code does not include any guidelines for code enforcement officials to tell them how to calculate the R-value of wall insulation.

    In some cases, the logic behind the numbers can be inferred (as you demonstrated). But for most of the country, for decades, code enforcement officers (and builders) have said, "The code says that you need an R-13 wall" or "The code says that you need an R-19 wall."

  34. Expert Member
    Dana Dorsett | | #34

    Yes, there is a wealth of ignorance out there...
    ... amongst builders, inspectors, and (apparently) code writers too.

    But I would presume that drafters of IECC codes prescriptions would have a handle on how to calculate the U-factors at typical framing factors, or use other energy use simulation tools (as is clearly going on in the mass-wall prescriptives) to come up with codes that make at least some sense, given that financially rational energy conservation is their mission.

    It's arguable that architects & building designers SHOULD be able to figure this out, but it's clear those that can are the exception. I have no hope of turning builders or inspectors into engineers, which is why the codes are so simplified.

    IIRC the energy aspects of building codes in Sweden are performance based rather than a simplified list of prescriptive-Rs, with penalties assessed on the architects &/or builders for underperforming buildings (depending on whether it's found to be a design vs. implementation issue). You can bet THOSE architects & builders can all run U-factor calculations in their sleep, (and don't get much sleep if a design under construction is too close to the line.)

    The simplification of using cavity-R for ~25% framing fraction walls as the shorthand vernacular for wall performance is part of what lets ICF & SIP manufacturers get away with using outlandish performance numbers. An R13 wall is really an R9-R10 wall, and R19 wall is really an R12-R13 wall when comparing performance to continuous insulation assemblies. It kinda matters that (almost) NOBODY would choose to build a 2 x 12 studwall with a 25% framing fraction for an "R40" wall (with R40 cavity fill), so it's not a credible assembly for comparing with ICF or SIP construction.

  35. anderslewendal | | #35

    R20+5 insulation code
    Lecole asked about using closed cell foam in the wall cavities with closed cell foam panels outside of the sheathing. As Mr. Holladay pointed out, that is a bad idea. You could, however, put 1" or 2" of closed cell foam panels on the outside of the studs and then spray closed cell foam against the foam panels giving you a solid 5" to 7" of insulation. You might ask where the shear wall goes. If your home is fairly simple, you can save 50% to 75% of the shear wall materials if you ask a structural engineer to spec out shear walls nailed on the exterior wall studs from the inside. I have done this a few times with very good results. Plus, no issues with dew points. I am also in zone 6. Some would call it outsulation. Good luck.

  36. Katy_Hollbacher | | #36

    Disaster waiting to happen
    Another timely post—I was shocked when I learned about the new code and have been bracing myself for the fallout that's going to happen. Some of our projects are in CZ 6 & 7, and the fact that code not only allows but requires exterior insulation along with interior VBs as before is true insanity! I already know of some projects that have installed 1-2" exterior XPS, with R-21 batts, in a climate with average winter temps in the teens. SCARY.

    Martin, I do have a question re: "...the code needs to distinguish rigid foam products that have a low vapor permeance (foil-faced polyiso, XPS, and EPS) from mineral wool panels that have a high vapor permeance..." Absolutely, and this was my first thought when I heard the code details—for the code not to take it a tiny step further, and clarify the TYPE of exterior insulation allowed, is just unbelievable.

    But my question is, do you really consider EPS to have "low vapor permeance" at modest, most common thicknesses of 1 or 1.5" (and their associated perm ratings of 5 or 3)? That's a far cry from what I consider to be low permeance (<1 perm) xps or vapor barrier polyiso, and i generally wouldn't expect to see issues with using only 1 1.5" eps even in very cold climates.
    Also, something I'd like to hear more about is how the inclusion of what I like to call "quasi-rainscreen" functionality between the drainage plane and back of rigid insulation impacts condensation risks. Grooved EPS, Tyvek StuccoWrap, HomeSlicker kind of thing (ie, an approach that provides some kind of air movement & related drying potential, but not so much that it ruins the thermal performance of the insulation). I have yet to come across a thorough study that has looked at this in-depth. Anyone?

  37. user-1096360 | | #37

    Excellent discussion, but...
    I'm a bit surprised no one has mentioned one solution would be to remove the organic material layer that can/will potentially mold (OSB or plywood). Why couldn't the code state that if exterior continuous impervious insulation is used, wood structural sheathing would not be allowed. Hence, older traditional forms of wall bracing could be used - i.e. let in bracing. I know this still opens up the opportunity for potential condensation at the wall stud, but won't it be very minimal or possibly non existent?

    My fear is for a builder like myself that uses a 1" continuos exterior layer of polyiso (R Max - formerly Dow SIS) and CC flash & BIB cavity, that now my method will become 'illegal' with the new code. I essentially have 3-1/2" of CC foam on the exterior of the wall with this system, but could potentially be non-conforming?!

    Something tells me a few PHIUS folks are reading through this discussion and laughing histerically.

  38. ErgoDesk | | #38

    Mold and Wood
    Insistence on the use of wood limits the solutions, if there were any in this situation. There is no help for those who won't listen to the solution.

  39. GBA Editor
    Martin Holladay | | #39

    Response to Anders Lewendal (Comment #35)
    Your suggestion -- that walls without exterior OSB, braced by alternate means, could safely include R-5 exterior rigid foam and R-20 spray foam between the studs -- is correct. It points to the fact that there are many ways to build a wall, complicating the task of writers of any prescriptive code.

  40. GBA Editor
    Martin Holladay | | #40

    Response to Katy Hollbacher (Comment #36)
    Q. “Do you really consider EPS to have ‘low vapor permeance’ at modest, most common thicknesses of 1 or 1.5 inch (and their associated perm ratings of 5 or 3)?”

    A. Depending on the type of EPS we are talking about, the permeance of 1 inch of EPS ranges from about 2.0 to 5.8 perms. A sample that measures 1.5 or 2 inches thick would have a lower permeance, of course.

    I'll rephrase the question: "Assuming that EPS has an R-value of R-4 per inch, and that a builder chooses to comply with the 20+5 code provision by installing 1.5 inch of EPS on the exterior side of a wall, would the EPS be permeable enough to keep the wall out of trouble?"

    That's a difficult question to answer without either (a) using and trusting the results of WUFI, or (b) building 100 houses in a cold climate and seeing what happens. As a builder, I wouldn't trust outward diffusion through the foam layer to provide enough drying every March and April to make up for the dependable wetting occurring every February.

  41. GBA Editor
    Martin Holladay | | #41

    Response to Ryan McCoon (Comment #37)
    Like Anders Lewendal, you have suggested that damp OSB can be avoided by building a wall without exterior sheathing, instead using alternate bracing methods. That method will work, as long as the builder doesn't attempt to comply with the 20+5 provision of the code by installing R-20 fluffy insulation between the studs and R-5 of rigid foam on the exterior side of the studs.

    The result of the scenario I describe is that condensation will occur on the interior face of the rigid foam sheathing. The moisture will either saturate the outer layer of fluffy insulation, or will run down the face of the foam and form puddles on the bottom plate.

    You are avoiding that scenario by installing 2.5 inches of closed-cell spray foam (rather than fluffy insulation) on the interior side of the 1-inch thick polyisocyanurate -- a similar approach to the one suggested by Anders. That approach will work. I agree with both you and Anders that whatever solution is arrived at by code writers, their solution should allow the type of wall assembly you describe.

    I'm not sure exactly what you meant when you wrote, "Something tells me a few PHIUS folks are reading through this discussion and laughing hysterically." Perhaps you mean that PHIUS has always advocated for walls with a higher R-value than the values included in the prescriptive code, or that PHIUS (unlike you) favors walls without much rigid foam or spray foam. If that's what you meant, you're right.

    Like PHIUS, GBA has always recommended that builders choose walls with a high R-value, and has also emphasized that any cold-climate wall assembly must avoid the kind of moisture accumulation that is possible in a 20+5 wall.

    But editors at GBA aren't laughing hysterically at the 20+5 problem, even though GBA readers are unlikely to be tripped up by minimum code provisions. I think that energy codes should make sense, so that even builders who aim to barely meet the code should end up with safe walls.

  42. GBA Editor
    Martin Holladay | | #42

    Response to George Hawirko (Comment #38)
    You wrote, "Insistence on the use of wood limits the solutions."

    You're right, of course. The 20+5 provision in the building code assumes that the wall under discussion is framed with 2x6 lumber. If a builder chooses (for example) to build an ICF wall, then the problems discussed in my article won't occur.

    One solution -- perhaps one that you would support -- is to outlaw walls that are framed with lumber. Since most homes in the U.S. have 2x4 or 2x6 walls, however, that suggestion is unlikely to gain much traction.

  43. wjrobinson | | #43

    It's not complicated for the
    It's not complicated for the code to have one prescriptive plan set as one example of a buildable home. This one set of allowed details would be in addition to the codes that already exist. Anyone could just attach the standard allowed detail sheets to their plan set. Done. Easy.

  44. user-626934 | | #44

    Prescriptive vs. Performance
    Building off of Dana's comment regarding Sweden - the direction our energy codes need to move is toward performance-based standards rather than prescriptive. Presumably, the goal of the International ENERGY CONSERVATION code (IECC) should be to conserve energy. Prescriptive requirements will point a project generally in the direction of energy conservation, but they can't overcome certain design decisions such as 40% window to floor area ratio or very poor geometrical efficiency in a cold climate. These kinds of design decisions will still lead to excessive energy use, even with relatively stringent prescriptive requirements for the various assemblies.

    Fortunately, the IECC is already headed in the direction of whole-house performance with the inclusion of an Energy Rating Index option (for instance, RESNET HERS Rating) in the 2015 code. You can read more about it here - While I have some issues with using the HERS Index for code compliance - for instance, it's a lot easier to get low HERS score on a house with a conditioned basement, compared to a similar house that is slab on grade - I think this is a huge step in the right direction.

  45. Katy_Hollbacher | | #45

    Impacts of rainscreen/rear-vented insulation
    Following up from my own question, which was essentially: Exactly how much does rainscreen functionality, when used in an assembly that includes exterior insulation, impact sheathing condensation risks? (and following from that, how should this information be incorporated into related building codes, types and levels of insulation allowed or req'd, etc.?)

    The following papers/reports are compelling, but what I am really interested in seeing is the same types of studies applied to constructions that include various types and thicknesses of rigid exterior insulation. I don't see how this particular debate can be had without considering this topic.

    The Effect of Air Cavity Convection on the Wetting and Drying Behavior of Wood-Frame Walls Using a Multi-Physics Approach

    Ventilated Wall Claddings: Review, Field Performance, and Hygrothermal Modeling

  46. GBA Editor
    MIKE GUERTIN | | #46

    Change the Code - 2018 next opportunity
    I got a copy of the 2015 IRC a couple weeks back. Just checked the vapor retarder and insulation tables. Pretty much the same as the 2012 IRC.

    Next opportunity for code changes will be the 2018 code hearings. Here's the ICC codes and standards calendar for the next few years:

    Given that many jurisdictions haven't even adopted the 2012 codes, it may be a long time before good building science based prescriptive insulation and vapor-control requirements make their way into common practice - unfortunately.

  47. GBA Editor
    Martin Holladay | | #47

    Response to Mike Guertin
    I agree that the three-year code cycle for updates and the cumbersome code change process means that we are going to stuck with these provisions in the International codes for quite some time.

    It is nevertheless worth mentioning that the IRC and the IECC aren't legally binding on any jurisdiction in the country until they are adopted by a local authority. They are what used to be called "model" codes -- that is, they are models or suggestions for local authorities to consider when establishing local codes.

    Any jurisdiction in the country has the option of adopting parts of the International codes while rejecting or modifying other parts of the code. As far as I understand, that's what's happening in Vermont right now. The Department of Public Service is considering the 2015 codes; if the authorities want to, they can adopt a modified version of the code.

  48. GBA Editor
    Martin Holladay | | #48

    Response to Katy Hollbacher (Comment #45)
    We certainly know that ventilated rainscreen gaps accelerate the outward drying of damp wall assemblies. Whether this accelerated drying is enough to tip a risky wall assembly with thin exterior rigid foam over the line -- making it an acceptable wall assembly -- is unknown.

    One thing is for sure: if you are installing foil-faced polyisocyanurate, the rainscreen gap isn't going to accelerate outward drying of the wall assembly (although it will have other moisture management benefits). If you have EPS, however, it might.

    While WUFI can provide an answer to your question, it's up the each builder or architect to decide whether WUFI results are enough to hang your professional reputation on. (For more information on this topic, see WUFI Is Driving Me Crazy.)

    My own decision is as follows: I'm going to continue to follow Joe Lstiburek's recommendations for minimum exterior rigid foam thickness, rather than adopting thinner foam in hopes that the rainscreen gap keeps me out of trouble.

    Like you, I would welcome more field studies that look into this question.

  49. eingleseboi | | #49

    Possible Solution

    I enjoyed reading this article, and think you’ve provided valuable information about the complexities of code and the issues that can arise from lack of understanding and misinterpretation. We are seeing an increase in questions related to assemblies with exterior rigid foam and our end users have concerns about the moisture content of this type of system.

    What are your thoughts regarding a wall assembly using a drainable-type, vapor permeable housewrap overtop of sheathing, and underneath exterior rigid foam?

    This would help to let trapped moisture escape while still maintaining the R-value of the foam. Would this type of assembly help to solve the moisture concerns associated with exterior rigid foam?

  50. GBA Editor
    Martin Holladay | | #50

    Response to Elyse Inglese
    As Joe Lstiburek explained in one of his articles (Mind the Gap, Eh!), a crinkly housewrap can provide some hygric redistribution, although at a small loss in thermal performance. This hygric reidistribution is especially important when the insulation between the studs (for example, closed-cell spray polyurethane foam) doesn't allow any inward drying.

    You are proposing another use of crinkly housewrap. It sounds as if you are proposing that builders might install crinkly housewrap in hopes that the hygric redistribution will tip a risky wall assembly (one with with too-thin foam) into safer territory. This proposal is similar in principle to the proposal made by Katy Hollbacher in Comment #45.

    Count me a skeptic -- until someone conducts a field study that verifies the idea. I don't think it's worth violating the rules that have been established concerning minimum foam thickness -- even when we include a few tricks that might lower the risks.

  51. Katy_Hollbacher | | #51

    Response to Martin re: rainscreen impacts
    Martin, to be clear, I have not proposed anything except the question, "Exactly how much does rainscreen functionality, when used in an assembly that includes exterior insulation, impact sheathing condensation risks?"

  52. Bond | | #52

    Performance vs Prescriptive
    When I left the UK, Scotland had just moved to a largely performance based building code that was a fraction of the length of the previous version. However, at the same time England introduced acoustic requirements. Builders were allowed a choice: follow one of the multiple prescriptive options (and pay a small fee to license the detail), or design themselves and have it reviewed. Choice is good!

    Interestingly my current 1981 home has 6" stud filled with glass fibre plus a 3/4" polystyrene under the stucco. In Northern Alberta we hit -40C, so I'm straining my head as to the best options to improve the performance, bearing in mind that building science is critical. Seems removing the insulation may be best!

    We are expecting the Energy Code to be ratified for Alberta this fall, I will have to see what's sought after here.

  53. JoeUSAF | | #53

    My builder has used engineered 2 x 6 wall studs for added strength in the new home which has an unvented cathedral ceiling. Our walls our actually R29, (no shrinkage of engineered lumber), and consist of vinyl siding, R5 foil faced polyiso, Tyvek, 1/2" plywood, 0.8lbs open cell spray foam, gypsum covered by permeable latex paint. This configuration will allow drying to the inside. Our blower air door test performed by a certified HERS rater was 0.29 ACH My question is: How risky is this configuration concerning sheathing rot and does the Tyvek installed interior to the polyiso aid in reducing condensation on the plywood sheathing. JOE in CT

  54. GBA Editor
    Martin Holladay | | #54

    Response to Joseph Poland
    You didn't tell me your climate zone, but you signed your comment as "Joe in CT," so I'm guessing that your house is in Climate Zone 5.

    The minimum R-value for rigid foam installed on the exterior side of wall sheathing in your climate zone for a 2x6 wall is R-7.5. This is explained in the following article: Calculating the Minimum Thickness of Rigid Foam Sheathing.

    You have chosen to install R-5 foam, which is too thin. Moreover, polyiso has an additional problem: in cold weather, it doesn't perform as well as its label indicates. For more information on this issue, see In Cold Climates, R-5 Foam Beats R-6.

    Q. "How risky is this configuration?"

    A. Well, you've broken the rules, which is never a good thing. The best approach at this point is to strive to keep your indoor humidity as low as possible during the winter. Monitor the RH with a hygrometer, and operate your ventilation system as necessary to keep the RH low.

    Q. "Does the Tyvek installed interior to the polyiso aid in reducing condensation on the plywood sheathing?"

    A. No. The Tyvek is a water-resistive barrier (WRB). The primary purpose of a WRB is to resist wind-driven rain.

  55. miroland | | #55

    ZIP Exterior and Interior Vapor Retarder w/ Cellulose?
    I know this string is about Code deficiencies, but I'd just like to confirm a related wall assembly question if I could. Double-stud with cellulose only will avoid the problem of trapped moisture and condensation against inadequate exterior foam, but would there be a problem against ZIP sheathing? Is it permeable enough? Also, am I correct that the IRC requires an internal vapor retarder, but that latex paint is acceptable for this purpose in the absence of some sort of sheeting or craft paper?

  56. GBA Editor
    Martin Holladay | | #56

    Response to Michael Roland
    Here is how building scientist John Straube has answered this question:

    "The vapor permeance of Huber ZIP is in the same range as commodity OSBs. OSBs that we test have quite a range of wet-cup vapor permeances, and both roof and wall ZIP are essentially the same.
    But ZIP, like all OSB, does not have the high permeability of a building paper, so one needs to design carefully. For best practice, that means insulation on the exterior to warm it up and to blunt thermal bridges at mudsills and floor joists. If you use a double stud wall, then I am concerned, but I have a low risk threshold since I am a forensic consultant too. :)

    "The back of the ZIP has a different texture because of the way the product is pressed, and you will notice that other OSB is like this too. There is no difference in performance as far as we have been able to see."

    Recent versions of the IRC only allow the use of latex paint as a vapor retarder if your wall has an adequate thickness of exterior rigid foam. Image #2 (reproduced on this page at the bottom of my blog) shows the table in the code that lists minimum R-values for exterior foam when Class III vapor retarders (like latex paint) are permitted.

    To read more about code requirements for vapor barriers, see Vapor Retarders and Vapor Barriers. In that article, I wrote:

    "The 2007 Supplement to the IECC and the 2007 Supplement to the International Residential Code (IRC) introduced a new vapor-retarder definition. (Of course, many jurisdictions in the U.S. are still using local codes based on the 2006 — or even earlier versions — of the IRC and IECC.) Vapor retarders are now separated into three classes:

    Class I: Less than or equal to 0.1 perm [e.g., polyethylene];
    Class II: Greater than 0.1 perm but less than or equal to 1.0 perm [e.g., kraft facing];
    Class III: Greater than 1.0 perm but less than or equal to 10 perm [e.g., latex paint].

    "Since 2007, the IECC has required (in section 402.5) that walls in climate zones 5 (e.g., Nevada, Ohio, Massachusetts), 6 (e.g., Vermont, Montana), 7 (e.g., northern Minnesota), 8 (e.g., northern Alaska), and marine zone 4 (Western Washington and Oregon) have a Class I or Class II vapor retarder — in other words, kraft facing or polyethylene.

    "... Further exceptions are allowed in section 402.5.1, which states that in climate zones where a Class I or Class II vapor retarder would normally be required, a less stringent vapor retarder — a Class III retarder like latex paint — can be used under the conditions listed in Table 402.5.1. ... Only certain types of wall assemblies are worthy of this exception; they must have either an adequate layer of exterior foam sheathing or “vented cladding.” "

  57. StLHandyMan | | #57

    Political Battles
    "The short answer is that the code was not written by building scientists; its provisions are historical accidents resulting from political battles and compromises."

    I haven't seen our Politicians get much of anything right for quite awhile now. Why would they start or help make "Better Buildings, that are energy efficient"...#imo its because Big Business dominates our political process-dictating what is best for them!

    Good Article and thanks again for the tips.

    Build Green,

  58. Ken Levenson | | #58

    something isn't clear
    In this post, all about the code requirements for "continuous insulation", you speak virtually exclusively of "rigid foam" code requirements. While you acknowledge that mineral wool fits the code too, you appear to say that there is no need to take it into consideration - leaving the discussion entirely about rigid foam - because, as you write above, "There is no need for exterior mineral wool insulation to adhere to minimum R-value requirements." But is that true? It's not clear to me. While the mineral wool is vapor open and allows drying outward where the foam does not so readily, isn't the problem location the first potential condensing surface? Isn't that surface, without sufficient exterior insulation, the interior face of the OSB or plywood sheathing? OSB and plywood in winter are great vapor retarders and exfiltration/humid air hitting that interior face, cold surface, whether or not there is vapor open insulation outboard of it, would seem to posse a real risk of condensation and attendant problems. If this is true then mineral wool thickness does matter and we shouldn't be solely focused on rigid foam. Yes? No? Curious....

  59. GBA Editor
    Martin Holladay | | #59

    Response to Ken Levenson
    Your question is a good one, and it's one that GBA has covered extensively. This is the "cold OSB" question.

    You can read about the cold OSB question in these two articles:

    How Risky Is Cold OSB Wall Sheathing?

    Monitoring Moisture Levels in Double-Stud Walls

    You are correct that exterior sheathing gets cold (and therefore damp) in February unless it is protected on the exterior by an adequate thickness of insulation.

    Most U.S. homes suffer from the cold OSB problem (or the cold plywood problem), and most have walls that aren't rotting, because the walls are able to dry to the exterior in April. Even though OSB and plywood are vapor retarders when dry, they become more vapor open when damp -- in essence, they are "smart" vapor retarders.

    Some types of cladding -- stucco in particular -- aren't vapor-permeable enough to allow much outward drying, and there are many reports of OSB rot in stucco-clad homes.

  60. GBA Editor
    Martin Holladay | | #60

    Response to Ken Levenson
    Your prediction ("As homes get tighter, per new codes or otherwise, the indoor RH will rise significantly") is interesting. It's possible that you are right, and that the interior RH of homes will rise in the future. I'm not sure, though, since most energy-conscious builders include some type of mechanical ventilation system. I guess we'll have to await the data to see if your prediction comes true.

    I'm also a little confused by your tautology (or perhaps it is a contradiction?).

    You wrote, "As homes get tighter ... the indoor RH will rise significantly as will ... the risks. ... And airtightness & vapor control inboard of the the insulation is a robust way to mitigate those risks."

    So improved airtightness increases the risk, and one way to mitigate this risk is increased airtightness? Walk me through your argument one more time, Ken...

    [Postscript: After I posted this response to Ken, he went back to edit the comment I was responding to, correcting the confusing sentence.]

  61. Ken Levenson | | #61

    Response to Martin
    Thanks for the links - good to read again. So while the ASHRAE 160 humidity levels may have been too high in the double-stud post you link to, might the 30% RH be too low to draw conclusions? I ask because as homes get tighter (per new codes or otherwise) the indoor RH can rise significantly and could regularly be in the 50% RH range, which while below ASHRAE numbers presumably, are significantly above 30% RH range referred to in the double-stud article. This likely higher (but not extreme) RH range should be accounted for in dealing with cold sheathing.

  62. Ken Levenson | | #62

    you respond too quickly....
    I edited my comment above to refer to one of your links.

    But I'll respond briefly: airtightness, like many things in life can be beneficial but then also harmful, depending on many factors. Water is a good example.
    So in this case, airtightness helps in innumerable ways - controlling indoor air quality, comfort, heating/cooling energy efficiency, moisture protection from air leaks etc.... Part of the health/comfort benefit is often higher indoor RH because the cold dry air isn't blowing through the building. But with higher indoor RH we have higher risks associated with assembly condensation. Therefore, we like to see strong vapor and air control inboard of the insulation. Better?

  63. GBA Editor
    Martin Holladay | | #63

    Response to Ken Levenson
    I certainly agree with you that building assemblies need "strong air control." Whether the primary air barrier belongs on the interior or the exterior side of the insulation, or both, depends on the type of insulation that the designer has chosen.

    And not all wall assemblies need "strong vapor control inboard of the insulation." Some types of wall assemblies do; others don't.

  64. D Builder | | #64

    2x6 in Climate Zone 5
    Am I missing something here? In the Chicago area, we are in climate zone 5, and Table 402.1.1 of the 2012 IECC says that we can insulate our frame walls with R-20 cavity batts (in our 2x6 walls) or R-13 cavity batts with R-5 continuous insulation. A lot of people use the 2x6 with R-21 batts as prescribed by the code, but unless R-7.5 rigid is placed on the outside, you are pointing out that the inside face of sheathing will get wet. Isn't that a big problem?

  65. GBA Editor
    Martin Holladay | | #65

    Response to Dee Wilson
    You are right that 2x6 walls without any exterior foam often have damp OSB sheathing in February. The reason that most of these walls don't rot is that they are able to dry to the exterior in April.

    In some cases, however -- when the homes have stucco cladding, or when an ignorant builder has installed too-thin exterior foam (especially vapor-impermeable foam like foil-faced polyisocyanurate) -- the ability of the wall to dry to the exterior is reduced, and the OSB can rot. That's why thin rigid foam is riskier than thick rigid foam. (If the rigid foam is thick enough, the OSB never gets damp in the first place.)

    For further discussion of these issues, see How Risky Is Cold OSB Wall Sheathing?

  66. D Builder | | #66

    Thanks to Martin
    Thank you Martin. Ventilated exterior (brick veneer with air gap or vinyl siding), plywood instead of OSB, and cellulose instead of fiberglass batts. Reduce risk. Got it.

  67. lutro | | #67

    How continuous, and an R-Value typo?
    I first learned from a GBA article that 4x8 sheets of foam can expand and contract enough with temperature change, to leave up to a 1" gap between adjacent sheets. I followed the GBA links to foam manufacturer's websites, and found similar figures. That's a substantial thermal bridge. Yet in most articles, including this one, sheet foam outside the OSB is referred to as "continuous insulation". I would really like to understand this apparent contradiction. I've asked about this before, and Martin suggested that double layers of foam with offset seams would help. That makes sense, but the current article is focused on builders who will apply one layer of sheet foam outside the OSB, to meet a code requirement. Is the code making a faulty assumption that foam sheets constitute continuous insulation? Or are the expansion/contraction figures given by the foam manufacturers not accurate for this situation? Or do we lack data on the actual temperature cycle gaps between foam sheets that occur in houses insulated as described in this article?

    In comment 40, Martin writes, "Assuming that EPS has an R-value of R-1 per inch...". I'm thinking that this is a typo, since the usual values listed for EPS are in the neighborhood of R-4 per inch.

  68. GBA Editor
    Martin Holladay | | #68

    Response to Derek Roff
    Rigid foam is available with tongue-and-groove joints or shiplap joints, which reduces the problem you refer to somewhat. Two layers of rigid foam with staggered seams is the preferred approach.

    One inch shrinkage gaps are rare, but 1/2 inch to 3/4 inch gaps are more common. Even with this type of shrinkage, exterior rigid foam is responsible for a huge improvement in the performance of a wall.

    For more information on foam shrinkage, see Using Rigid Foam As a Water-Resistive Barrier. The relevant paragraphs are near the end of the article, beginning at the heading, "Do rigid foam panels shrink?"

    Thanks for catching my typo. I have corrected it.

  69. lutro | | #69

    Thanks for your response
    I appreciate your response, Martin, and the link to the earlier GBA article. This is important information for me.

  70. dknowlt | | #70

    Seems to me
    Should take another look at T-Ply. If not using vinyl and shear/racking resistance in the wall assembly can be augmented to the needed loading then why not? It won't rot, provides the air barrier and eliminates the moisture/rotting concerns in OSB.
    I'm considering a wall built on 2x6 plates with 2x4 studs, t-ply (red) sheathing and double studded to the inside later using nice straight light (non-struct) metal studs for a thermal break. If laid out so the metal-wood offset is slight (1/4"), 15-1/4" batts can seat against the web of the metal stud and butt up against the 2x4s - foam the wood-metal gap for thermal break,then fill in over the 2x4 with open cell. Outside corners can be assembled to permit cut batts to fit and provide a good continuous envelope. Inside corners, where columns are typically located are tougher but do-able with iso board and metal corner nose for GWB backer.
    In my case the house will have vertical spruce shiplap ext. so the backer blocking (2x3 flat to the t-ply) needed stiffens the wall and no need for what is at the core of the problem as I see it - OSB.
    OSB is just not the right material anymore.

  71. GBA Editor
    Martin Holladay | | #71

    Response to Daniel Knowlton
    Before reading your post, I had never heard of "T-Ply." After I did some Googling, I discovered that it appears to be a brand name for a type of plywood.

    Your suggestion to switch from OSB to plywood is one of the possible approaches I discussed in my other article on this topic, How Risky Is Cold OSB Wall Sheathing?

    In that article, I wrote: "OSB is more susceptible to rot than plywood. So if you’re worried about the durability of your sheathing, choose plywood (or diagonal board sheathing) over OSB. One other possible (permeable) sheathing choice is structural fiberboard sheathing, which is available from International Bildrite and Georgia-Pacific."

  72. Jon_R | | #72

    Not so simple
    Does code really not consider Membrain a Class I or II vapor barrier? If allowed, then 1.5" of EPS (pretty permeable) + Membrain makes "Zone 6 , 20+5, Class I or Class II" not always "nuts". Even more so when combined with other things that help: rain screens, roof overhangs, interior AND exterior air barriers, plywood (or no wood sheathing), cellulose, indoor/outdoor pressure differential control, and indoor humidity control.

  73. GBA Editor
    Martin Holladay | | #73

    Response to Jon R
    Of course it is possible to come up with fairly decent wall assemblies that meet the 2012 code requirements. If a builder uses mineral wool (or perhaps, as you suggest, EPS) instead of polyisocyanurate, or uses MemBrain instead of polyethylene, it's possible to design a wall that will work.

    However, the prescriptive code is written is such a way that it amounts to a trap for the unwary.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |