GBA Logo horizontal烈火情人下载最新章节免费阅读_烈火情人下载最新章节列表,西班牙面积最新章节_西班牙面积无弹窗,2021秋霞伦影院理午最新章节免费阅读_2021秋霞伦影院 FacebookLinkedInEmailPinterestTwitterInstagramYouTube IconNavigation Search IconMain Search IconVideo Play IconAudio Play IconHeadphones IconPlus IconMinus IconCheck IconPrint IconPicture iconSingle Arrow IconDouble Arrow IconHamburger IconTV IconClose IconSortedHamburger/Search Icon
Musings of an Energy Nerd
Five Cathedral Ceilings That Work

As long as you pay attention to minimum code requirements for R-values, these cathedral ceiling details can be used in all climate zones

A vented assembly with fiberglass or mineral wool insulation. This is the traditional way to insulate a cathedral ceiling. Note, however, that this approach requires deep rafters or scabbed-on 2x4s to add depth. The illustration shows an assembly with R-38 insulation, which is enough insulation for Climate Zones 1, 2, or 3. If you live in a colder climate zone, you'll probably need insulation with a higher R-value.

When GBA readers post questions about cathedral ceiling insulation, they are often directed to a 3,800-word article from 2011 called “How to Build an Insulated Cathedral Ceiling.” While the comprehensive information in that article—including detailed discussions of vapor diffusion, condensation risks, and failure modes—is useful, it may represent information overload for many builders.

This article will cut to the chase by providing simple descriptions of five cathedral ceilings that work. While these five assemblies are probably the most common approaches used, note that there are other acceptable ways to insulate a cathedral ceiling.

R-value minimums. In Climate Zones 4, 5, 6, 7, and 8, most prescriptive building codes require ceiling insulation to have a minimum R-value of R-49. In Zones 2 and 3, the minimum requirement is R-38, while in Zone 1, it’s R-30. (Here is a link to a climate zone map.)

That said, local codes may differ from these general guidelines, so it’s worth asking your local building department about minimum R-value requirements in your community.

For more information on minimum R-values for cathedral ceilings, see “Three Code-Approved Tricks for Reducing Insulation Thickness.”

Rafter depth. Note that 2×10 rafters equipped with a 1-inch-deep ventilation space have room for only 8 inches of insulation. If you are insulating with fiberglass, that’s not enough space for the minimum R-values required by building codes unless you live in Climate Zone 1. So if you hope to insulate your ceiling with a fluffy insulation material like fiberglass batts, and you plan to install all of the insulation between your rafters, you’ll either need very deep rafters, or you’ll need to scab on some extra framing material (for example, 2x4s) to your rafters with plywood gussets to get enough depth for the minimum insulation levels required by building codes.

烈火情人下载最新章节免费阅读_烈火情人下载最新章节列表,西班牙面积最新章节_西班牙面积无弹窗,2021秋霞伦影院理午最新章节免费阅读_2021秋霞伦影院 Air sealing. Most cathedral ceiling moisture problems can…

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 andconstruction details.

Start Free Trial


  1. User avater
    Peter Engle||#1

    Excellent article, Martin. Thanks for condensing this all into one concise article. It's already bookmarked on my list for client advice.

    FWIW, I've used all 5 of these approaches on my own projects. #1 is certainly the easiest to install and most familiar for builders. It is also the one I see screwed up most often. I'm seeing a lot more of #4 lately. Aside from being expensive, it is simple and that seems to be its biggest selling point. I usually see this one with dramatically less R-value than code because of the high cost of SPF. Somehow, it still gets sold/approved. I am ashamed to admit that I'm using it on a large condominium weatherization project right now....

  2. User avaterExpert Member
    Armando Cobo||#2

    Great summary, thank you. The challenge now is for folks to do a search instead of asking the same old question “How do I build a cathedral ceiling?”
    I don’t know if it’s worth noting that the 2021 IECC increased the insulation values by quite a bit. Perhaps a mentioned of this should be included, or in a short time, this blog will need to be updated.
    On option #5, it reads “R-5 in Zone 1, 2, or 4;” Should it be R-5 in Zone 1, 2, or 3?
    The sections where it says that open cell foam against the roof decking is a bad idea, it should be in CAPS, BOLD, and higher pica. Maybe folks won't miss it! I won't hold my breath... 😝

    1. User avaterGBA Editor
      Martin Holladay||#3

      Thanks for notifying me of the typo concerning the climate zones in Assembly #5. I have corrected the error.

  3. Joe Braun||#4

    I feel like there should have been a sixth roof. What about something like the cool vent by Hunter Panels? You have 5.5 inches fluffy underneath, deck board, WRB, ridged insulation, an air gap, another deck board, and then the shingles.

    Doing something like this would get to an R-60 (the 2021 code requirements) cheaper than closed-cell foam. It would also be easier to retrofit. You would have 8 inches of poly (R-40) on the roof and 5.5 inches of mineral wool (R-22.5) in the rafter. Something like this would also mean you don't need heels on the trusses.

    I like the idea because, in essence, it's an overroof. From what I have seen here and over at Building Science, this is the roof you need in areas that get a lot of snow and are cold.

    Is there some reason not to do this?

    1. User avater
      Peter Engle||#6


      The vent cavity above the rigid insulation in that assembly is sort of a belt-and-suspenders thing. With the vapor retarder at the lower deck level, there shouldn't be enough moisture present under the shingles to need additional ventilation. It doesn't hurt but it does cost more.

      1. Joe Braun||#12

        Hi Peter,

        Thank you for your reply. It is good to know. BTW, I had to look up what the term "belt-and-suspenders" meant. Based on the context, I think it is a great way to explain it.

        involving or employing multiple methods or procedures to achieve a desired result especially out of caution or fear of failure

    2. User avaterGBA Editor
      Martin Holladay||#7

      In the second paragraph of this article, I wrote, "While these five assemblies are probably the most common approaches used, note that there are other acceptable ways to insulate a cathedral ceiling."

      In your comment, you wrote, "I feel like there should have been a sixth roof." Well, why not a seventh and eighth? I decided to stop at five, to keep the article simple.

      Your suggested roof is similar to Assembly #3 on this page, except for the addition of ventilation channels above the upper roof sheathing. Note that the ventilation channels are unnecessary, as long as the thickness of the rigid foam is adequate for your climate zone. That said, there is some evidence that ventilation channels above the uppermost layer of roof sheathing can reduce ice dams, and if you live in an area plagued by ice dams, the expense of the roof assembly you favor may be justified.

      1. Joe Braun||#13

        1. Be nice.
        2. If you can't be nice, be polite.

        1. User avaterGBA Editor
          Martin Holladay||#15

          Sorry. No intent to cause offense. It's easily possible to list ten different ways to insulate a cathedral ceiling, and yours is one of them. I accept your point that perhaps my article could have benefited from a sixth roof assembly. Sometimes, however, it's hard to know where to stop when searching for the balance between simplicity and clarity on the one hand, and completeness on the other.

          This article was aiming for simplicity and clarity, and I'm sorry if it fell short.

    3. Shawn Batt||#22

      It's fine, but perhaps in zone 5 I would not be worried about the venting over top of the rigid material. To be honest, I also skim over ideas that use proprietary words for basic things. Sometimes the trademark wins (Xerox). In this case, Hshield and Coolvent just make me frown and move on.

  4. Expert Member
    Malcolm Taylor||#5


    Great stuff.

    Could you maybe help me out with a related issue. The foam closure in assemblies #1 and #2 is shown sealed to the framing to stop air movement. Baffles are also shown separating the batt insulation from the air-space. After reading RDH's study of the effect of wind-washing, I'm having doubts about whether either in necessary, or if there is no need for the closures or baffles except contain the insulation. Does wind-washing decrease the R-value of these assemblies enough to justify air-sealing batt insulation in roofs?

    1. User avaterGBA Editor
      Martin Holladay||#8

      I've always advocated in favor of air barriers to reduce wind-washing near soffit vents, especially when builders insulate with fiberglass batts. For similar reasons, I think that ventilation baffles should be installed with attention to airtightness.

      Some research findings show that the labor to create airtight details in these areas isn't worth the trouble. Fair enough. It's possible that most builders lack the resources or time to focus on nitpicky air-sealing details in these areas; that the energy saved is too little to justify the labor; and that the type of meticulous attention I describe is better spent on other aspects of the building's air barrier. If you decide to make do with simpler details, I understand the decision.

      But I don't like the idea of the wind blowing through fiberglass batts. It just seems wrong.

    2. Tyler Keniston||#10

      Malcolm, do you have a link to that study? Is it the 2016 one? Isn't that regarding rigid mineral board (I didn't read it through; perhaps it gives mention to fiberglass cavity fill?)?

      1. Expert Member
        Malcolm Taylor||#11



        See tables 8 and 9.

        I don't quite know how to think about wind-washing. That's something I'm looking for help with. If you hire an insulation company here, their default method of insulating above the exterior walls is to fold over a batt and block the cavity below the baffle with it. How much of en energy penalty does that cause over air-sealed blocks? Similarly, cathedral ceilings almost never have baffles if there is sufficient height to allow a 3" air-gap to the sheathing. Installing continuous baffles in each bay is a significant task. Is it worth it?

        1. Tyler Keniston||#14

          Thanks. I'll have to look more into it.

          I'm being a bit lazy, in that I still didn't read the entire study, but above the table it says: "Two fibreglass batt samples (low density and high air permeability) were used as products with known susceptibility to wind washing to confirm the ability of the apparatus to measure wind washing effects." (Suggesting even the author feels wind-washing is a concern with FB).

          In table 9, it appears that an R-13 batt is derated R-1 at wind speeds of 3ft/s. I'm not sure that's 'insignificant,' but it depends.

          I also wonder if wind speeds might tend to be in the high range at that soffit location vs behind cladding.

          All that said, I take your point about whether it is worth it. I really have no idea, and could see how it may not be.

  5. Tom Wheeler||#9

    It does keep the air coming out of the ridge vent cleaner after flowing through the filter, I mean insulation.

  6. A2ZED||#16


    Thanks so much for this article. Not having been a member until today I had read your original article from 2010 and had some questions so I drew a diagram of my proposed combination of rigid under shingle and batt between rafters with a view to posting in the comments section. To my delight after joining I found this updated article which answered most of my questions. I still have a couple of remaining questions. In the lower sections where I am not adding rigid insulation beneath the shingle would you recommend installing roof and soffit vents to create an independent vented section of the attic that is separate from the unvented part above?

    Thank you in advance.


    File format
  7. User avaterGBA Editor
    Martin Holladay||#17

    Your sketch shows a kneewall. The best way to address this issue is to use the same insulation method on both sides of the kneewall -- in other words, to continue insulating the slope of the roofline all the way down to the eaves.

    This article explains more: "Insulating Behind Kneewalls."

    If you ignore my advice, and try to create a vented mini-attic behind the kneewall, you can. If you do that, I would opt for soffit vents and triangular end-wall vents in the cramped attic.

    That said, you'll end up with complicated air-sealing details if you do it that way.

    1. A2ZED||#20

      Thanks so much for your prompt and informative reply, Martin! Much appreciated.

      I for sure do not want to "ignore" your advice. But unfortunately it may come down to budget. To continue the insulation over all the roof doubles the price and takes me over budget. There is only one room that would reap the benefits of the better way to do it, if it were all the upstairs rooms I would think differently. I will speak with my roofer and engineer tomorrow.

      There is currently zero insulation so anything is going to be better than what is currently there!

      Thanks again!

      1. CarsonB||#27

        For a 12/12 it would only be about %40 compared to insulating the kneewalls. I take it that the other 60% would be because of needing to meet code minimum roof vs wall insulation? That does seem like a code annoyance... it may make more sense to drop insulation across the whole roof. For a vaulted ceiling, you may be able to get away with r38 depending on your code.

  8. Eddy Luttmer||#18

    There's this newer HFO blown closed cell foam with nominal R7/in. If applied to a depth of 7-in. for R49 (CZ5B, reportedly less is needed the thicker it gets) against the roof sheathing ( a 4:12 pitch) it will envelop the upper chord and a few inches of the web members of the wood roof trusses. I have worried a bit about the impermeable underlayment and impermeable foam trapping moisture in the sheathing if it gets there in the first place. Roofs leak. There is limited experience with the HFO blown foam. The wood truss is a thermal bridge. Do I need to lose sleep over this planned assembly?

    1. User avaterGBA Editor
      Martin Holladay||#19

      The following article should answer your questions: "Sandwiching Roof Sheathing Between Two Impermeable Layers."

      1. Eddy Luttmer||#21

        Thank you. I am breathing easier.

  9. Margie Lynch||#23

    Thanks for this information, it's timely for me. It seems like folks trying to minimize use of foam will end up with Assembly #1. What does an R-60+ cathedral roof assembly end up looking like? Is this a situation where inclusion of some rigid or spray foam is preferred for cost and practicality reasons?

    1. CarsonB||#26

      Depends on which batts you were using. For r4 mineral wool you would need at least 15 inches of insulation. So with ventilation baffles, you are looking at 16 - 18” ijoists to achieve r60. My understanding is foam could get you a smaller and safer assembly, but likely with higher labor and materials.

    2. Tom Wheeler||#32

      I have 2 x 12 rafters with 2 x 2's on top of them from the original roof system. Then site built 1/2 plywood baffles leaving a 1 1/2" gap. Then R30 rockwool, r15 rockwool, intello, perpendicular 2x4's on edge with r15 rockwool. Haven't decided on the final layer inside.

      At ~22' peak, I wasn't too worried about the additional depth of the 2x4s.

  10. User avaterExpert Member
    Armando Cobo||#24

    I’m designing at a new house in CZ4, with R60 cathedral ceiling, and I’m thinking of using 3” R18 min. polyiso rigid foam on top of the roof decking and 2x12 rafter with 11.25” R40 DP cellulose, or 11 7/8 TJI rafters with R42 min. DP Cellulose.

  11. CarsonB||#25

    Thank you for this article Martin. For #1, does the validity change when dense pack fiberglass is used instead of batts, and the “baffle” is a layer of osb with 2x4s laid on top? There was some debate if this had enough drying potential. Reading GBA makes me far more worried about these assemblies than my builder or insulation contractor seem to be. I feel a bit like someone reading webmd, and then staying up at night worried my roof will have cancer.

    1. User avaterGBA Editor
      Martin Holladay||#28

      I'm not sure whether the "baffle" you are talking about is a continuous layer of sheathing above the rafters -- in other words, the roof sheathing -- or narrow rectangles of OSB inserted between the rafters in the manner of conventional ventilation baffles.

      If you are thinking of cutting OSB into narrow strips, and using those narrow strips as ventilation baffles, don't worry. The ventilation channels will keep your roof sheathing dry.

      If you are thinking of establishing ventilation channels above a continuous layer of roof sheathing, you should read my article, "How to Build an Insulated Cathedral Ceiling." Scroll down to the section with the subheading, "Creating vent channels above the roof sheathing," and read the paragraphs that follow.

      1. CarsonB||#29

        continuous sheathing, albeit there would then be another layer of roof sheathing on top of the 2x4s for shingles. Why is a continuous layer of osb more dangerous than one between the rafters? Is it the increased likelihood of air leakage from the "baffles" that makes it a superior assembly? Or perhaps that you can get away with thinner osb? Why would you need a WRB applied to the osb sheathing but not the osb baffles? In my mind they are exactly the same assembly, one just uses additional 2x4s and one uses the existing 2x4s from the ijoists.

        1. Expert Member
          Malcolm Taylor||#30


          That's why I'm not sold on OSB or plywood as baffles. Any moisture that makes its way up through the insulation has to then move through the OSB or roof framing members by diffusion before being vented to the outside.

          On a roof with the ventilation cavity over the sheathing, all the moisture has to diffuse through the plywood or OSB that way.

          To my mind, much more resilient is an assembly where the vent channel is not separated from the insulation, or is separated by something high perm like say house-wrap to reduce wind-washing and contain the insulation.

          1. CarsonB||#31

            thanks for the reply Malcolm. So instead of: sheathing over the framing with osb, adding membrane, then adding 2x4 on top of the sheathing where the rafters are, then adding another layer of osb for roof sheathing; could one just cover the framing with taped tyvek, run 2x4s on top of where the rafters are, then the roof sheathing? This is to avoid losing the insulation depth and having to fasten some sort of baffle like assembly for the house-wrap.

          2. User avaterGBA Editor
            Martin Holladay||#34

            The continuous sheathing above the rafters (in your case, OSB) has several purposes, one of which is structural. The sheathing adds strength to the roof assembly, and a layer of housewrap won't provide that function. So if you want to pursue your idea, you'll definitely need to talk to an engineer.

            There's another problem with the idea, and it concerns buildability. It's very hard to stand on unsheathed roof trusses or rafters while wrestling with a large roll of housewrap, with the goal of fastening the housewrap to the trusses or rafters -- even on a day without any wind. And all those difficulties happen even before you attempt to tape the housewrap seams.

          3. CarsonB||#33

            my insulation contractor recently recommended #5, but without any sort of interior air barrier. They want to flash the roof sheathing, dense pack fiberglass bibs into the bays, and then wooden T&G directly over the blown in netting. My concern is interior air getting in and condensing, but then again I have a hard time getting my head around the issues with 4" layer of closed cell spray foam. It would seem that the spray foam would avoid any surface reaching dew point and seal any air leaks through the sheathing, which would appear to negate moisture and exfiltration concerns? My builder was concerned that exterior rigid foam would add too much labor compared to the flash and batt/pack method.

          4. User avaterGBA Editor
            Martin Holladay||#35

            In a flash-and-batt assembly, the closed-cell spray foam is the air barrier, and, if the contractor knows what he or she is doing, the spray foam layer is thick enough to keep the interior surface of the cured spray foam above the dew point in winter. So (assuming the spray foam job is flawless) you don't have to worry about air leakage or condensation.

            An interior air barrier between the tongue-and-groove boards and the fiberglass insulation isn't necessary if the contractor does a perfect job of installing the spray foam, but such an air barrier might be useful if the spray foam job is sloppy and there are hidden air paths between the interior and the ridge.

            For a real-world example of a sloppy spray foam job, see this article: "What Is My Roof Trying to Tell Me?"

          5. Expert Member
            Malcolm Taylor||#38


            475 Performance Building suggests a similar assembly. As Martin points out the problem is safely building it. I wouldn't want to tape that house-wrap, or work on a roof with only membrane covering the framing while I installed the 2"x4"s on the top of the rafters. The finished assembly would work very well though.

          6. CarsonB||#40

            malcolm, do you have a link to the 475 assembly? I was able to find this article: /blog/foam-free-der-roof-assemblies/, but it is for retrofits and doesn't seem to match what you were recommending.

          7. CarsonB||#41

            Thanks Martin, the foam failure was a very timely article. This issue though is if the interior air barrier is *typically* redundant, intello membrane on the bottom adds thousands of dollars which is expensive insurance. For buildability then if I follow correctly, the baffles for a cathedral ceiling, whether housewrap or thin fiberboard, are put in from below after the top sheathing goes on?

          8. User avaterGBA Editor
            Martin Holladay||#42

            Q. "The baffles for a cathedral ceiling, whether housewrap or thin fiberboard, are put in from below after the top sheathing goes on?"

            A. You've basically got it right, except for the fact that there is no "top sheathing" with a vented roof assembly, because a vented roof generally only has one layer of roof sheathing, not two layers. (Two layers of sheathing usually imply that you have an unvented assembly.) If the idea of ventilation baffles is new to you, you should read this article: "Site-Built Ventilation Baffles for Roofs."

          9. Expert Member
            Malcolm Taylor||#44


            I can't find the pictures, which I'm pretty sure were here 0n GBA, but you can see a similar assembly in this 475 video at 40 seconds and 1.06.

          10. Tom Wheeler||#45

            Just thought I would toss a picture out of a partial finished ceiling with site made baffles. I used 1/2 plywood and 1 1/2" thick eps strips.

            The second picture is almost complete. I plan to air seal it, even if some say it is a waste.

          11. CarsonB||#46

            you guys are awesome, thank you. The Mento plus does seem like a great idea to skip the extra sheathing or baffles, but Martin is a former roofer so he likely has a good nose for the labor involved.

          12. Expert Member
            Malcolm Taylor||#48


            If it was something like a single storey 4/12 roof with a simple shape I might be tempted to try it. I'd probably work in eight foot increments, stapling the underlayment, nailing the battens, and running two courses of sheathing. Wandering around on a few 2"x4's the way they are seems a bit sketchy.

          13. CarsonB||#55

            " the closed-cell spray foam is the air barrier". So out of these assemblies, only #1 and #2 would truly require an interior air barrier, with #2 the interior air barrier likely being the interior side foam. For #3, 4, and 5, the exterior foam boards or spray foam would be sufficient to prevent condensation issues, so interior air sealing is not required, only an extra precaution.

          14. User avaterGBA Editor
            Martin Holladay||#56

            Yes, you're right.

  12. stevesax||#36


    Would you add to your description of assemblies 4 and 5 that the roofing and roofing underlayment should both allow for outward drying? This would avoid "sandwiching roof sheathing between two impermeable layers". (I just reread that 2019 article.) I am currently planning on standing seam metal roofing and 8" of closed cell spray foam in the rafter bays with a ventilation channels in between the foam and the sheathing. In your opinion, is the ventilation necessary or advisable or not? I live in zone 5. Thanks

    1. User avaterGBA Editor
      Martin Holladay||#37

      I would urge any GBA readers who are worried about the issue you raise to read my article on the topic, "Sandwiching Roof Sheathing Between Two Impermeable Layers."

      I disagree with your attempt to reduce the discussion in that article to your proposed advice ("the roofing and roofing underlayment should both allow for outward drying"). The issue is not as simple as you imply, and plenty of builders have decided to install vapor-impermeable roofing above flash-and-batt roof assemblies or spray-foam insulated assemblies.

      If that approach makes you nervous, you are, of course, free to choose an assembly that makes you more comfortable.

  13. stevesax||#39


    I've read the article you mention and other related articles. I am not reducing the discussion; GBA has (thankfully) made me very aware that this discussion is complex. Nor, by way of my question, was I proposing advice. Rather I was checking to see if you forgot to include some of that complexity on impermeable sheathing or if you didn't mention it because you thought it was no longer a concern.

    Everything makes me nervous! As a new builder, about to put on a metal roof, directly after this snowstorm, and then insulate a cathedral ceiling, I am trying to make sense of things given what seems to be divergent expert advice and go with a sound design. I take it that venting the bays can't hurt and if anything makes the assembly more robust/resilient. Is that your take?

  14. User avaterGBA Editor
    Martin Holladay||#43

    You wrote, "I was checking to see if you forgot to include some of that complexity on impermeable sheathing or if you didn't mention it because you thought it was no longer a concern."

    This article deliberately tries to avoid complexity, although it includes links and a "Related Articles" sidebar that provide more information. Here's a fun fact about building: it's almost always possible to screw something up, and any attempt to provide instructions is necessarily incomplete, because no author can include warnings against every imaginable error or omission that an inexperienced builder might come up with.

    I'll include a link to my article, "Sandwiching Roof Sheathing Between Two Impermeable Layers," where relevant. Thanks for raising the issue.

  15. Yeldog||#47

    I have two thoughts ....

    If # 4 and 5 work why does #1 need to have the complex cold roof channel ? That -- Typically is not built correctly anyway. Why not have a proper VB (not plastic) ... if the fiberglass is thick enough ....what's the problem.

    I just had my property foamed last week. New addition has a cathedral ceiling --- Zinc SS roof ... recommended base sheet applied to deck is waterproof. I'm not worried about the plywood sheathing not drying -- where is the moisture coming from ?

    I had a guy years ago not wanting to foam a deck with a slate roof above -- he said the deck can't dry/. The guy had clearly never seen a slate roof go on.

    1. User avaterGBA Editor
      Martin Holladay||#49

      Q. "If #4 and 5 work, why does #1 need to have the complex cold roof channel? ... If the fiberglass is thick enough, what's the problem?"


      A. The interior of the home is warm and moist during the winter, and the roof sheathing is cold (and therefore a potential condensing surface). Moisture moves from the interior to the sheathing two ways: by diffusion (slowed by a vapor retarder or vapor barrier) and by air movement (slowed by an interior air barrier). The thickness of the fiberglass is irrelevant, because fiberglass is neither a vapor retarder nor an air barrier. [Later edit: Actually, the thinner the fiberglass insulation, the lower the chance of condensation. If there is no insulation at all, the sheathing will be warm, and therefore condensation is unlikely. Your proposal of using thick fiberglass actually raises the risk compared to thin fiberglass.]

      Experience shows (and field research confirms) that it is impossible to create a tight enough air barrier on the interior to prevent air movement through hidden pathways from reaching the cold sheathing if your chosen insulation is fiberglass. If you want to try, go ahead. But in 5 or 10 years your roof sheathing might be a soggy mess. It's not worth the risk. (For more information on some confirmatory field research, see "Filling Rafter Bays With Fluffy Insulation.")

      If you add a ventilation channel between the top of the insulation and the underside of the roof sheathing, the ventilation channel removes the incidental moisture that reaches the sheathing, keeping the sheathing dry and keeping the assembly safe.

      If you install closed-cell spray foam insulation, and if the insulation is installed conscientiously, the spray foam is an effective air barrier and vapor barrier that prevents the migration of moisture from the warm interior to the cold sheathing.

  16. Yeldog||#50

    Martin -- It's been my observation that failures are more likely when penetrations are present --- most often recessed lights and plastic VB. I question the drying ability of the channel in #1 -- I guess it does something. I'm not an expert .. but I have done about 50 projects and it's interesting when you rip apart the best and greatest from 25 years ago and you see where they fail. Plastic is the devil .... I wonder what ZIP type buildings are going to look like in 30 years .

    Have never seen rot in a filled fiberglass bay .... covered (closed) with sheetrock.

    Foam does have its issues -- but, it's simple to build and solves lots of problems . End up with a nice quiet -- energy efficient building.

    1. User avaterGBA Editor
      Martin Holladay||#51

      You haven't mentioned your climate zone or geographical location. I seriously doubt that the polyethylene vapor barrier was the cause of the failures you've seen; rather, the problem is probably air leakage.

      That said, building science experts in Massachusetts who have tried to put together careful assemblies with fiberglass and no vent channel have reported worrisome levels of sheathing moisture. Just because you haven't seen a failure doesn't mean that the approach you describe -- "a filled fiberglass bay" -- is an acceptable approach.

      If you live in a warm climate, you might consider an assembly that includes vapor diffusion ports.

  17. Yeldog||#52

    Have mostly done it mid-atlantic. Did build a house in VT and rebuild one my father built there in the 60's. Don't think I have ever ripped apart a house done with plastic w/o damage someplace ... it's all about leaking ... I agree. Most assemblies leak. The plastic traps any moisture .that does get in there.

    It's just been my observation -- these vented assemblies seem to bring in moisture VS solve moisture. Why not control the moisture in the house?

    1. Expert Member
      Malcolm Taylor||#53


      The other reason to include a ventilation channel in assembly #1 is that all the building codes I know of require it.

      Controlling indoor humidity does help, but that alone would never be enough to stop that warm interior air condensing on the sheathing - remembering that it's relative humidity we are measuring, meaning that air in a house kept at say 40% will easily condense at the temperature the roof will be in most seasons.

  18. Yeldog||#54

    Malcolm : Interestingly in NJ the energy audits are requiring sealed assemblies in order to get the rebates and loans ? NJ clean energy program. I have used it for two properties .... going on three .. all six audits.

    Are you speaking of the internal baffles -- that go from soffit to attic space ? I have seen so many damp attic spaces where the moisture came from outside .... same with crawl spaces.

    I agree with above -- it's all about the leaking.

  19. stevesax||#57

    When I renovated our house, I put asphalt roll roofing on a shed dormer over a bathroom. Underneath the plywood sheathing in the bays was a 3/4 vent channel, 1 inch of polyiso and 7.5 inches of cellulose and then a layer of poly and sheetrock. The sheathing was fine for several years. But when the roll roofing deteriorated I had it replaced with TPO. Within a year the plywood became mushy, but only in the two bays that terminated at a chimney at the ridge. Because of the chimney, these bays did not connect to the ridge vent. It seems moisture from the bathroom was able to escape through the asphalt roll roofing, even in these unvented bays. But when these bays were covered by TPO, moisture could not escape. The sheathing over the bays that reached the ridge vent have remained strong even under the TPO.

    At first I just had the mushy sheathing replaced and patched up the TPO, not understanding the cause of the problem. The roof quickly became mushy again. I then ripped out the sheathing again and added a back vent at the base of the chimney. The sheathing got mushy again. I finally realized, upon ripping out the sheath mushy a third time, that the framing around the chimney perpendicular to the rafters was blocking the vent channel from making it to the back vent. Drilling holes in this framing has finally fixed the problem. Not something I want to go through again.

  20. Ben Reese||#58

    In zone 2A new construction objective to construct one story house with a conditioned attic space. No onsite spray foam (except for small sealing applications on seams) can be applied. g Foil faced Thermax installed under the roof rafters is being evaluated. This would be a slight modification to the cathedral ceiling assembly indicated in this article. The area between the rafter would be left open for venting the underside of the roof decking, with foil face foam board used as a moisture seal at the bottom of the roof rafters. This eliminate the need for baffle vents and reduces installation cost by not having to make cuts to fill insulation between the rafters. The objective is conserve not only energy to heat and cool in addition this design minimizes the multiple layer of material that you find in exterior roof insulation (requires two sets of decking in most assembles). See the attached drawing the 4" thermax installed would yield R 26 which is in the acceptable range for zone 2A. If a higher R value is needed an additional 2" sheet R 13 could be added for a total of R 39. This eliminates on site foam application and the toxic emissions and potential for errors in field installation. One other benefit of this design in this super hot climate it allows venting the underside of the deck to minimize shingle damage. All joints of the foam board will be sealed with spray foam and taped to make a air tight barrier between the conditioned attic and the vented roof rafters. Not clear to me if the foil faced foam board requires a fire cover in this application? Also not sure why you don't see this assembly very often other than price?

    File formatFile format
    1. User avaterGBA Editor
      Martin Holladay||#59

      As I wrote in the second paragraph of my article, "While these five assemblies are probably the most common approaches used, note that there are other acceptable ways to insulate a cathedral ceiling."

      Q. "I'm not sure why you don't see this assembly very often other than price?"

      A. You suggested assembly has the following disadvantages: (1) As described, the assembly is only R-26, although you correctly note that someone could always add thicker foam. (2) This approach reduces the ceiling height of the attic to a considerable degree. (3) This approach requires a conversation with your local code official concerning whether or not the Thermax requires a thermal barrier (usually, 1/2-inch drywall). In most cases, builders can convince code officials that Thermax can be left exposed -- but the need to present documentation and to successfully convince a local code official often isn't worth the hassle.

      1. Ben Reese||#60

        Thanks for the quick response. In this case the lower headroom is less of a disadvantage because it is attic space that needs to be minimized. Would it be acceptable to run the foam board up the rafters to the collar ties (and across) to minimize the conditioned space (see attached). Are there any foam board products that are better at fire safety, and are there differences in production emissions? My novice reading on the topic is that control production of foam boards produces less environments impact than "field" applied spray foam that seems to be inconsistent in quality. Let me know if you disagree.

        File format
        1. User avaterGBA Editor
          Martin Holladay||#61

          The assembly you describe, including the version that includes rigid foam under the horizontal collar ties, would work, provided that (a) you are confident that you can secure the foam with large washers and long screws, and (b) the seams are sealed to make them airtight.

          Thermax is a brand of polyisocyanurate that many building inspectors allow to be left exposed. That said, it's impossible to predict how your local building inspector will rule on this issue. Talk to your local building department.

    2. Expert Member
      Malcolm Taylor||#62


      You could achieve the same R-value, with no cutting or baffles, by using 6" batt insulation in the cavities and only 1" of Thermax. The main advantage would be the ease of fastening the foam, which could be secured with cap nails.

      1. Ben Reese||#63

        Thanks Is there a minimum opening or clearance needed for air flow under the roof deck? I have 2 x 10 rafters (9.25 ") less 6" for batt insulation leaves 3.25 inches of air space. Also based on Martin's comments there may be a need to seal the assembly with some material to insure that we have an air tight systems, however it seems wasteful to duplicate decking material.

        1. User avaterGBA Editor
          Martin Holladay||#64

          I'm not sure what Malcolm is suggesting, but it sounds like he wants you to switch from an unvented approach to a vented approach. If you take Malcolm's advice, that means that you'll need soffit vents and a ridge vent. (At least I think that's what Malcolm is proposing.)

          Q. "Is there a minimum opening or clearance needed for air flow under the roof deck?"

          A. Most codes call for a minimum of 1 inch for a vent space, but 2 inches is better.

          Q. "There may be a need to seal the assembly with some material to insure that we have an air tight systems."

          A. I think you're referring to the need for a ventilation baffle to separate the top of your fiberglass insulation from the ventilation space. I certainly recommend the use of a ventilation baffle if you are planning to use fiberglass insulation in a vented roof assembly. Malcolm knows that many contractors skip the baffle, and just leave an air space above the fiberglass, with no air barrier to separate the fiberglass insulation from the moving air stream. You get to choose which approach makes sense to you.

          1. Expert Member
            Malcolm Taylor||#65

            Martin and Ben,

            First, A very merry Christmas to you both!

            Bens's foam-only assembly is vented. He just was proposing not to have any insulation in the cavity. Filling them with batts (our code requires a 3" channel if no baffles used) seems a lot easier that trying to install that much foam. I'm sure the R-value of the batts will degrade a bit due to wind washing (again I'm still unclear by how much), and the air-barrier will still be the foam, so no extra decking.

          2. User avaterGBA Editor
            Martin Holladay||#66

            Merry Christmas, and thanks. You're right, of course -- Ben's proposed roof assembly was always vented. My mistake. And I agree -- your approach would be easier.

  21. Ben Reese||#67

    Thanks to both of you. I appreciate the great guidance.

  22. Bill Vesely||#68

    Would the ambient air temp influence when you would consider performing option 4? We have a post-and-beam type cathedral ceiling in our ranch, so the bays are about 3 times as wide as the typical assembly. It's a small project area, so I am wanting to do option 4. Second question would be about dealing with potential overfill of the cavity. The beams are 6x6s. Maybe that won't be an issue. The cavity depth might just get me to a code-minimum R-value; would foam attached to the undersides of the beams cause a drying inward issue? Maybe I am missing the point on that. Thanks in advance for the help, Bill.

Log in or become a member to post a comment.

Sign upLog in



Recent Questions and Replies

  • |
  • |
  • |
  • |