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

Community and Q&A

Unvented metal roof

GGwk9xmLvb | Posted in Energy Efficiency and Durability on

In the process of construction… currently have a standing seam metal room installed, over synthetic felt, over OSB. No insulation installed yet. There will be ducting and other HVAC equipment in the attic. The plan is to spray open cell foam under the roof deck. This is the recommendation of all three local insulation installers. However, as the future owner, I’m beginning (maybe a bit too late) to read about unvented roofs and I’m wondering if this is really the best option? If it isn’t ideal, are we too late to take another route, given our current progress?

One other question, it appears the eaves, which extend 2.5′, won’t be insulated, not sure if that matters?

We are in climate zone 4.


GBA Prime

Join the leading community of building science experts

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


  1. Expert Member
    ARMANDO COBO | | #1

    Unless you are planning on spraying 11” (R38 for CZ4) open cell under your roof decking, you may have cold sheathing in the winter. Depending on your interior humidity and temperature, can lead to condensation under your roof sheathing. Your Architect/Designer, Energy Consultant or Insulator should have done a roof assembly analysis to determine the correct insulation and/or whether you should have a conditioned or unconditioned attic and how to approach it. It would also help you a little if you live on the dry side of CZ4.
    On your question about the eaves, as long as the wall sheathing continues to the roof sheathing and is sealed so there are no air leaks, then you’ll be ok.

  2. GGwk9xmLvb | | #2

    Wow, 11" is a lot. The plan is for 6" of open cell, again, apparently the norm here, as all three installers recommended that amount (5.5" - 6").

    - Any recommendations, given that the metal roof is already installed as unvented?
    - Is 11" of open cell the only option on the bottom of the deck? Is there possibly a more cost effective approach using dense pack cellulose?

    Thanks for the reply!

  3. GBA Editor
    Martin Holladay | | #3

    Your insulation contractors are planning to install only R-20 to R-22. That's not much. In fact, it doesn't even meet the minimum code requirements (2006 IRC) for Climate Zone 4, which calls for at least R-38.

    If you install less than the code minimum insulation, you will pay higher energy bills for as long as you own the building. Many builders prefer to exceed code minimum requirements.

    I disagree with Armando, who wrote, "Unless you are planning on spraying 11” (R38 for CZ4) open cell under your roof decking, you may have cold sheathing in the winter." Adding extra insulation won't make your sheathing warmer -- it will make it colder. The fact that your sheathing will be cold is a given. Adding insulation under your roof sheathing won't warm it up.

    There is some logic to choosing open-cell foam over closed cell foam, because the open-cell foam will allow some drying of the sheathing toward the interior. In your climate, wintertime moisture accumulation due to diffusion from the interior shouldn't be a problem; however, there is always the possibility of a small roof leak, and drying to the interior is a (small) virtue.

    Whatever you do, be sure to install at least R-38.

  4. Expert Member
    ARMANDO COBO | | #4

    You are correct Martin, I expressed incorrectly the “cold sheathing in winter”, my intention was to say that a better way would have been to install 1”-1.5” rigid insulation on the outside of the sheathing and the 5.5” open cell on the inside to avoid condensation, otherways you may have cold sheathing in the winter.
    I do disagree with “In your climate, wintertime moisture accumulation due to diffusion from the interior shouldn't be a problem”. I’ve seen many times condensation under the roof decking for the same reason in KS and MO (all CZ4). It’s surpricing to know that Topeka, KC and St Louis have a 1°F lower 3-month colder temperature than NYC and DC, and the same as Newark, Providence and Boston. (from Climate Contant 5).

  5. GBA Editor
    Martin Holladay | | #5

    Condensation can occur in that climate, but as long as the assembly has some ability to dry to the interior, any moisture accumulation should be more than balanced by drying.

    I certainly agree that it would have been better to install some rigid foam insulation on top of the roof sheathing.

  6. Expert Member
    ARMANDO COBO | | #6

    Yes, but when you see rot appear on the sheathing and rafters its a clear sign that other methods should be used in CZ4 and even in CZ3 as we see the new codes are going that way. I choose not to wait for the new codes.

  7. Expert Member
    ARMANDO COBO | | #7

    Chandler added that while codes appeared to limit foam thickness to no more than 6 in., he routinely recommended to his clients that they use more. “I advocate spraying 11 in. and certainly never less than 8 in. here in North Carolina,” he wrote, “so technically all of my homes are in violation of code on this issue.”

    From a recent article,

  8. GGwk9xmLvb | | #8

    If I could change things, we'd have rigid foam above the sheathing. It makes sense, and provides thermal breaks. So, at this point, it sounds like the best thing to do is 11" of open cell, if the inspector will allow it.

    Is condensation at the sheathing layer a concern with this roof, and consequently mold/rot?

  9. GGwk9xmLvb | | #9

    Also, I suppose I still have the option of insulating the attic floor with cellulose, instead of the ceiling. Would this be advisable over a 'conditioned' attic, given the circumstances? The air handler/duct work does throw a bit of a kink in that approach though.

    (Thanks for the very helpful input)

  10. GBA Editor
    Martin Holladay | | #10

    It is my opinion that 11 inches of open-cell foam can be installed in your climate on the underside of roof sheathing without a vapor retarder. However, it appears that Armando may disagree.

    If you have further doubts, contact the technical representative of the manufacturer of the foam you intend to use, and get the manufacturer's opinion on the question.

  11. GBA Editor
    Martin Holladay | | #11

    If you are planning to insulate between your rafters, it doesn't make any sense to also insulate your attic floor. Choose one location for your ceiling insulation, and insulate well.

  12. Expert Member
    ARMANDO COBO | | #12

    as Martin said, at this point your best approach is to install the 11" of open cell w/o a vapor retarder and forget the cellulose on the ceiling.

  13. GGwk9xmLvb | | #13

    Thanks everyone for the answers, much appreciated.

    One follow on question, (if this post still has some visibility): I found an informative article ( that suggests first putting in rigid foam, to form a gap between the roof sheathing and the insulation. This would allow for ventilation directly under the roof deck, but still to have a 'conditioned' attic.

    I'd really appreciate thoughts on this approach. Further, would it be advisable to use foil faced polyiso for this (if I remember correctly, foil faced polyiso, in conjunction with an air gap, can sometimes exhibit an even higher r-value), or must this be using an open cell, permeable, foam?

    Thanks again!

  14. GBA Editor
    Martin Holladay | | #14

    Your suggestion can work.

    First install some 1.5" x 1.5" sticks in the corners of the rafter bays to maintain the ventilation gap; then install your rigid foam.

    There is nothing wrong with using vapor-impermeable rigid foam for this purpose; the insulation will still dry to both directions (from the top, via the ventilation channels, and from the bottom, through the open-cell spray foam).

  15. Expert Member
    ARMANDO COBO | | #15

    This is from a link Martin posted today at:

    I don't believe that all attics should be unvented or conditioned, but looking at Bill Rose and Jeffrey Gordon's PPT, there are some good reasons to have conditioned attics. Look at slide 33 in particular.

    Slide 33:
    Wet, Cold, Coastal Climates
     Computer simulation indicates that attic ventilation results in higher sheathing moisture content, not lower
     Ventilation makes the attic colder, without lowering water vapor levels very much
     Colder means wetter

    Forest, Walker, Attic ventilation and Moisture. Final Report, CMHC, 1993

    Slide 34:
    Warm and Humid Climates
     No one has ever claimed moisture control benefits in venting attics in warm, humid
     Outside air is more humid than inside, and attic venting will tend to increase rather
    than decrease moisture levels in the attic

    Slide 40:
    “Even an unvented attic is very unlikely to develop significant ice dams unless there are significant heat sources in the attic or significant air leakage from below,…”

    Slide 60:
    Cooling season energy savings
     Venting will reduce the temperature in an open attic. The difference in attic temperature between a vented an unvented attic, with R-30 at the ceiling, translates into minuscule savings. No savings have ever been measured.
    Usually there is a penalty with venting because venting causes greater air pulsing across the ceiling.

    Slide 62:
    What have we learned about the impact of ventilation on 4 issues?
     Moisture control – can have a some impact, but it is small, and not the determining factor in an attic

  16. David_Gregory_CZ3_CA | | #16

    Thanks for posting/referencing the Rose/Gordon report...great resource.

    But I'm still struggling to understand some things:

    1. Why is the debate about 'vented' vs. 'un-vented'? This seems too polarized. It's clear that having an attic that is only vented when the wind happens to be blowing is a liability - still hot in summer (though admittedly unclear net energy effect), but can pull conditioned air out of a not-well-air-sealed house; and is clearly a liability in winter (for energy and condensation issues). But the de-pressurization issue could be easily handled by having two equal fans, one blowing in, one pulling out, with electric dampers (because simple 'backflow' dampers wouldn't be adequate). Moisture could be handled separately by a dehumidifier (but it seems that the only real moisture issue is from bathroom fans vented to the attic! Or other ceiling penetrations.) So the question should be the cost of a setup to vent when beneficial, vs. spending the money elsewhere. Any references? With costs?

    2. How do we know that _controlled_ venting as I suggest above (no de-pressurization/good air sealing, and only when attic is very hot) isn't effective after you have more than 6.5" of insulation in the ceiling? It depends on actual R-value, insulation density, solar insolation/roof color/transmission, attic/room air deltaT, and whether you have HVAC ducts up there! The report says in mixed climates venting is a trade-off between summer savings and winter losses; in cold climates it's a net loser. From this, might we extrapolate in _hot_ climates (at least sunny / dry ones) it would be a net winner?

    Using a roof as a (large, inefficient) passive solar collector in winter is an argument against insulating the roof of an un-vented attic. Intentional ventilation would reduce the summer liabilities of this assembly...

    Guess I'll stop there...have 5+ GBA tabs alone to read on the topic, before even getting to other sources...! Info overload... ;)

    [DOE3B/mid-dark roof/blown fiber barely over ceiling joists/HVAC all over/two rotary vents/bath fans vent to attic!! working on it...step by step...]

  17. GBA Editor
    Martin Holladay | | #17

    Your suggested set-up for an unconditioned attic -- one fan to pressurize the attic, balanced by a second fan to depressurize the attic, and a dehumidifier to keep the attic dry -- is wildly energy-intensive. It's off the charts -- it uses a lot of energy to solve an issue that probably isn't even a problem (that is, regulating the temperature and RH of a volume of air that is OUTSIDE your home's thermal envelope).

    Plenty of researchers have done a cost/benefit analysis on powered attic ventilators -- and I'll ignore your suggestion of running a dehumidifer outside of your home's thermal envelope for the time being -- and they all conclude that it takes more energy to run a powered attic ventilator than any conceivable energy savings.

    Here's some more information:

  18. David_Gregory_CZ3_CA | | #18

    Thanks for your response; and you may still be right [don't ever (mechanically?) vent or dehumidify an attic], but in defense of my green credentials (and the homework I've been doing) I think I wasn't clear in explaining 'my set-up'. It was actually two separate scenarios (thought exercises); not running two fans and a dehumidifier at the same time for indefinite / lengthy periods! ;) I was simply trying to isolate the potential issues of heat and humidity and solve them directly. [In all this I'm assuming that an 'un-vented attic' doesn't _require_ insulation/radiant barriers under the this correct?]

    The first point was simply that it makes absolutely no sense to use 'uncontrolled' venting (passive, or 'solar', which isn't much better calibrated) to deal with moisture, as many have shown. If you do everything else right (attic totally separate from house air, _and_ from outside air) and moisture was somehow _still_ a problem, focus on the problem: all you need is a dehumidifier with a sensor so it runs only when necessary - which would be rare. Very little operating energy cost. Given that the sealed hot attic will drive out moisture but not let it in easily, it's hard to imagine this being necessary, and probably not worth the capital cost of equipment.

    Which brings us to the completely separate second suggestion, about ventilation: Existing home. Ceiling air-sealed (retrofit) as best as possible, deep ceiling insulation (choose your R; but way more than 6.5 inch [Burch&Treado1978 via Rose/Gordon]). Original attic vents permanently sealed. No ceiling penetrations (especially bathroom fans!). Duct work in attic, well insulated also. No air is going in or out, moisture driven out by heat. But there's a lot of heat. High albedo roof is not an option. Mixed climate; any benefit to insulation/radiant barrier at rafter/deck will come with a winter penalty. Even though the ceiling is air-sealed, because it is a retrofit, we can't be sure it's 100% - and we definitely don't want to cool the attic by pulling conditioned air. Now the problem has been isolated to reducing summer attic air temperatures without reducing 'free heat' in winter or creating negative pressure. So why not replace 150+ degree air (dark roof, sunny climate?) with 80-90 degree air, only when there's a benefit, via two _SMALL_ fans, one in, one out, to eliminate any concerns about negative pressure? Positive closure dampers to avoid in/exfiltration when not in use. Operating cost small. Capital cost modest, but not insignificant. Benefits would need to be proven by the numbers. I'll work on it.

    The point was, I assume a completely sealed attic with ductwork and no insulation under the decking will get very hot in summer. But I hate to give up 'free heat' in winter. At some point, might it be cheaper to (neutral pressure) vent the space occasionally, rather than blowing in another (x) inches of cellulose / super-insulating the ductwork to deal with a 100deg(duct)/70deg(indoorspace) deltaT?

    Finally...I would argue that the air in an un-vented attic is _not_ 'outside' your home's thermal envelope. The roof is part of your home, and part of the _building_ envelope, and creates a different thermal condition (higher deltaT than the 'outside' air). Doesn't that make it fit the definition?

    Hope that clarifies and qualifies me to 'get out of jail free'... ;)


Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |