GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon
Musings of an Energy Nerd

Creating a Conditioned Attic

Five questions to consider before insulating a sloped roof

Image 1 of 2
Should you “cathedralize” your attic? Moving the insulation layer from the attic floor to the sloped roof brings the attic into the home's conditioned envelope and greatly reduces the energy penalties associated with attic ductwork.
Image Credit: Fine Homebuilding
Should you “cathedralize” your attic? Moving the insulation layer from the attic floor to the sloped roof brings the attic into the home's conditioned envelope and greatly reduces the energy penalties associated with attic ductwork.
Image Credit: Fine Homebuilding
Insulating the sloped roof brings the attic ductwork inside this home’s conditioned space.

UPDATED on August 13, 2017 with new concluding paragraphs

Millions of Americans live in states where residential HVAC contractors routinely install ductwork in unconditioned attics. In many cases, these attics also contain a variety of appliances, including air handlers, furnaces, or water heaters.

Because the disadvantages of this arrangement are fairly well known, I’ll mention them only briefly:

  • During the summer, attic temperatures often exceed outdoor temperatures.
  • Attic ducts almost always have thinner insulation than ceilings, in spite of the fact that the delta-T (that is, the temperature difference) between the air in the ducts and the air in the attic is even greater than the delta-T between the inside of the home and the exterior.
  • Most duct seams leak; as a result, supply ducts lose conditioned air to the attic, while return ducts suck in attic air — air which is hot in summer and cold in winter — and bring it to the air handler.
  • If access to the attic is through a hatch, servicing any HVAC equipment in the attic is awkward at best.

The bottom line: running ducts through an attic saves money for the builder, but costs the homeowners dearly in increased energy costs.

Ductwork belongs inside a home’s thermal envelope

Ideally, HVAC appliances and ductwork should be located inside a home’s conditioned envelope. In the northern half of the country, appliances and ductwork are routinely located in basements or crawl spaces. If your house has a slab foundation, HVAC appliances can be located in an equipment room and ductwork can be located in interior soffits.

Another solution is to move the insulation from the attic floor to the sloped roof, thereby creating a conditioned attic.

Assuming you want to create a conditioned attic — either during new construction or as a retrofit project — how do you go about doing it? Before getting down to the nitty-gritty details, you’ll need to answer at least five overlapping questions:

  • Will the insulation be installed above the roof sheathing, between the rafters, or below the rafters?
  • Will the insulated roof include ventilation?
  • Is it important to address thermal bridging through the rafters?
  • Will any combustion appliances be located in the attic — and if so, where will the appliances’ combustion air come from?
  • Is there any need to install a forced-air register to condition the attic air?

Question 1: Where will the insulation go?

There are at least three ways to insulate a sloped roof:

  • Install rigid foam insulation above the roof sheathing.
  • Install insulation between the rafters.
  • Install rigid foam insulation under the rafters.

A fourth approach combines between-the-rafters insulation with rigid foam insulation (either above the roof sheathing or below the rafters).

If you install rigid foam insulation above the roof sheathing

Installing rigid foam insulation above the roof sheathing makes a lot of sense. The foam keeps the roof sheathing warm, and therefore dry. Rigid foam also interrupts thermal bridging through the rafters — a big benefit.

Of course, there is a limit to how much foam you can install above roof sheathing; the main problem is the difficulty of hitting the underlying rafters when attaching furring strips or plywood above the foam with long screws. Several builders and homeowners have successfully installed 6 inches of rigid foam on top of a home’s roof sheathing, although some (including Alex Cheimets of Arlington, Mass.) have grumbled about problems driving long screws.

Two layers of rigid foam installed with staggered seams will perform better than one layer of rigid foam with the same total thickness.

Six inches of polyisocyanurate gives an R-value of R-39, which isn’t enough for cold climates. That’s why builders who install rigid foam on top of roof sheathing usually supplement the foam with additional insulation between the rafters.

One way of avoiding the hassle of installing thick insulation above roof sheathing — perhaps in exchange for a different set of hassles — is to install structural insulated panels (SIPs) or nailbase (basically, SIPs with OSB on only one side). It’s possible to order 12-in.-thick SIPs that measure R-48.

For more information on this topic, see How to Install Rigid Foam On Top of Roof Sheathing.

If you insulate between the rafters

Although I don’t recommend it, it is possible to install all of your sloped-roof insulation between the rafters. A few guidelines concerning insulation installed between rafters:

  • Air-permeable insulation (like fiberglass) is risky, since it allows moisture-laden air to reach the cold roof sheathing. If you want to use air-permeable insulation in this location, you have two choices: include a ventilation channel between the top of the insulation and the underside of the roof sheathing, or install rigid foam insulation on top of the roof sheathing to keep the sheathing warm. (Don’t do both, since these two solutions are incompatible.)
  • Insulation that is air-impermeable but vapor-permeable (like open-cell spray polyurethane foam) is risky in a cold climate (climate zones 5 or higher) unless the interior side of the insulation is protected with a vapor retarder. According to section R806.4 of the 2009 IRC, such a vapor retarder is required in climate zones 5 through 8 for open-cell spray foam installed to create an unvented attic assembly. Recent research shows that vapor-retarder paint is ineffective when sprayed directly on cured foam, so if you want to use open-cell foam on the underside of roof sheathing in a cold climate, the cured foam should be shaved back even with the bottom of the rafters, and the insulated rafter bays should be protected by a layer of drywall. At that point you can install vapor-retarder paint on the drywall. In a cold climate (climate zones 5 or higher), if you don’t plan to install any drywall, you should use closed-cell foam, not open-cell foam.
  • Recent research suggests that open-cell spray foam may be risky in all climate zones, so the safest spray foam insulation to use is closed-cell spray foam. For more information, see Open-Cell Spray Foam and Damp Roof Sheathing.
  • Remember that it’s important to meet or exceed minimum code requirements for insulation R-value; beware of spray-foam contractors who try to convince you that below-code-level insulation is adequate. For more information on this issue, see It’s OK to Skimp On Insulation, Icynene Says.
  • According to the 2009 IRC (Section R806.4), it’s possible to build an unvented roof assembly with a combination of rigid foam insulation above the roof sheathing and air-permeable insulation in the rafter bays. The code requires that “rigid board or sheet insulation shall be installed directly above the structural roof sheathing as specified in Table R806.4 for condensation control.” The table calls for a minimum of R-5 foam for Climate Zones 1-3, R-10 for Climate Zone 4C, R-15 for Climate Zones 4A and 4B, R-20 for Climate Zone 5, R-25 for Climate Zone 6, R-30 for Climate Zone 7, and R-35 for Climate Zone 8.
  • The 2009 IRC also allows another approach: it’s possible to use a combination of air-impermeable and air-permeable insulation in unvented rafter bays, as long as the minimum R-value of the air-impermeable insulation that is “applied in direct contact with the underside of the structural roof sheathing” meets the requirements for condensation control shown in Table R806.4.
  • The 2009 IRC defines air-impermeable insulation as “an insulation having an air permeance equal to or less than 0.02 L/s-m² at 75 Pa pressure differential tested according to ASTM E 2178 or E 283.” Although spray foam insulation and rigid foam insulation can meet this standard, dense-packed cellulose cannot.
  • Most rafters aren’t deep enough to accommodate enough insulation to meet minimum R-values required by code, especially if the rafter bays include a ventilation channel. For example, 2×10 rafters, which are 9 1/4 inches deep, provide room for only about 8 1/4 inches of insulation (about R-30 of fibrous insulation) if you ventilate the bay. Some builders solve this problem by sistering, furring down, or scabbing on additional framing below the rafters to deepen the rafter bays.
  • Most building codes require spray foam insulation to be protected by an ignition barrier or a thermal barrier. For more information on this issue, see Thermal Barriers and Ignition Barriers for Spray Foam.

If you install rigid foam insulation under the rafters

Installing rigid foam under the rafters has some of the same advantages, as well as the same disadvantages, of installing foam on top of the roof sheathing.

Attaching thick foam overhead is awkward, so few builders install more than 2 inches of foam in this location. That’s why rigid insulation below the rafters is almost never done in isolation; rather, it usually complements insulation installed between the rafters.

One popular insulation combination for sloped ceilings: 2-in. thick foil-faced polyisocyanurate attached to the underside of the rafters and held in place by 1×4 furring strips, with dense-packed cellulose blown into the rafter bays through holes in the rigid foam. (If you choose this method, be sure to include a ventilation channel under the roof sheathing.)

Question 2: Will the insulated roof include ventilation?

Insulated sloped roofs usually include a soffit-to-ridge ventilation channel. Soffit vents allow air to enter the bottom of these channels, and a continuous ridge vent allows air to exit at the ridge. Such ventilation channels work best on simple shed roofs or gable roofs; if your roof has hips, valleys, dormers, and skylights, it will be hard to ventilate well.

The 2006 International Residential Code (IRC) sets out roof ventilation requirements in Section R806. However, in Section R806.4, the IRC permits conditioned attic assemblies without any roof ventilation; according to the code, if ventilation is omitted, only “air-impermeable” insulation can be used in contact with the roof sheathing.

Although the 2006 IRC neglected to include a definition of “air-impermeable insulation,” that omission was rectified in the 2009 IRC, which defines air-impermeable insulation as “an insulation having an air permeance equal to or less than 0.02 L/s-m² at 75 Pa pressure differential tested according to ASTM E 2178 or E 283.” In other words, an air-impermeable insulation must meet the same airtightness standard as an air barrier material. Although spray foam insulation and rigid foam insulation can meet this standard, dense-packed cellulose cannot.

If you choose to install a fibrous insulation like fiberglass or cellulose between your rafters, you must include a ventilation channel between the top of your insulation and the underside of the roof sheathing.

Even if you insulate between your rafters with an air-impermeable insulation like spray polyurethane foam, you may want to provide a ventilation channel under your roof sheathing. The main function of such a ventilation channel is to separate the roof sheathing from the foam; this facilitates future repairs of sheathing rot.

Polystyrene “ProperVents” are too narrow, too shallow, and too flimsy. The sturdiest vent channels are site-built channels, using 1″x1″ sticks in the upper corners of the rafter bays and thin plywood, Masonite, or rigid foam. If the panels used to build the vent channels are caulked in place, it’s possible to create an effective air barrier to prevent wind-washing from lowering the performance of the insulation.

Remember: ventilated rafter bays are incompatible with rigid foam insulation installed on top of the roof sheathing. If you want to reduce thermal bridging through the rafters of a ventilated roof, you’ll have to install rigid foam insulation under your rafters.

Question 3: Is it important to address thermal bridging through the rafters?

By now, most builders know the difference between the R-value listed on a roll of batt insulation and whole-wall R-value. (The whole-wall R-value is always significantly lower than the label on the batts.) There are several reasons why whole-wall R-values are so low; the most significant reason is thermal bridging through the wall framing.

Thermal bridging can also be a factor that lowers the performance of an insulated roof. It makes little sense to “cathedralize” an attic — that is, to bring the attic into a home’s thermal envelope — in hopes of saving energy, if the roof insulation fails to address thermal bridging through the rafters.

In other words, if your goal is to save energy, do it right. That means installing rigid foam on top of the sheathing or under the rafters.

Question 4: Will any combustion appliances be located in the attic?

Once you’ve tightened up your thermal envelope and brought your combustion appliances indoors, you have to come up with a plan to provide your appliance burners with combustion air. The best solution is to install only sealed-combustion appliances in a conditioned attic.

Each sealed-combustion appliance has two big pipes: one is the flue, and the other conveys fresh outdoor air to the burner.

If you hope to convert an existing unconditioned attic to a conditioned attic, the presence of any atmospherically vented appliances (for example, a gas water heater or a gas furnace) complicates the retrofit work. If you can’t afford to buy new sealed-combustion appliances, you’ll probably be better off leaving your attic unconditioned.

Question 5: Is there any need to condition the attic with a forced-air register?

Okay, you’re coming down the home stretch now. You’ve insulated your sloped roof and your gable walls. You’ve protected any exposed foam with a thermal barrier like drywall. You’ve done your best to limit air leaks in the tricky area at the attic’s perimeter, where the rafters meet the top plates of your exterior walls.

You’ve just created a conditioned attic. Now your HVAC contractor asks you whether or not you want a supply register and a return grille in your attic.

The answer to this question depends on several factors. If you have insulated the roof assembly with closed-cell spray foam, you probably don’t need a supply register or return grille in your attic. If you’ve done a good job insulating and air sealing, the temperature of the air in your attic will approximate indoor conditions, even without a supply register or a return-air grille.

If you have insulated the roof assembly with open-cell foam and you live in a humid climate, however, there is increasing evidence that you probably need to install a supply register and a return grille in your attic in order to keep humidity levels low enough to avoid damp roof sheathing. For more information on this issue, see Open-Cell Spray Foam and Damp Roof Sheathing.

Conditioned attics have a few drawbacks

Creating a conditioned attic is a solution to a fundamental design flaw (locating ductwork or HVAC equipment outside of a home’s thermal envelope). But conditioned attics are not unalloyed blessings; they come with their own set of drawbacks.

These drawbacks include:

  • Installing insulation along a sloped roof is always more expensive than installing insulation on an attic floor.
  • It’s usually easier to perform air sealing work on the attic floor than to ensure that a conditioned attic is well sealed.
  • In a home with a conditioned attic, it’s much harder to locate roof leaks or repair rotten roof sheathing than in a home with an unconditioned attic.
  • Damp roof sheathing will always dry faster in a home with an unconditioned attic than in a home with a conditioned attic.

The bottom line: unless you have ductwork in your attic or you plan to convert your attic to living space, an unconditioned attic is usually preferable to a conditioned attic.

A radical approach: Abandoning attic ductwork

If your house has a vented unconditioned attic that includes ductwork, and you are thinking of converting the attic into a conditioned attic, you may be surprised at the high cost of the required work. Creating a conditioned attic is expensive.

In light of the high cost of the work, it’s worth considering a radical option: abandoning the existing forced-air system. In some cases, the cost of installing one or two ductless minisplits is less than the cost of creating a conditioned attic. Moreover, the minisplits will perform at a much higher efficiency than a conventional forced-air system with attic ductwork.

Author’s note: I’d like to provide credit to GBA reader Dana Dorsett for suggesting this last approach.

Last week’s blog: “The Pros and Cons of Advanced Framing.”


  1. User avater GBA Editor
    Allison A. Bailes III, PhD | | #1

    Great article!
    Great article, Martin! Maybe you could link this article to the one you wrote a while back about contractors putting in lower R-values when they insulate the roofline. That's an important issue that you kind of glossed over in this article but that generates a lot of discussion, especially from the SPF installers who insist on diminishing their installed thickness because of the diminishing returns of extra R-value.

    I agree with you that we need to pay more attention to thermal bridging. Here in the Southeast, SPF installers generally go in and spray about 6" of open cell foam between the rafters for an R-value of ~20. Sometimes they spray over the rafters to reduce that thermal bridging, and sometimes they don't. Rigid foam above or below is almost never installed here. The closed cell installers generally put in the same R-value (~20), but I've seen as little as 2" for R-14.

    One thing you didn't mention is whether or not you should remove the existing insulation in the flat ceiling if you insulate the roofline in an existing home. I haven't seen data on this, but I've heard that you should remove that insulation to keep the attic close to the temperature in the house, which helps prevent condensation on the underside of the roof insulation. Any thoughts?

    Finally, I think you used the wrong word in the last paragraph of your answer to Question 2:

    Remember: insulated rafter bays are incompatible with rigid foam insulation installed on top of the roof sheathing.

    Didn't you mean to say 'ventilated' instead of 'insulated' there at the beginning of the sentence?

  2. User avater GBA Editor
    Martin Holladay | | #2

    Response to Allison Bailes
    Thanks very much for catching my typo, which I have corrected in the text.

    Thanks also for all of the points you brought up:
    1. As I made clear in my earlier blog, It’s OK to Skimp On Insulation, Icynene Says, I believe that it is essential to comply with minimum code requirements for insulation R-value, even when using spray polyurethane foam. The main reason that spray foam contractors try to circumvent minimum code requirements for insulation R-value is because spray foam is so expensive that it has a hard time competing with other insulation products. I have taken your suggestion to heart and added a reference to this point in my blog.

    2. I agree that the Achilles' heel of many spray-foam jobs is that thermal bridging through the rafters is not addressed.

    3. In most cases, the existing insulation on an attic floor should be removed when an unconditioned attic is converted to a conditioned attic. However, this recommendation should be addressed on a case-by-case basis. Depending on the climate and the thickness of the existing insulation, it may make sense to leave the old insulation in place.

  3. John Nicholas | | #3

    Follow Up on Removing Existing Insulation
    How do you recommend removing the existing insulation? Are there other services out there with truck mounted vacs that could do this? Is is more of a hand and time job?

    What about removing extremely deteriorated R-2 Rockwool batts, before blowing cellulose?

  4. Allan Edwards | | #4

    Martin: You mentioned that
    You mentioned that for "many spray-foam jobs thermal bridging through the rafters is not addressed." Can you point me to any other articles by you or anyone else in the Building Science community on thermal bridging at the rafters? By the way, thanks for another great article.

  5. User avater GBA Editor
    Martin Holladay | | #5

    Response to John Nicholas
    John Nicholas,
    1. Contact cellulose insulation contractors for information on removing existing insulation with a vacuum. Many of these contractors offer this service.

    2. If you have severely deteriorated old batts, it's probably a good idea to remove them, if only to get rid of the rodent feces. That's probably an old-fashioned job requiring a dust mask, gloves, and many garbage bags.

  6. User avater GBA Editor
    Martin Holladay | | #6

    Response to Allan Edwards
    Thermal bridging links -- (by the way, the "search" function on our GBA site can help you track down the answer to questions like this):

    Thermal Bridging

    Thermal bridging with spray foam insulation

    Insulation Overview

  7. M. Johnson | | #7

    Pitfalls of different climates
    I appreciate your coverage of this very interesting topic. In the hot-humid South where I live, it is my understanding there are a number of lawsuits from insulated attics and I believe the #1 problem is rotting roof decking connected with undiscovered water leaks. Research is still ongoing but I do not see a consensus yet as to the wisdom, nor guidelines to avoid trouble.

    The methods you describe work well in some climates, but around the country there are several zones with special needs. The simple factor of outdoor dew point being chronically very high or low is important in many situations. Hot-humid climates make it normal to consume a lot of energy, and I would like to hear more from the perspective of a builder in such climates.

    Thank you for bringing up the attic subject, it is very interesting.

  8. Riversong | | #8

    Why Does "Green" Require Hazmat Suits
    I love the picture at the top of this article. It always amazes me that people would consider "green" any material or method that is IDLH (a NIOSH acronym for Immediately Dangerous to Life and Health) and requires a complete isolation suit for safe installation.

    And M. Johnson, you're quite right that studies have demonstrated the danger to the roof sheathing when roofs can't breathe in either direction and take literally months to dry after a leak event. Anything wooden needs to breathe.

  9. User avater GBA Editor
    Martin Holladay | | #9

    Response to M. Johnson
    M. Johnson,
    1. You are quite right that installing spray polyurethane foam directly against roof sheathing is controversial, because of the possibility that the foam can disguise roof leaks. That's why I wrote, "Even if you insulate between your rafters with an air-impermeable insulation like spray polyurethane foam, you may want to provide a ventilation channel under your roof sheathing. The main function of such a ventilation channel is to separate the roof sheathing from the foam; this facilitates future repairs of sheathing rot."

    2. You wrote, "The simple factor of outdoor dew point being chronically very high or low is important in many situations." If you follow the recommendations provided in this article, however, your roof sheathing will never be cold enough to allow for moisture accumulation in the sheathing due to a dew-point issue -- even (especially) in a southern climate.

    3. I disagree with our statement, "Hot-humid climates make it normal to consume a lot of energy." If you design your house well, you will have low energy bills, even in the south. For more information on energy-saving tips for hot climates, see Hot-Climate Design .

  10. User avater GBA Editor
    Martin Holladay | | #10

    Comment from Joe W
    [Joe W. was having trouble with the GBA spam filter, so he sent me the following comment by e-mail:]

    I've heard a couple of proposed solutions to increasing the accuracy of screwing SIPs and thick exterior insulation to rafters.

    One was to order truss rafters with a screwable flange -- this was advanced as an easily available option.

    Another was to double up truss rafters every four feet to provide a wider surface.

    Of course, with a timber frame set up, either rafters or purlins would be expected to be wider to begin with.

    Are these legitimate suggestions?

    Thanks in advance, and happy new year
    Joe W

  11. User avater GBA Editor
    Martin Holladay | | #11

    Response to Joe W
    If trusses can be ordered with wide top flanges, that seems like an easy solution -- although some truss-spacing jigs probably won't work with wide flanges.

    Doubling up trusses every 4 feet doesn't make sense to me, for two reasons: it's an expensive solution, and you'll want to space your screws closer than every 4 feet.

  12. Aj Builder, Upstate NY Zone 6a | | #12

    A home that uses half the energy of a code built home
    Seal up a home well somehow and stop the flow of heat thru the insulation and a home instantly will use half the energy of yesterdays leaky homes.

    Icynene does that and I can take anyone to a home and show you.

    The home will have less than R 38 above in complete defiance of Martin's rant that Icynene is selling nonsense. I say Martin who I agree with most of the time is barking up the wrong "R vs. air leaks" tree.

    Talk is cheap. I have Icynene installed and it is performing amazingly.

    Martin, come visit, I will give you a tour. Then come back to this thread and post your thoughts. I have never seen you post actual hands on experience with Icynene. Stop railing against it till you do have some personal knowledge of the success many of us are having with it.

    I am all for going green and entirely cellulose as to insulation. But I am not going to say Icynene does not do a great job. It does. It does for me. It does for my customers. It does for my spray company.

    Every good assembly has its' use. What Icynene did for us is cut the size of our HVAC system size in half. It reduced the number of AC run hours to 1/4.

    Robert will rant back as to myself and foam and Icynene and using less than "code" R. Rant away. The homes work, I am in them all the time and the HVAC is off most of the time. I have never been in a home that an HVAC system runs so little an amount of time. And they are sized smaller than the installers wanted to size them at my insistence. I will not counter-post Robert. You all do not need an extended warming from that.

    peace out. I suggest anyone that if you are trying any assembly that you learn about online to go to sites and learn in person. Talk to satisfied customers. Go to sites that are in the process of being built.

    Cellulose is by far the greenest insulation. Icynene is a less green product but it is better at reducing energy demands then Martin is willing to admit.

  13. Alex Rockas | | #13

    unvented attic spaces
    I am looking at a home in Santa Cruz, CA that is to be built. It has a rather cut up 4 in 12 truss roof with many areas of California framing. It is fairly complicated with fake rafter tails and a lot of odd framing details. The architect wants closed cell foam (R30) blown in under the roof sheathing with no ventilation. The roof is to be corrugated metal screwed directly down to the plywood sheeting. I am worried about condensation in areas where the foam may not be able to reach (California sections) among other things. Any helpful advice? Thanks, Alex Rockas

  14. User avater GBA Editor
    Martin Holladay | | #14

    Response to AJ Builder
    Did you accidentally post your comment on the wrong page? I don't know what you're talking about. Why are responding to "Martin's rant that Icynene is selling nonsense. ... Icynene is a less green product but it is better at reducing energy demands then Martin is willing to admit."

    I didn't attack Icynene insulation. For the record, Icynene spray foam is excellent at filling framing bays completely and does an excellent job of air sealing. For those reasons it always performs much better than fiberglass batts.

  15. User avater GBA Editor
    Martin Holladay | | #15

    Response to Alex Rockas
    Alex Rockas,
    1. Please define "California framing."

    2. Your architect's plan should perform very well, and shouldn't lead to any condensation problems, as long as the spray foam contractor can reach all areas of the roof sheathing. If some areas of the roof sheathing are not accessible to the spray foam contractor, you need to come up with a plan to insulate those areas.

  16. Iowa builder | | #16

    Ice damming with OC foam
    Good topic I will attempt below paste my post from a different thread.

  17. Iowa builder | | #17

    I am constructing town homes
    I am constructing town homes in Iowa where the zone is 6. The homes are slab on grade so it requires that all the HVAC to be place in the attic. The entire envelope of the home is done with OC foam. The underside of the roof has anywhere from 6" to 8" of foam and putting my HVAC in conditioned space. I am experiencing major snow melt and ice damming on these roofs when the are completely covered with snow with an outside air temp. at as low as 0 degrees.
    I'm feeling I missed a major element in the equation by not sealing the underside of the foam.
    I am still able too apply the sealer and am thinking about spraying it with Suppertherm or a good latex paint. But I do not want to trap the moisture at the roof sheathing or worse yet have it raining in the attic.
    Sorry to be long winded
    Thanks in advance for you comments

  18. Iowa builder | | #18

    Why Does "Green" Require Hazmat Suits
    The hazmat suit is for the over spray it sticks to anything and everything and is in the vapor until it is dry.

  19. User avater GBA Editor
    Martin Holladay | | #19

    Response to Iowa Builder
    Iowa Builder,
    Your roof has at least three problems.

    1. The biggest problem is that you installed less than the minimum code requirement for roof insulation in your climate zone. In Zone 6 (according to the 2006 IRC) you need to install a minimum of R-49 of ceiling or roof insulation. If you choose to install open-cell spray polyurethane foam, that means you need a little more than 13 inches of insulation.

    You write that you have between 6 inches (R-22) and 8 inches (R-30) of insulation in your roof. That's not enough, as your ice damming problems indicate.

    2. Your roof also has thermal bridging problems, since your rafters are uninsulated.

    3. Finally, unless you sprayed the cured foam with vapor-retarder paint, your roof sheathing is at risk of accumulating moisture during the winter.

    To learn more about solving ice dams, see Prevent Ice Dams With Air Sealing and Insulation.

    It's not impossible that your home has an air leak -- for example, at the perimeter of the attic, where the roof foam should join up with your wall's air barrier. Any air leak would certainly contribute to an ice damming problem.

    The solution is to beef up your roof insulation by adding rigid foam on top of the existing roof sheathing (and installing a new roof) or by adding insulation on the interior.

  20. Iowa builder | | #20

    Thanks Martin
    Thanks for the reply. The 6" would be cutting myself short but I included it in my range of thickness.
    I do feel that the thermal bridging is on the trusses you can see that on the frost pattern. I feel comfortable about air leakage for I am getting better blower door test on the total envelope than other homes of the same size. I do have foam over the trusses just not enough I guess primarily where the webs meet the top cord. I will need to correct the problem from the under side they are 4 homes tied together and thats 128 squares of roof. One more thing. It only seems to be bad where there is a large volume of space below the roof and it all freezes at night. I feel I am right at the threshold and sealing will do the trick but don't want two wrongs. I have 86 more homes two do.

  21. Alex Rockas | | #21

    California Framing
    California framing (also called blind valley framing) is when you frame a portion of a roof over a roof that has already been sheeted. Here is a link to a page with a rough sketch. It will be hard to spray foam in those areas. Is it feasible to try and seal those spaces with spray? There will be nooks and crannies that easily could be missed. Also, roofers usually seal the top of the sheeting with Grace fire and ice shield before screwing down the metal roof. It is a very effective vapor barrier. Thanks for your help. Alex

  22. User avater GBA Editor
    Martin Holladay | | #22

    Second response to Alex Rockas
    Determining whether a spray foam contractor can reach a space in order to install foam isn't rocket science; it's common sense. If a guy in a moon suit has enough room to get in there with his spray wand and move his spray wand in a smooth motion without bumping into anything, the work should be possible. When in doubt, talk to your spray foam contractor.

    However, I think you already know the answer to your question. Spray foam is not a good solution when you want to spray the underside of roof sheathing in tight crawl-space areas of attics.

    In your case, rigid foam on top of the roof sheathing makes more sense.

  23. Doug McEvers | | #23

    Iowa Builder
    Do the roof trusses you are using have an "energy heel", are you getting full insulation out over the plate line? My experience in using energy heel trusses, insulation to code or above and venting the attic space at both the eave and ridge, prevents ice dams. My projects have been in locations with 8,000 to 9,500 heating degree days.

  24. Aj Builder, Upstate NY Zone 6a | | #24

    Iowa builder experiences differ from mine
    Scratching my head as to his melting snow. I have had snow melt problems in the past mostly from their being very deep snow on a roof for more than a short period of time. The insulating effect of the deep snow is the issue along with it being on a building that is heated. We have plenty of "unheated" structures here that suffer not from melting snow but do collapse once in awhile due to weight. Lots of snow, rain and ice can take down any building in this area.

    If I ask what is different between our two experiences, the number one difference I see is that I do not have the HVAC or any ducts in my attic spaces. That makes me think they may just be pushing a lot of BTUs into his roof that my homes aren't doing, especially since he is insulated much more than my 5-6 inches.


  25. Aj Builder, Upstate NY Zone 6a | | #25

    Solutions for Iowa Builder type assemblies
    Given that snow is on the roof and is melting and that the HVAC and ducts are in the attic space, vent. Your insulation contractor should have a method that works. If he does not, get another or contact Building Science Corporation and pay for solution that is backed by an insurance policy and contract.

    Your project is certainly large enough to afford more expense to solve a possibly horrendous future liability.

  26. User avater GBA Editor
    Martin Holladay | | #26

    Response to AJ Builder
    Iowa Builder has a conditioned attic with insulation installed between the roof rafters. Your suggested solution is: "vent."

    Obviously, you can't vent a conditioned attic, so I assume you are suggesting that Iowa Builder install 2x4s on top of his existing roof sheathing and vent on top of the roof sheathing. That may help, but it's an expensive solution. If Iowa Builder ends up doing that, I hope he puts down some rigid foam under the 2x4s.

  27. Aj Builder, Upstate NY Zone 6a | | #27

    Different solutions for 4 of the 90 homes he is building
    My suggestions above are mainly to do with his next 86 units. As for the existing units, since I have no issues with snow melt and he does, I think his safest fix for the 4 built units is to add venting above. He could try spraying foam on all the HVAC system too to try to stop the heat flow that I do not have at my homes. I sprayed my ducts in the cellars and love the outcome.

    As to expense, yes... but if the melting is a real issue it must be dealt with. No way would I add foam and make a sandwich of any OSB board. Nope. Never ever ever ever. OSB waits for no one to rot. Junk. Has to be dry. Never sandwiched.

    For the same price as adding foam sheeting he can spray more foam in the attics which like I said should encapsulate all his ducting and even blank sides of his HVAC. Open cell comes off easy if ever a unit needs to be pulled and replaced. Just leave appropriate areas insulated with wrap style insulation.

  28. Aj Builder, Upstate NY Zone 6a | | #28

    How to vent on top
    To vent the first 4 units for the least cost I would go right over the shingles. Add 2x4s then Advantech Zip sheathing taped, then shingle it as per standard methods. The fascia would be lapped and doubled like we do now for thick roof assemblies. Vent added between fascias or choose one of many other set ups.

    Done... problem gone, start building the next 86. Worry not.

  29. Anonymous | | #29

    Great feedback thanks
    1st. Doug yes I have energy trusses and the exterior sheathing runs right too the deck and notched around the tails and foamed full.
    2nd. Aj Yes snow is an insulator but does not melt itself so I feel Martin is on track with the thermal bridging it don't take much water on a roof to make a lot of ice.
    A roof over a roof will not happen on a job of mine, the insulating contractor will correct it or remove his product.
    This is my first experience in using foam. My last project of 26 duplexes and same style build only vented,R-8 on the HVAC and blown fiberglass. Drove thru there the other day and had a consistent 16 to 20 inches of snow on the roofs and not one icicle.
    I wanna be green but I'm not slip sliding on a carbon footprint. ( Just scratching my head here)

  30. Iowa builder | | #30

    Sorry was my previous post
    Sorry was my previous post not anonymous

  31. User avater GBA Editor
    Martin Holladay | | #31

    More on minimum R-values
    Iowa Builder,
    You wrote, "The insulating contractor will correct it or remove his product."

    I think it's a good idea for you to hold the insulation contractor responsible for his work. In preparation for meeting with your contractor, I would suggest that you research your local code requirements for minimum ceiling R-values. It's probable that your contractor's spray-foam job doesn't meet minimum code requirements. If the installation doesn't meet code, this is a negotiating point in your favor.

  32. Brennan Less | | #32

    Duct leakage to outside in homes with ducts in conditioned space
    I thought it worthwhile to mention that moving duct work into conditioned space does not eliminate all duct leakage to outside. We have measured duct leakage to outside in numerous homes that have brought their ducts inside, and they still register a fair bit of leakage to outside. The reason for this is that the "conditioned" areas that the ducts now leak into are also connected to the outside. It is not difficult to imagine this happening in a drop ceiling chase that is imperfectly sealed. Now, this is certainly not as bad as having that air leak directly into an attic or crawlspace, because heat/cool is imparted to the indoor air before the leakage to outside occurs. But there is still an energy and air delivery problem. I'm sure it goes without saying on this website, that envelope tightness is essential in homes with duct work within the conditioned space. A tight envelope will ensure that duct leakage within conditioned space stays within that space. I add this comment because it is often thought that once ducts are in conditioned space, there is no longer any need to seal the ducts with tape/mastic. So, if you move duct work into conditioned space, be sure that the conditioned space and your ducts are well-sealed, otherwise you've accomplished little. Cheers.

  33. Danny Kelly | | #33

    Creating a conditioned attic without a HVAC unit in the attic?
    Seems to me, the only reason to create a conditioned attic is to bring the HVAC system and ductwork into the thermal envelope of the home. If there is no HVAC in the attic - why not just air seal and insulate at the attic level - much easier and less expensive. I am not a HERS rater but I would think adding all of that volume and extra surface area to the thermal envelope would actually decrease the efficiency of the home. Even if you are a foam lover - can't you spray your foam at the ceiling level? Any HERS raters have an opinion on how this would affect the rating?

  34. John Brooks | | #34

    Bingo Danny
    keeping the thermal envelope at the attic floor level does keep the exterior surface area lower...
    and it will use less energy.
    It will not change the HERS index because the HERS index will compare the house envelope to it's own geometric twin.

    I think the best solution is to keep the envelope at the attic floor with airtight drywall and use more thickness of less expensive insulation.
    Increasing the thickness and r-value will improve the HERS index

  35. User avater GBA Editor
    Martin Holladay | | #35

    Response to Danny Kelly
    If you don't have any HVAC equipment or ductwork up there, and you don't intend to use your attic for living space, then it is clearly best to leave the insulation level on the attic floor. I agree with you and John Brooks: that makes air sealing easier and insulating cheaper.

    It also makes it easier to spot and repair roof leaks, which is a big plus.

    I've added a few sentences at the end of the blog to address these issues.

    Thanks again for your comments.

  36. Danny Kelly | | #36

    Thanks Martin
    Thanks for the clairifcation. I thought your article was clear by the way - was mainly responding/questioning AJ's post/practice and wondering if would afect HERS. John - thanks for the reply.

  37. M. Johnson | | #37

    Construction details you alluded to
    Martin you said "Even if you insulate between your rafters with an air-impermeable insulation like spray polyurethane foam, you may want to provide a ventilation channel under your roof sheathing. The main function of such a ventilation channel is to separate the roof sheathing from the foam; this facilitates future repairs of sheathing rot."

    Do you have any articles with more discussion of that? Something with pictures is most welcome. I have been looking for a comprehensive discussion of how this might be done with low probability of trouble in a hot-humid climate.

    Thank you -- Mark Johnson

  38. User avater GBA Editor
    Martin Holladay | | #38

    Response to Mark Johnson
    Mark Johnson,
    No, I can't think of a specific article that discusses the concept.

    However, people have been building site-built ventilation chutes in rafter bays for many years, so we aren't discussing a new technology or technique.

    Although these ventilation chutes are usually used with fiberglass or cellulose insulation, there is no technical reason they can't also be used with spray polyurethane foam, as long as the panels are rigid enough not to collapse under the pressure of the expanding foam.

  39. Aj Builder, Upstate NY Zone 6a | | #39

    My spray company will do vented roof decks
    My spray company, as I remember said they would spray right over the store bought foam vent chutes. For odd sized spacing one would have to make these fit well or make your own like has been posted many times on this site. He had no issues with venting by the way, though I really want to try an unvented non heated attic assembly and it is working beyond my expectations. Much better than detractors state it should. Open cell if a tiny bit of water gets to sheathing will let the sheathing dry. That is a fact. Much less chance of rot verses closed cell. If I went closed cell I would purchase pressure treated sheathing. It would be worth every penny.

  40. Mark Parlee | | #40

    Spray foam on the roof deck
    I wrote an article that appeared in the JLC on Repairing a Rotting Roof June 2010 that dealt with a few shortcomings of the Icynene, when not installed considering all of the parameters. While this was not just an insulation problem but an indoor humidity problem as well, some things could have been done to alleviate this problem at the time of original construction.
    Iowa Builder feel free to get in touch with me for additional perspective.

  41. Green Mountain Realty | | #41

    Making the Attic a Conditioned Space
    Taking the building envelope to the roof is a good approach and will save energy - another approach is to use radiant coated deck sheathing, seal all ceiling attic penetrations and blow in R-60 insulation as it will cost less than spray foam and should give at least the same energy efficiency to the home.

  42. A.K. Harrison | | #42

    What about thermal bridging in the ceiling joist
    If I used 2x10s would I spray cellulose to the top of the joist and add an extra 2 inches? Would the 2 inches be enough to eliminate thermal bridging? What is the best barrier to use between the ceiling joists and the plywood for cellulose?

  43. User avater GBA Editor
    Martin Holladay | | #43

    Response to A.K. Harrison
    Q. "If I used 2x10s would I spray cellulose to the top of the joist and add an extra 2 inches? Would the 2 inches be enough to eliminate thermal bridging?"

    A. If you are using spray cellulose, you would need to install framing members perpendicular to the rafters to address thermal bridging. Two inches would go a long ways toward addressing thermal bridging.

    Q. "What is the best barrier to use between the ceiling joists and the plywood for cellulose?"

    A. I'm not sure what you mean. If by "joist" you mean "rafter," and if by "barrier" you mean "ventilation chute material," the answer is "plywood, fiberboard, or rigid foam." If you mean something else entirely, please explain.

  44. Gerard Celentano | | #44

    You need to cite your sources
    Sorry for long post. I find this article a bit more informative than previous ones but the opening premise is flawed (IMHO) and hence doesn't seem worthy of much discussion. I find it curious that I asked a question touching on some of the topics in this article a few months back. The advice you guys gave seems to contradict what this article is saying today.

    I have a problem with opening premise is that it costs "homeowners dearly". I'd like that quantified since I'm pretty sure that implementing the recommended upgrades would cost homeowners dearly.
    I tend to rely on my experience so forgive me if I can't quote DOE or IRC. I have a house in the mid Atlantic with the very scourge of energy efficiency you refer to (leaky air handler in the dreaded uninsulated attic/crawlspace). Your definition of costing "dearly" is about $1000 extra to run the AC in the summer. I can't see how I can effectively insulate the attic (as you suggest) with an ROI of less than 7 years. So even though it's costing me dearly, I have better places to put my resources.

    I have another house in New England the same energy scourge and I paid big bucks to spray foam the rafters. I calculate that I save about $0 (that's Zero) in AC costs. Maybe it's a hundred bucks a year, but with rising energy costs my ROI should be sometime before I expire (I'm in my late 40's).

    Your delta T argument goes down the right path but is meaningless without duty cycle; ie. it may only be for short duration. This is why a frugal New Englander (or one good at math) would not insulate his attic to save cooling costs. Even my house near the Nation's Capital peaks at about 50% during the hottest 6 hours of the hottest days. Thank goodness it's oversized by today's standards because it's a much cheaper solution to solve the issues raised in your opening argument than the ones you proposed.

    I hate to pick; but your thermal bridging argument only makes sense in the context of an opportunity cost or comparative analysis. Diminishing returns are valid issues that real people have to struggle with. Instead of glittering generalities you need to quantify how much energy you'll save jumping from R1.5 to R20 in contrast to R20 to R50? Then quantify the perceived comfort, up front and recurring costs, and ROI to substantiate your claim.

    Thank you

  45. Scott Sanders | | #45

    Good points Gerard

    I have to agree with the idea behind you points on the cost effectiveness of retroactively adding insulation to existing houses and also looking at the incremental savings of additional energy saving strategies in new construction (such as 2" of under slab insulation vs 16"). Regarding existing houses, a cost/benefit analysis definitely should to be performed, and unfortunately, the benefit is frequently not realized within the desired time frame (before an owner wants to sell the house for example).

    The two "devil's advocate" points that I would like to mention are
    1) The externalized costs of heating/cooling your house (the ambiguous environmental impacts of remote power generation and transmission) are not reflected in the price you pay for gas and electric. Therefore the cost benefit analysis may work out in more situations than we think.
    2) In new construction the incremental costs of increased energy efficiency seem to intuitively be cost effective because the labor costs will have little change. I can't name a study off the top of my head, but I bet there is one out there from LBNL or Building America that backs this up. There is of course a tipping point where they are cost in-effective, but I don't know of any energy modeling software that can accurately help determine this point.

  46. User avater GBA Editor
    Martin Holladay | | #46

    Response to Gerard Celentano
    Thanks for your post; you bring up many important issues.

    First of all, the point of my article is to provide advice to anyone who wants to create a conditioned attic, with the aim of making sure they do a good job. I never intended to encourage people to install insulation between the rafters of an existing home. That's why I wrote, "Creating a conditioned attic is a solution to a fundamental design flaw (locating ductwork or HVAC equipment outside of a home’s thermal envelope). But conditioned attics are not unalloyed blessings; they come with their own set of drawbacks."

    According to your own estimates, this fundamental design flaw is costing you $1,000 a year. For you, the investment in insulation to create a conditioned attic isn't worth it, because the payback period is 7 years. That's your decision. Some other homeowners are very happy to make investments in their homes with a 7-year payback.

    I am happy to stipulate that many energy-efficiency measures don't make much sense when analyzed in terms of payback. The main reason for these long payback periods is the fact that fossil fuels are quite cheap, and are sold at prices that don't include external costs like damage to the planet's climate.

  47. T Conover | | #47

    rigid foam under rafters - fire protection required?
    Thanks for the article. If you install the rigid foam under the rafters in an unoccupied/non living space attic, does the foam need to be covered with drywall?

  48. User avater GBA Editor
    Martin Holladay | | #48

    Response to Troy Conover
    The answer to your question can be found in section R316.5.3 of the 2006 IRC. If the attic space is "entered only for purposes of repairs or maintenance," you can dispense with the requirement for a thermal barrier and instead install an ignition barrier (a less stringent option than a thermal barrier). But ignition barriers aren't nothing -- many people choose drywall, which must be at least 3/8 inch thick to qualify as an ignition barrier.

    If you install Thermax foil-faced polyiso, some building inspectors will allow you to dispense with the requirement for an ignition barrier.

  49. Joell Solan | | #49

    rigid between rafters
    Thanks for thinking so much about insulation, i'm still new to thinking about this so please bear with me.

    I live in the Northwest, am planning to gradually finish a 600 sq foot attic space with cathedral ceilings, new roof put on about 5 years ago, before we moved in. my rafters are only 2x4 and i don't want to lose much living space. My tentative plan has been to put 2-2.5 in. polyiso between and 2 in. under rafters for now. This won't get us to code but would be better than what we have now. now i've been thinking that someday the roof will need to be replaced and i could add rigid on top.

    Which brings me to my questions:

    if "ventilated rafter bays are incompatible with rigid foam insulation installed on top of the roof sheathing" does this approach make sense?

    and since poly-iso is faced, can i use 2 layers on top of each other and if so do i create a sandwich with the faced surfaces or layer them with facing to the interior? thanks

  50. User avater GBA Editor
    Martin Holladay | | #50

    Response to Joell Solan
    Don't bother to include ventilation. Instead, fill the 3.5 inch rafter bays with 3.5 inches of polysio (one 2-inch layer, and one 1.5-inch layer). Seal all gaps with canned spray foam. Then install a continuous layer of 2-inch polyiso under the rafters to stop thermal bridging.

    Using several layers of foil-faced foam is fine -- there is no problem with making a sandwich like this, except that it is slow, fussy work.

  51. Gary Steinfeld | | #51

    Foil Faced Foam for Venting Channels
    In a 2x12 rafter attic, can I use foil faced foam (Dow Thermax) to create the 2" air channels in the rafter bays and then use fiberglass batt in the bays followed by sheetrock with latex paint. Will the low perm of the thermax cause any problems at this location. Also, shoud I add some rigid foam such as dow blue board below the rafters for a thermal break against bridging: say 1" foam?

  52. User avater GBA Editor
    Martin Holladay | | #52

    Response to Gary Steinfeld
    Yes, you can use rigid foam (for example, Thermax) to create a ventilation channel.

    If you have 2x12s with a 2-inch ventilation channel, 1 inch of Thermax will give you an R-value of R-6.5, and the remaining 8 1/4 inches of fiberglass will give you about R-30.5. So you end up with R-37. I don't know if that is enough to meet minimum code requirements; it depends on your climate zone and the local code.

    Yes, you can cut down on thermal bridging through the studs by adding a continuous layer of rigid foam insulation under the rafters. I don't think this type of assembly will have moisture problems.

  53. Gary Steinfeld | | #53

    Kraft Free Insulation
    My intention was to use batt insulation with out kraft paper. Is that the correct detail?

  54. User avater GBA Editor
    Martin Holladay | | #54

    Response to Gary Steinfeld
    I don't know your climate, your geographical location, or your local code requirements, so I can't answer your question.

    Your detail has to satisfy your local code official, so you can call your local building office for an answer. It rarely matters from a building science perspective whether your fiberglass batts have kraft facing or no kraft facing, as long as you pay attention to air sealing.

    Here's more information on vapor retarders: Vapor Retarders and Vapor Barriers.

  55. David Head | | #55

    Venting Cathedral Ceiling
    Like the previous poster I am thinking of using foil faced 1 inch polyiso for a site built ventilation baffle for 12 inch deep rafters, creating a 2 inch gap. Then spray foam underneath. Should I be worried about vapor condensing on interior foil so that the foam absorbs moisture like a sponge? Is one type of foam better than a another? The house is in climate zone 4c in Northern Ca. Thanks for you time.

  56. User avater GBA Editor
    Martin Holladay | | #56

    Response to David Head
    Q. "Should I be worried about vapor condensing on interior foil so that the foam absorbs moisture like a sponge?"

    A. No, for three reasons: (a) The foil layer, even if it is wet, prevents the water from reaching the rigid foam; (b) The foil will be warm (since it is on the interior side of the foam) and will therefore never be cold enough to become a condensing surface; (c) In most cases, the spray foam will stop moisture transfer from the interior to the foil.

    Q. "Is one type of foam better than a another?"

    A. Closed-cell spray foam has a higher R-value per inch and is less vapor-permeable than open-cell spray foam; for those two reasons, I would say that closed-cell foam is better than open-cell foam in this application. But either one will work.

    For more information on this topic, see How to Build an Insulated Cathedral Ceiling.

  57. David Head | | #57

    Cathedral Ceiling Venting
    Thank You very much for you quick reply, you have relieved some of my anxiety of this project. I take that "No, for three reasons: (a) The foil layer, even if it is wet, prevents the water from reaching the rigid foam" also means that water vapor can not entrap itself inside the spray foam and lead to rotting or mold growth on the rafter? Does the vapor just move to dissipate back into the conditioned space?

    Sorry for the basic questions, I just do not want to make any mistakes and understand the building science.

  58. User avater GBA Editor
    Martin Holladay | | #58

    Response to David Head
    During the winter, the indoor air is warm and humid. The outdoor air is cold and dry.

    The moisture held in the air indoors isn't supposed to go anywhere. It is supposed to stay put. You don't want to have any air or water vapor migrating through your building assemblies. That's why you have an air barrier -- to keep the indoor air in, and the outdoor air out.

    Q. "Can water vapor entrap itself inside the spray foam?"

    A. No. Closed-cell spray foam is vapor-impermeable. While open-cell foam is vapor-permeable, any water vapor inside the foam isn't trapped. It's free to come and go. Water vapor doesn't hurt spray foam.

    Q. "Does the vapor just move to dissipate back into the conditioned space?"

    A. Are you talking about outdoor water vapor during hot, humid weather in the summer? If you are, the water vapor should stay outdoors. You can make sure that happens by including an air barrier in your ceiling assembly -- in your case, that would be the foil-faced polyiso. The foil is also a vapor barrier.

    If you are talking about water vapor that is inside the house during the winter, the idea is to keep it inside, not allow it to flow through your walls and ceilings.

  59. Eric Schroeder | | #59

    Creating a Conditioned Attic
    Mr. Holladay,
    I am writing to you in hopes you that you may possibly add some clarity to a confusing situation for me. While renovating my house a couple of years back I used this sight for a lot of the recommendations in regard to insulating my house. I am in Westchester NY about a mile from the coast zone 5. The house is a wood frame ranch with a pitch of 6/12 with 2x8 rafters and new ¾ in sheathing. My main concern is the attic and roof insulation. Unfortunately, I have Hvac equipment in the attic and I might finish the attic in the future too. There is floor decking in the attic with R-30 bat between the ceiling of the first floor and the attic floor. The sheet rock ceilings were made air tight or as close as possible. I have a sealed door to the attic. My question is this, as per a number of articles and readings (one of which was “Five questions to consider before insulating a sloped roof”) I decided that I would have the contractor create an on-sight baffle to vent the underside of the roof sheathing by using 1x1 and 1” xps foam board to create a continuous channel from soffit vent to ridge pole vent. I then intended to either flash foam with closed-cell and then another type of insulation after that or fill the rest of the 2x8 with closed-cell. Either way I believe I would be short the required code value for insulation (R-38 zone 5?), but thought I would be close and the tighter envelope would still be better over all situation. The problem I have is that all of the South/East facing side of the roof has the 1x1 baffles installed, the other side was never finished, but looking to complete everything now. The info in the last 2 or 3 years has seemed to change (Dr. Joseph Lstiburek article about roof venting). My concern now is that the 1x1 channel with foam board is not large enough even though the IRC says 1” is the minimum. I believed from what I read that 1” would be fine especially since I wanted to give up the least amount of R-value as possible while doing it. The idea was minimum amount of R- value lost, no foam against the deck, and the attic included in the envelope. Now I’m reading all these negative things about potential for mold and other bad things if the channel under the sheathing is less than 2”. So do I spray over the existing baffles and cross my fingers or rip out the baffles and spray against the sheathing? I could try to take the foam board down and increase the channel size up to 1.5 “or 2” , but I would lose some more R-value? What would you do at this point. Any information would be greatly appreciated. Forgot to mention that I am having that side of the roof spec’d out for PV array, if that makes any difference?
    Thank you,

  60. User avater GBA Editor
    Martin Holladay | | #60

    Response to Eric Schroeder
    1. I wouldn't worry about whether or not your 1-inch deep air channel is adequate. It is. A great many attics have smaller air gaps than the one you have provided (for example, those that use Styrofoam Proper-Vents). You have done an above-average job.

    2. The most important way to ensure that there are no moisture problems in your ceiling assembly is to pay attention to airtightness. Because you plan to use closed-cell spray foam, you are unlikely to have problems on this score, as long as the spray foam installer pays attention to maintaining air barrier continuity near the soffits, where the air barrier in the sloped roof assembly meets the air barrier in the wall assembly. Since you are worried on this issue, you can always hire a contractor with a blower door to depressurize your attic when the work is done; that way you can check for air leaks and fix them.

    3. If you fill the rafter bays with closed-cell foam, you'll end up close to R-38. However, most closed-cell spray foam contractors don't fill framing bays (because the foam is dense and hard to trim when it bubbles past the plane of the framing) -- the framing bays end up "almost full." You might want to consider adding a layer of rigid foam on the interior side of your rafters, to address thermal bridging and to beef up the assembly's R-value.

  61. Chris Thomas | | #61

    Roof Assembly without sheathing
    This seems to be a common theme with my home - I asked a question in the 'Cut and Cobble' article. I need to create a conditioned attic due to the location of the air handler and ductwork. It would also be a great help to use it as storage as it's an old home with limited closets, etc. I'm located in Climate Zone 4. The home has a metal roof on perlins. I cannot insulate from the exterior. The roof has no ventilation (no soffit vents/ridge vent) except from holes from those pesky squirrels.

    My hope was to use fiberglass between the rafters and foil-faced foam on the underside. I don't think I would need to worry about wind-washing in the roof assembly.

    Any suggestions on how to properly condition this type of attic would be greatly appreciated.


  62. User avater GBA Editor
    Martin Holladay | | #62

    Response to Chris Thomas
    Everything you need to know is explained in this article: How to Build an Insulated Cathedral Ceiling.

    Here are the steps for the type of roof you describe:

    1. The air between the purlins (under the metal roofing) provide enough ventilation.

    2. The first step is to create an air barrier at the top of each rafter bay. You can either install rectangles of plywood, OSB, or thin XPS or EPS foam. No matter which product you choose, the rectangles have to be air sealed at the perimeter with caulk, canned spray foam, or a high-quality construction tape.

    3. You can insulate the rafter bays with fiberglass batts, blown-in cellulose, mineral wool, or spray foam. If the rafter bays aren't deep enough for you to get the minimum R-value of insulation required for your climate zone, you may have to extend the rafters inward (make them deeper) with scabbed framing members.

    4. It's important to install an air barrier on the interior of your roof assembly. The simplest air barrier at this location is gypsum drywall.

  63. Joe G | | #63

    New Roof & Venting Options
    I'm currently in the process of working with a roofing contractor for a new roof and thus looking into the ventilation options. Here are the details:

    - We would like to use stone coated steel shingles - for a variety of reasons
    - We currently only have 2 gable vents for attic ventilation and a large open attic that we would eventually like to convert to living space
    - We have no soffit or ridge vents and the eaves are stucco
    - We have cathedral ceilings in about 1/3 of the house, and while the roof is not extremely complex, it does have some hips and valleys, and 2 eyebrows.
    - We have recessed can lights everywhere - and in the cathedral ceilings
    - I'm pretty sure there's some sort of insulation in between the cathedral ceiling and the roof - most likely fiberglass
    - Los Angeles area
    - Built in 1975 - we just bought the house last year

    Since the metal roof would need adequate ventilation, and venting the attic would not easily allow us to convert it later, in addition to the area above the cathedral ceilings needing to be vented (for the metal), we are looking at alternate options to vent the roof. I am aware that adding recessed lighting in an insulated ceiling is completely backwards. The previous owner added the cans and we'd like to keep them - especially if using one of the two options below negates need for the cathedral ceilings to be insulated.

    So it seems we have two options:

    1. We can use the flat metal shingle we like best (they are only a direct to deck install) and add a second roof deck over the existing deck with battens in between to create an air space that we vent. The existing roof sheathing would have a layer of SharkSkin underlayment and radiant barrier. The new sheathing (2nd layer) would just have an underlayment.

    2. Create a counter batten system with a raised metal shingle and radiant barrier / SharkSkin.

    Both options would allow us to convert the attic and should cut down on a considerable amount of heat transfer into the attic. There are a few ways we can vent the roof near the eaves with venting options from cor-a-vent. We will also be adding 3 ridge vents. The shape of the house / roof is basically a U, though upside down and with hard corners. There's 3 parts to the roof, the largest which goes left to right and is the horizontal section of the U shape. The other 2 parts are the vertical extensions off the U shape and intersect the longer section. There's a space in between the vertical sections (our front walk). Then there are 2 small eyebrows, one on each of the extensions that faces out toward the street. I realize even with the above sheathing venting it will be difficult to completely vent the roof.

    Since we would much prefer to use the flat shingle, we'd be looking to add the second roof deck. The issue then becomes the added weight of the second roof deck (and the solar system we plan on adding ~820 lbs), and this being earthquake country, the lighter the roof the better. This is beyond the scope of your article but I mention it only to illustrate that we are having difficulty deciding which route to take and maybe you can offer an opinion unrelated to the weight issue that will steer us in one direction over another.

    I apologize for such a long comment - but I'm finding it very difficult to find anyone who really seems to understand all this, and would very much appreciate any insights and advice you might have.

    Thanks for the great article.

  64. User avater GBA Editor
    Martin Holladay | | #64

    Response to Joe G
    It seems to me that you are too concerned about venting, and not concerned enough about R-value.

    For more on venting, and why it isn't that important, see All About Attic Venting.

    I have never heard that metal roofing needs a ventilation channel. Is this a requirement of the roofing manufacturer? Here in Vermont, many builders install metal roofing over unvented roof assemblies.

    It sounds like your roof assembly is underinsulated, especially in the cathedral ceilings with recessed cans. The obvious solution (if you are reroofing) is to install a thick layer of rigid foam above your existing roof sheathing. For more information on this option, see How to Build an Insulated Cathedral Ceiling.

    If you want to install added insulation and a new vent channel, one possible solution is to install vented nailbase panels like the Cool-Vent product from Hunter Panel.

  65. Joe G | | #65

    New Roof & Venting Options

    Thanks for the response.

    I should add by the way that - you guessed it - our ducting is all in our attic.

    Yes, the metal manufacturer specifies that there should be adequate ventilation - not necessary directly below the metal shingles but at least in the attic from soffit to ridge. The company is Metro Roof Products. I'm assuming this is to prevent condensation issues.

    As much as I would love to use a rigid foam like Polyiso, I have many reservations due the inclusion of known carcinogenic flame retardants like TCPP and the fact that has a newly published / revised paper from January 2014 - updated in March ( which essentially states that TCPP is completely and utterly safe all around, based on a 2008 study. It goes on to quote the agency study saying there is no need for further testing as a result of the 2008 tests. There have been numerous studies as of late classifying TCPP as carcinogenic and it is listed on CA's Prop 65. The fact that Pima recently published this brief and the way the information is laid out sounds like propaganda and I just can't get behind such a product or organization.

    If we vent the roof deck or added insulation above the deck -- 1. Would it also be necessary or recommended to remove the existing insulation (if there is any) from the cathedral ceilings? We have plenty of insulation in the rest of the attic floor and walls between the cathedral ceilings and attic - which would most likely be moved to the ceiling if we converted the attic.

    I read your article on mineral insulation. Owens Corning Thermafiber is a mineral wool / rock wool product I just found that seems fairly safe. Like the Roxul product it seems to use a formaldehyde binder, according to their FAQ ... 12 ppb (

    I've been looking for the study Owens Corning was supposedly doing on Thermafiber UltraBatt, mentioned here:, and the formaldehyde levels but it doesn't seem like it's been released yet. Thermafiber has many different products and some of the MSDS list Phenolic Resin with no mention of formaldehyde, while others list mineral oil where the Phenolic Resin was listed, and then the MSDS for the SoundZero product lists Formaldehyde. I can't even find an MSDS for UltraBatt:

    It says Thermafiber can be formed into boards. 2. Would you recommended something like Thermafiber or Roxul for roof insulation? It looks like you can even get the Thermafiber with a vapor retarding foil facing.

    I should add that when we convert the attic we would of course insulate the ceilings and remove the attic floor insulation. Thus, we'd have a vented roof and insulated ceilings. Of course the Cathedral Ceilings would still be an issue. I wonder if it might make more sense to address the Cathedral Ceilings instead of adding insulation above the roof sheathing for the entire roof, when we have rafters in which we can insulate below - and they cover more of the house. I'm assuming there's no real way to insulated the cathedral ceilings under the roof sheathing, but above the drywall, WITH the recessed lighting. And from your article on that, it seems like the air gaps around the cans could cause some moisture issues on the insulation in there etc., especially since there's no venting.

  66. User avater GBA Editor
    Martin Holladay | | #66

    Response to Joe G
    Q. "[If rigid foam or mineral wool were installed above the roof sheathing,] would it also be necessary or recommended to remove the existing insulation (if there is any) from the cathedral ceilings?"

    A. No.

    Q. "Would you recommended something like Thermafiber or Roxul for roof insulation?"

    A. According to my understanding, mineral wool panels are only installed above roof sheathing on low-slope or flat roofs. Use of this product on a steep-slope roof would be considered experimental.

  67. Nick Schuler | | #67

    Conditioned Attic Needs a Little Air
    Hello, posted on here before about my 1800 sqft attic in Nebraska that was transitioned from a blown cellouse vented attic, to an icynene open-cell unvented space. Initially humidity was very high during sunny afternoons. After posting on here for advise, and speaking with the insulation company, the company came back filled gaps, made the insulation thicker, and used a spray on vapor barrier. They even paid for a local energy inspector to check it afterward. The energy inspector thought the work was really well done. Since then in the spring and fall in the late afternoon when the HVAC is off and the windows of the house are closed it can reach 80% RH at the very peak point of the attic which is 8ft to central peak from atttic floor. However as the sun sets the humidity goes back down to about 55% RH at night. If the windows are open in the house or the HVAC is on its lower, and in the winter it stays at 51% all day. The energy adviser recommended two options for the afternoon humidity and to provide a bit of air to the space. Option one was just a small ventilation feed from the house into the attic. Option two was placing a bathroom exhaust fan in the space that vents outside, and has a humidstat switch that turns it on when needed. The feed from the house would be difficult since the duct work is in the basement. What are your thoughts on the bathroom exhaust fan? Should I look for HRV or CRV system?

  68. User avater GBA Editor
    Martin Holladay | | #68

    Response to Nick Schuler
    The phenomenon you are describing -- high RH in an unvented conditioned attic after open-cell spray foam was installed on the underside of the roof sheathing -- is common. That's why my article notes, "Recent research suggests that open-cell spray foam may be risky in all climate zones, so the safest spray foam insulation to use is closed-cell spray foam. For more information, see Open-Cell Spray Foam and Damp Roof Sheathing."

    In that article, I noted, "If you have an unvented conditioned attic with high humidity problems, the best solution is to install a supply-air register connected to your forced-air HVAC system in the attic (and in some cases, a return air grille as well). If the attic is heated in the winter and air conditioned during the summer, humidity problems should disappear."

    I hope this solution works for you. Good luck.

  69. User avater
    Dana Dorsett | | #69

    Try a tiny ventilation fan running continuously.
    A Panasonic FV-05-11VK1 set to 50cfm draws 3.2 watts. If you set one up to draw from the attic and exhaust into the conditioned space, with a return grill to the attic from some other place the peaks in attic humidity will be managed by the air conditioning.

    If 50cfm doesn't quite do it during the summer it can be cranked up a bit. At 80cfm it draws 5.4 watts, at 110 cfm it draws 9.8 watts.

    It's available at some box stores for about $135. You'll have to buy a small exhaust grille/diffuser, a tiny amount of 4" round duct, and a return path grill too, but for under $200 in hardware plus some electrician work you'd be there.

  70. Ben Reese | | #70

    Rigid Foam Board Under the Rafter
    I am in a HOT climate near the Gulf Coast in Texas with the air conditioning unit is in the attic. My thought is to add 2 inch foil wrapped foam board under the rafters. The space between the rafter would be left open to allow air to flow from the soffit vent to the top roof vent. I would seal the air space between the rafter from the airspace in the attic. I have NO combustion in the attic from appliances (all electric). Also plan to leave the existing insulation on the ceiling below. This will create a space that is less than the current 150 F and over in the summer to something cooler but not as cool as the conditioned space under the insulated sheet rock ceiling. My question if you get comfortable with these plans is which ridged board material is best for this application. Keep in mind freezing weather is not a major factor in this climate. Part of the driver here is comfort, when the AC starts the first air out is hot, not cooler than the existing house. And to be clear i have installed the newest insulated duct work available however the broad side of the roof faces west and catches the major afternoon heat from the sun. And suggestions are welcome to improve the heat.

  71. User avater GBA Editor
    Martin Holladay | | #71

    Response to Ben Reese
    The best type of rigid foam to use for your project would be foil-faced polyisocyanurate. Make sure that you seal the polyiso seams with a quality tape (for example, housewrap tape).

    The approach you describe can't hurt -- R-12 insulation is better than nothing. But clearly, R-12 insulation is a far cry from code-minimum insulation.

  72. User avater
    Dana Dorsett | | #72

    How deep are the rafters? (@ Ben Reese)
    If you have 2x6 rafters you can install R13-15 batts between them, and add the 2" polyiso below to improve the performance. (Don't go with lower-density R11s, since at low density the translucency of fiberglass to infra red radiation renders it pretty useless during the cooling season.) If you have deeper rafters than that you can go even higher. If the roof pitch is ~3:12 or lower you'll want at least 2" of air space between the fiber and the roof deck, but at higher pitches it can be the code-min 1" gap.

    Foil faced polyiso is fairly fire-safe relative to polystyrene foams, but you may be required to use a fire-rated polyiso (eg Dow Thermax) if leaving it uncovered on the interior. Local codes vary.

    If you're going to skip the batts, you get a modest performance boost out of low-E "radiant barrier" paint with an IR emittance less than 0.25 on the underside of the roof deck. (see: ) If you're installing the batts that benefit is pretty small, but if it's facing a shiny foil facer it's real enough. Not all low-E paint is the same- be sure to look up the specifications for both emittance and coverage. Mop-on exterior coatings are typically higher emittance and only cover a few hundred square feet per 5 gallon bucket, and they're pricey. A 5 gallon bucket of the better class low-E under-deck paint starts at about $300 and covers ~2000 square feet, which is 15 cents per square foot. But if you don't have 2000 square feet of roof deck to paint it's higher on a per-square foot basis, and you're stuck with the leftover paint to dispose of.

    Bang per buck, R13 batts run 35-35 cents per square foot at box store pricing, but you don't need to buy 2000 square feet at a time, and the performance boost is higher.

  73. Ben Reese | | #73

    Conditioned Attic - Hot Climate
    Thanks for the responses, first the rafters are 2x6. I will look for foil faced polyisocyanurate board. And while this will this is less than code the insulation on top of the ceiling is 21 inches and meet code for the enclosed space. Again i am really just trying to make the AC system more comfortable due to all the long runs of ducting in a space that is over 160F today. My concern on adding insulation between the rafters is the potential to block adequate air flow under the roof deck. My shingles are fairly old and would prefer to no do anything to reduce the life of the shingles. Again I could be all wrong on my assumption so any help would be most appreciated.

  74. Dan H | | #74

    Mini split system for conditioned attic renovation?
    Thank you for the great articles, Martin. I'm working on an attic renovation project in Chicago where we are converting a previously unfinished and uninsulated attic into a living space with 2 bedrooms and 1 bathroom. The existing space has 8 passive roof vents installed between rafters close to the ridge and the roof does not have any eaves. After reading this and several other of your articles I have a couple questions that I'm hoping you can help me with. For context, the existing rafters are 2x6's and the gable ends are double wythe brick walls that have been inspected, are not good candidates to insulate, and are therefore being left exposed. Because there are no eaves and the roof has multiple valleys, we are spraying 3" of closed cell spray foam against the existing sheathing to control condensation, compressing a 3.5" batt into the rest of the rafter cavity, and then capping the interior of the rafters with a continuous 1" rigid insulation board. The final interior finish will be painted drywall. It would be difficult to extend the existing ductwork for the forced air furnace and AC into the attic (but we could do it if it made the most sense) and we are therefore considering a mini split system with 2 zones for dedicated heating and cooling for the attic. In total the attic is roughly 900 sq ft and the larger bedroom will be 400 sq ft when done.

    My first questions / concerns with this plan are regarding humidity and fresh air:
    I have not yet personally installed or lived with a mini split system but I have read that it should be able to alleviate any excess humidity that collects in the attic space during the summer. Is that accurate? Would another system be more appropriate? If we are not planning to extend the ducts or add additional ventilation, should we be concerned about a lack of fresh air in the bedrooms? The stairs will be completely open to below but one bedroom is 30' away and we don't have headroom to add a ceiling fan for additional air circulation. It's such a small space that perhaps it won't be an issue, but I'd hate to have the space feel stale when it's complete.

    My second question is concerning the existing passive roof vents that we are planning to remove in order to turn the roof into an compact unvented assembly with the close cell foam as I describe above. Should we be concerned that this plan could lead to interior moisture problems that were not present before when the roof was so freely "breathing"? There aren't any signs of moisture in the basement or existing walls, but Chicago can be quite humid and we would need to remove all of the passive vents in order to make the space livable as an unvented roof is really our only option.

    I really appreciate your insight.
    Thank you.

  75. User avater GBA Editor
    Martin Holladay | | #75

    Response to Dan H
    Your proposed insulation will total about R-31, which is probably less than minimum code requirements. That's not enough to invalidate your plans -- just something you should know.

    Sealing the roof vents (as part of the work to convert a vented unconditioned attic into an unvented conditioned attic) will not put your roof sheathing at risk. As long as your roofing and roof flashing are in good shape, your roof sheathing will be fine. Eventually, your roof will leak -- but that happens to all roofs.

    The indoor air in your attic shouldn't be any worse than the indoor air elsewhere in your house. During the summer, your minisplits will cool and dehumidify the indoor air.

    The question on ventilation is hard to answer. Whether or not the top floor of your home feels stuffy depends on several factors, including (a) whether you have air leaks on your gable walls or near your windows, (b) whether you like to open your windows a crack when you sleep, (c) whether or not your house has a mechanical ventilation system for the lower floors, and (d) the quality of the indoor air on the lower floors.

    If your house is tight and poorly ventilated, the solution is to install a ventilation system. If your top floor is stuffy, a good solution might be to install two Lunos fans. For more information, see Designing a Good Ventilation System.

  76. Dan H | | #76

    R-value followup question
    Thank you for the quick response, Martin.

    I'm curious why the R-value would be 31? I was estimating 36.5 with the following:
    3 x 6.5 for the closed cell = R19.5
    A 3.5" R15 compressed to 2.5" = R11 (per owens corning)
    1" polyiso board = R6

    Am I making a mistake in that calculation?

    We will most likely need to sit down with a code official on this issue in order to get our permits. The Chicago code has an exception where up to 500 sq ft can be less than the required R-49 (down to R-30) if existing conditions prohibit more insulation. And, the home owners really don't want to use more closed cell spray foam than we have to. We are maintaining as much of the original roof as possible which is why head room is at a premium and we don't have any space to furr down to add more insulation. The 1" board below the rafters is already cutting it close. The shingles are only a couple years old or else I would encourage them to insulate from the exterior but that would add a huge cost and destroy the budget.

  77. User avater GBA Editor
    Martin Holladay | | #77

    Response to Dan H
    Your R-value estimate may be more accurate than mine. You wrote "rigid foam" in your original post, so I assumed R-5 for the (unspecified) material -- polyiso probably would be R-6.

    I assumed R-6.25 per inch for the spray foam -- but R-6.5 is possible.

    I assumed R-9.5 for the 2 1/2 inches of compressed fiberglass -- probably more pessimistic than necessary, but compressed fiberglass is rarely perfectly installed.

    Then I made an addition error, reporting R-31 when I should have reported R-33. (Sorry.)

    So the ultimate R-value will likely be somewhere in the range of R-33 (my low-end estimate) and R-36.5 (your high-end estimate). That's OK for a retrofit job. Just so your eyes are open -- it's not R-49, but it doesn't have to be.

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |