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
Q&A Spotlight

Pondering an Attic Conversion in New York

A homeowner wonders whether the advice he’s getting on insulating his attic will lead to trouble

Image 1 of 2
The attic could use some updating, but the homeowner is unsure about accepting his contractor's recommendation that he turn it into a conditioned space by dense packing the rafter bays with cellulose.
Image Credit: Images #1 and #2: BuildingNewb
The attic could use some updating, but the homeowner is unsure about accepting his contractor's recommendation that he turn it into a conditioned space by dense packing the rafter bays with cellulose.
Image Credit: Images #1 and #2: BuildingNewb
The furnace in BuildingNewb's attic would be replaced with a more efficient model.

An energy audit on BuildingNewb’s upstate New York home has prompted a recommendation that he insulate the rafter bays with dense-packed cellulose, transforming what is now a ventilated attic into conditioned space.

The HVAC equipment that’s already in the attic would then be located in a conditioned space, Newb writes in a post at GBA’s Q&A forum. The contractor has suggested that existing vents at the ridge and soffit be maintained, but apparently doesn’t see the need for any ventilation space directly beneath the roof deck, which is completely covered on top with Grace Ice & Water Shield, a peel-and-stick waterproofing membrane.

“The contractor told me that when cellulose is dense packed to the proper density, air and water vapor shouldn’t be able to easily get through it from the interior,” Newb says. “In the event any moisture does get into the cellulose, it will disperse throughout the material and find its way to the openings at the soffit or ridge rather than trying to diffuse through the wood which obviously has a vapor impermeable membrane on the other side.”

Newb says that air-sealing the attic floor looks like it would be too complicated, in part because of the three cathedral ceilings that terminate at the attic.

His contractor’s approach should yield R-35 of insulation between the 2×10 rafters, and keep the underside of the roof deck warm enough (the contractor claims) to prevent any water vapor that reaches it from condensing. Spray foam insulation is off the table due to potential health concerns.

If Newb chose not to insulate the rafter bays and live with the leaky attic floor, he says he’d upgrade the R-4 flex ducts for his heating and cooling runs to R-8.

Is the plan offered by Newb’s contractor worth considering, or will it lead to problems? That’s the topic for this Q&A Spotlight.

The approach is risky

GBA editor Martin Holladay finds the contractor’s plan “risky.”

“Even though you have Ice & Water Shield installed on your entire roof, you can still include ventilation baffles under the roof sheathing to provide ventilation,” he says. “Of course, you can only create this type of vented roof assembly if your roof has no valleys or hips. You want a straight shot from the soffits to the ridge. If you can’t (or don’t want to) install these ventilation baffles, don’t use cellulose between your rafters.

“If you have your heart set on an unvented conditioned attic, you have to follow the rules,” Holladay continues. If Newb does not want to install ventilation baffles, he has two options that would work: install rigid foam insulation above the rafters, an approach that would require a new roof, or spray closed-cell foam on the underside of his roof deck.

A vented roof would be best, Newb says, but the problem lies in the three cathedral ceilings. The polystyrene ventilation baffles currently installed there would have to be removed and replaced with rigid baffles sturdy enough to stand up to dense packing, and that would require a considerable amount of drywall to be removed.

“The contractor (who has been dense packing roofs for many years in my area) swears that the baffles are unnecessary, since the hygroscopic properties of cellulose will bring moisture to the ridge/soffit,” Newb adds. “The contractor has been in business in my area for over 30 years and provides a lifetime warranty on the work any any potential damage. They do roofing, siding, and insulation.”

Consider a diffusion vent or spray foam

Dana Dorsett suggests that adding a diffusion vent at the ridge might help. That would require the replacement of a “significant amount” of sheathing at the ridge with a material that is more vapor-permeable than plywood or OSB.

“If going that route, replacing the sheathing at the ridge with something fairly vapor-permeable at the ridge such as MDF or asphalted fiberboard (top side only) and removing the Ice & Water Shield from that portion would mitigate some of the risk,” he writes.

Dorsett adds that Newb could insulate the bottom of the roof deck with a combination of closed-cell polyurethane foam and batts or cellulose — as long as he makes sure that a minimum amount of the total R-value of the insulation is foam. (In Climate Zone 5 where Newb is located, that percentage should be at least 40%.)

“In Zone 5 with 3 inches of foam (R-18) and 6 1/4 inches of fibrous insulation (R-23) you’d have R-41 total, with 44% of the R-value being the foam.”

“At 4 inches, closed-cell foam is pretty vapor-tight,” he adds, “but still not a true vapor barrier, and still provides a drying path to the interior for the roof deck. But it’s good to check the moisture content of the roof deck in several places before applying more than an inch of closed cell foam.”

Consider batts with rigid foam

If spray foam really can’t be considered, as Newb insists, Dorsett has another suggestion: insulating the rafter bays with batts and adding a layer of rigid foam below that.

“With a 9 1/4-inch cavity depth and standard spacing there is just enough room to accommodate R-30 high-density batts with a 1-inch code-min vent gap, or R-25 mid-density batts,” Dorsett says. “Ideally, there would be an reasonably airtight baffle on the exterior side of the batt to guarantee that the vent space remains open for the longer term.”

The next step would be to attach foil-faced polyiso foam, 1 1/2 inch or 2 inches thick, to the bottom of the rafters with cap nails. After the seams were taped, Newb could install 1/2-inch drywall with 3-inch or 3 1/2-inch screws, or nailed with 12d ring-shank nails. That assembly would yield R-34 at the center of the rafter bay, more than 9 1/4 inches of dense pack cellulose, with at least R-9 of insulation as a thermal break over the rafters.

“It’s probably cheaper than dense packing, too,” he adds, “unless the roof lines are complex or the space is too cramped to work in.”

The space is just too awkward

Indeed, Newb replies, the attic is just too awkward and difficult to work in, which is why dense packing from the attic and the soffit would be ideal — as long as that doesn’t cause any serious problems.

“Maybe it’s better to just replace the R-4 flex ducts up there with R-8 and roll out some unfaced fiberglass perpendicular to the current batts in the floor joists?” Newb adds. “Perhaps the lesser of the evils is just succumbing to the fact that the furnace is in the attic and just insulating the floor where I can.

“I’d love to make some sort of furnace room up there to accommodate a condensing furnace but the layout is just too awkward and would require substantial modification of gas line and location of furnace,” he says.

And even though Holladay’s suggestion that he consider rigid foam above the roof deck may have merit, the fact that he roof was recently done seems to put that idea out of reach as well.

“The contractor recommended Ice & Water Shield for the entire roof, which I now realize was a bad idea,” Newb adds. “Nevertheless, foam on the exterior is not an option.”

But as the discussion has unfolded, so has Newb’s thinking about the project. In an email accompanying the photos of his attic, Newb added this:

“The ductwork is an absolute mess which is why I’m looking into building an insulated mechanical room with a high efficiency furnace, installing all new ductwork with a longer main trunk and shorter runs of flex with R-8 insulation opposed to the existing R-4.2. Then the floor will be air sealed and blown in with R-60 worth of cellulose.”

Our expert’s opinion

Here’s how GBA technical director Peter Yost sees it:

There are essentially two problems to solve in this project:

  • Lousy HVAC location and ducting
  • Lousy attic building enclosure

To solve the first problem, the best practice would be to move the equipment and ducts into conditioned space, either by moving them out of the attic or by moving the conditioned space boundary up so that it includes the attic.

For the second problem, the best practice would be to ensure continuity in the following elements of the enclosure, in this order:

  • (1) The bulk water control layer (connecting your roof underlayment with all flashings at all penetrations).
  • (2) The air control layer (airtight membrane or insulation, sealed at all penetrations).
  • (3) The drying potential (designated and directional).
  • (4) The thermal control layer (insulation that eliminates or reduces thermal bridging).

For all HVAC elements to be truly inside conditioned space, you need #2 and #4 on the list above to be true.

You must assume, or know, that your new roof has continuous bulk water management as described above.

The best way to ensure drying potential of your roof sheathing — given that drying up and out can’t be done with continuous Ice & Water shield — is to vent from soffit to ridge. If you can’t configure this, then you must select components on the interior side of the roof sheathing that are vapor-permeable (with vapor permeances in the Class III or vapor-open range for optimal drying potential). And in my experience in cold climates, south-facing pitched roofs can tolerate quite a bit of moisture (because of subsequent drying driven by high energy solar gain), while north-facing pitched roofs tolerate very little wetting.

But back to #2 on the list above — the air control layer. In cold climates, roofs don’t get wet and stay wet from vapor diffusion; they get wet and stay wet from air leakage. In any successful scenario for this project, you need to configure the components for a continuous and complete attic/roof air barrier or air control layer.

Can any type of cellulose get you there? Unconditionally, no. Any type of cellulose is simply too air-permeable to be best practice. I have seen buildings get by with cellulose and no dedicated continuous air barrier, but only when occupant pressure and other significant sources of moisture are low, very low. But it’s really risky to count on occupant number and behavior to protect vulnerable assemblies.

Your contractor with 30 years of experience counting on cellulose to manage air-leakage-driven moisture should come visit some recent attics and roof assemblies I have been called on to inspect. One recent project was a dense-packed cellulose cavity installation where air leakage around a ceiling drywall penetration for a sprinkler head allowed warm, moist air to accumulate near the ridge. The moisture condensed and damaged framing and rafter hangers.

And note that while ridge diffusion venting could work, it would only work if there were a continuous roof assembly air barrier; otherwise that vent will make the air leakage and subsequent wintertime wetting worse. The vent needs to be for diffusion only, not an air vent.


  1. Jud Aley | | #1

    Un-vented Cellulose roof
    10 years ago when I knew even less than I do now an Insulation sub convinced me that dense packing the 2x12 rafter bays of an attic full of duct work and air handlers was the way to go. There were/are no ridge or soffit vents. because the house has a slate roof the insulation sub said venting was not need.
    The house is in Rye NY, about 25 miles outside of NYC
    So we went ahead with the work, almost immediately the owner said the house was more comfortable and less drafty and her utility bills dropped a noticeable amount over the following year.
    As time went on and I became aware of dew points & that vented roofs were to control condensation and prevent rotted roof sheathing I’ve made it a habit to stop by this house once or twice a year and stick my hand up thru a seam in the cellulose mesh, thru the cellulose and up to the underside of the sheathing. I’ve done this in at least a dozen places over the years, summer, winter, fall and spring and the cellulose and the underside of the sheathing are always dry as a bone.
    Now if we had sheet rocked over the cellulose this might be a different story and I might have had a moisture problem, but I figure with only the cellulose and the mesh to hold it in place it is a vapor open roof assembly that is performing very well.

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

    Response to Judson Aley
    Thanks for sharing your experience.

    GBA recommends that readers only install cellulose between rafters if either (a) there is a ventilation channel between the top of the insulation layer and the underside of the roof sheathing, or (b) there is an adequately thick layer of rigid foam above the roof sheathing.

    That doesn't mean, of course, that everyone who ignores our advice and installs dense-packed cellulose without a ventilation channel under the sheathing will end up with roof rot. It simply means that the approach is risky.

    Some roofs (like the one you describe) may do well without the ventilation channels. Others (perhaps with air barrier defects or dense-packing defects) will fail.

    GBA promotes insulation methods that are robust, not risky. That said, every homeowner has a different appetite for risk.

  3. User avater
    Dana Dorsett | | #3

    Shading factors matter. @Judson Aley
    Slate roofs are at least partially vented on the exterior side of the roof decking (unless covered by an impermeable layer rather than asphalted felt.) Even a little bit of sunshine can improve drying rates by quite a bit. North facing pitches or sections of roof shaded heavily by trees/hills/buildings will usually have measuarably higher moisture content than south facing pitches in full sun. So the orientation and shading factor of pitch with the hand-feel test matter.

    Rather than poking big hole in the mesh & insulation with a hand, use a 2-pronged moisture content meter, and measure the moisture content of the roof deck at several locations, making notes of the moisture content & date. The peak moisture content of roof decks with a class-III vapor retarder on the interior (latex paint on sheet rock) will usually be in late winter. In this case since there is no vapor retarder it may be shortly after mid-winter, but taking 2-3 measurements between Valentines Day and Tax Day would give you a better idea of the moisture content as temperatures rise to mold-growth conditions.

    To poke it through a foot of insulation you'll need to make up extenders for the probe prongs, but that's not a difficult project:

    Insulating the prong extenders with heat shrink tubing would be necessary for avoiding measurement error related to the moisture content of cellulose in direct contact with the extenders.

Log in or create an account to post a comment.


Recent Questions and Replies

  • |
  • |
  • |
  • |