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

Is This Insulation Too Good To Be True?

The owner of a house in a historic district wonders whether a certain ‘space age’ insulation is the answer to his problems

Not much room for insulation. An old house in an historic district has recently been re-roofed with cedar shingles over skip sheathing. Now the owner wants to bring insulation in the rafter bays up to code. But how?
Image Credit: Jan Verschuren

Jan Verschuren has a nicely roofed older house, and a problem to go with it. Cedar shingles have been installed over skip sheathing, making for a roof that’s not only historically correct but one that allows air to circulate freely beneath the roof deck. Verschuren’s next objective is to insulate between the 2×4 rafters, and here’s where he has run into a snag.

Building codes require at least R-38 worth of insulation in the roof. Ordinarily, that wouldn’t be a problem, but Verschuren says that he has only the 3 1/2-inch deep rafter bay to work with. He can’t afford to lose any headroom by adding insulation below the rafters, and he’d rather not tear off the roof to add extra insulation on top of the rafters.

He’s zeroed in on a type of insulation called Spaceloft with a reported insulating value of R-10 per inch. The distributor’s web site describes Spaceloft as “a flexible aerogel composite blanket.”

“So, 3 1/2 iches is all we have, but we’d like to get to an R-50 or thereabouts in our upgrade of this 1925 house built in Climate Zone 4C,” he writes in a Q&A post at GreenBuildingAdvisor. “Spaceloft seems the only option.”

Verschuren thinks he can get nine layers of this material in the rafter bays, followed by one or two layers over the bottoms of the rafters.

There are a few downsides, however: The cost is “exorbitant,” and Verschuren still has to figure out how to detail the installation so it will be vapor-open and able to dry out.

Or, does he have other alternatives? That’s the issue for this Q&A Spotlight.

Space what?

GBA senior editor Martin Holladay has never heard of Spaceloft, but has two other suggestions. “Three and a half inches of closed-cell spray foam would give you about R-22.7,” he writes. “That’s less than code minimum requirements. The solution is to thicken up your rafters or to install several layers of continuous rigid foam under your rafters.

“You’ll lose a little headroom, but at least you’ll have a roof that meets minimum code requirements.”

If Verschuren absolutely can’t afford to lose any headroom in the attic, Holladay adds, “it’s time to raise the roof. Again, we’re talking physics. Sometimes you just have to face facts.”

Nate G understands why Aerogel insulation performs as well as it does, but Spaceloft is another question.

“I am highly skeptical of the claims of this product because Aerogel is a rigid material, basically a mostly-hollow brick made out silicon,” Nate G writes. “This stuff [Spaceloft] comes in a flexible blanket. I have a hard time believing that it behaves the same as rigid Aerogel. The way Aerogel material works is by trapping a lot of air in billions of nanometer-sized pores. In a solid material, the integrity of these pores can be guaranteed. In a flexible sheet, how can it?”

And then there is the expense.

“But let’s say I’m wrong that this stuff really performs as advertised,” Nate G adds. “The best price they offer is $95/inch thick/square foot. By contrast, conventional insulation materials are literally in the ballpark range of to 100 times less expensive. At 3 1/2 inches thick, you’re paying $332 per square foot of roof. You could, like, demolish the house and build a whole new one for that price given an average-sized attic.”

To Holladay, that kind or price tag makes Spaceloft “fairly irrelevant to the everyday concerns of residential builders” even if it performs as advertised.

Do the math on how much you’d really save

The cedar roof cost $20,000 just five years ago, Verschuren replies, so removing it to put more insulation above the rafters isn’t an attractive idea — and it makes the Spaceloft option look a little less harebrained despite the expense. In addition, the engineer who’s designing a radiant-floor heating system for the house wants a healthy dose of insulation.

Charlie Sullivan has sharpened his pencil and done some math. After using new cost estimates provided by Verschuren, he calculates that Spaceloft would cost about $15 per board foot (1 foot square by 1 inch thick).

“That brings is back into the realm of a little eccentric, as opposed to completely wacko,” Sullivan says. “I’m comfortable with a little eccentric. My previous cost-effective calculation changes to 12 times as expensive as it should be, rather than 75 times.

“The extra heat loss from that, at a 60°F temperature difference (we still don’t know your climate, so that’s a guess) is only 265 Btu/h less than you’d get with an R-5/inch solution. The radiant engineer shouldn’t call himself an engineer (or radiant for that matter) if he thinks that an extra 265 Btu/h is going to undermine his design. That’s less heat loss than a 25 square foot high-quality U-0.2 triple-pane window.”

Nate G believes Verschuren would do himself a favor by being more flexible. “You seem to be getting yourself tied in knots because of the limitations you’ve set for yourself,” he adds. “Sometimes it’s time to admit that those limitations either preclude any good options or need to be removed.”

If Verschuren could lose just a few inches of ceiling height, he adds, he’s have many more options. Or, if he needs more space, he could always build a ground-floor addition that preserves the architectural character of the house.

Go with the spray foam option

One of Verschuren’s objections to spray foam is that it would defeat the purpose of installing cedar shingles over skip sheathing by limiting their drying potential.

But Holladay has a suggestion: “If you are committed to your plan,” he writes, “install some cardboard or housewrap between your rafters followed by R-20 of spray foam, and live with the imperfect thermal performance of that type of assembly (assuming, of course, that your local building department doesn’t require a higher R-value),” he writes.

Once Verschuren has cut up all that expensive insulation and placed it between the rafters, Holladay says, there is still the problem of thermal bridging through the rafters. “Once thermal bridging is accounted for, the improved performance compared to spray foam is minor, and certainly not worth the investment.”

Nate G looks at it this way: Verschuren’s planned installation of Spaceloft, with a total price tag of $10,000 for materials alone, yields a total R-value of 38. Holladay’s alternative of closed-cell foam over cardboard plus a single layer of Spaceloft gets the roof to R-29 at a cost of $2,480.

“So for less than 25 percent of the price, taking Martin’s advice and adding only a little bit of Spaceloft would get you 77 percent of the R-value. That seems like a no-brainer to me. The differences between R-30 and R-38 are not worth paying another $7,500, in my humble opinon,” he writes. “Comfort-wise, if this is conditioned space, you won’t feel the difference. But it’s your money!”

The coup de grace for this idea?

The discussion so far has assumed Spaceloft has an insulating value of R-10 per inch, and that apparently does apply for certain versions of this product. But Bill Dietze takes a close look at the fine print and throws this curve: “Jan, the link you provide mentions the uses of Spaceloft as ‘Great for home insulation, winter clothing, science projects,’ but the link to datasheets for material with a conductivity of 0.14 W/mK (R-10 per inch) is an undersea pipe product,” he writes. “Hardly residential. If you proceed, be aware that you are probably in the ‘science project’ category.

“The Spaceloft product rated for ‘Ambient temperature walls, floors and roofs in commercial, residential and institutional building’ has a lower thermal resistance (R-8.3 per inch) and no datasheet describing the application.”

If the lower R-value is correct, it makes the case for Spaceloft even less compelling, Holladay points out.

“But it is still ahead in many ways over ‘the rest’ to use,” Verschuren says. We have the money to do it but do not want to waste it, either!”

Our expert’s opinion

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

I decided I really needed to check in with Jan Verschuren on this one; just too many balls in the air to offer any helpful perspective. Having done that, I think the list of constraints (below) Jan poses or faces proves that if you narrow the boundary conditions on a problem enough, you can arrive at a singular solution.

Here are the factors to consider:

  • The newness of the current roof cladding means you can’t build up insulation topside.
  • Very low ceiling heights mean that there is little room to build up insulation on the bottom.
  • The ceiling is already being built down 13/16 inches for Warmboard-R panels, primarily for radiant cooling.
  • Radiant cooling for the attic space is being driven by lack of room for ducts and an already-in-place ground-source heat pump system, making for a fairly easily retrofit for radiant cooling.
  • The homeowners can’t really rely upon natural ventilation (open windows) for low-grade space cooling because of urban noise issues.
  • Spray foam insulation is not an option.

As Jan says, “…my hands are tied!”

So, if we can agree to accept these constraints, Jan is ending up with layers of Spaceloft in between the rafters and two layers (2 centimeters) of continuous Spaceloft to the underside of the rafters.

We confirmed that standardized independent R-value testing of Spaceloft yields about R-8.3 per inch. That makes Spaceloft significantly better per unit of thickness than any other readily available building insulation but also means the entire assembly is still not code-compliant and far from high performance.

At the end of the day, that phrase “readily available” may be the key to whether this solution will work for Jan and others considering this pricey approach to insulation: neither I nor Jan could get Aerogel to respond at all to our requests for technical information and information on the availability of Spaceloft for Jan’s project.

No magic bullet for this one, with neither the bullet or its delivery magical at all.

Editor’s note: After this article was published, GBA learned that the product formerly known as Spaceloft has be re-branded as Proloft. For more information on Proloft, visit Advanced Insolutions Inc.. Here is a link to a report on R-value testing of Proloft Aerogel Blanket.

58 Comments

  1. Meyer Brendan | | #1

    Our jurisdiction exempts
    Our jurisdiction exempts historic buildings from the Energy Code. Moisture of any kind under cedar shingels/sharkes or terra cotta roof tiles kills those materials. Ventilation is extremely important to their long term value. Turning the attic into a living space is the limiting variable in this equation.

  2. Bob Coleman | | #2

    this question is so formulaic, but still fun
    Questioner has some housing situation in which they absolutely can not change anything, even if they have woefully inadequate structure or need to spend a billion dollars on insulation.

    For the installed cost of this aerogel blanket, the questioner could likely tear the roof off saving the shingles, install a new structurally adequate roof with beams larger than 2x4s and cheaply blow in 10inches of cellulose insulation.

  3. D Dorsett | | #3

    Even if...
    Even if the R10/inch aerogel were available at an affordable price to hit R38 center cavity, the performance of the assembly thermally bridged by R4-ish 2x4 15-20% framing fraction is woefully short of the performance of R38 in an R12-ish 2x12 rafters. Even at a 12% framing fraction the 2x4s would be conducting as much heat as the 88% of the area that is aerogel.

    It would fail code max on a U-factor of U0.30 basis alternate to R38 code minimums, per TABLE N1102.1.3: http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_11_sec002.htm

  4. Caleb Crawford | | #4

    painted himself into a corner
    Another argument for bringing in a professional to establish a strategic plan before you start.
    Radiant heating? With a robust insulation and air sealing package it wouldn't be needed, and it would save a significant cost.
    Radiant cooling? I hope you know what you are doing. If you have humid days with a skimpy insulation package you'll have condensation on the floors. I hope this is in New Mexico.
    This aerogel insulation? I agree with the Dorsett that the thermal bridging will degrade the overall thermal performance.
    Definitely no spray foam; you'd have mold in no time.
    Any foam in between the rafters with a vented roof is a huge potential fire hazard. The retardants in foam don't really do anything.
    2" mineral wool between the rafters, and an inch of polyiso below the rafters, that will give R 15 and a meaningful thermal break.
    Add a robust air barrier like Intello plus, all joints taped and taped to window frames and subfloor. Yeah you lose an inch, but seriously, one inch? get rid of the radiant and you've got it.
    With the air sealing you probably need an ERV - one of those little Lunos through the wall numbers.
    Talk with an architect before you start next time.

  5. Alan B | | #5

    $10,000 buys a lot of heat,
    $10,000 buys a lot of heat, or solar panels...
    The OP needs to give up on doing the impossible and either live with his choice or use realistic options. The 8+R value does not sounds realistic, and as they say if something sounds too good to be true...
    Desperately hoping its not snake oil will not make it legitimate, no matter how much he can convince himself it does, but if he has 10k to blow i'd be happy to sell him snake oil plus and at least his money will have gone to a worthy cause.

  6. Larry Maloney | | #6

    Simple solution
    By law R values are tested by reputable third party laboratories. If this manufacturer could document such a high R-value they would have done it and readers and writers and consumers and builders would not have to speculate. The writer of this article should not have printed his story without getting the documentation first. He's promoting the product to the gullible.

    As a practical matter it sounds convoluted that building codes would require a "historical" home have the roof raised to accommodate added insulation. Since the depth only drops near the very edge, as is common in most all homes, why be concerned?. In forty years as an insulation contractor I've never heard such a silly concern about the edge of the wall needing to be a full R-38. Enforce that and all homes will need extensive and unaffordable ceiling modifications. What a silly article ...just to promote a useless product. Blow cellulose in and you won't have problems with icicles forming and the heat loss will be minimal. Spraying foam is not cost effective.

  7. User avater GBA Editor
    Martin Holladay | | #7

    Response to Larry Maloney
    Larry,
    You wrote, "The writer of this article should not have printed his story without getting the documentation first."

    However, if you had followed the link to the Q&A thread under discussion, you would have discovered links to technical specifications for Spaceloft and a third-party report on the product's R-value. Here are the links:

    Spaceloft specifications

    Report from a Slovenian laboratory

    Proloft evaluation

    The GBA article quotes Peter Yost, who correctly noted that "We confirmed that standardized independent R-value testing of Spaceloft yields about R-8.3 per inch."

    You wrote, "In forty years as an insulation contractor I've never heard such a silly concern about the edge of the wall needing to be a full R-38. ...Blow cellulose in and you won't have problems with icicles forming and the heat loss will be minimal."

    You seem to be assuming that this article is discussing a job where insulation will be installed on the attic floor. In fact, the article is discussing a job where the insulation will follow the sloping roofline.

    Finally, even if the job in question concerned the installation of insulation on an attic floor -- which it doesn't -- it's rarely a good idea to skimp on insulation thickness at the perimeter of the attic. The area above the top plates of the home's exterior wall are precisely where a thick layer of insulation is most needed, especially if a homeowner is concerned about ice dams.

  8. John Miranda | | #8

    This insulation kept the shuttle from burning up
    You might try contacting the folks @ http://www.Hy-Tech.com They have paint which can be sprayed. Of course, the paint has no insulation value. It's what they put in the paint, tiny ceramic spheres with a void (empty space, like outer space) that does not conduct heat or sound. However, I don't know if there's an R-value associated with it to meet the inspector's code. I know I've used it both on the roof of my house and in the basement. From my experience it works. It's also been my experience that you probably won't find an Architect or Engineer who will even look at the website, since they will all give you the knee jerk reaction that paint has no R-value. But, it was good enough for NASA to use on the heat shield to keep the shuttle from burning up upon re-entry. Of course, that's different. This is a house. How could you apply that technology to insulating a house? I believe the folks at Hy-Tech will tell you it can be done and has been done - very successfully, I might add, for several years, in spite of the ignore-ance of mainstream professionals, who won't even listen.

  9. Nathan Spriegel | | #9

    John engineers HAVE looked at such "paint"
    And they found it to be of little use.
    “The Florida Solar Energy Center has tested ceramic paints and found them to have no significant advantage over ordinary paint in terms of their ability to retard heat gains through exterior building surfaces."
    https://www.greenbuildingadvisor.com/blogs/dept/musings/insulating-paint-merchants-dupe-gullible-homeowners

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

    Insulating paint
    Before any GBA readers rush out to buy "insulating" paint, in response to John Miranda's comments, I strongly urge them to read this article: ‘Insulating’ Paint Merchants Dupe Gullible Homeowners.

    [P.S. It looks like Nathan Spriegel and I were typing our responses to John Miranda's comments at about the same time, but Nathan beat me to it.]

  11. Malcolm Taylor | | #11

    Keep an open mind
    Maybe ceramic paint makes you feel all warm inside?

  12. Eric Habegger | | #12

    radiant barriers
    I think people can get confused by applications for radiant barriers and those for actual insulation. Radiant barriers work terrific in space because there is no air, hence no thermal conductivity or convection. That's why vacuum panels work so well even though they are horribly expensive. On Earth there is no substitute for thick regular insulation unless one has more money than sense.

    I have radiant barriers attached to my roof rafters and I use it to keep the attic cool, not to keep the house cool. It works well in that application but I use straight attic floor insulation to keep the house cool. I can put my hand right next to the radiant barrier at mid afternoon and will feel no heat radiating from it. But when I touch it it's hot. The reason any thermal barrier that works on the radiant barrier principle is not very effective on Earth is because the air is always touching that hot radiant barrier. Those air molecules conduct that heat to other air molecules and will eventually heat the whole attic. It's effective to keep the attic from getting super hot because you don't live in the attic and it's cheap.

  13. D Dorsett | | #13

    With enough low-E layers and enough air-gaps...
    With enough low-E layers and enough air-gaps it's possible to meet almost any thermal performance spec, but like any approach there are diminishing returns. The cost & complexity of making high-performance assemblies with radiant barriers makes them of pretty limited use.

    The ceramic paint nanospheres are a triumph of marketing optimism over real-world optimism. The fact that a material can withstand the extreme temperature differences encountered during an earth orbiter re-entering the atmosphere is great if your house is in orbit, and re-entering the atmosphere. But if the goal is providing home comfort & energy savings for a stationary earthbound home the stuff has no practical application, marketing claims notwithstanding.

  14. Brian Hierlihy | | #14

    Proloft Aerogel Insulation
    "100 times less expensive". How exactly is that calculated?

    Assume $1 cost.
    $1 x 100 = $100.
    $1 - $100 = -$99.
    That is, the buyer actually 'earns' $99 for every dollar spent of the product?

  15. Jesse Ewing | | #15

    Proloft Aerogel Insulation
    Hello Everyone,
    I'm the Vice President of Advanced Insolutions Inc. and we're the owner/distributor of Proloft. We have re-branded Spaceloft as Proloft and Intertek has performed our third party testing under the name Proloft (see test results below) Please feel free to contact me with any questions and you'll find more info at http://www.advancedinsolutions.com. My cell is 250-551-3200 and/or email [email protected]

    Intertek has conducted testing for Advanced Insolutions Inc. on Proloft™ Aerogel Blanket to evaluate the thermal transmission properties. Testing was conducted in accordance with ASTM, following the standard methods of C518 (2013) Steady-State Thermal Transmission Properties by Means of the Heat Flow Meter Apparatus.

    R/inch using instrument measure thickness is 9.1 R/inch (Hr-ft2-ºF/Btu/in).
    R/inch using ASTM D1622 measured thickness is 10.2 R/inch (Hr-ft2-ºF/Btu/in).

  16. Jan Verschuren | | #16

    Martin, and others.
    Any

    Martin, and others.
    Any verdict on the R rating of this Proloft, Spaceloft?
    Where is the truth on this. Is it 8.3, 9 or 10.2. OR are all numbers off?
    R value being relative as it is, it is still good to have this info.
    Thanks
    Jan

  17. User avater GBA Editor
    Martin Holladay | | #17

    Response to Jan Verschuren
    Jan,
    Here is the link to the evaluation of Proloft Aerogel Blanket.

    The document notes, "R/inch using instrument measure thickness is 9.1 R/inch (Hr-ft2-ºF/Btu/in).
    R/inch using ASTM D1622 measured thickness is 10.2 R/inch (Hr-ft2-ºF/Btu/in)."

    I'm not sure how important the discrepancy between these two methods of measurement really is. I can't really imagine any application where the difference between R-9 per inch and R-10 per inch would be significant enough to matter.

    Remember, when you install any type of insulation -- fiberglass batts, cellulose, or spray foam -- installed R-value per inch can easily vary +/- 10% from published values.

  18. User avater
    Dana Dorsett | | #18

    The numbers are real enough...
    ... including the price tag.

    But no matter what you put between the framing it doesn't change the R-value of the framing fraction. If using expensive high-R/inch products in a building situation it is far better to use it in thermal breaks over the low-R/inch framing than as cavity-fill, but even there the price/performance often can't be rationalized, especially when priced against other solutions of equivalent thermal performance.

    If you're trying to meet code without sufficient rafter depth to meet it on a cavity-R prescriptive basis you can sometimes get there on a U-factor basis per Table N1102.1.3 (see: http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_11_sec002.htm ) using strips of higher R/inch goods on the rafter edges.

    R49 is code-min for zone 4C, which is not achievable with any existing cavity fill in a 3.5" deep cavity. But U-0.026 (R38 whole-assembly) would also meet code. Hitting U0.026 is not possible with the 3.5" deep 2x4 framing, even with R-infinity cavity fill due to the thermal bridging of the rafters. But with 2" deep aerogel thermally breaking the rafter edges and 5.5" of R6.5-R7/inch closed cell foam cavity fill, plus an inch of rigid EPS (or rigid rock wool) on the interior you'd be there. That adds 3.5" to the depth of the assembly.

    But is it worth it?

    There's nothing magical about R50 in the roof- the codes specify that since it's cheap additional performance when done in a more conventional matter. But with the added expense of the the more complex assembly using expensive materials it's literally a "pays never" situation, on either greenhouse gases (due to the high global warming potential of the blowing agents for the closed cell foam) future energy use savings in a 4C climate. A lot more performance can usually bought elsewhere in the house with that money (at much lower environmental cost to boot.)

  19. Jan Verschuren | | #19

    Dana,
    The problem we are

    Dana,
    The problem we are having is depth. We are going over top of the rafters with the Proloft to get the benefit of thermal breaks. That way we are now leaning to get to R48. (Contractor here mentioned 48 being the new code here) That means 4 layers now instead of two overtop of the rafters. The price is never ever going to be justified as in: return on investment. Our goal is to get the total chain of items to work as best in unison as possible. Radiant heating works most efficiently the more you can drop your water temp. In this old house that means meeting code is nothing wild. So that is what we want to do and that we spend that money is our decision. I am not saying everybody should do this. But when we now say: you are allowed to build below code as it is an historic, grand fathered in home, that is only that. We do not want to do that. It guaranteed will never pay itself back. THAT is not the question.
    The goal is to create an as comfortable home as possible, and with radiant you can do that ONLY when your insulation is in place.
    Let's say some people buy a Rolls Royce. WHY? Pointless in my opinion. But not to that person.

    You mentioned cooling being ridiculous for the NorthWest in another posting. With this heat wave we are having now we definitely disagree. Good sleep is crucial for healthy living. And summer has not even begun...

    Now, I copy my installation question from the other posting here. See if you are worth sending you that P85D one day.... :-)!

    Installation advice.
    Full WarmBoardR on the inside, so no venting to be had there. Our cedar roof installed as skip sheating is TOO vented. So best way to (no foam) seal this as airtight as can be: Tyvek for protectection around the rafters followed by MemBrain. Tyvek to protect the MemBrain.
    OUTward venting in climate 4C. Replacement of cedar roof (in 30 yrs we hope) will probably get plywood sheathing overtop of the skip sheathing, and will then hamper the ventilation to a degree. With 0.5" plywood and limited length staples it makes installation safer: to not penetrate the Tyvek and Membrain down the road.

    Jesse was helpful to advice the use of PL200 or PL400 construction adhesive to hold the Proloft in place during installation. I would need some help probably to temporary hold things in place, like a construction glue. Same for the Tyvek and Membrain. Stapling the Tyvek(?) but spot glueing the MemBrain?
    Thanks
    Jan

  20. User avater
    Dana Dorsett | | #20

    No comfort difference...
    "The goal is to create an as comfortable home as possible, and with radiant you can do that ONLY when your insulation is in place."

    The comfort difference between R10 roof vs R50 roof with a WarmBoard radiant ceiling below it is zero, even though the lower R assembly raises the heat load a bit. The roof is still only a fraction of the heat load, and at at R20 whole-assembly it's probably not even a very large fraction, despite the higher delta-T when using the ceiling as your radiating surface.

    The difference in water temp requirements at peak heating load is quite modest. With R15 rock wool between the rafters and 2" of rigid rock wool thermally breaking the rafters over the whole interior you'd have a whole-assembly R of about R20, or a U-factor of about 0.05. That is about half the heat loss of 2x4 fiber-insulated wall, and a quarter the heat loss of U0.20 triple-pane windows. Pouring the Rolls Royce luxury-cash into the ceiling with aerogel cavity-fill provides less energy savings (and certainly less comfort) in your case than pouring it into window upgrades.

    I don't recall ever calling the cooling loads in the PNW "ridiculous", but they are quite modest compared to other US climates, particularly the latent loads. In a tight higher-R house night time ventilation strategies usually work pretty well there. Even though a few of the recent nights only dropped to the high 60s overnight, the dewpoints barely hit 60F even during the most humid part of the day. (Pull up a dew point graph as well as temp: https://weatherspark.com/#!dashboard;q=everett%20wa%20USA ) With dew points in the 60s you can run into significant condensation in radiant-cooling, so while do-able, it's still a bit tricky some weeks. Most of the summer PNW dew points are in the mid-50s or lower, even during hot weather. Midsummer dew points in my part of New England average in the mid 60s, sometimes breaking into the mid-70s, but thankfully not so far this season, which makes radiant cooling even harder to pull off. (And this area's cooling loads are still pretty tame compared to the gulf coast from TX to FL.) Most of my relatives in the PNW who air-condition do it with window-shakers or mini-splits. But most of them don't bother. (I can't think of any who have central air conditioning.) Even during the warmer weather the peak & cooling loads are still pretty small. I've personally experienced a few 100F days on Puget Sound- it's more like 100F days in Phoenix than 100F days in Tampa, and more comfortable than 90F in NYC, which as a rule are accompanied by dew points well into the 70s.

    Skip sheathing under #30 felt & cedar shingles though air leaky, is not anywhere near "vented" from a code definition point of view, and can't be treated as such. You would still need 1" of air between the skip sheathing and the cavity insulation to meet code, which means you're down to about 2.5" of space for insulation between the existing rafters. Given that it's skip sheathing you can get away with 3" rock wool (sound deadening rock wool batts like 3" Roxul Safe'n'Sound), which has an R-value of about R12-R13). With a 15% framing fraction that comes out to about R10 whole-assembly, including the R value of the WarmBoard, cedar shingles, and gypsum. If instead you installed 3" of R10/inch aerogel that would only come in at about R15, despite the R31-R32 center-cavity R. That's a VERY expensive R5.

    If you install 1" of aerogel over the entire interior between the rafters and the WarmBoard with 3" rock wool in the cavities you'd be at about R20 whole-assembly, using 60% less aerogel.

    To fully hit the U0.026 code max would take 3" of continuous aerogel between the WarmBoard and the rafters as well as the 3" of rock wool in the cavities between the rafters. Center-cavity R would still be less than the R49 code-min, but thermally breaking rafters with R30-ish aerogel gets you there by lowering the U-factor of the assembly to under the U0.026 code-max.

    With a properly (or almost-properly) vented roof assembly you don't need the MemBrain, just make the ceiling gypsum air-tight. With the high density of rock wool batts wind-washing on the vented side is very minimal- even without exterior side air barriers it still performs. But it doesn't hurt to have Tyvek as an exterior air barrier, as long as you have at least 1/2" to the skip-sheathing (and more than 1" to the #30 felt between the skip sheathing planks.)

    If you run the heat load numbers of the 3" aerogel against other types of thermal improvements elsewhere in the house (using BeOpt or even a crummier heat load tool) you can make rational choices on where to shed load. Heat load is heat load- it doesn't matter if the losses are at the walls, windows or ceiling. The lower the heat load, the lower the hydronic water temp requirements- the heating system doesn't really care in what part of the room that heat loss is occurring, it's the total heat load per square foot of Warm Board that determines the water temp.

    Regarding the P85D, I understand there's now a heavily discounted smaller version:

    http://www.bloomberg.com/news/articles/2015-06-18/mattel-rolls-out-hot-wheels-tesla-model-s-for-fans-big-and-small

    (My kid informs me they were sold out on Father's Day, even after quoting him a $30-40 spot-market price on the phone.)

  21. Jan Verschuren | | #21

    Dana, thanks again for you in
    Dana, thanks again for you in depth help.

    The ceiling radiant is used for cooling. And as we have it up there anyway we can use it for supplementally heat as well. Central whole house dehumidification, top central on top floor. The heating upstairs will be staple up floor radiant. Not WarmBoard. So, in all: when it is 67 degrees outside, we'll need no heating system at all in the house. When we on the other hand do not want to run out of heat on the coldest and windiest days of the year, which are not frequent to be had here in the NorthWest though, but with radiant one has to measure the system properly. Hence the desire to have the roof, bedroom ceiling, not be at R20. It will only show it's money on the coldest and the warmest days. We are upgrading the windows, we are double studding the outside walls to true 8". We are minimizing the heatpump. So out of that background: R value is king. Plus airsealing. But must read all your thoughts in more detail. Very good food for thought to use the Proloft ONLY under the rafters, not in the cavity. But when I have no room to spare. Venting still required. YOU have not deserved even a Mattel version now... But I'll give you a free charge...

    Thanks again.
    Jan

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

    Outdoor temperature
    Jan,
    You wrote, "When it is 67 degrees outside, we'll need no heating system at all in the house."

    I sure hope not! The temperature should drop to the 50s before you need to turn on the heat -- assuming your house has a refrigerator and a person or two inside...

  23. Jan Verschuren | | #23

    Martin,
    Just illustrating why

    Martin,
    Just illustrating why we need good insulation: ONLY when the going gets tough, on the coldest and warmest days. And those days are less here over Missouri... We can live outside when it would always be 67 degrees. Well, 70-ish is nicer. With no wind. And no rain...
    But the illustration let to the radiant heating and cooling we do. THAT is where insulation is more critical over other types of heat delivery.

    Jan

  24. Jan Verschuren | | #24

    Dana,
    Better reading your

    Dana,
    Better reading your thoughts...
    You mentioned at the beginning:

    "The comfort difference between R10 roof vs R50 roof with a WarmBoard radiant ceiling below it is zero, even though the lower R assembly raises the heat load a bit. The roof is still only a fraction of the heat load, and at at R20 whole-assembly it's probably not even a very large fraction, despite the higher delta-T when using the ceiling as your radiating surface."

    I fear that your advise and thoughts got thrown off into a wrong direction.We do not heat through the ceiling, as on a slanted ceiling, radiant panels do not work well. According to our radiant engineer, Rob Brown from Rockport Mechanical in Maine, it was not recommended to heat through a slanted ceiling. Cooling is fine.

    "To fully hit the U0.026 code max would take 3" of continuous aerogel between the WarmBoard and the rafters as well as the 3" of rock wool in the cavities between the rafters. Center-cavity R would still be less than the R49 code-min, but thermally breaking rafters with R30-ish aerogel gets you there by lowering the U-factor of the assembly to under the U0.026 code-max."

    3" below the rafters is just not there to be had. 4 layers of Proloft 4 cm, 1.5" would give us R16 below the rafters. With 9 layers between the rafters, R36 would get to a layman's R52. You tell me what I would loose and end up with realistically it being between the rafters mostly.
    The Proloft would have similar characteristics like Rockwool I would think, being very permeable.

    I am hoping to retrofit these metal plates(pic below) in between the cedar shakes making the cedar roof absolutely open to air movement. Code or no code: with these plates (a canadian invention) my roof assembly will be vented. HENCE my worries that Tyvek alone will not give me enough air tightness to make my Proloft perform as advertised. So that is where I was thinking of adding MemBrain, but to the outside. To the Inside will be air and vapor tight with the aluminum WarmBoard R.

    And regarding Rockwool. I am a bit surprised you recommend to go with the Safe'N Sound batts. They are unrated for a reason I would think. They will definitely carry an R value, but what we do not know...
    But will use them between floors and interior walls where R value is needed but not critical.
    And sorry for mentioning you saying that cooling in the PNW would be "rediculous". When you did not maybe. It just was something ringing in my ear left over from another thread. Telling a texan we are cooling they may laugh. But the last few weeks now we definitely would love to have cooling here. So yes, loads are low, etc. BUT it is a very nice thing to have when weather hits like this once in a while and maybe more frequent in the future.
    A good night's rest is hard to get for a northwesterner, conditioned to rain, this time of year...
    Jan

  25. User avater
    Dana Dorsett | | #25

    Aha- I was assuming it was for both...
    If the ceiling is not being heated except when the radiant floor won't keep up, yes, the ceiling R makes a difference if you're talking about R13 between the rafters (~R10 whole-assembly) and something with thermal breaks that comes in at R20+ whole-assembly.

    But there's still scant comfort difference between the R20 whole-assembly (U0.050) and a fully code-compliant U0.026 assembly. We're talking less than a 1F difference in ceiling temp at your 25F-ish 99% outside design temp. The walls and particularly the windows will be much colder than the ceiling, and the much larger comfort concern.

    The density of sound abatement batts is a bit less than insulating batts, but not dramatically less. The somewhat higher density R15s are good for about R4.3 per inch, the lower density batts are good for at least R4/inch, even if they are not tested / labeled / marketed as such. Since it's thermally bridged by rafters even large differences in R/inch have only modest effects on the whole-assembly R.

    You can't put 9cm of insulation underneath the skip sheathing and still call it a vented roof, no matter WHAT you do with the shingles above. The shingle spacers vent the shingles to keep them from rotting, but the #30 felt layer above the skip sheathing has too low a vapor permeance to count on the skip sheathing to be able to dry to the exterior. You can probably get away with 8cm, but to meet the letter of code on vented roofs you'd need 2.5cm between the skip sheathing & insulation.

    Tyvek that is tight to the Proloft (or rock wool) is a sufficient air barrier for thermal performance purposes- don't sweat that one.

    Assuming you could get away with R36 in the cavities and R16 under the rafters behind the WarmBoard, if we assume the WarmBoard + gypsum adds up to R0.8 and the skip sheathing + shingles, with a 15% framing fraction you end up at about R34.5 whole-assembly, or U0.029, with a 12% framing fraction (not likely, but maybe) it's about R36.5 whole assembly or U0.27, still not quite the U0.026 code-max.

    If you back off to only 8cm / R32 to provide better roof venting that becomes about R32 whole-assembly /U0.031 or R34 whole-assembly/ U0.029 for 15% & 12% framing fractions, respectively.

    If you did the cavity fill with R4/inch rock wool (R12 @ 3") and added R16 aerogel on the interior that comes in at about R27 / U0.037 or R28 / U0.036 respectively.

    Even though U0.037 is below current code min, that is NOT a comfort problem in a Puget Sound winter, and will barely affect the cooling water temp requirements in summer, OR the heating water temp requirements in winter. The major heat loss is still elsewhere.

  26. User avater
    Dana Dorsett | | #26

    Other notes
    Read the ventilation free vent space & clearance requirements in chapter 8:

    http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_8_sec006.htm

    Note that a 2x4 framed wall with R15 rock wool, wood siding & sheathing (no exterior continuous insulation) and plaster or wallboard interior has a U-factor of between 0.08 and 0.10 depending on the actual framing fraction and other particulars.

    An insulated door has a U-factor of about U0.20-0.25. A pretty-good triple pane window is about U0.20.

    If you used 1.5" rigid rock wool for the continuous insulation (R6-ish) and 3" /R12-ish batts instead of aerogel in the ceiling, with the rest of your ceiling stackup you'd be at about U0.05, which is half that of your antique insulated wall, and 1/5 that of even a pretty good (way better than code) window. Any comfort issues here would still be massively the walls & windows.

    At the comparative prices of aerogel compared to rigid rock wool it might be cheaper to pull (and re-use) the cedar shingles, put R15 rock wool in the cavities, and add rigid rock wool both above roof deck and below the rafters sufficient to actually meet U0.026 code max. That would take about 6-6.25" of rigid rock wool (total), so with 1.5" or rigid rock wool on the interior you'd get there with 4.5" above the skip sheathing. Add a half-inch plywood for a nailer deck and it raises the shingles 5", which is probably do-able without creating other issues where it meets dormers, vents & other roof planes(?).

  27. Jan Verschuren | | #27

    Dana,
    I contacted the radiant

    Dana,

    I contacted the radiant engineer to get his thoughts on all this.
    Removing the cedar and raising the roof is very problematic let's say impossible.
    Other option is to do what we did with the steel roof. Loose 25% of the WarmBoard and use Membrain over the rafters, between the Proloft in between the rafters and the Proloft continuous underneath the rafters. Venting to the inside.
    Did not think venting would be a problem, the current #30 felt and cedar would be "like" our steel roof? It is vapor tight??
    Thanks
    Jan

  28. User avater
    Dana Dorsett | | #28

    That works as well as anything.
    If you used 1.8lb blown fiberglass instead of Proloft the difference in U-factor wouldn't be huge, and you'd get the thermal benefit and air-retardency of dense fiberglass between the skip sheathing, or you can just stack aerogel in there with a Tyvek top side air barrier on the underside of the skip sheathing. The #30 felt isn't completely impermeable (it's almost a class-I vapor retarder when dry), but in foggy-dew Puget Sound zone you won't get significant drying to the exterior. So yes, it's sort of like your steel roof section.

    It'll never be "...venting to the inside..." (venting is air movement), but with the MemBrain it can dry to the inside, whereas with a 100% coverage of impermeable WarmBoard it would not, and would absolutely need actual venting above the insulation for a drying path.

    (Tell Rob I said "Hi!" next time you talk to him! )

  29. Jan Verschuren | | #29

    R 30 minimum BUT air tight!
    Dana:

    Rob mentioned that anything above R30 becomes moot as far as insulation gains go. Smaller and smaller gains as the R value goes up beyond R30. HOWEVER, when staying closer to R30 the necessity of air sealing the cavity goes up.
    I fear that I have to stay with Proloft in the cavity as with anything else, I still have to go inches below the rafters to even get to R30.
    Rough calculating: Proloft in the cavity gives me R35. R4 below the rafters with a single layer of Proloft would bring it all to R39.
    Bring in the rafter losses and airseal imperfections I have the final things to resolve.

    Use still two layers of Proloft for an R8 thermal break to get to a R43, minus my losses.
    (OR Comfort BattIS R6 but that would be 1.5" over .4" or .8" with Proloft)

    How best to airseal the entire cavity. Tyvek only around the rafters? Spot glued in to hold it in, sealing the seams with tape or glue also? I have a pneumatic caulk gun for the Tremco Acoustic Sealant as by hand it is UNworkable to seal an entire house. Can spot glue or caulk seal with that in a heartbeat as well....

    Tyvek is vapor open. But isn't my Felt paper AS WELL? You mention having to see it as vapor impermeable as my steel roof. How can Felt paper than being rated as better than Tyvek on walls. Still I understand that I have no venting on my roof. But the laths would be between the Tyvek and the Felt would have ample vapor openness underneath the air open shakes, wouldn't it?

    Soggy Seattle (4C) IS rated as a dry climate. Rob even had to wrap his head around this:

    "You'll be interested to know that another client of mine doing radiant cooling in seattle has been functioning with a non-functional dehumidification system and maintaining 74 degree indoor without condensation through the latest heat wave. I still struggle with accepting the data on dryness in supposedly soggy seattle, but I'm getting there."

    I think it is because of the lack of humid air over prolonged warm periods. The rain dries quick after it's done...

    So do I create an airtight to the outside and vapor open, with the full WarmBoardR panels, seam caulked, to the inside a roof assembly that is good to go.

    Maybe keep the R value over the R 30 number to be safe with condensation risks regardless: two layers or Proloft for thermal breaks, because we'll be cooling in summer.

    Alternative would be to loose the WarmBoard by 25% and install MemBrain over the rafters, Proloft after that, WarmBoardR and vent to the inside as we did with the steel roof.
    But a cedar shake on skip sheathing with felt, I doubt can be considered as vapor impermeable...

    Thanks Dana!
    Jan

  30. User avater
    Dana Dorsett | | #30

    Tyvek & felt are radically different
    #15 felt runs about 0.5 perms when arctic-bone dry, ~5-6 perms when it's moisture content rises- it's a "smart" vapor retarder.

    #30 felt runs a bit lower permeance, but has similar characteristics.

    In combination with asphalt shingles it's about 0.1 perms, wet or dry, but under cedar there is at least a miniscule amount of drying toward the exterior, as long as the shingles aren't wet.

    Tyvek runs 30-50 perms, wet or dry.

    Seattle IS a dry climate, from an average summertime outdoor dew point perspective, with an average mid-summer dew point of about 55F, which is what makes radiant cooling possible. Summertime outdoor dew points rarely break 60F, and it makes front page news when it breaks 65F, which is roughly the AVERAGE outdoor dew point in Boston or NY, with many excursions in to the 70sF. So in Seattle the radiant panels can be set up to be 60F or even cooler without a condensation problem, at least MOST of the time. But in the northeast there would be copious condensation running panels that cool.

    But from a wintertime dew point point of view it is anything BUT a dry climate. Dew points are usually above freezing (~35F average) and fairly close to the outdoor air temperature, and morning dew/fog very common. (Does it ever rain in Seattle in winter, or is it just high tide all the time? :-) ) Unlike summertime rains, wintertime rains take a long time to dry. This is because the outdoor air's dew point is so close to the outdoor temperature in winter- a handful of degrees F. In summer it's common to have a 50F outdoor dew point on a 75F afternoon after a rain, which is a rapid-drying condition.

    When there is dew/rain on the shingles or fog outdoors there is very little vapor pressure gradient to move much moisture through 0.5-5 perm felt, since the temperature at skip sheathing would already be almost at the outdoor dew point. The felt layer might as well be true vapor barrier under those conditions, and in fact the vapor pressure under wet shingles is usually from the outside toward the interior. You' get a bit more (5-10x as much) drying out of Tyvek, but that only helps if the sheathing is vented below. With the #30 felt under the cedar and no venting under the skip sheathing to the outdoors it doesn't matter. But at least Tyvek won't slow down the drying rate toward the interior the way a lower-permeance product will.

    From a moisture accumulation point of view it's the winter INDOOR dew points relative to outdoor air temperatures (and to a lesser degree, outdoor dew points) that matter. Average wintertime indoor dew points are 45-50F in Seattle, whereas the average outdoor temperature is about 40F. So ~40F skip sheathing will accumulate moisture from interior side vapor drives if there isn't a sufficient ventilation to the outdoors to relieve that load, unless there is a low vapor permeance at some interior layer. That is why you need the MemBrain in an unvented stackup, and why it's not a bad idea even with a minimally vented roof. The ONLY way to get appreciable drying toward the exterior is with vented roof sheathing, and in foggy-dew Puget Sound even drying into the venting can be slow in winter.

    To air-seal the exterior Tyvek you can just wrap the rafters with the stuff rather than cutting it up. Staple it to the sides of the rafters at the depth where your cavity insulation will be. That way you only have to caulk or tape the edges and overlaps of adjoining sheets of Tyvek.

  31. Jan Verschuren | | #31

    Felt's wet permeability goes up to 60 perms??
    Dana,

    Looking on the web I see that felt #30's perm rates are different than what you say.
    http://www.jlconline.com/underlayments/what-happened-to-30-felt.aspx
    "Felt’s perm rating varies. Dry, #15 felt is rated at 6 perms, #30 felt at 5 perms. When wet, however, felt’s permeability increases to between 20 and 60 perms, according to Martin Holladay of TheEnergyNerd.com."

    Where is the truth of this?
    As with a high perm rating under a cedar shake, the roof SHOULD get time to dry out outwards in my 4C climate?

    http://buildingadvisor.com/materials/exteriors/sheathing-wrap/
    "Asphalt Felt. The old standby, often called “tar paper,” allows drying with a perm rating of of approximately 5 (more when wet) and blocks water effectively on a vertical wall, making it an effective sheathing wrap .
    Asphalt felt has other desirable qualities as well. If water leaks into a wall, the asphalt felt will tend to soak up the excess water (as will lumber and wood products like plywood), helping to store the water until drying conditions improve. Also, once wet, the permeability asphalt felt actually rises from 5 to as much as 60, further promoting rapid drying. This is the same characteristic engineered into some of the hi-tech “smart” sheathing wraps. In contrast, plastic housewrap does not absorb water and tends to trap liquid water (although it will permit slow drying by letting water vapor pass through.
    While plastic housewraps now dominate the market, some contractors have returned to asphalt felt as a weather barrier on sidewalls. In addition to its good water-management characteristics described above, some people find that felt paper is easier to install because of its rigidity and narrow 3-foot roll width and easier to properly tie into flashings. Felt, however, will tear if not handled carefully, and tends to get brittle and deteriorate from UV if left exposed to sunlight for
    extended periods. Otherwise, asphalt felt re¬mains a good option for modern homes.
    However, remember that what is sold as #15 felt today is typically half the weight of traditional 15-lb. rag felt, which weighed – guess what –15 pounds! To get something comparable today you need to buy unrated #30 felt which weighs 15 to 20 lbs/square, or special-order ASTM-rated #Type 1 felt paper which weighs 11.5 lbs/square (ASTM D226). For certain critical applications, you may want to use ASTM-rated Type 2 felt, which weights 26 to 27 lbs./square. The modern felt papers are made from recycled cardboard and sawdust rather than cotton rags, but have similar properties."

    Jan

  32. User avater GBA Editor
    Martin Holladay | | #32

    Response to Jan Verschuren
    Jan,
    You ask, "Where is the truth?" when it comes to figuring out the vapor permeance of #15 and #30 asphalt felt. Here's the truth: there is no single answer to the question.

    If you look up the vapor permeance of #15 or #30 asphalt felt, you'll find that different sources give different values.

    I addressed this issue in my article, All About Vapor Diffusion. In that article, I wrote:

    "If you try to create your own table of permeance values for building materials by collecting information from multiple sources, you will soon discover that different sources provide different values for common materials like #15 asphalt felt or 7/16" OSB. There are three reasons for this:

    • Materials produced by different manufacturers have different characteristics.
    • The permeance of some materials varies depending on moisture content, making permeance measurements difficult even when laboratory technicians try to use consistent test procedures.
    • The ASTM E96 procedure is difficult to perform and often produces inconsistent results.

    William Rose, a research architect at the Building Research Council at the University of Illinois, elaborated on this last point in his landmark book, Water in Buildings. Rose wrote, ‘It is very difficult to conduct this test [the ASTM E96 test] well, and it is easy to wet the specimen in doing the wet-cup test. ... Unfortunately, results from ASTM E96 are not particularly dependable. Several round-robin test of water vapor permeance have been conducted, resulting in a wide range of values.’"

  33. User avater
    Dana Dorsett | | #33

    The vapor permeance is variable
    The vapor permeance of asphalted felt varies with moisture content by about an order of magnitude from when it's ultra-dry to when it's wet. The "rated' values typify it's performance at a fairly substantial moisture content, not bone dry, but a 5 perm rating isn't particularly useful on a shingled roof.

    It's fine for allowing walls to dry, but not in roofs, which are constantly being re-wetted by dew & rain. Roof decks to not dry toward the exterior very much in any climate (with the exception of metal roofs mounted on purilns), but that's, ESPECIALLY true in a Pacific Northwest winter climate, where the wetting is almost daily, and the outdoor air is nearly saturated most of the time.

  34. User avater
    Dana Dorsett | | #34

    The vapor permeance is variable
    The vapor permeance of asphalted felt varies with moisture content by about an order of magnitude from when it's ultra-dry to when it's wet. The "rated' values typify it's performance at a fairly substantial moisture content, not bone dry, but a 5 perm rating isn't particularly useful on a shingled roof.

    It's fine for allowing walls to dry, but not in roofs, which are constantly being re-wetted by dew & rain. Roof decks to not dry toward the exterior very much in any climate (with the exception of metal roofs mounted on purilns), but that's, ESPECIALLY true in a Pacific Northwest winter climate, where the wetting is almost daily, and the outdoor air is nearly saturated most of the time.

  35. Jan Verschuren | | #35

    Not high tide all the time!
    Dana,
    Sure it is wet here. But is it really to that degree that there is outward drying on the skip sheathing, felt paper, cedar shake roof that is equal to a steel roof??
    And regarding Martin's publications on permeance. In the list the felt 15 is still rated as 60 when wet... But wet means that we have a roof leak! More relevant is the dry rating of 5.
    Regardless of the inaccuracies of the testing methods: there is insufficient vapor diffusion to be had... Again: is our cedar roof = a steel roof?
    ANY roof is in that case vapor tight.

    So we need to be vapor open to the inside. "He says reluctantly..."

    Do the same as what we did with the steel roof. Loose 25% of the Warmboard and span MemBrain over the rafters. Add insulation for thermal breaks. Install WarmBoard with gaps. Latex paint the drywall.
    Do not like to loose the WarmBoard and add the MemBrain, truly was not expecting that.
    But that is the end verdict here?

    And regarding permeance... At what point does permeance end and air openness begin? This as I see Tyvek having a 77 as permeance. Typar only 11. This in connection to creating an as airtight cavity for my ProLoft.

    Thanks. (sorta...:-) )

  36. Jan Verschuren | | #36

    Venting impossible
    Dana, Martin,
    A venting channel is only partially possible. See the pics attached. Venting from the gables is not possible either due to the only 4" deep rafters.

    With all these constraints it is now getting into an Apollo 13 mission... I never thought the venting would be a problem on a cedar roof. But reading this 457 newsletter: http://foursevenfive.com/unvented-foam-free-roof-retrofit-with-dolphin-insulation/ it states all can be done, EXCEPT on a vapor drive susceptible cedar shake roof...

    When I PineTar/Linseed Oil 50-50 ratio, treat the cedar roof. Every 5 to 10 year repeat, so this will be a user dependent solution now!
    The rainwater supposedly does NOT soak into the cedar shakes anymore but beads right off.

    Will that not GREATLY reduce my inward vapor drive to a point that my #30 felt (which I assume it is, Building Science rated #15 as vapor permeable?) will give me enough safety to not having to go an inward installed MemBrain smart vapor retarder plus reduction on the WarmBoard? Which may not work as well on this steeper section anyway maybe?

    With the skylights and the perpendicular front door overhang and walk in closet: venting is impossible to properly do.

    Jan

  37. User avater
    Dana Dorsett | | #37

    Don't push your luck.
    Blocking inward drying in an unvented wood shingled unvented roof in a PNW climate is guaranteed to eventually fail. Stick with the smart-vapor retarder + reduced WarmBoard area plan.

    It rains in winter, and frequently. In Everett WA it rains an average 163 days per year. That's more than three days per week on average, but it averages more than 4 days/week from October through April:

    http://www.wrcc.dri.edu/htmlfiles/wa/wa.01.html

    In an unvented roof when the shingles are wet there is zero drying toward the exterior, even if the shingles themselves aren't saturated with water. The outdoor dew point in winter is usually only a few degrees from the outdoor temperature, which doesn't induce very much drying drive even on those 2-3 days per week that it isn't raining. You might get away with it in a climate that rarely rained from November through April so that any wintertime load could dry toward the exterior, but that isn't Everett, WA.

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

    Response to Jan Verschuren
    Jan,
    You wrote, "With all these constraints it is now getting into an Apollo 13 mission."

    You're right. Many people have advised you to loosen your constraints and accept a lower performance goal. If you do, you'll save money. You'll sleep better. And your life will have fewer worries.

  39. Jan Verschuren | | #39

    No venting Martin
    Martin,

    I do not follow your comments. Even loosening my restraints, being: accepting lower R values, I still cannot have the needed venting.
    There is no route for any air go anywhere with this roof. Even if I would built inward or outward, which I cannot do.

    Dana:
    I put a down payment in for the X for you. If you know what I mean. ;)
    Thanks for clearing it all! Good lesson on my misinformation on a cedar shake skip sheathing roof...

    Now comes the execution....

    Jan

  40. User avater GBA Editor
    Martin Holladay | | #40

    Response to Jan Verschuren
    Jan,
    Unvented roof assemblies can be built -- they just require the use of foam insulation. I know that you don't want to use foam insulation; that's one of your constraints.

    I wish you luck, Jan. With Dana's help, you'll probably be able to pull off this Apollo 13 mission. But your constraints sure do make it complicated.

  41. Jan Verschuren | | #41

    Rain days...
    Dana,

    The amount of rain days is a rather crude way of measuring. In the old country I remember a stat using percentages of the time it would rain.
    That would give better info over "rain days" One little shower on one day would make that day a RAIN DAY?
    I remember the percentage of rain time was something like 7 in The Netherlands.
    We get similar rains here, by feel. But it all falls in the fall and winter for sure. No drying period like summer will be sufficient to endure the next rain period again?

    Jan

  42. Jan Verschuren | | #42

    Inward drying CAN be blocked by foam?
    Martin,
    Dana's comment: "Blocking inward drying in an unvented wood shingled unvented roof in a PNW climate is guaranteed to eventually fail.

    Foam is fine? Isn't that blocking inward drying? Skip sheathing and rafters would be fine behind the felt paper?

    The 475 article was all about non vented non foam use roofs...

    Thanks
    Jan

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

    Response to Jan Verschuren
    Jan,
    Back on June 1st, in Comment #13 on your original Q&A thread, I wrote, "If you are committed to your plan, install some cardboard or housewrap between your rafters, followed by R-20 of spray foam, and live with the imperfect thermal performance of that type of assembly (assuming, of course, that your local building department doesn't require a higher R-value)."

    I gave that advice a long time ago, when we were only two days into this long Apollo 13 mission. I still stand by that advice.

    If your cedar shingles have some air behind them, which they will if you follow my advice, the shingles will dry readily to the exterior. The shingles should easily last 25 to 30 years if they were installed well; that's how long my Eastern White Cedar shingles lasted on my roof. In any case, when the shingles finally need to be replaced, you'll be glad -- because you will finally be able to add some rigid foam above the roof, followed by new roofing.

  44. Jan Verschuren | | #44

    "some air"
    Martin,
    "If your cedar shingles have some air behind them"
    Yes, they will always have some air behind them, but that air has nowhere to go or come from. There is no venting to be created on the majority of the roof, as the pics of the roof lines indicate.

    How can they in that case: "dry readily to the exterior"?

    Jan

  45. User avater GBA Editor
    Martin Holladay | | #45

    Response to Jan Verschuren
    Jan,
    Q. "How can they in that case dry readily to the exterior?"

    A. Two ways: (a) By diffusion, since cedar shingles are extremely vapor-permeable, and (b) Through ventilation drying, due to the pumping action caused by changes in temperature. Cedar shingles aren't airtight -- they are leaky. On warm days, the air behind the shingles expands and escapes through cracks; when the weather turns cold (for example, at night), the air behind the shingles contracts and pulls outdoor air inward through cracks. This pumping action is a powerful drying mechanism.

  46. Jan Verschuren | | #46

    Entire assembly
    Hi Martin,
    Yes the shakes are fine and are more open than our front door at times...
    Are you taking into account our entire assembly at all?

    Thanks
    Jan

  47. User avater GBA Editor
    Martin Holladay | | #47

    Response to Jan Verschuren
    Jan,
    I have no idea what you mean by "taking into account our entire assembly."

    You still need to detail the assembly properly. You need an air barrier. You need insulation. You need interior finish materials.

  48. User avater
    Dana Dorsett | | #48

    The PNW weather is different from the Northeast Kingdom.
    On the west side of the cascades it's not unusual to see moss growing on cedar shingles. They are constantly re-wetted, with minimal drying during the winter & spring. In VT they may be covered with snow for months, but at a low temperature. When temperatures rise to moss-growing temp a true drying season begins it's an actual drying season, raining 2 times per week, not the every-other-day rain of the northwest. Roofs in Everett WA throughout the winter are nearly always at a growing temp for molds & moss (mean temps well above freezing mean) unlike VT.

    At the beginning of VT's March drying season they get 12 days of precipitation to Seattle's 16 days, and 192 hours of sunshine to Seattle's 178 hours. But the mean March temp for VT is about 31-32F, to Seattle's much more bio-active 50F. In April VT gets 12 days of precip to Seattle's 14 days with comparable amounts of sun, but the mean temp in VT is ~45F to Seattle's 50F.

    http://www.usclimatedata.com/climate/washington/united-states/3217

    http://www.usclimatedata.com/climate/vermont/united-states/3215

    The average dew point in March & April for St. Johnbury is about 20F and 30F respectively, compared to Everett's 38F and 40F. The late winter/early spring air in VT is much drier.

    https://weatherspark.com/#!dashboard;a=USA/WA/Seattle

    https://weatherspark.com/#!dashboard;q=st.%20Johnsbury%20vt%20USA

    The summer air in VT is more humid than summer air in the Puget Sound region, but in Everett it's in an active tempearture range all winter & spring whereas VTs drying season has been in effect for months by the time the temps reach high activity levels.

    With air both above and below the shingles the shingles can dry, even in the Northwest- skip sheathing usually provides at least a miniscule amount of air behind most of the shingle and those shingle-spacers Jan found provide even more. But that's not necessarily going to be the the case for the roof deck below. If the roof deck is vented to the exterior, or has at least some drying capacity toward the interior it has a chance.

    I'm not 100% convinced that closed cell foam directly on the underside of the skip sheathing would save it under cedar shingles in that climate. It would limit the interior moisture drive issues, but it would also limit the drying rate. Cedar is FAR more tolerant of high moisture content than the hemlock or douglas fir skip sheathing. As long as it was kept to an inch or so of closed cell foam and you didn't have full-coverage low-perm WarmBoard on the interior it would still have some drying capacity toward the interior, though far less than with MemBrain. But none of these stackups we have been discussing come anywhere near to the drying capacity of a vented roof.

    "I put a down payment in for the X for you. If you know what I mean. ;) "

    Echt waar? Te gek! Wanneer kan ik de sleutels opnemen? :-)

  49. User avater GBA Editor
    Martin Holladay | | #49

    Advice for Jan
    Jan,
    My advice is free. So is Dana's. You have to make your own decision.

    My own instincts tell me that cedar shingles on skip sheathing with air behind them can dry to the exterior -- with enough drying potential to keep you out of trouble.

    Dana paints a bleaker picture, and he may be right. If his description of the weather is close to accurate, I can't for the life of me figure out why you chose to install a cedar shingle roof (moss and all) instead of metal roofing. But it's too late to reconsider that decision.

  50. Jan Verschuren | | #50

    #30 Felt
    Martin,
    The question regarding the readily drying to the exterior, and whether you are taking the entire assembly into account were focused on the #30 felt that is over the skip sheathing. Plus no venting channels to be had.
    When it would be 15 weight, Building Science claims vapor openness. #30 would be semi permeable. But as you say: those findings are not absolutely set in stone.

    Dana,
    Not much moss growing on roofs though in general unless they are forestry camp shelters deep under cedar trees... But here and there we do have algae forming which is very local, and exactly linked to shade, morning warmth reflection or not. But the majority of the roof does not seem to linger in eternal water.

    Rob is looking at our WarmBoard needs for our specific cooling wishes. And as a full venting channel just cannot be had, but would absolutely always beat any other vapor route: it is safer to leave an inward vapor diffusion path by spacing the WarmBoard and using MemBrain.
    The fir skip sheathing would sit behind the Felt, and that would just not be that great when it would have alu plates on one side and rain and mist behind the other side of the felt. WHEN that felt would not allow for sufficient vapor diffusion to the outside.

    Sleutel, sleutel? Niks sleutel. Een slimme hollander zou toch beter moeten weten...!
    Maar je creëerde een giga lach hier met je nederlands einde op de laatste respons.. Je naam is verre van ook maar iets nederlands??
    ZEER onverwacht!

    Thanks
    Jan

  51. User avater GBA Editor
    Martin Holladay | | #51

    Response to Jan Verschuren
    Jan,
    Ik woonde in Nederland voor drie jaar toen ik nog heel jong was (1955-1958). Maar ik heb al vergeten mijn Nederlandse, en nu gebruik ik Google Translate.

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

    About the #30 felt
    Jan,
    There are two ways that cedar shingle roofing is usually installed.

    The traditional way is to install the cedar shingles on skip sheathing over a ventilated unconditioned attic. This method does not require any roofing underlayment (asphalt felt). That's what I did when I built my house in 1982.

    An updated version is to install plywood or OSB roof sheathing, followed by asphalt felt, followed by 1x4 skip sheathing.

    Your roofer chose a hybrid approach that I have never seen, with asphalt felt installed on the exterior side of skip sheathing. Did the roofer install the cedar shingles directly on top of the asphalt felt, or did the roofer include a plastic mesh product like Cedar Breather between the asphalt felt and the cedar shingles?

  53. Malcolm Taylor | | #53

    Martin
    In response to perceived problems with wind blown snow penetrating attics through shake roofs, our code was amended so that each row of cedar had to have a layer of building paper in between it and the next. Perhaps that's how jan's was installed? The result of the new method has been that shingle and shake roofs now last half the time they use to.

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

    Response to Malcolm Taylor
    Malcolm,
    The photo at the top of the article doesn't appear to show lapping strips of asphalt felt, with each strip installed for a single course of shingles as your describe. It appears to show that the felt was installed by unrolling the 36-inch-wide roll over the skip sheathing.

  55. Jan Verschuren | | #55

    Why Membrain when we dehumidify?
    Just wondering the point of MemBrain when we will be dehumidifying? All is focused now on drying to the interior on both roofs, metal and cedar shakes with felt roof.
    MemBrain opens on higher humidity: so INTERIOR humidity should never open the MemBrain for that humidity to enter the roof cavity. We have dehumidification planned in as that is going hand in hand with radiant cooling.

    Where does the MemBrain fit in now at all? Airtight drywall with latex paint is vapor open, right?
    So just control indoor humidity, which we have to do anyway...
    We get more vapor openness without the MemBrain? No need to "wait" for the MemBrain to get into it's zone when moist vapors need to get out.

    Thanks
    Jan

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

    Response to Jan Verschuren
    Jan,
    At this point, I don't know what type of insulation you are considering for your roof assembly, so I have no idea whether MemBrain makes sense.

    Even if you expect your assembly to dry to the interior, MemBrain can still be useful during the winter prevent moisture accumulation. During the winter, the vapor drive is from the interior to the exterior (from the warm, humid side of the assembly toward the cold, dry side of the assembly). In that season, MemBrain acts as a vapor retarder, reducing the diffusion of water vapor from the interior of your house toward the cold components of the assembly (in your case, your skip sheathing).

  57. User avater
    Dana Dorsett | | #57

    The MemBrain is necessary as
    The MemBrain is necessary as a drying path for the unvented assembly, since without it there is none. No matter how dry you keep the interior the skip sheathing and rafters can't dry toward the interior if it's covered with WarmBoard, and without vented sheathing it can't dry toward the exterior. You may be able to just use air-tight drywall & latex if you keep the interior moisture levels low, but it's riskier. With MemBrain any moisture that gets into the rafter bays (from any source) that reaches levels high enough to be a problem leaves about as quickly as it would with just air tight drywall & latex paint. But when the rafter bays & sheathing are relatively dry moisture migrates into the rafter bays from the interior space at 1/10th the speed as it would through air-tight drywall & latex.

    The average mid-winter temperature of the skip sheathing is about 40F, so as long as the average interior air has a dew point lower than 40F there isn't much moisture drive. But that is about 30% relative humidity at 70F, which is the very low end of what health professionals recommend. Most homes in the northwest would have air that's 40-45% relative humidity in winter, which has a dew point of 44-48F. While the moisture drive from the is low, the direction is from the indoors out to sheathing, and there is no reliable drying path for the sheathing.

    Ik ben blij dat je en giga lach gehad! (Dat WAS de idee, he'?) Maar ik ben amerikaans en hollands niet dus. De naam engels is zo ver ik weet- na de zuidwestelijk engelse wijk "Dorset County" denk ik. Boven vijf en twintig jaren geleden toen ik in nederland was ik een slag van de molen wiek gehad- daarna kon ik de helft van deze taal verstaan! :-) Het was slechts een tien maand verblijf maar herinner ik nog en woord of twaalf. Na een paar pilsjes ik kan bijna vloeient (helemaal vreselijk) nederlands praten- ik heb repetitie nodig. :-) (Ik vind het iets makkelijker dan het oerdoe/hindi tenminste.)

  58. Jan Verschuren | | #58

    Klap van de molenwiek
    Let Martin run THAT through Google translate!
    Got a touch confused when he replied with a few dutch words after your dutch sentences Dana! It was all getting Double Dutch.

    Membrain it is. Thinking it all through: what about this to hold the MemBrain in place over the rafters, to keep it air tight. Strip of acoustic caulk. And this over top: http://www.supplyhouse.com/Oatey-33927-3-4-100-Plastic-Hanger-Strap-Bagged
    The tack staples for the continuous thermal break insulation can punch right through. As can the WarmBoard screw. Drywall screws go into the WarmBoard.

    And regarding the application of acoustic caulk: FORGET a hand applicator when you have to use it a bit... First cartridge killed my hand from the start. Get a pneumatic applicator! NOW it's WAY better and easier to apply! Super control and steady hand. Use half the caulk and EXACTLY where you want it to go.

    Yes Martin, the Felt was rolled out over top of the skip sheathing. The shakes were than installed overtop of the felt. Don't ask me why. The roof was leaking, it was a phone call and go!

    Looked up the Cedar Breather, interesting product. What we installed on a new out building were these metal plates(pic attached) that create a space between the shakes. Plus creates a metal roof, vented, underneath the shakes. I like that stuff. Will have to see the Cedar Breather but these metal roof plates are a pretty smart invention as well. As with the Cedar Breather probably, best is to install with the new roof install. The metal plates can be installed afterwards but will need screws. And those do not fully penetrate, but I am trying to negate that for the existing roof. Any thoughts on that? They just need to be held in place and sustain the swelling and shrinking of the roof. I'll take some pics of a retrofit of these plates on a house in Canada. It may be the only way to do it but it's not great...

    Insulation of the metal roof will be Spider. For the shakes Proloft.
    Just spent the half day emailing with Johns Mansville and the local rep to look for an installer. None to be had. Only Spider blown in behind bibs. UNTIL I "ran into" a local installer who DOES install Spider with the acrylic binder, without any netting needed. Even for a ceiling install. Nice!
    BUT Johns Mansville now has the Spider Plus, no acrylic anymore but water....
    Tomorrow I'll hear what the full difference there is but so far it looks like the local Spider installer can only do the old fashioned acrylic install. Which is a relief that he CAN do that.

    And Dana: je bent een mooie ouwehoer! "Beter dat als een jonge hoer zonder klanten" zei mijn vader altijd! Hopelijk dat Martin me niet van de web pagina gooit nu... Maar mijn geschiedenis is 30 jaar in Nederland, daarna naar Canada geëmigreerd waar ik 20 jaar gewoond heb. Nu 4 jaar in de USA, getrouwd met een amerikaanse! Met een hoop werk aan een oud huis! MAAR met een 100A laad station in huis voor onze natte droom!

    Thanks for all your help and patience!
    Jan

Log in or create an account to post a comment.

Related

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |