Joe Lstiburek says: The only reason to vent a roof is to prevent ice-damming
The dean of building science, Dr. Joe Lstiburek, is out with a new* article where he says: “The only time you should vent a roof is to control ice-damming. And when I do a vented roof to control ice-damming I install the vented roof over the top of an unvented under-roof.”
The whole article is here:
https://buildingscience.com/documents/building-science-insights/bsi-141-shakespeare-does-roofs
This is a pretty radical departure from what seems to be the conventional wisdom. His introduction:
“Vented roofs blow off more than unvented roofs. Vented roofs burn more than unvented roofs. Vented roofs are energy inefficient compared to unvented roofs. Vented roofs cause moisture problems south of the Mason-Dixon Line and east of Interstate 35 in Texas. Venting a roof in a hot-humid and mixed humid climate is a very, very bad idea.”
I have to quibble with that last part, Washington, DC, is south of Mason-Dixon and has tens of thousands of houses build in the decades after the Civil War. Virtually all of them have vented roofs and they seem to last OK.
Something else he says that I disagree with: “In cold climates roof venting controlled attic moisture accumulation only in poorly insulated attics… In the winter the air outside is cold. Yup. Cold air is not capable of holding much moisture. So bringing cold air into your attic does not remove moisture unless there is heat loss from the building. The heat loss warms up the incoming cold air giving it the ability to pick up moisture and carry it to the outside.”
Even in the winter, when the sun hits a roof, the attic warms up. That warming is enough to drive air exchange. In my experience, even in well-insulated houses the wood in the attic is desert-dry.
*(OK, new to me, it came out in March).
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Two thoughts:
1) In a cold climate, because of the greenhouse effect you mentioned, vented attics can actually help reduce the delta T across the ceiling by having a warmer-than-ambient attic and reduce heating costs. But the reverse is true in hot climates, where the increased attic temperature above ambient is actually a cooling energy penalty.
2) The surface area of the roof deck vs the surface area of the attic ceiling difference matters, and must be combined with #1 above to determine which strategy works best for your roof type (gable vs hip), pitch, and climate. The difference on gable roofs is not as much as hip roofs.
"Washington, DC, is south of Mason-Dixon and has tens of thousands of houses build in the decades after the Civil War."
I expect that these would be homes that are very well "ventilated", i.e. have little or no insulation, skip sheathing, and allow air to pass through like a sieve. This is the case with virtually all 19th century wood frame construction, and to the extent that wood is used in brick homes, brick homes as well. Build a home like that today and you're still likely to have very little in the way of moisture/vapor related problems. Unfortunately you'll also have lots of drafts and sky high heating/cooling bills.
Thanks for the link to the article. Lstiburek always has interesting and useful things to say.
I'm not prepared to make a pronouncement from on high that conditioned attics superior in ALL cases. But "do the math" and I suspect they will usually win out, in all parts of the country. Building a new home today I would start with the assumption that the attic will be conditioned and work from there.
I'd love an unvented roof on my house to help seal it from Asian lady beetles and mice, however, 5" of exterior insulation is expensive, requires unusual details, and is slow/ difficult to install within a weather window. So to answer Joe, there is another at least one more reason for a vented roof, and that is money.
Figure 3 in the article shows r-60 above the decking. That's ten inches of polyiso or fifteen inches of EPS.
Jo has been known to write the code book I can barely read them but when I try, I fail to see the language to allow the unvented and unconditioned attic some have imagined this article is advocating.
The old quote from Jo that stick in my mind go something like the vented attic works in every climate zone and cost much less to build.
Yes winds could blow off your roof but is that really a common failure mode.
Yes a wildfire could get inside an burn down your house but not a likely event in my neighborhood.
I like my vented roof I think I will keep her.
https://www.youtube.com/watch?v=_lNfpQ_8tEw&ab_channel=MichaelChiu
Walta
I can't comment on the aspect of what happens south of the M-D line, but I do question the idea that venting roofs in cold climates is only helpful if there is heat loss from the interior.
I also have a HUGE reason why venting roofs is a good idea: it allows you to use insulation that has low levels of up-front carbon emissions, or even negative emissions, and to save money. Joe likes foamed roofs because they are easy and reliable. They are not inexpensive or low-carbon.
Michael,
I think another reason to use vented trussed roofs is their adaptability over time. Houses built that way here - especially those on crawlspaces - are very simple to renovate, as the living areas are sandwiched between two accessible service spaces.
At least in climate zones 1-3 you could use vapor diffusion port, and I assume at some point there will be something similar for climate zones 4 and up but who knows.
It also allows you to use less insulation for the same effect -- the ceiling line is shorter than the roof line.
I agree that carbon reduction, durability, and energy efficiency should all be prioritized. Can we have our cake and eat it too?
What about this unvented roof assembly-
rafter bays with dense-pack permeable insulation in contact with sheathing…
combined with air-sealed smart vapor membrane at the ceiling plane…
And a vapor diffusion port at the ridge.
This uses the same strategies we use for water management at walls and windows regardless of climate; prevent moisture from getting in, while allowing it to escape.
Climate zone would still dictate other details of course, such as roof color, interior moisture management, etc.
Assuming it’s detailed correctly, wouldn’t this assembly be low-carbon, durable (from a rot, fire and wind standpoint), and energy efficient?
Adam
475 seems to advocate for this, even in northern climates. However, the code does not allow such an assembly north of climate zone 3 if I remember right.
Is that because of ice-damming? We’re really only talking about unvented roofs where ice-damming is not a problem. But maybe there’s another issue with colder climates that I’m not thinking of?
I’m in Marine 4c, so technically it’s not allowed…yet we aren’t at risk of ice-damming. I’m not sure I understand why it this wouldn’t work here.
Adam
It's not due to ice-damming, but due to moisture accumulation in the sheathing. The colder the climate, the harder the moisture drive. They did a test in climate zone 5. The cellulose results seemed ok, but the overall assembly seems susceptible to slight errors in construction. A small air leak in this assembly can cause big issues.
https://buildingscience.com/sites/default/files/presentation-docs/2020-08-14_nesea_be18_ueno_unvented_roofs_for_pdf.pdf
With that said, I believe the owner of this house used 475's detail for intello on the inside and dense pack to the sheathing in the rafters. I guess it depends on your code official if you want to take the risk?
https://www.greenbuildingadvisor.com/green-homes/a-first-time-builder-demonstrates-a-keen-understanding-of-building-science-part-two
One final thought - if going that route, I think you could top vent the sheathing for some extra protection. First layer of structural sheathing, followed by 1" batons, second layer of sheathing for the roof. Not sure if this is code compliant or not. It doesn't buy you a ton other than that extra inch or two of space in the rafter bay.
Tim_O,
"if going that route, I think you could top vent the sheathing for some extra protection"
Aren't you really just back to a vented assembly then? Practically it functions just like a typical cathedral ceiling with plywood baffles.
I'm with you and Michael: Using permeable insulation in an un-vented assembly and relying on vapour diffusion ports to remove accumulated moisture is risky.
Yes, you're right, it's identical to using plywood baffles. Allows you to monopoly frame I suppose. You could just use the batons on the bottom side of your joists to achieve more depth too. Whichever you find easier to install.
There was a great report on this recently. The conclusion was that it's doable, but requires a lot of attention to detail, and it may still have some moisture issues due to the insulation settling and minor air leaks even with meticulous building practices. Thus it was only recommended in CZ3 and below. Added link: https://www.nrel.gov/docs/fy21osti/77518.pdf
Do you mind sharing a link to that report? That would shed some light on the issue.
https://www.nrel.gov/docs/fy21osti/77518.pdf
BSC tested similar assemblies in their test hut and found that they are not safe for cold climates. Maybe future developments will change that but I'm not optimistic about it.
Nearly any assembly will be ok for 5-10 years. The question is what happens after 20 years or more--is the house still safe from mold and fungus or is it being eaten while you (or others) live in it?
I'm not sure how significant this is but here's some recent research from RDH on the matter of heat loss affecting roof drying https://www.rdh.com/wp-content/uploads/2019/10/TB-10-Attic-Mold-in-Vented-Wood-Frame-Roofs-2015-12-17-3.pdf
Very interesting stuff. Money quote:
"In recent years, a trend towards more airtight and highly insulated roof assemblies (R-40+) has been pursued to reduce building energy consumption, and this trend seemingly correlates with an increased frequency of attic moisture problems. While airtightness improvements have decoupled the vented space from interior moisture loads, increased airtightness and insulation have also reduced heat loss during cold weather, reducing the drying capacity of the roof sheathing. Consequently, phenomena that may previously have been masked by excessive heat loss in older, poorly insulated and air-leaky roofs can become apparent in modern well-insulated and airtight systems."
I wonder if it is appropriate to say that this is similar to the experiences we have had as we have begun insulating walls better, or if walls and roofs don't behave similarly enough from a moisture management perspective to draw that comparison.
I'd say it's similar to what people sometimes report when beefing up the insulation on an old house. The walls let a little water in, but they leaked enough heat so it wasn't a problem.
Joe L has been known to retract his own statements/ideas at future times too. That's OK -- sometimes initial conclusions might not be correct, or something gets missed. He works out on the edge, trying new stuff, and it takes time to see what works or not. I can see some of his logic here, but in other areas I don't agree with him. I do agree with Michael (post #6), in that vented roofs allow more types of insulation to be used. My goal is usually more economic than carbon-related, but in this his and my goals are aligned: loose fill cellulose is just hard to beat for green-ness, cost, or performance.
Rigid foam is probably the most expensive way to insulate, and it's pretty high labor too -- especially compared to blowing loose fill insulation onto an attic floor. I'm wary of going too far over onto the side of ultimate performance because of what it does to bottom line costs and affordability of the resulting home. There needs to be a balance. There is a good reason why the "perfect wall" is rarely built after all.
Bill
He does seem to love foam and is moving toward the ideal home being a Styrofoam cooler.
He lays out his reasoning explicitly in many of his articles and lectures. He doesn't care what type of insulation it is so long as it is moisture tolerant and can be placed on the exterior. He repeats this in nearly every article and lecture he's delivered.
His reasoning is that keeping the control layers and structure dry, cool, and dark will help them last longer and improve source control for indoor air pollutants. Whether or not this is truly necessary is up for debate.
I don't know if anyone has tested whether it is cheaper to replace damaged exterior insulation or if it's cheaper to replace a damaged WRB. I can only speculate that the logic is that a building with neglected and damaged exterior insulation will last longer than one with neglected and damaged WRB. Or maybe it's the condensation control factor. We'd have to ask him.
If you start with an old house with t&g roof sheathing and clipped eaves; the article seems to imply you could do worse than adding an over roof and then dense packing to the original roof deck. Cheaper insulation and install than 5 inches of rigid exterior, and some drying in case of a moisture problem. Or am I thinking the vented over roof can do more than it actually can?
Assuming climate zone 5 and no current ice dam problems.