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

Community and Q&A

Roof Assembly for Very High Snow Load

lenkov | Posted in General Questions on
I want to design vented roof with 4″ of outside rockwool insulation in a snow country (Zone 6B, snow load: 250lb/sq.ft). I was considering the following roof assembly:
– 12″ TJI (with Fiberglass batts insulation between them)
– Sheeting
– Weatherproofing Membrane
– 4″ Rockwool mineral wool
– 2×3 screwed to the rafters with ~7″ screws (for venting)
– 3/4″ plywood decking
– Ice Shield
– Shingles
But I’m concerned will this handle the 250lb/sq.ft of snow load.
Are there better options to handle high snow load and without creating of thermal bridge?
Some details:
– new construction
– snow load: 250lb/ft^2
– roof slope 12:2
– climate zone: 6B

GBA Prime

Join the leading community of building science experts

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


  1. Deleted | | #1


  2. Expert Member
    Michael Maines | | #2

    That's a heavy snow load! The most I design for locally is 90 psf. 250 psf is 1.74 psi, which shouldn't compress Comfortboard too much. One alternative I have used on a retrofit, which might work for you: horizontal 2xs above the roof sheathing, 2xs on edge 24" o.c. (you might need 16" o.c. for your snow load) and infill with mineral wool batts instead of Comfortboard. You get an integrated vent channel and mineral wool is fairly resistant to wind-washing.

    One note: a roof slope of 2:12 is the bare minimum needed for venting. Is there any chance you can make it steeper? In any case, the deeper the vent space, the better.

  3. Jon_R | | #3

    This is a weak hybrid assembly, relying somewhat on warmer sheathing and somewhat on venting, with neither alone being adequate. So make sure the "Weatherproofing Membrane" layer is highly vapor permeable, allowing "flow though". Air seal the interior side well and consider a smart vapor retarder.

    Consider simplifying by not doing a hybrid and using only a large, under sheathing vent. Ie, > 12" TJIs.

    For ice dams, think about how you can avoid colder-than-roof eaves. The energy needed to do this is quite low.

  4. walta100 | | #4

    If that snow load is real one of the big problems will be the R value of that much snow will be much greater than anything you could but under it. Then the bottom layer of snow will get warm melt and refreeze on any overhang building huge ice dams.

    I remember see a Joe Lstiburek YouTube video about his work at ski resorts. If this is a real project with a real budget I say hire Joe. If you have no budget spend 10 hours on YouTube and watch all the videos you will learn a lot and have some laughs


    1. Expert Member
      Michael Maines | | #18

      I wouldn't say "much greater"--250lbs at 20 lbs/ft³ is 12.5' of snow depth. Snow averages about R-2 per inch, or less when compressed, which it would be 12' down. So at most, the snow would be R-24, well below the roof assembly R-value. BUT you are correct that this will likely lead to ice dams, because R-24 will shift the dewpoint location toward the exterior.

      1. Jon_R | | #19

        > at most, the snow would be R-24

        12.5 ft x 12 inch/ft x R-1/inch = R-150

  5. Expert Member
    Akos | | #5

    That is a lot of layers, makes for a complicated and expensive install.

    I would take the money of all those sheet goods, bump up your TJI to either 14" or 16" and build a simple vented roof with either loose fill or batts. The extra height would give you plenty of space for venting above the insulation without any of the complication of exterior rigid.

    I-joist have very narrow webs and have much lower thermal bridging than dimensional lumber, the assembly R value is about the same as batts, you don't need exterior rigid insulation. Make sure to order batts for metal studs as these are wider and will fit snugly between the OSB webs.

    I don't know of any shingles rated for 2-12 slope, most can go as low as 3-12 but even than it is questionable. When it comes vented shingle roofs in snow country, steeper is better.

  6. Expert Member
    BILL WICHERS | | #6

    +1 for deeper TJIs and a vent channel. A 2" vent channel may make sense here to help minimize ice damming too -- more airflow = less temperature differential, since the moving air will help to carry away any excess heat.


  7. kbentley57 | | #7

    I'm from the south east, so forgive me for asking!

    Could that snow load possibly be correct? I've looked at maps, and even in places in Canada I'm seeing no more than 150 lbs/ft^2. I might be in for a beating, but I'd need to see a source showing a 50 year snow load that was even half that.

  8. lenkov | | #8

    Thank you all for the feedback.

    Just to clarify few things:
    1. I'm trying to follow closely Lstiburek's advise -- the house is air-tight ( the sheeting will be covered with a peel and stick airtight membrane ) with continues control layers. My goal was to have cold roof above hot roof, in order for the hot roof to be airtight.

    2. I got the 250 lbs/ft^2 show load from the official site of Town of Truckee:
    When I click on the map and select my lot, I get 247 PSF.

    3. If I go with a simpler solution with just 16" TJI and venting between the insulation and the sheeting, I'll have to look for an alternative way to airtight the house. And considering that I won't have an attic (the drywall will be directly attached to the rafters and there will be plenty of penetrations for the recessed lights and fire sprinklers that you can't possibly air seal), this may not be trivial.

    Any feedback will be highly appreciated.


    1. kbentley57 | | #9


      I stand corrected! Thanks for the link.

      *Removes Truckee from list of retirement locations. :)

      1. user-2310254 | | #14

        The Donner Party probably had similar thoughts.

    2. Jon_R | | #10

      > I'm trying to follow closely Lstiburek's advise

      As I read it, Lstiburek's high snow load advice is a strong hybrid - all of the insulation above the sheathing and a vent. And a mandatory interior side air barrier. A little insulation above the sheathing, an undersized vent and no interior side air barrier is not equivalent.

      Consider an interior side service cavity (allowing a good interior side air barrier).

      Typical ridge vents (if that's the plan) with lots of snow over them aren't effective (for moisture removal or cooling to reduce ice dams). Consider alternatives like a mini-attic at the peak with gable exit vents. Hopefully the snow won't block them (or the soffit vents).

      Could be worse - some of your neighbors have to build for > 400 PSF snow loads.

    3. josh_in_mn | | #12

      For the recessed lights, consider LED disk lights, which can be installed in a regular octo box or round box, which can be air sealed. Also, they're much thinner so they don't displace so much insulation. Lastly, they're easier to install!

      If you do a service cavity, as others have suggested, then you could even use the ones that simply clip into a hole in the drywall.

  9. Expert Member
    Akos | | #11

    The simplest is to have a warm side air barrier under the I-joists. I usually connect this air barrier to the wall sheathing with a wide piece of flashing tape over the wall top plate.

    The air barrier could be either one of the many specialty membranes, or any sheet good that can be taped (foil faced rigid insulation, OSB/CDX).

    A 2x4 service cavity can be framed bellow this would let you install your lights and sprinklers.

    Even with a roof with hybrid insulation install you originally proposed you would have to do a decent warm side air barrier, the air barrier and service cavity is something you would need anyways.

  10. lenkov | | #13

    Thanks again for the insightful answers and ideas.

    One last question: If I decide to go with a hybrid assembly, and use a 4" exterior insulation. Can I use mineral wool (I'm sure it will compress a little bit under 250 lbs/ft^2) or I need to go with something like EPS (hopefully won't compress)?

    I'm planning on having a 2 1/2" air gap between the insulation and the roof deck ( will use a vertical 2x3, screwed though the 4" insulation into the rafters with ~8" screw).

    Thanks again!


    1. Jon_R | | #15

      You are way above 25% compression for Comfortboard 80 and around 10% for 25 psi foam (which your design shouldn't use because it restricts vapor flow-through). This is assuming the screw shank and plywood let the 2x3 move downward - meaning the screw head will pop up. And there are lots of other roof design issues when it comes to heavy snow load. So I'd hire an engineer.

  11. Expert Member
    Akos | | #16

    Assuming 16 oc strapping, your 250 lb/sqft snow load puts 350lb/linear foot onto your 2x3s. That means around 20PSI load. Comfortboard80 is around 3PSI for 10% deflection, so it will definately get squished. Comfortboard110 is only a bit stiffer, so won't help you much.

    Your best bet is to go for other type of rigid insulation and install the 2x3s on flat. This increases the surface area and reduce the amount of local pressure on the foam.

    The other issue with your proposed roof is you don't have enough exterior insulation for condensation control.

    If you look at the table here, you need 50% of your assembly R value to be above the roof deck:

    4" of rigid is nowhere near enough with 12" I-joists. You need to reduce the amount of fluffy bellow the deck and increase the rigid above it.

  12. kbentley57 | | #17


    There are several densities of rigid mineral wool insulation, with the density roughly corresponding to its resistance to compression. Rockwool and others make a roofing version with a density of 13.75 lb/ft³, that's about as high as it goes (that I've seen). They list the compressive strength at 11 lb/in^2 @ 10% compression, or 1584 lb/ft^2. I would feel safer if I could trial it out with a mock-up or make a conservative calculation to see how much it moves under the point-like nature of that load.

    Here's the link I found that information on. They list a point load of 30 psi @ 5 mm compression, fwiw.

  13. walta100 | | #20

    Seems to me when faced with such unusual conditions the safe move is to copy the time tested designs your neighbors have used for decades. They may not be as energy efficient as you would like but you can be sure they will not be a disaster.

    If you must try something new the smart move is to pay expert enough money for them to understand you local conditions and design something likely to work.

    Free internet advice is generally worth what you pay for it.


  14. Expert Member
    Dana Dorsett | | #21

    >"– roof slope 12:2"

    Do you mean 2:12 pitch? Snow won't stick to a 12:2 wall- it's too steep.

    At such a low pitch roof venting isn't going to provide much benefit AT ALL since the convection forces are weak. There is a reason roofs in snow country tend to vary from steep to very steep or even "lightning splitter" steep.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |