The following stumper is presented in BPI's February Performance Matters e-Newsletter:
Homeowners of a two story balloon frame home located in Watertown, NY wanted to eliminate ice damming on their roof. An insulation company was called in during the summer for recommendations and possibly to provide retrofit work. While there, the company discovered the entire home was un-insulated. The company proposed a comprehensive work scope, but the homeowners wouldn't bite – they agreed to pay for just one measure. This was to air seal and insulate the attic floor (the cavity between the 2nd floor ceiling and attic floor) with dense pack cellulose. During the next winter, the ice damming continued on the roof as with previous winters with equal snow fall. Both floors of the home were heated and had a semi-conditioned full basement.
Question: Why were ice dams continuing to form on the roof after air sealing and insulating the attic floor?
Think you know the answer?
Figure out what is going wrong with the house, write it up along with your prescribed solution, and share your wisdom by posting your answer right here in our comments section. You can also send it to us at firstname.lastname@example.org. If you’re the first person to get the right answer, we’ll feature you, your company and your answer in the next issue of Performance Matters!
hmmm, Maybe because air sealing the attic floor redirected stack losses at the top of the walls out the soffits, then up along the roof.
A: Warm air pushes hardest to escape at the attic level. Even though the attic level was air sealed, the balloon frame without insulation nor air sealed has the next greatest stack effect to force air leakage. Ice dams form when snow melts and runs down the roof to the eaves, when it refreezes and backs up into the eaves or the house, causing damage.
But shouldn't a good attic air seal include blocking off the tops of the walls, so that warm air would not then reach the soffits from the uninsulated walls. It seem unlikely that if the contractor had done the complete air sealing of the attic, that that would have eliminated the ice dams (though) there would still be extensive heat loss through the walls).
I really like the colander analogy!
Everybody is right on, so far. Stacy's question, though is probing the where and how. The air seal job was measured from inside the living space in reference to the attic. THAT was a tight seal. Those pesky walls, though, remain suspect, huh?
Those balloon-framed walls may or may not have had top plates. Flip a coin, but even if they had them, anybody want to bet they were enough to stop warmed up air from crawling up the inside walls on a cold winter night when the house is toasty, warming the top plate continuously and then the roof and melting the snow only for it to refreeze, as usual on its way across the overhang and then, well, you know the rest.
Some would say 'karma' got 'em' for trying to piece-meal with one single FIX a system interface more like a big-city interstate highway, railway, or seaport than he'd realized. A warm roof will always melt the snow; the source could be this or it could be...and now we know more than when we started so we're on our way, but not quite there until the attic is cold and the walls are the next likely target for cellulose dense-packed.
But why not an infra-red scan to see what we're missing in order to do it as 'right the first time' as we can? Checking the work is part of the job. And though sealed, the sealed and insulated attic was only a piece of the ice dam problem. The warm roof remains, and the walls remain the likely culprit with radiation and conduction the companions in crime, now that the convection has enjoyed its free reign over the wall cavities.
Thanks Joe, hope that will work for you.
Stacy, visualizing the way walls connect to attics is tough. Remember these are separate components.
The original path of least resistance may be top and bottom. But remember that leaks aren't just at the top and bottom of a component, they often leak throughout.
So even if the attic flat does connect to the wall tops (balloon frame, vs having the wall tops below the attic), the leaks are simply chased to a little lower point.
If the basement rim had also been addressed this job may not have been deemed a failure. Hitting primary entry and exit tends to have a much more dramatic effect than simply hitting one or the other. Get the big holes high and low first, right?
The attic floor is insulated and air sealed, we can all agree on that. We don't know the cavity depth so hard to say if enough R-value was added to slow down the heat loss enough to prevent the warming of the attic and then the ice dams as a result of that. Let's assume that the insulated floor is not the problem. That now leaves us with a perimeter of exterior wall plates with no insulation on top of them or below (remember this a floored attic) so now each empty wall cavity around the perimeter of the attic is acting like a chimney and causing the problem. Had the exterior sidewalls been dense packed or an effort made to target dense pack the outer wall plates during the floor insulation process the problem would have been resolved. I am confident that during the dene packing of the floor that the insulation was not packed all the way to to the exterior sheating. As mentioned those were balloon framed walls so the insulation would have just trickled down if it was even able to reach that far and more likely would have hit blocking leaving the outer cavity open. So now that should make sense why we still have that snow melt around the roof edge. R-0 maybe close to R-1 if you take the R-value of a typical 3/4 floor board around that crucial area. Solution is dense pack the exterior sidealls or a cheaper alternative target dense pack using the "bag method" at the perimeter.