![](http://images.greenbuildingadvisor.com/app/uploads/2018/07/24211528/Pat%20Dundon%202_0-main-700x824.jpg)
By Pat Dundon
Never tell anyone that you will stop ice.
I went to the building science conventions and thought I was bulletproof. But it turns out … not so much.
I experienced this revelation at a farmhouse near Syracuse, New York. It had a typical second-floor exterior wall, about 5 feet tall, then a 3-foot-wide sloping ceiling, and then a very tight attic. Someone had added a large two-story T-shaped addition with some sloped and flat ceilings to make it interesting. It also happened to be 70 miles out of town. The owner is a great client who happens to be a personal friend of a GC I do a lot of work for.
The owners had a big problem with ice. The addition had a Swiss-cheese ceiling, with slopes, valleys, an interior hidden Death Valley, and about 20 recessed lights — all insulated with fiberglass. I thought it was a lucky break that the recessed lights in the addition were “airtight” cans.
We’re hired to cure the ice dams
There was a single-story section with a bead-board interior ceiling that was an obvious problem that we felt we could fix. I thought it would be simple to do some air sealing, box the lights in the old part, get rid of the fiberglass, dense-pack the slopes, use foam to detail the ceilings, and be a miracle worker for them. So they invested about $5,000 with me and we did the deed.
We accessed the single-story attic through a second-floor wall. We removed the existing fiberglass, sprayed down an inch of foam, and blew in cellulose. That worked great.
A few surprises and stumbles
We had some trouble too. Several of the recessed lights in the slopes were too far down-slope for us to reach from the attic, so we tried dense-packing around them. That didn’t work.
The R-38 batts in the slopes were kraft-faced and stapled in place, so we could not remove them. We tried dense-packing them, but that didn’t work.
The Death Valley was a place we tried using foam to spray the part of the wall above the attic floor. That didn’t work.
On the older part of the house, where interior partitions intersected the roof, the studs ran up into the rafter cavities, so when we tried dense-packing the rafter cavities, we left a void at the partition top plates. Oops.
The ice returns with a vengeance
Other than the single-story section, we had total failure. The ice was as bad as before. We did the work in summer, so my blower door and IR camera weren’t as useful as I wish they would have been, but even in winter I had trouble locating some of the problems. I did find some and spent a day and a half with 4 guys and some foam and cellulose to fix those.
On conjecture, I decided to fill the interior partitions in the old part with cellulose to see if that stopped that ice.
We wanted to do the same with Death Valley, but the owners said they didn’t want the finishes in the nursery destroyed, so we didn’t do that. Pesky clients.
I could not think of any way to improve the recessed lights in the slopes without removing the roof. The owners said no way to that, so that was left.
That was on my dollar. Approximately $5,000 spent (invested?) there.
After Round 2, success was still elusive
The second round of work was a semi-failure. There was still one huge leak I was still missing — a leak I could not figure out in the older part. The lights were melting sufficient snow to leave lethal-sized icicles hanging off most of the addition.
The following summer, the client decided to remove the roof, and we were set to remove small areas to approach known problems only. The roofer said it would be easier to just take it all off and deal with it. The GC friend said, “Just do it all.” So the owners had the roof removed.
Round 3
With the roof off, we could see that the cellulose we tried to get below the recessed lights did nothing. It was not dense enough nor was it able to move laterally well enough to fill the gaps in the batts. None made it more than 1 foot beyond the lights. The slopes were about 5 feet long and the lights were about 1½ foot down from the top.
We were able to get to the partitions around Death Valley and fill them with cellulose.
We found the partitions in the old part were full, but the lath created a barrier to the cellulose that left a square cylinder the width of the partition and height of the rafter directly over each partition that was not reachable without removing the roof.
The big leak was a false ceiling. There was a bathroom in the original second floor that was remodeled. Apparently the plaster had fallen off the lath in that bathroom, so the remodeler put up furring and a new sheetrock ceiling, which left a space about ¾ inch thick, the same size as the bathroom.
You would think the foam we sprayed on top of that ceiling in the first place would have fixed that, but what I didn’t catch was this: one long side of the furred area dead-ended into the sloped ceiling. The original rafters were 3x4s on 30 inch centers. You can’t get a foam gun into the slope from the attic in this situation. If you don’t know that gap exists, you will not catch it from the attic.
We had tried to dense-pack the slopes. The attic was not vented, and the rafters were not vented, so the blower door was not pulling air into this void. Once the roof was off, this void was obvious, and we repaired it.
A costly lesson
The owners paid for the roof to come off and go back on. The owners paid us $4,000 to do the third attack on this roof. My job costing tells me I actually spent about $10,500 in labor, overhead, and materials, not to mention lost opportunity elsewhere, on this third approach.
My lesson is to look at these jobs much more suspiciously before suggesting a solution. I’ve learned to use my knowledge of the limitations of the planned approach to set client expectations at a deliverable level. That is a big limit to my business because I don’t know how to transfer the technical ability to a sales staff.
Pat Dundon, a.k.a. The Insulation Man, is the owner of Dundon Insulation in Windsor, New York
Editor’s note: A version of this piece was originally posted on one of the JLC Online Forums. GBA would like to thank Don Jackson of JLC for granting permission to reprint Pat Dundon’s piece.
Weekly Newsletter
Get building science and energy efficiency advice, plus special offers, in your inbox.
3 Comments
can lights
Did you try removing the can lights and sending the hose all the way down from there? If you couldn't go in from the ceiling, could you have gone into the sidewall, behind the cladding? Can lights must be IC rated or they have to be boxed.
original thread
It was an interesting conversation - you can find it here: http://forums.jlconline.com/forums/showthread.php?t=57657
reply
The can lights were airtight IC fixtures, so the part inside the ceiling is much larger than the hole. they were located in a rafter cathedral ceiling, so we could not get to them without removing the roof or the ceiling.
Log in or create an account to post a comment.
Sign up Log in