Joe Lstiburek said our poorly insulated roof is OK
tundracycle
| Posted in General Questions on
Our house is currently under construction. This is a finished attic so a cathedral roof with closed cell applied directly to underside of roof sheathing (with Durovent in some areas).
We are seeing significant telegraphing of trusses with every snow event. House interior is only about 60°f. More info here: https://bamasotan.us/2020/02/roof-insulation-problems/
Where we were able to access the underside of the roof behind the knee walls much of the closed cell was only about 6” thick so 1” shy of code requirements for R-49. We don’t know if the entire roof structure is like this but assume it likely is. Per this discussion, while it’d be good to have, that 1” itself may not be something to worry much about: https://www.greenbuildingadvisor.com/question/roof-is-r-42-instead-of-r-49-big-issue
Thermal imaging indicates significant cold air coming in through multi-ply trusses when the house was depressurized to -50pa. Thus indicating that under normal conditions with higher indoor temps and stack effect (4 stories) we can expect higher interior pressure driving warm indoor air through these same openings/cracks.
In the Southeast corner of the east facing roof (where the thermal image is from) we added the 1” of closed cell which also encased the multi-ply trusses (our builders insulator does only closed, not open, so this was the only choice). Through subsequent snow events this seems to have either very significantly reduced or eliminated the telegraphing and thus indicating that it likely stopped the warm & moist interior air from leaking out.
My concerns with this include; warm moist air traversing under the sheathing could cause the sheathing to rot, energy inefficiency, problems from very frequent ice dams (vs one every few years on typical new homes).
While the final inch of foam to bring it up to code doesn’t add much for insulating value, it does appear that its encasing the bottom/interior side of multi-ply trusses largely solves the problem, at least for now. (There is still a possibility that movement of the trusses could cause enough cracks in the foam that the problem would return?)
Our builder wants to leave it as is and not do the rest of the roof as that would involve cutting out strips of sheetrock (as yet unpainted), spraying and sometimes having to work around HVAC ducting and then re-rocking and taping.
Our builder had the top code guy from the State of MN out to look at it. At my request he (our code guy) also ran it by Joe Lstiburek who, through one of the code orgs, is a consultant to state code folks.
First, I’m surprised MN’s top code guy seems OK with the roof insulation not meeting code (R-42 instead of R-49).
According to our code guy Lstiburek thought all was OK and included this:
I asked joe two specific questions.
Has he ever seen this before where there is air leakage in-between ply’s of a Girder truss? Joe Answered “Yes, I have seen air leakage between the layers of girder trusses…but not seen any issues…” Having said that…it is pretty easy to address….caulk or seal the joints…..”
I just don’t believe the plywood is going to have a negative much less a detrimental effect from any air leakage in the girder truss plies. Am I nuts or on the right track? Joe replied “You are not nuts…. I do not think it will be an issue”
To me, given the extent of the telegraphing and what we saw in the thermal imaging, this is a significant issue. I’m not a building scientist though.
Our top code guy thinks our roof is OK as is.
I’d love to be convinced that this is OK and I can move on without being further concerned by it. I’m having difficulty getting there. Perhaps especially since this project was originally to have been very energy efficient with exterior insulation on walls and roof.
Thoughts?
Thanks (photos to follow),
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
Recent photo after extra inch was added on some lower parts. You can see that there is little to no telegraphing on these lower parts while the upper part still has problems.
Wow, Joe can see what's happening all the way from MA....
Hi W Ramsay.
Failures are difficult to predict. Some times houses get wet and then dry, no problem. Other times they get wet, can't dry and you end up with mold and rot. There are a lot of variables. That said, you obviously are loosing heat and likely air through the roof framing. If thermal bridging is the culprit, then you would need to install continuous insulation above or below the roof to mitigate the problem. If it is air leaks, you need to do some air sealing. Is this a new build? If so, now would be the time to have your builder improve the roof assembly, or accept it as is.
Thanks Brian. Our builder does not want to do any more on it and says that it's acceptable as is. I'm trying to determine if this is a big enough problem to continue pushing or if that's not worth it.
We've already severely strained our relationship with out builder over this and some other issues so I don't want to make a mountain out of a mole hill but I also don't want to ignore this if it is serious and s/b corrected.
I was quite surprised by the response from Lstiburek regarding our roof. So I'm also trying to determine if he was provided all information and he indeed says that what to me appears extremely poor insulation is actually OK or if he was not provided accurate info or his actual response was not as it appeared.
Thanks,
In addition to the above, the hottest spots of melting on the roof, the deepest indentations in snow and then the first spots to appear snow free, do appear to line up with the gussets on the multi-ply trusses.
It's called conduction....
W, I would trust Lstiburek if he says that your situation should not be a problem. He has seen and analyzed more assemblies with similar issues than pretty much anyone. That said, "should not be a problem" is not the same as "done properly."
For R-values: your math assumes that the spray foam is R-7/in. I'm sure that's what the spray foam sales rep told you. The problem is that spray foam's r-value reduces over time, to as low as R-5.6/in, as air displaces the blowing agents in the foam cells. If they used HFO-blown foam, R-7/in may be fair, or may be a bit high. If conventional HFC-blown, which most installers are still using, within a few years you will actually have R-34 to R-36 in your roof.
I've worked with many code officials who don't fully understand insulation and related requirements, and I've talked with several who don't believe they should even enforce insulation requirements, only life safety issues. So it's not a big surprise that they are ok with your assembly.
"He has seen and analyzed more assemblies with similar issues than pretty much anyone."....is that a fact?
Tom, I meant to add a qualifier. It's an educated guess.
"W, I would trust Lstiburek if he says that your situation should not be a problem. "
Yes, definitely. Given all of the issues with our roof and as many articles of his that I've read I'm quite surprised that he would say all is OK, but I could be wrong.
I wouldn't say that Joe says your specific roof is ok.
I have a different take on Joe's response...
He said that the laminated trusses themselves aren't a problem, and to caulk or seal them if they are leaking.
He didn't seem to address the insulation level. But the code officials questions seemed to be centered around the laminated trusses. It doesn't sound like your situation was fully explained.
IMHO, I don't see how Joe's response gives the builder a pass to allow leakage (especially when Joe mentioned sealing the lamination's), or under code insulation.
That being said, I'd rather have slightly undercode insulation with low leakage and no thermal bridging, than code insulation with leakage and bridging. More spray foam (enveloping the external truss members), will fix the leakage and thermal bridging, caulking will just fix the leakage.
Sounds like the questions were asking in such a way as to get a certain answer.