Cathedral ceiling flash & batt error? Some condensation.
Hi! I’m freaking out about my ceiling! I know that the “flash and batt, what have I done!” question comes up often, but I really need to hear from the experienced.
Our current unvented cathedral ceiling is set up as follows (from out to in)
Standing seam metal roofing
1.5″ wood furrings
2″ Rockwool Comfortboard
SIGA Majvest
0.75″ Plywood
3″ Closed Cell Spray Foam (nominal, could be closer to 2.5″ in places)
16″ Fibreglass Batt
We have yet to install a drywall or plywood ceiling finish. The flash does not fully cover the top chord of our trusses, as they are 2×6 material.
The house will be heated by in-floor radiant, with back up wood stove and high velocity air con system. There is a whole home HRV, which is currently the only power hungry item in operation. The wood stove is fed for 10 hours a day, and we have had a propane heater running for about 4 hours every two or three days, especially during the cold snaps.
Recently we noticed staining on the wall drywall, in small patches. On touch, the spray foam had condensation on it. We have since set up air movement with a small construction fan, and have not noticed further wetting. Things seem to be drying. From our (admittedly hopeful) data interpolation, we thing the wetting happened when the interior temperature was close to 85F, the humidity around 40% and the exterior temperature around 32F. This was followed by several days of extreme cold (0F) and 72F interior, 30% humidity, with the propane heat on. It’s hard to keep the house a steady temperature when relying on a wood stove that is meant for atmosphere and emergency back up only.
We’re in climate zone 6 (I believe, Canada you know). The winter average (Dec – Mar) from gathered temperature data is 20.1F. Doing dew point calculations on the stack up (with drywall finish), it looks like the ceiling is a point of failure for anything above an interior temp of 68F and 30% humidity. Or, it only works when the humans are uncomfortably cold and dry.
https://docs.google.com/spreadsheets/d/1NshHLIbpF8la3lQWATUIp9ytVDLEKoCay-inE6u8PWo/edit?usp=sharing
Since this is literally a one man crew working, and a large ceiling area to cover, removing the ceiling batts and adding 3″ more foam is the last possible solution we want to pursue. It would literally put us back a year in the timeline. I’m hoping that either adding an air barrier (Aerobarrier), or the addition of drywall/plywood finish to slow air movement into the batts, or the use of latex paint, removing batt insulation (please no) or having ceiling fans move air around is a potential “yes you’ll be okay” solution.
Looking forward to hopeful solutions. 🙂
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Hi Elizabeth,
You probably have some construction humidity that is causing the issues. You are monitoring humidity inside the structure, but at what level? Water vapor wants to rise in an assembly, 30 or 40% at the ground level might be 40 to 50% or more at the top of the cathedral. If you have a way to monitor humidity up there, you may find different readings. We typically run dehumidifiers after we begin heating a new structure (I'm in climate zone 7).
You also are off on the roof ratio for air impermeable insulation to fibrous insulation. Climate zone 6 requires a 51% ratio. To my calculation, you have about R-30 impermeable (the closed cell spray foam and Rockwool, which is above the CCSF) to R-50+ of fiberglass. Either reduce the fiberglass R-value to meet the ratio requirement, or you'll want to install a class II (responsive) vapor retarder that is very well air sealed. I would consider Siga's Majrex. If you are installing drywall in such a way that it can become an effective air barrier, a class II paintable vapor retarder should also help.
I'm curious to see what other readers have to suggest.
Sorry about your moisture issues,
Randy
Elizabeth,
I agree with Randy & Peter's advice. The roof will be fine if you add a class 11 vapour-retarder, and limit the amount of air that can get through the fiberglass.
What you are seeing is not unusual for a house under construction with no interior air-barrier. See comment #8 in this thread: https://www.greenbuildingadvisor.com/question/air-sealing-and-insulating-around-trusses
Thanks!
I want to add that you want to be VERY CAREFUL with your AIR SEALING DETAILS. I have an unvented cathedral ceiling in my house here, with ALL closed cell spray foam (8 ish inches, goal was R49 worth), and some skylights in it. I've had some condensation issues on the framing in that assembly, especially around the skylights, due to thermal bridging of the rafters. If you can keep the moisture out, you'll probably be OK, but you have to really do a diligent job with ALL the air sealing. Don't forget the ENDS of the rafter bays too. Be VERY careful EVERYWHERE.
Bill
Thanks Bill. We will be really conscious of the air sealing details. It's one of our guiding principles for this house.
Hi Randy,
We're monitoring the temperature and humidity at the reachable peak: i.e. the bottom of the roof truss.
We originally we thinking of installing Majrex. We hoped that the cost & performance ssociated with the Majrex would transfer to the spray foam, but we may have to discuss it again.
We are looking to be quite airtight as well. We have also discussed having Aerobarrier installed on top of drywall (which we know will have micro perforations from hanging pictures etc. ) Along with that, we're likely going to do drywall as airtight as possible.
Thanks !
First, your accidental "test conditions" would be challenging for any wall or roof system. As you seem aware, your air at 85F/40%RH has a dewpoint of about 57F. Your 72F/30% air has a dewpoint of about 37F. If interior air comes in contact with a reasonably impermeable surface that is below its dewpoint, moisture will start to condense. We try to avoid this by keeping the "first condensing surface" above the dewpoint of the interior air. We do this by changing the ratio of fluffy stuff inside the rafters and rigid (impermeable) stuff towards the outside. In your climate a ratio of 1/3 impermeable to 2/3 permeable is about right. You seem to be right on this ratio with your stackup. This means that you can use a type III vapor retarder on the inside surface of the ceiling and still be relatively safe.
I think I understand your spreadsheets, except for the dewpoints for each material. Are these the interior or exterior surface of that material? Obviously this will make a difference. Also, why is there a 1" air gap shown between the fiberglass and PU foam? Was this just to be conservative in assuming that the FG won't touch the foam everywhere? If so, that's probably OK for the analysis, but you do need to make sure the cavities are relatively airtight so that interior air cannot circulate between the two materials.
In your case, I suspect that the very high moisture levels with the 85F air caused condensation on the foam, especially when the temperature crashes like that. The general moisture issues are compounded by the use of a propane heater. Those produce about 3 lbs. of water for every pound of propane you burn, or something like that. And finally, since you have no interior finishes, there is nothing to stop water vapor from rising into the FG. Condensation like this is actually pretty common during construction. Application of drywall with a couple of coats of latex paint (Class III vapor retarder) is probably safe. If you want belt & suspenders, install a smart vapor retarder on the ceiling (Class II vapor retarder) before installing the drywall.
Hi Peter,
Yes, the test conditions were certainly extreme, and I mentioned them because I was hoping their extremity was the main issue. I think regardless of our final ceiling finish, a class II vapor retarder will be used. We may use the Majrex, that'll be a budget and installation consideration. But it's definitely on the radar.
As for my spreadsheets, I had an option for an air gap between the flash and batt in case we chose to decrease the batt amount, as then it would likely sag as it wasn't fully filling the cavity anymore.
And I cannot agree more about the propane heat. I do love how the reaction of burning propane has H2O as a by-product. It is both a testament to the chemical bond power between H and O as well as so opposite our assumed thoughts of combustion. We had previous moisture issues in the bunkie (which is essentially a much smaller version of the building principles being used) during construction because of using propane. However, being unconnected from the grid and relying solely on the solar we can generate does mean we need to supplement during construction with propane. Hopefully we'll be off the fossil fuel train as soon as the weather cooperates and the full solar system is commissioned.
After writing a Fine Homebuilding article about the flash-and-batt technique, for years afterwards I got comments from people experiencing condensation. In almost every case it was because the interior had not been covered yet. Flash and batt ratios only work if indoor air can't circulate freely through the batt or fill layer; the prescribed ratios assume that airflow is largely blocked with drywall or an interior membrane.
Hi Michael, I've definitely read (and re-read and double checked and gone back over) your article in both the initial consideration of the insulation system, the switch to flash and batt, and the subsequent condensation.
I am fairly reassured that the mass air movement through the batts is a large consideration. Reduction in mass movement through finishes and paint will obviously lower the amount able to condense. I think I was also just concerned as the calcs point to issues in many scenarios where even the drywall is installed (but as mentioned in many other posts and articles, dew point calcs are 2D, and do not account for perm ratings).
All good points. The ratios are based on Joe Lstiburek's work, and assume that indoor RH is within a normal range (under 50% at 70°) and according to my calculations that a small amount of condensation could occur, just not enough to harm the assembly or for long enough to result in microbial action. It's always safer to add a higher percentage of foam-to-fluffy insulation. An old rule of thumb that still works pretty well in CZ5-6 is that with at least 2/3 of the insulation on the exterior or in a foam layer, that you don't need to worry about covering the interior perfectly. My goal when using flash-and-batt is to limit the amount of foam, for various reasons, so I tend to stay close to the prescribed limit, and learned to quickly cover the interior.