TOPIC: I Have An Existing Metal Roof and Have A Question
Return to I Have An Existing Metal Roof and Have A Question
Return to Forum
Hi, All. I had a metal roof installed about 5 years ago, and at the same time, had skylights put in. This metal roof replaced a cedar shake roof.
The roof is a steep, vaulted ceiling with what I think is termed a "closed attic" or "unvented attic" - the ceiling (car-decking) is attached to the underside of the roof joist/rafters.
The insulation is dual foil-faced closed cell, air-impermeable, 4" poly-iso foam with no seams, or sealed at the seams with foil tape.
Last year, we started having water drips coming in around the upper side of the skylights. This would occur on the hottest days and when the sun had been shining directly on the roof for a few hours, and this past winter after very cold days, and only when the sun was hitting the roof.
The roofer told me this is normal with metal roofs, and not much can be done about it. The roofer claimed that the upper section of the inside of the house was way too hot and needed to be vented - they proposed installing an electric fan to vent the inside of the house to the outside, saying that would take care of the problem. The temps between the upper and lower sections of the house are only 2-3 degrees F, and humidity levels are 45-50%. Plus, I have ceiling fans running non-stop up there.
After looking at the IBC & IRC codes, and local state codes, it appears that my roof is not installed correctly. In the 2012 IRC, section R806.5 (Unvented Attic and Unvented Enclosed Rafter Assemblies), subsection 5.1, it states “Air-impermeable insulation only – insulation shall be applied in direct contact with the underside of the structural roof sheathing”.
This link explains this code pretty clearly: http://www.energy.wsu.edu/documents/AHT_Unvented%20attics%5B1%5D.pdf
I don’t believe this is the way my roof has been installed. I photo-documented the entire process at the time. The attached photo shows the roof after the cedar shakes were removed, and during the time that the roof decking/sheathing was being laid down in preparation for the metal panels. Note that the 1x4 sheathing strips used for the shakes are still attached, and the roof decking is being laid down over them. The poly-iso isulation is seen below the 1x4 sheathing strips. This has essentially created hundreds of unvented air-pockets between the insulation and the metal roof, because the insulation is not in direct contact with the roof sheathing.
The questions I have are:
My thought on #1 is that the fan won’t do anything but remove heat from my house.
My thought on # 2 is that the only way to solve it is to vent those spaces, which at this point is virtually impossible, OR eliminate those hundreds of air pockets that now exist between the insulation and the metal roof.
Am I correct in my thoughts on this?
Thank you. You have written a great narrative. Still, I wonder if it wouldn't be best if we could talk on the phone. Is there any chance you could call me at 1-800-543-8938 ext 201?
Hi, Todd. Thanks for the response. I called a left you a message a little bit ago.
Sounds to be like a bit of solar vapor drive that is getting into the interstitial spaces between what is the old skip sheathing and the decking for the roof. In that case, the slightest bit of moisture in the air will and can condense on the cold side of the assembly and condense as long as there is not an exchange of air or some underlayment to assist in the condensation.
Just as a quick sidebar, this would have likely happened regardless of the roof type in the right set of circumstances. Metal roofs get a bit more hammered on the condensation argument because they are impermeable and get colder which can exacerbate the situation, but this set of circumstances could have manifested in this issue regardless of roof type.
Thank you for your response & thoughts, Eric.
You mention that "the slightest bit of moisture in the air will and can condense"..."as long as there is not an exchange of air or some underlayment to assist".
Regarding the underlayment - could you explain where this underlayment should be in the layering of the entire roof assembly, and how that would work to reduce the condensation?
Your comments seem to indicate condensation cannot be controlled in metal roofs, no matter what.
I don't understand how "this would have likely happened regardless of the roof type in the right set of circumstances". Regarding the IRC code referenced earlier, it seems logical that if air-impermeable insulation is the ONLY insulation used, and having it in direct contact with the underside of the roof sheathing would leave no air space (or virtually none) between the roof layering for any substantial condensation to occur, and certainly not the amount I am experiencing dripping through my ceiling.
I realize condensation in metal roofs probably cannot be 100% eliminated, however I believe that even with metal roofs, proper construction techniques should reduce the condensation issue to a point of negligible concern. Am I totally off here?
So the description of underlayment in this case isn't quite correct. Normally when we see underlayment in an above vented style roof, it is when the roof is down on a set of battens that run both horizontally and vertically to allow for a vent space and a drainage plane if there is any condensation on the backside of the roof.
The issue here is the little dead space of air that is allowing for those surface temps to plummet and allow for a condensing surface. Ideally they would have been filled with foam or the roof strapped to allow for an above deck type of vent space. You are kind of in betweens'ville and neither is working.
James, I am sorry for not getting you called back yet. I have been thinking about this a lot. I am wondering a lot about why this only happened after a few years and I am curious exactly where the drips are coming through around the skylights. Is there any way to send me photos showing where the drips are coming in and also showing how the skylights are flashed? Were there any other significant changes made to the home during the four years when the drips were not occurring? My email is [email protected]
OK, I understand about battens and how they are used both horizontally and vertically in regards to ventilation. I agree the spaces could have been filled with foam to eliminate the dead space. It is interesting to see that vertical strapping laid down on the solid roof sheathing prior to the metal panels being laid down could have solved the problem as well. What I am wondering is why couldn't the old horizontal sheathing strips (or could be called horizontal battens - no?) have been removed and the solid roof sheathing laid down, directly in contact with the foam insulation? Would that have not solved the issue? The building codes seem to indicate that.
That would have most likely abated the issue if you went direct solid decking to the foam with zero air space.
Hi, Todd. Ya, that's a good question about why it took a few years to manifest itself.
The roof was put on in 2013. The first drips we noticed were from around the stove ceiling box in 2016 - nothing big - just several drips on very hot days, which I thought "no big deal". In early 2017, we saw some dripping originating around the upper part of the skylight, again several drips, but seemed to be more than what we saw on the stove. Late that summer, we had significant dripping from the skylight. Then early 2018 after a cold spell, more significant dripping - the first we'd experienced during cold weather.
It's almost as if the moisture is building up.
I will email you the photos I have showing the skylight install.
Could someone tell me what my options are to solve this issue?
I don't believe an inside>outside exhaust fan will do anything except suck the heat out of my house.
I either have to create a ventilation for those spaces, or fill those spaces with foam.
I think ventilation is not possible because the skip-sheathing blocks bottom>top airflow.
I don't think there's any way to fill those spaces with foam, unless the roof comes off.
The most obvious solution is to make it conform to IBC & IRC, which I think means ripping off the roof, removing the skip-sheathing and laying the roof decking right down on the insulation. It seems to me, that's the only solution.
Does anyone have any recommendations?
The roofer sent some of his workers out this past week to pull off the ridge cap and cut the underlayment that wraps over the peak of the roof, thinking this may vent the moisture. The roofer proposed doing this before he wants to install an inside>outside fan.
What we saw underneath the underlayment was a lot of wet, black, mouldy wood, lots of fungus growing, and some places the older skip sheating was already rotted (photos attached), and some of the newer solid roof deck plywood was starting to delaminate.
Also attached is a photo of the house & roof I took this past winter after it snowed. Note the area between the middle and right skylight, and stove pipe - the snow is melted there and the melting is trending to the right mostly - not upwards. The left skylight shows melting to the left. That indicates to me that inside air is coming up through the skylights, and then following the horizontal nature of the old skip sheathing to exit at the far left and right edges of the roof.
Any thoughts on this?
Bump. My last 2 posts weren't posted here yet - they were from last week. Overlooked, maybe?
Hi James. Sorry for the delay. I know that several of us have been traveling a lot lately and it has been tough to keep up with things. Yes, your skylights are probably causing the snow melt. So, you mentioned cutting back the underlayment at the ridge ... the ridge cap on the metal roof is not vented, though -- correct? I honestly think the roof needs to come off and I am going to still be a strong proponent of getting some vertically oriented ventilation in place. The issues here all seem to stem from a lack of ventilation. Avoiding condensation requires at least two of the following three things: vapor barrier behind the ceilings; significant insulation; ventilation. You can sometimes get by with just two of the three but you can never get by with just one of the three. I am sorry again for not keeping up with things better.
That is a mess up there.
This is definitely going to be a bit more of a project than a surgical fix.
Thanks for replying, Todd & Eric.
Understood about "Avoiding condensation requires at least two of the following three things...". In the photos of the skylight-install I emailed you (I hope you got them), you can see that there actually is a vapor barrier behind the ceiling & below the insulation - this appears in photo 0016.jpg (also attached). That photo is right before they installed the skylight framing, and the black vapor-barrier can be seen around the edge of the cut if zoomed in.
In another photo that I sent you, it can be seen that they stuffed air-permeable insulation (fiberglass) around the skylights (also attached). I feel this is the cause of the melting of the snow in those areas and no where else - essentially, there is no longer a vapor barrier and air-IMpermeable insulation in that area, and it therefore is allowing heat to easily migrate THROUGH the fiberglass insulation and out under the roof decking.
I wonder if they should have used expanding foam to maintain the air-impermeable insulation instead of fiberglass insulation to seal those areas around the skylight framing.
Foam around the skylight (window and door specific that is low pressure and closed cell) would be preferable there.
Thanks Jay, Yes, I received your photos. I was unsure whether that was a vapor barrier or if it was a roofing felt or something. Of course, to work, a vapor barrier has to be sealed tight -- seams all taped, penetrations all taped, etc. I would agree that air is getting up through the fiberglass insulation. Based upon what you discovered when you opened the ridge, I think moisture has been migrating to the peak continuously and, unable to escape, it has condensed, slowly rotting everything away. Eventually it got to the point of rot and moisture saturation spreading downslope on the roof to where it hit the uphill edge of the skylights and then trickled into the house.
Hi, all. Been a few years since I started this thread, but I wanted to share what has happened. After I presented the roofer with all the info I had amassed, they were pretty agreeable to remove the roof and correct it. Even though they didn't admit that what they did caused the issue, I think it was very clear from what we saw during the removal that what I was clamming was exactly the issue - I think they knew it too, otherwise they wouldn't have been so agreeable to do this, right?
Anyway, the decking around the skylights was saturated, and starting to rot. Fortunately, there was no damage to the structure. They removed the skip sheathing, laid down new decking, roofing paper and covered the entire decking in something called "fire & water" or something like that.
We put expandable foam around the skylights, and fire-rated expandable foam in the Stovepipe roof box. So far, no problems, and I don't expect any.
See the photos! Pretty bad!
I want to thank all of the advice I got from you gents - thank you!
Been over a year now
If you would like to reply to this thread, please
If you do not have an Ask the Experts forum user account,