Metal roof–condensation or leakage?
This is not so much a green building question as a building science question. I know a lot of you are quite fluent in this area and would appreciate your comments.
I was asked to look at a building with a “roof leak”. It is a slab-on-grade metal building, a bolt-together red iron frame with metal siding and roofing, and “metal building insulation” behind the siding and under the roofing (wide, rolled fiberglass between two sheets of heavy, white plastic sheeting). It is a very large building, roughly 50′ x 250′ and has a very low-slope roof, at least for metal, about 1:12. The metal roofing is a high rib profile and has no pipe penetrations or obvious openings. It does have some dried-out sealant along the ridge cap, but it didn’t strike me as a big deal.
The first bay of the building I looked at had water dripping on the floor, but that wasn’t the only area of the insulation holding water. The obvious drip was near the ridge and also near the exterior wall. Another area of water (not yet dripping) is along the opposite wall a little further from the ridge. OK… so maybe some water blows in at the ridge cap and runs down along the top of the plastic until it finds an opening, and then ponds up on the lower plastic sheet, soaking the insulation.
Later, the user of another bay mentioned that he also had water on the floor. This one is near the bottom of the roof slope. There are no penetrations anywhere near this. The gutters are clean and working. There are two other “ponds”, one also near the gutter, and a second one up higher, next to an exterior wall.
The entire building is unheated. Yesterday the dew point was about 32F inside, and the air temp was close to 40F. No condensation on anything. No sources of moisture in the building at all, aside from the slab itself and an area of concrete wall that is below-grade outside the building. In wintertime we get overnight temps in the 30s and the sky is sometimes clear.
Thanks for reading this far. Any thoughts?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
David, I have a tool shop with a tin roof and beaten earth floor. In the fall and spring, it sometimes "rains" inside my shop when there is snow on the roof and the air under it is moisture laden.
Was it any colder outside than inside?
How well sealed is the "metal building insulation" against air infiltration?
Is it possible that relatively warm and humid air is passing through the insulation to condense against the cold steel skin before running down the sides to pool on the floor?
I suppose it must have been for there to be any contrast in your thermal images.
Steel skins make efficient condensers.
How well ventilated is this building?
David,
What climate? Must be relatively mild if you are around 40 in the middle of winter.
In an unheated slab on grade structure the concrete slab itself can be cold enough to act as a surface of condensation. In my climate 6/7 this is a common summer morning phenomena when humid air hits the cooled down garage floor surface and forms puddles.
I'm not smart enough to know for sure but a 32 degree dew point at 40 degree air temp sounds like relatively humid air. You did not mention any precipitation events that occurred around the times the leaks were discovered. Even though the interior of the building doesn't have any activities that produce vapor (aside from the slab maybe) this type of unconditioned utility building likely has an interior RH similar to outside conditions. If the air is humid there is enough moisture to condensate in the interior and especially on metal. Does anyone notice condensation on the metal structural members?
Do the soils around the building have a high clay content? High water table?
Is this a new building? or an older building and the problem just recently occurred?
BSD-119: Summer Condensation Problems in Ice Arenas
http://www.buildingscience.com/documents/digests/bsd-119-summer-condensation-problems-in-ice-arenas
Some answers: the building is 12 years old. It is partly built into a bank, and the land rises for quite a ways beyond that, so it's safe to say the soil against the concrete wall gets wet, and I assume it's wet under the slab too, and probably without a vapor barrier. This is climate zone 4A. There really isn't any ventilation. The building has four airplane hangars and each has a large door that's probably about as airtight as a garage door... not great, but the wind is not whistling through there. The sun had come out on the day I took the IR images, so there was plenty of thermal contrast, with the wet spots staying cold the longest. There aren't any signs of condensation on the structural members or anywhere else. The only clues are the ponds of water accumulating in the insulation.
One thing I don't know for sure... does this type of metal building insulation usually have plastic on both sides of the fiberglass, or just on the bottom? As far as I can tell, if this water IS from condensation on the underside of the metal, it would have to run down along a top sheet of plastic until it found an opening, and then collect on the lower sheet. If there's no top sheet I don't see how it would all be collecting in a few places (I assume that condensation would be somewhat uniform on the underside of the roof, at least around the perimeter of the building and/or the seams in the plastic).
Thanks for the input.
David, John Straube details the fix very well. But it sounds like you are not familiar with a commercial build such as you are looking at. BSC does do consulting, depending on the budget.
I think if you print out the BSC doc and get up into the ceiling of your building you will start to get a feel for your next steps. Myself, budget first. If they had the budget, fly BSC out to spec it. IF not, cut the holes they suggest, (now that is low budget) at low points water can drain out, air can get in. Time could straighten all out. The fan idea of forcing air into the fiberglass sounds smart too. More money involved, but would have faster results and sounds like more of a sure thing to me. I truly am just thinking this all out with you. Having Straube involved to me is by far the way to go.
Oh and the insulation you are asking about is available where I live at commercial roofing supply outlets that deal in it all, such as EPDM, commercial products etc or steel and pole building companies.
http://www.steelbuilding.com/buildings/accessories_fiberglass.htm
http://www.steelbuilding.com/buildings/accessories_ventilation.htm
David my bet is on a condensation problem given the lack of ventilation you describe, ground moisture, and torn plastic sheeting.
Check out some of the photos in this article I found:
http://www.designandbuildwithmetal.com/Columnists/Writers/danny_wirth_7_13_09.aspx
Possible remedies for condensation issues:
- Increase ventilation
- Replace any sagging insulation with a tight-fitting insulation (Roxul makes rock-wool insulation products for this application)
- Look for and seal any breaks in the plastic sheathing (air/vapour barrier)
- Eliminate thermal bridges
- Seal the foundation
This probably explains why so many commercial steel buildings are insulated with CC spray foam.
BSC has a great PDF that I can't find to do with metal buildings and water filling up the roll insulation.
The attachment is NOT the one I am looking for but is very interesting to read none the less.
http://www.buildingscience.com/documents/insights/bsi-044-pressure-is-on/
See if you can find the right article, it is out there somewhere.
Lucas, if you read the attachments above, the cure is the opposite of what you suggest. What makes you think you are giving good advice?
This stuff stumps all but the BSC crew in the 30 years that I have been in construction.
I defer ultimately to BSC. David should really contact BSC IMO.
AJ, what cure would you suggest? Air-sealing the insulation would probably be #1, although very difficult due to the who-gives-a-damn way they build these buildings. Honestly, the best approach would probably be to remove the metal, install thick polyiso nailbase over the frame, and then reinstall the metal. The FG is a woefully bad attempt to insulate and prevent condensation.
I have not read the BSC articles but I will. Ultimately, there is no need to call out the heavy artillery. The only real issue is water on the floor and I think we can deal in a couple of ways. The owners won't spend millions to fix this, it is a minor nuisance and not an issue of major damage.
If this were a house, it would be a completely different picture.
David, not that it matters but where is this and why is there insulation if they don't heat?
As to a fix, cheap fix per BSC, cut round holes in the plastic right where the water is trapped, low point. That will drain the water, and may stop the drip because air will then get in to keep what condensates drying out pryer to it buidling up to dripping. If like BSC says that doesn't work the hole can be taped back up.
That is the cheap easy fix.
If you read the BSC posts, you will see that instead of a better air barrier they sometimes say the oppposite. They suggest a pin holed plastic actually in a section of the posting.
Just trying tohave a good time with a little puzzle offered up by Mr. Meiland AJ.
I suppose you consider "I defer ultimately to BSC. David should really contact BSC IMO." to be good advice?
Don't be offended if I'm not crushed by your critisims AJ.
Seems like the net effect of that is to render the insulation almost worthless.
The building is not heated and probably this one never will be. In other scenarios people use these buildings for lots of things, and even build finished spaces inside them, so heating might be applied and dripping might be a real problem. As far as I can tell, the builder installed the insulation mostly to try to keep the roof from dripping on people's heads.
David, if the building is to remain unheated, possibly the easiest first step would be to increase the rate of ventilation. In that article I linked before, it says that livestock enclosures should be ventilated at 6-8ACH(Nat?). By your description, it sound like this building has very little to none.
To start, why not get the owner to try ventilating with some portable fan units for a while? See if that makes any difference and if so, move to a more permanent ventilaton strategy.
I was thinking of putting some dataloggers in one of the units and seeing how often they get dewpoint conditions. Slit the plastic and tunnel thru the insulation, tape one to the roof deck, tape up the plastic. Tape one to a column up near the ceiling, tape another to the same column a few feet off the floor.
Any ventilation strategy would require cutting holes in the building and possibly installing electrical wiring. I'm not sure I want to recommend that yet, although I do think that ventilating would help if there were enough of it. It would certainly add to the electric bill.
David,
I am working on a sketchup solution to problem
and I will not even charge you
Peace out
That's what I'm here for Bill... free stuff!