Sealing/Insulating exposed ducts for evaporative cooler in Tucson AZ
I’m just getting done helping relatives move into a house outside Tucson, AZ. The house has an evaporative cooler on the roof ducted to each room. The ducts run out of the side of the evap cooler and across the roof splitting a few times and then heading down through the roof into each room. The ducts are rectangular metal ducts, sealed with mastic and then lightly painted brown.
Given that it’s 110° outside during the day and the sun beats down directly on the ducts, my thinking is that this is like AC ducts in a hot attic, but worse 🙂 A quick inspection shows that the ducts feel cool to the touch close to the cooler but warm to the touch at the far end. Basically the swamp cooler is working really hard to cool the ducts, instead of the house!
In a perfect world, I don’t see why any ducts would run outside and, worse, be in direct sunlight. The medium term plan might involve ductless mini-splits, but the evap coolers have great advantages in the desert, including helping raise humidity in the house (3% relative humidity here today) and allowing/requiring doors and windows to be open, which is a welcome change from a sealed up AC house.
But in the short term I think there are two tasks: 1. Air-sealing the ducts 2. Insulating the ducts.
They aren’t too badly air-sealed, but there are a few leaks here and there. From quick inspection the mastic looks to have dried out and cracked in a few places. Does anyone have advice for sealing ducting in brutal sunlight? Considering metal backed tape; presumably any seam that is getting sealed with tape would have to have the paint and mastic stripped off before putting the tape on?
The insulation plan is to wrap the ducting with 2 layers of 1 inch polyiso overlapped and woven at the corners, sealed with metal backed tape. Any thoughts on which tape, given the brutal sun?
A more radical approach would be to remove the ducting and drive the air only into one room, but there are concerns that that would mean that one room would be freezing (the evap coolers can really drive cool air in the desert overnight).
Thanks,
James
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
James,
I think you are on the right track. I would consult local HVAC companies. They would be the best ones to recommend tapes, insulation products, and insulation protection that will stand up to the weather conditions in Tucson.
The most common scenario is to introduce the air into a hallway near the bedrooms- just one place. But the ducts eliminate the need to keep the bedroom doors open. A finish layer of paint-lock sheet metal screwed together to cover the duct insulation would be pretty durable. Tape exposed to sun and water won't hold up very long.
Thanks to both Martin and Kevin. Interesting. I get the issues with the tape in the sun over the long term, and they do actually like the ducts going into each room. But I think covering the whole ducts again with metal is not going to happen, due to hassle and cost. But it does get me thinking, perhaps something could be arranged that just covered the corners and was pulled together every few feet by a cross-piece across each side attached to the edge pieces with screws. I'm thinking aluminum angle pieces on the corners, perhaps with some epdm rubber gaskets (like drywall gasket) underneath them, the flat joins could be covered with the cross pieces. That would hold things together pretty tightly and protect the tape from the weather, but without the need to wrap the whole thing in metal again.
Martin, talking with local HVAC people seems like a good idea, but it's pretty worrying just how many systems like this I saw just driving around Tucson and surrounds for a few days. Very standard way of doing things, it seems, although the long above roof runs on this building are longer than most, almost every evap. cooler I could see had some length of run above the roof and no sign of insulation. I expect that those contractors with real insight would do their installs and ducts runs in conditioned space.
Unlike compressor driven AC, with evaporative coolers there is ZERO chance of getting moisture condensing on the ducts through air-permeable insulation. The laws of physics limit the temperature out of the swamp cooler to something above the dew point of the air- when the temp is at the dew point it can't evaporate more moisture, since it's saturated. For moisture to condense on the outside of the duct would require higher moisture level in the attic/outdoor air, than in the conditioned space, but that would be both an unusual, and short-lived condition once the evaporative cooler is running.
Since the ducts can't create a condensation problem, it would be both cheaper & better to just bury the ducts in blown cellulose, while adding a bit more R to the attic. Adding even 3"/R10 over the top of whatever batts or blown is in there is usually cost effective on it's own merits, and makes a real difference in comfort (unless you're already over R40).
Cellulose is the fiber insulation of choice here, since it is far more air-retardent than fiberglass at open-blow densities, is opaque to infra-red (radiant) heat transfer (f.g. is translucent), and adds some thermal mass to the ceiling/roof. But if the local outlets don't carry it, use rock wool or if that's not available fiberglass will do- just go a few inches deeper.
Before insulating,air seal the attic floor (particularly where the ducts penetrate the ceilings) first- partiularly where the ducts penetrate the ceilings below. If you heap at least 4" over the tops of the ducts and you'll have as much or more R on the ducts as the 2" of polyiso.
Thanks Dana,
I don't think I quite conveyed the situation. The ducting runs above the roof, not in an attic, so there's no way to bury the ducts at all. I don't have a picture but you can see a similar thing here:
http://images0.estately.net/43_21311121_0_1366922596_636x435.jpg
Actually this one shows something more like it more clearly:
http://sphotos-a.xx.fbcdn.net/hphotos-prn1/s320x320/551090_501919419824768_1425987918_n.jpg
(I think I attached that one too).
I'm not an experienced swamp cooler installer in AZ (which is who you should ask), but I will take a stab at this:
During the summer, it's too hot to run a swamp cooler during the day. Thankfully, they abolished daylight savings time in AZ, so you turn it on as soon as the neighborhood starts cooling down after sunset. The heat picked up inside the ducts just due to still ambient air temperature outside the ducts is negligible.
The white paint seen in the photo is supposed to reflect 70% of the sun's heat when it's clean. That's good, but not good enough to run it when the sun is beating down. At night, the emissivity is 91%, radiating to a night sky at -200F, which also helps balance any heat picked up from the ambient air.
Your insulation band aid proposed above doesn't sound very easy on the eyes, but if it isn't visible from the street I guess aesthetics aren't an issue.
If the ducts are completely outside of conditioned space and exposed to the weather, whatever you insulate them with has to be tolerant of rain/UV/birds, etc. or a protective cover added.
Is there any possibility of a re-design, installing new ducts inside & scrapping (most) of the ducts up top?
I think that bringing the ducts inside is a medium term option, I'll bring it up with them. There's no soffit space, the roof construction is something like (outside-inside): white mop-on sealant, 4? inches foam, ?, tongue and grove sheathing, 2x10 exposed structural beams. There's a decent amount of head-room, it's a shed-roof structure, 8' at one side, probably 12' at the other. Ducts currently run down middle of the house (above the roof), could bring them directly below, which would run down a corridor.
But all that would definitely come down to aesthetics, perhaps exposed round spiral ducts below the beams, entering each room with a vent above the door (would require cutting vent hole in the internal walls, which are all exposed adobe bricks. The airflow requirements (cfm) for evaporative coolers are high too, so the ducts have to be large. Probably not going to happen.
The additional exposure issues that you point out (birds, UV etc), do suggest fully enclosing the insulated ducts in another weather-proof layer. Replacing the ducts entirely outside with something similar to a SIP construction duct is a possibility, I suppose. Does anyone make pre-insulated metal ducts?
Yes, there are manufacturers of insulated metal ducts with weatherized finishes, but I don't have a list I can refer you to. Try some web searches, or ask local HVAC contractors who do commercial buildings.
Interesting. Dana's comment lead me to do a little googling. I came across this presentation on ways that exterior ducts have been insulated and weather sealed.
That promotes a peel and stick product called Alumaguard. That looks pretty interesting; of course it's primarily industrial. AFAICS it goes on after and over the rigid foam. They include a white version of the wrap that might work well in this application.
Another option seems to be ArmaTuff, which is foam insulation with a laminated exterior shell that is applied using adhesive.
Anyone have experience with these kind of exterior ducting claddings?
Alumaguard is (hopefully) cheap, but it's also nearly useless on it's own. Like any radiant barrier product, while it reflects 97% of the solar radiation, it has no R-value, and when it's 110F outside , 130F+ in the micro-climate of a sun-baked flat roof, R-value still matters.
A mop-on/spray-on ACCA rated "cool roof" finish with a high solar reflectance index would be about as effective (and recommended). A 1-1.5" shot of closed cell foam on the outside of the ducts with a cool-roof finish/UV-barrier to protect the foam would work. The foam itself is weather resistant, but needs the coating to avoid degradation over time by the UV component of sunlight. For ducts 1.5-2lbs density closed cell foam is just fine, but 3lb foam with UV resistant top coatings is sometimes used for both the roofing & insulation over the whole roof. See:
https://www.greenbuildingadvisor.com/blogs/dept/guest-blogs/spraying-polyurethane-foam-over-existing-roof
Note, the global warming potential of the blowing agents is significant, and needs to be paid attention to when doing an entire roof at even code-min R values, but wouldn't be a big issue when sealing & insulating HVAC ducts at 1-2". Figure on about $1.50-2.00 per square foot for R8-R10 on the ducts, with high-SRI UV resistant paint.
ArmaTuff has a rated U-factor of about U-0.256, which is about R3.9, way better than nothing, and I think you'd probably want about twice that as a minimum.
Just a little followup. They completed wrapping the ducts in 3/4 inch polyiso, taped seals with aluminum duct tape (UL-181 stuff). They took measurements as they added more insulation over a few days (side by side, I think). See attached pic for impressive drop in temperatures throughout the house, 10 degrees cooler nearer the start of the duct run, about 15 degrees cooler near the end of the runs. I don't think there was a vent right at the start of the run, there was a few feet from the cooler to the closest vent (explaining why so much heat loss even in the start of the runs).
They aren't concerned about birds (not many in the desert?), but after reading this thread will probably use a UV protective paint or wrap.
If you are using foil-faced 3/4" polyiso about half the benefit is from the highly reflective surface, so anything you cover it with needs to be a high-SRI material or you'll be undercutting the performance. Aluminum won't degrade under UV, but it will corrode (and thus lose effectiveness as a radiant barrier) over time, so yes, putting something more weather resistant yet highly reflective would seem necessary.
Anything with a 3-year SRI in the 100 range or greater would be about as good as it gets for mop-on cool-roof coatings:
http://coolroofs.org/products/results/search&channel=products&orderby=cf_product_init_sri+desc&cf_product_slope=Low&cf_product_pri_prod_type=%22Field-Applied+Coatings%22