Metal Roofing Below Hydronic Heat in Joist Bay...

swihart_art

I've been researching installing pex in the joist bays to heat the floor above and the aluminum plates are pretty expensive that are on the market... from my research I've learned that the plates are there to transfer the heat from the pex to the subfloor and also prevent it from escaping downward... I would put at least r-30 in the joist bays beneath to keep the heat against the subfloor, and the house is on stilts so it is exposed underneath to the outside air in PA, why I want to heat the floor to begin with... this will supplement the oil furnace, the main hvac and keep feet nice and warm...

So I dug deeper thinking of how to source a cheaper way to make some plates when I realized that metal roofing is flat with channels every 9" of the 36" width that would nicely fit 1/2" pex tubing... these steel panels are cheap compared to aluminum and would install fast in long lengths (with a few people helping)... so I looked even further and learned that steel is a not as good of a thermal conductor compared to aluminum but I wonder if it would still pull enough heat out of the pex on it's way back to the pump ... if the steel was installed below the pex and then the insulation was installed, the heat would be trapped and have nowhere to go but up into the subfloor... the cost is about half to use steel vs the aluminum plates... and cut a steel panel in half and you have full coverage between joists with channels for pex...

The goal is to not actually heat the room with the floor but rather keep our feet from being cold on a subfloor with just insulation in the bays exposed to the weather, the steel may not even be necessary but I'm exploring the options, we will use the on demand water heater for house hot water and run a branch that controls the radiant heat in the floor (open system)...

thoughts or advice on this application?...

Zman

Dozens of people come to this site every year trying to save a few bucks using tankless heaters, open systems and other shortcuts. If you dont need radiant for primary heat, you can install plateless radiant. For the sake of your health; do not install a open system. A tankless heater will work, it is just not the right tool for the job.

swihart_art

The open vs closed is a hot topic and I'm not dead set on open, a small water heater is cheap if I decide to go closed... which is probably the best way anyway but there's time to sort that out.. lots of info out there on that and how to deal with it...

I'm working through the plate issue and how necessary they are in my application...

Ironman

Without plates, you'll get about 7 btus per square foot output from the floor; with good plates, about 20 btus per square foot.

Like many, I think you're missing a key factor about radiant heat and heat transfer: conduction is by far the most efficient means. That requires intimate contact between the conductors.

The slots in coragated roof panels are not round but square. That means that the tubing will not be in contact with most of the slot in the panel. That will translate to poor heat transfer.

Heat moves toward cool; the greater the temperature difference, the faster it moves. Again, conduction (intimate contact) is the objective. The only place that a radiant floor should radiate heat is from its surface to the surrounding mass of the structure. Through all of the other components (water to tubing to plates to subfloor to finished floor), heat needs to be transferred by conduction.

swihart_art

The stuff I get locally is this panel.. the ribs are actually a rounded shape rather than square and 3/4" deep so pex lays beautifully in there... I tested that with a scrap of each... so the steel would have 40 ish % contact area... or be in very close proximity to the steel and as a benefit continuous steel coverage should make less noise because the pex can expand and contract within the channel...



One con I see is that steel doesn't transfer heat as fast as aluminum but I would think over time the steel would have some thermal mass and absorb heat from that heated space above the r-30... and of course release to the cold floor above - as you describe...

I have a bunch of leftovers of various colors laying around and the cost of using steel would even be less since I have about 700 sq ft of panels that need used but don't match...

Solid_Fuel_Man

A piece of 22 or 24 gauge steel wont transfer much if any heat. You can do it, but likely it will be about 5% better than no plates at all. You have a furnace now, but want warm floors?

If you design (and pay) for good thick aluminum plates and a boiler, you can eliminate the furnace and save fuel forever. Pay now or pay forever....

Solid_Fuel_Man

Also, your water heater temps wont do much to keep the floor warm, especially with no plates, or metal roofing

hot_rod

Certainly aluminum would be the better metal for heat transfer. Copper or gold better yet
That being said, staple up tube systems have been and continue to be installed, Watts Onix for example.

Wirsbo promoted suspended tube for many years, and the UltraFin product clips on to suspended pex to enhance convective currents in the joist bay. Plenty of options these days.

Your metal plates would be somewhere between a pex or EPDM, bare tube staple up and an extruded aluminum plate I suspect.

The heat output from any radiant panel is directly related to the surface temperature of the floor, or wall. How you get the floor surface temperature there is more about distribution efficiency. The floor output would be the same from suspended tube to the best extruded plates IF the floor surface temperature was the same.
But the SWT to get the output will be lowest, and heat transfer optimized with the best conductor.

I have seen a number of homemade transfer plates, flashing aluminum formed in a wood jig is a common DIYer method.

Silicone in the grooves of the metal would help transfer. The two critical components of transfer plates is the tube to plate contact, and the plate to floor contact. Thicker metals tend to spread the heat better.

The more insulation you can get below the better in those"stilt" homes. I'd add a layer of rigid foam board below the batts and tape the seams to reduce infiltration, air movement could be the biggest energy thief.

Zman

You are contemplating an experiment with the pro panel roofing that I would not be comfortable with. I doubt it will transfer heat very well and there is an excellent chance that it will be noisy, very noisy.

You might consider omega plates as a budget option
https://www.ebay.com/i/183505597277?chn=ps&var=690843427140&norover=1&mkevt=1&mkrid=711-117182-37290-0&mkcid=2&itemid=690843427140_183505597277&targetid=797689825373&device=c&mktype=pla&googleloc=9029014&campaignid=8086082588&mkgroupid=82257584063&rlsatarget=pla-797689825373&abcId=1139466&merchantid=7999894&gclid=CjwKCAiAuqHwBRAQEiwAD-zr3VpSTxCSpNtoPPI5enlK4cvryItUGZZqRByDYuTZGz3pVJ91i4-W4BoCdDAQAvD_BwE

hot_rod

some test data of different transfer plate materials.
https://radiantdesignandsupply.com/theheatexchanger/2013/3/4/comparative-infra-red-study-of-heat-transfer-plates

hot_rod

The main issue with thin metal plates, regardless of the metal is “oil canning”. That pop sound an oil can makes when used.

One manufacturer went as far a suggesting you fasten just one side If the transfer plate to allow the tin to move without the squeak or pop.

So much for conductive transfer.

I’d be most concerned about the heat load, how many btu you need to deliver. That can help determine installation options.