DIY Hydronic Floor Heating

[AC_Hacker]

Craig,

Thanks for that information. It's just what I was looking for.

I guess the last piece of the picture might be a few words about the insulation in your house, because that greatly affects what your heat loss rate is.

Did you do any kind of Manual-J analysis, or previous fuel-use estimate or some other method of estimating what your heat loss would be?

Seems to me that with the feed temps you have and your performance description, you hit the target pretty much on the bulls eye.

-AC

[Mobile Master Tech]

I have builder-stock R30 blown in insulation in the attic and R13 in the walls/floors. I did a manual J for the master BR since it has a high window to floor area ratio (pretty accurate) and did some research on exactly how much heat you get out of your loops (very innacurate). This is why I started with staple up and no plates-big mistake!

See details of my current energy usage in the Solar Storage Tank thread. I haven't yet figured out how to add your cool hyperlinks, AC!

[Mobile Master Tech]

I have used up to 350 therms of natural gas previously in Jan without heating the basement. Using 150 while heating the basement is a big relief! I sealed up lots of leaks, added the floors, added the efficient water heater. Attached is a pic of the biggest sealing improvement: a foamboard panel that fits exactly in the fireplace opening to prevent air leakage.

[Vlad]

Originally Posted by Mobile Master Tech

Here goes the first installment of the "RADIANT FLOOR HEATING MANIFESTO"

Extruded or Omega plates work better, but I found a great ebay vendor for u-channel plates, Gehman Iron in Knoxville PA, whom I then purchased from directly. They offered 2' and 4' plates at less than $1/ft-the best prices, reasonable shipping and great service. You could also get extra heavy duty aluminum foil on ebay that is 0.0015” or 35 microns thick to do something like what Vlad did(or even make an HRV!) This is 2-4x thicker than standard foils. 24” by 500’ is $112 shipped from pactogo_inc.

I ran a small bead of 100% silicone caulk in the groove before stapling up, which fills the gap around the bottom half of the pex. Silicone is an excellent thermal conductor, so avoiding the tiny air gap makes a huge difference and negates most of the benefits of omega or extruded plates.

I didn't make my own plates I bought them @

PEX Heat Transfer Plates (Aluminum), 2ft long

They cost 68.5c a foot + free shipping . You can't beat this even if you buy just material. They have very uniformed omega shape and touch pex tub all around. There is no gaps to fill with silicone. Also they hold pex tight against plywood and keep flat surface on top.

The only thing I didn't like was the end of the plate. It was just straight. So I took a piece of steel rod and opened each end a bit. I know this is just paranoia. Pex tube is tough and straight edge is nothing for it.

[Mobile Master Tech]

Great find, Vlad! I saw those but got the ones from Gehman because they were cheaper including shipping and I didn't know the pexuniverse ones were omega-they don't describe them as such. Gehman offers them in 4 ft lengths also, ideal for "access hole" retrofitting. I probably would have chosen these omegas for my 2ft length plates had I known. I used staggered spacing with my 2ft plates when layout and obstacles allowed. This breaks up the "strip of heat" effect:



I think using a smear of 100% silicone or other heat conductive compound is a good idea even with omega plates-it greatly improves the heat transfer, as only the peaks of each surface touch each other (surfaces look smooth until seen under a microscope) and the resulting airgap really cuts down on the thermal conductivity. Heat sinks for electronics always use a contact paste for this reason.

Radiusing the edge to prevent pipe chafing is a great idea. The plates from Gehman already came with a slightly expanded section at each end of the plate, but I touched it up using a 1/2" socket wrench extension similar to what you did because I wanted a little extra insurance.

I like webaware's grouted corrugated metal approach if one is doing new construction on joists. Perhaps commercial metal roofing available here would line up with common joist spacing?

I forgot to mention that I bought $300 worth of Ultra-fin plates, thinking that by getting good heat transfer from the pipe and then warming the air in the joist bay, I would have just as good heat transfer as with staple up plates and a more even floor temp. NOT!! Even spaced every 12", I couldn't get enough heat out of them. However, they are very useful for adding extra heat to the ends of the joist bays where you can't use plates due to curves or where the pex runs through holes in the joists. Note the different pex types. I tried 4 kinds before settling on the translucent yellow-white kind from blueridgecompany.com due to price & installation ease:



And here is a pic of the final product with radiant barrier ready for basement finishing:




The radiant barrier is stapled 1" down the sides of the joists and droops slightly, creating a slightly parabolic trough to reflect the heat through the subfloor without pointing too much at the joists. I made a little more droop just before the end of the joist bays so the barrier doesn't touch the pex, then stapled the end of the barrier tightly to the subfloor to keep warmed air from sneaking out the end towards the rim joists. The rim joists are sealed with touch n' foam, then insulated with R26 fiberglass (a doubled over layer of R13).

In hindsight, there is another idea I would have tried. Line the underside of the subfloor in aluminum foil such as the 35 micron thick stuff I described earlier, mist it with black spray paint, then install the pex & plates normally. I wonder if this would give slightly more even surface temps since the foil will touch the plates as well as the floor & allow a little better conduction, while the black lets the foil pick up heat within the joist bay as well?

Vlad, didn't you use aluminum foil or something between the plates?

Craig

[Vlad]

Originally Posted by Mobile Master Tech

Vlad, didn't you use aluminum foil or something between the plates?

Craig

No, I didn't. I think foil is too thin to transfer heat.

I used foil under tubes as a reflector. If you look at my pictures closer you will see that tubes are suspended and don't touch the subfloor.

I don't know how dust will affect "reflector" performance in long run.

[berniebenz]

The emissivity and conductivity of dust is quite low, and will kill the reflectivity of foil not in the vertical plane or facing downward. Waisted effort, IMO.

[Robert Bean]

Greetings all,

Just wanted to drop in to acknowledge a number of our graphics have been pulled from our website and posted here in the forum…we are ok with this and what we would like to suggest is if you want to use the materials just let us know… and more importantly if there are any comments regarding the images that need clarifying please let us know via our linked in forum.

One concern we have relates to a reply by a “Ko_deZ” in regards to images located at this URL:

ecorenovator.org/forum/renovations-new-construction/728-diy-hydronic-floor-heating-16.html

In which it was stated that the illustrations were not accurate. What was not posted from our website along with the illustrations was this text which states explicitly: “Please note: any of the illustrations you see here, represent a hypothetical case for a specific set of criteria. Another application might behave completely differently so it's best to model each case on its own criteria.”

The illustrations are in fact 100% accurate for the simulation for which they were modelled but they do not represent other models with different criteria.

It's also important to understand the differences between steady state and transient conditions when modelling just as its equally important to understand the model only represents the conditions tested.

With regards to, “Obviously, the thermal imaging is faked”…this is not a thermal image ergo it’s not a fake…it is a coloured contoured plot based on outputs from the partial differential equations used in the simulation.

On another note: If you wish to self-study on radiant design may we suggest you read the design process from these two pages:

ASHRAE Design Graph for Sensible Heating and Cooling with Floor and Ceiling Panels

Infloor Radiant Design Guide | Hydronic underfloor heating | About Floor Heating

We’d be happy to answer any other questions regarding these and any other images and applications associated with our online educational resources via our technical support forum.

Kindest regards,
Robert Bean, R.E.T., P.L.(Eng.)
Registered Engineering Technologist - Building construction (ASET)
Professional Licensee (Engineering) - HVAC (APEGGA)
Building Sciences / Industry Development
Journal of Indoor Environmental Quality
Indoor Environmental Quality :: Radiant Based HVAC :: Healthy Heating

[Ko_deZ]

You twisted yourself out of that quite well Robert, but I am not convinced. I think this is a simplified illustration, not a simulation(?). If it is a simulation, the the parameters you have used is nowhere near a real world example, am I right? Would not a real world example look better on your webpage?

[Robert Bean]

You can purchase the program from FlexPDE, the input string from us or write your own strings and then run the various simulations.



The objective of the exercise pulled from our website was to define isotherms at the mid length temperatures across various tube patterns in the interior of a building under steady state conditions using a 10C (18F) delta T along a single 76.5m (205ft) loop at 300mm (12"o.c.) spacing using a constant decay rate calculated as dt/L.

Very much a real world example.

For transient work with variable decay rate see these two papers:

Karlsson, H., Embedded Water-based Surface Heating Part 1: Hybrid 3D Numerical Model, Journal of Building Physics, 2010

Karlsson, H., Embedded water-based surface heating part 2: experimental validation, Journal of Building Physics, 2010