Sandwich Radiant DIY Pour Over or Metal

[mtnxtreme]

Just built a new house and we are all setup to put 1/2"radiant tubing 8"OC on top of the subfloor, in between 1 " high nailers, with 1/2" plywood overlay, and plan to put anything from 3/4" ( thats all the height we have left) to thin prefinished wood, to tile, and or vinyl as money dictates and/or each room needs. I am still undecided wether to pour 1" of sand and cement between the nailers, worried abt. weight and high cost of the mat./install. I have 2x12 16"OC with the longest span being 15' and 2" x 12" TJI's spanning 22', so I think it should be no problem as the mix is supposed to add approx. 10lbs. per sq. ft. My other option if making up aluminum pans ( I have a large 8" commercial metal brake and can make nice profiles with pretty tight dimensions) to hold 2 rows of tube and go up on the nailers for heat transfer, still up in the air on the exact profile, would probably be a little less expensive, faster, and less messy. I've heard the slabs hold the heat, but taking longer to heat up, and the aluminum won't hold it, but heat up much fatster. Anyone have any ideas as far as pros/cons of either, or experience with any of theses install?

 

[DaveBP]

A quick read and a commercial source for comparison to your own fabrication.

http://www.radiantengineering.com/PlatelessInRadiantville.pdf

http://www.blueridgecompany.com/radiant/hydronic/316/rht-heat-transfer-plates

Aluminum plates and an outdoor reset control that varies the temperature of the water in the tubes rather than just shutting the water on and off.

Why 1" nailers (often referred to as sleepers in this use)? 3/4" is more than ample. If you're using ripped plywood 3/4" will save your a lot of money.

 

[SE Iowa]

Just built a new house and we are all setup to put 1/2"radiant tubing 8"OC on top of the subfloor, in between 1 " high nailers, with 1/2" plywood overlay, and plan to put anything from 3/4" ( thats all the height we have left) to thin prefinished wood, to tile, and or vinyl as money dictates and/or each room needs. I am still undecided wether to pour 1" of sand and cement between the nailers, worried abt. weight and high cost of the mat./install. I have 2x12 16"OC with the longest span being 15' and 2" x 12" TJI's spanning 22', so I think it should be no problem as the mix is supposed to add approx. 10lbs. per sq. ft. My other option if making up aluminum pans ( I have a large 8" commercial metal brake and can make nice profiles with pretty tight dimensions) to hold 2 rows of tube and go up on the nailers for heat transfer, still up in the air on the exact profile, would probably be a little less expensive, faster, and less messy. I've heard the slabs hold the heat, but taking longer to heat up, and the aluminum won't hold it, but heat up much fatster. Anyone have any ideas as far as pros/cons of either, or experience with any of theses install?

I would not do the cement. I have 1.5" of gypcrete on top of my floors with pex on 12" centers. It weighs ~15lbs/ft. I had to have special joists made to handle the load. My living room has a span of 18' and so I had to have 2x4" engineered web floorjoist that are 12" on center! YES! there is 8.5" in between each joist. They said I had to have that beefy design so that the floor did not vibrate and bow with all the weight, andBTW, it still vibrates and bows slightly.

 

[SE Iowa]

Also, forgot to mention that MENARDS sells radiant floor panels with pre-formed channels for pex tubing. They have a whole DIY design center to help you build manifolds and add pumps, controls, instantaneous water heaters, etc. If I was to do my project again I would never use gypcrete and would go with these panels or probably I would use Warmboard despite the cost. Just my 2 cents.

 

[mtnxtreme]

A quick read and a commercial source for comparison to your own fabrication.

http://www.radiantengineering.com/PlatelessInRadiantville.pdf

http://www.blueridgecompany.com/radiant/hydronic/316/rht-heat-transfer-plates

Aluminum plates and an outdoor reset control that varies the temperature of the water in the tubes rather than just shutting the water on and off.

Why 1" nailers (often referred to as sleepers in this use)? 3/4" is more than ample. If you're using ripped plywood 3/4" will save your a lot of money.

I got a good deal on the full 1" battens, and I figured 1" would hold that much more heat than the 3/4.

 

[mtnxtreme]

Those RHT heat transfer panels look like the ticket, I can use the double tube ones and mount right over 3/4 boards and they will sit tight to the plywood above, I wonder how much efficiency I will lose over the thermal mass or will it be a wash since these react quicker.

 

[DaveBP]

I wonder how much efficiency I will lose over the thermal mass or will it be a wash since these react quicker.

I don't know that the efficiency (as far as fuel use goes) is improved by high thermal mass in a floor. In older systems, before microprocessor controls were available, the high mass of the gypcrete floors helped to even out the on-off (sometimes called bang-bang) cycling of the boiler supply water which was always way hotter than required. But now with modulating boilers and electronically controlled water temps through the tubing that is determined by the actual heat loads for the current conditions the need to soak up the extra heat is just not there. That is, if you are willing to put the money into the controller and mixing valve that are the heart of this kind of setup.

Basically during the middle of winter the circulator runs almost all the time and the temperature of the water going through the tubing varies to match the calculated load determined by a sensor that keeps track of the outdoor temps. Either one works. You'd probably need to get actual quotes from contractors in your area to know what the cost comparison is.

Most contractors will use plywood or OSB sleepers instead of solid wood boards. They are much more stable under the temperature and humidity extremes that a radiant floor subjects its materials to.

 

[Floydian]

Have you had a heat loss calculation done? If not, that is step one. I am planning on a dry sandwich install soon myself and I am considering having a pro design my system then I'll take it from there. A room by room heat loss is important as rooms can vary quite a bit and sometimes there is not enough available clear floor area to deliver x btus/ft2(Bathrooms and kitchens). If you want a high performance system this part has to be right. In a lot of cases it seems lightweight stamped plates can't deliver the necessary btus/ft2 so you want to make sure before you spend the $$$. Also with lightweights pex-al pex seems to be the choice to minimize the potential for noise.

A couple of good sites:

http://www.heatinghelp.com/Forum

http://greenbuildingtalk.com/Forums/tabid/53/view/topics/forumid/12/Default.aspx

Good luck with the install.
Noah

 

[DaveBP]

Here's a webpage that compares Taco and Grundfos pump curves for a bunch of different models, including the multi-speed Grundfos circulators.


http://www.builditsolar.com/References/Pumps/PumpCurves.htm