Underslab insulation comparison

hot_rod

Always wanted to compare the various methods. 3 panels:
2" Blueboard foam, 4" slab, tube suspended 2" below top surface
2" "topping slab" knob slab insulation, 2" pour, this would be poured over a typical 4" or so structural slab
2" knob insulation, tube at bottom of pour.

Tube 6" OC, same length and flow, 105F SWT, 65° ambient temperature.

On the green panel, if you want a true 4" over the top of the knob, you end up with tube at the bottom of a 5" slab
With then grey knob panel, the tube will be 6" below the surface.
Next I did a time laspe with the IR camera, over a 90 minute run.

Solid_Fuel_Man

What did you find out?

I just did a job, 5" slab with tube 3" deep, some 6" OC, and some 12" OC. Poured over a total of 7" of EPS (white styrofoam) per the customer's request. Never used anything but 2" of XPS (blueboard) under a slab before. 

For the 6"OC stuff, I did two 12" loops offset by 6". This is will be for a post and beam structure with 28' to the peak. The 4 tubes which are not sleeved are for slab temp sensors. 

I'm curious to see how it performs. Have to get a year's worth of data though. 

mattmia2

I especially like the tube as a conduit for the temp sensors

hot_rod

Next step is to remove foam from below, maybe slice some sections open. That is why you see the bolt in the center to lift the panels.
I'm curious how much of the pex is actually in contact with the concrete in the knob panels, and voids around the pour at the knob.
The depth of the tube and possible lack of concrete contact should be accounted for in the design.
The energy transfer is conduction, if 1/3 of the tube is not touching the concrete??

As expected start up time was considerably longer on the panel with the tube 5" below the surface. I predict somewhere around an 8-10° higher SWT to get equal same performance from the deeper tube position.

And of course the flywheel effect once the heat input is stopped was quite a bit longer.

I like the concept of the knob panels, be nice to have some actual data on performance.

Solid_Fuel_Man

mattmia2 said:

I especially like the tube as a conduit for the temp sensors

I've always done it that way. Easy way to use up the scraps. Old guy taught me that some years ago. I just tape up the ends, use an old electrical snake with the sensor taped to the end. Push it as far as it will go, and cut and curl the snake a foot or so out of the tube. Easy to change out in the future if need be. 

hot_rod

This article show the simulation that closely matches my testing.

All things being equal, that 30% reduction in output with tube at the bottom of the slab seems accurate.
Knowing that design days will certainly show the output limitation.
So we all know what happens next, the installer cranks the SWT 10° -15°. Which drives down the efficiency of the boiler of course, and may limit the use of heat pumps or other low temperature heat generators, in the future.

https://www.pmmag.com/articles/87540-depth-perception