Hi.  I have a basement in a townhouse in New York City that I want to heat with radiant hydronic heat.  My contractor was all set to pour the cement for the rat slab, when I stopped him to tell him he needed to put some insulation under it.  I met with him today, and he pointed out that the ground is very uneven due to the fact that there are large rocks in the ground.  He broke them up as much as possible, but there are a couple of spots where they will be sticking up into the rat slab.  Basically, he spread a bunch of gravel in the cellar to even the ground out as much as possible, and was then going to pour the rat slab to even everything out.  When we started talking about putting rigid foam board on top of the gravel and then pouring the concrete, he told me that would be a bad idea, as the ground is uneven where those rocks are (they come above the level of the spread gravel) and there would be fairly substantial voids under the insulation which could lead to cracking later.  This made perfect sense to me, so we decided to pour the rat slab then put some foam board on top of that, wire mesh, pex tubing and more concrete.  This sounded like a reasonable compromise, but I was just reading another post in this forum that talked about using 2lb spray foam insulation under the slab.  That probably would work better for me as I could then have both layers of concrete together rather than separated by a layer of insulation.  Basically, I'm wondering if anybody has any thoughts on which setup would be better.  My main concern with the "sandwich" approach that we're taking is that the concrete above the insulation will have to be pretty thick in order to keep from cracking.  Any ideas as to how thick that would need to be?  I'd love to make it thinner than 2 inches to get a little extra headroom in the basement, but the contractor advised against it for fear that it would crack.

Thank you,
James

Hi James,

minimum thickness for concrete overpours is typically 1.5", and 2" is better.

Spray foam is a pricey way to go, but it very well might work out better for you.

though, can't you just keep the rat slab really thin, just to level it out, and then do your insulation and a thicker top pour?

Actually, that's what I asked as well. He said I really shouldn't do less than 3" for the rat slab or I'd risk having it crack, but I didn't think that would be such a big deal as I figured the insulation would prevent the crack from traveling up into the overpour. I'll talk to him about that again, but can you see any structural and/or heating efficiency problems with that "sandwich" approach assuming I do use a 2" overpour? The contractor is very skeptical of having concrete sitting on something that is only rated as having 25psi compression strength.

25 psi is standard for slab insulation, but 60 psi is available as well.

I can't imagine why it would matter if the rat slab cracked, but I'm not a structural engineer either.

Ok, how about a foil type insulation. I just came across Ultra CBF which is a foil sandwiched in between two layers of bubble wrap to protect against corrosion. Apparently it can be used under the slab or between the bottom slab and the overpour as the foil is protected on both sides. They claim to have the same insulation properties as 2" of foam board, but I'm still a bit skeptical. Anybody have any thoughts on this?

If you can put a 3/4" air gap between the bubblepack and the earth below, and another air gap between the bubblepack and the slab you're good to go- you'll get at least R6, maybe R8-equivalent out of it.

But if it's in contact with both the earth and the slab, the ZERO percent of the heat transfer is radiant energy, ALL of it is conducted, and you're stuck with the R value of the plastic itself (which should be around R0.05 ) If the bubbles aren't crushed in the pour, you'll get about R1.5 out of it before they inevitably collapse.

Radiant barriers have some (but still limited- it's not always cost effective) value in under-insulated attics in cooling dominated climates, but there is NO evidence of ANY thermal performance under slabs.  To get any value out of 'em you need big delta-Ts, no conductive thermal path between the RB and the radiating or absorptive surfaces, and minimal convective thermal transfer.  Bigger air gaps are better (your bubbles don't do squat), and direct contact takes it out of the equation almost entirely.

When your slab construction starts to approximate the ASTM C 236 guarded hot box then you'll be gettin' somewhere!

http://www.ornl.gov/sci/ees/etsd/btric/hotbox.shtml

(ASTM C 236he test by which Ultra CBF scored a whopping snake-oil calculated R 3.8* ask for the test report & data to verify the test temps & configuration to find out how relevant they are to YOUR app.)

A more appropriate test for a full-contact application would be ASTM C 518, where it would most likely score a zero (or "as good as...")

just so dana is not alone, the products are total, absolute crap, and if they claim an R-10 for under slab applications in writing, please call the federal trade commission and file a complaint, as shysters like that only understand one thing, and that is a smack right in the pocket book. that sounds harsh, and I can only say I am not kidding, and it would not be the first time a concrete insulation product such as that was sued.

Yeah, I thought it sounded too good to be true. What really made me suspicious is that they don't actually say that it has an R-10 value they just say it has the same insulation properties as 2" foam, which I suppose is not quite as absolute as saying R-10. Sneaky. Anyway, I'm thinking of putting just 1" of foam board on the rat slab and under the overpour with radiant tubing. I figure we're far enough below ground that frost won't be an issue and it may be nice to have the room be a little cooler in the summer. Am I making a mistake? What about putting a foil sheet on top of the 1" insulation? Would that help at all?

The ground temps 20" (50cm) down in NYC doesn't go below freezing even in the parks:

http://clic.cses.vt.edu/icomanth/22-AS_Temperature.pdf

Underneath a heated building adjacent to another heated building, a floor 40+ inches below is probably pretty reasonable- 50F+, maybe even approaching 60F year round.  (Drill a hole a foot or so below the slab and measure it.)  1"/ R5 XPS should be pretty reasonable.

Foil sheets have NO insulating value when in contact with solid materials on both sides.  XPS is are not very transparent to radiated heat (which is the same as light, at a much lower frequency/redder-color.  Not a perfect experiment since the absorption will vary by color, but a cut a thin slice, hold it up close to the light- you may get some light through, but the intensity is much reduced.  Since your eye sensitivity is logarithmic, not linear, if it looks even half as intense as staring directly into the light, the foam has absorbed or reflected over 90% of the energy in just that thin slice.  Then, take a 1" thick sample and hold it up to the light...  The remainder is all you'd be able to block with a piece of foil IF YOU HAD AN AIR GAP.  If it's in contact with the XPS and the concrete, the temperature of the foil is the same as the surface of the XPS and concrete, and 100% of the energy transfer is conducted, none radiated.

I'd like to see the purveyors of under-slab bubble-pack radiant barriers submit a sample of concrete slab with their product applied for an ASTM C 518 test, which would give the TRUE total R value of the slab + product in the prescribed-use configuration, with a 45-50F cold side and a a 75-80F warm side. Those numbers will give a far more accurate measure of the heat flux/loss to expect using it under a radiant slab.  (My expectation is that it would be almost as good as putting 15# felt under the slab. ;-) )

Thank you for the detailed replies, Dana. I really appreciate you taking the time to give me all of the particulars. I feel good about my set up now, and I'll definitely stay away from the bubble wrapped tin foil stuff. Maybe I'll try that felt suggestion though. ;)

This is a little late maybe, but don't get wrapped around the axle about concrete cracking. Why? Because concrete will crack, period. It's very difficult, if not nigh impossible, to keep it from cracking. That's why you see crack control cuts in slabs, sidewalks, driveways, etc. In house slabs about the only way to keep cracking from being a problem, even if it does happen, is to use a technique called post tensioning, cables run through the slab and tightened like a banjo string after the concrete hardens.

Rather than trying to keep your rat slab from cracking your concrete man should be looking at ways to control how and where the cracks will form.