In reading John Lstiburek's article "Slab Happy", he mentions that the under slab insulation only reaches 4 feet inwards from the perimeter of the slab. I thought the insulation would have to lay underneath the entire slab or you are losing heat down into the ground. I'm wondering if this has to do with the fact that he also recommends insulating the INSIDE of the foundation/stem walls. Is this how he gets away with not putting insulation under the entire slab? Does the ground underneath a heated home (no in-floor heating system) stay that warm? Right now our plans call for insulation on the OUTSIDE of the foundation/stem wall, and continuous insulation under the entire slab and around the inside vertical edge. It would likely be less expensive to do it Lstiburek's way, but I don't know which is a better idea. Thoughts? We are in Zone 4c (Vancouver Island, Canada).

Heat transfer through the ground is one of the hardest to grasp heat transfer problems for a house, so folks are often reluctant to hazard a guess. The usual assumption is that the majority of the heat lost through the slab must go to the outside air. Thus, heat lost more than four feet from the edge of the slab has to pass through four feet of dirt before it can begin heading upward to reach outside air temperatures. A fellow on this forum provided estimates of soil thermal conductivity from 0.5 to 10.0 hr-ft^2 degF/Btu per foot of thickness, with the lower values for wet soils and the higher values for dry sandy soils. Therefore, you might adjust the 4 ft. recommendation based on the soil type, with a higher value if you are on a wet soil.

The assumption behind the above argument is that you don't have an underground river under your foundation that is going to be carrying away the heat. Otherwise, it is assumed that the under-slab temperatures will tend to stabilize over a long period of time, with the main heat loss being to the outside air.

Thank you for explaining that, Lee. Our house is situated at the top of a small hill, so there is little concern that water is running anywhere close to under the slab. With that said, sounds like it might be better to be safe than sorry and just insulate under the whole thing. The ground is always going to be colder than inside the house, even if the ground is relatively warm. That is my thinking, anyway.

Yes, your average annual temperature is about 52 F (sorry for the units) so your deep ground temperature is similar. Without insulation you'll gradually warm the ground under the house, but you'll still lose some going down. As Lee says, it's very hard to calculate, even if you know the details of the soil type, but for the combination of comfort and low heat loss, insulating under the whole thing is a good idea. Also, in the summer, your outdoor dew point gets above 52 F, so you might have condensation potential on the slab if its uninsulated.

Second law of thermodynamics: Heat Flows to Cold. A cement basement slab will draw heat out of the house to the 50o earth. Constantly. 24/7/360. The idea that your oil burner will warm up the miles of earth under your house is dreaming; won't happen. Insulate the slab.

Thanks, Bob. That was pretty much my thinking too.

Posted By Bob I on 05 Oct 2015 10:54 PM
Second law of thermodynamics: Heat Flows to Cold. A cement basement slab will draw heat out of the house to the 50o earth. Constantly. 24/7/360. The idea that your oil burner will warm up the miles of earth under your house is dreaming; won't happen. Insulate the slab.

Bob, I have a different understanding of this heat transfer problem than what you have described. First, if you want to consider heating up miles of earth under your house, you would need to consider the geothermal gradient, which is the rate of temperature increase with depth into the earth. It is about 1 degF for each 70 ft. (https://en.wikipedia.org/wiki/Geothermal_gradient), so if the ground temperature were 50 degF at the surface, it would be 125 degF at one mile, so you would not be losing heat into "miles of earth," but rather gaining heat.

So why doesn't this underground heat reach the surface? Because the soil is a pretty good insulator. In my area, the soil is about R9 per foot of dry sand. So at 10 ft. below the gravel on the bottom of the conditioned crawl space, the heat is looking at R90 of "insulation." Detailed calculations will show that the gravel or slab does not stay at 50 degF, but both the slab and the earth under it rises in temperature, with the only significant heat losses going to the air at the ground surface.

Take my house as example. For heat traveling through the floor of the crawl space floor 4 ft. from the perimeter, it must travel that 4 ft. plus another 4 ft. to reach the surface. At R9 per ft., the heat must travel through R72 worth of insulating soil, plus R5 worth of foam to reach the air. That is the reason for the recommendation that the insulation is only needed on the outside 4 ft. around the perimeter inside the edge of the foundation.

It is a bad assumption that the earth under a home is an infinite heat sink with zero thermal resistance. A good designer does not waste precious resources duplicating what is already there in the earth, in this case, good insulation provided by the soil.

ok Lee, that may be true where you are, but I doubt that here in New England we have much area where there is any amount of R9 dry, insulating sand. It's mostly cold damp clay, silt and ledge; hardly good insulating material.

Even if the ground under the house warms up, it doesn't warm to room temperature. It is still a heat loss and still the coldest area in the house, hence the condensate zone.
Insulation is cheap.

We're sticking with insulation under the whole slab.

I know how to fix your problem, try to use thermal leak detector or heat loss camera they start from $35 but give you some information about your cold spots.