Currently in the final planning stages of my new house ("raised ranch" style), and the first floor is going to be slab-on-grade. My current insulation plan has 2" of XPS under the slab across its entire area (with a vapor barrier and 4" of 2B stone beneath). However, when I went to put the compliance information into ResCheck, that application only had parameters for entering perimeter insulation (e.g. from the edge of the slab, vertically down to the footers). That leads me to believe entire-slab insulation isn't that common.

My question is, from a thermodynamic standpoint, is it pointless to insulate the field of the slab if the perimeter is insulated, or do you still get a lot of energy migration down through the stone into the dry earth below? Does most of the energy loss in a basement occur from conduction of the slab to the foundation walls to the exterior, or is there still a significant amount of vertical conduction into the soil?

  • When we built, what you might consider a super insulated house - 30cm in walls and roof, we put 20cm under the complete floor. Builder said that that was the biggest thickness he had ever been told to use. But only needing 2m^3 wood per year for all heating...
    – Solar Mike
    Commented May 21, 2020 at 4:53
  • What it does tell me is that the software is not capable - there are better ones.
    – Solar Mike
    Commented May 21, 2020 at 4:54
  • Depending on where you're located, the earth (below the frost line) is something like a constant 50 deg F year round (or something close to that). So without any kind of under-slab insulation, you're going to lose some heat through that slab in the winter. But conversely in the summer, that same heat loss is a plus because it reduces the cooling load. You should be able to do a rough hand calculation of the heat loss through the slab, given it's thickness and base, and with a 50 deg F ground temp and 70 deg F house temperature.
    – SteveSh
    Commented May 21, 2020 at 13:32
  • I only insulate when I am heating the slab , yes the slab is a heat sink but I have found Un-heated slabs are still cold even when insulated , probably because heat rises add a thermal break on top like carpet pad and carpet and I did not notice a difference when not heating. 16x26 , 16x32 , 16x36 slabs same heated area only the center one heated all 3 insulated. When the hydronic was turned off all 3 slabs were the same temp. .
    – Ed Beal
    Commented May 21, 2020 at 14:16

2 Answers 2


Depends on your climate. Ground temps more than a few feet downtend to be close to the average yearly temperatures. Here in Alberta (10,000 degree heating days/yr) that means that 8 feet down it's about 7C. This makes for a cold basement floor. Eventually the soil below the floor will warm up -- but that can take years. And the equilibrium will be somewhere between room temp and normal soil temp for that depth

Usually perimeter insulation schemes are used to make an less expensive floating slab foundation. In a seriously cold climate, however, a perimeter insulated foundation will still freeze if the building is left unheated.

A first order approximation is fairly easy. If you have a 1000 sq foot basement and a 20 degree F differential between inside and ground temps then with R1 of concrete it is taking 20,000 BTU/hour. This ignores:

  • Lots of people like to keep basement cooler. Maybe it's only 15 F differential.
  • The ground will warm up some, reducing this. The articles I've seen about passive annual heat storage mumble about 1 to 5 years to get the ground warmed up enough. And if you have ground water (sump pump...) then all bets are off. Water moves a lot of heat.

If you put in R10 worth of foam, this is reduced to 9% or about 1800 BTU/hour.

20 BTU/hr = 1/4 million BTU/day or about 2.5 Therms. or about a quarter of a gigajoule.

Here natural gas costs about 3-4 dollars per GJ, so to heat that slab is about $1000/year instead of $90/year with R10.

Is raising the insulation from R10 to R20 worth it? Probably not, unless you have a bunch of foam that fell off the truck. An extra R10 would save out about 90% of that $90 would would only be another 81 bucks.

Here a common practice is 4" of foam for the top 4 feet of basement, and 2 inches below that, and under the slab. Not sure how it's detailed where wall meets floor.


Concrete 4" thick has a r-value of roughly 1 (a lot depends on the weight of the concrete). 2" of XPS foam board has a r-value of 10. So roughly speaking, using the 2" of foam board under a 4" slab is going to reduce your heat transfer (loss) by ~ 90% (going from a u-value of 1 down to 0.1).

If you know the area of your floor, you can then estimate the amount of heat flow through the floor into the ground (winter scenario), in BTU's per hour.

Then you can do a similar calculation to estimate the heat loss through the perimeter, compare the two, and see how much the under-slab insulation buys you.

  • I understand the heat transfer through the slab, but what I don't know is how well the heat will transfer through the stone and soil to the deep thermal sink under the house.
    – ereisch
    Commented May 21, 2020 at 23:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.