# How much foam to put under a basement floor?

| Posted in General Questions on

I’m trying to determine how much foam to put under my basement floor in a house design, and I’m having a surprisingly hard time coming up with a clear answer. This is climate zone 7, and the ground temperatures down deep are somewhere around 40-45F. Most homes here do 2″ of foam. My house design is a walkout basement, so one side of the house would be close to the ground and the other 3 sides would be under about 8 feet of soil. ICF walls.

It’s pretty simple to come up with a rough model for heat loss through walls, for example: area * U-value * delta-T. But when I apply this same equation to a basement floor, if I use R-10 for the floor (~2″ of foam) and a ground temperature of 45F, even though the 45F temperature isn’t too far from the heating set point compared to being -20F outside, the basement floor ends up being quite a large contributor to heat loss — significantly more than the ceiling even though the ceiling is exposed to much colder temperatures.

But soil is not a heat sink the way air is, and so the ground under the house would not be 45F continually — the heat from the house will warm the soil and reduce the heat loss. But how much?

I have been using BEOpt to model heat loss but there surprisingly isn’t an option for basement floor insulation. I found this thread on their message board: https://beopt.nrel.gov/node/124 which basically says that they don’t think basement floor insulation has a significant effect.

Also, if I’m reading the manual J guide correctly, I think they only factor in linear feet of perimeter around the slab to figure heat loss through the slab? And it seems like they figure it would be quite low.

So I’m wondering — has this been studied? Will going more than 2″ of foam under the slab help anything, either in reducing heat loss or increasing the temperature that one feels when walking on the slab? How do I figure how significant the heat loss really is at 2″, is it truly R-10 or is there some way of coming up with a figure for the heat loss equation? Thanks.

## Join the leading community of building science experts

### Replies

1. GBA Editor
| | #1

Nick,
Your research has been thorough. Briefly, modeling heat loss through a basement slab is extremely complicated. Moreover, assumptions about this type of heat loss are controversial. PHPP software often shows higher levels of heat loss through slabs than energy modeling programs written in North America, and the discrepancy is unresolved.

For a fairly thorough discussion of this issue, read this article (along with the 96 comments that follow): Can Foam Insulation Be Too Thick?

It should also be pointed out that the answer to this question is heavily dependent on your energy costs.

For a rough guide, I advise you to follow the advice of building scientist John Straube.

Straube wrote, “The cost of insulation becomes more than the cost of generating energy for the walls in a typical house in a 7,200-HDD climate at about R-60 (using the Building Science Corporation approach), and slabs [on grade] at about R-20 to R-25, depending the cost of placing EPS (which costs around 10 cents per R per square foot). Basements have less heat loss [than slabs on grade], so the cut-off point is more like R-15 to R-20 for a basement slab. Heating a slab with radiant tubes increases the temperature of the slab from around 68°F or so to 80°F or so on average, so the insulation levels need to be increased by about 50% over this for radiantly heated slabs.”

Other experts have used other calculation methods to come up with recommendations for much lower levels of insulation under basement slabs. For example, Gary Proskiw and Anik Parekh have calculated that even in Yellowknife in the Yukon Territory, a basement slab requires no more than R-10 vertical insulation at the slab perimeter. Full details can be found here: Optimization of Net Zero Energy Houses.

Finally, the main reason to install at least 2 inches of rigid foam under a basement slab in your climate zone has nothing to do with saving energy. The main purpose of the rigid foam is to make sure that the slab stays warm enough to prevent condensation. This will (a) make your basement smell better (because it won't be damp or moldy) and (b) make your slab warm enough to allow you to install carpeting if you want to, without worrying that mold will grow under the carpeting.

2. | | #2

"the heat from the house will warm the soil and reduce the heat loss." Not quite.
If you build a well insulated slab, you can raise the temperature of the slab by adding heat, but the soil under your house is thousands of feet deep, not a few inches, and the temperature never changes, so the heat you deliver to the sub-slab soil will be transferred to colder soil in an ongoing process. Your insulated frost walls at the walk out will help with slab temperatures. 4" of EPS (minimum) will give you about R16; 6" will give you R24.

3. Expert Member
| | #3

The soil under the slab has both R-value and thermal mass, both of which vary considerably by soil type and moisture content, making the modeling complex. With the exception of active volcanic areas or flowing artesian water streams the temp of the soil 50-100' down is close to the average outdoor air temp over the past 50-100 years in the area. The temp several hundred or thousands of feet down is much warmer, due to the high core temp of the planet, but that temp is completely irrelevant to this discussion.

The time constant of the distributed soil-R and soil-mass over a season or at most a year is the relevant time period from a slab insulation perspective, and there we are talking a couple tens of feet down at MOST. The slab losses DO raise the temp of the proximate layers of soil, reducing both the rate of loss and the total loss from the slab. The more insulation you put under the slab the less the soil is heated, so the soil stays colder, and the thermal mass of the soil then has diminishing effects on energy use. Putting R10 down instead of only R5 doesn't cut the heat loss by half- not even close, and the difference between R10 and R20 sub-slab foam is even smaller.

This can be modeled with much higher precision if more of the local soil characteristics are known, but the cost of fully characterizing a site's soil is more than the cost of adding a couple more inches of foam.

The amount of sub-slab R necessary to mitigate summertime mold and musty-basement issues is pretty modest- R5 -R6 (1.5" of EPS) would be fine even in zone 7. There's still a long term or lifecycle financial energy cost savings argument for R12-R15 or so even for basement slabs, and for on-grade slab edges even up to ~R20 in that climate, but not R25 (unless you expect future energy costs to be several times the levelized cost of solar electricity at today's PV prices.)

4. | | #4

Great answers above. I'll add that because the ground temperature is much warmer than the outside temperature, the heat loss to the ground varies strongly according to how warm you keep the space, more strongly than heat loss through your walls. So if you plan to let the basement stay cool in the winter, you might choose a little less; if you plan to fully condition it, a bit more.

Definitely use EPS, not XPS--you get more insulation value per \$, as well as much lower environmental impact.

5. | | #5

Thanks for the responses!

6. | | #6

When planning (and during) our house build, I devoured as much information about other house projects as I could. Anecdotal evidence shows most low energy or net-zero builds in the Northeast US are using at least R20 under slab. Many up to R30 or more.

4" of EPS seems like a safe choice (versus 2") if planning an energy efficient house in zone 7.

7. GBA Editor
| | #7

Brian,
You didn't tell us whether you are talking about a basement slab or a slab on grade.

Nick: basement slabs can get away with much thinner sub-slab insulation than a slab on grade.

8. | | #8

Yes, the slab I'm talking about is below a basement so under the frost line. Except for the edge by the walkout which is more like a slab on grade, but if I insulate the frost wall there does that make it perform more like a basement slab even at that edge? I'm wondering if I should just do the frost wall as ICF since the rest of the wall there will also be ICF?

9. | | #9

Correct, I was just offering some general insight into what I learned as a first time home builder/owner and I don't have a solid answer.

If Nick was paralyzed by indecision, bogged down in modeling/calculation, and needed to make a decision to get material on site.....I'm thinking hitting R15-20 would be a good call in his situation?

• |
• |
• |
• |