GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon
Q&A Spotlight

Details for Insulating a Slab Perimeter

Keeping all of the insulation to the inside of the foundation wall offers at least two advantages

In this insulation detail at the perimeter of the slab a total of 6 inches of rigid foam insulation separates the interior slab from the outdoors. A plywood bottom plate helps to stabilize the double-stud wall. Illustration by Chris Roche.

In New Hampshire, Chris Roche is pulling together plans for a new energy-efficient home. The slab-on-grade design calls for double-stud walls insulated with dense-packed cellulose, what Roche believes is the most economical and practical approach for getting the walls to R-40, or beyond.

Roche has paused at an important detail: the layer of rigid foam insulation that separates the concrete foundation wall on the outside from the slab on the inside.

Roche writes in a recent Q&A post that just about every foundation design for a slab that he’s seen either puts all of the rigid insulation on the outside of the foundation, or splits it half-and-half between inside and outside. In either case, Roche sees a problem.

The aim is to insulate the foundation wall with a total of 4 inches of expanded polystyrene (EPS) insulation. If the foam goes on the outside of the wall, it creates a difficult flashing detail, plus he’ll have to protect the outer surface of the foam from damage. If he puts 2 inches of foam on the outside and 2 inches on the inside, the concrete may stay to cold, and transmit that to the interior slab.

Roche saw a solution to this dilemma in a recent GBA article on double-stud walls by Dan Kolbert in which the entire perimeter of the slab was insulated with EPS to the interior.

“This would effectively create a cold foundation wall but warm interior slab,” Roche says. A potential weakness with this approach is that the interior stud wall would be partially resting on foam, so Kolbert had suggested placing plywood under the bottom plates to provide some additional support.

Roche has followed that up with a design detail of his own, shown in the illustration at the top of this article. A video from a Canadian builder seems to suggest the same approach, and Roche now asks GBA readers for their views.

“The whole point of this post is that I am just wondering if anyone else has done a wall/foundation assembly such as this and also if anyone foresees any problems with building a foundation in such a [manner]” Roche says.

That’s where we start this Q&A Spotlight.

Your plan is mostly sound

With a few caveats, Malcolm Taylor likes Roche’s approach.

“No reason your detail won’t work, as long as the outer wall is the load-bearing one,” Taylor says. “If the walls are deep enough to cover the insulation I agree it’s [a] much simpler solution than exterior insulation with all the attendant problems of protecting and detailing it.”

However, Taylor questions whether Roche really needs a full 12 inches of rigid insulation underneath the slab, as his drawing shows, and suggests that he leave the foundation wall a full 8 inches thick, rather than reducing the upper part of the wall to a 6-inch thickness to allow for an additional 2 inches of insulation.

“You don’t need to thicken the slab edge or pin it to the foundation,” Taylor adds. “It isn’t load-bearing, and pinning it means any differential movement between it and the foundation walls will cause cracks.”

Finally, Taylor suggests that if Roche plans to built and stand the walls together, the plywood bottom layer should extend under both 2×4 bottom plates.

“I agree the 12 [inches] does seem a little excessive,” Roche replies. “I could possibly shrink that to 8 inches, however I was thinking the extra foam doesn’t hurt and will make the backfill a little easier. My aim for this house is to be .5 ACH or less, R-40+ walls and R-75 attic. I have done some heat loss estimates, but haven’t played around much with foam thickness.”

Will the inner wall sink?

Jon R raises another concern: the slab, which is the finished floor, may sink, taking the inner wall with it.

“If the EPS foam is of the correct density,” Roche says, “I would think that the backfill compaction (potentially settling) and concrete shifts during curing would be the biggest concern for movement of the slab.”

Taylor doubts there would be any appreciable settling, assuming the right kind of fill has been compacted properly. He adds, “For the very small amount of movement that occurs in both settlement and seasonally, decoupling the slab from the foundation wall helps allow this to happen without real consequence.”

The design is on the right track

A nameless GBA reader (User-6935182) assures Roche that his planned insulation detail will work. The reader built and lived in a similar double-stud wall house for many years without incident, and enjoyed very low heating bills.

“Mine was also in [Climate] Zone 6 (at 7000 ft. elev.) in AZ,” the reader says. “All electric house (2400 sq. ft.). No heat bill ever exceeded $105 for an entire year.”

More to the point, the slab was insulated with 4 inches of extruded polystyrene (XPS) insulation, not the 12 inches of EPS that Roche is planning.

“Maybe Arizona has higher soil temperatures than N.H.,” the reader adds, “but I assure you that your design is on the right track.”

Other suggestions

Mike Kolder raises another issue, the possibility that rigid foam insulation will become a “highway for bugs to get inside the house.” The problem is far from unprecedented, and even though insect-resistant versions of rigid foam are available, it’s apparently not always easy to find.

For that reason, Kolder suggests that Roche use rock wool instead of foam.

Finally, there’s the question raised by Tyler Keniston of whether finished concrete floors typically get control joints. Cracking can be minimized by using lots of reinforcement in the concrete, Taylor replies, but control joints are more common.

“If you don’t like open joints,” he says, “you can grout them or fill them with sealant.”

Our expert weighs in

This is how GBA Technical Director Peter Yost sees it:

First, the guidance given by others in this exchange is per usual GBA standards—on the mark.

Second, transferring compressive loads and eliminating thermal bridges with slab-on-grade foundations is bedeviling. Especially in cold climates where the thickness of the vertical leg of insulation isolating the slab gets thicker and harder to manage. It’s a shame that appropriate levels of insulation are hardest at the slab perimeter, since according to at least some studies  about 60% of slab heat loss in cold climates is at the perimeter (insulating the rest of the slab is pretty easy in comparison).

Third, quantifying and accurately modeling heat loss and gain to the earth, which is an infinite sink, remains challenging and still is debated (see, for example, “Do We Really Need 12 Inches of Foam Under Our Slab.”)

The general hygrothermal rule for all assemblies is to push conductive materials (like concrete) either completely out of, or pull completely into, the conditioned boundary. The tough part comes with the adverb, completely.

I work with architect Steve Baczek quite a bit so I like to refer to his best-practice details whenever I can. Here is a detail from Steve showing the foundation wall pushed completely out of the conditioned space. The slab is on the inside of the thermal boundary.

In this scenario, the concrete foundation is pushed completely outside of the thermal boundary. Illustration courtesy of Steve Baczek.

Here is another detail from Steve in which he pushes the slab completely out of conditioned space.

In this instance, the slab is outside of the thermal boundary as well. Illustration courtesy of Steve Baczek.

On the issue of just how much we need to consider change in volume of EPS insulation, of any type (compressive strength), I checked in with my EPS manufacturer buddy on this and he told me that there is an ASTM standard for this (ASTM D2126) and that this standard permits 10% response to “aging.”

He added that for geofoam—EPS used below grade for infrastructure such as runways, roads, etc.—there is a different standard (ASTM D6817) that permits only a 1% response. I am not aware of builders or architects sourcing or using geofoam for foundation assemblies because it responds so much less over time, but this could be because our actual assemblies are not experiencing anything near 10% response over time with “standard” types of EPS and perhaps the typical dimensions of geofoam blocks are not suited well to residential construction.

Perhaps one way to think about this whole issue is that the compressive strength of concrete is incredible overkill for the loads of typical single-family residential structures, but transferring loads to a material that bugs can chew through with ease is nothing to trifle with.

Here are a few more resources on this topic:

—Scott Gibson is a contributing writer at Green Building Advisor and Fine Homebuilding magazine.


  1. Robert Swinburne | | #1

    Sometimes, more is easier when it comes to sub-slab insulation. When the excavator has prepped the area you can do some fine tuning. I recommend seeing what you can get for used rigid insulation. There are a few places in Mass that sell used insulation. Below slabs is an excellent place for this.

  2. Steve Knapp CZ 3A Georgia | | #2

    Great article. Re the bug issue. We have a lot of those down south, so I used borate infused EPS under the slab of my last house. In planning for my next build, I came across Glavel foam glass gravel. Has anyone tried this product? Combining gravel and insulation seems to simplify things, and I don't imagine the bugs will want to set up home in the stuff.

    1. John Clark | | #3

      Seems like a great idea for the termite kingdom AKA the SE US.

  3. User avater
    Carl Seville | | #4

    For those of us who don't live in the frigid north, it is instructive to recognize that R40 walls and 6" of XPS slab insulation does not provide enough benefit for the investment. 1-2" of slab insulation around the perimeter (none needed under the entire slab if not using radiant heat) and R-25-30 exterior walls is all you need for a very comfortable, cost effective structure.

    1. Steve Knapp CZ 3A Georgia | | #5

      Good to know, Carl. I assume you are okay with placing the foam around the inside of the footer.

      1. John Clark | | #8

        Looks like he is (See Gallery Photos)

        "I installed Owens Corning 1-inch XPS board around the perimeter of the slab and below the slab, 24 inches wide around the perimeter. This should give me all the insulation I need in Climate Zone 3, where it rarely gets very cold.".

    2. User avater
      Michael Maines | | #11

      Carl, I assume you are referring to projects not built in the north? I can assure you that in heating-dominated climates we want more than 1-2" of insulation at the slab perimeter and we definitely want insulation in the floor. Walls are another matter--R-30 CAN make sense, but it depends on the heating system.

  4. User avater
    Stephen Sheehy | | #6

    12" of foam seems way too much . What's the point? I can't imagine that 4" isn't adequate .

  5. Malcolm Taylor | | #7

    Illustrations 4-4 in the ORNL link are quite odd. They show a stem-wall reduced to about 3" at the top. I don't know how y0u would pour it, and the result would be very unstable.

  6. Seth W | | #9

    Great Article. I just built a slab edge scenario like this. Worked like a charm. I considered doing a plywood connector, but I just used a 2x6 PT bottom plate for the interior wall and put the 2x4 wall to the inside face of it. I held the 2x6 off the exterior wall plate, and used ccSPF as a flash coat over the bottom plates and bottom of wall. Exterior door thresholds are also tricky with this detail.

  7. Seth W | | #10

    I would also add that when working with thicker EPS - take extra care to make sure your compacted stone layer is dead flat. I had some areas that weren't as perfectly flat as they could have been, which magnified into EPS sheets not sitting dead flat.

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |