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
Green Building Blog

Another Take on a Concrete-Free Slab

Northern Minnesota designer Randy Williams omitted the concrete under the floors in this new house–an interesting new trend among builders of green homes

Randy Williams of Willcon Inc. recently contacted me regarding a project he is wrapping up in northern Minnesota. I was intrigued because he has eliminated the concrete slab foundation altogether, which, given the material’s carbon load, is a design decision I fully support. There were additional aspects of his approach that were equally interesting, including his use of cutting-edge products in a residential building market generally unfamiliar with them; his success building directly off insulated concrete forms (ICFs); and his choice of a plenum truss for running duct work. To inform his decisions, Williams had several conversations with Steve Baczek and Jake Bruton, both of whom are experienced in the science of the “slabless slab.” Ultimately, in lieu of a slab-on-grade foundation, Williams used two layers of 2-in. Type IX EPS, a 6-mil. reinforced poly vapor retarder, and two layers of 3/4-in. AdvanTech subfloor.

In large part, Williams’s approach was a response to the clients’ desire for a heated floor. “I wanted a system that didn’t need heat in the floor,” he explains. “In our climate, the heat moves downward and warms the ground, not just the upper living spaces. The system I used creates a warm floor without having to incorporate a heat source.”

Though similar to Michael Maines’s frost-protected shallow foundation, Williams’s approach starts with excavating down 5 ft. Additionally, Maines puts sleepers on top of the EPS foam, and his flooring extends to the exterior walls. Here the floor system butts up to the double bottom plate of the wall framing. “My concern with Maines’s method was moisture,” says Williams. “We couldn’t build it fast enough to get it shelled in before the floor assembly would get wet. With two layers of AdvanTech, two layers of foam, and poly, we would have trapped in all of that moisture. So we built up the stem wall and had everything dried in by the time we set the floor.”

Foundation and floor system

After excavating, Williams compacted the soil using both water and mechanical forces over a three-day period. Furthermore, the site sat for a month before construction got underway during which time several rainfalls aided compaction. Sand was the substrate of choice. Why not gravel? “We could have used all crushed rock but the sand compacts more tightly,” Williams notes, adding that it was unwashed and locally sourced. Footings for ICFs were poured and cured; and drainage was handled with corrugated pipe covered with gravel and a filter membrane.

On the interior, a layer of gravel was put down as part of the radon mitigation system, followed by sand topped with rock fines, one layer of 2-in. EPS foam, a layer of 6-mil. poly, the second layer of EPS, and two layers of AdvanTech glued and screwed with 2-in. square-head decking screws. A small gap was left around the perimeter to allow for expansion, and the vapor retarder is sealed to the bottom plate.

To achieve an 8-ft. finished ceiling, Williams combined a double-bottom plate with insulation over the ICF foundation. He explains that he went over the top plate with tape to connect to the ceiling air-barrier. “In this climate, we need something a little more closed than OSB for a vapor [retarder], so I chose Intello—its variable perm rating means it will open, if needed, and still get a decent air tightness value.”

To protect against the possibility of a burst pipe, Williams added floor drains in the wet areas, which included the mechanical room with washer/dryer and water heater. There he laid a porcelain tile floor over Schluter-DITRA to waterproof the assembly; the edges of the Schluter membrane go up the wall. The same treatment was given to the master bathroom with one difference—he used a curbless shower so water will find its way to the shower drain. And in the kitchen, he added a floor drain beneath the refrigerator.

Wall assembly and insulation details

The basic framing was conventional 2×6 walls on 16-in. centers—where the assembly differed was in building directly off the ICFs. As Williams points out, it’s more common to start framing off a slab or a floor system. Interestingly, the interior walls are not bearing on anything other than the AdvanTech. Asked if there is any concern around that, Williams answers: “There would be if the ground wasn’t compacted enough, then the whole floating slab inside the ICF would drop. That’s why we spent three days compacting.”​

Given the climate zone (7a), Williams prioritized vapor and thermal movement. “I knew I wanted to have a vapor-open assembly on the exterior, which is why I chose ROCKWOOL,” he says. “And I wanted to have insulation at the thermal boundary. We don’t do that very often up here, which is unfortunate. Builders aren’t willing to take the chance of putting insulation on the exterior because of the labor and costs.” (Pro Clima’s Intello was used for the interior air/vapor control layer.) Additionally, though air-tightness is typically achieved with the interior poly vapor retarder in his region, Williams wanted to move the air-sealing to the exterior, so he used ZIP System sheathing. He also went the route of a ventilated rainscreen on the exterior using Cor-A-Vent products.

Williams points out another atypical detail: “We strapped the ceiling with 2x4s—which is unheard of in this area of Minnesota—so the electrical wiring in the ceiling didn’t penetrate the air barrier.”

In addition to the continuous exterior insulation of R-8 ComfortBoard, the walls were insulated with R-21 fiberglass batts, and ThermalBuck was installed at the rough openings to allow windows to be in the same plane with the siding after 2 in. of Comfortboard 80 was installed.

Air-sealing electrical openings

“I knew we wouldn’t be able to strip the exterior walls to keep any of the electrical inside the building envelope,” Williams notes. “During blower door tests, I could still feel air moving through those outlets. So I did some research and found AirFoil boxes. They have a slot above and below the box where the electrical wire penetration into the box can be sealed with expanding spray foam. The limitation is that they come only in single two-gang and round models, so if we needed a three- or four-gang amp, we needed a bigger outlet box, so we had to use traditional foam gasket-boxes in two places.”

HVAC and mechanicals

A plenum truss was used to keep ductwork inside the heated space. “This was my first time using or even seeing a plenum truss,” Williams recalls. “It’s definitely something I would do again, especially for a slab-on-grade house. I can’t believe more builders in this climate don’t use them.” The plenum measured 8 ft. wide by 18 in. and ran the length of the building for a total of 56 ft. Nearly all of the HVAC ducts were concentrated in this space. (Ironically, to his regret, while Williams was attending the Fine Homebuilding Summit, subcontractors put some ducts in the attic space outside the plenum.)

Mechanicals included a ducted Mitsubishi air-source heat pump with PVA multi-position air handler, which has a rated capacity of 24,000 BTU, and features an integrated electric strip heater for when the heat pump can’t keep up with heating needs. Williams placed the water heater on an off-peak heating program that the local electricity provider offers; water is heated only in the middle of the night when the electricity rate is less than half the standard rate, and it is controlled by a radio receiver.

A negative-pressure blower door test performed on the shell during construction measured 1 ACH50; a positive-pressure test registered .41 ACH50. “I attribute most of the difference to the staples holding the Intello in the ceiling and garage/house common wall,” Williams notes.

To the Leviton load center Williams anticipates installing WiFi-enabled smart breakers to monitor the electrical consumption of the air-source heat pump and plenum heater.

Lessons learned

As with many first-time projects, Williams came away with ideas of how he might do things differently in the future. After applying Prosoco FastFlash for air-sealing between the ZIP sheathing and ICF foundation, for example, he remembered something Mike Guertin had done. “I saw a great detail from Mike that I wish I had done on this assembly,” Williams says. “He added a 3-in. metal plate over the exterior EPS foam on the ICF foundation. “Normally it’s lumpy where the ICFs lock together, and trying to air-seal to that is tricky. Putting that metal plate there before we poured would have provided a smooth surface to seal to.”

He might also opt to use rigid Comfortboard under the floor assembly. It is a non-petroleum based and therefore greener product than the Type IX EPS he used—though he notes he would consult an engineer to ensure it would be comparable.

Additionally, upon reflection, he might have gone with a different vapor retarder for the ceiling. “In our climate most builders use poly on both walls and ceilings but I worry about the walls being able to dry in both directions—that’s why I chose Intello,” he remarks. “This project had a vented attic. In the future, I might use a poly on the ceiling to save on cost. I’m not worried about it being wet up there. If it is, it’s because the roof is leaking. There shouldn’t be a vapor issue up there as long as it is airtight.” He also learned a lesson about the Intello itself: “You can’t use tape on the backside of it; it doesn’t stick. I had to add a second layer of tape. If I had used poly, it wouldn’t have been an issue.”

Of course, the duct work mishap was lesson-rich. Williams says, “Next time, I will have a hard work plan in place to ensure no ducts need to be run outside the plenum—even if I have to make a box and run them between two trusses.”

Finally, he adds: “I might choose to do a pressure-treated foundation system—it gets rid of the concrete altogether and it’s carpenter-friendly, meaning you don’t need another crew.”

Despite potential revisions to improve this assembly, it’s commendable that Williams is adopting building science principles in a region where they are not commonly understood.

-Kiley Jacques is design editor at Fine Homebuilding magazine. If you have a project that might be of interest to our readers, please send a short description and images to [email protected]. Photos courtesy of Randy Williams.


  1. Expert Member
    Rick Evans | | #1

    Cost Savings?

    First- thank you for this article. I am so glad to see that more attention is being paid to reducing the carbon footprint of our buildings up-front and not just energy savings over time.

    I really like Randy's design. By having two layers of EPS, Randy can protect the vapor retarder from damage. Obviously with a normal concrete slab, this isn't issue as concrete is an air barrier. But in this case, his vapor retarder is his air barrier. The upper layer of EPS combined with sleepers and subfloor should prevent any fasteners from penetrating.

    My Question:
    I wonder if one could opt for Type II EPS rather than Type IX? (Not talking about footings here, but the "slab" area.) There is probably a 25-30% cost difference between the two. With a conventional slab, Type II will suffice just fine. If you could also use Type II EPS for this arrangement, then I imagine that you could start to see real cost savings here. Perhaps wider sleepers- like 2x6 or perhaps strips of Advantech would make this possible?

    Great article! Thank you Kiley, GBA Team.

    1. Randy Williams | | #5

      We compared the cost of the system we built with a concrete slab and the tubing for the in-floor heating. We were a couple hundred dollars cheaper using the Advantech and EPS. Type II would save additional costs. I feel better using the type IX, never know what a future homeowner might load onto the floor system.

      1. Jason D | | #11

        Matt Risinger briefly shows a similar detail in a recent video:

  2. Expert Member
    Malcolm Taylor | | #2

    A really nice, well thought-out project.

    One comment on compaction: Doing it for a set period of time is a bit like acclimatizing wood flooring on site for a week rather than using a moisture meter. Instead of guessing whether three days is enough (or maybe one or four?), you can take the uncertainty out of it by getting a Proctor test. Then you know exactly where you stand.

    1. Randy Williams | | #7

      I am confident with how we compacted this foundation, but I agree, If we do another system like this one, we will be testing.

      1. Expert Member
        Malcolm Taylor | | #8


        You know what you are doing. My comment was more a general one for people attempting the task for the first time.

        I know this article is about construction techniques, but I'd love to see photos of the finished project.

        1. Deleted | | #28


    2. Jon R | | #22

      I'm interested in the validity of a Proctor test for measuring compaction when it is apparently a "optimal moisture content" test.

      "The Proctor compaction test is a laboratory method of experimentally determining the optimal moisture content at which a given soil type will become most dense and achieve its maximum dry density."

      1. Expert Member
        Malcolm Taylor | | #23


        You are saying you don't think the standard engineering tests to measure compaction under foundations work?

        1. Jon R | | #24

          At the end of the video, you see that the result of the Proctor test is a curve of moisture content vs compaction - ie, properties of a soil. But that says nothing about how well some specific site is compacted. So I'm curious how one proves that some specific site is adequately compacted to the right depth. Some kind of soil boring test?

          1. Expert Member
            Malcolm Taylor | | #25


            If you start with Proctor test you get the amount the fill can be compacted and how best to achieve that. It takes the guess work out.

            Our warranty provider required confirmatory compaction tests on the fill before the slab could be poured. They used several probes. Some seemed to core the fill, others measure resistance.

          2. user-7647820 | | #26

            A nuclear densometer checks the density in the field. Very likely the same company that did the proctor can provide the field test.

  3. James Someone | | #3

    I'm concerned in this quest for the greenest approach to building.

    I'm concerned about insects invading foam board in general outside of the building envelope and eliminating a concrete slab(insect barrier) comes with some risk.

    I couldn't feel secure about a foam/plywood slab. It's seems not robust enough against insects. Also, a floor drain under a refrigerator is likely to be forgotten and a source of IAQ problems from septic gases. Did I read that wrong?

    It's 70F in Southwest Virginia today. Wasps and bees are landing on my houses' south wall. Makes me wan't to live in a concrete bunker.

    1. Expert Member
      Malcolm Taylor | | #4


      living in a region with destructive carpenter ants I share your concerns, but not so much in this situation. If the insects get through 5 ft high concrete foundation walls they are already inside, and I'm not sure it matters much whether the foam they destroy is sub-slab, or under the subfloor. I'd be more concerned with an ICF foundation in areas where insect are a problem.

    2. Randy Williams | | #6

      I have yet to see any signs of insect problems in below grade foams in my climate. No termites, some carpenter ants but I have only seen them inside homes that have moisture issues. Our winters tend not to be conducive for most insects. One advantage to living in a climate with 10,000 heating degree days and at least a few days every winter with 100 degree differences between inside and outside temperatures.

  4. Cm Mm | | #9

    Great article! Does anyone have a link to the Mike Guertin metal icf foundation cover? Thank you in advance for your help. - Chris

    1. Randy Williams | | #10

      It's a tip he posted on his Instagram account, @mike_guertin. The post appeared October 30, 2018.

  5. WillTe | | #12

    I think the pressure-treated foundation is an interesting idea in combination with the slab-less slab. Would this be pier and beam? Or something like this:

    1. Expert Member
      Malcolm Taylor | | #15


      I think it would have to be the second, as pier and beam foundations don't have slabs to replace.

      The impetus to switch from concrete foundations to pt wood ones seems to have stalled about a decade ago, and I'm not sure carbon concerns are going to be enough to revive them.

  6. David Argilla | | #13

    Never seen Advantech OSB. What is the advantage over exterior grade T&G plywood ? Is it more resistant to water damage, or is it flatter and less likely to warp?

    1. Expert Member
      Malcolm Taylor | | #14


      It is impregnated with wax making it more resistant to water damage, and stiffer than plywood of equal thickness. It doesn't act as a variable perm vapour-retarder the way plywood does though.

      1. David Argilla | | #16

        Ok, thanks Malcolm. I found a local seller, and the price is essentially the same as Exterior T&G plywood. Leaning towards using this floor system instead of a concrete slab. Just need to find out how the building dept would react.

        1. Expert Member
          Malcolm Taylor | | #17


          I've been wondering the same thing. I pester my building inspector with annoying questions all the time. Maybe I'll try this one on him.

  7. George Hawirko | | #18

    Amazing how people prefer to overthink a good idea and waste money in the process.

    1. James Someone | | #19

      George, Can you elaborate?

      1. Expert Member
        Malcolm Taylor | | #20

        You can get good idea by looking at his posting history.

  8. George Hawirko | | #21

    There, Malcolm knows all and confirms my statement.

  9. Jon R | | #27

    If I understand the floor correctly, it consists of moist soil, a .06 perm poly vapor barrier, wood, then 0.006 perm Ditra. Nobody is concerned about the wood eventually rotting in this moisture trap?

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |