GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Any experience with an unheated (but insulated) slab as a finished floor? Will it be uncomfortably cold?

emturnerNWO | Posted in General Questions on

We are building a 1000 s.f. single-story home in Northwestern Ontario (climate zone 7) on a very tight budget. It will be well-insulated (walls R-40, roof R-60), and have triple-glazed windows, etc. An energy efficiency consultant calculated our heat loss as 16,800 Btu/hr at -32.8 F. The foundation is a short ICF stem wall with a 4″ concrete slab (R-20 insulation underneath). Heating will be accomplished with the combination of a high-efficiency woodstove and electric baseboard, and we might add a mini-split down the road.

We would like to use the concrete slab as our finished floor for two reasons: 1) cost–we won’t need to pay for additional flooring, and 2) to act as a thermal mass, absorbing some of the heat from the woodstove and from the south-facing windows. However, a number of contractors have expressed concern that the floors will be intolerably cold, and have suggested we lay PEX tubing in the slab in case we want in-floor heat in the future. On the other hand, we have also been advised that in-slab heat will be too slow to respond to temperature fluctuations (especially with the wood stove) and therefore lead to overheating, and that the slab will feel cold even if heated (because the heating requirements are so low).

Does anyone have experience with an unheated (but insulated) slab in a cold climate? We do have a number of area rugs we can put down–that should help. My only point of reference is the mud room of our current home–it has a concrete slab (uninsulated) on top of bedrock, and standing on it without shoes for more than a couple of minutes is *not* enjoyable.

To summarize, our options are:
1) Unheated concrete slab as finished floor–simplest, least expensive option, but too cold?
2) Heated concrete slab as finished floor–adds complexity/expense we don’t want, slow to respond?
3) Unheated slab with a floating floor on top (i.e., wood, cork)–simpler, but do we lose the thermal mass benefits?

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. STEPHEN SHEEHY | | #1

    We're in the middle of building a house in Maine, design temperature a balmy 0°F. Like you, we opted for an unheated concrete slab as the finish floor. We put 4" of reclaimed XPS foam under the slab.

    I doubt putting heat in the floor will do much. Your heat load is pretty small, so any floor heating system will hardly ever be on. We have in-floor radiant heat in our present house and the toasty floor concept is largely a myth, since either the wood stove or sunshine usually warms the space, not the radiant system.

    I'd invest in a nice pair of socks and use the money saved by not putting heat in the floor to vacation in the Bahamas if your feet get too cold in January.

  2. emturnerNWO | | #2

    Thanks Stephen. That was my thinking as well, but it's hard to convince contractors--most seem to be really keen on in-floor heat, and only have experience with un-insulated slabs like basement floors. If we're spending so much to get a really good building envelope, I keep thinking we shouldn't have to install an expensive, complex heating system (plus there is the whole argument about in-floor being unsuitable for small, well-insulated houses). On our tight budget, every penny counts, so adding just-in-case PEX tubing means we'd have to cut some costs elsewhere.

  3. charlie_sullivan | | #3

    It is legit that a well insulated concrete floor, sitting at 68 F, will feel colder than a wood floor, also at 68 F. That's because it conducts heat away from your foot faster. Put a rug on it, or but on thick socks, and the effect disappears. And in any case, it's nothing like a 50 F concrete floor. And as Stephen says, putting heat in the floor isn't the solution. In fact, a heated concrete floor at 72 F probably feels colder than an unheated wood floor at 68 F.

    An argument that the contractors might buy is that if you are unhappy later, you'll put wood on top of the concrete. That's easier than putting pex into a slab that's already been poured.

  4. dickrussell | | #4

    I can speak from experience with a similar basement floor design. I have a superinsulated house, with R20 foam under the basement slab. With that much sub-slab insulation, the surface will be fairly close to room temperature, But as Charlie said, the more conductive concrete will feel cold to bare feet than a wood or carpeted floor at the same temperature. In the "living space" area of our basement, we wound up with fairly thin commercial carpeting in the "living room" and Pergo wood flooring in the bedrooms. Both areas feel just fine in stocking covered feet.

    Save the money otherwise spent on radiant floor heating. In a very well insulated house, you wouldn't be able to run the floor temperature up high enough to feel warm underfoot (upper 70s), or else your whole room would approach that and feeling stiflingly hot.

    I might add that originally we thought that we'd have the concrete floor downstairs polished as a finished surface. The contractor couldn't get a good result on a test area back in the utility space, claiming something like the concrete was too hard. I don't know if that was he case or he lacked expertise. Ultimately he applied a two-coat epoxy finish for the same price. Well, that wasn't done to what you would call a good finish for living space (too many applicator squeegee marks), so all of that later got covered with the carpeting or snap-together wood flooring. That final result is very comfortable, and very dry I might add, what with the foam, vapor barrier, and epoxy-coated concrete sandwich.

  5. krom | | #5

    It seems foolish to not put pex in the slab, it would be less than $500 in materials (even in CDN$)

    There is no way a 68 or 70 degree floor will overheat a room.

    Skip the baseboard get a small electric boiler, one ecm circulator pump and plumbing for it (should be 2k or less)

    Then set the thermostat and don't worry about ever having cold feet (you won't have a tosty warm floor, but it won't be freezing)

    Some fancy work with a controller and you might even be able to move some of the woodstove heat into distant rooms.

  6. STEPHEN SHEEHY | | #6

    Dick Russell: we're having our concrete floor polished and the polishing guy recommended a harder than usual(4000 psi) concrete mix.

    Trevor : the floor will be 70° without any heat in it. Why bother with pex?

  7. Expert Member

    Beth, the floor under my feet is exactly what you are proposing. If you are comfortable on a tile floor the concrete will feel the same. I'd suggest saw cutting the slab and grouting the joints. Use a water based sealer so that years down the road when it is time to re-finish the job isn't too onerous.

  8. krom | | #8

    I don't see the floor ever reaching the same temp as the air above it, unless it is heated, the floor will always be cooler

  9. emturnerNWO | | #9

    That's great Malcolm...alright, I'm getting more comfortable with this plan. As for the concrete finishing, that's a whole other can of worms. We live in a small town and, from what I understand, nobody here does interior finished concrete (although I'd be happy with something fairly basic). Thanks for the tips about the joints and sealer. I'll have to do some more research on the concrete and I'm sure I'll be asking more questions down the road. Another option, if we wanted to keep the thermal mass exposed (or if we really messed up the concrete) are the long ceramic tiles that look like barn boards--very cool looking, about $3/s.f.

    Trevor, I know the slab will be colder than the room, and I'm ok with that. I just don't want it to be *really* cold, like in our current mud room (your feet start to hurt after about two minutes). If it's about the same as ceramic tile, that's ok with me. And we could always cover it up later with wood if needed. I guess the tough part about all this is saying 'we will probably never have in-floor heat' when most contractors (around here, anyway) consider it to be the best heating option for every house.

  10. Expert Member
    MALCOLM TAYLOR | | #10

    Trevor, Surely it is also true of every other surface in the house? I know it's tempting to throw some pex into the slab, but you don't really end up with a useful radiant heating system. In for penny in for a pound. It needs to be designed, zoned and sized so that it integrates with whatever other heat source is going to be used. And at they point you are into something beyond a cheap add on.

  11. emturnerNWO | | #11

    That was exactly my thinking, Malcolm. It was suggested that we rough-in a really simple in-floor system without many (or any) zones to keep the cost down. That seemed like a really bad match to the woodstove (whereas with baseboard, each room will be zoned separately, and it should help even-out the the temperature variations causes by the woodstove heat). The quotes we got for the rest of the system (pumps, boiler, etc.) were several thousand dollars (we also live in an area with extremely high labour costs)'s definitely beyond a cheap add on. If in-floor was truly life-changing (or if it was substantially more energy efficient), then maybe I could justify it. But as it stands, I'm not sure it makes sense.

  12. Expert Member
    MALCOLM TAYLOR | | #12

    Beth, There are various levels of finishes for concrete floors. You can pay a premium for polishing, but i'd bet you will be happy with a good power trowelled surface that any of the concrete finishers that do garage slabs in your area would do. That's what I have been putting in houses for a couple of decades with great success.

    I agree with you; wood stoves and baseboard heaters are a very good combination allowing zoning of the rooms and complementing each other in the event when one of the heat sources isn't being used. I would suggest getting your electrical contractor to quote radiant cove heaters instead of the baseboards. If the upgrade wasn't too great, it might be worth considering.

  13. STEPHEN SHEEHY | | #13

    Beth: As Malcolm said, you need control joints so when (not if) the concrete cracks, the cracks don't show. You can use the joints as a design element. Whether to grout the joints is up to you. We're not planning any grout. We were adamant that the joints be cut precisely and the concrete sub did a really nice job. You can also place control joints under where interior partitions will go.

    You can always have the floor polished later if you can't find someone now. The guy who is doing mine showed me some floors that he polished after the slabs had been in place for many decades.

  14. Airithol | | #14

    Malcolm (#12),

    Regarding the varying finish levels, is there a significant durability / maintenance difference between the (finely) power trowelled slabs and the fully polished (burnished & hardened) slabs?

    We are considering the same things as Beth, but are not overly fussy on the level of gloss/shine - we could be happy with either. For us the most important point is the frequency of re-sealing necessary and the chemicals used.

    Thus far, I have been led to believe (by various concrete contractors, yes) that a higher degree of polish = less maintenance. The ones who do not offer higher degrees of polish, not surprisingly, suggest that power trowelling would be fine, and only require occasional resealing. The ones who offer polishing services suggest that a properly burnished and hardened slab should require no maintenance for 'many' (between 40-100 years) except perhaps a "re-buffing/sealing" in high traffic areas. Being a chronic skeptic, I suspect the truth is somewhere in the middle.

    My own research on this topic is, at this point, inconclusive. I would welcome any input you (or others have).

    Also, as a side, every contractor I spoke with (6 or so now) suggested that the control joints be filled with 'caulk', though when pressed about the actual product, I was told they would have to get back to me with specifics. When I mentioned grout (my original preference) to any of them, they all pretty much dismissed the notion out of hand.


  15. STEPHEN SHEEHY | | #15

    Jason: Has anyone explained why they recommend filling the control joints? I suppose the joints may fill with crud, but grout needs to be cleaned and caulk can get dirty or moldy. Our plan is to leave them open. They are only about 1/8" wide, nice and clean, with no chips or ragged edges. The caulked joints I've seen look awful.

  16. Airithol | | #16


    No, they haven't, but to be honest I never asked. As you say, they may fill with debris, which neither my wife or I find desirable. We have a few more in our area to talk to still, I'll ask if they recommend filling the joints, and if so, why.

    One of the contractors I spoke with said he normally left them unfilled, but would be happy to fill them with 'caulk', or whatever. I got the impression he would have used ketchup if I had asked (and paid him to). For this, and other reasons, he isn't on our short list.


  17. emturnerNWO | | #17

    I much prefer the look of unfilled joints. The key is probably to avoid cutting them too deep. Now that I think about it...I believe I asked our contractor (the one who did our septic and driveway) a while back about joints in the concrete and he said we didn't need them. But from what everyone is saying, it sounds like they are absolutely necessary. Yikes...I'll need to look into this some more, and maybe talk to some other people. The art of concrete finishing seems to be shrouded in mystery--it's been the most confusing part if this entire project.

    Thanks for the radiant cove suggestion, Malcolm. I'll definitely check those out. It would certainly free up space for furniture, etc.

  18. Airithol | | #18


    Based on what I've read, the depth for a cut is supposed to be 25% of the depth of the concrete (1.5" for a 6" slab). I have seen this many places, here is one.

    In practice, I am not sure this happens all the time. Most of the contractors I spoke to said they would be around 1" deep, even though I had not yet told them the proposed slab thickness. Also, embedded items (rebar, reinforcing mesh, PEX, etc) can interfere with cut depth if not planned for or placed correctly.

    With respect to unfilled joints, I do think they look nicer. I guess we can always fill them later if we are unhappy, though the work would likely be cheaper, not to mention less disruptive, prior to occupancy.


  19. Expert Member
    MALCOLM TAYLOR | | #19

    The wear surface of a concrete floor is the sealant, regardless of the degree to which the concrete is finished. When the sealer wears off the concrete will absorb stains and take on oils. I suppose if anything a case could be made that the sealer would adhere to a power trowelled floor better than a highly polished one, but really that's just splitting hairs. The only appreciable difference in the life of the sealer is use and traffic. After 18 years our kitchen needs re-sealing as does the front hall. Much of the rest of the main floor looks just as it did when first completed.
    Caulk is used to fill control joints in industrial buildings (look at the floor in HD or Costco next time you are in one). From the perspective of a structural engineer this makes sense, allowing full movement of the slab. But in a house, with good substrate preparation and no real loading, you simply aren't going to get much movement and grouting the joints or leaving them empty s largely an aesthetic choice.

  20. Airithol | | #20

    Thanks for the input, Malcolm.

  21. brp_nh | | #21

    We are starting our first full winter in a house with a monolithic slab, insulated with 4" XPS below and on the sides. The location is in the mountains of New Hampshire, zone 6. We considered running pex in the slab, but decided it didn't make sense with such a well insulated house. I'm glad we didn't and even that "small" amount of money would be better saved or put towards other areas. The floor feels comfortable.

    For budget reasons and a desire to continue on the DIY path, we are staining and sealing our floors with products from Eco Safety Products. While I'm sure professionally polished floors would be more durable, we like working with the product and the results. It's time consuming to prep (clean, etch, etc) the floors, but somewhat fun and satisfying.

  22. Ikillyou | | #22

    Great Thread, I just finished my plans for 1200 sq ft Coach House, lower level 3 bay garage/shop and 2nd floor 2 bed plus loft with cathedral ceiling. Full South exposure on the gable end and lots of glass with large overhangs. Minimal windows on the other elevations. Thinking double stud upper and 2x6 lower with insulated slab. Heating, Zuba Central heat pump with electric auxiliary heater and woodstove/fireplace upper and only fireplace lower to heat the shop area. My plan is to close in under the deck and use Sunspace WeatherMaster vertical window system (clear flex vinyl) and locate the heat pump in the closed in area 26' x 12'. I think with the solar gain and/or the heat from the fireplace will enhance efficiency.

    Looked at radiant and geothermal and decided against either.

    Planning on insulating the slab too just concerned about the garage doors and air infiltration although the upper and lower will be isolated and foamed. Any thoughts on the garage doors or suitable candidates welcome. Building near Britt one hour South of Sudbury on the Magnetawan river near Georgian Bay.


  23. Expert Member
    MALCOLM TAYLOR | | #23

    Stephen, I think you are right in that, beyond perhaps accumulating mung, there isn't much wrong with open joints. The grout does hold up much better than that around tile though. I've never had a lot of success with grout sealers but those for concrete are more like a urethane finish on wood floors and actually provide a surface as opposed to completely sinking in.

  24. Expert Member
    MALCOLM TAYLOR | | #24

    Jason, Cheers. Hope your house build goes well.

  25. user-917907 | | #25

    I know it's a little bit off-topic, but could someone explain why a slab can't be poured such that it won't crack? In years past concrete-paved highways had expansion joints every fifty feet or so. As you drove down them your car make an annoying ka-chunk, ka-chunk, ka-chunk. But now they pave roads with continuous pours of concrete, many miles long. And no expansion joints. I'm not even sure if they have re-bar in them either? So how is it they can now pour 10 miles of concrete with no expansion joints, but they can't pour a garage floor or a basement floor without putting in joints?

  26. GBA Editor
    Martin Holladay | | #26

    Modern highways are paved with asphalt, not concrete.

    Concrete still requires control joints. It shrinks as it sets.

  27. user-917907 | | #27

    I can assure you that there are modern highways still being paved with concrete. Since I asked the question I did a little research and found they may be more common in warmer climates than colder ones. They last longer than asphalt paving, but are more expensive, so apparently tend to only be used for major highways. I also found that all types of road-building concrete do crack, and thus require either expansion joints or steel reinforcement, or both. However, the type of concrete roadways that I've seen without expansion joints apparently has more steel reinforcing than those that require expansion joints.

    Continuously reinforced concrete pavements do not require any transverse contraction joints. Transverse cracks are expected in the slab, usually at intervals of 3 to 5 ft (0.91 to 1.52 m). These pavements are designed with enough steel, 0.6–0.7% by cross-sectional area, so that cracks are held together tightly. Determining an appropriate spacing between the cracks is part of the design process for this type of pavement.

    Continuously reinforced designs may cost slightly more than jointed reinforced or jointed plain designs due to increased quantities of steel. Often the cost of the steel is offset by the reduced cost of concrete because a continuously reinforced design is nearly always significantly thinner than a jointed design for the same traffic loads. Properly designed, the two methods should demonstrate similar long-term performance and cost-effectiveness. A number of agencies have made policy decisions to use continuously reinforced designs in their heavy urban traffic corridors.

    So it appears that if a person didn't want expansion joints in their concrete slab floor adding sufficient rebar to equal .7% of cross-sectional area might result in acceptably small cracks. That might work out to a 1/2" rebar every 12" for a 6" slab, or 1/2" rebar every 18" for a 4" slab.

  28. STEPHEN SHEEHY | | #28

    Jack: first you'll need to define acceptably small cracks. What's an acceptable crack in a highway may be totally unacceptable in your living room. I'm hoping my concrete floor only cracks at the control joints, but it was only placed a few months ago.

    Of course highways in frost prone areas may suffer unacceptable damage if water gets in cracks and then freezes, expands and creates bigger cracks. Typically around here (Maine) only some bridge decks are concrete and as far as I know, they have joints, filed with sealant that keeps the water out for a while.
    You never see new roads paved with concrete around here.

  29. user-917907 | | #29

    I actually don't know how large the cracks are in continuously reinforced concrete pavement, but I didn't see or feel the cracks when I drove over them at highway speeds.

    I should also point out that my arithmetic was faulty in my earlier post. For the re-bar to be about .7% of the cross-sectional area would require 3/4" re-bar @ 12" centers for a 6" concrete slab, and 3/4" re-bar @ 18" centers for a 4" concrete slab.

    But this is for roads that have to support heavy trucks. Your average basement slab won't have that type of stress (or probably the temperature differentials that roads do), so it may not need that heavy of reinforcement to minimize cracks. However, this is all just a guess on my part.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |