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

installing radiant heat over existing basement slab

user-7498632 | Posted in General Questions on

We are planning on upgrading the heating system in our 1998 built scribed log home.  The builder put in five separate dual vent propane heaters (plus a wood stove) which are all aging, take a lot of space, and aren’t very efficient.  Our plan for the walkout finished basement is to put 1″ EPS foamboard covered with PEX, then wire mesh, then pumped concrete to minimum 2″ thickness above it all.  We’ll eventually do an acid stain then seal.  We lose about 1.5″ of ceiling height vs going with tile over concrete board but save thousands of dollars and like the look/durability.

Our questions are:  is it necessary to use a denser foamboard then the foamular 250 carried locally to support the weight, and also would a more dense foam be recommended or necessary to help in stabilizing the plastic foamboard staples holding the PEX tubing in place.

thank you


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. Expert Member
    Dana Dorsett | | #1

    A wire-reinforced 2" slab above foam fully supported by a structural slab below isn't going to flex enough on 12-15 psi rated foam to crack, but staple retention becomes an issue at less than 2lbs density.

    Foamular is XPS, not EPS, and though it's labeled R5 @ inch you'll read in the fine print that it's only warranteed for R4.5, and as it loses it's climate damaging HFC blowing agents over decades it drops toward R4.2 at full depletion, the same as EPS of similar density. Type IX EPS is nominally 2lbs, and holds plastic staples well, and is R4.2 @ 1" now and forever.

    In your situation 1" Roth panels on top of 1" EPS (cheaper 1.5 lbs density Type-II is fine) under a 3/4" subfloor or backer-board + tile may be a better choice than the thin slab approach. Roth panels have molded-in tracks for rounting half-inch PEX (no staples needed) and aluminum coating for heat spreading. That would eat up less headroom and provide a higher R between the tubing/aluminum and slab.

    There are other pre-molded EPS products designed for routing tubing without staples cheaper than Roth Panels that would be more appropriate if sticking with thin-slab concept, eg:

    In US climate zone 5 or colder locations it's really lossy to go with less than R10 under a radiant slab. An inch of foam and a thin slab without the radiant floor, going with a radiant ceiling or right-sized euro-style panel radiators can be more efficient and nearly as comfortable.

    From a heating cost point of view it's almost always cheaper to go with ductless air source heat pumps (mini-splits) than with condensing propane. Using electric mesh radiant floor operating off a floor thermostat (set only high enough to keep bare feet comfortable on concrete) and using the heat pumps to control the room temperature can still be cheaper, more comfortable heat than condensing propane using only a radiant floor for heat.

  2. Seabornman | | #2

    I don't know the pros and cons of fitting pex and wire into a slab that thin, but I've had a bad experience trying to achieve a quality job on a thin slab like that. The concrete didn't want to stay in pace, the edges curled when it cured, and there was a hollow feeling to it when it was done. I was able to patch the worst spots as I was putting carpet over it. You'll need some form of crack control.

  3. Expert Member
    Dana Dorsett | | #3

    An inch of XPS is only warranteed to R4.5, and will eventually hit R4.2 at full depletion of it's HFC blowing agents, the same as EPS of similar density.

    Type-II EPS doesn't hold the staples well- Type-IX does, and graphite loaded Type-IX EPS would give you the true R5.

    But even that's not really good enough from an efficiency point of view.

    There are several purpose molded EPS products with knobs/posts for holding the PEX in place without staples and putting a half inch to inch of EPS between the bottom and tubing. That can be fattened up with sheet EPS as-needed. Those systems can deliver R6 or better of foam between the tubing and slab below.

    Installing 2" of EPS under a 5/8" subfloor and using electric mesh radiant will be less lossy. In many markets the operating costs of condensing propane and resistance electricity are comparable. A mesh radiant floor operated off a floor thermostat with a heat pump maintaining the room temperature (or just the foam & subfloor, no radiant, just the heat pump) is going to have a much lower operating cost.

    If sticking with the 2" slab concept a basalt fiber reinforced concrete is your best bet for limiting cracking potential during curing and use. (I don't know how easy it is to get it in the small quantity you're talking though.)

  4. user-7498632 | | #4

    my apologies - we intend to use extruded polystyrene (which is abbreviated XPS not EPS - my bad). Foamular XPS 250 (type IV) has the same compressive resistance as EPS type IX (25 kPa). Does anybody have experience enough to know whether 1" of the XPS 250 is dense enough to hold the clips well or should I special order XPS 400 or 600 which is denser?

    The insulated moulded panels (Roth panels etc) cost over $5/sq ft and don't appear to add much to the R-value per the system per the radiant system supplier.

    thanks for everybody's input


    1. Expert Member
      Dana Dorsett | | #5

      >"Does anybody have experience enough to know whether 1" of the XPS 250 is dense enough to hold the clips well or should I special order XPS 400 or 600 which is denser?"

      Yes, Foamular 250 (or even the lower density Type X XPS such as Foamular 150) will hold the plastic staples for PEX adequately. But neither is really good enough thermally at just one inch.

      >"The insulated moulded panels (Roth panels etc) cost over $5/sq ft and don't appear to add much to the R-value per the system per the radiant system supplier."

      Roth panels are often installed over flat foam to boost the R, but has full channels for the tubing and built in aluminum heat spreader which makes them more effective for use under wood subfloors but also more expensive. I've never heard of a project paying more than USD $3.50 per square foot for Roth panels (which would still be less than CDN$5.)

      The simpler molded EPS, such as Ampex (Amvic) or Nudura or Thermal Foams Inc, which are often designed to be the top layer over foam of arbitrary thickness of your choosing, not the total R, and are substantially cheaper than Roth panels.

      R10 Ampex knobby PEX panels designed for use under slabs run less than USD $2/sq ft even at the big orange box store, but are probably thicker than what you're looking for.

      Have you run a Manual-J load calculation on this yet?

      What is the intended heat source?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |