Resistances in series, and why the film layer matters
Heat leaving a radiant floor crosses everything between the tube and the room air, and those layers add up in series like resistors. The concrete or subfloor above the tube is one. The covering, its underlayment and any pad are the next. And there is always a third that people forget: the still air clinging to the floor surface, worth roughly R 0.6 for heat flowing upwards, which is present no matter what the floor is made of.
The output per square foot is the temperature difference between the water and the room divided by that total resistance. Everything that follows on this page is that one division, done twice — once without the covering and once with it — and compared.
The film layer is worth understanding because it is why a bare slab cannot deliver unlimited heat. The floor surface has to sit above the room to push heat through that film, and the hotter the surface, the more heat crosses it. Once the surface hits the temperature people will tolerate underfoot, that is the ceiling, and raising the water temperature past it only makes the floor uncomfortable rather than making the room warmer.
Where the numbers land for common coverings
| Covering | Rough R-value range | Effect on output |
|---|---|---|
| Ceramic or stone tile | Very low, well under 0.2 | Almost none. This is what radiant is designed around. |
| Vinyl plank or sheet | Low, but the attached pad is not | Small on its own, meaningful with a thick pad |
| Engineered wood | Moderate, roughly half that of solid | Noticeable, generally workable |
| Solid wood | Higher again, and it moves with heat | Both a thermal and a dimensional question |
| Laminate with foam | Moderate, the foam does most of it | Depends far more on the underlayment than the plank |
| Carpet with a pad | High — the pad is often worse than the carpet | Large. This is the case the page exists for. |
Those are orientation ranges and nothing more. Two carpets that look the same can differ by a factor of two, and the pad underneath varies more than the carpet does. Use the actual figures for what you are buying, and add the pad and underlayment into the same number, because the heat has to cross all of them.
The three ways a covering penalty shows up
The first is output, which is the headline. A covering that doubles the total resistance roughly halves the output at the same water temperature, and carpet with a pad over a slab does much more than double it.
The second is water temperature. To recover the lost output you have to raise the average water temperature by the load multiplied by the added resistance. A load of 15 BTU per hour per square foot across an added R of 2 is 30 degrees of extra water temperature, which is often more than the heat source has to give — and it is a great deal more than a heat pump wants to give, because efficiency falls as the water gets hotter. A system designed for low temperature water and then carpeted is not just weaker, it is more expensive to run.
The third is response. Every layer of resistance is also a layer that has to warm up before the room feels it, and a slab is slow to begin with. A carpeted radiant slab takes a long time to react to a setback or a change in the weather, which is why setback strategies that work on forced air often make radiant floors worse.
What to do when the covering is not negotiable
Sometimes carpet is the requirement and the floor still has to heat the room. The options, in the order most people should consider them:
- Reduce the load. Every BTU the envelope does not lose is one the floor does not have to push through the carpet, and insulation and air sealing are cheaper per BTU than anything on the mechanical side.
- Choose the lowest R-value pad the carpet will tolerate. The pad is frequently more resistive than the carpet above it and it is the easier of the two to change.
- Increase the area doing the work, if there is uncovered floor available, or accept a supplementary heat source in that room.
- Raise the water temperature last, and only after checking that the surface temperature cap and the heat source both allow it.
What does not work is treating the covering as a detail to be chosen after the system is designed. The covering R-value is a design input, on the same footing as the tube spacing and the design load, and the time to fix a mismatch is before the slab is poured.
Questions people ask
Can you put carpet over radiant floor heat?
Physically yes, and it is done. What the calculator shows is what it costs. A carpet and pad totalling around R 2 over a slab takes the total resistance from roughly R 0.74 to roughly R 2.74, so the output at the same water temperature falls to about a quarter of what the bare slab delivered. Recovering that means raising the average water temperature by tens of degrees, which the heat source may not permit and which pushes a heat pump into much worse efficiency. If carpet is required, the honest approach is to design the system around its R-value from the start, and most radiant manufacturers state a maximum total covering R-value for exactly this reason.
What is the maximum floor surface temperature for radiant heat?
Occupied areas are generally kept below the mid-80s Fahrenheit, with slightly higher figures sometimes accepted for bathrooms and perimeter strips where people do not stand for long. The reason is comfort rather than safety: sustained contact with a floor much above that is unpleasant, and there are circulation concerns for people who stand on it all day. The practical importance of the cap is that it limits output independently of the water temperature. With a 68 degree room and a still-air film around R 0.6, an 85 degree surface can only push out around 28 BTU per hour per square foot no matter how hot the water is. A room whose design load per square foot exceeds that cannot be heated by its floor alone.
Does engineered wood work better than solid wood over radiant?
Generally yes, on two counts. Its thermal resistance is lower for the same thickness in most constructions, so less output is lost. And its cross-laminated construction moves much less with moisture change, which matters because a radiant floor is deliberately drying the wood from below and cycling it with the heating season. Neither point makes solid wood impossible; it makes it a product-specific question with tighter conditions attached, usually including a maximum surface temperature, a narrower board width, and moisture content limits at installation. The dimensional side is worked through on the expansion gap calculator.
How do I find the R-value of my floor covering?
From the product data for the specific item, and add up the whole stack rather than the visible layer. A laminate plank might contribute little while the foam underlayment beneath it contributes several times as much. A carpet with an attached backing plus a separate pad is three layers. Where a product does not publish a figure, the supplier can usually obtain it, and it is worth asking before buying, because this is one of the few numbers where a wrong guess by a factor of two changes the design of the heating system rather than just the shopping list.
Is a thicker slab over the tubing better or worse?
Worse for output and better for evenness, which is a genuine trade rather than a mistake. More cover is more resistance between the tube and the room, so output at a given water temperature drops. But it also spreads the heat sideways before it reaches the surface, which reduces the striping between the tubes and lets you space the tubing further apart. Thicker slabs are also slower to respond, which is an advantage for steady loads and a disadvantage for anything that changes quickly. The cover depths in the assembly selector reflect common practice; the tube spacing side of the trade is on the radiant floor loop calculator.