Footage from spacing, and where the extra comes from
The tubing in the field is a straightforward geometric result. Runs at a given spacing cover a strip that wide, so the length of tube per square foot is twelve divided by the spacing in inches. At twelve inch centres that is one foot of tube per square foot; at six inch centres, two; at nine, one and a third. An 800 square foot room at twelve inch spacing needs 800 feet in the floor.
What that formula misses is everything between the manifold and the heated area. Each loop leaves the manifold, runs to where the pattern starts, and comes back, so the leader distance is counted twice per loop and once for every loop. Four loops with a twenty foot leader add 160 feet, twenty percent on top of the field footage in this example, and that tubing produces no useful heat where it runs through a corridor or a joist bay. It still counts against the loop length limit, which is the part people forget.
Why loops have a maximum length
Water going round a loop loses pressure to friction, and the longer the loop and the smaller the tube, the more pressure it takes to move a given flow. Past some length the pump cannot deliver useful flow, or can only do so at a head that makes the pump absurd. The commonly quoted ceiling for half inch tubing is around 300 feet, and this page uses that as a default because it is the figure most people will meet, but it is a working convention rather than a physical constant. The real limit depends on the tube inside diameter, the flow you need through it, the temperature drop you are designing to, and what the pump can produce against the rest of the system.
| Tube size | Default limit used here | Typical use |
|---|---|---|
| 3/8 in | 200 ft | Staple-up, thin slabs, tight bends |
| 1/2 in | 300 ft | The usual choice for residential slabs |
| 5/8 in | 400 ft | Larger areas, fewer loops |
| 3/4 in | 500 ft | Large commercial pours, snowmelt |
Treat the column as a starting point to check, not an answer. The tubing manufacturer publishes pressure drop data for its own product and the pump publishes a curve; the intersection of those two is where the real limit lives.
Flow, and the 500 in the formula
Flow follows from the heat you are moving and how much the water cools on the way round. The relationship is that BTU per hour equals 500 times gallons per minute times the temperature drop in degrees Fahrenheit. The 500 is 60 minutes times 8.34 pounds per gallon times the specific heat of water, so it is not a fudge factor, it is the mass of water you move per hour per GPM multiplied by how much heat a pound carries per degree.
An 800 square foot floor at 25 BTU/h per square foot is a 20,000 BTU/h load, and at a 20 degree drop that needs 2 GPM total, half a gallon per minute per loop across four loops. Those are small numbers, which surprises people who expect radiant to need a large pump. What radiant actually needs is enough head to push that modest flow through several hundred feet of small tube, which is a different requirement.
The temperature drop is a design choice with consequences. A wider drop means less flow and a smaller pump, but a larger temperature difference between the start and end of each loop, which shows up as a warm side and a cool side of the room. A narrower drop evens the floor out and costs more flow.
Balancing, and why equal loops matter
Water takes the easy path. Put a 150 foot loop and a 290 foot loop on the same manifold and, with the balancing valves open, the short one gets most of the flow and the long one gets starved. Manifolds have balancing valves precisely so this can be corrected, but correcting it means throttling the short loop until it is as hard to get through as the long one, which wastes pump head and takes patience with a flow meter. Laying the loops out at similar lengths in the first place is cheaper than balancing them afterwards. That is why this calculator reports the longest loop and the headroom left rather than only the count: if it tells you the loops came out short of the limit, that is usually the right outcome.
The other layout decision is pattern. A serpentine run puts the hottest tube at one end of the room and the coolest at the other. A counterflow spiral alternates supply and return runs so that hot and cool tubes sit next to each other, which evens the floor temperature out considerably for the same footage. The tubing quantity is identical either way, so the choice costs nothing but bending effort.
What the tubing quantity does not settle
Output per square foot is capped by floor surface temperature, and that cap is about people rather than about hydronics. Beyond a certain surface temperature a floor becomes unpleasant to stand on, and no amount of tighter spacing or hotter water gets past that. What sits above the tube also matters: bare concrete, tile, engineered wood and carpet transmit heat very differently, and a thick rug over a section of slab is a working insulator that the design did not know about. If a room needs more heat than the floor can produce within the comfort limit, the shortfall has to come from another emitter, and the time to find that out is before the pour.
For the load itself, work through the heat loss calculator to see the conduction and infiltration terms, then check the envelope side with the insulation calculator. Under-slab insulation deserves particular attention on a radiant job, since heat driven downwards is gone. If the heat source is a fuel-burning boiler, its installation, venting and combustion air are licensed work rather than a weekend project, for reasons set out on the mini-split sizing page. Running cost estimates live in the heating cost guide.
Questions people ask
How much PEX do I need per square foot?
Twelve divided by the spacing in inches gives feet of tube per square foot, before leader runs. Twelve inch spacing is one foot per square foot, nine inch is about 1.33, eight inch is 1.5 and six inch is two. Add the leader runs on top: each loop needs the distance from the manifold to the start of its pattern, counted twice, and with several loops that is not a small addition. Then add a waste allowance, because tubing gets cut short, kinked on a tight bend, or damaged during the pour, and a splice buried in a slab is exactly the thing nobody wants.
Why is 300 feet the loop limit for half inch tube?
It is a convention that comes out of pressure drop, not a hard number. Friction loss rises with length and with flow rate, and falls sharply with tube diameter, so a long loop of small tube needs a lot of pump head to move a useful flow. Three hundred feet of half inch happens to land in a comfortable range for typical residential flows and ordinary circulators, which is why it is repeated everywhere. If your flow per loop is unusually high, the practical limit is shorter; if it is very low, longer loops can work. The tubing pressure drop tables and the pump curve settle it for your job.
Can I run different spacings in the same room?
Yes, and it is common practice rather than a compromise. Tighter spacing at the perimeter, along exterior walls and under large glazing puts more output where the heat loss is concentrated, and wider spacing in the interior of the room where losses are lower. It also evens out the cold-feet effect near an outside wall. Calculate the tighter and wider zones separately here and add the footages, and keep in mind that a tighter zone consumes loop length quickly, so the perimeter band may need its own loop rather than being folded into a field loop.
What flow rate does each loop need?
Divide the loop share of the load by 500 times your design temperature drop. For a typical residential slab that lands somewhere near half a gallon per minute per loop, which is far less than people expect. The number that actually stresses the pump is not the flow but the head required to push it through several hundred feet of small tube, plus the manifold, valves and any mixing arrangement. Size the circulator from the total flow and the head of the worst loop, not from flow alone.
Does the calculator account for what is on top of the slab?
No, and that is a real limitation rather than an omission you can ignore. Floor covering resistance sits directly between the tube and the room, so the same tubing under tile and under thick carpet with a pad deliver noticeably different output at the same water temperature. A design that assumed tile and got carpet will underperform in that room. Decide the finish before the design, tell whoever is doing the water temperature calculation what it is, and be cautious about adding heavy rugs over a floor that was designed to the limit.