Some radiant installation can afford to push the envelope.

What's the most commonly used tube spacing for floor heating systems in the United States? Most of you will no doubt answer one foot.

But have you ever wondered why? Why should the performance of a heated floor be tied to a dimension that, at one point in time, was defined as one-third the length of some English king's arm!

Certainly the fact that the foot is a standard unit of length in the U.S. construction industry has a lot to do with its common use in tube spacing. In slabs, it allows the tubing to be secured easily to standard 6-inch by 6-inch welded wire reinforcing mat. It also allows tube spacing to be coordinated with other standard building materials, such as 4-inch by 8-inch sheets of plywood and drywall.

Pushing The Envelope

Have you ever contemplated installing tubing spaced well beyond the sacrosanct one foot? The foremost incentive would be to reduce cost by decreasing tubing footage and installation time. The chief reluctance is not knowing if the floor will perform properly.

From a manufacturer's standpoint, there's little incentive to encourage tube spacings beyond one foot. Wider tube spacings decrease tubing sales. Incorrectly applied, they also increase the chance of complaints, especially those due to inadequate heat output or excessive variations in floor surface temperature.

Still, I believe there are floors where tubing spaced 2 feet apart can provide performance that would be deemed acceptable by a vast majority of the building's occupants. The interior floor areas of large industrial, retail, or office buildings would seem to be ideal candidates. People don't walk barefoot on these floors.

Slightly greater variations in floor surface temperature probably would not even be noticed, much less voiced as a complaint. The heating loads of the spaces above such floors are often minimal and thus don't require heat outputs on par with those of a typical residential floor heating system. If only we could know how the system would perform.

Since there's very little information available on the performance of floors using tubing spaced 24 inches on center I developed my own model using heat transfer simulation software. I used this model to compare the performance of a heated slab with tubing spaced 18 inches and 24 inches apart, relative to a floor with tubing spaced 12 inches apart.

The floor I modeled was a 6-inch thick concrete slab, with 1-inch (R-5) underside insulation and no floor covering, which is typical of an industrial building. The soil temperature under the insulation was assumed to remain at a constant 65 degrees F.

Figure 1 shows floor surface temperature profiles for tubing spaced 12 inches, 18 inches, and 24 inches on center based on this model. The room temperature assumed for this particular run was 70 degrees F. The water temperature in all tubes was 100 degrees F.

Not surprisingly, the slab with tubes spaced 12 inches apart produces a significantly higher average floor surface temperature and lower peak-to-valley variation in floor surface temperature compared to the slabs using wider tube spacings.

The heat output from the floor slabs with these tube spacings and 100 degree F water temperature is given in Table 1.

At this water temperature, the output of the floor with 24-inch tube spacing is about 40 percent less than the floor with 12-inch tube spacing. However, even outputs of 15 Btu/hr/sq. ft. can be adequate for spaces with low heat losses due to limited exterior exposure ¡ even at design load conditions.

Figure 2 shows floor surface temperatures versus the T between average circuit water temperature and room temperature for tubing spaced 12, 18 and 24 inches apart. For 24-inch tube spacing, 100 degree Fahrenheit water temperature, and 65 degree Fahrenheit room temperature this variation is about 10 degrees Fahrenheit. This is less than that of a 1.5-inch gypsum thin slab with tubes spaced 12 inches apart and operating at 120 degree F water (about 12 degrees F). ItOs also considerably less than many "plateless" staple-up installations (15 degrees F to 20 plus degrees F). Furthermore, the variation is spread out over a much larger horizontal distance than in these other installations.

Crank It Up

So what happens if we need higher heat outputs, (say 30 Btu/hr/sq. ft.)? To answer this, I reran the simulations at higher water temperatures and used the results to develop the graph of heat output shown in Figure 3.

To use this graph, calculate the difference between the average water temperature in the tubing and the room temperature. Find this value on the horizontal axis. Draw a line up to the appropriate sloping line (based on tube spacing). From this point, draw a line over to the vertical axis to determine the floor's heat output under these conditions.

The average circuit water temperatures needed to deliver a heat output of 30 Btu/hr/sq. ft. to a room at 65 degrees F are given in Table 2.

The supply water temperatures would likely be 5 degrees F to 10 degrees F higher than these average temperatures depending on the flow rates of the circuits.

A supply temperature of 141 degrees F is not beyond what a thin slab with higher resistance floor covering or staple-up tubing would operate at. However, this temperature does preclude the use of heat sources such as a water-to-water heat pump. It's also higher than desirable for systems using condensing boilers.

The peak-to-valley variation in floor surface temperature of the 24-inch spaced floor at 30 Btu/hr/sq. ft. is about 19 degrees F. I feel this is a bit high from the standpoint of the owners' expectations of how a heated floor should feel, especially in residential settings. A peak-to-valley variation of 10 degrees F to 12 degrees F seems to be a reasonable limit that most owners will tolerate. There are plenty of systems out there (thin-slabs and plateless staple-ups in particular) operating with such floor temperature variations under design load conditions.

Keep in mind that all these numbers are based on a bare slab. If the slab were covered with a finished floor, the peak-to-valley variation in floor surface temperature decreases while the required average water temperature increases.

So What Can We Conclude?

If we accept the premise that floor surface temperature variations of 12 degrees F are acceptable from an occupant comfort/satisfaction standpoint, then there are opportunities for installing heated floors with tubing spaced 24 inches on center. The heat output from the floor would have to be limited to about 19 Btu/hr/sq. ft. The average circuit water temperature under such conditions would be about 107 degrees F for a room maintained at 65 degrees F.

This level of heat output could easily handle situations such as the interior areas of manufacturing plants or other large commercial spaces, especially when combined with closer-spaced perimeter circuits. It also would handle the design load of many basements.

Another good candidate would be interior office spaces that often have very little heating load other than ventilation. Over the years, I have specified tubing spaced at 18-inch centers in such spaces. On many such circuits, the balancing valves ended up nearly closed to prevent overheating. Why install more tubing than necessary and then choke off most of its flow?

I would not recommend 24-inch tube spacing for floors that will be repeatedly wet and then expected to dry quickly and uniformly. An example would be the parking area for snow plows in a highway garage.

I can, however, vouch for the fact that floors with tubes spaced 18 inches apart do dry well in such installations.

I would also avoid 24-inch tube spacing in residential buildings with the possible exceptions of basements that will be heated but not necessarily finished into living space, or select areas of very well insulated houses that need little more than a trickle of heat.

Is tubing spaced 24 inches apart a legitimate form of value engineering? When applied judiciously, to situations with low loads or tolerances for somewhat higher floor surface temperature variations, I think it is. Used discriminately it can reduce installation costs without sacrificing perceived quality. And that's a combination upon which all designers should capitalize.