Last fall I received an inquiry that went something like this:

“I'm planning a radiant floor heating system for my 3,000-sq.-ft. shop. One of the installers I've spoken with wants to install 500-ft. circuits of 1/2-inch PEX tubing because he says it makes the installation easier. I know from some reading I've done that this is quite a bit beyond the recommended 300-ft. maximum circuit length for 1/2-inch tubing. My installer says he will compensate for the longer circuits by using a high head circulator. Do you think this is a good idea?”

After reading this, some of you have surely rolled your eyes, thinking, “Here we ago again with another bad radiant job.” Perhaps, but let's look at this a bit deeper and put some numbers behind a decision to support, or recommend against, this proposed installation.

Is It Doomed?

We've talked about the implications of long radiant floor circuits in the past. They do cut down on hardware, such as manifold connections. They also speed the installation. But what about thermal performance? Can the use of a “high head” circulator correct for the potential thermal degradation due to an extra long circuit?

Although it's possible to investigate these questions using manual calculations, it's not an easy or fast process. Instead, I used the Hydronics Design Studio software to get the flows, temperatures and heat outputs for various scenarios being considered.

Let's assume design option No. 1 is to install six 500-ft. circuits of 1/2-inch PEX tubing at 12-inch spacing. This would cover about 3,000 sq. ft. of floor area. All circuits will be embedded in a 4-inch-thick concrete slab with no floor covering. The supply water temperature at design load will be 110 degrees F. The design load of the building will be 75,000 Btu/hr.

Given the long circuits, the circulator for option No. 1 will be a high head circulator, such as the Grundfos UP26-99.

Running this hardware configuration through the Circuit Simulator module is straightforward. The screen with the results is shown in Figure 1.

Each of the six 500-ft. circuits yields a heat output of 12,615 Btu/hr. The total heat output for the system is 75,690 Btu/hr. carried along on a pump flow rate of 7.52 gallons per minute. This output meets the assumed design heating load.

Divide And Conquer

For option No. 2, let's assume the 500-ft. circuits are cut in half, so we end up with 12 circuits at 250 ft. long. With no other changes in the system, the same UP26-99 circulator now delivers a flow rate of 17.3 gpm, more than double the previous flow rate. Even though the total length of tubing in the system is the same, the circulator “sees” the 12 parallel 250-ft. circuits as having far less hydraulic resistance compared to the circuits used in option No. 1.

The higher flow rate increases the average fluid temperature in each circuit, which, in turn, leads to greater heat output. Notice that the total heat output of the 3,000 ft. of tubing is now up to 91,035 Btu/hr. That's a 20 percent increase in heat output due entirely to higher flow rate. There was no change in supply water temperature or total amount of tubing in the slab. The screen showing these results is shown in Figure 2.

But since the heat output of the six-circuit system in option No. 1 (75,690 Btu/hr.) is sufficient to heat the building, and thus the higher output of option No. 2 isn't needed, the next logical question is: could I use a smaller circulator and still get a nominal 75,000 Btu/hr. output? This would definitely reduce installation cost.

It only takes a couple of seconds to change the circulator selection in the software and answer this question (see Figure 3).

In this case, the circulator was changed to a Taco 007 and all other parameters were left the same. The result is a total heat output of 82,914 Btu/hr., slightly higher than the heat output of the six-circuit system (option No. 1), and more than enough to meet the design heating load.

Most of you know that a Taco 007 (or comparable zone circulator) is significantly less expensive than a Grundfos UP26-96 (or comparable high head circulator). Most of you also recognize that the manifold hardware to supply 12 parallel circuits will cost more than a six-circuit manifold. Perhaps these cost differences are a wash: the savings associated with the small circulator is reclaimed by the higher manifold cost to supply 12 circuits.

For a moment you think back to using the high head pump and the longer circuits, since it doesn't appear you're going to save in hardware cost by going to the 12-circuit system and the smaller circulator.

Life Beyond First Cost

What some of you might not think about in weighing these options is the difference in operating cost of the high head circulator relative to the smaller zone circulator, especially over the long haul. Here's a peek at that difference.

The UP26-96 circulator has a listed power demand of 245 watts. The 007 circulator has a power demand of about 80 watts. If electricity is assumed to cost 10 cents/kilowatt-hour, the difference in operating cost for these two circulators is can be seen in Formula 1.

A savings of less than two cents per hour doesn't seem like much, does it? But what happens when you look at the life cycle cost of keeping these two circulators running 3,000 hours per year over the next 20 years? The numbers become a bit more impressive (see Formula 2).

Keep in mind this number doesn't assume any increase in the cost of electricity over the next 20 years. Anybody who thinks that's unlikely to happen should probably look for work in another industry. Even a conservative estimate of a 4 percent annual increase in electrical cost pushes the 20-year operating cost of the 245-watt circulator $1,474 dollars higher than that of the 80-watt circulator. Now we're looking at dollars comparable to the annual heating cost of a modest house in a cold climate.

Although the savings in using the smaller circulator may be a wash with the higher cost of a larger manifold station, there's no getting around the difference in operating cost.

Fortunately, few if any homeowners will ever question your selection of circulators. That's because you're supposed to be the professional that knows what's needed. Perhaps if they knew your selection would end up costing them an extra $1,500 in unnecessary operating expense over the next 20 years, they wouldn't be as trusting.

My point is to carefully scrutinize options that appear to reduce system installation cost or time at the expense of operating cost. In the long run, systems with high distribution efficiency almost always win the total cost battle over a hasty decision that only effects first cost.

High head circulators have their place in systems where the equipment imposes significant and unavoidable flow restriction, or where static lift head is necessary. But in the case of radiant panel circuits, you can control the flow resistance of your system. Keeping your circuits shorter and your circulator wattage lower is usually a wise choice.

Hydronic designers should keep this in mind on every project. We need to move our industry in directions that further reduce distribution energy requirements without sacrificing thermal performance. The systems we craft are not “disposable” entities. They should always be designed with the long term in mind.