Figure 1

Last spring I visited a town highway garage heated by a radiant floor system I designed four years ago. Although the building was comfortable and the occupants happy with the system's performance, I couldn't resist a few minutes of hydronics contemplation in the boiler room.

The dual oil-fired boiler system was supplying heat to a short primary loop under partial load conditions. The only load was an injection pump operating at a relatively low speed. Based on the outside temperature, I estimated the space heating load at around 25 percent of the design load, or roughly 36,000 Btu/hr. One of the two boilers was operating with cycles on the order of one minute. Such short operating cycles are not good for any boiler. They lower overall efficiency, increase emissions and certainly accelerate wear that eventually causes some component to fail.

The boilers were piped and wired as shown in Figure 1. Each boiler got its on/off signal from a set of relay contacts in the staging controller. This controller monitored the supply temperature of the primary loop, and compared it to a "target temperature" that was continually calculated based on outdoor reset logic.

The differential on the staging controller was set to 20 degrees F. A call for boiler operation occurred when the supply temperature sensor dropped 10 degrees below the target temperature. Boiler operation stopped when the supply temperature climbed to 10 degrees above the target temperature.

The circulator on each boiler operated from the (C1/C2) terminals on the boiler limit controls. Each boiler's circulator ran whenever the staging control called for operation of that boiler.

formula 1

Bouncing Btus

Assuming the injection pump was extracting 36,000 Btus/hr. (or 600 Btus/minute) from this loop, the rate of temperature drop is astonishing:

formula 2

You can probably guess what happened next. That's right, another very short cycle as the primary loop temperature quickly shot up to the turn-off temperature of the staging control.

While in the mechanical room I did try widening the differential on the boiler-staging controller. This didn't help much because the thermal mass of the primary loop was just so small compared to the rates of energy input and output.

It soon became apparent that lots of heat was left stranded in the boiler each time the staging controller terminated its call for heat and the boiler circulator stopped. Each boiler contained about 550 pounds of cast-iron and 12.5 gallons of water. This gives each boiler a thermal mass of about 170 Btus/degree F. The water in the tubing between the boilers and the primary loop added another 42 Btus/degree F, bringing the total thermal mass to 212 Btus/degree F. That's over 14 times the thermal mass of the primary loop.

If the boiler circulator could remain on after the staging controller stopped the burner, the heat stored in the boiler and its connecting piping would be available to the primary loop. The combined thermal mass of these components would be about 227 Btus/degree F. Assuming the same 600 Btus/minute heat extraction, the rate of temperature drop at the supply sensor would now be:

Figure 2

So what's the best way to operate the boiler circulators? Although wiring them for constant operation during the heating season would be simple, it would defeat the cardinal rule of multiple boiler systems: Don't pump heated water through unfired boilers. Doing so merely uses the unfired boilers as heat emitters. A better solution was to run each boiler circulator for a "post-purging" time interval after the staging control turned off the burner. Time for some creative wiring.

Figure 3

Hangin' In There

With a typical DOB relay, line voltage is always applied across the "input terminals." When an external contact called the "control switch" closes, the main contacts instantly move to their operating position. The time delay period doesn't begin until the control switch opens.

This is not an ideal match when the circulator is operated from a typical high limit control that simply turns line voltage on and off. The work-around is to install a "pilot relay" to provide an isolated contact closure to the control switch inputs on the DOB relay. The pilot relay would use a line voltage coil wired to the (C1/C2) terminals in the boiler limit control as shown in Figure 2. You can buy both relays for this circuit for about $65.

An even better approach is to use what one manufacturer calls its "true off delay timer." Such a device doesn't need a control switch or a pilot relay. Instead, it uses an internal capacitor to continue supplying power to its own relay coil for a set time beginning when power is interrupted. This "holds in" the normally open contacts allowing the circulator to continue operating. One of these devices would be installed on each boiler in a multiple boiler system.

One specific product that provides this action is the GT3F-1AF20 from IDEC Corp. (800/262-4332). This device, shown in Figure 3, can be set to keep the boiler circulator operating for up to 10 minutes after the line voltage is removed from the (C1/C2) terminal in limit control. The relatively simple wiring for connecting this device to a typical boiler high control is shown in Figure 4.

Here's the operating sequence: When the boiler staging controller opens the (T/T) circuit of the boiler limit control, line voltage is removed from terminal (C1). This starts the timing period on the GT3F-1AF20. The relay contact in the GT3F-1AF20 is held closed by power from the internal capacitor allowing line voltage to reach the circulator and keep it running. When the timing interval set on the GT3F-1AF20 expires, the contacts open and the circulator turns off. This same process repeats itself each time the staging control turns off a boiler.

This particular device also allows any time delay period to be interrupted by providing a momentary switch closure between the "reset" terminals 1 and 4.

The GT3F-1AF20, along with an eight-pin mounting socket, costs about $75. It's small and rugged enough to be mounted in a junction box adjacent to the circulator. The electrical contacts can handle currents up to 5 amps and voltage up to 250 VAC. This should handle just about any boiler circulator used in a residential or light commercial system.

Figure 4

Who's Driving Anyway?

When setting up a system having high limit controls on each boiler, as well as a boiler staging controller, be sure the temperature on the limit controls are set high enough to prevent interference with the staging controller. If its temperature setting is too low, the limit control may interrupt boiler firing before the staging control opens its contacts. This often leads to unpredictable burner firing as the two controllers jockey for who's in charge.

In multiple boiler applications, it's best to think of the high limit controls as over-temperature safety devices rather than operating controls. They should open the burner circuits only if the staging control malfunctions and fails to open its contacts. This is also the case for any manual reset high limit device that may be present. In other words, let the boiler staging control do the driving and keep the limit controls as emergency brakes.

This control modification, when installed in the system, will correct the short cycling situation at a relatively low cost, certainly less than adding a buffer tank to the system for additional thermal mass. It's a detail likely to see future use in other multiple boiler jobs. The bottom line: Don't overlook the stabilizing effect of thermal mass when it can be "harvested" through creative design.

Side Note