Twinsprings Research Institute

I mentioned there were several things wrong with the first installation of boiler piping. When tested, water was not getting through the system and I believed the pump was just not strong enough to circulate through all the radiant loops. That would be solved by changing the simple single looped system to the more common and complex primary/secondary piping.
After pulling out the pipes and rerouting into the new configuration, there was still a problem with the water getting through the system. It turned out that the valves are configured to open only one way. At some point I was looking for the recommended pump location and got into my head that I could assume the valves flowed down. I did remember the valves should be on the cold side to limit wear and tear from the hot water. But it was not obvious to me which way the valves opened. I thought since the lever that holds them open manually was on the bottom of the valve, that the flow was from top to bottom. Incorrect assumption!

Honeywell Valve Diagram

Of course I was not looking at this obvious diagram, but at this installation.

Valve Piping Covered by Insulation

If you know what you are looking at in the above photo, you can see that the air scoop and expansion tank are on the lower pipe. Typically they are on the hot side, but NOT ALWAYS in older, existing installations. So I assumed (again) that the higher pipe was the hot supply.

The flow is shown in this photo and is obvious now that you know the valves flow upward.

Valve Flow

The lower pipe is the hot supply with the water flowing directly into the radiant/registers. On return the Honeywell flow valves are controlled electrically based on thermostat temperatures. The return water can also bypass the valves to bleed the system of air.

The major error in the piping was that twice both in the simple loop and in the first attempt at primary/secondary, I plumbed the hot into the upper pipe. After some consideration, I realized the error and replumbed all the piping so that the lower pipe was the hot supply. This meant remounting the secondary pump and I had to buy more 1″ copper and more fittings to reconfigure everything. The pipes seem more complicated than they should be but I could not figure out how to clear up what is a typical rats nest created by an inexperienced plumber, but at least now the plumbing is correct and the water flows properly.

New Challenger Boiler Piping

This is the new primary/secondary piping. It follows several boiler piping rules of thumb, although the 009 pump is now a bit over-sized for the primary loop. There is a secondary 011 pump that also is capable of pumping at high head because a radiant loop system consists of several turns of pipe that create a high head condition.

New Boiler Piping Illustrated

Primary Secondary System Piping

In place of the simple in/out system piping we now have a primary loop that circulates from hot to cold directly as well as a secondary piping loop that draws hot water from the primary loop and returns it to the primary loop. This is similar in function to the simple system except that the primary loop allows the boiler to circulate at a different and constant speed, while the secondary loop that feeds the heat emitters (radiators and radiant) can vary in speed and pressure without reducing the pressure to the boiler.

The main feature of the primary/secondary design is the “closely spaced tees” that cannot be more than 12 inches apart but should be only 4 pipe diameters (4″ for 1″ pipe) apart. The hot water line length to the tees is 8 times the pipe diameter while the cold water line needs only be 4 pipe diameters long. The Taco 009 pump runs only the primary loop while another Taco 011 pump runs the secondary loop. Both pumps are stainless steel even though cast iron is acceptable in a closed loop system, I was concerned about water quality for boiler longevity so made the whole system brass, copper, and stainless steel as well as using oxygen barrier pex.

I added several gauges to the return line: temperature, pressure, and flow. That way I can compare the pressure differential between the hot water supply as it leaves the boiler and the cold water return. This difference is called a delta T and has a role in the system design. I originally was planning for a delta T of about 20 but it appears we are getting closer to 10. Sometimes when only the radiators are calling for heat in this combination system, the boiler is not firing because the return temperature is higher than the outdoor reset curve. That is how it should work–unless the radiators are not hot enough to warm the space. So I have to keep an eye on these temperatures.

Some of the other components would have been added to the simple system as well. The boiler feed valve is a Caleffi and maintains a pressure of about 10-12 psi in the system. The valve can be closed to keep new water from entering the heating system. There is a Spirovent air separator. When the system was first purged of extra air (by opening the drains at the Honeywell valves and running the loop water into a bucket until no more air bubbles appear) the system was still airy enough that water was burping out of this vent so I attached a pipe to drain it to the sump, later I removed the pipe.

The system also has a Spirotrap dirt catcher on the cold water return side. There is an acid neutralizer for the condensate overflow so that the water from the boiler condensation process is treated with limestone before entering the sump. There is a boiler pressure relief valve that is piped to the sump too. There are shutoffs before and after each pump and a boiler drain valve as well as a hose (green in the photos) that is hooked up to the Honeywell control valves in the manifold that drain each loop as it is opened.

Single Loop System

The original boiler piping design was going to be a single loop system. Return water would be pumped through the system to the boiler and hot would flow out to replace it. One Taco 009 pump was chosen for the job due to its high head pump curve. Seen below, the 009 pumps fewer gallons per minute but at higher head in feet than the other larger pumps and higher than the smaller pumps that pump similar gallons per minute.

Taco curves 009-0014

This graphs also compares the 009 with the smaller pump sizes.

Taco pump curves 003-0014

This was a beautiful simple system. However, when tested, I could not get the water to pump fast enough through the piping to keep pressure in the boiler. Actually, there were several things wrong with this first installation that might have been changed one at a time and this simple loop system would have worked. I did not know that at the time, and only found the issues one by one. My first step was to abandon the simple system for a more complex and more usually recommended primary/secondary boiler piping design.