The hydronic system that heats my small office building was installed in 1999. It started out as an oil-fired cast-iron boiler supplying a combination of radiant floor slab on the ground level, combined with some “had to have” radiant wall and ceiling panels on the second story. It’s also set up with a panel radiator and towel warmer. I wanted all these different heat emitters present so that I could provide “show & tell” to prospective clients over the years when I was designing systems.

About nine years ago, I added a monobloc air-to-water heat pump, which now serves as the main heat source with the boiler retained for those really cold (sub-zero) nights, or as backup if the heat pump was down for service. After nine years, I’m happy to report that the heat pump has operated without incident, and it has been exposed to some brutal winter weather. The heat pump also provides chilled water for cooling using a wall-mounted fan-coil in the office space.

It’s time

This fall, I started noticing the sound of air in the piping when a heating cycle began. Air inside the pipes meant that water had to exit somewhere. I feared a leak, and searched for it wherever I could see the piping. Nothing… Everything was dry.

In retrospect, what I should have done was to start looking at the drain pipe for the system’s pressure relief valve. That finally occurred to me one night as I lay in bed thinking about what I might have to do to find and repair a leak. At that time of day, my mind tends to conjure up a worst case scenario. Let’s just say it involved a jack hammer…

After realizing that I should have started at the PRV, that was the first thing I checked the following morning. I had forgotten that I had placed a small plastic pail under the PRV discharge pipe, and sure enough, it had a couple quarts of water in it. Some of you can relate to my sense of relief - and even joy at this discovery! Most of it stemming from not having to get a jackhammer involved.

As soon as I spotted the accumulated leakage from the PRV, I started tapping my fingernails on the lower portion of the diaphragm expansion tank. There’s supposed to be air in that portion of the tank, and if there is you’ll hear a nice hollow “ping” when you tap the tank.

What I heard was a muffled “thud.” I gently rocked the tank and noticed its inertia. The final test is to push in on the stem of the Schrader valve at the bottom of the tank. If water comes out the tank, it’s toast. Then, the only remedy is a new tank.

What’s going on?

When the diaphragm in an expansion tank fails, the tank will eventually fill with water. The air that was under the diaphragm will rise up out of the tank, and eventually make its way to the upper piping in the mechanical room, or higher if the piping creates a suitable path. If make up water is allowed into the system - either automatically or manually - that air will eventually be captured by an air separator or float vent, and get expelled from the system. All this eventually leads to the tank becoming “waterlogged.”

Since there’s no longer an air “cushion” in the system, the water pressure rises quickly whenever the heat source operates. In my system, the pressure easily made it up to the 30 psi rating of the pressure relief valve and thus some water was dribbled out of the system on each heating cycle.

But, if the system was waterlogged - as evidenced by the water ejected by the PRV - where was the air that made the gurgling sounds coming from? Here’s a hint: The pressure gauge on the boiler dropped to zero whenever the water in the system cooled down.

This is where it’s important to understand that a float-type air vent or automatic air separator can act as a vacuum relief device if the water pressure at the vent or separator’s location is lower than atmospheric pressure. When the water in the system cools following a heating cycle, its density increases slightly. Since this occurs within a rigid container (e.g., the piping system) that’s now completely filled with an incompressible liquid (water), the pressure at the air separator dropped below atmospheric pressure, and air got sucked into the piping.

An easy changeout:

Fortunately, I had installed lots of valving in the system, making it relatively easy to isolate the failed expansion tank.

I was able to break loose a 27-year-old threaded joint (making sure I had a backup wrench to prevent over stressing the 1/2” tee supporting the tank).

I checked the air pressure in the new tank while it was cold. It read 7 psi. A tad low, but fine for my relatively “short” system. Keeping the air pressure at 7 psi provided a bit more margin against the possibly that the PRV was going to start dribbling water at around 25 rather than 30 psi.

I doped up the threads on the new tank and screwed back in place. After opening the valves and enabling the fill system, water entered the system and the air separator began hissing. All good signs. The system is back to its normal quiet operation.

Take-aways

I wanted to write about this encounter because it demonstrated all the classic behaviors that pointed to a waterlogged expansion tank. The gurgling sounds in the piping, the repeated discharges from the PRV on each operating cycle, and a “thud” rather than a “ping” when tapping the lower portion of the expansion tank.

The original expansion tank served the system for 27 years. A very reasonable life, in my opinion. If I could roll back the clock to the original installation, I probably would relocate the tank from the hot to the cooler side of the system, and perhaps install a 1/2” ball valve directly above the tank connection to further reduce any water loss when the 2nd tank eventually needs replacement. I didn’t have quite enough clearance to other piping to do this when I swapped the tank.

If the second expansion tank lasts as long as the first, I doubt if I will be the one doing the eventual changeout. Still, this experience reminded me to think about the “long game” when designing or installing hydronic systems. Make the systems as easy as possible to service. Those who follow will appreciate and respect you for doing so.