Radiant Heating - Taco Pumps Ran Dry for Few Hours - #30 Pressure Relieve Valve Popping Off Now

[cloves]

We have a closed radiant heat system and the supply to the pumps was turned off overnight and the thermostats called for heat. The following day when I got home I heard the pumps and they didn't sound right so I shut off the system. I checked the radiant side of the system and it was almost bone dry at one of the outlet valves. I proceeded to evacuate the air out of the system and then tested to see if pumps were still working ( I had flow when thermostats called. I noticed that pressure gauge (On the radiant side), which never has passed 12psi was now running higher 18+. Today I noticed that the #30 pressure valve was going off enough to fill a half bucket of water.

Water supply has been turned back on but something is clearly wrong with the system. We have a Taco 008 SS on the water side and a Taco 011 iron on the radiant side. Yes, they are high head pumps because our radiant guy was a moron who oversized the pumps. But that's a whole other issue.

Anyone have any insight on what could be causing the excessive pressure? The radiant side of the system is the one popping the #30 relief valve. I suspect my Taco 011 is damaged and causing excessive heat but having trouble wrapping my head around the why. The system uses a heat exchanger and since the supply side was the one starving for water not exactly sure if the problem is Taco 008 instead.

 

[Jadnashua]

You may still have some air trapped in the radiant side or, your expansion tank has failed and there's no place for the heating, expanding water to go except to almost instantly raise the pressure, then leaking out the now open safety relief valve to relieve it.

 

[Sponsor]

 

[Dana]

If the expansion tank loses charge the pressure can spike during burns. It needs to be pre-charged with air to the desired system pressure plus 1-2 psi when there is no pressure on the water side of the tank. If the expansion tank's bladder is shot it will dribble or spray water when pressing the pin on the Schrader valve (the air valve.)

If there is an auto-fill valve on the radiant side that is seeping the system pressure will rise over time. Most systems will have a full shut-off valve plumbed in series with the auto-fill, and closing the shutoff valve after filling to the desired system pressure will prevent a seeping auto-fill from overpressurizing the system.

The -0011 isn't a particularly high head pump (the -007 pumps more water at 13' of head than the -oo11) , it's a high volume pump. The -008 is indeed a high head pump:

With some radiant systems with gazillion loops plumbed in parallel a high flow pump might be called for, but there's almost no point to putting high-head pump on the potable side. It might need to be higher-head to deliver enough heat if it's on a high pumping head water tube heat exchanger tankless water heater, but an -008 would probably over-pump the water heater to the point that it's lifespan is shortened considerably.

What is the source of heat here?

Before just swapping out expensive pumps try running some napkin math on the system, starting with estimating just how much heat is needed by correlating wintertime fuel use against heating degree-days between meter-reading or fuel-fill-up dates using this methodology.

Running the heating system at anything more than 1.4x the 99% design heat loss is overkill, but up to 1.7x isn't going to change efficiency or longevity by very much. A typical tight 2x4 framed 2500' in NJ will have a design heat load of 30,000 BTU/hr @10-15F outdoors if it has an insulated foundation to as much as 40,000 BTU/hr with an uninsulated somewhat air-leaky basement or crawlspace. At a delta-T of only 10F on the potable side of the heat exchanger 40,000 BTU/hr that takes about 8gpm, which would wear out any tankless water heater before it's normal lifecycle in a heating application. At a delta-T of 20F it takes only 4 gpm, and a lot less pump.

But run the fuel-use based load numbers and report back what those are, along with more specifics on what you're using on the potable side, the size & condition of the house (air tightness, window & insulation specs, etc.) What water temperatures are you running on the radiant side, and what water temperature on the potable side?

 

[cloves]

Thank you the replies. To further add some details. We are using an AO smith vertex to provide both domestic hot water and heat. While changing the anode rods I turned off the supply and return valves on the vertex to the taco 008. I had the thermostats down to 50 and didn't expect the temp to drop overnight the way it did which is totally my fault. Yesterday, I swapped in another expansion tank I had on the radiant side, prefilled it to 12psi as I had in the past. I then turned on the water supply and opened the upper bleeder valve by the air eliminator and made sure I had water coming out followed by cracking open the cold return on the manifold. Took a few minutes before my taco 008 started to sound normal but it did eventually.

So I watched the system for an hour. The temperature slowly started to creep up to 16 psi. Before this the system ran a solid 12psi. Last year I added a slightly larger expansion on the domestic side because occasionally I was seeing the psi on the radiant side also climbing, but after the larger expansion tank everything was running smooth.

I only have 3 loops currently running on the system as I am still working on the house. I have attached a rough drawing of my system.

@jadnashua I swapped in a new expansion tank on the radiant side thinking perhaps the pumps running almost dry for who know hows long damaged the tank but seems like that the old expansion tank was ok and holding pressure after I took it out. Since I had the water shut off to the radiant system (supply/return) when this happened I don't think the domestic water expansion tank could have been compromised.

@Dana interesting point on the feeder valve. My feed line (as shown in picture) is linked to the return line. Last night after restarting the system, I turned off the shut off valve on the feeder to see if that would stop pressure from going up. It still went up slowly. I only tested it for about an hour.

Here are links to larger images if the forum ones are too small to see.

 

[Mliu]

I have attached a rough drawing of my system.

First, let me say that you've produced one hell of a rough drawing! Very nice. (I'm saving a copy for future reference.)

On to your interesting puzzle...

You said you shut the supply and return valves from the water heater overnight and you didn't discover the problem until you got home the next day. So I suspect the system was running like this for more than just a few hours; probably a lot closer to 20 hours. The difference is significant.

When you closed the supply and return valves, you created a closed system on the heat-source loop. Then, when your thermostat called for heat, the Taco 008 started running. But it was trying to pump in the closed loop: nowhere for the water to go (there is an interconnection to the radiant loop, but the PRV would have blocked water from entering the radiant side).

So the pump was furiously churning the water for ~20 hours. That energy had to go somewhere and where it went was in heating the trapped water. But there is no expansion tank nor relief valve on that loop because it's never expected to operate with the supply and return valves closed. Water is essentially incompressible so as it heats, it expands, and when there's nowhere to expand to, the pressure spikes.

I suspect that what happened in your system that a crack developed in the heat exchanger between the heat-supply loop and the radiant loop. This would explain why your pressure keeps rising in your radiant loop and why your 30# relief valve is blowing off.

I'm at a loss to explain why your radiant loop was "almost bone dry at one of the outlet valves." How did you determine that: did you manually open one of the unused valves on your radiant supply manifold?

 

[Dana]

Thank you the replies. To further add some details. We are using an AO smith vertex to provide both domestic hot water and heat. While changing the anode rods I turned off the supply and return valves on the vertex to the taco 008. I had the thermostats down to 50 and didn't expect the temp to drop overnight the way it did which is totally my fault. Yesterday, I swapped in another expansion tank I had on the radiant side, prefilled it to 12psi as I had in the past. I then turned on the water supply and opened the upper bleeder valve by the air eliminator and made sure I had water coming out followed by cracking open the cold return on the manifold. Took a few minutes before my taco 008 started to sound normal but it did eventually.

The pumping head of the Vertex is extremely low and the pumping head of the (as yet unspecified) heat exchanger isn't much. The Taco 008 is totally not the right pump for the application.

So I watched the system for an hour. The temperature slowly started to creep up to 16 psi. Before this the system ran a solid 12psi. Last year I added a slightly larger expansion on the domestic side because occasionally I was seeing the psi on the radiant side also climbing, but after the larger expansion tank everything was running smooth.

First "...16 psi..." is a pressure, not a temperature. The verbiage is a bit confusing.

Is that another pressure gauge on the potable side, right next to the Watts backflow preventer in your diagram? If yes, is that the place where pressure that was creeping up? (If the potable side is that low a pressure the expansion tank shouldn't be pre-charged to 60psi.) Or was the pressure gauge on the heating system side? There also appears to be another pressure gauge at the expansion tank & air eliminator(?).

When the -oo8 is running the pressure at that point will change based on the amount of pumping head at flow through the heat exchanger, the ball valves, the Y-strainer and plumbing. If I had to guess the pressure at that gauge would drop at high flow when the 008 is running, but it can be tweaked by partially closing the ball valve on the return branch to the Vertex.

The radiant side needs to be pre-pressurized to 12 psi. If there isn't enough pressure to begin with on the potable side it will never get there. I'm not sure what the minimum pressure drop is on the Taco PRV, but I'm pretty sure it's not zero. Was the pressure on the radiant side the full 12 psi before turning on the system?

interesting point on the feeder valve. My feed line (as shown in picture) is linked to the return line. Last night after restarting the system, I turned off the shut off valve on the feeder to see if that would stop pressure from going up. It still went up slowly. I only tested it for about an hour.

If the pressure is continuing to rise when the isolating ball valve is closed the most likely cause is a leak between the potable and heating system side of heat exchanger. If you've been pumping the hell out of it with the Taco 008 it may have created a leak by eroding the metal at high flow.

As a general rule it's better to turn that ball valve off after filling and fully pressurizing the system side, and only open it up if the system pressure is running low, which can happen if it develops a leak, or has purged a lot of dissolved air over time.

Normally changes in pressure on the potable side will not affect the pressures on the system side, and conversely- they are isolated from one another from that point of view. Many things can affect the pressure on the potable side, but on the heating system side it's generally only temperature, or in this case (apparently), seepage/leakage from the potable side.