Hot water heating system pressure

[wwhitney]

If I fill the system to 19psi when cold it does reach 30 psi by the time the water reaches about 160-170 degrees. But if I drop the pressure much below that, I find the top floor radiators will take in air and wind up cold. I feel like we shouldn't have a pressure change that large.

I have no experience but recently installed my first hydronic system (air source heat pump) and have wondered about the pressure vs temperature behavior of a closed hydronic system. So here's a simplistic analysis from first principles, corrections welcome:

For the pressures involved, I believe treating water as incompressible is accurate enough. If you raise the water from 20C (68F) to 75C (167F), then a random website tells me the volume increases 2.4%.

If the closed system is made of all steel components, then steel has a linear coefficient of expansion of around 1*10^(-5) / degree C. The volumetric coefficient of expansion is 3 times that, and so for a 55C change in temperature, the components should expand in volume by about 0.16%. So let me ignore that volume change. [If much of the volume of the system consisted of plastic pipes, the effect would be an order of magnitude larger.]

There's air in the expansion tank, which is rising from 19 psig (34 psia) to 30 psig (45 psia). That means the air volume has changed by a factor of 34/45 = 75%. I.e. from the 20C start, the air volume has gone down by 25%, while the water volume has increased by 2.4%.

I conclude that at 20C, the volume of water in your system is about 10 times the working volume of air in your expansion tank (25%/2.4%). Is that at all close to being true? If so, seems like the pressure increase you are seeing is within expectations.

Cheers, Wayne

 

[Jadnashua]

https://www.omnicalculator.com/physics/boyles-law may help.

If you estimate the volume of the water in the whole system, and know what the min/max temperatures are, you can figure out the volume change, and then plug that into the Boyle's law calculator to see what the pressure change will be. Note, that assumes constant temperature, and you can get more accurate if you include that, but that should give you the lower bond of pressure change.

 

[Sponsor]

 

[DIYorBust]

https://www.omnicalculator.com/physics/boyles-law may help.

If you estimate the volume of the water in the whole system, and know what the min/max temperatures are, you can figure out the volume change, and then plug that into the Boyle's law calculator to see what the pressure change will be. Note, that assumes constant temperature, and you can get more accurate if you include that, but

https://www.omnicalculator.com/physics/boyles-law may help.

If you estimate the volume of the water in the whole system, and know what the min/max temperatures are, you can figure out the volume change, and then plug that into the Boyle's law calculator to see what the pressure change will be. Note, that assumes constant temperature, and you can get more accurate if you include that, but that should give you the lower bond of pressure change.

You mean assuming all the compression and expansion occurs in the tank?

 

[DIYorBust]

I have no experience but recently installed my first hydronic system (air source heat pump) and have wondered about the pressure vs temperature behavior of a closed hydronic system. So here's a simplistic analysis from first principles, corrections welcome:

For the pressures involved, I believe treating water as incompressible is accurate enough. If you raise the water from 20C (68F) to 75C (167F), then a random website tells me the volume increases 2.4%.

If the closed system is made of all steel components, then steel has a linear coefficient of expansion of around 1*10^(-5) / degree C. The volumetric coefficient of expansion is 3 times that, and so for a 55C change in temperature, the components should expand in volume by about 0.16%. So let me ignore that volume change. [If much of the volume of the system consisted of plastic pipes, the effect would be an order of magnitude larger.]

There's air in the expansion tank, which is rising from 19 psig (34 psia) to 30 psig (45 psia). That means the air volume has changed by a factor of 34/45 = 75%. I.e. from the 20C start, the air volume has gone down by 25%, while the water volume has increased by 2.4%.

I conclude that at 20C, the volume of water in your system is about 10 times the working volume of air in your expansion tank (25%/2.4%). Is that at all close to being true? If so, seems like the pressure increase you are seeing is within expectations.

Cheers, Wayne

Wayne, one thing another poster mentions is that air bubbles or pockets in the water loops could compress and their pressure could increase when heated.

However the real question is, given what you are saying, do we think a bigger tank would solve the problem? Because an increase of 12 psi per heating cycle means either the tp valve will blow when hot or the radiators will absorb air and become airbound when cold.

 

[Jadnashua]

There is no downside to having a larger ET. A larger ET will lower the system pressure changes. The difference in price isn't much, either unless you get excessive.

 

[wwhitney]

Note, that assumes constant temperature

Good point, I inadvertently assumed constant temperature. So while the pressure change is a factor of 75%, the temperature change (293K to 348K) is a factor of 348/293 = 120%. So the total volume change should be a factor of 90%, i.e. a 10% reduction. Compared to a 2.5% increase in water volume, that would mean the starting water volume is 4 times the starting air volume.

[Don't trust this answer very much.]

Wayne, one thing another poster mentions is that air bubbles or pockets in the water loops could compress and their pressure could increase when heated.

Yes, the air volume in the above analysis is sum of the volume of air in the expansion tank and the volume of any undissolved air bubbles in the system.

Cheers, Wayne

 

[Jadnashua]

In an ideal system, with a working air extraction system, and, over time any oxygen will get trapped in rust, there should not be any gasses in the system. The 'air' in the ET shouldn't touch the water, so is just the buffer. If you want to get more exact, PV=nrT, but for this, it's not worth the effort.

 

[DIYorBust]

I finally got a chance to work on this today, as it was warm enough not to worry too much about shutting off the heat for a couple hours. I drained down the system and removed the Amtrol ex-30 tank and tested it and charged it to 22psi. I could not find anything wrong with it so I decided to reinstall it along with the new ex-60 tank I was planning to replace it with. I also charged the ex-60 to 22psi. The original installer had not made any effort to support the tank, so I made a little shelf the to rest both tanks on and plumbed them in. So I know have 2-3x the expansion capacity.

I refilled the system, and it had been at least 5 hours since I shut it down, so I'm certain all components had time to cool. I adjusted the PRV and the boiler read 21.7 psi. After waiting 15 minutes or so to make sure the pressure did not increase further, I plugged in my gauge to the bottom of the dirtmag, which is below the boiler, and it read about 22psi, the dial is not accurate enough to get a better reading than that.

So I turn on the thermostat, and as soon as the boiler starts up and suddenly reports a jump to 29.5 psi. The gauge at the dirtmag reports no increase. I let it run, and eventually, when the water temp reaches 180 the boiler is reporting 33 psi, and the gauge about 28 psi. However the T&P valve did not blow, at least so far.

What do you pros think? Is boiler pressure sensor wrong? Did it make any difference that I added the bigger tank? Can I leave it like this? Did I disprove Boyle's Law?

Any thoughts appreciated!

diy

 

[Fitter30]

40' in height from boiler to heat emitter? Take a pic of the boiler name tag expansion tanks ,pumps ,boiler so we can see how its piped. Have never seen a residential boiler good for over 30 lbs. 22 lbs is 50.8', 20 lbs is 46' more than enough water.

 

[John Gayewski]

There should be no pressure change at the dirt mag. This is normally near where the expansion tank is connected to the system. The point where the expansion tank is hooked to the system is actually called the point of no pressure change. The pressure can't change there. This is no suprise. Hydronic systems are dynamic you'll have different pressure readings at all different points in the system. Boyle's law stands.

 

[DIYorBust]

There should be no pressure change at the dirt mag. This is normally near where the expansion tank is connected to the system. The point where the expansion tank is hooked to the system is actually called the point of no pressure change. The pressure can't change there. This is no suprise. Hydronic systems are dynamic you'll have different pressure readings at all different points in the system. Boyle's law stands.

That makes sense, but that is also where the boiler connects to the system. So why would the boiler sensor report a 5psi increase, but not the gauge at the dirtmag? Perhaps from the primary pump starting up?

 

[John Gayewski]

That makes sense, but that is also where the boiler connects to the system. So why would the boiler sensor report a 5psi increase, but not the gauge at the dirtmag? Perhaps from the primary pump starting up?

The pressure in the system will depend on where you measure it in relation to the pump, yes.

 

[DIYorBust]

The boiler is rated for 30 PSI, so if it's really experiencing 33 psi that would be over the spec for the boiler, but I can't find a point any closer to the boiler to measure the pressure.The dirtmag is indeed very close. The purge valves always read the same as the dirt mag. Do you think it's actually 5 psi higher inside the boiler? Do I need a much bigger tank like one of those huge floor mounted ones? The tanks are close to the dirtmag, so I'm thinking the tanks aren't maxed out

 

[DIYorBust]

I think I see the problem. The pump is pointing directly into the boiler and when it runs increases the pressure by 5 or 6 psi, and sometimes blows the t&p valve. It's a 3 speed pump and lower the speed on the primary loop pump does reduce the pressure a bit. Will it reduce my heating capacity or cause problems if I turn it down to speed 1?

 

[John Gayewski]

I think I see the problem. The pump is pointing directly into the boiler and when it runs increases the pressure by 5 or 6 psi, and sometimes blows the t&p valve. It's a 3 speed pump and lower the speed on the primary loop pump does reduce the pressure a bit. Will it reduce my heating capacity or cause problems if I turn it down to speed 1?

It will reduce the heating capacity.

Why don't you go to a lower pressure? 15 should be enough.

 

[wwhitney]

Why don't you go to a lower pressure? 15 should be enough.

If the highest emitter is 40' above the boiler (= 17 psi) and the pressure at the boiler is 15 psi gauge, wouldn't that make the pressure at the highest emitter -2 psig? Would that be a problem, or is that fine, as it would still be 13 psia.

Cheers, Wayne

 

[John Gayewski]

If the highest emitter is 40' above the boiler (= 17 psi) and the pressure at the boiler is 15 psi gauge, wouldn't that make the pressure at the highest emitter -2 psig? Would that be a problem, or is that fine, as it would still be 13 psia.

Cheers, Wayne

Yeah I didn't scroll back that far. Generally you'd just figure the psig. I wasn't considering the closed system pressure, but that does raise an interesting point.

Either way he's starting out too high and finding out why it's not a good idea to start out high or pump toward the boiler conventionally.

I'm also kind of doubting thr 40' elevation.

If I was having this problem I would just get the water to the top and have someone up there with a phone to report when I dialed in the fill valve. Then just crack a little more pressure into it. Set the expansion tank based on that.

 

[wwhitney]

I'm also kind of doubting thr 40' elevation.

Yeah, seems to imply a 4 story house with the boiler in the basement. Or a 3 story house with 12 foot ceilings, including the basement. Or maybe ceiling radiant on the upper floor of a 3 story house with normal ceilings.

Cheers, Wayne

 

[Fitter30]

Wheres the pictures? All the gauges read within 2 lbs of each other with system off pumps not running.

 

[DIYorBust]

OK, I'll finally got the picture resized and here it is.

The house is a 4-story rowhouse with the boiler in the cellar, so yes it's approximately 40' and this is consistent with the plans and surveys and a reasonable visual inspection. Now I would say a PSIG of -2 is not okay based on experience, because the radiators become airbound. This is not a calculation, but an observation. With the pressure it 22 cold at the boiler, the top floor radiators stay hot. I'm not sure if the minimum needed is 20, but it's not 17, at 17 the radiators fill with air. I'd say a pressure of 22(cold at the boiler) keeps the radiators hot, it's hard to test exactly, in part because I have to climb 5 stories to observe conditions at the top.

Do you think it would work to simply turn the primary loop pump upside down and pump away from the boiler?
Edit: I see this would not work. Could I move it to the flow side and pump away? Any other options?

To be clear, I did not do the install or choose the pump locations, it was a professional install by a licensed plumber who is an authorized navien dealer and has done many installs including commercial jobs like small hotels etc. However my guess is those have the piping spec'd by an engineer. In fact I requested a different configuration, but he said this would pass inspection and everything is fine.