Expansion tank question

1daveman

I would appreciate some help with my expansion tank. I hope this is not a dumb question, but I have a radiant floor system that has a 70 gallon water tank that is heated with solar batteries. The water flows from the water tank to a propane tankless. If the solar batteries do not heat the water enough, the propane tankless heats it up. The water then flows through a valve that is opened / closed by the controller based upon the thermostat. (the controller opens the valve before turning on the pump, then closes it when the pump turns off). The water then goes through a manifold, through the floor loops, and back. It then goes through an air eliminator that has an expansion tank installed on it, then through the pump and back to the water tank. See attached schematic.

Question. The pressure increases in the hot water tank as the temperature increases, unless the valve that is controlled by the controller is open. If that valve is closed, the pressure increases beyond the pressure relief valve. If I open that valve manually, the pressure equalizes and does not increase with higher temperatures. I cannot just keep the valve open because the controller shuts it when the pump is off…

Should we install the expansion tank elsewhere? I would appreciate any suggestions.

Dave

PC7060

The closed valve and the pump IFC are blocking the heated boiler water from reaching the expansion tank.
Reconfigure to pumping away configuration; ie propane water heater → expansion tank with air eliminator → pump → floor loops.
Not sure why you have the valve on a single loops. The pump tied to thermostat call for heat is all you need for temperature control.

1daveman

Thanks - that is easy enough and I really appreciate the direction.

Before I ask my plumber to do that, I would really like to understand why that is a better design. Would you be able to point me to a reference or something that would help me understand the reasoning behind it? If I understand it, I will be better able to handle ongoing issues if they arise.

Thanks again!

Dave

EBEBRATT-Ed

I agree with @PC7060 that you probably have a check valve in your pump (most newer pumps have them. When the valve closes and the pump shuts down water is trapped between the valve and the expansion tank. If residual heat or the propane tankless or solar batteries heat this water while the pump is off and the valve is closed the pressure will rise.

Don't see why you need a valve either unless it is to prevent gravity flow.

Before doing anything I would make the thermostat turn the pump on and leave the valve in the open position and see if that solves the problem.

What controls the heat in the loop? Does the solar and propane work independently from the thermostat and maintain temp all the time?

GroundUp

It's better because as it is, the expanding volume of water is disconnected from the expansion tank by the closed valve and the IFC. The expansion tank can not "see" the expansion because it's separated from it. As for the valve, it serves no purpose. The circulator alone should be switched on with a heat call and initiate flow. When the heat call stops, the circulator and therefore the flow also stop. It'd be simple enough just to manually lock the valve open and leave it like that, then everything should function appropriately without cutting anything apart.

1daveman

EBEBRATT - excellent thanks. I did not realize the pump would have a check valve, so with that then I see the expansion tank would not do anything if the pump is off and the valve closed.

The only reason the valve is there is because I could not figure out how to wire my Taco controller to turn the pump on when the thermostat indicated without opening a valve first, so we added the valve. It does not do anything else.

The solar heats the water in the tank to a set temperature when the battery capacity allows for it, independent of the thermostat. The thermostat turns on the pump when the room temperature is below the set point. If the water in the tank is hotter than 120F, the tankless allows the water to flow but does not heat anything. If the water in the tank is less than 120F, the tankless heats it up to 120… That part all works really well.

Thanks again

1daveman

Groundup - thanks so much. That makes sense and I should be able to just lock that valve open. Your explanation really helps.

hot_rod

Do you have solar thermal collectors heating the tank? Or an electric element running off a PV array?

When the tankless runs, you will heat the 70 gallon tank to the return water temperature from the radiant. So you limit some solar potential by warming the tank that way.

To maximize the two heat inputs you could run off the solar tank until it reaches a low temperature, say 90°. Then bypass the tank and heat with a tankless. Although a tankless is not an ideal radiant heat source.

Although unless you have a very small space, 70 gallons will not supply a lot of heat energy.

1daveman

Hot_Rod

I have electrical elements that heat the water when the batteries are fully charged, and an Aquastat that turns that off / opens a relay, when the water gets to a set temperature. I plan to set the aquastat to about 150 - 160 degrees. I understand that is not a lot of heat capacity, but wanted to use the extra PV capacity since I have it. While we are at 5000 ft elevation we live in Arizona so the cold temps are not too severe. We have about 1500 sq feet o the ground floor. I will look into your suggestion of bypassing the tank when it gets down to 90 or so. Thanks!

PC7060

https://forum.heatinghelp.com/discussion/comment/1829184#Comment_1829184

agree with @EBEBRATT-Ed and @GroundUp - no need to reconfigure since you already have pump configured for pumping away from the expansion tanks. Air removal would perform better at highest / hottest point in system but not worth the bother; just lock the valve open.

hot_rod

can your pv system heat your domestic hot water also ? That is a good load for solar as you use hot water most days

Sun Bandit is sn example of PV direct to an elements in a tank. No need to involve you battery bank this way

https://sunbandit.us/how-it-works/

GroundUp

Assuming you have a Taco ZVC controller with 4 terminals per zone valve, I believe you can run a jumper between 3 & 4 (as a 2 wire valve) which would turn the circ on when the thermostat calls. Or swap it out for a switching relay such as an SR501.

1daveman

Hot- rod - Yes my hot water tank has a coil that runs through it with the DhW - that helps us get better use of that excess PV.

Thanks!

1daveman

Ground Up - That is the controller I have. I was just getting out the user manual to try to figure out how to skip the valve - so if it is as simple as aa jumper, that should be straight forward. I appreciate the suggestion.

You guys have really been helpful - thanks so much.

Dave

GroundUp

@1daveman take the cover off the controller and there will be a diagram inside the cover. I'm pretty sure you can just run a jumper between 3&4 to skip the end switch and still operate the circ, but I also believe you could disconnect the valve altogether and lock it open- just using the circulator function.

1daveman

Ground up.

I did that this morning, I removed the wires going to the valve, connected 3 to 4, and it worked like a charm. Easy!

As soon as my solar batteries charge up this morning, I will check to confirm, but I expect the pressure will be fine as the water heats up.

Thanks again

Dave

1daveman

I thought I would return and post the results. I jumped 3 and 4 in the controller, manually set the controller valve open, and heated the water today.

The pressure did still rise, but not nearly as much as earlier. I think at some point in the future I will move the expansion tank to be closer to the hot water tank so it can be more effective, but at this point all is good.

Thanks again everyone.

EBEBRATT-Ed

Moving the expansion tank nearer the heat source will not do anything. It is normal for the pressure to rise when the water is heated. No one sizes an expansion tank for "0 pressure rise" you would need a huge tank that would be $$$$$.

The question is how much does the pressure rise and what is the system designed for?

In a typical house HW heating system for example the cold fill pressure is usually 12-15 psi. The boiler usually has a 30 psi relief valve.

An expansion tank would be selected to allow the water pressure to increase as it is heated from cold (say 50 deg) to hot (say 190 deg). In this example the water pressure may rise to 20-25 psi. Personally I would like to stay at 20-22 max. To me 25 is getting to close too the 30psi relief which may open at less that 30 psi. They are not a control valve just a relief valve and are not 100% accurate.

Your design may require something different.

But if your pressure rise is not going over 80% or so of the relief valve setting you should be ok.

1daveman

Oh shoot. My pressure rose more than that. My relief valve is at 75psi. The pressure in the system is set to about 40psi with the water temp at 78. The expansion tank was preset at 40 psi.

The pressure rose to about 55psi when the temp got to 105.

Right now the expansion tank is before the pump, and after the tankless and all the loops in the floor. Since I was told the pump had a check valve, I thought I should put the expansion tank right where the water exits the tank so it sees the pressure...

I would appreciate suggestions.

Thanks

hot_rod

The higher the air pre-charge that you set, the less expansion capacity.

Using the Amtrol tables for Critical Sizing Method

You have a 70 gallon tank, misc piping and floor loops, lets just estimate 85 gallons of water total. If you know exactly, shout it out.

If it goes from 60F to 180F use the factor .02708 x 85gal. =2.3 gal expanded water divided by

acceptance factor 40 starting pressure to 70 psi max. allowable, the factor is .354

2.3 divided by .354= 6.5 gallons of expansion tank acceptance

So 3- #30 tanks, or 1- #90 are needed to contain the pressure increase.

Screenshot 2024-12-28 at 9.56.12 PM.png

Screenshot 2024-12-28 at 9.53.46 PM.png

Screenshot 2024-12-28 at 10.01.35 PM.png

hot_rod

Take the fill pressure to 10 psi, set the tank pre-charge to 10 psi

2.3 gallon divided by .708 (10 psi to max 70 psi) = 3.2 gallons of acceptance volume

2- #30 tanks would cover that

1daveman

Wow, I am going to study that so I can understand how to read those charts. Very helpful.

When I dropped the psi that low, my alpha pump flow volume really decreased, which is why I have it higher. Should I just let it run and see if it increases? 2 #30 tanks would be better than 3!

PC7060

I use a pair of 30# tanks for my converted gravity system. The wall mount hangers make for clean installs.

Jamie Hall

https://forum.heatinghelp.com/discussion/comment/1830490#Comment_1830490

Wait wait! Why are those pressures all so high? Most heating boilers have a maximum pressure rating of 30 psi to begin with, and relief valves to match. What you are quoting sounds much more like a domestic hot water system?

30 psi is adequate pressure for a six story building.

There's more to this puzzle than we are seeing, I think.

1daveman

I don't have a boiler, instead I am using PV electric to heat water in a tank. The water goes through a propane tankless before going into the floor. The propane tankless is used only for back up when the PV does not sufficiently heat the water.

Does that make sense?

hot_rod

one other thing adds pressure to a closed loop system, that is the circulator. What model circ do you have? Some tankless water heaters use a high head model, which could add 10 psi or more when it runs

this system isn’t a combined heating and hot water? The water in the tank is not used for the home, potable water?

The static fill pressure should not have any bearing on the circulators ability to move flow. The same gpm will circulate at 10 psi or 40 psi

You can run 40 psi as long as the expansion tank is sized and pressurized correctly

Lower psi= smaller expansion tank

1daveman

Thanks

I am using an Alpha 2 pump. When everything is working correctly I get about .4 gpm in each of my 6 loops, (~300ft of 1/2 pex) at about 40 psi. The tank has coils for the DHW, but the floor water is separate

At 20 psi the flow is reduced maybe 20 percent for some reason. maybe I am doing something wrong there.

I need to get a larger expansion tank anyway, so think I will just do that and keep things going, as long as that sounds OK

Thanks again

Jamie Hall

I won't comment on using solar PV to heat water. Except to note that direct solar is close to 100% efficient, while solar PV is 20% efficient in terms of heat gain per square foot of collector.

Now. First, using the high static pressure in your system is just insane. There is no reason to do that, except stress the components. And it, by itself, has no bearing on how much the pump can pump, since the pump cares about pressure differentials in to out, not static pressure. Something else is happening there.

Now there is one possible rather serious problem: tankless water heaters have remarkably high head losses in comparison with purpose built boilers, and some have flow volume restrictors as well. If you are measuring you pressure at the expansion tank, which is at the inlet to the circulating pump and the outlet from the tankless heater, however, that should not make a difference either — but the pressure at the outlet of the circulating pump or the inlet of the tankless may be much higher, due to the head loss through the tankless — and that indeed could be 20 to 30 psi higher, depending on the overall head losses and flows.

1daveman

Thanks

I understand the PV heating element is not nearly as efficient. Since it is only used after my batteries are fully charged, I would lose that power anyway. I am using power I would just throw away, which is why I thought it made sense.

The only place I am measuring static pressure is the gauge on top of my tank. Since you point out that can cause unneeded stressed on the system, I will reduce that static pressure and see how it goes. It has been a while since I tried it at a lower pressure. It might be that when I had a lower pressure something else was reducing the flow, and I was mistaken

Thanks for the suggestion.

hot_rod

What brand and model is the tankless? I'm surprised the Alpha has enough "fizz" to pump through that and the loops.

If the floor heats well, then the .4 gpm is great. Common 1/2" loops design around .5-.65 gpm.

Here is a graphic of how a circulator causes fluid to move. I'd expect the numbers are close to what that Alpha could add.

The other graphic shows how that energy the circulator adds gets used to overcome the resistance of all the components in the loop. The circ adds 9 psi to the static fill of 10 psi. So the first, closest to the pump, gauge would read 19 psi, dropping as you go around the loop.

The pressure changes throughout the loop. But never drops below that static fill psi.

I'd guess the pressure at the tank would be 6-10 psi higher than the fill pressure with the pump running, cold system.

As the water heats the expansion will raise the pressure throughout the entire system as you have noticed.

40- 70 psi is no sweat for any of your components, probably most are rated to 150, except the tankless? It just will cost you more expansion tank capacity

If the system works best at 40 psi fill, so be it. Just watch the pressure as it heats and add expansion capacity as needed.

Screenshot 2024-12-29 at 1.45.53 PM.png

Screenshot 2022-12-13 at 8.43.11 PM.png

1daveman

Interesting charts. They make sense and help me understand what is happening. The tankless is a Navian. I am not sure which model and am not at my home for a few days to check.

So far the floor has been awesome. I am at 5000ft in AZ, so the temps are not extreme, but we have had lows in the the mid twenties and it has worked wonderfully. The floor pex loops are in 6in concrete.

Thanks again for all the advice

hot_rod

Here are the 3 steps to calculate the tank size. If you have some actual numbers to put in, try some sizing. Maybe I was a bit high with the 180° max. temperature? Or starting temperature is maybe higher than 60°?

The biggest guess may be the system capacity? Figure a gallon for every 100' of 1/2" pex.

But I don't think you will pull it off with a single #30. Cheapest and easiest is to just pipe two together.

Hard to beat a radiant floor for clean, quiet, comfortable heat.

Screenshot 2024-12-29 at 3.27.52 PM.png

Screenshot 2024-12-29 at 3.28.07 PM.png

1daveman

Thanks

I did some calculations using 70F to 140F, 90 gallons (1800feet of 1/2 pex + 70g allons), and pressures from 40 to 70 psi.

I got .01352 x 90= 1.2196 net expansion

.354 acceptance so 3.44 gallons of expansion.

Looking at the chart, unless I am reading it incorrectly, #30 is 2.5,

#60 is also 2.5? that seems odd…..

Regardless I should be able to get by with 2x 30#.

Does that all sound about right?

Thanks again!!

Dave

hot_rod

The # 60 tank has the diaphragm crimped in at the same location. However it has a much larger air space, so it will have a slower pressure increase.

The tank manufacturers don’t want to see the diaphragm stretched more than @65% That is why the acceptance is lower than the tanks actual volume.

Undersized tanks fail quickly for that reason, the diaphragm is over stretched. Most folks blame the tank when they fail. I suspect many tanks are undersized, installers too lazy to use the 3 steps to properly size the tank. Or use a cheat sheet that sizes by just btu of the boiler😞

You can buy full acceptance expansion tanks, they have a bag or bladder inside instead of a diaphragm, so they can accept close to the tanks actual capacity. They tend to be more $$

Your math looks correct, but do you want to limit the tank to 140?

The hotter the tank , the more “free” energy you have harvested.

Id run it to 180 as long as the tankless can tolerate that temperature running through it?

1daveman

That is a good point about the 180 degrees. My Navian has a max temp of 180, so I don't think that will be a problem.

With that high of a temperature I assume I would need a mixing valve to mix the hot into the manifold with the cold return, so the floor does not see that high temps? Is that correct?

When this was installed I purchased and provided a mixing valve for my plumber, but he literally refused because he said I did not need it. I have a new plumber who also does not know much about radiant floor heating - do you think I should ask him to install a mixing valve if I go for that high of a temp?

( I am in a remote area so plumbers are hard to find!)

hot_rod

are you running 150 to the radiant now? If it is a slab, that is way too warm. If it is a slab, rarely do you need more than 100f supply.

For two reasons you want a mix valve on the radiant. To supply the accurate mix temperature, and to extend the use-ability of your buffer.
If you ran the tank to 180, and could use it down to 100, that is a huge 80 degree delta to leverage.

You paid for the PV, try to use it to the max. Arizona must have close to 300 days of sunshine?

Do you get$$ or credit from the electric utility for feed back? What do you do with excessive PV in summer months?

To properly pipe that system with a mix valve and better tankless flow would be a major repipe. Two circulators needed. Especially for a novice plumber.

You might just sort out the expansion issue and run with what you have?

Here is how I would pipe it.

Either the solar or the Navian would heat the tank, or both if needed. The tank is all they see as their load.

The Nav gets its own pump to get full flow, full out put.

The radiant just pulls from the buffer tank via the 3 way valve and its own pump. You don’t need or want the solar heated water flowing through the Nav, as it becomes a small cooling tower😎

IMG_0903.jpeg

1daveman

Wow - that really changes the lay out. I understand now why that would work better than what I have now. I am not comfortable with my current plumber taking that over, but I am going to keep that for future reference, in case I find someone who can do that. I assume the Nav would be controlled by a thermostat in the buffer tank?

There is no grid where my home is, so no one to sell the excess power to!. In the summer I have a couple of minisplit a/c units that use power, and also heat the tank for the DHW. Currently I have a Nav filter in front of the tankless that has a max temp of 100f - so until I get rid of that I am using primarily the Nav. I recently learned about the DirtMag, and ordered one to replace that filter so I can increase the tank temps.

I think I will get the #90 expansion tank for the possible future change in the design, and take it from here.

Thanks so much for all the advice.

Dave

1daveman

I hope you don't mind me asking one last question. As I mentioned earlier, I am not comfortable asking my plumber to do the redesign you suggested, at least at this point. However I was wondering if it would make sense to add a mixing valve to my current system, so I can increase the temps?

I attached a schematic of where I am thinking I would put the mixing valve, but I am afraid that the water would just flow back to the pump rather than through the valve. Or should the valve be in front of the pump? Or is there an alternative that would work? Or should I just post this as another discussion? Or just try to get my plumber to implement the redesign?

I would appreciate if you don't mind providing some additional insight.

Thanks again

20241223_191503 v2.jpg

hot_rod

To do it correctly? You need a pump pulling out of the 3 way valve. That allows it to pull some hot and pull some cool return into the valve to blend the desired temperature.

So if you set the valve to 100°, the tank is sending 160°, the low return temperature from the radiant, call it 85° blends with the 160° tank water to get you 100°or whatever you set the mix valve to.

So you need a second circulator to make all this work properly. You cannot pump into three way mix valves.

Sorry, there is no wrong way to pipe this right.

I would think any plumber that can read a drawing should be able to connect that loop, throw in two tees 8" apart, a circulator and the mix valve?

Whatever control turns on the pump you have now, wires to both pumps.

OR, if you pipe it like my previous drawing, the thermostat only turns on the radiant pump. The radiant pump only sees, only needs to see the buffer tank. It doesn't care what is heating the tank.

The boiler pump would only need to run when the tankless is fired. That is how you would use the solar heated tank without involving the tankless.

So you have bad, better (attached drawing) or best options (previous post drawing) :)

Screenshot 2024-12-30 at 6.27.28 PM.png

1daveman

Thanks - I will review better and best with the plumber, and see how he feels about it.

I see you are with Caleffi, so I will make sure I purchase the mixing valve and anything else I can from them.

Take care

Dave

Kaos

Here is how I would plumb it with a single pump.

You can set the tankless to fire a bit bellow the mix valve outlet temp, this way it will only ever run if the buffer tank runs out.

The nice part is you don't need any extra controls, wire the circ to your thermostat and all heat is managed automatically. This also means the return water temperature of the tankless is pretty low and it will be condensing all the time so you'll get pretty high efficiency out of it.

The one drawback of this layout is you are using one pump to push through a tankless and your emitters, so overall a pretty high pressure loss. If your loads are low enough, it will probably work with a single ECM circ. If you need more flow, the high head ECM circs tend to be expensive, you are probably better off with two 007E in series.