Heat on inactive zones when separate zones active

Smf_000

Hi,
I recently split one main hot water radiant heating zone into 4 new zones with zone valves. The new main zone has 4 convectors and the 3 additional zones each have one convector (bedrooms). I recently did this work and when testing the system, I noticed that when one or more zones are active an inactive zone will have some heat at the radiators. I put the zone valves on the return on advice from a plumber friend. The main zone uses 1-1/4" pipe while the 3 others are 1/2". All zones/radiators are branched off of one main supply line but the returns are separate now and go to the zone valves, to a manifold and single circulator. I've included a diagram of my system below.

Any advice on why the other zones are getting heat would be greatly appreciated.



Thanks,
Steve

pedmec

Because you have one flocheck for multiple zones. Although only one zone is getting a forced loop thru on call for heat they are all sharing the same supply after the flocheck. you have hot water common to all zones. one or two things are happening and they both could be happening together. your getting ghost flow traveling thru the supply lines or your getting conduction thru the water on the supply. according to engineers you can get ghost flow thru the same pipe depending on the size. Heat seeks cold, warm water rises.

EdTheHeaterMan

the piping design of the supply and returns that are indicated in the red will determine if Ghost flow is your problem. I believe it is. On pager 12 of Zoning Made Easy is an explanation of the phenomenon. Colder water and hotter water move in opposite directions in the same pipe. http://media.blueridgecompany.com/documents/ZoningMadeEasy.pdf

EdTheHeaterMan

To verify the Ghost Flow I would need to look at the actual pipe location and elevation as built. Basically the pipes not on your diagram. the ones between the 2 red boxes in the diagram.

Those pipes will tell the real story.

Did you read about Ghost Flow. We get a lot of that around Halloween. Kinda' Spooky

Smf_000

Thanks all for the replies.

I agree with the assessment of ghost flow since the inactive zones are cooler than the active ones. Having a main pipe to feed all the radiators with some on separate returns seems like it would be reasonable to assume heat transfer to the water/pipes on other zones.

Here is an updated diagram showing the radiator locations and pipe sizes.

Any idea how to correct for this without having all the zones with separate supplies and circulators? Keep in mind the house was built in the 50's and getting access to the piping isn't easy. The main reasoning for doing this is I had refurbished the 3 bedrooms and 2 of them upstairs were always cold being all on the one zone so I figured I would separate them using zone valves based on advice from an experienced plumber friend. I guess it's all about the details.

Smf_000

BTW, the link you posted doesn't work. I did find another source though. https://www.xylem.com/siteassets/brand/bell-amp-gossett/resources/technical-brochure/fh-z100b-bg-zoning-made-easy-2.pdf

hot_rod

Move the two circs to where the flow control valves are, circs with integral checks. Remove the l flow control valves and use them as boat anchors🤓
this gives you 100% shutoff on returns via the zone valves, ghost flow protection with the checks in the circs 

Two bonuses you will now be pumping away instead if towards the expansion tank. And removing any weighted type flow checks reduces head in the circuit = more flow

Id use a delta p circ on the multi zone valve circuit 

Smf_000

Thanks for the replies Pedmec, Ed and Bob.

Forgive my ignorance on the subject matter and questions. Relating to Bob’s reply, how will moving the circs with check valves help ghost flow on my multi-zone circ path? Isn’t it still pushing water/conducting heat on the supply to all zones before the ZVs? I would guess the CVs would help between circ1 and circ2 paths.

I did a little reading on a delta P circ but not sure what the advantage is. Seems that in Europe they run this type all day long but it can be run on-off like we do here. Please help me understand the need for this.

More reading shows that putting the circs on the supply side with CVs along with expansion tank on the hot side is the preferred method but I’m working with an old existing system with circs on return and expansion on the cold inlet side for good or bad. I’m not looking to correct the entire system as it’s worked for the most part with the exception of previously having cold radiators upstairs on the same zone as the first floor (bad design in any book).

Would adding CVs before every radiator help the ghost flow even though it’s still trying to push hot water one way on the closed paths?

hot_rod

The circulators with internal checks prevent froward flow when not running. The zone valves on return prevent flow in either direction.

With 4 or more zone valves on a standard circ, you should have a pressure bypass, or a delta p circ that changes output flow as zone valves open and close. Either method assures proper flow regardless of which zones are open or close.

Some graphics of what that looks like.

mattmia2

I would think you need an individual flow check after each suppl branches off to the separate zone.

I think the flow is hot water rising and falling within the same pipe, not through an entire circuit as most ghost flow is.

You could try using the isolation valves in each zone as balancing valves to get them to all heat at more or less the same rate so that it wouldn't be running a long time trying to heat up a cold zone after some zones were satisfied and overheating a hot zone, maybe if you can get it balanced the bit of extra heat from the ghost flow won't cause overheating.

Er, never mind, if your isolation valves are gate valves, you can't use those for balancing, only ball or globe valves.

Smf_000

I think I understand. So, with my current setup, if I have only 1 zone with the 1/2” return active, with my current single speed circ, could I achieve over pressure and blow out water from the relief valve? Will my current setup cause any damage?

Can I put a delta p circ on the return side in place of the old one? Can I keep the second circ where it is regardless of where the new zone 1 circ goes? Can I keep my ZVs on the return side? Sorry for all the questions.

I’m not being lazy or cheap, I just think it’s more work than I can get done before the weather gets much colder.

mattmia2

I guess the other possibility is your circulator in that configuration is exceeding the pressure rating of the zone valves when only one or 2 zones are calling and a delta p circualtor would solve your problem. See if the zones still heat if you close the isolation valves. If they do it is the zone valves not holding under the pressure. What model are the zone valves?

Smf_000

Taco 570 for the 1/2” lines and 573 for 1-1/4” line. What do you mean by isolation valves from my drawing?

I just turned on my upstairs Master BR zone and am getting heat in that radiator but also in the 1st floor 4 radiators (just as hot). Other bedrooms are cold as they should be. Tested ZV on 4 rad line and end switch is open. And the 1-1/4” return pipe for that zone is cold.

4 rad zone return pipe is cold, ZV is closed, rads are hot, does that indicate ghost flow?

Mattmai2, my diagram is wrong the gate valves after the ZVs are actually Pro-pal ball valves with drain.

Also, I put gate valves in my drawing for all the old valves but I’m not actually sure if they’re gate or globe. Don’t know the difference visually.

Smf_000

Question about comment “ See if the zones still heat if you close the isolation valves. If they do it is the zone valves not holding under the pressure”

Do you mean if the isolation valve is closed and they still heat it doesn’t necessarily mean the zone valve is leaking but rather it’s an issue of ghost flow since the path is completely closed by ZV and iso valve.

What is the different between a flow check and check valve? Which should be used on a branch loop?

hot_rod

Smf_000 said:

I think I understand. So, with my current setup, if I have only 1 zone with the 1/2” return active, with my current single speed circ, could I achieve over pressure and blow out water from the relief valve? Will my current setup cause any damage?

Can I put a delta p circ on the return side in place of the old one? Can I keep the second circ where it is regardless of where the new zone 1 circ goes? Can I keep my ZVs on the return side? Sorry for all the questions.

I’m not being lazy or cheap, I just think it’s more work than I can get done before the weather gets much colder.

Look at it more from a gpm standpoint. Suppose with all zones open you are moving 10 gpm. Everytime a zone valve closes the pump is still trying to move 10 gpm. So flow keeps increasing in the remaining open zone. when you get down to only 1 zone, the circ is still moving at the same speed, so you can end up over pumping, excessive flow and velocity= noise and tube wear.

The bypass valve of variable speed circ just flattens the pump curve to prevent over pumping.

The bypass valve just diverts a portion, a parasitic device.
The variable speed circ does it by reducing the gpm, so you save electrical energy also.

Common zone valves can close off against 8,10 up to 75 psid, so I doubt a standard residential 1/12 hp circ would push past a closed zone valve.

The best fix depends on how much piping change you want to do to assure you get it fixed on one attempt.

Smf_000

Bob,

I’m not opposed to changing the system to make it work properly but maybe not to all current practices.

1) Sounds like a delta p circ would fix the flow across multiple zones. Can it be on return to keep my current piping? I know you said supply is better for reasons you stated earlier.
2) Need check valves on each supply branch for circ1 to fix ghost flow. I think I can do this easily with smaller spring CVs.

EdTheHeaterMan

Your diagram of the missing pipes is helpful. Is it a proper elevation (Side View) of the vertical piping design? This illustration shows the problem on the return on zone 2 when zone 1 operates.
Zone one may also have the problem to a lesser degree as the heat in the shared return piping can easily ghost up the other.

I believe that you may have the problem as a result of a hot horizontal larger diameter manifold pipe passing by those 1/2" branches, like this.

One simple fix would be to add a thermal trap to all pipes that leave the manifold in a vertical downward direction (about 12" on those 1/2" takeoffs) before turning to go up to the radiators like this.

Smf_000

Hi Ed,

Thanks for the help and diagrams. My diagram doesn’t really show the piping exactly. I have a center stair Cape with 4 main rooms on 1st floor and 2 BRs on 2nd floor. The boiler is in the basement on far side of the house. The main supply line starts with 1.25” pipe vertical from boiler to first 90deg and goes to 1” to cross the entire basement. All branches are 1/2” lines. The pipes for the two 2nd floor BRs go right up the outside wall at each end. All the branches for the 1st floor branch off the main supply, run horizontal to the radiator above then go vertical. All the returns for 1st floor 4 radiators, go the same route as supply with 1” pipe to ZV. New zones, MBR, kids BRs, use 1/2” returns back to ZVs. The vertical length of the 1st floor pipes for each radiator is only the distance of the floor joist ~10”. The 2nd floor pipes would be ~10ft.

Hope this helps.

I can’t really do the 12” vertical loop due to basement ceiling constraints.

EdTheHeaterMan

My understanding is that you have some horizontal pipes in the basement near the basement ceiling. From those horizontal pipes you have 1/2" branches that go 10" up to the first floor and 10' up to the second floor. Do I have that picture in my mind correct?

If so, then my proposal is to make those pipes to the first floor go down 12" then make a U turn Up 22" to the first floor radiators. for the second floor risers leave the horizontal pipe going down 12" then go up 11' to the second floor radiators.

Does that make sense?
That would be a thermal trap.


It would be very easy to cut the copper pipe and place sharkbite couplings on each end then add some oxygen barrier PEX in a 12" loop to the pipe

This is an inexpensive way to test one loop to see if that solves the problem before repiping everything at a greater expense.

Smf_000

Yes, you have that correct. Unfortunately I can’t do the 12” loop due to ceiling constraints in the basement. Also, pipes for the radiators on the external wall don’t really go vertical easily. They kind of jog over the sill plate then go vertical. This house makes everything difficult, lol.

hot_rod

Not much I like about that piping, if you want to stay simple I would add some hydronic spring checks. I'd like to fix all the glitches with one fell swoop..but if that is not an option.
Pumping at an expansion tank is my pet peeve and can setup a number of problems that never go away.To add insult you have a bunch of high pressure drop valves that can add up to a negative pressure condition when pumping at the PONPC.

I am not a fan of those weighted type flow checks, a lot of pressure drop added to the pumped circuit. 83' for a typical 3/4" size.

I put 6' deep thermal drops on a problem job once and it wasn't enough to solve a ghost problem. It is buoyancy dependent and a tall vertical line can win the battle of the ghosts. It's the basics of how gravity heating systems worked.

Spring checks are considered a bubble free seal, 100% shutoff.

As piped, a ∆p circ would not fix the ghost issue. It would help correct the over pumping, save some electricity.

EdTheHeaterMan

Are you saying that the basement ceiling is a finished ceiling?
I always assume that a basement ceiling is exposed floor joist and visible pipes and electric lines. Although I grew up in a home with a finished basement. OH... What parties we had there

Smf_000

hot_rod said:

Pumping at an expansion tank is my pet peeve and can setup a number of problems that never go away.To add insult you have a bunch of high pressure drop valves that can add up to a negative pressure condition when pumping at the PONPC.

Can you explain this to me more? What are the high pressure drops you refer to? What is PONPC?

hot_rod said:

I am not a fan of those weighted type flow checks, a lot of pressure drop added to the pumped circuit. 83' for a typical 3/4" size.

I know this system isn’t ideal but it has worked fine since built with the exception of the 2nd floor not being on its own zone. Again, not saying it’s right nor am I saying you’re wrong. I wish it was better but I’m not willing to scrap the entire system. Moving the circs isn’t easy but doable. CVs definitely doable on each of the 7 branches. I don’t get ghost flow between the 2 circs on the return side.

BTW, circ1 is a B&G NRF-22.

Smf_000

EdTheHeaterMan said:

Are you saying that the basement ceiling is a finished ceiling?
I always assume that a basement ceiling is exposed floor joist and visible pipes and electric lines. Although I grew up in a home with a finished basement. OH... What parties we had there

Yes, the ceiling is finished with a drop ceiling in most of the basement and too low in the unfinished portions. I grew up in a similar Cape type house with a finished basement also and did mine in a similar way. If my parents only knew the parties I had there. Lol, ah youth! I should have done all this work when I was a younger newly married man with a whole less crap to move to do all this work, 😂

hot_rod

Smf_000 said:

Pumping at an expansion tank is my pet peeve and can setup a number of problems that never go away.To add insult you have a bunch of high pressure drop valves that can add up to a negative pressure condition when pumping at the PONPC.

Can you explain this to me more? What are the high pressure drops you refer to? What is PONPC?

I am not a fan of those weighted type flow checks, a lot of pressure drop added to the pumped circuit. 83' for a typical 3/4" size.

I know this system isn’t ideal but it has worked fine since built with the exception of the 2nd floor not being on its own zone. Again, not saying it’s right nor am I saying you’re wrong. I wish it was better but I’m not willing to scrap the entire system. Moving the circs isn’t easy but doable. CVs definitely doable on each of the 7 branches. I don’t get ghost flow between the 2 circs on the return side. BTW, circ1 is a B&G NRF-22.

I guess it comes down to you getting a few opinions on fixes and upgrades, or learning all the fluid mechanics involved in why the fixes are recommended. That gets to be a very long thread, beginning with hydronic Fundamentals

If so download the 31?issues of Idronics, start with # 12

If you want the easiest fix. Add some hydronic specific spring checks, not more weighted flow check would be my suggestion.

If the system happens to have chronic air or noise, the more involved fix suggestions may be worth considering

Smf_000

Thanks for everyone’s help with this. I’ve learned a lot but still have more learning to go and will explore all options including moving my newly made ZV manifold to the supply side and disconnecting the separate zones from the main supply feed. Will also look into a delta p circ. If you have a suggestion on a model with correct specs to match my nrf-22 or system needs, I would appreciate that.

pedmec

Add flochecks on the beginning of each loop under the baseboard covers. While not ideal as the supply water before the checks can get warm it will prevent the warm water from reaching the fins and heating up the rooms not calling. And as bob has pointed out, adding a variable speed circulator works best with a zone valve system.

hot_rod

All the brands have delta p circs in that size, I use the Grundfos Alpha 15-55

Smf_000

pedmec said:

Add flochecks on the beginning of each loop under the baseboard covers. While not ideal as the supply water before the checks can get warm it will prevent the warm water from reaching the fins and heating up the rooms not calling. And as bob has pointed out, adding a variable speed circulator works best with a zone valve system.

All my radiators are the old style convectors and not baseboard. I may just move my ZV manifold to the supply side and eliminate the need for CVs unless I need them before the return manifold. Will look into the delta p circs on the supply side as well.

Was looking at my setup this morning and adding things to the supply side will be a bit challenging due to space and piping contraints but I need to study it further.

Got a lot of reading to do. It's interesting nontheless even if it's not entirely applicable. Always willing to learn.

ronbugg

If you have space on a wall you can repipe new circulator there where it would be easier to service Maybe install manifold there also and run 1/2” pex from there to across basement and pipe in to copper going to 2nd floor Ask Hod Rod to send you some pics of jobs upgrades, that will give you ideas of easier ways to upgrade supply system. If you do that, when you replace the boiler it’s a matter of connecting 2 pipes. The circulator and manifold could be prepiped and wired 

Smf_000

I wish I had the room to do that. I’ve seen many pics online and a neighbor’s beautiful and nicely organized setup. It would be in my nature to do that but you can see what I’m working with below. You can see the new ZV manifold I just added in the first pic. I’m exploring the possibility of 1) testing 1 CV on a branch to see if it resolves my ghost flow issue before doing all 7 or 2) moving the ZV manifold with a delta p circ to the supply side and feeding all the zones individually then redoing the returns to a manifold. Have considered PEX for the run across the house but the pipes closest to the boiler I can just solder in or crimp. Just need to consider cost of tools for a one-off job. If I can rent or borrow I’m all in.

Smf_000

Here's the piping layout

MikeAmann

Would these help? They are easy to install.
https://www.supplyhouse.com/SharkBite-U2008-0000LF-1-2-Sharkbite-In-Line-Check-Valve-Lead-Free

Smf_000

I did consider those but not sure how well they would work on existing lines. One side would have to pass through further to connect to the other pipe if both are fairly rigid. I’ll check out some videos to see how installation would work. Thx

Just watched a video using different methods with shark bites. With copper to copper you need a little flex in the pipe to pull it back a little to get one end in. Another way is to use a slip type coupling and this requires making 2 cuts in the pipe, one for the slip fitting side and another for the CV on the other side, then slide the slip fitting and cut pipe towards the CV. Can also use a section of pex. Using both pex and shark bite you need to bridge the copper sections with a bonding wire. All very interesting and possible.

I may go with a crimp style fitting this way if I get the tool I can use it for the future mods to the whole system

hot_rod

Cast iron radiators? A TRV on each would be a great upgrade and may even fix the ghost flow.

Make sure any valves you add are rated for hydronic temperatures 180F or higher. Most hydronic devices, circs, etc. are rated to 225F.

Smf_000

Nope, just the normal, half in the wall, steel covered convectors from the 50s.

I did read #31 of idronics. Good read and informative. I’ll get to the others in time. Not sure I’ll need to do head loss calculations but if I have nothing to do, maybe, lol. I looked into the alpha2 15-55 and the taco Viridian. Seems the taco is variable over 3 selected ranges vs. the alpha which is infinitely variable of the whole range which I like better.

I think after reading that guide, while not exactly specified, I can keep my basement circ2 on the return side and move circ1 to the supply side and still maintain circ separation effects if the common piping is the same or close.

MikeAmann

Smf_000 said:

I think after reading that guide, while not exactly specified, I can keep my basement circ2 on the return side no move circ1 to the supply side and still maintain circ separation effects if the common piping is the same or close.

This is what I was thinking also. Try it to see if it solves the problem.