Ghost Flows from incorrect tee spacing?

Dave_144

End of last Jan we had our old cast iron boiler replaced with a wall mount combi boiler. We have 3 active zones and whenever one calls for heat the other two ghost flow.

Plumber who installed spaced the tee's on the 1-1/4" dia manifold pipe at 8-1/2" apart. As I understand it 5" should be the maximum spacing for these tees? Each zone has a Grundfos circ pump with a internal check valve, but they don't seem to be able to stop the reverse flow.

Would it make sense to have the tees moved closer together or maybe install check valves on the return lines, or both? Thanks for any input

Ironman

I doubt that changing the spacing between the Tees will change anything since most diagrams allow up to 12” between them. A check valve in the return may work, but there’s a reason it’s called “ghost flow”.

Can you post some pictures?

hot_rod

it really depends on that secondary circuit. If it is a very low pressure drop, maybe large piping and large radiators, there is more potential for ghost flow.
The bigger concern is when the secondary is straight up off the primary, then you get thermal siphoning and would need a check on both S&R for absolute protection.
Use a soft seat.low pop hydronic check only.

mikeapolis

Can you show some pictures of how the piping was done? Here's a good article that might help.

https://www.pmmag.com/articles/98167-how-to-eliminate-ghost-flow-and-recirculation-issues

EdTheHeaterMan

Read page 12 in this
http://media.blueridgecompany.com/documents/ZoningMadeEasy.pdf. Good explanation, Brief and to the point, Simple diagrams

hot_rod

The key point is in the last paragraph, Lose the Loop

Dave_144

Thanks for all the replies, appreciate the input

Here's a pic of the system. I think it's called the Moose Antler design?

All zones are baseboard. From the left the first 2 zones are 3/4" and are basic loop systems. The 3rd is also 3/4" but it's not connected and was added to be used for a future addition. The fourth zone to the right is 1" and original to the house (other 2 are later additions) and supplies the main first floor heat using the monoflow system.

In reading page 12 of the link Ed provided, it seems the 1-1/4" manifold pipe is undersized for the 3 active zones and will be undersized for sure when the 4th zone is eventually in use. Could this be part of the ghost flow issue too?

hot_rod

What size is the boiler? The pipe sizing depends on how many BTU and GPM you need to move.
1-1/4" copper can easily move up to 18 gpm, so 180,000 BTU/hr at a 20 degree delta. 1-1/4 would be sufficient for up to a 199,000 input boiler.
Ghost flow is usually due to a check valve in the circulator not shutting off tightly.

I cannot quite see the piping under the boiler, is it like this?

Dave_144

Thanks Rod, here's a pic of the piping under the boiler.

Yes piping is like that diagram and it's a 180k BTU boiler.

The circ pumps are new with this installation. Are the internal check valves pretty robust? Given the ghost flow is happening in all the zones, could it be an issue with the valves in all 3 pumps? I can take them out and give a look if there would be something obvious.

Dave_144

https://s3.amazonaws.com/s3.supplyhouse.com/manuals/1249544647674/Grundfos_ups15-58fc_Product_Data.pdf

I believe this is the circ pump used

hot_rod

Gosh, everything looks fine with the piping.

In some cases you can get ghost flow caused by a two direction flow in the return side of the zone.
Hot water rises up the center, cooler water falls down. It is usually a slow or gradual overheat. Maybe feel the return and supply see if in fact the return is warmer than the supply. Your system is a classic example of return pipe ghost flow "potential" as the pipes rise straight up off the top of the header. The higher the rise the more flow potential.

This was the concept behind the old gravity systems, the taller risers had to be throttled back as they induced more flow.

The fix is to add spring checks, similar to what is in the circulators on the return.
You are positive the circs have checks installed? If not, or if they are stuck open, here is what can happen. That might be the first, simple check.

Dave_144

Yes it's definitely doing what I call back heating. When I hold my hand on the 2 return pipes above the manifold for the zones that aren't calling for heat, very shortly after the call, I can feel the hot rising from the return manifold into those 2 pipes. And if I move my hand I can feel it back traveling. The supply side of the pipes is cool until the hot water makes its way backwards through all the baseboards.

I thought initially it may have been a strong pull created because the tee's are wider than 4 diameters apart but it doesn't sound like that's the case?

I'm going to take the circ pump off of the zone that isn't connected and see if I can spot the check valve and then take a look at the other 3 pumps if it's straight forward.

I don't think the plumber flushed the system so I'm wondering if maybe some debris from the old cast boiler may be keeping the check valves from closing completely. Picked up some of the Fernox F1 and F3 and figure it might be good to use esp if the system has to be opened for the spring check valves?

Canucker

You need check valves on your returns. As @hot_rod has pointed out, the hot water returning from the zone or zones calling for heat wants to go up. The way your system is piped, it wants to go up your returns regardless of whether they're in use or not, causing the ghost flow. Put some checks in your returns, most likely the problem is solved

hot_rod

Here is a check from one of those circs. As you can see, doesn't take much to keep one open.
Nice to have a check serviceable, it can be a maintenance item. We use that exact same check in in this product.

Dave_144

I took the check out of the inactive zone pump and there seemed to be a fair amount of muck on it. I'll check the others and see how they look. Suspect they may be stuck open a bit from debris.

Wondering if it would be worth adding a wye strainer to the manifold in addition to adding the serviceable check valves on the return line if it doesn't take much to keep one open? Does a wye strainer decrease flow rate at all?

Ironman

The strain won't restrict flow until it starts to become clogged.

Handyman 242

Two issues; the first is that the system should have been cleaned with a chemical flush before installation. A Y strainer will only do so much. We have experienced frequent failures of Mod/con boilers being applied in older cast iron systems if the old iron oxides have not been removed. Typically takes around 2 years for the boiler to fail. Chemical treatment and conditioning is the answer.
The 2nd issue with pumping away installations is the gravity flow up the return as other contributors have stated. An easy solution is to install a "thermal trap" on the return connections to the manifold. Each of the 4 returns will be extended 8" to 12" below the return manifold and then a double elbow connection up into the manifold. That will stop the gravity flow. No need to add check valves that will fail in the future. Being in the industry for over 50 years I try to learn from my mistakes, I have made a lot of them.

Smitdy

Kind of off the subject of what you are talking about but that boiler sounds huge even if it has a 5 to 1 turn down like most of those boilers do, you may want to do a heat loss to the house or measure the baseboard and see if the boiler is within 15% of the heating load. Also would want to check if the smallest zone is smaller than the lowest firing rate of the boiler. If either of these doesn't match heating loads you might need a buffer tank to make the boiler run properly (not short cycle) & get the efficiency out of it. If the boiler isn't running for 10 minutes on heating cycle the efficiency goes out the window on those condensing units.

Geosman

Yikes, dirt in the system! you will need to do a thorough flush. I use our geo flush pump with bag filters and divert all radiator water through the bag filter before introducing it back into the system. I use Fernox F3 system cleaner followed up by Fernox F1 as the corrosion inhibitor. Always use a magnetic separator either Fernox or Caleffi work well. With your system you may find you need to remove all the pump check valves until the system is cleaned up and then reinstall them to help prevent future ghost flow issues.

advhtg

It's definitely not a problem of incorrect manifold sizing, dirty non-return valves, muck in the system or spacing of tees!

The other guys have hit it on the head as a thermosiphon up the return pipes, so yes:

- an anti-gravity loop on each return MAY fix it (does anyone out there have concrete examples of anti-gravity loops failing where it was definitely not caused by a small net flow?? I've heard of it but not experienced it) , or

- a second spring loaded non-return valve on each return will definitely fix it, or

- replacing all six non return valves (in the previous bullet point) with 2-port motorized valves will also fix it, but... it is expensive to install, introduces six new sources of electromechanical failure, thankfully reduces the power required and noise generated by each of your circulators (as they don't have to waste effort overcoming two NRV's on each loop), and is possibly less susceptible to dirt settling on the valve faces depending on the valve seal design.

For the choice of two-port motorised valves I prefer the motor-on-motor-off designs of valve by Sunvic and others because they move into position and then electrically shutdown. Big sellers from Honeywell etc are spring-return designs that constantly strain their electric motors while the valve is open, which is an unwanted risk on multiple fronts. A third and super cheap option is to use wax actuators as seen on underfloor heating systems, atop a plunger type valve (like a radiator valve but higher Kv), which just as effective as a "proper" motorised valve but silent, only consumes 2W instead of 4W, and it's opening/closing time of 2 minutes may not matter to you !

mferrer

Gentlemen,
Are flow control valves good to use in this scenario? I see a lot of discussion about NRVs. Holohan's books refer to flow control valves, like the Bell & Gossett cast iron valves. Are they no longer used?
Just wondering.

hot_rod

mferrer said:

Gentlemen,
Are flow control valves good to use in this scenario? I see a lot of discussion about NRVs. Holohan's books refer to flow control valves, like the Bell & Gossett cast iron valves. Are they no longer used?
Just wondering.

Same function, some use a weighted plunger, some a spring.

hot_rod

advhtg said:

It's definitely not a problem of incorrect manifold sizing, dirty non-return valves, muck in the system or spacing of tees!

The other guys have hit it on the head as a thermosiphon up the return pipes, so yes:

- an anti-gravity loop on each return MAY fix it (does anyone out there have concrete examples of anti-gravity loops failing where it was definitely not caused by a small net flow?? I've heard of it but not experienced it) , or

- a second spring loaded non-return valve on each return will definitely fix it, or

- replacing all six non return valves (in the previous bullet point) with 2-port motorized valves will also fix it, but... it is expensive to install, introduces six new sources of electromechanical failure, thankfully reduces the power required and noise generated by each of your circulators (as they don't have to waste effort overcoming two NRV's on each loop), and is possibly less susceptible to dirt settling on the valve faces depending on the valve seal design.

For the choice of two-port motorised valves I prefer the motor-on-motor-off designs of valve by Sunvic and others because they move into position and then electrically shutdown. Big sellers from Honeywell etc are spring-return designs that constantly strain their electric motors while the valve is open, which is an unwanted risk on multiple fronts. A third and super cheap option is to use wax actuators as seen on underfloor heating systems, atop a plunger type valve (like a radiator valve but higher Kv), which just as effective as a "proper" motorised valve but silent, only consumes 2W instead of 4W, and it's opening/closing time of 2 minutes may not matter to you !

In one case I had to deepen the thermal drop after trying the suggested 18" from this Sigenthaler article, 11/2000.
I used pex and went right down to the floor, about a 48"

I'm not sure of a calculation that would give you an exact drop dimension. Depends on the forces that are trying to induce the flow, temperature and height I'd imagine.
I suppose in some cases 6" may be enough drop.

PaulShlasko

The way I learned it, your answer is flow control valves. If you used zone valves, this would not be an issue. Since you used separate circulators, you need the flow control valves. This is nothing new....

clammy

WHen ever I install zone pumps for zoning I will pipe thermal traps for each zone as cheap insurance against ghost flow . The suggestion of cleaning the system is one I would agree with especially on a mod con install you want your system and fluid system clean . What type of heat emitters is the connected to your zones ? Do any of the return pipes run through unconditioned space I have seen piping exposed to out side drafts ( which end up getting fairly cold ) induce thermal migration when any other zone would operate .. remember the laws hi pressure / temp to low pressure / temp . In one case even a flow chk on the return would not remedy until I found they had a 2 Nd floor bath room gutted w no Sheetrock or insulation and about it was about 40 degree in the room . I explained the hows and why’s and insulated the piping the ghost flow ended . In your case it s all the zones . If you are sure that the migration is coming from the returns not bypassing the pump ch valves the sure fire cure is to install flow checks on the returns personally I would use a weighted chk valve danfoss which is brass or Taco . On a side note I get into arguments on this but I always install ifc pumps and on indirects and reverse indirects I add a flow chk on the return especially on low temp or systems using Outdoor reset no sense have migration to the system from the tank and then short cycling the tank because of the migration . Peace and good luck clammy