How to eliminate "GHOST FLOWS"?

[Donl]

I have a Primary/Secondary configuration. The primary loop is about 20 feet of 1 1/4" B.P. with fittings,powered by continuous running Taco 007 pump. The two secondaries are 1 1/4" pex connected to the Primary through closely spaced tees. These secondary loops, round trip, are about 160 feet in length. The secondaries use Grundfos 15-58 pumps that are activated only when there is a call for heat. When there is no call for heat the idle secondary will experience ghost flows.

I don't know how much waste is being created here. I don't know what rate the ghost flow is in gpm. However, if the temp starts out at 180 F, it returns at about 120 F during ghost flow.

Any ideas as to get ride of these flows? Do you think that maybe the Taco 007 is causing too much fplow in the primary? Do you think weighted flow checks in the secondaries may help? Anything?

Anybody experiencing similar problems with Primary/Secondary configs?

 

[heaterman]

Ghost flow in a properly set up P/S arrangement in a testimony to the power of thermal action caused by light, buoyant hot water doing what comes naturally. It wants to rise to the point where it equalizes in temperature and then begins to return. The only way to stop this natural action is with a positive stop device such as a flow check, either spring loaded or weighted. Choice of which depends on the orientation of the device, the amount of increased head the system can tolerate and of course personal preference. A circ with a built in check or a set of pump flanges with the same will usually take care of it.

It can be a really troublesome thing to deal with and not just on P/S systems. I have seen ghost flow establish itself in a single pipe. You could actually measure and feel the temp difference on each side of this particular vertical pipe. The water was traveling up one side and down the other.

 

[Donl]

A circ with a built in check or a set of pump flanges with the same will usually take care of it.

The 15-58 pumps I am using on the secondaries do have built in checks. Is this what you mean? The ghost flow direction I am experiencing flows in the same direction as the pump would normally move it. If I install a spring loaded check valve such as http://www.pexsupply.com/Matco-Norca-525T05-1-Spring-Loaded-Threaded-Check-Valve-5564000-p is the idea that the spring pressure will stop the ghost flows, but the pump when activated will overcome this spring pressure to allow adequate flow? I currently operate the 15-58 pumps on low speed. Therefore, if necessary I could increase their speed to overcome the spring pressure.

Thanks, Don

 

[kabbott]

A circ with a built in check or a set of pump flanges with the same will usually take care of it.

The 15-58 pumps I am using on the secondaries do have built in checks. Is this what you mean? The ghost flow direction I am experiencing flows in the same direction as the pump would normally move it. If I install a spring loaded check valve such as http://www.pexsupply.com/Matco-Norca-525T05-1-Spring-Loaded-Threaded-Check-Valve-5564000-p is the idea that the spring pressure will stop the ghost flows, but the pump when activated will overcome this spring pressure to allow adequate flow? I currently operate the 15-58 pumps on low speed. Therefore, if necessary I could increase their speed to overcome the spring pressure.

Thanks, Don

Don, I am not familiar with the flow checks in the 15-58 pumps but the taco built in checks take a bit to open and I would not think that the pressure drop
across two close spaced tee's would be enough to open them. Do both loops have the ghost flow? If only one could there be debris stuck in one?
How far apart are the P/S tee's?

I get some flow on my tank charge/discharge loop with NO checks but it is caused by thermosiphoning not pressure drop, if I close the valve
for an hour or so and then re-open it the flow stops. This is with no flow checks and an 007 pumping through 1 1/2 inch manifold on my
hydraulic separator.

If you are positive there is flow and not just thermal action up one side and down the other as heaterman describes then I suppose the solution
would be another flow check or an electronic ball valve/ zone valve.

 

[Dune]

Heaterman nailed it. Sometimes even flowchecks do not stop it. It can be very frustrating.

 

[Donl]

[
Don, I am not familiar with the flow checks in the 15-58 pumps but the taco built in checks take a bit to open and I would not think that the pressure drop
across two close spaced tee's would be enough to open them. Do both loops have the ghost flow? If only one could there be debris stuck in one?
How far apart are the P/S tee's?

I get some flow on my tank charge/discharge loop with NO checks but it is caused by thermosiphoning not pressure drop, if I close the valve
for an hour or so and then re-open it the flow stops. This is with no flow checks and an 007 pumping through 1 1/2 inch manifold on my
hydraulic separator.

If you are positive there is flow and not just thermal action up one side and down the other as heaterman describes then I suppose the solution
would be another flow check or an electronic ball valve/ zone valve.

Yes both loops have the ghost flows. If I don't activate either secondary and get them hot, there will be no ghost flows through them. If I do run it and heat it up I will get ghost flows. After several hours if I don't activate the secondary the ghost flows will gradually stop.

Both secondaries are clear of any debris and I get very good flow through them.

I have known since last year about this problem and it does not cause a major problem. It just make the overall system less efficient which tends to bug me. I think I will try a spring loaded check valve on each secondary. My hunch is that Heaterman is correct (as usual) on this.

Don

 

[tigermaple]

I also get ghost flows using 007 taco IFC's in a similar situation. What do you guys mean "2 closely placed tees"? Is the supply and return closely placed?
Pat

 

[Donl]

I also get ghost flows using 007 taco IFC's in a similar situation. What do you guys mean "2 closely placed tees"? Is the supply and return closely placed?
Pat

Yes, The supply and the return are connected to the primary loop with about 6 inches between them.

 

[heaterman]

I also get ghost flows using 007 taco IFC's in a similar situation. What do you guys mean "2 closely placed tees"? Is the supply and return closely placed?
Pat

The supply and return for a given zone are both on the same pipe, situated as close as possible to each other. The rule is no more than 4 pipe diameters apart center to center. Less is better, anything more and the pressure differential between the S & R will cause circulation without a pump running on that zone. The supply and return coming from the boiler just serve to make water available for all the secondary zones. No zones are directly connected to the boiler, only to the short/primary loop.

Primary secondary goes like this.. The primary loop is where all the btu's "get on the train" as it leaves the boiler. This loop goes nowhere except around a little loop (short little track if you will) and head directly back to the train station (boiler). If the btu passengers need to go somewhere in particular, say to your living room, a second track (secondary loop) would take them there with it's own engine (circ). Some of the btu passengers don't get off in the living room so they circle right back to where they came from on the short little track where they join up with the rest of the crowd. The off and on points are right next to each other so the btu passengers simply hop on the main train and ride along being pushed by the main engine (circ) instead of the little engine (circ again) that took them out to the living room.

Gettin pretty late for story time.............I'm headin' to bed........been a long week. i can tell when I start to babble like that......

 

[snowman49820]

I have the exact situation. I didn't have this problem last year but this year it's all the time. My supply and return are only 10" apart. Had no problems until I switched to bigger 3/4" pex. I'm thinking I need to separate them further.

 

[Donl]

I'm thinking I need to separate them further.

Doing that will only increase the pressure difference between the Secondary supply and return, resulting in additional unwanted flow.

 

[leaddog]

I'm thinking I need to separate them further.

Doing that will only increase the pressure difference between the Secondary supply and return, resulting in additional unwanted flow.

Yes I think he needs to put them CLOSER. I know that sounds backwards but that is how it works. The closer the better
leaddog

 

[leaddog]

Another way to help is to have the supply dip atleast 12in as that will stop the flow unless it is fast. Also don't insulate the first part of the supply. Anything that lowers the temp fast in a small section will stop the thermo-flo.

 

[tigermaple]

Thanks for the bedtime story Heaterman. My problem then is my tee's are way to far apart (with a oil boiler in between). No problem last year, but I streamlined all my zones in a massive remodel this year, taking out yards of extra copper and replacing with pex. I thought I was making things more efficient at the time. You would think running water through pipes would be simple.

 

[heaterman]

I have the exact situation. I didn't have this problem last year but this year it's all the time. My supply and return are only 10" apart. Had no problems until I switched to bigger 3/4" pex. I'm thinking I need to separate them further.

What's the diameter of the main? Remember that the P/S rule is not more than 4 pipe diameters apart. So if the main is 1", 4" apart would be max. If it's 1-1/2" then you could go up to 6", 2" main would be 8" apart and so on.

If you have a 1-1/4" main with 10" spacing you will indeed have some unwanted flow even with an FC style circ. Leaddog brought up another good point and I have used it a couple times on "bleeders" that were really stubborn. That is, when you want to come off the main vertically, elbow your secondary loop down about 12-18" first befroe making a 180 bend and heading back up. This acts as a thermal trap if you will and will sometimes do the trick when nothing else will. Switching to the 3/4" pex made the zone flow easier and that's why you have the problem now.

 

[Gooserider]

From the hardware standpoint - to do closely spaced tees most effectively, put them together using a "close nipple" on black iron. On Copper, a section of tube that only lets you see about 1/8" of pipe between the two fittings. On PEX the shortest length that will let you assemble the two clamps...

IOW, while the MAXIMUM is no more than 4x pipe diameter, the optimum distance is as close as you can mechanically get the fittings... Keep in mind that you are looking at a mechanical minimum of about 2 pipe diameters just from the length of the tee bodies, and you really don't want to add anything to that.

(I'm mildly surprised that the fitting makers don't offer a pre-manufactured "close spaced tee" as an actual fitting - or maybe they do and I've just never seen it mentioned)

Gooserider

 

[heaterman]

From the hardware standpoint - to do closely spaced tees most effectively, put them together using a "close nipple" on black iron. On Copper, a section of tube that only lets you see about 1/8" of pipe between the two fittings. On PEX the shortest length that will let you assemble the two clamps...

IOW, while the MAXIMUM is no more than 4x pipe diameter, the optimum distance is as close as you can mechanically get the fittings... Keep in mind that you are looking at a mechanical minimum of about 2 pipe diameters just from the length of the tee bodies, and you really don't want to add anything to that.

(I'm mildly surprised that the fitting makers don't offer a pre-manufactured "close spaced tee" as an actual fitting - or maybe they do and I've just never seen it mentioned)

Gooserider

I use and sell a ton of these. The ball valve in the middle makes it possible to purge the secondary loop through the same ports used for the primary. Webstone rocks

http://www.webstonevalvespurgetee.blogspot.com/

 

[Gooserider]

From the hardware standpoint - to do closely spaced tees most effectively, put them together using a "close nipple" on black iron. On Copper, a section of tube that only lets you see about 1/8" of pipe between the two fittings. On PEX the shortest length that will let you assemble the two clamps...

IOW, while the MAXIMUM is no more than 4x pipe diameter, the optimum distance is as close as you can mechanically get the fittings... Keep in mind that you are looking at a mechanical minimum of about 2 pipe diameters just from the length of the tee bodies, and you really don't want to add anything to that.

(I'm mildly surprised that the fitting makers don't offer a pre-manufactured "close spaced tee" as an actual fitting - or maybe they do and I've just never seen it mentioned)

Gooserider

I use and sell a ton of these. The ball valve in the middle makes it possible to purge the secondary loop through the same ports used for the primary. Webstone rocks

http://www.webstonevalvespurgetee.blogspot.com/

That is SLICK :coolgrin: I hadn't even thought about putting a ball valve in the middle, but I can see how that would make for easy purging and such. At the same time, I suspect it might be difficult to make an equivalent out of discrete components without exceeding the 4x pipe diameter guide, not to mention the extra flow resistance the added fittings would cause... Webstone seems to be really good at this sort of specialty fitting, I think you or other folks have pointed at them for other stuff a few times as well...

Gooserider

 

[MNBobcat]

Let's say you have 180 degree water coming in. On your secondary loop, why would you not take the supply off of the supply but have the return go into the return line going back to the boiler instead of dumping the now cooler water back into the supply? By dumping back into the supply wouldn't the next exchanger on the next secondary loop receive much cooler water?

 

[Gooserider]

Let's say you have 180 degree water coming in. On your secondary loop, why would you not take the supply off of the supply but have the return go into the return line going back to the boiler instead of dumping the now cooler water back into the supply?

If you did that, it wouldn't be P/S plumbing any more...

By dumping back into the supply wouldn't the next exchanger on the next secondary loop receive much cooler water?

Yes it would... However part of the theory of P/S plumbing is that the upstream loads shouldn't take so much heat out of the system that the downstream loads are "starved" - so as long as they can deal with the lower temperature, it doesn't matter - This means that part of the strategy is to put the high temp demands like a DHW tank early on the loop, and lower temp tolerant loads on the latter part. The other part is to use loads that are flexible about what kinds of temperatures they need, or that need temps lower than what the loop can supply - for instance a radiant loop that only needs 120°F water, and is fed from a mixing valve, or a panel rad with a TRV on it, that can open and close to change it's volume need depending on how hot the water coming into it is...

If you have a bunch of loads that all insist on getting the same temperature supply, you could also go to a "Parallel P/S" configuration, where you split the primary loop into a bunch of parallel branches (using big enough header pipes that the branches are "hydronically equal") and tap one secondary off of each branch...

I will admit that while I agree that there are some nice things about P/S plumbing, I'm not completely sold on the idea that it is better from an efficiency standpoint than traditional supply / return configuration. However that doesn't mean that I don't get how it's supposed to work...

Gooserider

 

[dogwood]

Gooserider, Taco "Twin Tees" serve the same closely spaced tee function as the Webstone fitting, without the ball valve advantage however. Try http://www.taco-hvac.com/uploads/FileLibrary/100-6.8.pdf. Saved this info from an old thread.

Mike

 

[Gooserider]

Neat, but for some reason I don't get the same "Thats cool" feeling from the twin tees that I did from the Webstone valve... Nonetheless, I think that one needs to go in the tidbits sticky along with the Webstones...

Gooserider

 

[rkusek]

[
Don, I am not familiar with the flow checks in the 15-58 pumps but the taco built in checks take a bit to open and I would not think that the pressure drop
across two close spaced tee's would be enough to open them. Do both loops have the ghost flow? If only one could there be debris stuck in one?
How far apart are the P/S tee's?

I get some flow on my tank charge/discharge loop with NO checks but it is caused by thermosiphoning not pressure drop, if I close the valve
for an hour or so and then re-open it the flow stops. This is with no flow checks and an 007 pumping through 1 1/2 inch manifold on my
hydraulic separator.

If you are positive there is flow and not just thermal action up one side and down the other as heaterman describes then I suppose the solution
would be another flow check or an electronic ball valve/ zone valve.

Yes both loops have the ghost flows. If I don't activate either secondary and get them hot, there will be no ghost flows through them. If I do run it and heat it up I will get ghost flows. After several hours if I don't activate the secondary the ghost flows will gradually stop.

Both secondaries are clear of any debris and I get very good flow through them.

I have known since last year about this problem and it does not cause a major problem. It just make the overall system less efficient which tends to bug me. I think I will try a spring loaded check valve on each secondary. My hunch is that Heaterman is correct (as usual) on this.

Don

Resurrecting an old thread. Don L. did the spring checks work? Did you put them on just 1 of the lines on your secondaries (one w/o the circ) or on both lines? Did you have to run the circs higher than low? I'm guessing a spring check is better than a swing check for this purpose since Heaterman suggested it. Is a swing check mainly concerned with stopping the flow in the opposite direction where a spring check will stop the ghost flows because of the force required to open it? I think I want to add these to my connections to storage and also where the boiler and furnace water to air HX circ tie into. My closely spaced tees are all within the 4 pipe diameter rule since I used closed nipples but I don't want to have to change something later. Also, anyone with a P/S setup using the 2 circs pointed at each other or did you keep the charging and drawing taps separate? Since I'm using existing tank connections I'm trying to figure out a way to make it work with 1 set of pipes, maybe using a couple high flow zone valves and these check valves.

 

[Donl]

[
Don, I am not familiar with the flow checks in the 15-58 pumps but the taco built in checks take a bit to open and I would not think that the pressure drop
across two close spaced tee's would be enough to open them. Do both loops have the ghost flow? If only one could there be debris stuck in one?
How far apart are the P/S tee's?

I get some flow on my tank charge/discharge loop with NO checks but it is caused by thermosiphoning not pressure drop, if I close the valve
for an hour or so and then re-open it the flow stops. This is with no flow checks and an 007 pumping through 1 1/2 inch manifold on my
hydraulic separator.

If you are positive there is flow and not just thermal action up one side and down the other as heaterman describes then I suppose the solution
would be another flow check or an electronic ball valve/ zone valve.

Yes both loops have the ghost flows. If I don't activate either secondary and get them hot, there will be no ghost flows through them. If I do run it and heat it up I will get ghost flows. After several hours if I don't activate the secondary the ghost flows will gradually stop.

Both secondaries are clear of any debris and I get very good flow through them.

I have known since last year about this problem and it does not cause a major problem. It just make the overall system less efficient which tends to bug me. I think I will try a spring loaded check valve on each secondary. My hunch is that Heaterman is correct (as usual) on this.

Don

Resurrecting an old thread. Don L. did the spring checks work? Did you put them on just 1 of the lines on your secondaries (one w/o the circ) or on both lines? Did you have to run the circs higher than low? I'm guessing a spring check is better than a swing check for this purpose since Heaterman suggested it. Is a swing check mainly concerned with stopping the flow in the opposite direction where a spring check will stop the ghost flows because of the force required to open it? I think I want to add these to my connections to storage and also where the boiler and furnace water to air HX circ tie into. My closely spaced tees are all within the 4 pipe diameter rule since I used closed nipples but I don't want to have to change something later. Also, anyone with a P/S setup using the 2 circs pointed at each other or did you keep the charging and drawing taps separate? Since I'm using existing tank connections I'm trying to figure out a way to make it work with 1 set of pipes, maybe using a couple high flow zone valves and these check valves.

I placed spring loaded check vales in the secondary return lines. This resolved the ghost flows . No noticeable effect overall flow with either taco 007 or Grundfos 15-58.

 

[rkusek]

Sounds great, I will just put them on the return side. Don is your storage connected differently now? I saw your response in a post last January to Huffdawg system and you mentioned you eliminated the 4 way valve but the new diagram of your system shows a 4 way valve on the storage. Did you have a 4 way valve somewhere else? Where did you find the 4 way and what are the specs on it? With only 1 set of existing taps on my tanks I am looking to go that route instead of separate charging and discharging loops connected to the primary loop. Does the valve have limit switches that verify it is fully switched over to either the charging or discharging position? If I can't find a 4 way valve I may have to use two 3 ways or something similiar with zone valves and checks.