Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit https://heatinghelp.com/contact-us/.
Click here to Find a Contractor in your area.
unwanted flow in zones not calling for heat
in Gas Heating
I upgraded my heating system to a new condensing boiler. When I did that I had the installer replace the old zone valves with individual circulator pumps, one for each zone. When he did that he pumped the return, my old system pumped the feed. They are all controller by a new Taco SR506 relay system. We confirmed that it is working correctly. I have 5 zones. My problem is that when one zone calls for heat, I get flow in two other zones. The zone to my den and to a finished room in the basement seem to behave ok. It appears that pumping any of the three zones in question , causes pull through in two other , but not all five. The circulator pumps are Grundfos 3 speed units with and integrated check valve. The installer verified the check valves are installed. I am stumped.. I would have expected the check valves to prevent this zone cross talk. I do not understand why three do it and two do not. My installer now thinks I need flow control valves and not check valves. Any suggestions would be appreciated. I attached a sketch of my system.
Most mod con boiler recommend primary secondary piping or a hydro separator. Do you have the installation manual? If not pull one online for the boiler you have and see if you have anything close to the recommended piping. Doesn't look like it?
It is possible to get ghost flows when the distribution is above the boiler piping and checks may be needed on S&R of every zone. Sounds like more than ghost flow.
What model circs? All on the same speed?
The checks supplied with the circus should be plenty, if required use spring checks on the other side.
Good zoning info here, especially the section on different sized circs.
http://www.caleffi.com/sites/default/files/coll_attach_file/idronics_5.pdfBob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0
The pumps are Grundfos UPS15-58FC/FRC units. I believe the checks in them are spring checks but not sure. I had them on different speeds. Big loop upstairs on high, small loop to den on low , others on medium. I set them all on low to see if it helps.
I do have the manual, I will look. My installer does 3-4 per week and says he has not seen this. He did not replumb the house, he used existing circuits. The only obvious item is that the S side is a single 3/4 pipe that runs down the middle of the basement and splits out to the four zones from the end. Not sure how that might effect things, logically it is the same as have a manifold closer to the boiler but not sure. Does say he thinks flow controls in return side will solve. ( return side much easier to do plumbing). Is there a definition of just what a ghost flow is? I suspect my problem is being caused by one pump pulling through another or by gravity feed taking place. the Grundfos check valve has very little forward resistance and I suspect the pump itself is not much of an impediment. I am not sure what a hydo seperator is or how it might help. Can you enlighten. Thank you for taking the time to reply. As I have spent a fair bit on this new system, I am frustrated it is not behaving as expected. One other note. The unit is a Lochnivar and it default boiler setting was 120 degrees on heat 145 for HW. The technician upped the boiler setting to 170 degrees. This is when I noticed the issue. Perhaps the higher temp is inducing a gravity feed. As the flow is significant enough to heat the room, I am not sure the gravity story is the answer. I am inclined to suspect a pull through.0
Take an hour and read through that tech journal, maybe send a copy to the installer. It shows and explains potential issues with zoned systems.
Ghost flow is basically un-wanted flow. It can be caused by pumps conflicting, flow going up the opposite end of the loop, buoyancy induced flow where hotter water rises up inside the pipe, cooler water falling.
The checks installed in the circs should be plenty to provide 100% shut off, an additional flow check should not be needed.
I'm pretty sure all the Lochinvar mod cons suggest P/S or hydraulic (low loss header) separation in the I&O manual. Is it a Knight or Cadet model? Here is a caution from the Knight I&O
A pic of the piping might help, I'm curious where the expansion tank connects into the system.
Circs are wired correctly to the proper thermostats?Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0
It is a Lochnivar Knight WHN110
The thermostat wiring is correct. the Taco SR506 is operating correctly
I have provided a PIc of the cluster of pipes where 4 of the 5 zones break out. They are at the end of a 24 foot run from the boiler down the middle of the basement and are on the supply side. ( originally on return side with old Weil McLean and Taco zone valves.)
I also added a pic of the system as installed on my basement wall and where the expansion tank is. it is hung off the bottom of a spiral air removal device in the supply side of the system
Thank you again
I will read the manual
Zman Member Posts: 7,446Never mind"If you can't explain it simply, you don't understand it well enough"
Don't see how P/S piping will correct his problem, I do however agree that some of those zones look pretty skimpy and the boiler should be piped that way to maintain proper flow through boiler.
There are shut offs on the inlet to each pump. You can simply close them off on the zones the ghost feed and if that solves the problem then the checks or flow control should also work.0
Just because your installer does 3 or 4 install a week, doesn't mean they are right or correct. He's just lucky.
If it was just a replacement, he should have installed the boiler with all the trim with a primary pump pumping IN to the boiler, looped it over or piped it through the primary side of a Hydraulic Separator, and connected the existing system through the secondary side of the Separator. All the issues that can come up will be resolved in the Hydraulic Separator. Almost every installation manual for high mass or Mod Con boilers I have seen since 2000 have shown Primary/Secondary piping as the method of choice. Hopefully, your installer will realize this and start doing his 3 or 4 install per week with P/S like that.
If he just hung a boiler on the wall, and dressed it with all the required fittings, into a Hydraulic Separator, then, connected the secondary side to the Hydraulic separator, there would be a lot less problems.
I really think that some are having a hard time visualizing the mixing going on in closely spaced tees or hydraulic separators.
The only half @$$'ed fix for that would be to tie the supply and return AFTER the last circulator and to the return and put a differential flow regulator to bleed off the excess pressure, being forced/sucked around in the unused zones. What Hydraulic Separators do all by themselves.0
The manual is pretty clear regarding the piping options. Also the pump selection and speed setting.
With that type of fire tube HX possibly a single pump for the boiler and zones may work, it could cause some excessive velocity in the small, short zones however, since the circ needs to be on speed two.
Check the checks, maybe something is stuck in one or several.Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0
As above ^^ Check the checks for proper checking.
Maybe I didn't see the photos correctly, but it looks to me like the boiler is piped p/s.
Honestly, I believe the problem lies in the common supply and the zone branch junction. If any zone calls for heat it will bring hot water to that junction. From the junction, it looks like the branches go vertical. No doubt one or two of them go vertical straight to the first emitter in the loop. If that's the case, you will get ghost flows heating up that emitter even when the zone isn't calling.
Easy to diagnose. Turn on only one zone at a time and feel the supply branches. See how far up the pipe they heat up. You may have to put flow controls on several or all of the branches. Or bring the branches back for a longer horizontal run and pipe in a heat trap on each branch.
IMO and understanding, because the boiler isn't pumped as a P/S, and no Hydraulic Separator is used, a over pumped small zone is trying to push through the high restriction of the high restriction boiler. The pressure has no other place to go on the trip back. When using the Hydraulic Separator, you have constant flow initiated through the boiler and the H-S, and the excess flow pressure is given up in the H-S.
Just for BM's and laughs, I always added a second Tridicator gauge to Mod-Cons so I could see in and out pressures. A benefit was I noticed that you can get fluctuations between the P/S systems.
Take up sailboat racing. You'll really come to understand what slight pressure differentials can do.
And if you take the OP's drawing, you will find that something isn't piped exactly as drawn. Its how he THINKS it is piped. A photo shows it all. Its hard to fool a photograph.0
There is a pump underneath the boiler that could be P/S piping.0
The OP's drawing didn't show the pump under the boiler. The photo doesn't show any P/S connection. The connection (if there is one) may be improper. If it is a properly piped P/S system, it should be working. A photo is needed of the pump below the boiler and the connections.
A hydraulic separator would have resolved all issues.
I got a chance to read the manual as suggested earlier. It definitely has a primary, secondary set up. From what I could tell, I have two differences from the recommended piping.
1. The flow is reversed , the pumps are pumping toward the manifold. This causes the heated water to travel 33 feet to where the zones split at the other end of the house ( 33 ft away)
2. I do not have a second manifold as implied by the Lochnivar schematic. Kind of the same issue as above.
There are two zones that do not participate in the ghost flow. The first is the basement zone. It does branch at the cluster which is on the basement ceiling , but it goes down to the basement floor and then up to the basement ceiling to return. The other is the den zone. The only visible difference with this one is its return is closer the the boiler by 20 ft.
As I am a electrical guy, as I look at this as I would an electrical circuit, I don't see a major difference. As I am not a fluid dynamics guy, I do see the positioning of the splitting of the zones so far from the furnace as quite a difference from the schematic, although they do not specifically speak to this in the install documentation.
I have added the original schematic and an edited version to show what I found as delta's.
I have also adds some better pictures of the primary loo, the long run and the split out.
The service guy who came to look at this said he has not seen a system behave like this in his 15 years. He felt the check valves should prevent this.
They plan to come back next week and remove the check valves on the output side of the pumps and install flow control valves the input side.
-will this work?
-Should I ask them to install three more pipes back t the boiler and pipe with another manifold approach?
-Any ideas why the den and basement behave and don't cause ghost flow in any other zones? Do the ball valves on the basement and den help in any way? three problem zone do not have down stream isolation valves as these do.
-Does the direction of flow contribute in any way? not sure why they did this.
- will this require a hydraulic separator to get right? ( i assume I would need to add the 3 pipes to do this. I did not see a requirement in the documentation, unless the reference to a low loss header in the schematic is such.
Again, I very much appreciate everyone willingness to help me understand this and ultimately help my contractor get it right.
The contractor is coming Monday 1-5-2015to install flow controls. Because the supply side is a 33 foot long single 3/4 "pipe before it breaks out into individual zones, he is putting them on the return side of the circuits. It is the return side that is connected individually to the circulatory pumps. The pumps push the the water through the long common pipe. I have seen the recommendation that the flow controls be put on the supply side. I am wondering if the return/supply installation makes any difference? Given we have not been able to determine a specific technical explanation why I get ghost flows in some zone and not others , I am nervous that the proposed approach will not actually work. I fear that each zone will need to piped back to the boiler to resolve this. Additional comments appreciated.0
Sounds like some air causes and air solutions. Solutions pulled out of thin air.
Problems like yours require some lengthy and serious face time with the problems. IMO, those little dinky IFC's found in circulators are manufactured with small teeth that can bite you. You hardly ever see anything like that happening with those big old ones that the cost cutters did away with.
Something to be said about old.0
Without putting hands on the system myself, I can't be absolutely sure... However, if the problem is ghost flow as I believe it to be, putting the flowchecks on the return will do absolutely nothing to cure the problem. I realize there is a bit more work involved with putting them on the individual supply branches, but that is where they belong.
It'll cost you less to do it right the first time.2
I believe the contractor's theory is the additional flow controls will add enough resistance to stop the gravity flow. The guy from Grendfos thinks the weight of the water above the boiler is holding the checks open. Again, as I am not a plumber I have no experience with this. My personal view is that the 33 foot long supply side pipe has something to do with it. The three zones that do get ghost flows when any one of them is calling are the originial to the house when it was built 50 years ago. The two that do not participate in the problem were added with additions to the home say within 10 to 15 years. What baffles me is that when one pump in pumping into the manifold , it should be putting back pressure on all the other zones attached to the manifold. Flow through any off pump should only happen if there is a pressure,differential across the check valve. I can't figure where it is coming from. The water weight alone should not be able to overcome the back pressure supplied by the one pump that is pumping. In theory only another on pump could do that. One responded suggested that the on pump wasomehow creating a venture effect. Perhaps something like that could be happening. I would suspect that to happen at the common breakout point 33 feet away. One other speculation I had was that the water being heated along the 33 foot pump is expanding enough push open the checks in the other zones. Not sure is water expansion plus the weight of the water could over come the back pressure of the pump. I have read these pumps are just water movement facilitators so the actual pressure the product is not high. I have seen that if the boiler is turned on with the isolation valve to the expansion tank closed, the pressure relief valve will trigger in short order. I am speculating again. I do appreciate all the helpful comments I am getting0
I'm late to the party, but here's my $.02 worth:
1. As Harvey, Ice and Hot Rod (some of the best around), have said, ghost flow can occure in a pipe from the buoyancy of the water being heated and causing the hotter water to rise up the middle of the water column while the cooler falls back down the perimeter.
2. Your are very much over-pumped with the number of circ's on the zones.
3. Your common piping in the header is too small for the volume of flow that the zone circ's can produce when two or more are running.
4. The business about the column of water forcing the check open seems unlikely to me since it would be exerting equal pressure upon both sides of the check valve in a closed loop. An open loop would be different.
5. Your observation about the hot water traveling 33 feet before splitting is correct and would be a contributing factor in what I described in #1 above.
Therefore, the best solutions that I see would involve:
1. Installing flow checks in the supply lines, not the returns
2. Set all your circ's to low speed
3. Re-pipe the common header to at least one sizer larger (1.25" ?). Or use a low loss headeder as mentioned. This may not be necessary if 1 & 2 solve the issue.Bob Boan
You can choose to do what you want, but you cannot choose the consequences.0
"" The guy from Grendfos thinks the weight of the water above the boiler is holding the checks open. ""
What? If everyone stands on one side of a Carousel, it doesn't start moving by itself does it? If a Ferris Wheel has 20 cars around the wheel, and only 4 are filled but they are all together, if the operator stops at 1 O'clock, and disengages the transmission, the wheel rolls so that the 4 cars are on the bottom. Its called gravity. Did the guy from Grundfos tell you that you need a bigger pump to push the water up to the second floor? If he did, he needs to come here to get schooled in how pumps work.
If there are IFC's on the pumps, they are so dinky with that dinky spring, it doesn't take much of anything under the seat to set up gravity or ghost flow.
Not all tech support people are equal. Sometimes you learn from them. Sometimes, they learn from you. The best experiences are when you both learn from each other.0
3 to 4 WHN's a week?Steve Minnich0
This contractor is a large company with many crews. They do both commercial and residential with specialized crew for each. They focus on Lochnivar unless a combi. The team that did mine ( 5 zones) put it in in 1.5 days and the team lead told me he does a 1-2 zone system in a day. All his team does is boiler replacements in residences. I watched them, they were like a machine. On line reviews were overwhelming good and they have been diligent in trying to solve the problem although they admit they have not seen my issue before. Thay is why I have been reaching out to the forum to see if someone out there actually understands what is happening. As a retired electronics engineer, I have an appetite for specifics, although I am learning with fluid dynamics that is not as simple to com by. Each implementation can behave differently. Thanks for replying0
If it was me...
Option 1 (simplest) - I would mount each circ so they are pumping into each zone. That way the zone's friction loss is acting as a flow break (like a partially throttled valve) before any manifolds, "T's" etc at the end of the zone. Run each circ at as low a speed as possible to provide comfort (check with system under as full a load as possible, all zones calling. Adjust the circ speed to get close to 20 Deg delta T).
Option 2 - Reinstall zone valves and use one circ for all zones (this will save a substantial amount of electricity as well).
Option 3 (most expensive) - To eliminate the PROBABLE over pumping cause of the ghost flows, use all delta T circs (sorry if this sounds like a sales pitch - merely a suggestion and one that would work).
Remember, water is lazy and stupid - takes the path of least resistance.0
"" Remember, water is lazy and stupid - takes the path of least resistance ""
Just like electricity. Direct Current electricity.0
Some may disagree with this and that's fine, but the main issue is that the flow checks are on the return because they're internal to the circ's and the circ's are on the return. Flow checks belong on the supply (so do circ's normally).
If the installers had put the circ's on the supply to begin with you wouldn't be having ghost flow. Circ's should pump away from the point of no pressure change (PONPC) which is the point where the expansion tank connects to the system.
I'd highly recommend that your installers get a copy of Dan's book "Pumping Away" from the book store on this site and carefully read it until they grasp the principle.Bob Boan
You can choose to do what you want, but you cannot choose the consequences.0
100% Agreed - Pumping Away is a classic. Should be mandatory for anyone in the "Comfort" business...0
My contractor is returning tomorrow and is going to eliminate the single feed line by running new Individual lines across the basement to complete the individual loops all the way back to the boiler via an additional manifold. I did discuss the Don Holohan book on pumping away and they were familiar with Dan's books but did not see the direction of flow as a part of my problem. They pointed out I was not pumping into the PONPC because of the way the primary/secondary coupling was set up. They told me they also consulted with Lochnivar and the supplier concurred the direction of flow should not be the issue. As the long and undersized single pipe is what is truly unique to my implementation, I am optimistic this will do it. If not at least we will have eliminated the unique feature. In fairness to my contractor, the piping in the house is what was done 60 years ago. I wish they spent more time an analyzing the existing piping before the did the install. That said they have been diligent in attempting to get it resolved. Hopefully my last post will be ..."it's fixed"0
The contractor has been there for 3 days.0
The service tech came when I reported the issue. As it was not really a service issue, the company supervisor came to look at it. The supervisor called a few experts includingnLochniva directly and came back with the forms hydronic expert. I did mention Dan's book and he was very familiar and said he had most of Don books. Based on this on site review the decided add 5 seperate line across the basement to a ,amid old at the boiler. Since I last posted, they have also decided to reverse yhe flow, "pump away"0
When you say the checks fail in the open position, what exactly does that look like? Is there a piece of crud holding them open? Broken parts? Melted from overheating a solder joint against the circ?
It a very simple device those internal checks. I'd guess the same type, maybe even the same brand is used in all the different colors of circ pumps. The color of the plastic may change from one brand of circ to another.
It's also the same check used in many of the hydronic style brass check valves you buy from major hydronic manufacturers.
They are engineered to do an excellent job providing a nice cone shape, friendly to flow, soft seats, for noise-free operation, and a low pop pressure spring so they open even with low speeds common with VS pumping now a days.
At some point a motorized valve may be require if the "driving force" for that un-wanted flow exceeds the stopping power of the check. That's not the check valves fault
If checks are failing open, perhaps look at adding a y-strainer or better yet a dirt separator.
If they are bouncing, perhaps you have a flow restriction, looking at the design of these checks there in not much to make noise, plastic cone against a rubber o-ring, unless, again something metallic is stuck in the check, they should operate silently.
As hydronic components get more efficient you will find tolerances are getting tighter. Probably circ pumps will get refined hydraulics to get to the next mandated efficiency level. I suspect that will require closer tolerances, less "hydraulic slip" to get the additional performance, best be thinking about devices to protect the components in hydronic systems from the smallest particles that may be swimming around inside your systems.
Magnetic separation will become more and more common as the magnets will grab the super fine particles that are tough to "strain" or filter out.
I don't think it's wise to add a flow restrictive device at the suction side of a circ, didn't one of the pump manufacturers try that a number of years ago?Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0
- 121.3K All Categories
- 84K THE MAIN WALL
- 2.9K A-C, Heat Pumps & Refrigeration
- 53 Biomass
- 417 Carbon Monoxide Awareness
- 40 Chimneys & Flues
- 1.7K Domestic Hot Water
- 4.7K Gas Heating
- 119 Geothermal
- 155 Indoor-Air Quality
- 3K Oil Heating
- 56 Pipe Deterioration
- 771 Plumbing
- 5.4K Radiant Heating
- 362 Solar
- 14K Strictly Steam
- 3K Thermostats and Controls
- 51 Water Quality
- 625 Buy, Sell, Barter
- 38 Industry Classes
- 73 Job Opportunities
- 17 Recall Announcements