Grundfos UPS26-99FC internal check valve rattling, installed too close to boiler?

rdbowden

I recently bought a house with a Knight WB105 boiler. The Grundfos pump was a little noisy and shortly after we moved in it got real noisy. I called a plumber and he said the internal check valve had come loose. He replaced the whole pump. Even after he replaced the pump, it was still noisy. A couple of years later I called out another plumber who installs Knight boilers to see if there was some way to quiet it down. He said the pump was too close to the inlet of the boiler and the valve was rattling because it was too close. He said he never has a problem but he always installs the pump further away. For some reason the original installer plumbed everything right underneath the boiler and not off to the side. Before I pay to have the pump moved, has anyone else seen this problem?

Big Ed_4

I am assuming it's the boiler pump ? Best post photo of boiler and pipe layout

EBEBRATT-Ed

If it was me I would remove the internal check valve and install a regular "Flow Check " in the piping (if you even need a check valve) Your plumber will know.

This is a common complaint with internal check valves. The velocity of the water coming out of the pump is at it's highest point at the pump and there is always a lot of turbulence there.

Maybe you can post pictures of the boiler and piping around it. We may spot something amiss

rdbowden

Yes, it is the boiler pump.

EBEBRATT-Ed

I don't think your piped right.

The Grundfoss should pump water through the boiler. It does that. But the pipe going into the boiler return (I think it's the return) from the Taco Pump on the right the tee it feeds into shouldn't be there. That tee needs to be moved. Could be part of your issue shouldn't be there in my opinion


Does the pump on the right feed your domestic hot water??

rdbowden

Thanks for the reply. The pump on the far right feeds the DHW. It is a superstor. The green pump hidden by the spirovent is for the heat. The supply manifold is on the left, the return manifold is on the right. I think the tee for the DHW pump is okay, at least it seems to be in the correct spot according to the Knight installation manual. I have attached an image. I did notice last night that when the call for heat from the thermostat shuts off the heat circulator and the zone valves shut off, but the grundfos keeps running for a short time, which it is supposed to. The noise stays the same. Also, when the DHW is on, the system is silent, can hardly hear it.

GroundUp

I also feel the piping is not ideal, but should work just fine for the most part. There is absolutely no reason for the 26-99 to have an IFC installed at all, it should have never been put in. They come shipped without the IFC installed, so somebody went out of their way to mess this up for you. Being "too close" to the boiler isn't a logical theory either IMO, there is 5 feet of piping upstream.

rdbowden

Thanks for responding. The piping diagram from Knight indicates a check valve is needed on the output of the boiler pump. I think it is there so the DHW pump doesn't flow back into the secondary loop. Also, the length of pipe I am talking about is to the left of the red pump. It is only about ten inches to the elbow and then 7 inches from the elbow to the boiler. The integrated check valve is on the left side of the red pump. I looked at pictures online of actual installs and the primary pump is usually mounted on a wall, well away from the boiler. The piping diagram seems to indicate this as well, although it just may have been how Knight drew the diagram. I am just a homeowner so have no other experience with other installs.

EBEBRATT-Ed

I don't think the pump near the boiler is an issue. I would remove the check valve from the pump and install a weighted flow check valve near the pump. If there isn't room they will have to re pipe it to make room. Nothing else you can do.

A regular flow check has an internal check that is heavier, should stop the rattle

Ironman

You need a flow check on both the boiler and indirect pumps as indicated in the diagram. However, you should NOT have one in the system circulator which is behind the air separator. Zone valves and flow checks should never be used together on the same circuit or loop. Doing so causes a portion of the piping to be isolated from the expansion tank when the pump is off and the zone valves are closed. As that portion heats up or cools down being isolated from the tank, the pressure in it can go way up or down. When the pump starts and a zone valve opens, that pressure difference meets the pressure in the rest of the system - often with great force. That may be your problem.

Make sure that there's no flow check in the system pump.

HomerJSmith

I have never installed a condensate neutralizer in a vertical position.

rdbowden

The Grundfos was supplied by Knight with the unit. The heat exchanger has a pressure drop according to Knight. There are curves in the installation manual. The Grundfos is set a low speed and according to curves, pumping at a rate of around 7 GPM. The system was installed by an HVAC outfit hired by the previous homeowner.

Ironman

@HomerJSmith
The 26-99 is the correct circulator for that boiler and the one that Lochinvar supplied with it. The HX has a resistance of 27 ft. of head @10 gpm.

HomerJSmith

I looked it up, your right. 50 to 150 input btu is a UPS26-99F with a temp rise of 20 to 25 degs, 27'head. Factory provided. I'm surprised.

I find it hard to believe that an in volute check valve is rattling with a flow of 9.8 gpm less piping pressure losses. You need that check valve to keep the indirect pump from pushing into the primary sys.

By the way it seems that the instructions were followed, I mixed up the outlet and inlet of the boiler. The fire tubes are the opposite. I didn't realize it was a water tube. I will correct my previous post.

Perhaps a stand alone spring check would be quieter or shift the pump to the right.

GroundUp

Oooooh crap, yes I was missing the indirect piping for some reason. Check valve is necessary, sorry. Either way, it shouldn't matter one bit how much pipe is downstream of the circ. The upstream side is where things can get funky and cause a rattling check. Is this a constant issue? Like at all times the boiler runs, it rattles? Or just certain times?

HomerJSmith

Is the programing on that boiler set for the indirect on priority mode? It should be, I think.

Running the indirect at the same time as the space heating mode is operating may affect the check valve in the grundfos pump and maybe cause it to rattle.

Does the noise occur when only the space heating is operating or does it occur when both modes are operating together?

My thinking is that when both pumps are operating at the same time there is pressure from both pumps on both sides of the check valve causing the check valve to flutter. Just an idle thought.

Might check the pump wiring and sensor wiring at the control board for the correct connections.

rdbowden

I appreciate all the helpful comments. It is wired such the DHW, I think that is what you mean by indirect, is running, the Grundfos is off and the pump on the far right is running providing heat to a SuperStor indirect tank. It is extremely quiet in this mode. The DHW has priority, the heat will shut off the DHW is on. When the heat is on the Grundfos is on, one or more of the zone valves will be open and the system pump will run. When the thermostats shut off, the zone valve closes and the system pump shuts off. The Grundfos runs for about ten seconds or so and then shuts off. At all times when the Grundfos is running the rattle is there. I thought is might be due to the system pump and the grundfos running at the same time, but it isn't that.

I was just surprised that Knight would spec such a noisy solution and then that got me to wondering if the install was wrong. It seems like everything is okay and just try a mechanical check valve instead.

Ironman

I question whether the "rattle" is coming from a spring loaded plastic check valve? It may be possible, but it sounds like maybe something else.

Has anyone checked for debris in the impeller?

HomerJSmith

This is the second pump that was installed and is doing the same thing.

Go to your auto part store or Harbor Freight and get yourself a mechanic's stethoscope and put it on the pump and sys and see where the sound is coming from. Try and isolate the sound source.

Take more pics from different angles and further away and post them.

Your sys load may not be that great and the UPS 15-58FC might solve your problem.

You have a sys sensor installed on that sys and where is it installed. It should be on the output of the sys pump that goes to your zone valves where the water flow is turbulent.

infloorradiantheat

We had our boiler replaced in 2016. A Grundfos UPS 26-99FC was installed on the main boiler loop. We've had it looked at as to why it was making noise and was told it was the check valve rattling. At first we thought there was something in it so we flushed it. Didn't fix it. It was surprising as all our other circulators are quiet. The ones on the old boiler were quiet. Eventually we gave up looking into it.

rdbowden

The system sensor was never installed. I have attached a couple more pics. I guess I will just have to live with it.

Ironman

It would be very simple to remove the check and see if that stops the noise. Most likely you won't get it out without breaking it, but that's not a big deal: they're cheap.

rdbowden

At HomerJSmith's suggestion I took another look at the noise source. It is actually the heat exchanger. At the top left. I have attached a pic with the area circled in red. The strange thing is when the DHW pump is running, there is no noise in that area. The DHW pump is a three speed taco with integrated check valve. It was set at middle speed. I set it at high speed and still no noise. The grundfos is set at the slowest speed.

Ironman

I'd be curious as to what would happen if you swapped the position of the two pumps. What model is the Taco?

HomerJSmith

What is the model # of the Taco pump? I would like to compare the pump curves of the Taco and Grundfos.

So the noise is coming from the boiler output of the HX and not from the check valve in the Grundfos volute? I'm relying on your observation.

Yes the Grundfos pump is a 3 speed pump, I can see it on the name plate. The lowest speed may still be too much velocity.

I am still of the opinion that it is a velocity issue. Since you have a sys pump and heating manifolds, you may not need to suck every BTU out of the boiler. What is the supply temp and what is the return temp to the boiler output and the input. I'll wager that it is very, very close. How often does the boiler cycle, on the space heating side compared to the tank heating side?

rdbowden

The Taco pump is a 00R-MSF1-IFC. The supply and return temp are about 5 degrees apart once the system reaches equilibrium. The boiler does not cycle, It does modulate down to around 30 percent with one zone on, and 60 percent with two zones on. I did check the curves on the taco pump and if the Taco is set to high setting then it should match the Grundfos set to low. The Taco is about 3 feet further away and it is running the HX in the water tank. The primary loop for the Grundfos is very short. Knight specs for a 20 foot loop. So think you are right, water is just moving too fast.

archibald tuttle

been following this without a lot to add but i have developed some theoretical questions regarding the recommended piping schematic.

It does appear that the noise is isolated to, what is that, the feed manifold to the heat exchanger?

no moving parts. If I'm guessing right that is the beginning of the source of high head so you got lot of flow and higher pressure coming up against a restriction. If I mistake which manifold that is, then it is the outlet of a lot of flow freed from restriction. In the NFN department, these relatively large pumps are speced for the high output and i confess that is how i've pumped similar heat exhangers in other companies packages. But wouldn't it be better in theory to use a variable speed that matches more to the turndown output. That might not stop the noise at the highest output but, inter alia, with those low deltas and depending on what emitters you are using and temperature you are running for other than DHW you are never going to condense (so whether the neutralizer is best horizontal or vertical would be kind of moot.)

and if i'm reading your recitation of symptoms correctly you only get noise when the grundfos is running without the TACO?

I'm kind of assuming if its got DHW priority that the grundfos goes off when the DHW TACO circ goes on? and/or why wouldn't they have just piped the DHW secondary as well.

The more I see these systems the more I think an ondemand for DHW is a better choice. Can talk about whether the gas supply is sufficient for both turned up but for all the software we've got on these things why not just have an On Demand and Heating Boiler that talk to each other so the one ramps down if the other ramps up. then you could get more condensing out of both by optimizing the operation and the hydronic piping is less problematic.

Sounds like you still might have the problem insofar as noise at high speed of a variable pump and maybe the longer loop would help although what does the longer loop do really, except add a little bit more head (and tiny bit of heat loss to the utility room) thus slowing the water slightly for given pump speed, but what about just slowing the pump if it can keep up with the BTU's. Isn't that the same approach, again suggesting variable speed pumping, so long as the top end can handle the top BTUs.

maybe it is just a fact of nature that having all that restriction on the boiler side and very little on the system side (used to be the other way round in the old days) could be amended by a balancing valve, maybe even an automatically adjustable valve (although that is something else to break).

finally, assuming that the basic piping design stays more or less the same, on the check valves, which now don't seem as likely to be the problem, how much restriction is there to the non operating pump? I guess you might get some reverse flow through those systems. Just wondering about the head of a non-operating pump (in this case in reverse, not sure it would be that much different although maybe because maybe any flow would actually induce the impeller to turn, although I think the design of these impellers for efficiency at high speed operation wouldn't favor much of that.) I'm kind of wondering both for this configuration and for a totally separate project where i'm looking at using a circulator to move preheat through an engine and wondering when not in operation how much resistance it would pose in the heater line I intend to install it in.

a novel a day keeps the doctor away

brian

Alan (California Radiant) Forbes

Have you tried restricting flow to see if the noise goes away? You can experiment with that ball valve on the discharge side of the pump.

Fascinating thread!

HomerJSmith

Well, looks like a velocity issue. I would put in a UPS 15-58FC for the boiler pump and choose a speed that works for your space heating. That should increase the delta T by providing a better balance between the secondary boiler circuit and the primary space heating circuit.

You're not doing yourself a favor by running the boiler in this condition. Velocity issues have consequences.

GroundUp

HomerJSmith said:

Well, looks like a velocity issue. I would put in a UPS 15-58FC for the boiler pump and choose a speed that works for your space heating. That should increase the delta T by providing a better balance between the secondary boiler circuit and the primary space heating circuit.

You're not doing yourself a favor by running the boiler in this condition. Velocity issues have consequences.

How many times must it be said that his boiler requires a 26-99 by the MFG? 15-58 is too small for the necessary flow rate and will not allow proper flow through the HX, causing spot boiling.

rdbowden

Yesterday I had a plumber in from a plumbing company recommended by Lochinvar. It is a large outfit and he specializes in the Lochinvar installs. He said the pump was too big for this install he also said the pump was in bad shape, which surprised me, since it was brand new 4 years ago. Makes me wonder if the previous plumber billed me for a new pump but only replaced check valve. Just sayin. He replaced the old grundfos with a Taco 007e. The boiler is now completely silent. Heat output is the same or a little better, the boiler runs at about 33 percent rate per zone. Used to run at 31 percent. I have attached the pressure drop versus flow tables from the manual. It looks like an 007 is allowed. It is listed under the 35 temp rise chart. These charts are based on 20 feet of piping, my install is nowhere near 20 feet. The model is the WB-105. Thank you all for your suggestions.

Alan (California Radiant) Forbes

It looks like the 007e just squeaks by.

HomerJSmith

Gee! So it was a velocity issue.

I'm not a fan of putting an ECM pump on the boiler secondary piping. I know it saves electrical energy, but I'm concerned about the flow. The chart suggest a fixed speed circulator. Perhaps, my understanding is incomplete.

The reason one installs a condensate neutralizer in a horizontal position is that it take time to neutralize the condensate. It doesn't happen instantly which is why they are designed to hold the condensate before it drains out of the neutralizer. When it is installed vertically the water just runs thru the neutralizer.

rdbowden

When the plumber replaced the circulator, he replaced the neutralizer and condensate pump with a combined unit. A PurePro Condensate neutralizer pump.

hot_rod

Keep a close eye on it, my Knight would go into lock out occasionally with a 15-58, "excessive delta" error code. You may notice it when it goes into high temperature mode, for an indirect call for example.
I assume it was a single speed 26-99? The 3 speed would run down close to a 15-58 on speed 1.