Replacement circulator help?

northlakes

I’ve been trying to find a backup pump to replace a discontinued Grundfos UP 43-75F (sounding like it’s going downhill) and getting lots of different recos from distributors. Confused. Trying to avoid repipe and replace with same port to port length. 

2 floors 2000 sq ft of hydronic above boiler, another 1500 sq fr on boiler level. Primary secondary.  

They still make the same pump in bronze but seems like unnecessary expense. Ups43-100f is a 3 speed that appears to consume more power at lowest speed than my current pump… which is probably already a tad oversized. 

Robert_25

Is this the primary circulator on your boiler? If so, what does the boiler manufacture recommend?

Supplyhouse.com has a recommended replacement for your 43-75F: https://www.supplyhouse.com/Grundfos-52722373-UP43-75F-Circulator-Pump-1-6-HP-115-volt-5583000-p

GGross

43-100 will be the same port to port length, but is a bit more pump than the 43-75

26-99 is probably a better fit, but is a different port to port length.

I think if you were to go to the trouble of repiping a different sized pump in there it would be worth your time to double check the pump requirements of your system. 43-100 would be the easiest to just swap out though.

northlakes

GGross said:

43-100 will be the same port to port length, but is a bit more pump than the 43-75 26-99 is probably a better fit, but is a different port to port length. I think if you were to go to the trouble of repiping a different sized pump in there it would be worth your time to double check the pump requirements of your system. 43-100 would be the easiest to just swap out though.

Thank you, yes, 43-100 is oversized for what I need. Looks like a repipe will have to happen. 

To be honest I’ve struggled with circ sizing calculations:

- If supply and return pipe are larger 1” and 1 1/4” pipes but loops drop to 3/4”, does that mean system is calculated as 3/4”? 

- To calc head what is meant by longest run? Longest possible distance in pipe from circ back to circ? *EDIT: Is TDL all pipe in system not just the longest run?

- When calculating gpm using btu, I get confused as my boiler has a 10x turn down and seldom runs at full. Do you size it for design day max output of boiler? 

EBEBRATT-Ed

You size for maximum output.

If there is more than one piping loop or zone you only need the head loss of the zone with the most resistance (usually the longest zone.

Once you have the resistance for the pipe in the longest zone you ad 50% to that # to allow for fittings. You have to size for the complete loop supply and return leaving the boiler and returning to the boiler. You size the circulator for the GPM you need for all zones but only use the head Resistance) of the zone with the highest head.

northlakes

EBEBRATT-Ed said:

You size for maximum output. If there is more than one piping loop or zone you only need the head loss of the zone with the most resistance (usually the longest zone. Once you have the resistance for the pipe in the longest zone you ad 50% to that # to allow for fittings. You have to size for the complete loop supply and return leaving the boiler and returning to the boiler. You size the circulator for the GPM you need for all zones but only use the head Resistance) of the zone with the highest head.

I think I finally get it, thanks! Flow is cumulative, head is just max resistance…? 

If my longest loop has tees at supply and return, would the calc be whole perimeter of the loop or just longest run between the tees? When calculating head, does pipe size matter? About 45’  of a 150’ run (supply and return runs nearer to boiler) is 1 1/2” and 1 1/4“ pipe, baseboard runs are 3/4” does that figure into head loss?

Still a bit confused on pipe size and flow. My calc says flow would be 16.4 GPM (I used 500 btuh x 315’ of emitter plus a tad extra for quite a few bare copper pipe runs in enclosures.) how does this calc relate to max flow of 4 GPM in 3/4” copper pipe?

northlakes

Robert_25 said:

Is this the primary circulator on your boiler? If so, what does the boiler manufacture recommend? Supplyhouse.com has a recommended replacement for your 43-75F: https://www.supplyhouse.com/Grundfos-52722373-UP43-75F-Circulator-Pump-1-6-HP-115-volt-5583000-p

If I understand the terminology correctly it’s the secondary circulator (for 3 zones with valves)

pedmec

16.4 gpm would be your overall required flow to satisfy the load demand and keep your desired delta t. the 4 gpm is the max velocity recommended to keep velocity noise down and erosion within the pipes.

you should not have a piping system that has a single loop of 315' of 3/4 copper. you will have a pitting in the copper piping and thinning of the copper walls.

northlakes

pedmec said:

16.4 gpm would be your overall required flow to satisfy the load demand and keep your desired delta t. the 4 gpm is the max velocity recommended to keep velocity noise down and erosion within the pipes. you should not have a piping system that has a single loop of 315' of 3/4 copper. you will have a pitting in the copper piping and thinning of the copper walls.

That’s the total emitter in all zones (4 loops, 3 ZVs, one zone loop is always open for freeze prevention.) Does gpm accumulate with each loop? 

northlakes

GGross said:

43-100 will be the same port to port length, but is a bit more pump than the 43-75 26-99 is probably a better fit, but is a different port to port length. I think if you were to go to the trouble of repiping a different sized pump in there it would be worth your time to double check the pump requirements of your system. 43-100 would be the easiest to just swap out though.

Assuming a repipe:  

@GGross I had read somewhere @hot_rod said UPS 26-99 wasn’t a great match for zone valves (this system circ pumps through 3 of them) so I was also looking at the Alpha 26-99. 

Current calc says flow would be 16.4 and max head 9.

- To calc flow I’m tallying total feet of fin tube emitter, = 313’ over 4 loops (one is always open for freeze prevention) (313*500) / (500*20) = 15.6, I added a bit more for several bare pipe runs through baseboard enclosures, not sure if that’s correct to do but seemed to make sense, = 16.4 gpm

- On longest run to calc max head I’m assuming if a loop is “split” by a tee at either end, I’m only including the longest “side” of that run, and not the full perimeter of that tee’d loop? Not sure how/if pipe size figures in to this run, all baseboard is 3/4 but about 50’ of supply and return pipe nearer boiler is 1 1/2” and 1 1/4”. Just accounting for length alone that longest run comes out to 150’  x 1.5 x .04 = 9 max head 

When I see an oversized pump like this on an older system I guess I wonder if it was needed for something I’m not considering. E.g. would it be harder for air to be pumped through and removed with a smaller pump? 

hot_rod

Idronics 16

https://www.caleffi.com/sites/default/files/media/external-file/Idronics_16_NA_Circulation in hydronic systems.pdf

This issue takes you through the steps to selecting the best circulator. In an existing building you have to make some assumptions if you cannot measure all the pipe and count fittings

If you have multiple zones with zone valves, a delta p type ECM will help prevent over pumping and excessive velocity

northlakes

hot_rod said:

Idronics 16 https://www.caleffi.com/sites/default/files/media/external-file/Idronics_16_NA_Circulation in hydronic systems.pdf This issue takes you through the steps to selecting the best circulator. In an existing building you have to make some assumptions if you cannot measure all the pipe and count fittings If you have multiple zones with zone valves, a delta p type ECM will help prevent over pumping and excessive velocity

Thanks for that, very informative.

With my calculated results I’m finding either  “high head” or over/underpowered pumps that would supposedly not be a good match for zone valves, as well as info that says ECMs get quickly clogged up with debris due to the magnetic nature, have unexpected failures (maybe user error?) due to their sensing features, etc.

Maybe at the end of the day there’s just not a great match for this kind of system!
 

Robert_25

northlakes said:

Robert_25 said:

Is this the primary circulator on your boiler? If so, what does the boiler manufacture recommend? Supplyhouse.com has a recommended replacement for your 43-75F: https://www.supplyhouse.com/Grundfos-52722373-UP43-75F-Circulator-Pump-1-6-HP-115-volt-5583000-p

If I understand the terminology correctly it’s the secondary circulator (for 3 zones with valves)

Can you post a picture of the boiler and piping?  

hot_rod

The circ has to be able to do the job, regardless if it is a high head. It’s really more of a flat curve circ that you want for zone valves, if it is a single speed type circ.

Back when cast boilers with zone valves were the standard system, a low head, flat curve circulator like the B&G 100 was a great match.

With todays electronic type circs, they can adjust somewhat to the changing requirements of the system as zones open and close.

northlakes

hot_rod said:

The circ has to be able to do the job, regardless if it is a high head. It’s really more of a flat curve circ that you want for zone valves, if it is a single speed type circ. Back when cast boilers with zone valves were the standard system, a low head, flat curve circulator like the B&G 100 was a great match. With todays electronic type circs, they can adjust somewhat to the changing requirements of the system as zones open and close.

Thank you, I’ve been looking for a flat curve on the Grundfos pumps. Started looking at the delta p options like Alpha, was hesitant after reading about erratic performance and that I’d need to add some kind of magnetic trap… Started looking at the Taco 00s and the 0012 curve seemed to intersect without much room for error. 

northlakes

Not visible here is a zone controller, backup boiler (not in use) and an 1 1/4" pipe which feeds zones 2 and 3. Pump that needs replacing is the red one.

Robert_25 said:

northlakes said:

Robert_25 said:

Is this the primary circulator on your boiler? If so, what does the boiler manufacture recommend?

Supplyhouse.com has a recommended replacement for your 43-75F: https://www.supplyhouse.com/Grundfos-52722373-UP43-75F-Circulator-Pump-1-6-HP-115-volt-5583000-p

If I understand the terminology correctly it’s the secondary circulator (for 3 zones with valves)

Can you post a picture of the boiler and piping?  

northlakes

Below is how I'm assuming I should measure for longest run in loops with tees. Is this correct that I'd just measure the path through one side of the tees?

hot_rod

What size is the boiler? It looks like the boiler has a circ, so it is piped primary secondary? We need to see the piping above the boiler.
The 43-75 is a pretty good sized circ, 1/6 Hp, 215W
I’d be surprised the you need that much circ. No need to pump more than the boiler can put out.
A 150K boiler, 15 gpm perhaps?
Two things you need, gpm required, and the head of the distribution.

northlakes

hot_rod said:

What size is the boiler? It looks like the boiler has a circ, so it is piped primary secondary? We need to see the piping above the boiler.
The 43-75 is a pretty good sized circ, 1/6 Hp, 215W
I’d be surprised the you need that much circ. No need to pump more than the boiler can put out.
A 150K boiler, 15 gpm perhaps?
Two things you need, gpm required, and the head of the distribution.

The boiler is oversized at 199k, according to installer the next one down would have been marginal for design, and the turn down can account for the extra size (I understand this isn't always ideal for efficiency of condensing etc.) Current circ appears to be significantly oversized.

It's piped as primary secondary, attached are more images. My calculations for flow came to 16.5, and max head 9. That's if my understanding about how to measure longest run with a tee that "splits" the loop is correct...(posted that sketch above in the thread) If I measure the whole loop perimeter from both sides of the tees, it's 13. I was also unsure about the calcs because a fair amount of that piping length before and after the 3/4" fintube loops is larger pipe, 1", 1 1/4", 1 1/2".

hot_rod

Looks like Taco has a circ that matches the 16 gpm @ 9', the 0010. You might find an ECM type in that range also.

northlakes

hot_rod said:

Looks like Taco has a circ that matches the 16 gpm @ 9', the 0010. You might find an ECM type in that range also.

Thanks! For whatever reason the chart I have from Taco seems to show different curves?

I was also looking at the 0012. If the curve extends far beyond max flow like this, is that undesirable?

Also saw the 0012 3-speed variation.

hot_rod

Ideally you hit the middle 1/3 of the curve
medium speed on that model looks close

This all assumes your calculations are accurate

northlakes

hot_rod said:

Ideally you hit the middle 1/3 of the curve
medium speed on that model looks close

This all assumes your calculations are accurate

This 3-speed seems close if I'm calculating that longest run correctly. If not, I'd guess the higher speed would probably cover it.

I'd be happy to go ECM if I could get a sense for whether or not I'd need some kind of filtering system to avoid the magnetic bits getting crudded up. I'm all for efficient tech but if it's less reliable and more complicated I'm less inclined

Teemok

0012 on med seems right. I hope the branch loops in this circuit are well balanced. If you split 16 gpm by 4 paths and they are all 3/4 copper, the water velocity in a perfectly balanced loop is 2.9 ft/sec. If they are imbalanced one could be flowing close to the limit of what you want in 3/4 copper.

northlakes

Teemok said:

0012 on med seems right. I hope the branch loops in this circuit are well balanced. If you split 16 gpm by 4 paths and they are all 3/4 copper, the water velocity in a perfectly balanced loop is 2.9 ft/sec. If they are imbalanced one could be flowing close to the limit of what you want in 3/4 copper.

Thank you. @Teemok I don't think I understand what is meant by well-balanced branch loops, could you explain or link me to something to read?

Zman

I honestly haven't tried this but you might be able to make up the flange to flange dimension by tossing the check that comes with the circ an putting in one of these. https://s3.amazonaws.com/s3.supplyhouse.com/product_files/1370-sub.pdf

Teemok

In your drawing, If you follow the supply water out of the boiler it meets three Tee's. At each Tee the water will follow the path of least resistance. More water will go the shortest path until the increased flow creates more head until the two paths equalize in pressure at the tee but the branches will be unequal as far as flow rate is concerned. 16GPM splits in a perfect world to 8GPM each branch and then those branches both split to 4GPM. Because of piping differences/imbalances, each split will have a branch that has more flow that the other. Say 10 and 6. When the 10 splits, it's shortest path might take 7 gpm while the longer one only gets 3GPM. 7GPM in 3/4 copper is moving at 5.08 ft/sec and that's over what you want. I'm not saying this is happening. A slight imbalance is typical. Good designers and installers try to make the paths equal.

Teemok

A ups26-99fc 3 speed, a coupling, chunk of pipe and male adapter is the value move. Are the loops you drew zoned with zone valves or is it all a single zone. If it's zoned an ECM pump is a way better choice. The 0026e is a good pump and it has a black iron oxide barrier for whatever that's worth.

northlakes

Teemok said:

A ups26-99fc 3 speed, a coupling, chunk of pipe and male adapter is the value move. Are the loops you drew zoned with zone valves or is it all a single zone. If it's zoned an ECM pump is a way better choice. The 0026e is a good pump and it has a black iron oxide barrier for whatever that's worth.

It would be the pump for 3 zone valves, with 4th loop always open for freeze protection. Do I need to add some kind of magnetic filtration trap when adding an ECM to an existing system? 

northlakes

Zman said:

I honestly haven't tried this but you might be able to make up the flange to flange dimension by tossing the check that comes with the circ an putting in one of these. https://s3.amazonaws.com/s3.supplyhouse.com/product_files/1370-sub.pdf

That’s interesting, thank you, looks like 2 9/16” 

Teemok

The industry says yes a add a mag. filter with ECM pumps. That Taco pump has a filter on it meant to protect just the pump but the juries still out on the effectiveness. It's flange to flange is 6.5" You said you need 16GPM@ 8ft of head. The pump will do that and more if needed. But if you are zone valving you will use the constant pressure mode and it will ramp up and down as needed to maintain a set pressure range as zones open and close. It changes itself to meet the needs of the system as it changes without wasting electricity pumping for nothing.

Teemok

So 3 zones controlled by a thermostat that opens a zone valve for those areas and one that is always calling for heat firing the boiler? Or it just gets flow when one of the other zones is calling?

northlakes

Teemok said:

So 3 zones controlled by a thermostat that opens a zone valve for those areas and one that is always calling for heat firing the boiler? Or it just gets flow when one of the other zones is calling?

3 zones with tstats and ZVC, and a 4th loop in first floor where the ZV was removed, and my guess at this was freeze protection since there’s a pipe run across a garage door in a trench in the concrete floor and pump is currently wired to run continuous.

I’d like to modify this so that it only runs continuous <30 outdoor temp. One step at a time 

northlakes

Teemok said:

In your drawing, If you follow the supply water out of the boiler it meets three Tee's. At each Tee the water will follow the path of least resistance. More water will go the shortest path until the increased flow creates more head until the two paths equalize in pressure at the tee but the branches will be unequal as far as flow rate is concerned. 16GPM splits in a perfect world to 8GPM each branch and then those branches both split to 4GPM. Because of piping differences/imbalances, each split will have a branch that has more flow that the other. Say 10 and 6. When the 10 splits, it's shortest path might take 7 gpm while the longer one only gets 3GPM. 7GPM in 3/4 copper is moving at 5.08 ft/sec and that's over what you want. I'm not saying this is happening. A slight imbalance is typical. Good designers and installers try to make the paths equal.

When I'm calclulating max head for longest run, I'm only including the length of the path in red in my drawing. Is that correct, or should I be adding in both sides of the split loop?

Teemok

Just the longest path considering all the restrictions to flow (valves, fittings, tubing size and lengths radiation size and type). The drawing is an abstract representation that doesn't include the amount of radiation or number of fittings or sizing. When lots of radiation is put in series on the radiation at the end of the loop can see much lower water temperatures. A designer might chose to increase the flow rates in that loop to counter that but most of the time balanced design is what we want. With zone valves you don't need to worry about it most of the time. Sounds like an ECM is a good fit. That frost control constant pumping and I assume with that constant pumping the boiler has a constant call for heat. The flow rate and heat load of that wild frost loop might have that boiler cycling a lot. Do you have glycol in your system? A power outage might mean a fast frozen pipe in a that trench.

northlakes

I mean to ask does a split loop get added in for length calculation on both sides or just one side (a fish would only swim one side of the split loop to get back home...) A significant difference in head calculation depending on the answer to that.

There's no constant call for heat, (about once an hour, ODR curve could prob be improved) current pump just circulates the open loop at whatever leftover temp. Yes a power outage would be bad all around, no glycol, we have a generator...

Teemok said:

Just the longest path considering all the restrictions to flow (valves, fittings, tubing size and lengths radiation size and type). The drawing is an abstract representation that doesn't include the amount of radiation or number of fittings or sizing. When lots of radiation is put in series on the radiation at the end of the loop can see much lower water temperatures. A designer might chose to increase the flow rates in that loop to counter that but most of the time balanced design is what we want. With zone valves you don't need to worry about it most of the time. Sounds like an ECM is a good fit. That frost control constant pumping and I assume with that constant pumping the boiler has a constant call for heat. The flow rate and heat load of that wild frost loop might have that boiler cycling a lot. Do you have glycol in your system? A power outage might mean a fast frozen pipe in a that trench.

Teemok

You have that right about the split. You're looking for the longest branch splits to make sure the pump can move the right amount of water through it at the same time as it moves water through all the other paths. My uncle used put balancing valves on the short branch returns. Using the very imprecise method of feeling for hot water on a cold start up with all zones calling we would restrict the short branches until they got warm at about the same time as the long ones. There are far more precise methods for balancing. Like I said constant pressure and zone valves makes any imbalances negligible. The fastest loop to heat the return pipe with all zones open on a cold start up is the one to restrict first if you want to try to balance or test the branch balances.

northlakes

Teemok said:

You have that right about the split. You're looking for the longest branch splits to make sure the pump can move the right amount of water through it at the same time as it moves water through all the other paths. My uncle used put balancing valves on the short branch returns. Using the very imprecise method of feeling for hot water on a cold start up with all zones calling we would restrict the short branches until they got warm at about the same time as the long ones. There are far more precise methods for balancing. Like I said constant pressure and zone valves makes any imbalances negligible. The fastest loop to heat the return pipe with all zones open on a cold start up is the one to restrict first if you want to try to balance or test the branch balances.

Thank you for all this advice, very helpful... Tempted to go ECM but as you say the jury is out and not sure a mag filter is an option at the moment. Will look into it more.

Teemok

If you don't go ECM, you will have a single speed pump big enough for all flow responding to a single short loop calling for heat alone. The velocity of flow will likely be in excess of good design in that loop. The wild loop helps counter this situation. Designers use a spring loaded valve called a pressure bypass valve that lets excess supply flow bypass back to the return. The setting of the bypass the same as the longest loop head pressure. Anything over that pressure and the valve begins to bypass reducing flow in the shorter loops. The ECM pump is the elegant solution. If you have an existing black iron oxide problem than a filter is needed. If not it the "need" for a filter is debated. If the pumps is the only iron in the system it will be the sole thing that corrodes due to corrosive factors in the system. It is the target. The $242 Caleffi mag-filter is a worthy investment. It defiantly costs less than removing all iron and going with stainless pumps.

northlakes

Teemok said:

If you don't go ECM, you will have a single speed pump big enough for all flow responding to a single short loop calling for heat alone. The velocity of flow will likely be in excess of good design in that loop. The wild loop helps counter this situation. Designers use a spring loaded valve called a pressure bypass valve that lets excess supply flow bypass back to the return. The setting of the bypass the same as the longest loop head pressure. Anything over that pressure and the valve begins to bypass reducing flow in the shorter loops. The ECM pump is the elegant solution. If you have an existing black iron oxide problem than a filter is needed. If not it the "need" for a filter is debated. If the pumps is the only iron in the system it will be the sole thing that corrodes due to corrosive factors in the system. It is the target. The $242 Caleffi mag-filter is a worthy investment. It defiantly costs less than removing all iron and going with stainless pumps.

I've looked at a couple of models, some look more limited to horizontal runs like these:
5463 Series


Others have options for vertical or horizontal:
NA545356



With the tangled piping near boiler it looks like the vertical right ahead of boiler circ would be an easier job, options look limited on the horizontal

mattmia2

The head of larger diameter pipe per foot is less, you should calculate it in sections.

You would only use the longer half of the split loop, that is the idea of the split loop, it can keep the head lower with smaller pipe by feeding the loop from the middle.