Circulator Pump size and Qty

mpdport

Pump help?
Just trying to validate what is happening on my install. 
have a new radiant system going into an new structure attached to my main house. 

NE cold climate.  
Plumbed already. 

5 radiant loops approx 1000’ of 1/2 pex spread across 3 zones.  All on the 1st floor. 
Ecowarm radiant heating boards. 
5 port manifold
Plus long 90’ run of 1” pex to/from the manifold = 180’
So close to 1200’ of pex

Can this all run on one pump.  

Currently a Grundfos Alpha1 15-55F. 

I think they are seeing low flow issues 

I don’t know how to calculate….. (I am the homeowner)
do I need bigger pump ?   Can we add another in series, near the manifold, after the long 1” run to the boiler?

Any high level advice would be appreciated 
Thanks !

Big Ed_4

Are some sections heating better than others ?

mpdport

There is a second pump with only about half the pex length on it.  It is doing pretty well. 

All 5 loops on the 5 port manifold seem to be lacking flow. 
System has Been purged 

EdTheHeaterMan

Pump sizing is fairly simple. You just need to know a few factors.

1. What is the design temperature of the radiant system?
2. What is the design temperature drop of the system? Sometimes called ∆T
(if that information was not established by the folks designing the radiant system, then you will need to figure them out by way of the heat loss of the space, and the amount of insulation below the tubing and the material above the tubing. Basically how is the floor heat designed.)
3. That will determine the gallons per minute of each loop of heating.
4. Then you add up all the GPM of the 5 loops. That will give you the GPM capacity the pump must provide.
5. Then you use a basic calculation to determine the pump head requirements. That is the total friction that must be overcome to move the water through the pipes.
6. Then you select a pump that will provide the needed GPM at the pump head required by looking at the performance curve of the pump you are selecting to determine if the pump hac the necessary capacity to do the job.

Come to think of it... that sounds a little complicated for the novice.

EDIT: What pump are they using? And to be clear there is a 1" pex line from the boiler to the manifold and all of the zone loops are at 200 feet or less, Are there other systems connected to that boiler? High temperature radiators, air handler coils, or domestic hot water?

EdTheHeaterMan

Based on the total length method of determining pump head, you need a pump that will have the necessary flow rate at 22 ft of head. That sounds like your pump is too small for the application. The Alpha1 15-55 has a maximum head of 19 feet.

You need a Alpha1 26-99F

Zman

What are the symptoms? What is the difference between your supply and return temps?How is this piped? Primary/secondary?

hot_rod

So pumping from a boiler or header to manifolds? Like the attached drawing?

If you want .75 gpm per 5 loops = 3.75 gpm is what you are trying to move.
200' of 1/2" pex at .75 gpm flow = 4' head

Call the S&R 200' 1" pex, adding some fittings
200' 1" pex at 4 gpm= 5.05'

So 4 gpm at 5' is what I come up with. The Alpha is plenty?

The free calculator at PPI is handy for calculating pressure drops

A heatload and design would tell you how many gpm/ 1/2" loop .50- .75 gpm typically.

Jamie Hall

One thing which isn't always obvious: the proper length of pipe to consider when figuring head loss is the longest individual loop -- not the total of all the loops. Your loops probably -- or should -- average around 200 feet each, so your total length of pipe to consider is about 400 feet, not 1200 feet. On the other hand, for flow, you need to figure the total of the flows in all the loops.

And one more thing: pumps in series are not a good idea. Better to have one pump with the head and flow capacity needed.

mpdport

Thank You
You all are some very kind people, who are very generous with your knowledge!
Here are the lenghts of the Loops, calulated by the EcoWarm designers
The BTU values came from the heating contractor.
The Viessmann Gas Boiler also has high temp water going to the the 2nd Floor HW baseboards, and the DHW tank. The mixing valve for the radiant is controlled by the boiler. Not sure of the Delta T.
Sounds like we need a bigger pump.....

hot_rod

Do your manifolds have flowmeters to see how many gpm each loop will flow?

Here is an example of an Uponor design sheet. It indicates the flow required to each loop, and gives you the pump spec gpm and head. Also useful info like required supply temperature, notice the carpeted rooms in this example needing high temperature SWT.
Do you have a sheet with flowrates and pressure drop indicated?

Call your longest loop 253" (243' loop plus 10' of leader.)
At .75 gpm headloss is 5.8 ft so that is the highest pressure drop loop

What we don't have is the gpm required for each of your loops, but with a total load of
30,281 BTU/ hr for that manifold, you can work the numbers backwards.

Here is a pump curve for a 15-55 Alpha. At the Grundfos site you can open this and move your actual requirement on the curve and it gives you the data.

So at a 5 gpm flow (5 loops at .75 gpm) you have 13.43' of head available.

Since you have the actual lengths no need to use the Rule of Thumb sizer that Ed showed which could oversize pump by quite a bit.

A pic of boiler piping would help see if for some reason you have a high pressure drop there? Like a tankless water heater

Alan (California Radiant) Forbes

The Grundfos Alpha I pump is ideally suited for your application. The urge to fix your problem by installing a bigger pump is a common mistake. There's something else going on.

Is the pump installed downstream of the mixing valve? If the Viessmann boiler is controlling the mixing valve, has the correct information been programmed, i.e. heating curve? Has the outdoor sensor been installed?

Pictures of your system would help.

GroundUp

What is giving the illusion that there is not sufficient flow? Are there flowmeters on each loop? The 15-55 is more than adequate to supply triple the loops that you have here, it is not too small. Are they running it on constant speed mode, or AutoAdapt perhaps? If the latter, it should be changed to either constant speed or constant pressure. WHat is being used for a heat transfer fluid? If glycol, perhaps it needs to be diluted? What type of manifolds are being used? What model mixing valve? Does the mixing valve have screens? There are many variables here that may be causing an issue (if there is actually an issue) but zero of them are an undersized circulator.

mpdport

You all are asking some questions I cannot answer. But I appreciate the.
I am currently in another state from the construction.
I will try and answer a few of them. To see if they give you any other clues.
I think the plumber has done a beautiful job on the pumps, pipes, valves at the boiler. Pix attached.
The crude simplified sketch is mine. Please note the 90’ run to Manifold #1. So 180’ round trip. It is in a whole new structure. Does this need to be added into the length of the longest loop and the head calculations for Manifold#1? I think there is 40% glycol in system ( does that sound correct?). The plumber noted he was not seeing the flow the way he expected on Manifold #1. Flow gauges on each manifold loop, temp gauges on both supply and ret side. The pix shows the Mixing Valve before the motorized housing went on. Not sure about MV screens. Not sure if the outdoor reset is hooked up yet. To not “a shock” the engineered wood floor (that had been cold for months), the plumber has been bringing radiant temp up slowly. I think it is only at 85degrees so far. Uponor manifolds. Separate ports off the Viessmann feed the 53 gal Viessmann DHW (this is not going yet). Thanks again for looking at this.

hot_rod

 “The plumber noted he is not seeing the flow he expected
What is he seeing, and what is he expecting?

mpdport

Hard to say. As I am not directly asking him. From a distance. I am exercising a “trust but verify” mode. He is telling me some of what he is finding. I am trying to verify is he on the right path. I think he said he is getting less than half the flow (and heating ability) that he is getting on the other Manifold #2. Which has a very short run to the boiler and only about half the length of total per hanging off it.

hot_rod

But we are all missing  the verify part if this quest?
How is he confirming he is getting 1/2 the flow?

we don’t know how much flow he should have, or how much he does have?

Throwing a larger pump on is not a trust builder, in my mind. So the trust and verify concept seem suspect 

with the info we have it should work fine. A flow restriction, kinked tube, valve partially open, something stuck in the S&r piping.

Derheatmeister

The Piping looks clean...However:
When Auto Delta T/Delta P "Smart" Circulating pumps are piped such as your system has been piped issues arrise due to circulating pumps seeing and fighting each other..
Do you have a Low Loss Header (Black Device with 4 ports and a sensor well ) on the left hand side mediating between the (Viessmann?) boiler and the distribution side ?
Are the Flow checks installed in the Circulating pumps?
Personally i would have :
1. Piped it 1 1/2" off of the LLH to be reduced by a 1 1/2" x 1 1/4" x 1 1/4" tee
2. Installed >>>One<<< Circlating pump on the Low temp/Mixing valve side and have the zones controlled via Zonevalves or TRV's.
3. Installed >>>One<<< Circulating pump on the High temp side.and again and have the zones controlled via Zonevalves or TRV's.
just my 2 cents.

Teemok

There's a 4 way mix on the drawing but a 3 way in the photo?


5.2Ft for 180ft of 1" @ 4gpm + 6.67ft for .75gpm @ 250' 1/2" = 11.8ft Ok but added restrictions could approach the pumps limit and expecting much more 1.2 gpm per loop is a disappointment in advance (my definition of an expectation)

No other flow restriction factors? Fittings in the 1" pex?
Have the individual loop flows been balanced at manifold flow meters for .75 gpm each with the pump on fixed high speed?
Loops under 200' could be closer to .6 gpm

hot_rod

Derheatmeister said:

The Piping looks clean...However:
When Auto Delta T/Delta P "Smart" Circulating pumps are piped such as your system has been piped issues arrise due to circulating pumps seeing and fighting each other..
Do you have a Low Loss Header (Black Device with 4 ports and a sensor well ) on the left hand side mediating between the (Viessmann?) boiler and the distribution side ?
Are the Flow checks installed in the Circulating pumps?
Personally i would have :
1. Piped it 1 1/2" off of the LLH to be reduced by a 1 1/2" x 1 1/4" x 1 1/4" tee
2. Installed >>>One<<< Circlating pump on the Low temp/Mixing valve side and have the zones controlled via Zonevalves or TRV's.
3. Installed >>>One<<< Circulating pump on the High temp side.and again and have the zones controlled via Zonevalves or TRV's.
just my 2 cents. </p>

In the left side of the pic it looks like the Viessmann header? Shadow of an air vent up high?

mpdport

I think what you saw on left is the big black insulated box. I think for air purge ?
I am not sure the answer to most of the other recent questions

GroundUp

That "insulated box" is indeed a LLH. We need some information about the manifolds and circ mode in order to answer your questions. Either the loops need to be balanced, or the circ mode(s) needs to be adjusted.

Derheatmeister

LLH it is !...
What size(In BTU/HR or KW) and which model is the Viessmann ? (100 or 200 series )

Again....The distribution side of the piping arrangement is Nice looking but not very functional/ Practical.
Overseas we call this a Pumpaholic arrangement which even worse with the Grundfos Alpha causes all kinds of Flow related issues,Comfort level Issues,Increases your Carbon footprint and energy consumption.

No one likes to hear this but if we walked into a piping arrangement such as this we would repipe on the distribution side (Right of the LLH) with one circ on the high one circ on the Low side/Zone valves..

If you want to continue down this piping arrangement path you could try to Install the Flowchecks on the circs..Dumb down the circs to a fixed speed, install and adjust flowsetters on each of the Baseboard side and adjust the flow to each loop on the manifolds for the best Delta t .

Hope this helps..

Zman

Your system seems to be lacking purge points to remove air. Is it possible that some loops are airlocked?
A common mistake when mixing is an incorrectly sized mixing valve. Do you know the model number (specifically the CV rating)?

hot_rod

The header coming off the right side of the LLH needs to be sized to the total of all the circulator gpm attached to it
Ideally at 2 fps velocity. That way there would never be pump fighting or interference.

I prefer a single  pump with zone valves  but zone pumps are workable. 

Your drawing and design sheet shows a 3 and 5 port radiant manifold, but the pic shows 3 circs off the 3 way valve? Is there another low temp zone?

Again we need better info to help out. What is the total gpm the 3 way valve needs to supply?

It is a vane type mix valve, I suspect the cv is fairly high, it should be adequate and not present a high pressure drop

The attachment shows  how pumps can cause flow reversal if not checked 

EdTheHeaterMan

Total length of longest run = (loop 2 is) 243 + 90 (to the manifold) + 90 (back from the manifold) = 423 feet
Add 50 % to this: 423 + 212 = 635 total equivalent length.
Multiply this by .04. . . . . . . . . 635 x .04 = 25.4

25.4 sounds like it is more than 19 feet of pump head.

Now this is a rule of thumb and by no means exact but it is close! Did I do the math wrong? The Alpha 1 15-55 has a maximum head of 19. So how is that pump NOT too small? I picked the Alpha1 26-99 because it has a max head well above the 25 required. there may be another that is less expensive that may work. I just know that one will work on setting 1

hot_rod

EdTheHeaterMan said:

Total length of longest run = (loop 2 is) 243 + 90 (to the manifold) + 90 (back from the manifold) = 423 feet Add 50 % to this: 423 + 212 = 635 total equivalent length. Multiply this by .04. . . . . . . . . 635 x .04 = 25.4 25.4 sounds like it is more than 19 feet of pump head. Now this is a rule of thumb and by no means exact but it is close! Did I do the math wrong? The Alpha 1 15-55 has a maximum head of 19. So how is that pump NOT too small? I picked the Alpha1 26-99 because it has a max head well above the 25 required. there may be another that is less expensive that may work. I just know that one will work on setting 1

Why add 50%? There are no fittings in the radiant loops. I’d imagine. 

And what is the .04 based on? Why not .03 or .02😗

Also the S&r is 1” pex, much lower pressure drop, especially at a 3 gpm flow rate

When you have all exact (as built) lengths you can get the exact answer, why use a “padded” rule of thumb?

2” and  larger pipe is sized by pressure drop/ hundred feet for example. Smaller pipe is sized by velocity, 4 fps max. for hydronic work. The rule of thumb may work on larger sizing and unknown fittings in a circuit?

I ran the numbers with the  PPI calculator, the Uponor design sheet, and checked with my Circuit Syzer wheel. The wheel was a tad bit higher head, maybe the Carlson fudge factor

My shop has seven 250’ loops running on a small Alpha

mpdport

Well, I feel like I opened a can ‘o worms. I’ll address a few things. I pushed for the alpha pumps as they seemed considerably more efficient than the previous generations. is this not true? If it does not take too long to describe - what is the cause of the negative interactions from one alpha to the others. Is it moslty due to needing flow checks? Is the issues moslty the undesired mixing of supply and return water? It they were added, would they have to be on each pump, Or on manifold loops ?
I have admit, I groaned heavily at the thought of redoing big portions of this… if you only knew how long to took to get to this point. Where we are located, there is a severe lack of trades and those that are there are wayyy overscheduled.
Since I already own all the stuff, I may need to address the most obvious causes of problems, and focus on tweaks versus anything too labor intensive. Of course this assumes we can get to 90-95% operational (my goal line).
I too wonder about air in system. The plumber has purged several times. The manifolds are just below the lefel of the floor that the pix is installed in. I hope this is not part of the problem when trying to purge. I am not sure where he purges from. I will try and find out when I get up there next.
One of the cheapest things for me to do may be to swap out that one pump for the larger one…if that is a contributing factor. I see Hot Rod and Ed discussing the merits of that need, or not…. Other than the cost of the bigger pump, is there a down side or that?
We still are not up at what I think the eventual Radiant water temp will be. The Plumber did not want to shock the floor. I think only at 85 or so. Does that mean the system will be too unpredictable, until the water had enough heat energy in it.? Next week we’ll go up to 95. I am assuming based on what the Ecowarm guys say we will end up at about 105 to 110degrees. Anybody have expereince with EcoWarm to Warmboard-e?
Thanks again to all involved here !

hot_rod

I’m still not clear on the question or problem? Is that zone not getting up to temperature? If not, you mentioned you are running the SWT, 20 - 25 colder than what the design called for?
Is the floor covering in line with what the design considered. The type of floor covering can be a factor, throw rugs over wood flooring design, for example.

The two questions are, what is the required flow rate, and the required SWT?
All we know is the room loads in btu/ hr and loop lengths, so far

With the assumed gpm per loop at .65- .75 gpm and the loop length, the pump you have should work fine. Who speced the pumps to begin with? What data did they use?

The ECM circulators will use 50%, less electricity to move the same gpm. If you can modulate it, it can save up to 80% according to the pump manufacturers.

mpdport

…warning..Long post!
Ok, I finally got up to the house so I could measure and observe some of these things. I can’t seem to get answer on design spec. So, I am kinda reverse engineering here.
We have gotten the SWT up and the system is started to perform as I would have hoped.

We had to go to about 130degrees – to get the bigger/further manifold are to reach set point at night. We have the Outdoor reset set as flat as we can at the moment to reduce variables while testing.
So, the 130swt is at 32outdoor. It was 10 last night so the current curve would have taken it to about 136.
5 Port Manifold (90’ from boiler). Arrives at manifold at 120degrees. 35k btu load.
3 port Manifold (10’) from boiler. Arrives at manifold at 130 degrees. 17k btu load.

Still tweaking, but also trying to understand best practices for efficient operation.

A couple specific notes.
1- The 5 port Manifold. With the Alpha2 15-55F pump set on Constant speed 3. If only two telestats are open, the flow gauges read about .9 pgm. When all 5 open, it drops to .5gpm or so. There is about 31k btu of load on this one, plus I added 4k for the 90’ run from the boiler (a currently uninsulated)= 35k btu. The 5 loops there range from about 4K btu to 9k btu. The 9k btu Zone (single loop) struggles to hit its set point at night. This is a tile hallway with a LOT of glass facing the water and south. when the sun comes up it is more than fine. But, my question: when the flow drops to .5gpm - is that enuf pressure ? (9k btu/10,000 per gal=.9 is that correct?). That 9k btu zone called for heat about 20 hrs yesterday (cloudy day). Most of the other zones ranged between 7 and 12 hrs. This structure has 9’ and one 16’ cathedral ceiling.

2- the 3 port Manifold area seemed quite happy at 120degrees that we started at, before we raised to 130 to get the 5 port area happy. Now I am wondering will those area still be ok with the extra 10 degrees of SWT temps? I tried running that pump at speed at constant pressure spd 2, but on initial test it started to struggle to hit set points. So I have turned back to spd 3. This area is moslty 7’ ceilings.

3 – I have 2 engineered wood zones that I have to be careful not to exceed floor temp of 81 (a manufacturer spec) doing this with TekMar smart t-stats. And cross checking with a heat gun to validate/adjust for the floor sensor positioning variability. The other 3 zones are ceramic tile

4 – QUESTION - do most of you find it is better to run with lower SWT, and let the pumps run slower and more often. I have a Viessmann 200 series that modulates it’s flame based on who is calling for heat.
Or, is a higher temp and/or faster speed on the pumps with less boiler run time more efficient. I realize that question probably has a few more variables (that I don’t understand yet). But trying to gauge both propane and electrical efficiently.
Currently, it is looking like I am going to have to run the SWT higher for 5 port areas higher heat loss areas. And I suspect another mixing valve may be in my future to reduce that higher temp for the 3 port manifold? But wondered if I got more flow through the 5 port manifold should I work on that first ?

5- FYI have 5 basic Alpha pumps that are taking full temp water to the upper floors HW baseboards. they are all running fine.

6 – we also have a 53 gal Viesmann DHW tank hanging of boiler also. It is not on yet

7- obviously you can tell I am just trying to understand all this stuff , and possibly getting much of it wrong ? (I am a Broadcost engineer).

Thank you for listening to what my wife would say is my new obsession!