> >
>

# trying to determine flow rate

#1
12-20-11, 12:02 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
trying to determine flow rate

hello; i am trying to determine two things on a currently installed solar array that shares a boiler loop. There are 5 zones on this system and there is a grundfos 15-58 installed in the bottom of the boiler loop . There is also another grundfos 15-58 installed directly into the loop that feeds the solar array directly overhead about 18'. The boiler loop is 1 1/2" and the solar loop is 1". ( There is also a circulator in the boiler itself, but only comes on when the boiler fires)
First,
Looking at the pump specs for this circulator pump, it seems one pump would be hard pressed to make this lift and i am wondering if the current setup with 2 pumps is better than if i had a single pump with a different spec for the height it can pump.

Second, i am trying to figure out the correct flow rate so i can accurately determine the btu production of the solar array. Both pumps come on together and go off together.

#2
12-20-11, 02:23 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
Measuring flow rates isn't all that easy. Here are some ways:

Install pressure gauges on each side of a pump. (You can use a differential pressure gauge, but they are expensive.) Knowing differential pressure across the pump, use the pump's characteristic curve to pick off the flow rate.

Plot the head-flow of the the piping system that the pump supplies. Find where the pump curve and the system curve intersect. This is a quite complicated process for all but the simplest system.

Install a flowmeter in the line.

If you know the input heat to the system, Btu/hr, measure the temperature differential, supply minus return, and calculate the flow rate.

#3
12-20-11, 02:26 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
I don't understand how a solar array shares a boiler loop.

Anyway, it doesn't sound like a drain back system?
So the height of the array doesn't effect the "lift" of the pump because the water coming down equalizes the gravity head to zero.
The head will only be the total of all the pipes, fittings, collectors, etc. at a given flow rate.

Peter

#4
12-20-11, 02:38 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
the array is not a drain back system. either the boiler or the array can be running, not both. you are right about circulator vs pump, but the performance curves for a pump show the lift and gpm. the lift in my system for the 15-58 appears to be at its max for a single pump. that's why i am trying to figure out the effect of the second pump.

#5
12-20-11, 02:46 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
Is this a closed system? If so, what Peter said is correct. For a pump curve, "lift" can refer to either a static or dynamic pressure drop, or both. For a closed system, the pump sees only the dynamic head caused by the friction loss through the system - physical altitude differences make no difference.

P.S. An open system will have a head-tank, open to atmosphere, located at the tip-top of the system. A closed system will pressurized with an expansion tank, but nowhere open to atmosphere.

Last edited by gilmorrie; 12-20-11 at 02:57 PM. Reason: P.S.
#6
12-20-11, 02:49 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
How did you determine what the "lift" of your system is?

If you wish i have a program that can help figure it all out.
We would have to know all pipe sizes and lengths plus how many, more or less, of each type of fitting there is.
Also need the rated head loss rating of the collectors. And what flow rate you need to achieve.

Peter

#7
12-20-11, 04:37 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
i have heard about those programs. i will measure everything tomorrow. do you want me to post it here?

#8
12-20-11, 04:49 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
Originally Posted by mtmtntop
i have heard about those programs. i will measure everything tomorrow. do you want me to post it here?
No, we want you to answer this question, as previously asked here by Peter: "How did you determine what the "lift" of your system is?"

#9
12-20-11, 06:01 PM
Member
Join Date: Nov 2005
Location: USA
Posts: 2,459
1" piping to a solar array should have very low head loss, unless the panels are 300 feet of pipe away from the boiler/system. Panels (if that's what they are, as opposed to evac tubes) have low head loss. A 15-58 is plenty.

1 gpm is a common target flow rate.

#10
12-20-11, 06:49 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
Originally Posted by mtmtntop
i have heard about those programs. i will measure everything tomorrow. do you want me to post it here?
Sure i can try and help.

Peter

#11
12-20-11, 06:53 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
Hello Xiphas,
If he had an Alpha or Stratos set for a fixed flow rate and a couple of accurate temperature gauges, then i would think a fairly accurate measurement of the solar BTU production would be possible.
Otherwise, some sort of flow meter, as Gil mentioned should work.

Peter

#12
12-21-11, 02:39 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
first, i think i may have confused the issue. this is a 75 evacuated tube array, no flat panels.

here are the different pipe configurations of only the solar loop, through the boiler loop and back to the solar array

1" copper-135'
1 1/2" copper-11'
1 1/2" 90's-4
1" 90's-18
1" 45's -2
1 1/2 x 1 1/2 x1 T-2
1/1/2 gate valves (quarter turn) -2
1" gate valves (quarter turn) -2

Regarding how i figured head, i thought this was the height it had to go between the top and bottom of the system, ie from the array pump to the top of the evac tube array.

in this loop, there are 2 15-58 grundfos pumps and one zone valve. the pumps operate at the same time.

Here is some more info so you can get the idea on what the circulator setup is.

other zones include a 90' foot run to an air handler (heat dump) (180' total distance), 30' to a shell and tube exchanger for an indoor pool, and a hydronic floor in a 450 sq ft room. the air handler (90 feet away also has its own circulator that goes on when that zone opens)

there is usually one zone open, sometimes two. it would be unusual if 3 zones were open.

the boiler is on a timer so it never operates when the solar array zone valve could be open (daylight hours).

let me know if i left out critical information. I am not sure how to figure the resistance in the evac tube manifold. That is 28' and has 1" plumbing into and out of it.

#13
12-21-11, 05:40 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
Hello,
I'm guessing the fluid is 50% Glycol?

Looked up Apricus tube collectors.
They call for just under 1 gpm for a 30 tube collector.
Thus it would seem that 2.5 gpm would be needed for 75 tubes.
I see that 1.27 gpm has 5 inchesw of head, so guessing about a foot of head, double it for safety?

Be nice to see a drawing of the system.
Kinda hard for my small brain to follow the setup.

More later,
Peter

#14
12-21-11, 06:24 PM
Member
Join Date: Nov 2005
Location: USA
Posts: 2,459
Originally Posted by mtmtntop
First,
Looking at the pump specs for this circulator pump, it seems one pump would be hard pressed to make this lift and i am wondering if the current setup with 2 pumps is better than if i had a single pump with a different spec for the height it can pump.

Second, i am trying to figure out the correct flow rate so i can accurately determine the btu production of the solar array. Both pumps come on together and go off together.
Like Peter, not sure I follow all this. Is there a problem with the solar array?

If you want to measure solar BTU production, use something like this:
DLY Water Meter for Measuring Water Flow in Solar Systems

#15
12-21-11, 08:32 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
reading these answers, i see a few things. first, i was measuring static head (height) ,not pump head ( moving liquid past resistance) anyway i googled till i found this piece of instruction:

i think i have most of the info i need to figure this out, and i see i have to figure it out for every loop to find the loop with the greatest head (resistance)

what i did learn in this process is determining flow rate is not something easily solved in a forum; it can be done, step by step in your house, if you use the link i have shown. (or installing a flow meter, which i may do) . thanks for the help.

#16
12-22-11, 03:40 AM
Member
Join Date: Nov 2005
Location: USA
Posts: 2,459
This is a simple exercise in a spreadsheet. Good luck. Post back with results. The calcs will also be conservative, and probably underestimate the actual flow rate.

#17
12-29-11, 03:05 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
back with calcs

i have several zones in this system, and as it turns out there is no problem with the solar loop. (pump specs vs equivalent length of pipe) I also thought my grundphos 15-58 ran anywhere between the high and low speed detents because of the way the dial/handle is set up. a Consult with the company verifies i was wrong in this assumption, regardless of how the dial appears, there are only 3 distinct speeds.

anyway, it appears the zone with the highest value for head is my hydronic floor loop with a staggering 30, probably because it has 300 feet of 1/2" pex. the pex is in two loops, but both feed out of a single zone valve off the boiler loop. the only pump that would be activated for circulation to the floor would be the grundphos 15-58 which i have on the medium speed.
The taco publication i referenced earlier says that two pumps will double the flow rate but does not say it will change the head. what would you suggest? I am thinking that the floor loops need to either be open only one at a time (make another zone) , or add a pump into this loop. It does not appear that adding another pump to the existing loop would solve the problem. I appreciate your advice.

#18
12-29-11, 03:10 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
it has 300 feet of 1/2" pex. the pex is in two loops, but both feed out of a single zone valve off the boiler loop.
Is that 300' in ONE loop? or is that the total of BOTH? i.e. two loops at 150' each?

#19
12-29-11, 03:57 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
two loops. actually one was 125' and the other was 175 by the way it laid out in the floor- bay window area. but the two loop come from a single 3/4 line /zone valve then are split via a T and then reducer from 3/4 to 1/2". i should also mention it never seemed like this setup worked . part of the floor, closest to the boiler would be warm while farther out in the loop, it was not warm.

#20
12-29-11, 04:41 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
The head loss then for that loop is not for 300' of tubing, if that's how you calculated it. The head on that loop is the two in parallel which is less than in series...

part of the floor, closest to the boiler would be warm while farther out in the loop, it was not warm.
It does sound like not enough flow. I would think that the 15-58 should be man enough to do that job...

I guess there are no thermometers on the supply and return of the loops?

It's always a good design practice to make the radiant loops from a single supply all the same length to insure equal flow in each of the tubes.

#21
12-29-11, 04:43 PM
Member
Join Date: Nov 2011
Location: God Bless America
Posts: 122
Did you figure the amount of radiation required from the pex to determine the flow rate?

30 seems way high even for pex. I have 5/16 pex running ~200ft and I am using a Grundfos 15-10 happily.

The issue is the flow rate that is required, as more flow = more head loss.

#22
12-29-11, 04:51 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
How did you arrive at 30' of head for those loops?

At 1 GPM 1/2" pex is about 4 ft of head / 100' ...

At 2 GPM it's about 12 ft / 100 ...

big difference...

#23
12-29-11, 04:51 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
the only pump that would be activated for circulation to the floor would be the grundphos 15-58 which i have on the medium speed.
Why not just try turning the pump up to high speed?

IMO tha radiant zone only needs to be figured for 175 of length.

Two pumps in series will add their head for the same flow rate..

Two pumps in parallel will nearly double the flow rate if the head is kept the same as with one pump.

Again, some pictures and or a diagram of the system would be extremely helpful.

Peter

#24
12-29-11, 04:57 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
the min /max flow through 1/2" pex is 1.2/2.3 gpm. i just assumed 1.5 gpm to be able to use taco's table (see link in earlier post) My cacls were all done using that taco publication.

HL = k x c x L x (f1.75)
Where:
k = a number based on tubing type/size (found in Table 3)
c = correction factor for fluid type and temperature (found in Table 4)
L = total equivalent length of piping circuit (feet) (from Step 3)
f1.75 = flow rate through piping (gpm) raised to 1.75 power (selected values found in Table 5)

for me k= .0374 , c=1.349 because i am using glycol.,( probably greater than 50/50)
l= 300' pex +44 more feet from fittings and f1.75=2.033 (1.5 gpm flow rate)

so that is how i came up with 30. is this calculation correct?

#25
12-29-11, 05:16 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
One thing that complicates matters is that the actual flow is a 'moving target'.

If you took a pump curve chart and then derived a 'system curve' and superimposed that on the pump chart, you could determine the operating flow rate of the system.

so that is how i came up with 30. is this calculation correct?

Think of it this way, the cross-sectional area of the two loops is larger than a single tube, so it appears to be a single larger tube as far as the pump is concerned.

There is an electrical analogy... two resistors in parallel are always less resistance than either of the individuals. Resistors in series ADD.

I don't know that the analogy is exactly direct, but the formula for resistors is:

R1 X R2
-------
R1 + R2

So if you had two resistors, one 10 ohm (ft of head) and one 15 ohm (ft of head) in parallel, the total resistance (ft of head) would be

10 X 15 divided by 10 + 15 = 6

So the PUMP would see less head than either of the two parallel loops. There WILL be more flow in the shorter loop.

#26
12-29-11, 05:31 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
I don't know that the analogy is exactly direct, but the formula for resistors is:
Hello NJT,
According to John Siegenthaller's book it is very similar to Ohm's law but there is a different factor used for the current.
Ohm's law: votlage and current are directly proportional.
Hydronic is a different:
The flow rate is raise to the 1.75 power.
In other words, if the flow rate thru a piping device is doubled, the head increases by a factor of 3.4

Peter

#27
12-29-11, 05:43 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
The glycol is murder on the head situation.

For 10,000 bt's at a 10 degree drop:
100 ft of 1/2 pex at 2gpm, with water gives 12.2 ft of head and flows 3.5 ft per second.

The same with glycol is
2.2 gpm 19.8ft of head and 3.8 ft per second
===============

with a 15 degree drop:
water: 1.4 gpm, 6 ft head, 2.3 ft per second
glycol: 1.5 gpm, 9.7 ft of head, 2.6 ft per second

Peter

#28
12-29-11, 06:10 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
The flow rate is raise to the 1.75 power.
Thanks Peter...

1.75 eh? any explanation as to the significance?

I wonder if the exact number is 1.732 ? (hint: square root of 3)
or maybe it's 1.77 ? pie are not square, pie are round.
and Siggy is just getting close enough for gummint work.

#29
12-29-11, 06:24 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Here's an article that's right in line with the discussion! Thanks for the inspiration to look Peter!

Determining Flow Rates in Parallel Piping Systems Constructed of Smooth Tubing - Archives - PMEngineer

#30
12-29-11, 06:27 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
mtmtntop: min /max flow through 1/2" pex is 1.2/2.3 gpm. i just assumed 1.5 gpm to be able to use taco's table (see link in earlier post) My cacls were all done using that taco publication.
The minimum flow through that pipe is zero, or at least below 1.2 gpm. You can't assume a flow - since that is what you are trying to calculate in the first place. The way to do this is to plot, graphically, the system's head-flow curve vs. the pump curve, and find the intersection - as previously posted.

Last edited by NJT; 12-29-11 at 06:37 PM. Reason: added quotes
#31
12-29-11, 06:37 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
mtmtntop: Looking at the pump specs for this circulator pump, it seems one pump would be hard pressed to make this lift and i am wondering if the current setup with 2 pumps is better than if i had a single pump with a different spec for the height it can pump.
Is this a closed system? If so, then what lift? The pump(s) only need to overcome the flow resistance through the piping. The static "lift" is provided by the pressure setting of the system.

Last edited by NJT; 12-29-11 at 07:16 PM.
#32
12-29-11, 07:55 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67

yeah, in my original post i did not know the difference between static lift and , i guess what you would call head, or resistance. i get it now. or at least i think i do.

i guess where i am now is the head , or resistance of the two loops , using that calculation i posted, derives 30. the pump curve for the 15-58 at no flow is about 19 ft. and it goes down from there. so something must be wrong with the calculation but...

re another post, would the pump see the two loops as the equivalent of 150 of 1" ? then the calcs would change dramatically, like to 0.917'( 1 ft) instead of 30'.

finally, a few people have suggested constructing the system flow curve. i have no idea how to do that but am thinking i have to measure pressure on both sides of the pump. i don't see any place to insert a bypass or guage and i have to admit, soldering is my very weak suit. It takes me several times to get a joint that does not leak, although i can do it. I am not too excited about cutting into the loop to get a guage installed.

finally, someone said why not turn up the pump? my original goal was to determine the flow rate so i can put that number into the solar controller , and that flow rate is used to derive the btus produced. that is what i was after originally. the btu production of the solar array.

if i increase the pump speed to high, it seems to cycle the solar array off and on, whereas if i have it set on medium, the array produces enough heat to stay at about 125 all day without ever shutting off. this goes through a heat exchanger, heating an indoor pool

i have seen some great diagrams in other posts. is there somewhere you can point me to where i could make a diagram that someone besides me could understand? as you can surely tell, i am not a plumber...

#33
12-29-11, 08:01 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Just for the heckuv it, as a mental exercise, I used that article I found and ran the numbers for 125 and 175 feet of 1/2' pex in parallel and plotted a system curve on a 15-58 pump chart...

Keep in mind that I did not include any of the common piping... this is ONLY the pex, and none of the head from any of the other pipes, fittings, etc... this added resistance would shift the curve toward the left, and a bit steeper.

OH... and this is for water NOT glycol!

The blue lines are the three speeds of the pump, the black line is a system curve for the tubings mentioned.

The intersection points are the flow which would occur at that pump speed.

#34
12-30-11, 09:41 AM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
There are two must-have "bibles" for anybody who wants to get into flow calculations:

"Flow of Fluids Through Valves, Fittings, and Pipe," published by Crane Co.

"Cameron Hydraulic Data"

Both books have been around for eons and are available in various editions at Amazon. Best to buy used copies and earlier editions - brand new ones can be expensive.

#35
12-30-11, 04:41 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
NJ Trooper
Just for the heckuv it, as a mental exercise, I used that article I found and ran the numbers for 125 and 175 feet of 1/2' pex in parallel and plotted a system curve on a 15-58 pump chart...
That approach is worthy of an honorary doctorate in mechanical engineering. Actually, it's worth more, because quite a few MEs probably couldn't have done it (but I could have, when I had higher energy and mental faculties, but then I ain't an ME).

#36
12-30-11, 04:47 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Thank Microsoft for 'Windows Calculator' ! It did take me a while to figure out the x^y key!

#37
12-30-11, 05:23 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 2,980
Sounds good. How do I get to Windows Calculator? I'm running Windows 7.

#38
12-30-11, 05:33 PM
Member
Join Date: Feb 2011
Location: US
Posts: 552
All Programs,
Accessories,
Calculator.

Peter

#39
12-30-11, 05:59 PM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
It's really nothing to rave about... just a little graphic calculator...

#40
12-31-11, 03:54 PM
Member
Join Date: Apr 2007
Location: us
Posts: 67
i got it now. you solve for flow at different head values to make the curve for the pump. i have to brush up on my math skills. i haven't tried to solve an equation like that in years. thanks for your help.

one last question. can you tell me where i can get ( hopefully free) software or symbols to diagram this system so i will be prepared the next question i have to ask?

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Pingbacks are Off
Refbacks are Off