Help w/Pump Sizing for multiple branches 1

[7897]

I'm trying to size a pump for use in an aquarium system. The pump flow will be 2100 gph (35 gpm). The pump suction will come from a sump. The discharge of the pump will have three branches. One branch will be to a filter at 900 gph (15 gpm) and the other two branches will return water to the display tank at 600 gph (10 gpm) per branch. Piping will be 1" SCH 40 PVC. Can someone provide assistance as to how to determine what the TDH of the pump would be (35gpm @ XX TDH)?

Thank you

 

[Artisi]

To size a pump you will have to advise the length of pipe on each leg,any valves and bend etc., the difference in level from the sump to each point and the head loss thru the filter at the required flow rate.
A 1" pipeline might be too small if you have long runs as the head loss could be excessive.

 

[7897]

I understand that I have to account for all the friction losses in the piping. I wasn't sure if I had to look at each branch independently and come up with a TDH for each branch then sum them up for the total head.

 

[Artisi]

calculate the head loss for the maximum - probably the leg with the filter - this will be the head needed for the pump, and as it will probably be excessive for the other 2 points you will need to fit a control valve of some sort to each point to control flow. Again check for excesive head loss in the 1"line - you maybe need to go to 1.25" or 1.5"

 

[pleckner]

What Artisi wrote is correct but I would like to clarify what he is implying and answer your question directly. You take each branch individually and size as if that branch was the only one. Your pump is designed for the branch giving you the greatest TDH, you DO NOT add them together.

 

[HEC]

7897,
To add to pleckner's post, the "use points" should be considered separately to get the pump discharge head. The flows should be added to give the pump capacity. Consider the pressure drop to be in parallel. A visual picture would be if teh single pump were to pump the full flow to a header tank from which the various use points draw from.
Cheers

Mark Hutton

 

[7897]

HEC:

I did look at each "leg" of the piping. As it turns out the "leg" with the filter, 15 gpm resulted in a 7 ft of head. The "leg" that goes to the aquarium, 20 gpm resulted in 14ft of head. So, even though I'm looking for only 15 gpm to the filter and 20 gpm to the display aquarium, you're saying I should look for a pump sized for 35 gpm at 14 ft. of head? is that correct?

 

[pleckner]

I'm glad Marc came in with his post because he added something that I was intendig to write but left out due to distraction.

Your pump head is indeed based on the 14 ft TDH you calculated BUT your pump capacity (flow) will be determined by the maximum suction flow. For example, if I have 5 branches but only three are used at the same time, then my total suction flow will be equal to the total of whatever I need for those three brances only! This becomes your pump design flow (with a safety factor please).

For frictional losses, you first calculate the frictional loss between the pump discharge and the first branch, using the total flow for all simultaneous branches. Then you would calculate the frictional loss in that particular branch AT THE BRANCH FLOW RATE, NOT THE TOTAL. If this is what you did then you did it correctly. If not, take another look.

 

[7897]

Pleckner:

I follow the first paragraph of your response. I have three branches, but only two are used at the same time. It's your second paragraph that's got me confused. Are you saying that I should calculate the frictional losses of the filter branch at 35gpm and then calculate the frictional losses for the filter branch at 15gpm? If I do that, I get the following:

TDH of branch that goes to aquarium: 14ft. at 20 gpm
TDH of 1st branch that goes to filter: 7ft. at 15gpm
TDH of 1st branch that goes to filter: 25ft. at 35gpm

So if I'm understanding you correctly I should look for a pump sized for 35gpm at 25ft. Is that right?

 

[Artisi]

No - the flow rates per point are now different to the original posting - you now have a total flow of 70gpm.

 

[7897]

No, it's still 35 gpm. If your read Pleckner's last post, he stated that I should look at frictional loss at the first branch using the total flow (35gpm). Then, look at the frictional loss at the first branch using the branch flow (15gpm). At the end of my post, I was just list the 3 different ways I calculated the TDH.

 

[Artisi]

Ok - my mistake, if the total head needed to operate the system is 25ft and the total flow is 35 gpm then the pump will be as you have specified 35gpm at 25ft head.
One more thing to look at, what is the difference in level from the sump water level to the highest point in your system, this needs to be included in your head calcs if not already included.

 

[pleckner]

I'm trying to size a pump for use in an aquarium system. The pump flow will be 2100 gph (35 gpm). The pump suction 7897:

According to your original post, you have three branches, one at 15 gpm and the other two at 10 gpm each, but your post of 23 Mar 07 13:32 states that only two operate simultaneously but you give different flows.

Anyways, here is the way you do it. The two largest simultaneous flows are one to a filter, say 15 gpm and one return to the tank, say 10 gpm. Therefore the total flow the pump needs to put out at any given time is 25 gpm.

What you want to do for the head loss calculation is to take the maximum total simultaneous flow of 25 gpmand perform the friction loss calculation from the pump discharge right up to where the first branch occurs. Let's assume the first branch is to the filter. To get the frictional loss from this junction to the filter, you use the 15 gpm flow rate. Then I assume this flow returns to the tank after being filtered. If that is all it does, then continue with the frictional loss calculation from the filter back to the tank. You need to account for final elevation differences between the pump discharge and the tank as brought up by Artisi. BUT, you also need to take into account a pressure drop through the filter. This may not be easy to come by.

If the flow from the filter joins back up with the other branch before returning to the tank, then the frictional loss of that part of the pipe is again performed using the maximum simultaneous flow of 25 gpm.

You then do the same thing with the recycle branch, using 10 gpm. This will give you another discharge head loss.

The discharge head loss for the pump will be the greater of the two branches. Now, you need to do a similar calculation on the suction side to get the TDH. The suction side is of course using the maximum simultaneous flow, 25 gpm.

Don't forget to also take into account any valves you may have in your system and of course the tee fittings that make up the branch take-offs.

Is this clear to you now?

 

[7897]

I think that's much clear. Is there a way to post a jpg picture in a post so I can show you what the piping arrangement really looks like?

 

[pleckner]

Don't know. I don't even know how to put in bold or italics, etc. Perhaps someone can help both of us.

Let's try this:


25 gpm 25 gpm 10 gpm
-------Pump>>>>>>>>>>>>>O>>>>>>>>>>>>>>>
| |
| TANK
15 gpm | ^
| ^
FILTER >>>>>>>>>>>
15 gpm

 

[pleckner]

It didn't quite work out like I hoped!?!

 

[Artisi]

we also need the lengths of pipe run - suction to pump and diameter, pump to first branch, branch to discharge point, distance to the next branch and then to discharge point etc.
and the differences in levels.

Without this information we are wasting our time trying to establish the head loss .

Draw the system - scan the drawing and post it using the following link if you have a jpeg picture you can load it directly and send it.

http://www.freeimagehosting.net/

 

[HEC]

7897,
Draw the system as a rough "P&ID" ie show the sump, then a suction pipe to the pump then to a discharge header, tehn show each take off in order. From this you can find the flow rate in each section of the "pipe network". From your pipe arrangement for each section the pipe losses (pressure drop) can be determined. Add the individual section losses for each section and discharge point. The largest pressure drop will be the design point for the pump. Per the following

1. Pump suction: flow 35(us)gpm, .5' head
2. Pump discharge to filter off take: Flow 35gpm, 10' head
3. Filter take off to filter: Flow 15gpm, 2' head
4. Filter differential: Flow 15gpm, 2'(min) 4'(max) head
5. Filter discharge to sump: Flow 15gpm, 1' head
6. Filter take off to first aquarium outlet: Flow 20gpm, 5' head
7. First aquarium outlet: Flow 10gpm, 1' head
8. Firt aquarium outlet to second aquarium outlet: Flow 10gpm, 3' head

Filter branch total head is .5+10+2+4+1=17.5 (max) 15.5 (min)
First aquarium branch total head is .5+10+5+1= 16.1
Second aquarium branch total head is .5+10+5+3= 18.5

Pump specification will be in this example 35gpm, 18.5'

Note that the example system will require the following "balancing valves". The first aquarium outlet will need a fixed valve to balance the flows, the filter branch will require a back pressure valve to ensure constant flow to the aquarium outlets as the filter becomes loaded and back pressure increases.
Reckon that covers it!

Mark Hutton

 

[vpl]

7897 and plecker

Click on the link called "Process TGML" right above the submit post. It will tell you how to post a picture, add bold and italics, put quotes in a nice box so you know it's a quote and a bunch of other stuff that makes posting here a lot easier.



Patricia Lougheed

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[7897]

I've been away from the forum because I've been sick with the flu. I'm going to try HEC's method and see what I come up with.