Sump Pump capacity 1

[Jasper21]

Hi All,
I have a question on whether or not the 4 sump pumps I have installed, if operating at the same time, will work. I have (2) 4400gpm pumps leading to two separate 1.25in lines that combine to a wye flange which leads to a tee. I have a 3500gpm pump leading to a 1.5in line that leads to another end of the tee. And the 3rd end of the tee is fed by a 2500gpm pump and a 1.5in line. So this tee is fed by 3 pumps with (2) 1.5in lines and (2) 1.25in lines. After rising 2 feet, the tee leads to a 2 inch line out of the building and then to a 3 inch line. I tried modeling with the hardy cross method but keep getting lost!
Will this 2 inch line handle these flows if all pumps are on?

 

[tkall]

14,800 GPM through a 2" line?

 

[LittleInch]

Please sketch this out and check your data. This reads as if there is no outlet. You also needs to quote pump pressure or head curves as there is something seriously wrong with your info.

Any parallel pumping system is problematic and any small difference between the pumps can result in one pump or other not pumping.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Jasper21]

All pumps running, 13,000 gallons per HOUR through the 2" line

 

[Artisi]

would you like to resend your message with corrected flowrates for each pump.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[77JQX]

This is impossible to definitively answer without the pump curves, pipe materials, and elevations, (a sketch would also be nice). But I'll go out on a limb and say your pipe diameters are probably too small. Shoot for average water velocities in the 5-10 feet per second range to have manageable head loss.

 

[fel3]

Assuming "gpm" should be "gph" in each case, your pipe flow velocities are still far too high.
- 4400 gph in a 1.25" pipe = 19.2 fps
- 3500 gph in a 1.5" pipe = 10.6 fps
- 2500 gph in a 1.5" pipe = 7.6 fps
- 13000 gph in a 2" pipe = 22.1 fps (BTW, I get 14,800 gph for a straight sum, but total production will obviously be less than a straight sum).
I would normally target something around 3 to 5 fps for sizing the pipes. At your velocities, water hammer and even pipe wall erosion at the higher velocities must be considered. In addition, if your pipes are anything but VERY short, headloss will eat you alive.

A couple of other points:
- Hardy Cross is not how I would try to solve this, especially if you're doing it by hand. EPANET is free and will handle this type of problem (

http://www.epa.gov/nrmrl/wswrd/dw/epanet.html).

If you want a more robust user interface and the ability to handle multiple scenarios in one model, then a commercial program such as WaterCAD may be your best bet (

http://www.bentley.com/en-US/Products/WaterCAD/).

There is a learning curve, of course, and it will depend on your level of experience with hydraulics and network modeling as to how fast you master it.
- You indicated that you want to see if a 2" pipe will work with all four pumps running (it wont based on velocity and probably headloss too). However, you also need to consider what happens if fewer pumps are running. You have four parallel pumps of different sizes (well, three different sizes) and I suspect that their performance curves are not perfectly matched to each other. In addition to checking the four pump condition, I suggest you model this with the pump curves and see how the different combinations of pumps work. For example, let's say that all four pumps work OK together (after the pipes are upsized, of course). That doesn't mean that any one pump, operating alone, or any two or three pumps, operating together, will be happy. Especially with just one pump, you may find it wants to operate beyond the right end of its pump curve, in the region where cavitation may occur.
- In the past, I and others have posted information about similar and related problems. Here are links to some of these threads:

http://www.eng-tips.com/viewthread.cfm?qid=325862

http://www.eng-tips.com/viewthread.cfm?qid=325862

http://www.eng-tips.com/viewthread.cfm?qid=320430

- "gpm" is generally the preferred unit for pumps of this size, not "gph"

I hope this helps.

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[LittleInch]

A star for fel3. Very well said. The only thing I would like to add is to check the actual ID of your pipes, especially flexibles as they can be important at small diameters. A 2" nominal pipe can be 2" OD or 2" ID depending on your spec and try to avoid tees as the friction losses can be surprisingly large if your pipe lengths are quite short.

Also while I assumed you meant US gals, there are two gals about which can complicate matters so always best to specify exactly.

However if you've got 13,000 gph then your lines must be quite short and hence you seem to have answered your own question....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[bimr]

You probably will be able to get 85 00 to 90 00 gpm out of the 2-Inch pipe. I don't think there is any program capable of modeling that system. It would be easier to just run separate pipes for each pump.

So the answer is no, the 2-Inch pipe will not be able to handle the combined flows. Note that small sump pumps like you are describing have limited ability to pump up elevation and when the operating point on the pump curve moves left, the pump capacity will diminish.

Next time, add some decimal points so you don't get the boys all roused up.

 

[fel3]

bimr…

Any good water modeling program should be able to handle Jasper21's situation, assuming—of course—that the model is set up and applied correctly.

The "normal" use for water modeling software is modeling distribution systems. I have several man-years of experience in this one area with water systems serving populations up to more than 100,000 people. However, in the last ten years or so, I have spent at least as much time using EPANET or WaterCAD to model pumping stations for design or to figure out why an existing pumping station isn't functioning correctly. Even something as simple as a one-pump sewage lift station (suction-side "reservoir", pump, discharge pipe, discharge-side "reservoir") is faster to analyze with software than by hand. I have also analyzed multi-pump situations that are analagous to Jasper21's system, though it was with municipal well pumps all discharging to a common point and much larger piping, so I know it can be done.

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[bimr]

What I meant to imply was that it is not practical to model such a small system. Engineering rates will be over $100 per hour and the pumps are worth about $60 each.

 

[fel3]

But, a system needs to be modeled so the proper pumps can be selected and I can do it faster with software than by hand.

At least a dozen times in my career I have dealt with situations where the original design engineer selected the wrong pump because he didn't do the necessary hydraulic calculations (or he screwed them up). These pumps have ranged from 5 hp to 750 hp.

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[bimr]

Understood, but even your own post limits your modeling expertise to 5 HP. Sump pumps are typically less than 0.5 hp.

This topic fits in with adage about knowing more and more about less and less until you know everything about nothing. An engineer with modest pump experience should be able to make a reasonable estimate of the pump capacities based on just the sump pump operating curves.

 

[fel3]

My reference to 5 hp was for pumps I evaluated for operational problems, not for design.

I have, in fact, designed several sump pumps as small as 1/2-hp, mostly for loading dock drainage. In my case, these small pump designs were done for gov't agencies (municipal projects and prison projects) and the clients required calcs to support the designs (seriously), just as they did for the bigger stuff I designed on the projects (e.g. water distribution systems, pavement sections, retaining walls, etc). I can model a simple sump pump system in about 5 to 10 minutes, so I find it's worth doing anyway even if I don't have supply calcs for review. Just because the pumps are cheap does not mean that failure of the pumping system is similarly inexpensive. Regardless, the OP has 4 pumps in parallel operating in different combinations. There is no easy way to evaluate the hydraulic operation of such a system without a model, which in this case would probably take less than an hour's effort.

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill