Efficiency curve pumps in parallel 1

[noreddine79]

Hi Friends
Please how to draw the efficiency curve of different pumps in parallel having the efficiency curve of each one.
I m not hydraulic engineer may be the question is simple, but excuse..
Thank you very very much in advance.

 

[LittleInch]

The efficiency of each pump does not change, the thing that changes compared to single flow is the actual flow through each pump.

You need to develop a system curve and then see where the joint flow curve meets it. If you have pumps with different curves this is not easy or straightforward, but can be done with a bit of iteration.

If you have the pump curves and can post them that will explain a bit more, plus you need a bit more description here.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[noreddine79]

Dear LittleInch,
Thank you very much for this answer. What I want exactly is to estimate a general efficiency of the two pumps in parallel. I don't have curve here, I will plot them ASAP to further explain the problem.
Thanks a lot

 

[bimr]

Place the pump performance curve on to the system curve to determine what to expect.

Then overlay the system curve with two pumps operating

To find the pump efficiency, go back to performance curve of each pump using the operating points obtained on your graph.

 

[noreddine79]

Thank you very much bimr,
What if the two pumps are of different performance curves?
Many thanks

 

[bimr]

Same thing, expect you will find that it is much more difficult to operate unequal sized pumps in parallel.

 

[noreddine79]

In my case I have 9 pumps in parallel 03 of them are 5000 m3/h six are 12500 m3/h. ir order to modulate the flow rate, i need to operate sometimes small pump with big one..what are the consequences to expect from this?
many thanks again dear friend

 

[LittleInch]

We await your plot, but without the pump curves it is not possible to determine the efficiency of any pump.

That is a LOT of pumps in parallel, I can only image that your system curve is not a fixed line, but some sort of thing which has multiple flow paths or users as well.

at those flow rates I'm guessing this is some sort of water distribution system?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[noreddine79]

yes indeed the system is water pumping system to feed a big city. not all the 9 pumps are running in the same time since we have some restrictions in flow capacity of the treatment plant.
will plot the curves once back to my office (don't have them here ).
Many thanks friends for your help.

 

[LittleInch]

The issue you will have is that you cannot create a system curve as the flow versus pressure keeps varying wildly as the number of users and flows in each branch of the system changes continuously.

Many of these sorts of systems work on maintaining pressure at some point (discharge or somewhere in the feed system) within a range of pressure and then bring on or drop off pumps to maintain the pressure within that pressure range.

only by looking at the head across the pump can you estimate flow in each pump and hence the efficiency at that flow rate.

Only by looking at the shape of the head curves and what your operating range is can you start to think about what is happenng.

Your pump curves at the operating range will need to have a reasonable slope to them otherwise you could be getting wildly different flows in each pump, even if they are "identical".



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bimr]

I thought you were only talking about a couple of pumps

In actual practice, you are essentially pumping into a constant backpressure. To arrive at your pump efficiency, just take range of discharge pressures at the pump discharge, and read the pump efficiency of each pump off of the individual pump curves.

I suspect that backpressure that you are pumping against is in a range of maybe 10-15 psi. It is likely that the pumps were selected to operate in that sweet spot.

 

[Artisi]

If this is an existing operating system,what is the point of establishing individual efficiencies? You can establish an overall station efficiency by measuring flow output against power input over the various operating periods. This might be a more useful study.

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.)

 

[Softedge]

i agree with artisi given the complexity of your system. since the feed and discharge piping configurations will be different, even slightly, for each pump a theoretical approach will not provide actual operating results. if you have the ability, run and log flow and input power results for different pump configurations for the same flow condition; do this for each flow condition. over time you can establish best operating configurations and procedures for the station under specific, or close to, demand conditions.

 

[racookpe1978]

Seems like his differences are so great (12,500 vs 5500) that the smaller centrifugal pump has almost no flow at all though: The higher-flow pumps are going to be pushing almost all of the water, the little ones will be circulating most of their water and not pushing it through.