System Curve - Multiple Pumps in Common Pipeline

[catplank]

Hi Guys,

Hoping you can point me in the right direction of some literature to read.

Basically i have a main pump station and pipeline of 2300m (500L/s). after the first 1000m another pumping station will also be discharging into the pipeline (100L/s). I'm not really sure how this interaction will work when creating systems curves and looking at pump selections.

If someone could point me in the direction of something to read/watch i would be much appreciated

Cheers.

 

[LittleInch]

Well what you have now is a network and the simple system curve vs pump curve idea gets complex.

What you need to do is work out a series of combinations and flows.

The thing to remember is that at the junction of two systems the pressure cannot be different.

So this is a quite simple network:

You have a joint flow system downstream of your junction with the second pump.

Hence you can develop a system curve for this.

your portion between this and your initial pump is a separate system curve added to the downstream one.

But it's the different combinations which you need to generate.

So you could have Pump 1 only, pump 2 only, pump 1 and 2 at max flow.

In terms of pump selsction it might be that depending on how often the combinations of pumps and flow change that a VFD driven unit may be required or if the flows and combinations are quite simple then fixed speed units can be used.

Each system is different and has its own peculiarities so difficult to say much more.

This will generate your envelope, but for any specfici flow you will need to look at combined flow if both pumps are on.

Complex networks like water distribution systems and gas networks for example now run real time simulations to allow the controllers to understand the impact of turning pumps or compressors on or off.

However at 2300m long your pressure drops won't vary hugely. Are there any fixed pressures or static heads to overcome? It may be varying flow won't have a big impact? Who knows - we don't.

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

 

[catplank]

Thanks,

I appreciate the in depth response. This is where i was kinda of heading in my mind. Good to have it spelled out.

As this is what will likely be the beginning of a greater network I think it will be worthwhile getting my modelling skills up to scratch and develop a system model!

 

[LittleInch]

If this is water have a hunt around the water treatment forum. I think EPANET comes up quite often.

https://www.epa.gov/water-research/epanet

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

 

[hateMech]

Have a look at this, particularly pg 5, "Dissimilar Pumps in Parallel":

http://documentlibrary.xylemappliedwater.com/wp-content/blogs.dir/22/files/2012/07/TEH-1109A.pdf