Pumps in parallel vs series - Flow vs head 3

[ghensky]

From this Link I get,

1. Pumps in series: Head is added, not the flow rate
2. Pumps in parallel: Flow rate is added, not the head

However, for parallel case, not sure what I'm missing but find it not so intuitive to visualise why heads are not added. How the energy is dissipated in this case, by overheating the pumps?

There could be two cases in parallel case,
I. Both pumps are identical
II. One pump is bigger (higher head and flow capacity)

Why heads are not added? Which law of physics is working here? Thanks in advance.

 

[SandCounter]

Head is added in both cases. In the diagrams, you need to follow the system curve from points 1 to 3.

I used to count sand. Now I don't count at all.

 

[hydtools]

In the case of parallel pumps, the point in the piping where the separate flows combine must be an equal pressure condition in the system, therefore equal head seen by the pumps.

Ted

 

[mk3223]

In the parallel case, there are two options, as you stated, that the pumps may or may not be the same head. So, the total flow is added from two pumps based on the discharge pressure of each pump. If both pumps with the same head, the total flow is theoretically equal to two pumps flow. The reality is that the discrete pressure may be increased a little to accommodate the increased pressure drop because of the higher flow.

 

[georgeverghese]

An easy way for non process engineers to visualise this is to equate pump head with electrical voltage across the terminals on a battery, while flow is equivalent to current.

 

[lilliput1]

Parallel pumps must be similar in size and duty. Otherwise they will fight each other.

 

[LittleInch]

A common misconception when looking at parallel centrifugal pumps is that the pumps pump a fixed volume, e.g. say you had two "750GPM @ 150 ft hd" pumps, connecting them in parallel WON'T give you 1500 GPM @ 150 ft because the resistance to flow in the pipe increases.

If the pumps are identical then they might do, say, 1200 GPM @ 165ft, 600 GPM each. Hence they will use LESS power than if they were operating individually. So there is no "energy being dissipated", because it is not being used in the first place. However there will be an energy increase in frictional losses because you're flowing more fluid faster through the same size pipe. It all depends on the intersection with the system curve. If you added in my example a third 750 GPM pump, total flow might only go up by 200 GPM due to the system curve, hence <500 GPM each.

In general you tend to see series pumping when there is either a high static head or pressure to overcome and one pump can't supply that head or when the head requires changes but the flow doesn't.

Parallel pumps work well when the system curve is quite flat or where the static head portion is quite high compared to the frictional losses.



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

 

[wayne3298]

Draw a parallel pump curve using actual or proposed single pump curves. Calculate the actual or design system curve and plot it on the parallel pump curve. If pumps are not the same design you can't draw a parallel curve which tells you don't operate them at the same time. Some pump selections for parallel pumping results in one pump delivering 85 to 90 percent of design flow with one pump running. Depends on pump selection and how close actual operating conditions match design.

 

[bimr]

McNally has a discussion on parallel pumps:

http://www.mcnallyinstitute.com/15-html/15-01.htm

 

[TenPenny]

wayne3298 said, "Draw a parallel pump curve using actual or proposed single pump curves. Calculate the actual or design system curve and plot it on the parallel pump curve. If pumps are not the same design you can't draw a parallel curve which tells you don't operate them at the same time"

I don't know what you meant by this, can you clarify?