9 replies to this topic

[A mukherjee]

Hello everyone,
We were taking offer from vendors regarding a centrifugal water pump.The design flow rate through the pump is 10m3/hr ,head requirement 100 metre.
We found that initially the vendor offered a single stage pump,which has got a efficiency of 21%.But when we mentioned that we have issue regarding power,then the same vendor quoted offer for a multistage pump(7 stages),which has got a efficiency of 52%.
Can any one explain to me what is the reason of this huge increase in efficiency due to larger no. of stages?

[breizh]

Hi ,
Hope you can find answers to your query using this resource:

http://www.pumpfunda...e.htm#download6


Breizh

[kkala]

A single stage centrifugal pump has a head up to about 300 ft (91 m) according to Pump selection chart, found in Perry, 7th edition (1997), Section 10, Chapter: Centrifugal pumps. Multistage centrifugal pumps can approach up to 10 times the mentioned head limit, according to same reference (flow rate about 10 m3/h).
A 100 m head is quite on the limit for a single stage pump; my interpretation is that volumetric loss is excessive, due to rather high back flow (leakages) from impeller spaces, see http://www.engineeri...ency-d_633.html. This is smoothed out in a multistage pump due to better design of each impeller (resulting in less total leakages).
Apparently supplier had selected a single stage pump far from its best efficiency point, just to satisfy requirements. Other single stage pumps may have better efficiency than 21%, but you had better stay with multistage.
I assume that you have not required a single stage pump (as a practice we let vendor specify it), but vendor has offered it due to probably lower capital cost.

Edited by kkala, 12 April 2011 - 12:04 PM.

[A mukherjee]

kkala:

Does the volumtric loss that you refer to correspond to internal recirculation?

[James K]

The multi-stage pump will have better efficiency then the single stage pump is most likely due to the impeller design. The single stage pump may have a higher specific speed impeller and is not as good a hydraulic fit as good as compared to the multi-stage pump. The relatively low flow rate with a medium pressure requirement is a good choice for a multi-stage pump. Please be aware that if there are any abrasive particles in the pump, you will quickly wear out the fabricated design multi-stage pump.

Good luck.

[breizh]

Hi ,

let you consider this resource , similar to Kostas's reference

http://webwormcpt.bl...-selection.html

hope this helps

Breizh

[A mukherjee]

Thanks Breizh.the link is wonderful and of great help.
I just have another querie.
For a centrifugal cryogenic pump which has got capacity of 267 lpm & head of 91 metre,vendor has furnished curve for which NPSH required is intially(for arnd 40 lpm flow variation) decreasing with flow,then it is increasing steadily.Normally NPSH required increases with flow.
Can anyone tell me why is this fluctuation?

[Zauberberg]

Why don't you ask the vendor? They should be able to furnish you with all the required information.

[kkala]

kkala:Does the volumetric loss that you refer to correspond to internal recirculation?

It could be considered so, since this is the leakage from discharge to suction through impeller free space due to pressure difference. But another loss is considered as "recirculation loss" in a centrifugal pump.
Centrifugal pump losses are understood as follows (W. Badger & J Banchero, Introduction to Chemical Engineering, 1955).
1. Mechanical: Friction in bearings, in stuffing box, etc.
2. Leakage (or volumetric): Parasitic flows back to suction. The losses in concern.
3. Recirculation: Not quite uniform velocity between two adjacent impeller vanes creates parasitic turbulence, hence loss of energy.
4. Hydraulic: friction of handled liquid on the casing, impeller, sudden velocity reduction of liquid when leaving the impeller.

[A mukherjee]

I have already asked the vendor.But they are not explaining.They are telling that the pump curve is software output.