Applying affinity laws 2

[PalomaP]

An horizontal Between Bearing, Axially Split, Single Stage Pump @ 1000m3/h, 38m, 1480 rpm, 119 kw (impeller 368 mm). The system curve is H=26+12.10^-6.q^2
When changing duty conditions to (500 m3/h, 29m) checking online pumps charts, changing speed impeller to 1200 rpm, new power is 52,4 kw (ef.75,3%). However changing the impeller diameter to 310 mm, new power is 55,2 (ef.71,6%).
Why this difference in the efficiency? changing speed or diamater should be the same...(according to affinity laws).

In stead of using curves program on line from the supplier, how could i use the affinity laws? How can i check the values of new impeller diameter or new speed for new conditions?
Thanks in advance

 

[BigInch]

You should use the curves, whenever the curves are available!

Efficiencies shouldn't change much, but they do anyway.

Affinity calculations follow the theory of geometric simularity and should ONLY be used when better information is not available. Additionally, the range of applicability is small, generally being a maximum of +/- 10%. You're already out of the recommended range of applicability, so stick with the curves.

http://virtualpipeline.spaces.msn.com

 

[BigInch]

Affinity based online pump performance estimator is here,

http://www.tasonline.co.za/toolbox/pump/powaffinity.htm

http://virtualpipeline.spaces.msn.com

 

[PalomaP]

BigInch, Thanks! Your information is always very helpful!

 

[Scipio]

When you change the impeller speed, you aren't changing the geometry of the impeller tips or the clearances between the discharge edge of the impeller and the cutwater (or diffuser vanes, depending on construction). Decreasing the diameter, the vanes frequently get a bit thicker, and the vane angle at the impeller discharge changes. Also, the increased clearance between the rim of the impeller and cutwater/diffusers means that the pump isn't translating the kinetic energy coming out of the impeller into potential energy as efficiently, some of the energy is lost in recirculation and turbulence.

Affinity laws are fairly reliable for speed variation, but when you start changing diameter, there's no real way to estimate the resulting efficiency, vendor curve families for different diameters are frequently based on historical performance testing, which is why, as Biginch pointed out, applying the affinity laws to diameter changes should be kept to a small percentage trim adjustment.

 

[BigInch]

Scipio, thanks for the clarification. I made no distinction between changing speed and changing diameter in my answer.

http://virtualpipeline.spaces.msn.com