NPSHR convention 3

[sshep]

This could be easy. An engineer came to me with a KSB pump specsheet which gave NPSHr in units of meters H2O for a pump that pumps light hydrocarbon at well above 100C. The question was: what temperature (density) is the meters of H2O referencing. She wants to convert this to meters of suction fluid to compare with her NPSHa calc.

I have to admit I think it would be some ambient temperature water as per a test rig, in which case the difference between references like 20C or 60F would be negligible, but is there a convention for this sort of (stupid in my opinion) reporting? A quick google gave me no hint of this (meters H2O) convention being in normal use.

Thanks and best wishes,
sshep

 

[sshep]

After thinking about it some, I think that the water is just adding confusion as most of the specsheets in my experience just give the value in meters of head. The value despite referencing water is more generally interpreted simply as meters of process fluid- i.e. no density correction is required.

Any comment on that thinking?

You would think that after 25 years in the industry I wouldn't have any questions about a concept like NPSHr, but apparently not.

thanks again,
sshep

 

[katmar]

sshep, I agree with your revised thinking - it is the head that is important and not the density. There is probably some convention in a code somewhere about the temperature of the water used for testing.

From somewhere I recall that using water based test data for design of hydrocarbon systems is slightly conservative because of water's low molecular mass. When a HC vaporises it generates less volume than water does because the vapor is more dense. But HC's tend to have lower heats of vaporisation, so this would counter that argument.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[zdas04]

That sort of chart always has water at a density of 1,000 kg/m^3 (62.4 lbm/ft^3). It would be less confusing if it was in pressure units, but the conventions of an industry die hard.

It is easy enough to convert it to a gradient of 9.807 kPa/m (0.433 psi/ft) by multiplying density times g (and fixing the units). Then with the gradient of your actual pumped fluid you can recreate the NPSH-a chart.

David

 

[SNORGY]

I, too, hate the "metres of water" or "feet of water" that are commonly reported. Head is head, metres are metres, feet are feet.

I have to this point been led to understand that the NPSH reductions for hydrocarbon fluids afforded via the Hydraulic Institute Standards are - at least in part - due to their significantly lower Vg/Vf ratio. In the case of vapour bubble collapse during cavitation, much less energy is released by a hydrocarbon liquid than for water. The cavitation still occurs but the damage is much less severe.

Regards,

SNORGY.

 

[eadwine]

Has this chart been set up for hydrocarbons? If it is just a typical catalog tear sheet or supplied pump curve,it probably is based on H2O at std. pressure and temp. conditions. To convert, I'll leave that up to you.

 

[bchoate]

sshep[/]
The NPSH requirements of centrifugal pumps are normally determined on the basis of handling water at or near normal room temperatures. Lab tests (various parties) have shown that pumps handling certain gas free hydrocarbon liquids will operate satisfactorily with harmless cavitation and less NPSH than required for cold water.
"Cameron Hydraulic Data", Ingersoll Rand, 1984
The reference has a chart (p.1-46) graphically showing NPSH reductions for propane, isobutane, butane, and methanol. No reduction shall exceed the smaller of 50% of the NPSH for cold water or 10 ft.
e.g. Pump has stated NPSH of 16 ft on cold water. Service will be propane at 55 F with a vapor pressure of 100 psia. The chart shows a reduction of 9.5 ft. The corrected NPSHR is 8 ft (50% of the NPSHR for water).
WEC

 

[shahyar]

agree with bchoate.
"Centrifugal pumps don't care about density", but it is true just for DH and not NPSHR.
The other diagram to convert NPSHR(w) to NPSHR(any other liquid)is Fig 12-5 of GPSA.
If you want to be more conservative, you can skip converting it.

 

[sshep]

Many years ago I must have read about this reduction in NPSHr for hydrocarbons and elevated temperatures, and then forgot. Funny that neither my Durco Pump Manual or Gould Pump Manual (GPM) seems to mention it. I found the diagram in my GPSA manual (my Cameron has disappeared but I think it must be similar) as referenced by shahyar and bchoate. I give you both stars for good references, but thanks to everyone else as well. Any other comments on the theory of this are welcome.

best wishes,
sshep

 

[ione]

I have taken the attached artiche from the link below (which sometimes works and sometimes doesn’t): it is quite interesting

http://www.pump-zone.com/pumps/cent...rbons-stepping-npshr-to-different-speeds.html

 

[atayto]

As Zdas sais"That sort of chart always has water at a density of 1,000 kg/m^3 (62.4 lbm/ft^3). It would be less confusing if it was in pressure units, but the conventions of an industry die hard." But a little bit different 1020kg/m3 at 101.6 kPa and 15 degrees Celcius.