NPSH margin calculation

[uselessengineer]

Pump manufacturer npshr curve are always given in absolute meter of water column.

Question: npsha is always calculated or expressed in meter of pumping liquid. When calculating npsh margin, do you convert your npsha from meter of pumping liquid to meter of water?

 

[pierreick]

Hi,
Let you study this material, and you will find your answer.
You should always refer to the fluid in service. Margin is addition of ft or m of fluid.
Pierre

 

[uselessengineer]

What i gathered from the material is that npsha in m of pumping liquid is used for margin calculation, even if npshr is given in m of liquid water. This is what i have always seen in all engineering datasheets.

Just something i thought about this approach - what if your pumping liquid has a very low density compared to water. In that case your npsha in meter of pump liquid could be ample but it is actually less than the npshr in absolute pressure terms.

For example:
Pump liquid density is 300kg/m3.

Npsha is 3.0m of liquid. Npsha in Pressure = 300×9.81x 3.0 = 8.8kpa

Npshr is 1.0m of water. Npshr in Pressure =1000 x 9.8a x 1 = 9.8kpa

Technically, npsh margin = 2.0m which is ample.

But in absolute pressure terms, npsha is actually lesser than the npshr?

if your epc accepts this pump basis that npsh margin is sufficient, will you accept it?

 

[pierreick]

Share with us a sketch and relevant information about your process to validate your calculation of NPSH available.
Share your pump data sheet and curves.
your reply is very confusing, you must compare NPSH available and required, both expressed in m or ft of liquid.
In other words, you need to compare apples with apples, not carrots.
Pierre

 

[goutam_freelance]

According to the document attached above:

(NPSHA) is defined as:
The total suction head in feet of liquid absolute, determined at the suction nozzle and corrected to
datum, less the vapor pressure of the liquid in feet absolute.

So NPSHA and NPSHR are defined in terms of the working liquid property. It is to be noted that 'total suction head' above includes the velocity head also.

But, since the pumps are tested using cold water, the pump suppliers usually specify the NPSHA/R in terms of liquid water column. Whenever mwc is specified by the pump supplier, it is to be interpreted as mlc in terms of working liquid.

 

[LittleInch]

Pump manufacturer npshr curve are always given in absolute meter of water column.

Question: npsha is always calculated or expressed in meter of pumping liquid. When calculating npsh margin, do you convert your npsha from meter of pumping liquid to meter of water?

Useless (not a great user name tbh).

You are not correct. NPSH is metres. It is not water, simply whatever fluid you have.

When you calculate NPSHa, the first thing you do is calculate is the equivalent metres head of what ever pressure is above your liquid so that takes care of the density. Stick to metres head and forget pressure or the fact that they test in water.

It does take a while to realise that metres head, whether differential or NPSH is the same whether the SG is 1 or 0.3 or 1. 4. Pressure and power changes with density but not metres.

 

[georgeverghese]

Pls note that for low vapor pressure liquids, folks tend to interchangeably use metres of water liquid column as metres of actual liquid in service for NPSHr. This simplified approach can be a risk when operating with volatile liquids.
Preferably, with volatile liquids, you should insist that the pump manufacturer run the NPSHr test with the service fluid and NOT with water - the results may be quite different. With these volatile liquids, in critical services, it is better that a witnessed test be conducted with actual service fluid ( witnessed by Plant Owner rep or his appointed agent) -they will most likely charge you a lot for this.
For approx extrapolation of NPSHr value from water service to some other liquid, use Fig 10-25 in Perry Chem Engg Handbook 7th edition - valid only when extrapolating to pure component fluids.

 

[LittleInch]

The other key thing to remember is that NPSH is NOT the cavitation limit.

The way NPSH is measured is when the differential head falls by 3 percent as they restrict the inlet head usually by throttling the inlet. The pump can be busy cavitating like mad!, but it's the 3% drop which is what the vendor is looking for.

Also the margin between NPSH and onset of cavitation increases as you go past BEP.

Testing in volatile product is very difficult for the vendor and for smaller pumps is likely to cost you multiples of the pump cost to test it.

 

[Artisi]

Pls note that for low vapor pressure liquids, folks tend to interchangeably use metres of water liquid column as metres of actual liquid in service for NPSHr. This simplified approach can be a risk when operating with volatile liquids.
Preferably, with volatile liquids, you should insist that the pump manufacturer run the NPSHr test with the service fluid and NOT with water - the results may be quite different. With these volatile liquids, in critical services, it is better that a witnessed test be conducted with actual service fluid ( witnessed by Plant Owner rep or his appointed agent) -they will most likely charge you a lot for this.
For approx extrapolation of NPSHr value from water service to some other liquid, use Fig 10-25 in Perry Chem Engg Handbook 7th edition - valid only when extrapolating to pure component fluids.

Running performance tests with the service fluid is a bit of a stretch considering any respectable rig could contain an enormous amount of product plus the considerable costs involved in hazard protection for volitle liquids.
There is sufficient data available to extrapolate for other than water.

 

[uselessengineer]

From all points above, i gathered the normal practices is just to take the npshr tested with water by Vendor.

Npsh margin = npsha in meter of pumping fluid - npshr value provided by Vendor (tested using water).

There is no need to express the npsha to equivalent head in water column.

 

[shvet]

There is sufficient data available to extrapolate for other than water.

Could you please share some related to the liquids having viscosity less than water?

 

[LittleInch]

From all points above, i gathered the normal practices is just to take the npshr tested with water by Vendor.

Npsh margin = npsha in meter of pumping fluid - npshr value provided by Vendor (tested using water).

There is no need to express the npsha to equivalent head in water column.

Correct IME.

If you're getting down to a margin of less than 2 to 3m then you might need to investigate further.

 

[pierreick]

Another resource page 3-8 ,3-9
Pierre