NPSH test report 5

[asifraza0]

Dear Expert:
Need your help
I'm looking over a NPSH test for a centrifugal pump. The vendor had tabulated this report for 8 points over varying suction pressure. The cut-off is 3% dH/delta H. At this point the NPSH is 1.969 meter.

The vendor had done this test using water while the pumping fluid is LIquid Argon. vendor published NPSH (req) is 1 meter.

So I'm using the test results (at 3% loss in head) of 1.969 meter. I've converted 1.969 M into Argon head of liquid using Argon specific gravity of 1.37. This gives me a
NPSH of 1.43 meter.

Am I doing this right? Is this is the correct way to interpret this data?

Thanks in advance for your help



Asif Raza

 

[BigInch]

Probably, but nobody cares. Head of argon in that context is a meaningless number. The pump still needs at least 1.969m of head, if it's pumping water, gasoline, or argon. And it's probably best to add a little bit more to that as well. 1.43m won't do it.

What is your available suction head (of argon)?

you must get smarter than the software you're using.

 

[LittleInch]

Are you doing this right - I don't think so. Whenever I do this calculation for lower SG products I don't multiply he NPSH figure by the relative SG, I allow for any density change by making the head equivalent of atmospheric pressure higher than that of water, not by changing the head figure. If you don't do it for the discharge head then you don't do it for the NPSH figure.

Working to a very tight NPSH limit is a bad idea in my experience as there is nothing here about when you get onset of cavitation. It may be the same or more likely higher than the NPSHR limit.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[asifraza0]

Big Inch: The calcualted available NPSH is 1.82 meter.
Little Inch: The NPSH test was done on water and the NPSH at 3% head loss (Onset of cavitation) was 1.969 meter. All I'm doing was converting this head of water into head of Argon liquid.

Asif Raza

 

[BigInch]

Fine, but the pump needs at the very least 1.969m and should get more.

Why are you calculating an equivalent head of argon. Because you can?
That's not needed for anything.

you must get smarter than the software you're using.

 

[LittleInch]

Asif,

We can see what you're doing and we're saying it is not correct. You don't modify the head figure.

Also your statement " NPSH at 3% head loss (Onset of cavitation) " is fundamentally incorrect, but is a common misconception. Cavitation can commence at inlet heads considerably above the NPSH figure. The test for NPSH had to be based on something measurable (3% drop in differential head), because cavitation can be affected by things such as small amounts of entrained air and relies on noise - difficult to do in a test environment. Ask the vendor for the onset of cavitation curve (if he has it) and you'll see the difference.

ending with 1.82m of NPSH will give you a serious issue whichever type of pump you choose, so perhaps you need to do a bit of a rethink or allow for changing out the impellors on a regular basis as they get chewed by cavitiation....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[1gibson]

NPSH tests can be hard to read. At what flow was the NPSHR 1.969m, and was that a higher flow and your design flow where 1.8m was expected?

You should have multiple flow points (each with multiple suction pressure points, each making an "L shaped curve if you rotate the L 90* clockwise, with the angle representing the 3% head drop.)

Connect the angles of the L's with a curve. Do you have a proposal curve with an NPSHR shown on it's own graph, below the head/flow? If so, the curve you drew on the NPSH test report should match.

The pump does not need 1.969m NPSHA minimum, it will require less NPSHA at lower flows, and more at higher flows. Check the flowrate on the test report where you saw 1.969m. If it is higher flow than design, then maybe it is acceptable.

This is not the end of the world if you have been sent the test results for approval. Modifications can be made to the impeller to improve the NPSHR, the only cost to you will be a schedule hit. Talk to the vendor about it.

 

[asifraza0]

NPSH test results attached

The test was done at the same flow rate for all points. This flow shown in the test is the design flow. vendor mentioned in his quote / NPSH curve that the NPSH required is 1 meter. The last point is 3% (onet of cavitation?), the NPSH required is 1.969 meter at design flow.

Need help in undestanding the test results whether my pump with NPSH available of 1.82 will work or not?

Thanks




Asif Raza

 

[1gibson]

Ok, now you can start to worry. Test report indicates that there is an inducer (do a little research on this.)

With an inducer, the NPSHR increases at low flows as well as high flows, because the inducer will be designed to optimize (reduce) NPSHR the rated flow. An inducer is only used when the pump (without inducer) has too high of an NPSHR requirement for the application.

The rated flow may very well have the lowest NPSHR anywhere on the curve, so I'd say that no, it will not work. You can review what you purchased and see if NPSH tests at multiple flows is a requirement, if so request it.

No, you can't use this pump with 1.82m NPSHA. Vendor should review the inducer design, and/or do some work to the impeller inlet. If they push you to accept as-is, push right back and ask for additional testing, and assurances. Make it is easier for the vendor to modify the pump and meet the original requirement, than it is for them to satisfy your requests for additional testing, information, extended warranty, etc. that you would need before you accept it as-is.

 

[asifraza0]

Ok, thanks, this helps..
I know what is a inducer and what is does. But thanks anyways for letting me know.
The vendor initially said that 1 meter woud be required and we designed our system to give 1.82m to make it work.
I will contact the vendor to modify the pump & impeller design so that NPSH (req) can be decreased to 1 meter.



Asif Raza

 

[chicopee]

Looking at the physical properties of Argon, at 1 atm., the boiling point is -302.6dF and the freezing point -308.6dF. The difference is about 6dF which is a small temperature difference that would not take much to turn the liquid argon into a saturated vapor. Do you think that a 1.82m NPSHA would be enough and not merit an increase?

 

[asifraza0]

I agree that 1.82 meter is low for this application. But that's all we have based on the layout. The Argon column cannot be elevated more than what is is now.


Asif Raza

 

[GPRnD]

It would help if you could state where the pump will operate on the Argon vapor pressure curve. The reason being that high vapor pressure fluids demonstrate a noticeable thermodynamic effect (also known as the hydrocarbon correction factor). This can reduce the effective NPSHr you need. Refer to HI 1.3 for more discussion on this.

I'd also want to check the minimum flow because as chicopee noted, small temperature rises through the pump can cause flashing if not properly managed. We typically run this calculation for the full pump flow range to establish recommended limits.

The NPSH test is rather poor IMO in that they only tested at one flow point. You thus have no idea what is happening at higher or lower flows. Inducers typically exhibit U shaped NSPHr curves so unless your pump will only ever operate at 1380 l/min, I would strongly recommend asking for more test points.

 

[chicopee]

RurhRpumpen offers cryogenic pumps and claims to have extremely low NPSH requirements. Check them out and here is their site

http://www.ruhrpumpen.com/product/VLT

Cryogenic.

 

[1gibson]

Yes, the solution would be a VS6 (vertical can) pump. Whatever NPSHA you need, can be built into the pump with distance between the suction flange and the 1st stage impeller.