NPSHr and Max. Allowable flow verification

[gilus02]

Hi;

I have recently received a vendor datasheet for a Cent. pump and the pump has a problem on NPSH margin at Max. flow. Then vendor replied us we will reduce the Max. flow in order to see you project Spec.
I got confused!
Is it possible to reduce a pump Max. flowrate?
As this pump is API, is there any problem in performance test?
How can I verify the proposed NPSHr is correct?

You can find the curve as attached.

Thank you all.

Mohammad

 

[LittleInch]

Mohammad - you again(!) I'm just as confused, but I think something is getting lost in translation. What your vendor has done is give you a bigger pump where your duty requirement is met further to the left on the pump curve and hence ower NPSHR.

There should be no problem with performence test so long as the pump supplied matches your requirement.

You can only check NPSR in a pump test, but this is notoriously difficult. NPSHR is actually defined at the point where the differential head drops by 3% at the same flow rate when you throttle the suction. This is NOT the same thing as the onset of cavitation which can start from 1 to 5m higher than the NPSHR limit, but is very difficult to define. If you're worried about NPSH, ask the vendor for an onset of cavitation curve which could prove very illuminating...

Often you end up with a larger pump running at 1440 or slower using a 4 pole motor if you have NPSH problems.



My motto: Learn something new every day

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

 

[gilus02]

Little Inch where is BIG inch? I missed him!

I'm following your motto and try to learn something new every day but not today yet because I still don't know how a pump manufacturer calculate NPSHr and Max. flowrate?

 

[LittleInch]

He doesn't calculate it - he designs it to 610 or something simialr, builds it at max / min impellor size and then tests it. As I said above, I think when the vendor said "reduce max flow" I think what he really meant was "reduce the flow duty point compared to max flow", but might not have used all those words.

This (moving left on the pump curve) normally results in lower NPSH, but as I said, be careful because NPSHR figure is only a guide as to what point cavitation starts at and that's the key issue when it comes to low suction pressure. A pump will actually work whilst cavitating, just not for very long before your impellor disintegrates....

My motto: Learn something new every day

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

 

[TenPenny]

It depends.

'Max flow' can be defined in various ways. If you have a pump where the max flow specified by the manufacturer is 7500 m3/h, and this results in the NPSHr being too high, the manufacturer can now say the max flow is 7250 m3/h.

What do YOU mean by 'max flow'? Is it the maximum flow RECOMMENDED by the pump manufacturer? If so, then what the manufacturer is doing is simply recommending a maximum flow rate so that the NPSHr does not go too high.

I can tell you that most manufacturers define it that way, as a recommended flow rate that shouldn't be exceeded, and therefore the manufacturer can change that number based on your application.

What exactly do YOU mean by 'max flow', and what exactly does the manufacturer mean by max flow?

 

[1gibson]

Minor clarification, NPSHR is 3% drop of the head from the first stage impeller. If you use 3% drop of total differential head on a multistage pump then it will make the results more favorable.

Max flowrate is whatever the manufacturer specifies, just make sure it is clearly identified in all documents (especially the proposal curve.) If you do not need to run at end of curve, and end user can prevent this operating scenario, then it is just numbers on paper, it can be changed to suit.

There are no problems with reducing the max flowrate on an API pump, it just needs to be specified. Definition per API:

allowable operating region
portion of a pump's hydraulic coverage over which the pump is allowed to operate, based on vibration within
the upper limit of this International Standard or temperature rise or other limitation, specified by the
manufacturer

NPSHR is one possible "other limitation."

 

[JJPellin]

For an evaluation of NPSH margins, I would suggest making sure of your definitions. NPSH available was determined for some assumed set of conditions. Our specifications require the use of level at bottom tangent or bottom nozzle elevations. This could be several meters below the actual operating level.

NPSH(a) was also based on some assumption for vapor pressure (or product temperature). Be careful if the analysis was done based on a group of worst case conditions (highest temperature, lightest product, lowest level). This worst case combination might never exist.

NPSH required (at 3% head loss) is most commonly used. But, some customers insist on a 1% head loss definition. Be sure you know which was used.

Most customers would require a specific minimum NPSH margin. We require 5 feet for water and 3 feet for all other services. Was this margin added onto the NPSH requirement? Make sure you are not “double-dipping” on the margin. If it was added to the NPSH required and you use it again to evaluate if the margin is adequate, you are applying it twice. I have caught contractors making this mistake.

I do not offer these suggestions because I like to purchase pumps with a lower NPSH margin. I like a big margin. But, if the analysis drives me to make a bad decision, it should be questioned. If it drives me to purchase a pump that is too large just to run it at the left of the curve, this could be a mistake that might create more problems than it solves. It if drives me to purchase a very large, very slow pump, it might result in lower efficiency which will cost me much more in energy over the life of the pump.

Take a look at the NPSH margin using expected, actual operating conditions just to give yourself a second point of reference. If the NPSH margin using the definitions required by the specifications was 4 feet, but the NPSH margin using the actual operation conditions was 20 feet, I might want to consider deviating from my minimum NPSH specification in order to avoid purchasing a pump that would be inefficient or unreliable.


Johnny Pellin

 

[Pumpsonly]

The max flow you are referring to in your case is normally referred as end of curve flow rate.
The end of curve flow can be manipulated by the pump vendor by just limiting the flow range to be plotted on the performance curve. They often do that just to overcome having to use a larger motor in case purchase specified the motor must cater for end of curve flow/power.
As 1gibson pointed out,the NPSHR test should be carried out on the first stage impeller only for a multistage pump as the 3% head drop should be based on the generated head of the first stage impeller.

 

[mimentzos]

graph... with the answers
From Europump guide " NPSH FOR ROTODYNAMIC PUMPS"

 

[mimentzos]

and one more

 

[gilus02]

I am grateful to those who shared their experiences with me. You have almost answered my question but two more points are still remained:
1- Because of NPSH problem, it isn't possible to work at end of the curve (let NPSHa increment be off the table), therefore vendor promised to reduce the maximum allowable flowrate just on curve and no modification for the pump (like impeller trim). In this case vendor will deviate API-610 preference for 120% BEP and put the maximum flowrate on 105% BEP.
What is unclear for me isn’t why the pump is so big or why we haven’t chosen a more suitable one. I’m thinking that could we run the performance test according to API-610 at 120% BEP for this pump?
How can I interpret this discrepancy between the declared max. allowable flow and the max flow that is necessary for performance test (120%BEP)?
2- On the other hand, as this pump must be started with open delivery valve and you know (better than me) the pump will start working from end of the curve, where NPSHr > NPSHa, so however the manufacturer make maximum allowable flowrate smaller but in every periodic start (as we have two pump, one operating and one standby) the pump will reach its real max. flowrate (where NPSHr is higher than NPSHa).
According to my reasoning I don’t know how the pump manufacturer can limit Max. allowable flowrate despite the performance test and the start condition?
Thank you all for your helpful posts!

 

[LittleInch]

1) You put the vendor in a difficult place and now expect him to solve your problem. If you insist then test the pump to 120%, nothing bad will happen.
2) Why must the pump be started with open delivery?? You can see there is a problem so fix the problem. The start condition is your problem to fix, not the pump vendor. You can't expect him to design your system, only give you a pump to your data sheet.

My motto: Learn something new every day

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

 

[1gibson]

Larger pump might be selected to reduce the head rise to shutoff, and reduce the MAWP. Ask the vendor why.

You need to realize that there is no "real" max flow. You can trim the end of the curve, or extend it, doesn't mean the pump will obey. The original end of curve flow is not going to be enforced somehow if the pump runs without enough system resistance.

 

[mimentzos]

According to your data the operation point is at the midle of the curve where the NPSHr is almost flat.
The system curve is OK ?