Centrifugal Pump 2

[esperantes]

Hello Pump Experts,
We have an installation where a 160 k W 2900 rpm 10 bar centrifugal pump pumps river water from a tank to a centre pivot used for irrigation purposes. The discharge valve is throttled to maintain 6-7 bar to the center pivot. This pump was operating fine this way until it was replaced by a 160 k W 2900 RPM 12 bar centrifugal pump. It is worth mentioning that a 12 bar rated pump was selected because in some rare occasions we require to pump river water to a very distant location. The pump was replaced because both impeller and casing were worn after several years of operation. The volute casing, impeller and shaft of the new pump are made of stainless steel. The shaft sealing is packed gland. After one week of operation, the two bearings of the new pump failed. We were told by the pump supplier that the throttling of the discharge valve caused the bearings to fail. We were told to trim the impeller diameter to 260 mm to achieve 6 bar as per the pump curve. We did so and put back the pump into operation. The discharge valve was kept fully open and the discharge pressure was 6 bar but the bearings failed again after 12 hours of operation. What could cause the bearings to fail? The pump curve and drawings are attached.
Thanks
Esperantes

 

[LittleInch]

don't mention the flow rate? you show a "duty point" of 270 m3/hr, but what is your actual operating flow?
What are the characteristics of the irrigation system in terms of pressure at the centre vs flowrate?

Also what is the inlet condition of the pump?

What is the difference in head between tank water level and pump inlet level?

How far is the pump from the tank and what diameter pipework?

What pressure is at the pump discharge if you're controlling 6-7 bar at the centre pivot?, i.e. how far away are you doing this?

What sort of control valve are you using?

However if your discharge pressure was 6 bar (60m) and assuming your tank water level was say 1-2m above the inlet, you appear to be running off the end of the curve of your 260mm impellor. This leads to high vibrations and bearing failure. but failure in 12 hours is pretty bad. Is the pump suitable for "river water"? Are the bearings sealed from silt and dirt in the water?

Basically you need to control on flowrate, not discharge pressure.

In essence it looks like you have the incorrect sized pump for your duty.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Artisi]

How is the pump driven, direct coupled to the driver or belt driven?
12 hours operation is a disaster?
Is the pump operating correct direction?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

Suspect it's close coupled.

The wrong direction one is always worth looking at when you change units alright.

Also what was the cause of the worn pump? Lots of grit in the water? incorrect operation?

BTW the pump supplier "We were told by the pump supplier that the throttling of the discharge valve caused the bearings to fail." excuse is utter rubbish.

It's all about the flow through the pump. Throttling the discharge per se has nothing to do with it, but operate the pump beyond it's limits and the pump will fail.

I suspect that your new pump is not built to the same standards and thickness of your old one and won't handle any level of abuse.

Do you have the pump curve of the old pump to compare them?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[pierreick]

Hi,
Who is the manufacturer ? Which country ?
2900 RPM is quite high!
Pierre

 

[LittleInch]

"2900 RPM is quite high!" Eh? This is the speed of a 2 pole 50htz motor. Very common pump speed for this size of unit.

But yes, a "no name" 160kW pump is a bit suspicious alright.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[pierreick]

Hi LittleInch,
To me 1450 rpm is a more standard value for this type of application ( references Moret, KSB )!
Based on experience .
Note : you may want to calculate the Specific suction speed and comment . My calculation is about 14000 for 270 m3/h, 2900 rpm and NPSH R of 5 m , NSS (gpm) is 12000 far above the upper limit of 9000 . High risk of deterioration for the pump .

https://www.engineeringtoolbox.com/suction-speed-pumps-d_638.html

Pierre

 

[esperantes]

Dear LittleInch,
Thank you for your reply. I'll try to answer the points you raised.
don't mention the flow rate? you show a "duty point" of 270 m3/hr, but what is your actual operating flow? 240 m3/h @ 5 bar. We throttle the discharge pressure to around 6 bar so that the pressure at the delivery point is around 5 bar.
What are the characteristics of the irrigation system in terms of pressure at the centre vs flowrate? The nozzles for the center pivot requires water at no more than 5 bar pressure.

Also what is the inlet condition of the pump? The pump inlet is connected to the tank through a gate valve.

What is the difference in head between tank water level and pump inlet level? 2 metres

How far is the pump from the tank and what diameter pipework? 1.5 metres, 200 mm

What pressure is at the pump discharge if you're controlling 6-7 bar at the centre pivot?, i.e. how far away are you doing this? The discharge pressure is 6 bar and there is one bar pressure drop to the delivery point.

What sort of control valve are you using? Manual butterfly valve

However if your discharge pressure was 6 bar (60m) and assuming your tank water level was say 1-2m above the inlet, you appear to be running off the end of the curve of your 260mm impellor. This leads to high vibrations and bearing failure. but failure in 12 hours is pretty bad. Is the pump suitable for "river water"? Are the bearings sealed from silt and dirt in the water? Yes.

Basically you need to control on flowrate, not discharge pressure.

In essence it looks like you have the incorrect sized pump for your duty.

 

[LittleInch]

esperantes,

I appreciate you're giving the best information you probably have, but the numbers and the data are not tying together.

For the case now at 260mm impellor, if your flow rate was actually 240 m3/hr (Is this actually being measured or it is derived from something else?) then I would expect your discharge pressure on the pump itself to be more like 8 bar and not 6 bar. how accurate is your pressure measurement / how are you measuring. If this is a guage, what is the span of the guage (0-10 bar or 0-16 or??) or a digital guage?

If on the other hand you think the discharge pressure is more likely (more accurate) to be 6 bar with the valve fully open then the flow rate would more likely be about 280-290m3/hr and well off the end of the curve, hence high vibration and damage to the bearings.

Manual butterfly valve is not too bad for this duty, but is not the best for long term throttling. How is the valve controlled - variable via an actuator or a lever with set positions?

If you repair your pump and still want to operate it like this then you need to keep the pump discharge pressure between 7.5 to 8 barg inoredr to keep the flow rate within the max allowable.

It's difficult to know what the pump was doing with the big impellor as you haven't said what the pump discharge pressure was during that operation.

It could also be that your impellor or some other part is not properly balanced. Or indeed as artisi says, is it going round the right direction? 3 phase motors can easily go the wrong way. Someone needs to check, not assume.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[esperantes]

Thank you LittleInch. The flow rate is measured by a mechanical flow meter. You're damned right about the discharge pressure, 8 bar instead of 6 bar. I am worried that throttling from 8 to 6 bar did cause the bearings to fail again. The pump is running in the right direction. The discharge pressure was 12 bar with the original impeller. The valve is a manual one with a hand wheel. We have been instructed to trim the impeller to 260 mm to get the duty back in the middle of the curve.
Thank You.
Esperantes.

 

[esperantes]

Hello LittleInch,
One information which I believe is important is that the required NPSH for the pump is 5.1 m(as per the pump datasheet) while the difference in head between the pump inlet and the tank is only 2 metres. Do you think this could be the cause of the pump cavitation?
Thanks.
Esperantes.

 

[LittleInch]

No.

You need to read up what NPSH means but with a water supply higher than your inlet by 1 to 2m and a very short inlet line, your NPSH is about 11m.

If you're running at 240 m3/hr @8 bar discharge then you're in the right hand end of the curve but not so far that it should be a problem.

The other issue to check is the pump and motor alignment. Has this been done each time the pump is installed? If not you need to do it every time you do something to the pump. Is the pump foundation solid enough?

What sort of coupling do you have?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[esperantes]

Dear LittleInch,
Please excuse my ignorance, how did you arrive at a system NPSH of 11 m? NPSHs=source pressure + atm. pressure-losses-vapour pressure of fluid
Esperantes

 

[pierreick]

Hi,
If you want to learn about centrifugal pump , consider to study this document .
Pierre

 

[LittleInch]

Try this for details

https://www.eng-tips.com/faqs.cfm?fid=1598

But this all works in absolute terms in metres head.

I should have said NPSHA - Available.

So source pressure = 0 ( I assumed this was an atmospheric tank)
Static head above pump inlet / midline is say 2m
Atm pressure at sea level is 1 bara so ~10m
Losses - say 1m for entry losses
Vapour pressure of the liquid - water at 20C is virtually nothing, say 0.5m

So NPSHA = 0+10+2-1-0.5 = 10.5m

Your pump NPSHR (Required) is 5m. So your margin is 5m. this should be more than sufficient to avoid cavitation ( usually add a margin of at least 2m above NPSHR to avoid cavitation)

Now again this assumes as you haven't said otherwise that there are no other losses in the inlet - no filter, no constriction, no blockage, no valve half open. Have you checked that the tank isn't full of silt or leaves and that the inlet line is clear?

currently I suspect the pump is aligned, but maybe it is. Or maybe it's just a cheap rubbish pump?
It's very rarely the pump that's at fault though - check everything else first.
And don't believe the pump supplier - they don't sound very knowledgeable to me and are just looking for excuses to avoid the blame. IMHO.


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.