Hello Forum members, I have little

[esperantes]

Hello Forum members,
I have little knowledge in pump engineering and would like to have your honest judgment. We are having an issue with a 2900 rpm centrifugal pump which we just purchased. The pump is designed to operate at 12 bar 5.1 m NPSH. Our application is such that we have sometimes to throttle the discharge valve to get 8 bar on the discharge line. We understand that this is not an effective way to operate the pump but we had been operating this way for years with the old pump without any issue. The only reason we replaced the old pump was that the impeller was worn out after 30 years of good service. When we operated the new pump with the discharged valve throttled the double ball bearings on the pump failed after two days of operation. The bearings are oil lubricated, and we could see that the thermal degradation of the oil caused the bearings to fail. The manufacturer is claiming that a throttled valve condition caused lot of stress on the pump shaft and bearings, due to excessive internal circulation in the pump volute casing and this caused the bearings to fail. They are now proposing to put a variable speed drive to reduce the speed of the pump. This is obviously a very expensive solution.
Any advice?
Thank You
Esperantes

 

[LittleInch]

You need to post the old pump curve, the new pump curve and tell us what flow rate you're working at.

If the flow rate you're operating at is a low percent (<30%) of its BEP then your pump is probably too big, but let us see the data.

At the very least tell us the duty point of the pump, power etc

Is this the same as your post last year about a similar pump?

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

 

[Valvecrazy]

Your old pump lasted 30 years. Was it being throttled as well? Throttling should not hurt a pump/motor. Sounds like those double ball bearings were radial bearings and not thrust bearings as they should be?

 

[georgeverghese]

There should be a min flow recirc line on the pump immediate discharge ( going back to pump source tank or vessel) to keep flow above the min flow point when the pump is being throttled. What are you pumping btw?

 

[LittleInch]

It sounds very similar to his last post.

https://www.eng-tips.com/viewthread.cfm?qid=497765

So "river water".



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

 

[georgeverghese]

If you are running at 240m3/hr on a pump with BEP at 270m3/hr, that is well clear of the min flow which I guess is that line at about 80m3/hr on the 311mm impeller (which is about right for min flow at 30% of BEP flow).
My guess is that the internal river water supply line for cooling the oil flooded shaft bearings is probably clogged up with dirt, since you say there is thermal degradation of the lube oil at the bearings.
I have doubts if what you say is 240m3/hr is correct for a change in developed backpressure, downstream of the throttle valve, from 12bar to 6bar. A 50% reduction in developed backpressure implies flow has dropped to a quarter of 270m3/hr = 65m3/hr, which is below min flow.
Anyway, damage to the bearings within 2 days tells me this is more about bearing cooling.
Since the flow curve is flat at min flow, the only way to make min flow work is by flow control (FIC-FCV) and is much cheaper than VFD from a first cost point of view. A VFD soln saves you money in the longer term on operating cost by reducing power demand.

 

[georgeverghese]

" I have doubts if what you say is 240m3/hr is correct for a change in developed backpressure, downstream of the throttle valve, from 12bar to 6bar. A 50% reduction in developed backpressure implies flow has dropped to a quarter of 270m3/hr = 65m3/hr, which is below min flow."

This observation is not correct, since the 12bar backpressure on the pump refers to another location much further down this line. So this extrapolation is not valid.

 

[esperantes]

Hi,
Thanks for your replies. Attached is the pump curve of the new pump.
Esperantes

 

[LittleInch]

Flowrate you operate at?

Also do you have a sectional drawing of the pump?

Is this river water again?

Do the bearings have a seal in contact with the pumped fluid?.

Was the pump aligned before you operated it?

Was it going the right way around?

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

 

[esperantes]

Hi Littleinch,
My replies below.
Flowrate you operate at? 250 m3/h

Also do you have a sectional drawing of the pump? I'll look for one

Is this river water again? No, clean water form a borehole.

Do the bearings have a seal in contact with the pumped fluid?. The bearings are oil lubricated

Was the pump aligned before you operated it? Yes

Was it going the right way around? Yes

Esperantes

 

[pierreick]

Hi,
Did you check for vibrations? Can you share a picture of the installation of the pump with the piping around?
Can you share your specification sheet associated with the purchase of this pump? min, normal and maximum operating condition.
Can you share the sketch or PID/PFD associated with this pump?
Can you share the specification sheet provided by the vendor and model selection?
Who is the manufacturer and what is the reference of the pump?
Why did you specify 2900 RPM?
etc....
Note: Should be part of the scope of a mechanical engineer or a chemical engineer.

My view.
Pierre

 

[LittleInch]

Sorry, last questions - what is the inlet pressure during operation?
what size impellor have you got fitted?

The bearings might be oil lubricated but is there a gap between the bearings and the pump? i.e. if the seal doesn't work well can the water get into the bearings?

And yes, it might not be ideal to operate the pump at discharge of 8 bar, but it is still often a lot cheaper both in OPEX and CAPEX to do so rather than using a VFD.

But providing you have sufficient inlet pressure, at 240m/hr / 66 lps, you pump would seem to be operating right where it should so your manufacturer would appear to be talking rubbish. IMHO.

But send details of the ump itself or the sectional drawing as it could be something else.

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