Thrust bearing failure 1

[Marke]

I recently came across a submersible pump where the thrust bearing had failed and decided to investigate the cause.
The response was below.

This issue is driven by experts in drives, who have no understanding of hydraulics and pumps. With a bore hole pump it is critical to keep the pump operating within it's hydraulic limits. Just using the pipe line fill function does not necessarily do this. Each installation needs to be checked to ensure that there is sufficient system pressure to ensure that the pump is operating within the hydraulic envelope that the manufacturer advises, at all times. Typically we see situations where the drive sales person convinces the installer/owner that by utilising the line fill function they can do away with other safety devices such as pressure sustaining valves.

A classic case is when the pump is filling the lines from empty. The pump will start and ramp to minimum frequency. The line fill function takes over and 'manages' the pump from there. Unfortunately if the pump is operating at too lower pressure (system pressure too low) the pump will upthrust and damage the motor thrust bearing and pump components. I have attached a pump curve showing the hydraulic limits (of that particular model). If the pump operates any where below the shaded area, it is at risk of upthrusting.

If the system pressure is too low during the pipe fill function we recommend that a pressure sustaining valve be fitted at the bore head.

While this description does make sense, I have never seen/heard the warning being given.
By far the majority of installations that I have seen rely on the VFD controlling the flow to fill the line with no artificial head applied while the line is empty and this does not, to my knowledge, seem to have been a problem in the past.
Any thoughts?
Best regards,

Mark Empson
L M Photonics Ltd

 

[BigInch]

Can you post the diagram with the shaded limits and the model number to see if it looks like the limits that follow,

Constant speed centrifugal pumps should ideally be run between -20% and +10% of the BEP flowrate to avert potential bearing failures. VFD operated pumps should attempt to duplicate those conditions for the corresponding flow and head outputs they are generating at any given time.

The VFD operated corresponding limits would thus be,
maintaining flow withinb -20 to + 10% of Q_vfd = Q_bep / RPM_bep * RPM_vfd
while head is held around,
H_vfd = H_bep / RPM²_bep * RPM²_vfd



**********************
"Being GREEN isn't easy" ..Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com/

 

[JJPellin]

I am not familiar with bore hole pumps. But, if this is a multi-stage pump, then up-thrust can be a serious problem. The VFD will (I assume) slow the pump down to some minimum speed. But, at any speed, a multi-stage vertical turbine pump can produce up-thrust if it is operated at high flow (low head). Perhaps in most applications, this up-thrust is tolerable and will not result in failure. The pump manufacturer may be able to provide a thrust curve at the VFD minimum speed. We typically have these curves for our deep well water pumps. At very high flow, the thrust goes negative and can result in buckled shafts or thrust bearing failures. At lower speed, the up-thrust will be less than it would be at higher speed. So, determine the maximum up-thrust for the given pump at the minimum speed and compare this to the thrust load capability of the inactive thrust bearing in the motor. If the shaft is long and flexible enough to buckle, then the thrust bearing might not be the weak link. But your post describes a thrust bearing failure, so I assume that shaft buckling is not possible.

Johnny Pellin

 

[PUDCourt]

We recently had a similar situation where an irrigation pump motor prematurely failed due to thrust bearing wear (pumping routinely beyond the desirable limits). The pump technician who replaced the motor has recommended we install a pressure sustaining valve (PSV) to prevent a repeat. We plan to install a VFD in the short term to help provide a range of flows and are debating whether or not the PSV is needed. Our PSV supplier doesn't believe their product is suited for our particular application and is recommending we consider a motorized butterfly valve. Perhaps there's multiple ways to "skin the cat"? Seems like a VFD would be all that was needed.

 

[1gibson]

Also be aware that the pump manufacturer may only be giving you a flow limit that the pump goes into upthrust.

The motor manufacturer should advise a minimum thrust required to keep the thrust bearing from skidding (assume this is your mode of failure). Bigger the bearing (higher thrust capacity) the higher value for miminum thrust.

And of course, reduced speed will reduce the head ^2, which has a pretty direct effect on the pump thrust.

You might do better with higher speed during fill. Even if it shifts some of the abuse from the motor to the pump.

 

[NGLENGR]

Thrust is a huge headache in some of our multi-stage vertical can pumps. We don't have thrust balanced pumps, so the motor bearings have to be pretty beefy or we have to use impeller balancing tricks which can kill efficiency.

We had a 22-stage vertical pump launch the thrust bearing housing after running in a low/no-flow condition for about six hours. This caused the pump to up-thrust and the bearings began to skid. Eventually the balls locked completely and the shaft actually snapped in a ductile mode due to the heat generated where the bearings had seized.