Experienced 5 NDE back to back angular contact bearing (SKF 7311 BEBCM) failures in two years (2006 to 2007) on Pump”A”. From mid 2004 to March 2006 ran trouble free. A second “B”unit (exact duplicate) was installed near end of 2005 and till today this pump has run trouble free

The two pumps are both an 8 stage David Brown Union horizontal pump for water injection 85 bar 200 m3/hr, Driven by reciprocating engine and 3.179 gearbox.

I got involved in the middle of all this. When I came into the picture we had already experienced 3 failures and the pump was in bits and pieces in the shop. The failed SKF bearings looked as if someone took a cutting torch to them. One of the worst I have seen

Initially I was led off track as was told the pedestal mounts were flexing. Remember I never saw the unit running at that time. It was a non OEM pedestal and did look flimsy and the blocks where the pump feet set and are bolted to the pedestal were very corroded to the point they required changing. So with the pump in the shop we stiffened the pedestal. When we started up there was a resonance problem at 3x engine speed. The pump 1x was 0.5 mm/sec rms and all pump related harmonics looked low. We then had to cut off a portion of the stiffeners and lowered the overall vibrations from 5.1 to 4.4 mm/sec rms which is thought to be acceptable. The engine 3x harmonics were significantly lowered. However an engine 2.5x dominated the frequency. We no longer allowed operations to run in two pump mode (we are going to up grade the process control system at a latter date. We thought then we had resolved the problem but failed another bearing in Nov 07. After the shop repair when we modified the throttle bush we have ad two failures on these bearing and each time the same, the cage broke

When the pump was in the shop and due to the observed angular contact bearings “fried” condition I consulted the pump OEM. Also there was correlation that the pump started having problems when going from a single pump to two pumps in parallel (when B was installed). We observed that the pumps were often running close to dead head conditions. The min flow valve for these units are an ARC type that are of marginal design to handle the pumps min flow requirements (50 m3 pump min flow versus 46m3 ARC design). The OEM recommended to change the throttle bush diameter to lower thrust loads from 1200 lbf drive direction to 100 lbf drive direction rated flow and 290lbf at low flow condition away from the driver. Question- Could we have lowered the thrust too much?

We have installed a vibration safety shutdown to protect from catastrophic failure and it had paid for itself. However we still cannot find the physical cause of the bearing cage failing only after a few weeks in operation. Can anyone offer advice on why cage failures in our case and what I should be looking at.
Regards, Pokey

rca_for_web.ppt (2,383 Kb, 46 downloads) rca

I think the "M" in BEBCM indicates a machined brass cage.

Cages wearing to failure are often lubrication problems, since so much sliding occurs in the cage.

Cage breakage without lots of pocket wear would make me suspect the balls are not tracking correctly, and exceeding the pocket clearance.
For the balls to run true, the races must run true, and that >>requires<< that the Bearing seat diameters and shoulders ON the shaft and IN the housing must be gorgeous, flat, and smooth, and run true within ~ 0.0001 inch. That first bearing failure may have damaged the shaft in ways that no amount of hand work with a fine stone (or machine base rework) can repair.

If the pump is running now with fresh but crooked bearings I'd expect various hi-frequency anomalies, and maybe vintage Spike Energy over 1.0.

At the next bearing replacement I'd get the shaft and housing dimensions re-qualified by someone using, and familiar with, micrometers and dial indicators reading in 0.0001 inch increments. No disrespect to the maintenance gang, who are probably under immense pressure to get that pump back on line, but not many maintenance shops are equipped to make those kinds of measurements.

Is a new shaft available?

I agree with Dan. Cage failures are usually lube or brg seat alignment. Brg journals and shoulders should be checked for runout relative to each other. Shoulders are often overlooked. Sometimes the housing brg cap spigot is the shoulder for the outer ring so this must be checked also. If all this has been checked and os ok try swaping pump positions and see if the problem follows.
Cheers