Hi all,

I have a centrifugal pump/motor driven by a VFD that makes a very disconcerting sounds as it speeds up (hear attached video)

The motor had alleged never been regreased (note Zerk fittings) and sits in a location where it is is often exposed to steam. Given this context, I think that it's not unlikely that the sound is coming from the motor bearings. However,  removing and sending the motor off to be fixed is quite a hassle and expensive here in Spain and I'd feel a lot better the more confirmation I got on my suspicions of the cause... anything recognize this noise?



Videos (1)
Original Post


Do you have vibration data to go with the video?  Also, you state that its "doubtful" it has never been greased. Is the video/sound taken in that state? If not, have you tried to regrease and see if it still sounds the same?

You probably realize this now but this machine and any other machines should be on a documented lubrication program.

Jim P 

Yes....the video and sound played fine.  You're not running Windows Vista are you?  See link below

The motor/pump was running at some speed and ramped up to a higher speed.  You could hear a somewhat high pitch sound which got louder and with a higher pitch as you might expect.  It was hard to figure out (at least for me) whether it truly was from the bearings or perhaps some electrical noise.






Hi Jim and Walt,

Thanks for trying and watching the video. Since then, I have done some things. First of all, taking off the pump head (but not adapter and coupling) = sounds worse. Then, I spoke with the operator who had said he had regreased the motor for the first time in so many years last month. Turns out he used the food grade aluminum soap grease we use elsewhere in our process... So I went and purchased some NLGI 2 lithium grease (manual recommends 3 but that's too hard to find here) and through-flushed until clean grease came out the other side. The grease that came out at first was pretty dirty and kinda smells burnt. I turned the motor on to 20% speed again, and to me it might have sounded a little better, but it's hard to say. The sound is definitely still there. When I rotate the fan by hand, it moves quite freely - however, I came back in the middle of the night when there were no other sounds in the plant and only then can hear a faint scraping sound when rotating by hand. 

At this point, I'm thinking it is definitely something with the motor and will send it off to be taken apart in a shop. Will let you know if it's the bearings!

Oh and don't have vibration data on it or any of our other motors. They're all less than 10 kW but I was thinking about getting some of the new app-linked sensors for our flagship facility in California - they would spot this, yeah?

Ok....most likely you have a bearing fault problem but this type of diagnosis is VERY hard to confirm without quantitative vibration data. Based on your response, sounds like you have no vibration monitoring program at all....or, you do but this machine not monitored because it is less than 10kW ?   In any case, the decision to monitor a particular machine should be based on more than its size; the fact you posted this issue on this forum indicates this machine probably needs to be part of your PdM program.

To answer your question, yes....trended vibration data should have given you a clue as to the root problem.

Let us know what you find. Feedback is often lacking but important to those of us who provide input.


Jim P


I am surprised that you are trying to run a centrifugal pump at 20% speed and get any flow out of it. If the motor nominal speed rating is 1800 RPM, then 20% speed is only 360 rpm. It would be a little difficult to listen for bearing noise at such a low speed; especially with the electrical noise from the VFD. I'm also not sure what you mean by taking the head off the pump running it. If the motor bearings have not been lubricated in years, and with the wrong grease, then how can one assume the rest of the machine has been properly maintained? Sending the motor to the repair shop without any diagnostic information and without any attention to the pump is simply Reactive Maintenance; like waiting for the shaft to stop rotating.


Hi y'all,

You're right at 20% there is no flow and in fact here in Europe with 50 Hz it's only 1500 rpm (1455 on the plate). I was only running it this slow (essentially just the motor - removed pump casing, impeller, and m/seal but there is what Alfa Laval calls an adapter instead of a coupling that I left on) because the noise is quite evident even at this speed. When you say noise from the VFD, I'm a little unaware of what you mean... I'm definitely still learning the practicalities of motors, as I know more about distillation (and pumps to an extent - I studied chemical engineering and worked in a chemical plant before switching over to the wine industry) but my company just has one engineer and that's me and therefore I get to be in charge of it all! I've only been here one year and yes am definitely still chasing the lack of (preventative) maintenance from the person who left the job opening for me. Although, here in Spain, the case is much worse than back in California. We do grease moving parts there if they can be... I'm definitely, for example, going to through-flush the rest of the motors with the new lithium grease and set up a schedule for the operators to do it when I'm back in the States. Do you think it will be a big deal to use the NLGI 2 vs. NLGI 3? (As in, what could be the consequences?)

It's also true that we have no vibration monitoring program at all. When I took this position that includes maintenance in addition to distillation design and operation,  I did read about it and wanted to get the GTI iAlert2 but the local salesman kinda ignored me and I was forced to move on to other tasks! I think I will use this as an example to my boss to restart the convo to implement these. Our similarly-sized motors back in California have the hermetically-sealed, pregreased bearings and I'd think that those would likely need to be monitored even more than these through-flush ones (once they're on a lubrication schedule). The only other moving parts are drives that spin a shaft inside the distillation column itself. Much of the PM is, for example, cleaning HXs which also was taught to the Spanish team for the first time during my visit here...

It's difficult to tell if the sound is abnormal, because we don't know what it sounded like before. I hear the tell-tale whine of a VFD carrier frequency (the rate at which the tiny little DC pulses fire), but if this is above and beyond that, you may have EDM damage to your bearings because of the VFD output and poor grounding. EDM damage is the result of a voltage build-up between the stator and rotor as a result of the high frequency pulses from the VFD causing what's called "common mode" electrical noise, meaning it is referenced to ground (earth). CM noise always wants to return to its source, in this case the VFD, so it looks for a path through ground. If none exists, the voltage builds up a charge on the rotor and it eventually exceeds the dielectric (insulating) strength of the grease, then discharges across the bearings to the frame to ground. As it does so, it causes microscopic "pits" of voltage damage that is essentially the same as Electrical Discharge Machining (hence the name) such as a welder or plasma cutter. Once a pit is made, it is easier to find the next go around and the process escalates to destruction.

Here are some images from someone else in this forum as to what this looks like.


The solution is to use a "shaft grounding ring" on any motor run from a VFD, or the next time you replace the bearings, replace the seal with one that includes a shaft grounding system. Inpro/Seal offers them  as do others. (I don't work for any of them them.) Another alternative is totally isolated (ceramic) bearing assemblies that don't actually stop the voltage built-up, but it stops the bearing damage.

If this noise has always been like that, it's just what I mentioned earlier, a whine that is the side effect of being run from a VFD. The high speed pulses that comprise the pseudo-sine wave output has a magnetostrictive effect on the steel laminations in the motor core. You can increase the carrier frequency rate in the VFD output via programming, but there is no "free lunch" in that increasing the rate will make the motor appear quieter (it actually just moves the sound above human hearing), but causes more heating inside of the VFD, so you have to de-rate the VFD. If you have no head room in the VFD size, you can't do it.



Add Reply