We have a non-vfd vertical motor with spherical roller bearing on top and deep groove ball bearing on bottom. Both upper and lower bearigs are designed to be insulatd form the motor frame. 800hp, 900rpm, 4kv.

This motor drives a pump with three water-lubricated sleeve bearings above the impeller (somewhat brackish water). Those sleeve bearings are made of some kind of steel-like metal on both the shaft sleeve and the bearing.

Pump had some abnormal symptoms regarding seal water, so we went in to take a look and found the middle pump bearing severely pitted. Metallurgist and pump guy say it looks like electric current damage.

Checks of the motor showed bearing insulation megger is bad. Inspection showed that the upper bearing thermocouple well which is supposed to be insulated from the bearing outer race had visibly deteriorated insulation. When this was corrected, the megger passed indicating both bearings were now (after fixed) insulated from the motor frame.

With this in mind, we can visualize a loop flow-path of current which would have existed. It would go from the shaft, through the motor upper beairng, through the grounded upper bearing themmocouple well, to the motor frame, to ground, to pump casing, through spider, then jump accross pump bearing back to shaft, up through rigid coupling to motor shaft to complete the loop. Why the current chose to flow in the middle pump bearing vs top pump bearing... might perhaps be a result of shaft bending mode... caused lower clearance at the middle bearing than the top.

The part I was asked to double-check was if the motor looked consistent with this scenario. Electrically, the ground we found was consistent. But we also inspected the upper spherical roller bearing and found just about nothing unusual on the races or rollers. The only thing we did note was a very light circumferential ring around the exact center of the outer/lower race which had absolutely no depth (couldn't feel it with your fingernails). I don't put much significance on that light mark.

The thing I wonder is why we didn't see some of the typical current patterns on the bearing (washboarding). I guess I can come up with some theories why current is more likely to mark the pump sleeve bearing than the motor rolling bearing (higher current density at the small contact on the sleeve bearing, as well as more intermittent contact at the pump sleeve... making and breaking the circuit creates marks at that location). But theories don't mean much.

my question: Has anyone had a similar experience where faulty motor rolling bearing insulation resulted in damage to the driven pump sleeve bearing? Did any damage show at the motor rolling bearing?

(note - for most motors, only the top/outboard bearing is insulated, in which case, the return current would be through bottom/inboard bearing and not pump).

This message has been edited. Last edited by: electricpete,

Electricpete,

This sounds more like electrolysis than the electric currents produced by the motor. Motor bearing currents are pretty rare in non-vfd motors. In fact, I have not personally experienced bearing currents in a not-vfd motor.

John J

I have not experienced something like that and have not seen the actual damage but I'd have questioned metalurgist's conclusion unless he is 100% confident that it is a clear electrical signature.

Electric current results in surface damage different from pitting produced by metal mechanical fatique. Secondly, it has to be another place somewhere in the loop showing similar damage pattern.
Is it reasonable to suspect that damage is corrosion related?

Electricpete, It is quiet uncommon that steel is used for water lubricated sleeve bearing in this kind of pump. Normal practice is to use hard rubber compound, Carbon filled PTFE or graphited alloys and they are not good conductors. Sleeves are definitely of steel alloys. Please confirm sleeve bearing material

If water contains solids it is likely that lubrication will not be poor and bearings wear prematurely. If these bearings give life of more than 5 years it is normal. I have not heard of electrical pitting for this kinf of bearings.

Thanks for the comments. A very large and capable team of people has been investigating the pump bearing failure, but that team doesn't include me. I haven't even seen the pump bearings and don't have a lot of details surrounding the pump but I will try to find out some details such as the exact alloy of the bearings and sleeves.

I was not asked to drawn any conclusions about whether there might be other causes of the observed damage (electrolysis etc). My assigned task is to evaluate whether the scenario the pump team has come up with is likely, plausible, unlikely or impossible from the standpoint of the evidence we saw at the motor. (I was first-hand involved in the meggering of the motor bearings and inspection of the upper bearing.).

The only point for current to complete the loop was the upper bearing that we inspected and saw pretty much nothing. My initial reaction is the same as David's... that I certainly would have expected to see at least some visible evidence there at the upper bearing if enough current were flowing through there to significantly damage the pump bearing (and I've heard it's pretty severe pitting). But machines often don't do what I expect, so at this point in time I have to put it in the "plausible" category.

I was hoping to get some experience of people who had seen pump sleeve bearing damage claimed to be associated with a rolling bearing motor. This week I was at an EPRI motor user's group conference and I asked at least 15 people this question. One had seen pump bearing damage blamed on currents from the motors, but didn't know many details. The rest had never heard of pump bearing damage caused by a motor. These were mostly nuke guys with not many vfd's. Most non-vfd motors have only the outboard beairng insulated, so a failure of bearing insualtion on that type would return via the inboard motor bearing and not affect the pump. It seems a pretty rare scenario but still interested if any other experiences or thoughts.

This message has been edited. Last edited by: electricpete,

The pump bearing materials: are as follows:
(rotating) shaft sleeve = Martensitic 440B steel.
(stationary) bearing = metallized carbon.
The shaft sleeve was found with the severe pitting which was postulated to be a result of electric current. The bearing was pretty well destroyed.

Photo's of the motor upper bearing inspection are attached.

Now a new piece of info that I was not aware of before - the lower bearing also had a similar problem and was grounded at the thermocouple, but bearing was not easily accessible and not nspected.

So a new potential ground loop through motor upper and motor lower bearing, still would have to go thru upper bearing where we didn't see any evidence. More importantly it seems less likely the current would go all the way to the pump if it could simply return at the lower bearing.

Any more comments on whether the motor bearing insulation problems may have contributed to a pump bearing problem?

EW_Motor_1B_upper_bearing_inspection.ppt (4,626 Kb, 37 downloads)

Based on the erosion damage pattern why the conclusion was made that pitting resulted from passage of electric current? Was it done just on the basis of found raptured T/C insulation or on the damage pattern?

May be the preloaded T/C has experienced shock and vibration which ultimately caused insulation damage and then mechanical pitting?

We just installed 4 vertical lift pumps tht were specially ordered with tungsten carbide on tungsten carbide bearings in the pump. They were chosen over cutless rubber due to the product they are pumping. It is papermill effluent. Shaft currents have long been an issue with DC motors. As for shaft currents on AC motors, I don't recall that ever happening and we do not have insulated bearings in AC motors. A failure we had on startup was an assembly error by the pump manufacturer. The middle shaft was installed backwards so the hardened surfaces did not mate.

Have a good one,

Gary B

In our plant, all motors that are "large" (above NEMA-frame, approximately above 500 hp) are specified to have an insulated bearing on the outboard end. If it were a vfd application, they would also be specified to have an insulated bearing on the inboard end.

This is fairly common practice as far as I know. I don't see it in NEMA MG-1, but it is captured in API 541-95 (large motors for the Petrochemical Industry) and in ANSI C50-41-2000 (large motors for the Power Industry)

I thought it was standard practice in all industries, but based on the above standards, it seems plausible that people outside the power and petrochem may get large motors without insulated bearings?

The motivation for insulating the outboard bearing even in a non-vfd motor is that small assymetries in the core (for example the tiny airgap between portions of a segmented core) can create circulating current. If we insulate the inboard bearing, there is still a possible loop through outboard bearing to ground and through the coupling to driven equipment bearing to ground. If we insulate the outboard bearing, there is no longer any loop that includes the core. For VFD, the high frequency capacitive components create a path to ground, so one bearing is enough to form a loop and both bearings need to be insulated.

This message has been edited. Last edited by: electricpete,

If there is suspicion of shaft currents or voltage potential accross the shaft why not verify it directly by measuring shaft voltage?
PS
(Check your email).

There was no window of opportunity to do this:
1 - We didn't know there was any bearing ground prior to taking the motor out of service.
2 - We repaired the bearing thermocouple insulation prior to returning the motor to service.